[Senate Hearing 112-756]
[From the U.S. Government Publishing Office]
S. Hrg. 112-756
COMPUTERIZED TRADING VENUES: WHAT SHOULD THE RULES OF THE ROAD BE?
=======================================================================
HEARING
before the
SUBCOMMITTEE ON
SECURITIES, INSURANCE, AND INVESTMENT
of the
COMMITTEE ON
BANKING,HOUSING,AND URBAN AFFAIRS
UNITED STATES SENATE
ONE HUNDRED TWELFTH CONGRESS
SECOND SESSION
ON
EXAMINING COMPUTERIZED TRADING VENUES
__________
DECEMBER 18, 2012
__________
Printed for the use of the Committee on Banking, Housing, and Urban
Affairs
Available at: http: //www.fdsys.gov /
U.S. GOVERNMENT PRINTING OFFICE
80-273 WASHINGTON : 2013
-----------------------------------------------------------------------
For sale by the Superintendent of Documents, U.S. Government Printing
Office Internet: bookstore.gpo.gov Phone: toll free (866) 512-1800; DC
area (202) 512-1800 Fax: (202) 512-2104 Mail: Stop IDCC, Washington, DC
20402-0001
COMMITTEE ON BANKING, HOUSING, AND URBAN AFFAIRS
TIM JOHNSON, South Dakota, Chairman
JACK REED, Rhode Island RICHARD C. SHELBY, Alabama
CHARLES E. SCHUMER, New York MIKE CRAPO, Idaho
ROBERT MENENDEZ, New Jersey BOB CORKER, Tennessee
DANIEL K. AKAKA, Hawaii JIM DeMINT, South Carolina
SHERROD BROWN, Ohio DAVID VITTER, Louisiana
JON TESTER, Montana MIKE JOHANNS, Nebraska
HERB KOHL, Wisconsin PATRICK J. TOOMEY, Pennsylvania
MARK R. WARNER, Virginia MARK KIRK, Illinois
JEFF MERKLEY, Oregon JERRY MORAN, Kansas
MICHAEL F. BENNET, Colorado ROGER F. WICKER, Mississippi
KAY HAGAN, North Carolina
Dwight Fettig, Staff Director
William D. Duhnke, Republican Staff Director
Dawn Ratliff, Chief Clerk
Riker Vermilye, Hearing Clerk
Shelvin Simmons, IT Director
Jim Crowell, Editor
______
Subcommittee on Securities, Insurance, and Investment
JACK REED, Rhode Island, Chairman
MIKE CRAPO, Idaho, Ranking Republican Member
CHARLES E. SCHUMER, New York PATRICK J. TOOMEY, Pennsylvania
ROBERT MENENDEZ, New Jersey MARK KIRK, Illinois
DANIEL K. AKAKA, Hawaii BOB CORKER, Tennessee
HERB KOHL, Wisconsin JIM DeMINT, South Carolina
MARK R. WARNER, Virginia DAVID VITTER, Louisiana
JEFF MERKLEY, Oregon JERRY MORAN, Kansas
MICHAEL F. BENNET, Colorado ROGER F. WICKER, Mississippi
KAY HAGAN, North Carolina
TIM JOHNSON, South Dakota
Kara Stein, Subcommittee Staff Director
Gregg Richard, Republican Subcommittee Staff Director
Catherine Topping, FDIC Detailee
(ii)
?
C O N T E N T S
----------
TUESDAY, DECEMBER 18, 2012
Page
Opening statement of Chairman Reed............................... 1
Opening statements, comments, or prepared statements of:
Senator Crapo................................................ 2
WITNESSES
Joseph Mecane, Executive Vice President and Head of U.S.
Equities, NYSE Euronext........................................ 4
Prepared statement........................................... 27
Responses to written questions of:
Senator Reed............................................. 51
Daniel Mathisson, Head of U.S. Equity Trading, Credit Suisse
Securities LLC................................................. 6
Prepared statement........................................... 29
Responses to written questions of:
Senator Reed............................................. 65
Eric Noll, Executive Vice President and Head, Nasdaq OMX
Transaction Services........................................... 7
Prepared statement........................................... 35
Responses to written questions of:
Senator Reed............................................. 68
Robert C. Gasser, Chief Executive Officer and President, ITG..... 9
Prepared statement........................................... 40
Responses to written questions of:
Senator Reed............................................. 74
(iii)
COMPUTERIZED TRADING VENUES: WHAT SHOULD THE RULES OF THE ROAD BE?
----------
TUESDAY, DECEMBER 18, 2012
U.S. Senate,
Subcommittee on Securities, Insurance, and Investment,
Committee on Banking, Housing, and Urban Affairs,
Washington, DC.
The Subcommittee met at 9:31 a.m., in room SD-538, Dirksen
Senate Office Building, Hon. Jack Reed, Chairman of the
Subcommittee, presiding.
OPENING STATEMENT OF CHAIRMAN JACK REED
Chairman Reed. Let me call the hearing to order, and let me
welcome our witnesses. I thank my colleague, the Ranking
Member, Senator Crapo, for joining us this morning.
This hearing will be focused on how computers, market
regulation, and competition have dramatically changed equity
markets in the United States. In recent decades, the United
States equity markets have experienced tremendous technological
innovation and intensified competition between an expanding set
of trading venues, some of which have been driven by regulatory
changes.
As of December 2012, there are 13 official stock exchanges
registered with the Securities and Exchange Commission. In
addition, there are nearly 40 ATSs, alternative trading
systems, through which broker-dealers can access the market.
Included among the ATSs are the so-called dark pools that do
not publicly display quotes or prices. There are also more than
200 broker-dealers that execute orders in-house in a process
known as ``internalization.'' In short, liquidity has
fragmented over numerous trading venues as competition has
intensified.
Two SEC regulations dramatically changed or accelerated
changes in the structure of our financial markets. In 1998, the
Securities and Exchange Commission adopted Regulation ATS to
encourage the development of innovative new market centers.
Regulation ATS exempted these alternative trading systems from
having to register as exchanges. And in 2005, the SEC adopted
Regulation National Market System, or Regulation NMS, a series
of rules and regulations designed to modernize and strengthen
the national market system for equities. Implemented in 2007,
Regulation NMS is credited by some to have caused the biggest
change of the two by requiring that orders be sent to the
trading platform with the best price.
The adoption of Regulation ATS and Regulation NMS led to a
proliferation of new trading platforms. It also put a premium
on speed, giving an advantage to firms that could place orders
first and take advantage of minuscule price differentials
between the trading venues. Such high-frequency, computer-based
training has grown in recent years, representing about 50
percent of equity trading in the United States. High-frequency
trades employ many different automatic strategies with rapid
orders and short-term holding periods. High-frequency trading,
advanced computer technology, and trading strategies have led
to changes in the marketplace that minimize latency or the time
it takes to send and execute a trade.
In response to the demand for faster execution, exchanges
are allowing trading firms to place or collocate their service
in the exchanges' data centers. Some trading venues also allow
direct access through which certain trading firms access the
exchanges' matching engine directly. Exchanges also offer
direct data feeds on a proprietary basis to their customers.
These feeds cost more, but the data arrives sooner. It has been
reported that some exchanges and trading firms have developed
an ultra-fast micro wavelength to provide even faster speeds.
The growth of new trading venues and the increasing use of
automation and advanced computing technology have raised
questions about the effect that these changes have had on the
stability, fairness, and integrity of our marketplaces. Flash
Crash, the BATS and Facebook IPOs, and Knight's trading debacle
all bring into question the very rules that govern our market
structure.
Do market rules reflect the realities of today's market
structure? Are the markets fair and transparent? Is the
marketplace too complex or to fragmented? Should all markets
have the same rules? Is there a great potential for systemic
risk propagation as a result of the interconnection of highly
computerized markets? Or, essentially, the ultimate question,
do the rules of the road need to be changed?
I look forward to our witnesses' testimony and a robust
discussion of these issues.
I will now turn to my colleague Senator Crapo and thank him
for his indulgence for my long opening statement. Thank you.
STATEMENT OF SENATOR MIKE CRAPO
Senator Crapo. Thank you very much, Mr. Chairman. I
appreciate you holding this important hearing today. It is our
second in a series of hearings on computerized trading.
At our last hearing, we heard from witnesses representing
the buy side: high-frequency traders and the research and
advisory firms. Today we are going to hear from witnesses
representing exchanges and alternative trading venues.
Because there have been too many market disruptions caused
by software errors since the Flash Crash of 2010, there has
been a lot of discussion on how to fortify our markets during
times of market stress.
In October, the Securities and Exchange Commission held a
Technology Roundtable on how to minimize trading errors and
market malfunctions as well as how to respond to them in real
time. The use of kill switches, automated switches that would
turn off trading at securities firms when their volume exceeded
preset maximums, appears to be the first choice of many market
participants. Our market infrastructure should be able to
handle a trading error or technology failure in an appropriate
way to minimize disruptions.
I am interested in learning more about how effective kill
switches can be and what their effectiveness is with regard to
other safeguards.
I also look forward to hearing from our witnesses about how
the industry, working with the SEC, has made progress in
implementing kill switches since the Technology Roundtable. In
addition, I look forward to a discussion about the changing
landscape of our securities markets and how different trading
venues impact investor protection, market integrity, and
capital formation.
Our witnesses today provide a deep expertise on these and
other issues, and I appreciate them being here and bringing us
their expertise and their thoughtfulness in addressing these
questions.
Again, Mr. Chairman, I appreciate you, not only our working
relationship but holding this hearing, and I look forward to
working on this and other issues with you as we move forward.
Chairman Reed. Thank you very much, and I also look forward
to working with you, Senator Crapo.
Senator Corker, do you have a statement?
Senator Corker. No.
Chairman Reed. Thank you very much, Senator Corker.
Let me introduce our witnesses now and then ask them to
present their testimony.
Our first witness is Mr. Joe Mecane. Mr. Mecane is an
executive vice president with NYSE Euronext and oversees
strategy, business development, and operations for the
company's U.S. equities platforms. Prior to joining the
company, Mr. Mecane was managing director in the Equities
Division of the UBS Investment Bank. Thank you, for joining us.
Mr. Dan Mathisson is the head of Advanced Execution
Services at Credit Suisse. AES is a full-service suite of
algorithms, strategies, tools, and analytics for trading
securities throughout the world. He is on the board of
directors of BATS trading, the Credit Suisse Global Equities
Operating Committee, and is a member of the New York Stock
Exchange Electronic Traders Advisory Committee. Thank you, Mr.
Mathisson.
Mr. Eric Noll is executive vice president, Transaction
Services, of the Nasdaq OMX Group, Inc., the world's largest
exchange. Mr. Noll oversees the trading operations of all U.S.
transaction services businesses. Mr. Noll joined Nasdaq OMX
from Susquehanna International Group. Thank you very much, Mr.
Noll.
And our final witness is Mr. Robert Gasser. Mr. Gasser is
chief executive officer and president of Investment Technology
Group, Inc., ITG. Before joining this organization, Mr. Gasser
was head of U.S. equity trading a JPMorgan. Thank you very
much.
Gentlemen, all of your testimony will be made part of the
record. We ask you to summarize within the 5 minutes allocated
to you, and we will begin with Mr. Mecane. Thank you, Mr.
Mecane.
STATEMENT OF JOSEPH MECANE, EXECUTIVE VICE PRESIDENT AND HEAD
OF U.S. EQUITIES, NYSE EURONEXT
Mr. Mecane. Chairman Reed, Ranking Member Crapo, and
Members of the Subcommittee, my name is Joe Mecane, and I am
EVP and head of U.S. Equities at NYSE Euronext--a leading
global operator of financial markets and provider of trading
technologies.
While the U.S. continues to have the most liquid markets in
the world and remains at the forefront of innovative technology
used to conduct electronic trading, the infrastructure used to
operate the markets each day has grown so sophisticated that
few fully appreciate how well our markets actually operate in a
highly competitive and complex environment.
This has made it difficult for market participants,
regulators, and Congress to determine if the growth in the
number of trading venues and use of automated trading is
beneficial.
However, in light of the market events that have occurred
in recent years, I would like to focus on how technology and
our market structure have created both unnecessary complexity
and mistrust of markets; and, relatedly, what we believe the
industry, regulators, and Congress should be doing to address
it.
Electronic trading has added tremendous benefits to the
capital markets, including lower costs of trading, faster speed
of execution, and in some cases greater transparency. However,
there were several regulatory changes that drove the market to
become more electronic. One significant factor was
decimalization of the markets in 2001, which had an effect of
decreasing average spreads by roughly 38 percent, directly
benefiting end investors. The narrowing of spreads led to a
huge expansion of electronic trading because human traders
could no longer effectively make markets in this environment
and because institutions and brokers began relying more on
algorithmic trading to access the markets and reduce their
costs of trading. This process started before Regulation NMS.
In fact, almost all reductions in spreads occurred during the
pre-Regulation NMS period.
In 2007, just as the technology was becoming more
sophisticated, the SEC adopted Regulation NMS through which
exchanges began to compete by establishing the NBBO, and speed
became the competitive differentiator based on one market's
ability to set the national best bid or offer faster than a
competing market.
Regulation NMS also established the Order Protection Rule
to protect visible orders and encourage displaying quotes, yet
today over 3,000 securities have 40 percent of their volume
occurring off-exchange. This level of off-exchange activity
erodes the incentive for market makers to continue to trade the
less active securities and threatens to further decrease the
incentives for companies to go public.
Today, there are 60 execution venues in the U.S. markets.
Exchanges find themselves competing more directly with other
venues that are able to employ different practices than
exchanges with less oversight and disclosure. As a result of
these advantages, broker-dealers continue to move more order
flow into their own trading venues before routing to the public
markets.
As you can see, the technology and the rules that govern
the U.S. equity markets have resulted in the creation of a
trading infrastructure primarily focused on speed and
complexity through which traders can identify and access
liquidity. To accomplish this, exchanges, brokers, and vendors
have had to build expensive networks with robust capacity, as
well as learn how to interact in a very complex ecosystem.
In response to this new flow of orders, exchanges have
developed new order types and evolved their market structures.
Regardless of the reason for the specific order type, most are
premised on the goal of attracting liquidity to the public
markets, and all must be reviewed by the SEC and published for
public comment, something unique to national securities
exchanges.
The bottom line is that our market structure incentivizes
these various levels of increased complexity. Our main message
today is that if we want to reduce the complexity of technology
and the related framework of our markets, we should simplify
the overall market structure.
NYX believes that the SEC is best suited to propose
meaningful market structure changes, and regulators in other
global markets, including Canada, Australia, and Europe, are
already taking action. With congressional oversight, the SEC
should continue with the holistic review it began in 2010 with
the Concept Release on Equity Market Structure by proposing
changes that will promote transparency, fairness, and long-term
capital formation.
We believe that changes to be considered should include a
review of market maker obligations, the Sub-Penny Rule, the
Order Protection Rule, tick sizes for illiquid securities, and
addressing the conflicts and overlap between broker-dealers and
exchanges, including the obligations and responsibilities of
each when providing like services.
The Consolidated Audit Trail, proposed by the SEC, also is
a vital component to ensuring effective surveillance in a
highly fragmented marketplace.
NYX also believes that, in light of the existing complexity
of the markets and the trading glitches that have occurred this
year, all trading venues should ensure a robust set of policies
and procedures around their systems development life cycle.
Although testing may not be the most exciting part of the
industry's initiatives, the hyper-competition that exists in
the securities markets lends itself to excessive levels of
change to remain competitive and compliant with new regulatory
requirements.
In closing, I want to reiterate our belief that although
our capital markets are the best in the world, there remains
room for improvement. Our recommendations for change have a
simple premise: implement market structure changes that enhance
transparency for investors and level the playing field for
trading venues.
Thank you for inviting me to testify, and I look forward to
your questions.
Chairman Reed. Thank you very much.
Mr. Mathisson, please.
STATEMENT OF DANIEL MATHISSON, HEAD OF U.S. EQUITY TRADING,
CREDIT SUISSE SECURITIES LLC
Mr. Mathisson. Thank you, Chairman Reed, Ranking Member
Crapo, and Members of the Subcommittee. My name is Dan
Mathisson, and I am the head of U.S. Equity Trading for Credit
Suisse. Credit Suisse's U.S. broker-dealer unit, formerly
called First Boston, has been in operation in the U.S. since
1932, and today Credit Suisse employs 9,200 people in the
United States. We are one of the largest U.S. broker-dealers,
having executed 12.4 percent of total volume in 2012, and we
own and operate two alternative trading systems: Crossfinder,
which has been the largest ATS in the U.S. since 2009, and
Light Pool, a newer ATS that publicly displays bids and offers.
I have been working in the equity markets for over 20
years, and it is a privilege to have the opportunity to speak
here today.
Credit Suisse believes that the U.S. equity markets are
better than they have ever been and remain the envy of the
world. We recently published a broad survey of market quality
where we found that the markets have improved in almost every
measurable way. We believe that Regulation ATS, decimalization,
and Regulation NMS were successful at making the U.S. markets
more efficient, fair, and equitable. However, while they are
good, the markets are not perfect, and we have seen several
market disruptions that became big news stories in the past few
years. But we believe that the SEC has moved aggressively and
thoughtfully to fix these issues, and Credit Suisse fully
supports the pending Limit Up/Limit Down rule, the consolidated
audit trail, the market access rule, the tighter marketwide
circuit breakers, and the creation of the SEC's new Office of
Analytics and Research. We applaud the SEC for these actions
and believe they will serve to significantly reduce the risks
of disruptions in the future.
In addition to those SEC actions, we have a further
suggestion that we believe will decrease the likelihood of
disruptions. We recommend that the SRO status of the exchanges
be examined. The trading errors that occurred on the day of the
Facebook IPO last May served to highlight a peculiar quirk of
the U.S. market structure: that exchanges do not have material
liability for their failures. As SROs, exchanges have long been
considered by courts to be quasi-governmental entities and are
therefore entitled to the common law doctrine of sovereign
immunity, which protects Government entities from liability
judgments.
Yet exchanges today are clearly not governmental entities.
They are for-profit private companies that are not particularly
different from broker-dealers. While they still have a few
vestigial regulatory functions, the majority of their
regulatory responsibilities are typically outsourced to FINRA.
We believe that exchanges should not have been allowed to
retain their SRO status when they converted to for-profit
entities 6 years ago. Businesses or individuals are inherently
more cautious when they have the potential to be found liable
for their actions. It is time for policy makers to correct this
mistake.
Our second suggestion is that it is time for the regulators
to do a comprehensive review and overhaul of the market data
revenue plans. The Consolidated Tape Association has a legal
monopoly on providing a stream of real-time data from our
Nation's stock markets. The CTA is the only source of this
data, and it sells it at a high cost to the investing public.
After all of its operational and administrative costs, the CTA
makes a profit of approximately $400 million a year, which is
then rebated to the exchanges based on a complex formula.
Historically, the SEC has justified granting exchanges this
exclusive right to sell market data as a form of user tax to
fund the exchanges' regulatory expenses. However, based on what
we can glean from earnings reports and SEC filings, the amounts
earned by the exchanges today far exceed the regulatory
expenses, and tape revenue acts as a major profit center for
the exchanges. It appears that somewhere along the way market
data revenue became a Government-granted windfall at the
expense of the investing public. The current market data system
was passed in November 1972. After 40 years, we believe it is
time to give it a fresh look.
In summary, removing immunity from exchanges would increase
reliability over the long term and, therefore, reduce market
disruptions, and overhauling the tape revenue system would
reduce costs for the investing public.
Thank you for the opportunity to appear here today, and I
will be happy to answer any questions that you may have.
Chairman Reed. Thank you very much, Mr. Mathisson.
Mr. Noll, please.
STATEMENT OF ERIC NOLL, EXECUTIVE VICE PRESIDENT AND HEAD,
NASDAQ OMX TRANSACTION SERVICES
Mr. Noll. Good morning. Thank you, Chairman Reed and
Ranking Member Crapo, for the opportunity to testify today on
computerized trading. My name is Eric Noll. I am the executive
vice president of Nasdaq Transaction Services for Nasdaq OMX.
Computerized trading is a fact of life and the default
method for billions of trades over the past several years--
billions of trades that happen without any concern or problem
in the market every day. While there are issues, computerized
trading has a proven track record of delivering benefits for
investors, equalizing the information advantage that was the
staple of manual markets, lowering trading costs, and allowing
the market to handle trade traffic that would freeze otherwise
manual markets.
In light of recent events, do not forget the unique role
exchanges continue to play in the U.S. equity markets. All that
your constituents associate with ``the market'' starts with an
exchange. The iconic public companies--Apple, Google, eBay, and
Amazon--must satisfy exchanges listing standards and
regulations against corporate fraud and abuse. Only equity
exchanges carry the important responsibility to support that
growth and wealth creation from those companies.
Exchanges have heavy responsibilities to create a safe
market for investors, characterized by fair access,
transparency, and efficiency. No other market participant is
charged with or even permitted to undertake that burden. One
need only look at the list of SRO responsibilities that
registration triggers to understand why so few of our lightly
regulated competitors voluntarily take that step.
All of the buying and selling and active trading in the
markets is not a grand game of speculators. It has real job
creators and investors relying on the market's best information
to make rational business and investing decisions. Price
discovery, the best bid and offer, and visible liquidity are
very important to public companies seeking activities like
secondary offerings for expansion and to use their stock as a
currency in the marketplace to achieve strategic goals like
acquisitions.
Exchanges are not in the business just to see trading for
trading's sake. We are in the business to produce transparent
quoting and trading that helps price discovery, helps add
liquidity, tightens spreads, and benefits the continuous
market, ultimately helping job creation and economic growth.
U.S. markets are complex, fragmented, and interconnected.
We are working tirelessly to ensure that markets are strong and
fair, and that as the pace of trading accelerates, so too does
the pace of regulation and investor protection.
For example, after May 6th, the SEC and the exchanges
worked quickly and cooperatively to devise new protections and
reform rules for breaking trades, instituted single stock
circuit breakers, and will implement updated marketwide circuit
breakers and the Limit Up/Limit Down mechanism in the first
quarter next year. Nasdaq also fully supports and is helping to
lead and define the implementation of Peak Net Notional
Exposure levels, or what is commonly known as ``kill
switches.''
On the issue of high-frequency trading, many vilify HFT as
a business model. We urge caution against that sweeping
criticism. It seems the tenor of the debate about HFT has
become too broadly negative toward it as a business model.
Academic evidence, like the British Beddington study, suggests
that HFT trading tightens spreads and adds valuable liquidity--
certainly positives for the market.
It is not enough simply to vilify fast trading. Regulators
and exchanges are working to identify and address specific bad
actors and specific bad outcomes such as false, misleading, or
deceptive practices.
To improve our own regulatory program and the regulatory
programs of exchanges around the world, in 2010 Nasdaq acquired
The SMARTS Group, the leading provider of automated
surveillance for post-trade high-speed trading.
As you have heard, our markets are complex: 13 exchanges
and 40-plus alternative venues. Each venue has its own systems
and competes for orders. Each has its own order types and is
continually developing new ones with input from investors.
While some order types have come under scrutiny, let me be
clear: Nasdaq OMX order types do not provide advantages to
certain users allowing them to jump ahead in line at a given
price level. We believe that order types should be designed to
make our markets better, improve transparency, and improve
price discovery. We go through a rigorous process to get our
order types approved and to expose our innovative ideas to the
market through the notice and comment period at the SEC, often
allowing competitors to mimic our ideas and beat us to the
market. That is part of our SRO burden. To help members
understand our order types, we recently posted on our Web site
a plain-language list and a description of Nasdaq's order
types.
We are passionate about the critical role we play in
capital formation, and while it presents challenges to
everyone, ultimately we believe technology is an important part
of that solution. We appreciate the opportunity to testify. I
look forward to your questions.
Chairman Reed. Thank you very much, Mr. Noll.
Mr. Gasser, please.
STATEMENT OF ROBERT C. GASSER, CHIEF EXECUTIVE OFFICER AND
PRESIDENT, ITG
Mr. Gasser. Chairman Reed, Ranking Member Crapo, and other
Members of the Subcommittee, thank you for the opportunity to
testify this morning on the topic of rules of the road for
computerized trading venues. On behalf of a leading agency
broker, my goal is to offer an unbiased, fact-based view on the
current state of U.S. equity market structure.
ITG is a New York Stock Exchange-listed company with 17
offices across 10 countries and nearly 1,100 employees. As an
agency broker, ITG provides trading services, technology,
analytics, and research to a wide array of leading asset
managers. Throughout our 25-year history, we have worked in
partnership with major mutual funds, pension funds, and other
institutional investors, innovating to improve trading and
investment performance. In my testimony today I would like to
offer a brief overview of current market structure, discuss
some recent events which have impacted investor confidence, and
look at some ways to restore this confidence. This morning we
hope we can infuse some data and analysis into the debate.
Competition amongst market centers and broker-dealers
spawned by the passage of Regulation ATS in December 1998 has
led to intense competition for liquidity and ultimately to
fragmentation. This fragmentation has undoubtedly introduced
complexity into our marketplace but has been a positive force
in reducing execution costs.
Technology has provided market participants, including
retail investors and mutual funds, with the tools necessary to
aggregate liquidity and derive the full benefit of free market
competition for order flow.
Global asset managers, as fiduciaries, have an obligation
to achieve best execution. The global market standard requires
all asset managers of size to measure the quality of their
execution and its effect on the investment process. ITG is the
world's largest provider of TCA, or transaction cost analysis.
We measure millions of trades executed on behalf of hundreds of
global asset managers. Our TCA data clearly demonstrates that
institutional investors have benefited greatly from the
evolution of U.S. market structure. Over the past 12 years,
there has been a 70-percent decrease in average total equity
trading costs in the U.S. As the data indicates, market
structure is not broken. The current ecosystem of displayed and
dark markets has resulted in significantly reduced costs that
in almost all cases have been distributed back to investors.
There is no evidence to suggest that competition and
fragmentation have damaged price discovery or harmed capital
formation.
ITG is not a market maker, and we do not take on
proprietary positions. In other words, we do not have ``skin in
the game'' when it comes to the debates around broker
internalization, as our system provides ``meaningful price
improvement'' to buy-side investors as described in Regulation
NMS.
Based on our data, we would conclude that broker-dealer
internalizers, or ``broker-dealer dark pools,'' as they are
sometimes known, provide a useful permeable layer between the
client and the displayed markets. Brokers have a fiduciary
responsibility to their clients while exchanges do not, and
these liquidity pools would not exist unless benefit was
derived by the customer.
Most recently, Australia and Canada have imposed
regulations around internalization that will provide us with
data sets to examine when considering the implications of
potentially taking similar action here in the U.S. Early
returns do not look promising in terms of the effects on
liquidity and trading costs. Regressing to an oligopoly of
exchanges is clearly not the answer.
Unfortunately, the evidence also suggests that the
investing public has become disenchanted with equities.
According to the Investment Company Institute, over half a
trillion dollars has been pulled from U.S. equity mutual funds
since the start of 2008.
Much of this can be attributed to the reduced risk appetite
of baby boomers and the relative safety of bonds supported by
easy monetary policy.
The May 2010 Flash Crash, the Facebook IPO, and Knight
Capital's trading debacle this past summer provide little
comfort that U.S. equity markets are a safe place to trade or
invest. Add in the suspicions that the investing public has
about high-frequency trading and its perceived impact on the
quality of markets, and you have a recipe for anecdote and
conjecture overcoming facts and reason.
In our opinion, we can focus on five tangible initiatives
to accomplish restoring confidence:
The SEC's Consolidated Audit Trail, if implemented properly
and cost effectively, will give investors confidence that
regulators can police bad actors and predatory strategies.
The consistent application of the market wide circuit
breakers and the Limit Up/Limit Down Plan to all market centers
would likely prevent a market disruption of Flash Crash
proportions.
Costs should be borne by market participants who create
excessive quote traffic without executing order flow.
Market data should be distributed to all market
participants equally.
Marketwide risk should be monitored at a central
clearinghouse that would have the ability to terminate a
broker-dealer's connectivity to the national market system in
the event of a rogue program released to the market.
These five measures would give the investing public the
protections they need to confidently invest in the world's
strongest and most resilient market while still deriving all of
the cost savings and liquidity benefits which have been
achieved over the past decade.
Last, as the regulations called for by the Dodd-Frank Act
begin to take hold across other asset classes, the lessons we
have learned in equities will be applied to those markets.
Price discovery, central clearing, transaction cost analysis,
and pre- and post-trade transparency will become as deeply
integrated into foreign exchange and fixed-income markets as
they are in equity markets. And innovation will come more
quickly to those markets because of the lessons learned. For
this reason, our equity market structure is all the more
important to our broader financial system.
Thank you again for the opportunity to share our views on
these important questions. I would be happy to answer any
questions at the appropriate time.
Chairman Reed. Thank you very much, gentlemen, for your
excellent testimony. One of the major driving forces of this
discussion is technology, and one of the things about
technology, it tends to make good things better and bad things
worse. So when it comes to markets, there has been, I think,
evidence showing a decrease in spreads and savings to
investors, but I think our job is to applaud those good things,
but also look very seriously about what the costs might be
because of this new technology and new markets. And it raises a
threshold question about--and you all alluded to it, but let me
start with Mr. Gasser--the complexity of markets. I think the
dark pools, internalization, in fact, I think every one of the
exchanges--and correct me, Mr. Noll and Mr. Mecane--has their
own dark pools. Is that accurate?
Mr. Noll. We do not have----
Chairman Reed. You do not have a dark pool? Do you?
Mr. Mecane. We have a small investment in an ATS.
Chairman Reed. But the trend is that the proliferation of
dark pools, the internalization, and it goes to the issues
which--your question about investor confidence, investor
enthusiasm. Price discovery, is that--even if you are saving
prices, in the dark pools is price discovery for the average
investor being compromised or complicated? What about capital
formation for new companies? A lot of this trading is being
done in a very small range of stocks, not all the stocks that
are listed. So can you talk, again, accepting the benefits,
what are the costs, the complexity, and what things might we
do--and I would like everyone to respond--to make it a little
simpler and a little more easy to bear?
Mr. Gasser. Thank you, Senator. I think as we think about
complexity, as I alluded to earlier, I think we need to retain
the benefits that we have derived from current market structure
while making the markets less opaque to retail investors. There
is no evidence to suggest that the market share of
internalization or trades that are executed away from the
exchanges is harming price discovery. There has been a lot of
academic focus on that. So we tend to go back to transaction
costs and how they affect the institutional investor, which is
our core constituency.
But, clearly, the complexity, it has given rise to this
view that there is a certain opacity to our marketplace. I
would say that, you know, the IPO market in the U.S. has
actually been leading the globe in terms of capital formation.
It is not a great number this year, but I would submit to you
that it is a function of the economy, of Sarbox, of some of the
things that have been certainly imposed on growth companies
that make it difficult for them to think about the critical
mass they need in terms of market capitalization, to really
think about the public equity markets as opposed to a private
equity market or a recapitalization somewhere else on their
balance sheet.
So at the end of the day, I think there is a lot of market
forces, a lot of macroeconomic forces that come to bear as well
on that equation. But I think in terms of speed and latency, we
have gotten to a place now where it is a law of diminishing
returns. Do we need to trade in micro seconds? I do not think
so. That might be heretical for someone who has spent almost
their entire professional career trying to improve markets
through technology at delivery. But at this stage of the game,
micro seconds are, I think, a vast overachievement, if you
will, technologically in terms of the benefits that people
derive from it, and it just creates more and more suspicion
that the market is controlled by dark forces sitting in a data
center and deploying more and more servers and software to
attack a business motive.
So I think at this stage of the game, there are some very
clear things--consolidated audit trail, clearly. If folks have
confidence that the SEC can enforce and regulate--and I think
by their own admission it has been very difficult to do that
given the complexity of the market--I think it will restore
quite a bit of investor confidence. So I think the CAT, as it
is commonly referred to, is very, very important.
Chairman Reed. Let me ask Mr. Noll and the whole panel to
respond to this question. Then I will ask Senator Crapo for his
questions, and then we will do a second round.
Mr. Noll, the same issue, the general issue of complexity,
and also I think when you talk about the SEC, there are two
issues: one is the authority and one is the resources. So we
have to make sure that if they get authority, they also have
resources, but that is an aside.
Mr. Noll, please, your comments.
Mr. Noll. Thank you, Chairman. A couple of comments around
this.
First of all, I do think the markets are incredibly
complex. I disagree with Mr. Gasser, respectfully, that the
increase in off-exchange trading is not harming price
discovery. We have started to see some evidences of that. I
want to be clear, though, however, in responding to that. We do
think there is an important role for off-exchange trading. We
think there is an important role for ATSs. We think there is an
important role for dark trading.
Our concern is really about the primacy of price discovery
and price formation. And to the extent that those are being
negatively impacted by current market structure, we think that
those issues should be revisited as we move forward.
Other issues, you know, clearly to address more directly
your question about the SEC, I do believe that they have the
regulatory authority here. I do think that they are incredibly
interested in this topic, concerned, and working very hard to
address some of these issues. Some of the steps that we have
taken as an industry, like the adoption of CAT, will give them
some of the tools that they need in order to correctly assess
what is going on in the marketplace as we move forward.
Clearly, they are also deeply embedded in some of the other
issues that we are debating and rolling out and implementing.
Circuit breakers were the first of many. They are intimately
involved with the Limit Up/Limit Down scenario currently being
worked on by the industry as we roll that forward, and we have
to address Senator Crapo's earlier comment. The exchanges,
working with the industry, are working with the SEC to adopt a
format of kill switches that will prevent runaway algorithms
from damaging the marketplace.
So I do think that they certainly have the regulatory
authority, they certainly have the commitment to be involved in
this space. And to finalize my answer here, I think one of the
important things to look at is complexity in and of itself is
not a bad thing. Technology in and of itself is not a bad
thing. What we should be doing around market structure is
making sure that we can recover from incidents and that we do
not create scenarios whereby incidents are much more likely to
happen.
And so much of the efforts that we are working on at Nasdaq
and with the industry and our fellow exchanges are designed
really to either recover from incidents or prevent new
incidents from occurring. But the competition around the
development and innovation around new market structure, new
developments in the marketplace, is important. I think it adds
value, and I think it brings real benefits to investors.
Chairman Reed. Thank you.
Mr. Mathisson, please, and then Mr. Mecane.
Mr. Mathisson. Thank you. Mr. Noll just commented on the
move to off-exchange trading, and I would actually dispute the
data that there has been a move to off-exchange trading. In
October of 2012, 33 percent of the volume was traded on what is
called the TRF, the Trade Reporting Facility, which is the
trading that is done off-exchanges. If you go back 5 years to
October of 2007, 30 percent of the volume was traded on the TRF
or off-exchange. We have seen a fluctuation between 25 and 35
percent of the volume done off-exchange every month for the
last 5 years, with no exceptions. There is no trend if you look
at the long term of flow actually moving off the exchange. You
have to get pretty creative in how you cut the data to come up
with a trend that the flow actually is moving off. So, you
know, I think that there has not been an increase in off-
exchange trading. That is important to acknowledge.
Now, I concur with Mr. Gasser that we do not want to lose
the benefits of competition, and we have seen many of them, as
every panelist acknowledged in their testimony. At Credit
Suisse, we are not in favor of setting any type of a speed
limit or doing anything drastic. We would not be in favor of
setting a 55-mile-per-hour speed limit across the board. We
think the solution is to put proper safety mechanisms in place,
to put seat belts and air bags in the cars, but not necessarily
to say that people are not allowed to drive at the speed that
they want to drive.
We believe that some of the things that the SEC has already
done--the Limit Up/Limit Down rule, marketwide circuit
breakers, market access rules, and the discussion of kill
switches--are the right solutions and do put the right safety
mechanisms in place to allow the markets to continue to enjoy
the benefits of competition.
Chairman Reed. Thank you.
Mr. Mecane, please, and then I will recognize Senator
Crapo.
Mr. Mecane. Thank you. While I think we would all agree on
this panel that markets for the most part have improved
significantly, we are concerned about the direction of some of
the evolution of our markets. And I think it is also important
to acknowledge that while we have had a lot of practices evolve
over the period that markets have improved, not all the
practices that existed necessarily contributed positively to
some of the changes that we have had in the marketplace.
And, while we are supportive of competition, some of the
evolution that has happened has been a little bit unintended,
and the two things that we have highlighted in a lot of our
testimony and statements is around ensuring that from a public
policy standpoint we have the right incentives in place for
people to display orders publicly in the public markets, and
ensuring that we have similar levels of rules and oversight for
activities that occur on different venues.
You know, I would respectfully disagree with some of Mr.
Mathisson's statements about the level of off-exchange trading,
and it gets into a little bit of micro structure complexity
around the ECNs that did publicly quote back in the early
periods, and we are happy to address that offline. But, I think
what it highlights is our markets have become extremely
complex. I think the SEC and the industry have done a good job
of responding to some of the crises that we have seen over the
last few years, but clearly there is more that can be done.
One of the things that we would encourage is because the
markets are so complex and because when you do try to address
one issue in isolation, it raises a host of other questions and
issues. We have suggested that a holistic review of market
structure, as the SEC had intended back in 2010 with the
Concept Release on Equity Market Structure, is probably the
best way to try and address all the issues that are being
raised on this panel because it is difficult to address any of
these issues in isolation.
We would also suggest that perhaps one area that is
discussed but not fully addressed yet is having testing
standards in the industry. I think we need to be careful not to
be too heavy-handed with that type of approach. But certainly
having some best practices, some requirement for firms, trading
firms, trading venues to have policies and procedures that
address their testing standards, how they roll out new code,
and setting that as a standard in the industry could certainly
be very helpful. And as Mr. Noll highlighted, all of the
Members of this Committee are involved in the industry
discussions around potential kill switches that Senator Crapo
mentioned before and working with the SEC about ways to
potentially implement safeguards along those lines.
Chairman Reed. Thank you very much.
Senator Crapo.
Senator Crapo. Thank you, Mr. Chairman.
Let me follow up on the kill switch question. I guess I
will start with you, Mr. Mecane. What progress is being made
right now in implementing the idea of the kill switches since
the SEC Roundtable?
Mr. Mecane. Sure. It is a good question, and there is a
very active dialog around the kill switch concept. As you know,
all the SROs and a number of industry firms, including the
firms on this panel, were involved in discussions leading up to
the roundtable about a possible framework for a kill switch.
And subsequent to the roundtable, we have all engaged in a much
more detailed assessment of how a kill switch could work. We
have been working with DTCC, and, you know, as they say, the
devil is in the details. It gets very complex once you start
designing it and outlining it, and I believe we are getting
close to a possible framework as we iterate through the
different possible ways to implement something like that in the
industry, you know, what parts should be mandatory, what parts
should be optional, whether it should be at the clearing level,
whether it should be at the SRO level. All these types of
issues are under discussion, but I think we are hopeful to have
something to report in the first quarter next year.
Senator Crapo. Thank you.
Let me turn to you, Mr. Gasser. You indicated that you felt
it would be very helpful for a central clearinghouse to monitor
market risk. Could you explain that a little better?
Mr. Gasser. Yes, Senator. You know, all of us have a
substantial amount of capital committed to the business of
clearing client trades, and they are held within the national
clearinghouse, and there is a central clearing facility here in
the U.S. which is actually very effective and much different in
terms of global market structure. So it is very efficient, very
effective. We have quite a bit of capital on deposit to support
all of our trading activities there.
One issue, I think, that is a relatively scary one for the
industry is that Knight gave us some interesting lessons in
terms of a potential for a firm to blow through their capital
and for an unprecedented call on other firms' capital at one
stage or another to fill the gap. So that event I think had
some interesting implications for not only the national market
system, but for individual broker-dealers and their
participation within the central clearinghouse.
So when I alluded to that--and I know that Joe and Eric and
others, and Dan and our firms, have been involved in this
dialog. One of the things I think folks keep coming back to is
that notional limitation that we should all be held accountable
to. When you blow through that, then you are now at a place
where you may substantially affect other firms and their
capital on deposit.
Senator Crapo. Well, thank you.
Mr. Noll, I will ask you this question--well, actually, I
think Mr. Mecane and Mr. Noll both in their testimony indicated
that you support a pilot program for an increased tick for the
smaller capitalized companies, so I suppose the question would
be for Mr. Gasser or Mr. Mathisson. Do you agree with that
approach as well?
Mr. Mathisson. We would support a pilot program to test it
out. We think it would be a significant change to the market
structure, and as such, it should be done cautiously. But we
would definitely support a pilot program if explicit metrics
and goals were set out ahead of time to define success.
Senator Crapo. Thank you.
Mr. Gasser.
Mr. Gasser. We would also support that, Senator, and think
that from the perspective of capital formation it has
potentially positive or could have positive effects.
Senator Crapo. Thank you. And I guess this question is to
everyone but Mr. Mathisson since I am going to ask about his
recommendation that the immunity of SROs be addressed in this
system. Do you all agree with Mr. Mathisson's recommendation?
Mr. Mecane.
Mr. Mecane. Sure. The concept of being an SRO is founded on
or historically based on the idea of having obligations to the
market and getting certain benefits for meeting that
obligation. Some of the obligations include the rule-filing
process, the oversight that we have, the regulatory obligations
that we have, and some of the benefits that we accrue are
economic in nature. And we would encourage a review--which,
again, I think falls into the SEC's holistic review of market
structure--of the benefits and obligations of an SRO and what
an SRO means.
I think it is important to note that SROs by their nature
do not have blanket immunity. SROs have immunity when
performing certain of their regulatory functions. And beyond
that--and most of those regulatory functions are around our
function as a listing market. And when you expand beyond our
regulatory responsibilities, we have immunity that is set
contractually with our members.
We would encourage, again, a holistic review of all the
costs and benefits of being an SRO. Most of the items Mr.
Mathisson raised are focused on reducing the economics for
exchanges and transferring a lot of that benefit back to the
investment banks. Again, I think that should be part of the
holistic review that would potentially be done.
Senator Crapo. Mr. Noll.
Mr. Noll. Not unsurprisingly, I disagree with Mr. Mathisson
on this. You know, I do think that there are several things
about the exchange SRO model that are important to recognize as
we talk about our immunities and our limitational liabilities,
not the least of which is our obligation to provide fair access
to all investors under all circumstances. So exchanges by their
very definition exist to perform the price discovery function
by gathering trading interest from all participants who want to
participate and conduct that activity in a fair, transparent,
accessible manner for all market participants.
Under current market models, no other market participant
has those same obligations, and that also performs what I think
is a very critical role in the marketplace of establishing
price discovery, providing liquidity, and providing information
to not only investors and traders but issuing companies about
how capital should be formed, how it should be allocated, and
the risks in the marketplace. And I think because of the value
of that function in the marketplace, the immunities are well
established and should be enforced.
Senator Crapo. Mr. Gasser.
Mr. Gasser. Yes, I would agree with the examination as part
of a holistic examination of U.S. market structure. We were
actually, I think, very close to that--obviously, this issue
was not on the table. It was pre-Facebook, back in 2010. But we
were very close to what I thought was the end of a long comment
process, industry engagement, exchange engagement, and that was
obviously a pushback as a result of the Flash Crash. I think
bringing that back on board and engaging the same way with the
Commission amongst all market participants I think would be a
good thing.
In terms of immunity, I mean, I think we are all--as a
broker, as an agency broker, I think we are all feeling the
same thing. I do not want to speak for all brokers, but we are
all feeling the same thing. There has got to be consequence for
a system-wide failure of the type that we have experienced in
the Facebook circumstance. Our clients suffered. Other broker-
dealers suffered. Clearly, there were some decisions made that
were, with all due respect, the wrong ones in terms of opening
that stock. That is a very, very sharp contrast to the BATS
failed IPO where actually they walked away when they realized
that the software was not performing properly.
That does not mean that electronic markets cannot open
IPOs. I mean, they have done so successfully. Archipelago,
which Joe's firm acquired many years ago, did basically an
electronic auction and was very successful.
So the devil is in the details in terms of the technology
that we all operate today and its effect on the outcome, but if
the outcome is significantly negative and folks are harmed,
there has got to be some consequence felt. We certainly would
feel that consequence if we had a failure within our system and
it was attributed to our technology and acting as a fiduciary
on behalf of a client.
Senator Crapo. Thank you.
Chairman Reed. Thank you, Senator Crapo.
Senator Hagan, please.
Senator Hagan. Thank you, Mr. Chairman, and thank you for
holding this hearing. I think it is very interesting and
timely.
I just wanted to ask a couple of questions, one having to
do with the dark pools. Mr. Mathisson, this question is, first
of all, for this part directed to you. I have got definitions
of dark pools, but I would like to have you give me your
definition of a dark pool.
Mr. Mathisson. Sure. A dark pool is a trading system that
is organized under Regulation ATS that does not show bids or
offers. So it is a computerized trading system where people
enter a bid or an offer, say a buy of 100,000 shares of IBM,
that would not be displayed publicly. It just sits within the
computer. If somebody happens to come in with a sell order of
100,000 IBM, then the two cross. But the difference between a
dark pool and a displayed ATS or a displayed order on an
exchange is that the displayed order gets shown to the world,
so everybody sees there is a 100,000-share bid for IBM. In the
dark pool, it just sits in the hope that somebody might bump
into it.
Senator Hagan. How much of the trading is done on dark
pools now, percent-wise?
Mr. Mathisson. The best numbers out there are considered to
be Rosenblatt survey numbers, and Rosenblatt puts it at about
14.5 percent of volume.
Senator Hagan. And has that been growing since they have
been introduced after 1998, or however long?
Mr. Mathisson. That is right. It was probably----
Senator Hagan. Has it been growing in the last----
Mr. Mathisson. It has been growing, yes. It has picked up a
lot of volume at the expense of the TRF, the Trade Reporting
Facility. Broker internalization is down from about 26 percent
to 15 percent, and that 11 percent of the market moved to dark
pools. So we have seen broker internalization go down. We have
seen dark pool volume go up from approximately 4 percent 5
years ago to 14 percent today.
Senator Hagan. And what is typically the average trade in a
dark pool?
Mr. Mathisson. The average trade sizes are small. I mean,
they vary for different pools. Some pools are specifically
aimed at crossing blocks. Mr. Gasser runs a pool that has a
fairly high average crossing size. Some pools are as low as 200
shares, their average crossing size, which is comparable to
exchange average size crosses.
Senator Hagan. OK. Thank you.
Mr. Gasser, ITG has observed in the past that the traders
in these dark pools will often experience adverse selection,
and a trader will see a stock move in his or her favor
immediately after he or she executes a block trade in a dark
pool. The question is: Can you discuss why the adverse
selection is occurring in dark pools? And how does the lack of
transparency in these venues contribute to adverse selection?
Mr. Gasser. Yes, Senator. Actually, ``dark'' is probably an
unfortunate, nefarious term.
Senator Hagan. I agree with that.
Mr. Gasser. But I----
Senator Hagan. Why do we call it ``dark''?
Mr. Gasser. Because these are orders and executions that
are not exposed to the lit market, so basically a quote that
you can find on the screen. So I think that was basically the
origin of ``dark.''
ITG has been operating a dark pool for 25 years, and that
is POSIT. POSIT is the granddaddy of dark pools, if you will.
It is a system that brings together institutional investors,
allows them to interact with one another directly and
anonymously, so they do not know size, they do not know who
their counterparty is, and we improve price. So, in other
words, we trade at the midpoint, and then we report--and this
is very important. We report that trade immediately to the TRF.
So there is complete transparency in terms of the trade having
been done and now reported to the tape.
As Dan alluded to, our average execution size in our block
trading system is 29,000 shares, so it is significantly in
excess of the average size cross, and this is because it is an
institutional constituency.
I think over time what has happened over that 25-year
period that brings us to today is that internalization pools
have become more multiconstituent, if you will; in other words,
folks are not only crossing institutional flow, they are
crossing retail flow. They are acting as market maker. There is
high-frequency folks or, you know, strategies that are sitting
within those pools and interacting, and they look much more
like exchanges. So I think that has been a very steady
evolution over time.
To get to your point about adverse selection, one of the
things that we find is that, given the opportunity for
institutions to interact with one another directly, that is the
highest--and I think Dan would acknowledge that as well. That
is the highest-quality execution that one can achieve.
Now, one order may be larger than the other. That is
unavoidable in terms of negative selection. In other words, if
you are a buyer and the seller is ten times your size,
ultimately the price is going to go down if you extinguish that
order. But that is what I would describe as organic to the
environment we operate in.
What is much more difficult to control is that once you are
forced into the lit market and you are forced into other
trading venues, you now have less control over who you are
interacting with. And so adverse selection can come from dark
pools. It can come from lit pools. It can come from a variety
of different places. And there are predatory strategies that we
find every day, folks that are trying to, you know, basically
probe liquidity pools and find the size of your order, the
limit of your order, and are trying to trade against that
order.
Senator Hagan. Do you think that occurs in transparent
pools versus the dark pools? Where is it most likely to occur?
Mr. Gasser. I think they are interchangeable. I think they
come in and out of those markets, you know, looking for--you
know, it is----
Senator Hagan. Do the rest of you all agree with that?
Mr. Noll. I have a slightly different view of that. I
think, you know, Mr. Gasser said something I thought was
illustrative, which he said dark pools look a lot like
exchanges, except I would make some real distinctions around
that. You know, one is they do not have to display prices that
are accessible by outside investors. They are not required to
take all investors on an equal and fair basis. The rules are
not public. Their order types are not filed for a public notice
and comment. So there are some significant differences, and I
think that is where our concerns as exchange operators come in,
which is to avoid the negative externalities of off-exchange or
dark trading, which I will reiterate has real value for
investors, and POSIT is one of the real benefits for
institutional investors in the marketplace. It does what it is
intended to do, which is allow buy-side investors to find
liquidity in an efficient, information-less way.
But I do think that our concerns are really about
asymmetrical information leakage, fair access, and
nondiscriminatory activities.
Mr. Gasser. Just for clarity, when I said some dark pools
look like exchanges, I was actually alluding to the average
execution size, so other dark pools that are outside the POSIT
system.
Senator Hagan. Mr. Noll, let me ask you a question having
to do with the minimum price variation. In the U.S., the MPV
for all stocks of a dollar is one penny, and in Europe, their
minimum price variations are less uniform. How do MPVs impact
high-speed trading? And particularly I am looking at like a
stock--at Google at $670 a share versus B of A at $9.50 a
share.
Mr. Noll. I think one of the things that we have noticed in
the marketplace is that actively traded securities, not
necessarily limited to stock price, so that certainly is a
factor to consider here, but actively traded securities trade
extraordinarily well in our national market system today. There
is price discovery, there is liquidity, there is activity.
I think where we become very concerned is outside of those
actively traded names, and European markets, as you alluded to,
Senator, use what they call an ``intelligent tick size regime''
to determine what the appropriate tick size is based on a
variety of characteristics for the underlying security, a lot
of which have to do with liquidity in addition to price amount,
available float, size of the bid-offer spread. And so where we
are interested in looking at this as a pilot--and I agree with
the rest of the panel here that a pilot is a good thing for us
to do here, particularly if set up well--is to uncover ways in
which we can get better liquidity in less actively traded names
and smaller cap securities. And I think minimum price variation
will help us do that. I think a pilot will help us address that
as we go forward.
Senator Hagan. Thank you.
Thank you, Mr. Chairman.
Chairman Reed. Well, thank you very much, Senator Hagan,
for your excellent questions.
Let me just begin another round, and if Senator Crapo
returns, obviously, I would invite him to join, too.
I think we all recognize that the interrelationships of
these different issues are such that requires across-the-board,
marketwide approach, holistic in Mr. Mecane's terms, but I
think we also have to look at some sort of specific issues and,
therefore, we have been talking about dark pools and ATSs, et
cetera.
Just with respect to stepping back, the national market
system, my sense, the motivating force was to make sure that
orders sent to the trading platform were the best price,
regardless of where the orders originate. And with these dark
pools in particular, that does not seem to be the case. Is that
a fair--since to be a member of a dark pool, that is not
something that is--anyone can do or--let me use that as a
prelude to ask Mr. Gasser to comment and others to comment.
Mr. Gasser. I keep coming back to the notion of a broker
being a fiduciary on behalf of the client order. So we have
very, very--the U.S. has very prescriptive regulations about
how we report our execution quality, how we relate that back to
customers. Obviously transaction cost analysis is core to what
we do as a firm. So, you know, the highest priority is not to
internalize order flow within POSIT for us. It is to find the
best price. So that could exist in the lit market, other dark
markets, or our system.
Now, fortunately for us, I think we can demonstrate a long
history of very, very substantial cost savings, trading cost
savings within our dark pool. But that is part of an ecosystem
that we are forced to aggregate with technology every day, and
we do not aggregate it based on being able to internalize it.
We aggregate it based on the ability to find the best price,
and we know that we are all--as brokers, we are going to be
measured by the customer at the end of the day, and the
customer votes typically with their feet in terms of quality of
execution.
Chairman Reed. Mr. Mecane, do you have a comment. Then Mr.
Mathisson, because this is----
Mr. Mecane. Sure. Just to your question, Senator, I think
it is a slightly different issue, at least from our
perspective. As far as I know, most executions, dark pools,
wherever, do happen at what the best price is in the public
markets. From our standpoint, there is a slightly different
issue that we think should get addressed, which is whether the
person who is setting that best price in the public markets is
properly incentivized, meaning the person who is creating that
best bid or offer price very frequently, roughly one-third of
the time, will see trading happening at the price that he is
displaying. And while the executions that are happening away
from the public market are at that best price, the concern that
we have from a broader public policy standpoint is whether
there is diminished incentive for people to display their
orders publicly if a significant amount of activity happens in
front of their displayed price. And so when we have talked
about incentives for people to display liquidity, whether the
right balance is in place to ensure that the person who sets
the best price, and Regulation NMS in its main mission set out
properly incentivizing the public display of orders as its
objective, we would just highlight that we should probably step
back and review whether we think the balance that has evolved
is the right one.
Chairman Reed. Mr. Mathisson? And then I have another
question. I will change the topic slightly.
Mr. Mathisson. Yes, when traders decide whether to put out
a bid order on an exchange or on a displayed ATS or put one
into the dark, on a dark pool, they are making a tradeoff
decision. If you show it publicly, then you get what is called
NMS protection, meaning that nobody can trade through your
price. So, in other words, if I show a bid at 27.50 into the
national market and I choose to display it, then a trade cannot
occur at 27.49 or lower anywhere. It cannot occur in a dark
pool, it cannot occur anywhere. So I have what is called
protection. Now, that is the plus side.
The negative side of doing it is that I show this bid which
might scare the price up, and that is a tradeoff that traders
decide to make, and they are constantly weighing whether it is
better to display it and get protection and potentially attract
a seller or whether it is better to put it in a dark pool, take
on a lower fill rate, but have less chance of scaring the
price. And that is essentially what trading is. It is a
tradeoff between what we call signaling risk and fill rate. And
that is what trading algorithms do. That is what people,
traders, are deciding. And so the system allows traders to make
that choice.
Now, Regulation NMS does ban the trade outside of the trade
price. Once you put that price up, trading cannot occur in a
dark pool at a different price, and the dark pool trades are
reported to the tape immediately, so the whole world does know
about them after the fact, and they do occur at or within the
national best bid or offer.
Chairman Reed. Thank you. This testimony and the questions
have been very thought-provoking, so you can anticipate
additional questions in writing, in fact, some detailed
questions, and I know you will respond.
Let me change, again, quickly from my last set of
questions, and then I will recognize Senator Crapo again.
We do understand this is a systemic approach--we should
take a systemic approach, but issues crop up. One of them is
the order type issue, which is getting a lot of attention. The
proliferation of order types adds to the complexity, raises
lots of questions, and it goes back to sort of the public
perception of markets, too. Are these orders being written to
induce people to come into your market to give them tailored,
bespoke approaches to the market which other particularly
retail investors do not have? Are there too many of these?
Should there be a standard set of order types that every market
alternative as well as exchanges follow? And this general
question of order types I think is important, at least to
broach at this point. So let me start quickly--and, again,
because I want to recognize as briefly as possible Mr. Mecane
and then we will go down the panel.
Mr. Mecane. Sure. It is a good question, very topical. I
think there are two main points that I would draw out. One is
that it is important to recognize why a lot of these order
types have evolved. Some of them are to comply with Regulation
NMS; some of them are to guarantee economic results; some of
them are to compete with some of the practices, customer
segmentation, et cetera, that happens off-exchange. So I will
go back to one of my earlier points, that the order type
evolution is largely because the market structure that we have
creates the need or the demand for different order types to
replicate certain behaviors, some of which used to happen
nonelectronically. But, again, my first point is if we want to
review the order type issue or simplify the markets, we should
simplify the market structure that they operate in, and there
will be less need for these order types.
The second point I would just make is that it is important
to recognize that all exchange order types are publicly filed,
reviewed by the SEC, put out for public comment, disclosed, and
can be utilized by any member of an exchange. Now, if there are
any order types that are not adequately disclosed or not
described accurately in the filings, that I think is a separate
issue that needs to be addressed. But at least in terms of the
intent of the process, all the rule filings should describe
order types and order type behaviors, which, again, is
something that is very distinct from how nonexchanges operate.
Dan's and Bob's firms have algorithms. There are probably
more algorithms for institutional customers in the industry
than there are order types for broker-dealers to use. So I
think it is just the evolution of our complex market structure.
Chairman Reed. Mr. Mathisson, just for clarification, is
there a requirement for ATS to submit order types or algorithms
for review?
Mr. Mathisson. No. I mean, they do get filed in what is
known as a Form ATS, but it is not reviewed and subject to
approval. You just tell the SEC what you are doing. So that is
correct.
Chairman Reed. All right. The order type issue.
Mr. Mathisson. I agree with Mr. Mecane that complex
orders--that the line between algorithms that broker-dealers do
and the line between complex order types that exchanges offer
is blurry, and there are a lot of similarities between the two.
But I think that the broker-dealer is the right place to have
that complexity, and the reason is because the broker-dealer is
a fiduciary to the client. The broker-dealer has an error
account, and the broker-dealer has liability. And complex order
types are more likely to fail than simple matches. I would be
in support of a market system where ATSs and exchanges only
offer very simple order types, just buy-and-sell matches,
market on close, market on open, just the traditional, most
simple order types, and put that complexity within a structure
where there is an error account, there is a fiduciary
responsibility, and there is liability if things go wrong. The
more complicated an order type gets and the more complicated an
algorithm gets, the more likely it is to fail, and so the
question becomes who eats the error when that fails. With an
exchange, the answer is--you know, the exchange does not eat
the error. The error gets borne by the broker-dealer who used
the order type.
Now, finally, the order types, while they are reviewed--all
the exchange order types are reviewed by the SEC. The Wall
Street Journal put out a piece that said every single order
type that has ever been submitted by the exchange has been
approved. There has never been one that has been rejected. You
know, so I would question how thorough and meaningful a review
that actually is.
Chairman Reed. Thank you.
Mr. Noll? And then again, to allow Senator Crapo, your
brief comments.
Mr. Noll. Well, factually what Mr. Mathisson says about
order types, from my own experience around the SEC review of
order types and order type introduction, we have withdrawn many
more order types at the suggestion of the SEC than we have had
approved. So while the public notice and comment period does,
in fact, exist and so I do think historically order types have
been improved, many, many order types, many variations on order
types have, in fact, been asked--we have been asked by the SEC
to withdraw them for a variety of reasons, having to do with
their view of what is the appropriate market structure. So they
do not go unreviewed I guess is the point I would try to make.
I agree with Mr. Mecane on most of what he said. I think
order types have evolved in various ways, some due to
regulatory reasons, Regulation NMS being the primary one; but
also because in an electronic market--and make no mistake about
it, our markets are electronic today--order types fill the
function and allow the function of what used to be done
manually to be done electronically and with technology. And
order types are a tool to do that. And I think it is important
to recognize that they need to be applied fairly, they need to
be made available to everyone, and they have to be used so that
they are not favoring one type of party over another in an
unfair way. And I think we have strived to achieve that, and I
think for the most part we have been successful with that.
Chairman Reed. Mr. Gasser, your comment, quickly.
Mr. Gasser. Senator, you heard in your first hearing, I
think there is a significant cross-section of the institutional
investor community that has become frustrated with the
permutation of these order types. At the end of the day, I
think our view is--as Dan alluded to, as a fiduciary, it is our
job to make sure that their orders are properly represented and
that we are taking advantage of all the technology available to
use, and it is all in the public domain. So I think it is--from
the perspective of preventing competition and further
innovation, I would be hesitant to say that we need to prevent
further innovation there on that front.
Chairman Reed. Thank you very much.
Senator Crapo.
Senator Crapo. Thank you, Mr. Chairman. I just had one
final issue to get into.
Mr. Mathisson, in your written testimony, you indicate that
so-called quote flickering has increased with decimalization.
And, similarly, Mr. Gasser, you indicated in your testimony
that you recommend that market participants who create
excessive quote traffic without executing order flow should
bear the costs of this activity.
How does quote flickering work? And what is its impact to
the system? Mr. Mathisson first.
Mr. Mathisson. Well, quote flickering is when somebody
shows a bid or an offer for a very short amount of time and
then keeps changing it repeatedly. Now, the cost of it on the
system--I mean, the reason it is done is it would be typically
done by high-frequency trading systems that are just making
decisions very rapidly and they are deciding they want to buy
something and the price of something else moves and they
decide, you know, a thousandth of a second later to pull the
bid or the offer. And so the result is that they can get quote
flickering. It is empirical fact that quote flickering has
increased. The number of quotes per day is up dramatically
versus what it was about 5 years ago, and it is up, you know,
something like 100 times from what it was 10 years ago.
Now, the cost of quote flickering is primarily in
technology infrastructure. When you have a lot of quoting going
on, people need to build networks and computer systems that are
capable of digesting hundreds of millions of quotes a day. And
so it is an expense, essentially, on everybody. All trading
systems by broker-dealers and typically by mutual funds and
pension funds and other investors read all these quotes and
respond to them, and so they all have to spend a great deal
more money on technology than they otherwise would.
Senator Crapo. Thank you.
Mr. Gasser.
Mr. Gasser. Yes, in 2010, we gave a speech to an industry
trade group and recommended, you know, some type of tax on what
is--you know, there is no benefit to the broader national
market system of having folks come in and basically flicker on
the bid or the offer, as Dan alluded to. We spend a tremendous
amount of money on infrastructure to support that, as I said,
with no discernible benefit to investors or anyone else in the
ecosystem.
While equity volumes have steadily declined, the duress or
the stress that it places on our infrastructure has increased
logarithmically, so we are getting less liquidity, less volume,
and we are being forced to support this activity in terms of
our own investments in our business.
Senator Crapo. Thank you.
Mr. Noll and Mr. Mecane, do you want to comment.
Mr. Noll. Sure. I agree that quote stability is an
important goal for exchanges and for the marketplace. You know,
we at Nasdaq have tried to take some steps on our own to bring
more quote stability to the marketplace. So, you know, the
first thing we have done is introduce an order type called the
minimum life order type that creates economic incentives to
have an extended quote as opposed to a short quote. We have
also introduced and I believe we are the only exchange today
that has an excessive messaging fee for rapid quote changes
without trades and without taking on discernible obligations in
the marketplace.
And then, last, we introduced a new market model called
PSX, which is a price six exchange, which was designed really
to get away from being at the top of the book because you were
there first in terms of your quote ability, but being there in
size as the incentive for the market participants. I wish I
could say that that has had more success in the marketplace in
developing as a real alternative to some of the other market
models. That has yet to do that. But I do think that there is
room for innovation here. I do think achieving quote stability
is an important goal in the marketplace. I think it makes the
markets better for everyone.
Senator Crapo. Thank you.
Mr. Mecane.
Mr. Mecane. I would just add two quick points. I think we
would all agree that message traffic in general is a high cost
for the industry and for all of us. The two points I would
make, the first is just a recognition, again, that some of that
message traffic or a lot of that message traffic arises not
only because of Regulation NMS but because there are 50 venues
out there for people to trade on. So in a lot of cases, traders
will put their quotes in a subset of that many venues. They get
executed on one venue, and then they pull their quotes from
everyone else. And so some of the market complexity does
contribute to the amount of message traffic. So, again, that
points back to the holistic review of market structure.
The second is while the consolidated audit trail is still
very much a development in process, one of the discussions is
around whether there is a mechanism through the consolidated
audit trail to potentially align some of the costs of
surveillance and producing the consolidated audit trail records
back to the person who has created the quotation. And so there
is potentially another mechanism down the road to help address
message traffic.
Senator Crapo. Thank you very much. I really appreciate the
testimony of these witnesses. It has been very helpful.
Chairman Reed. It has been very, very helpful, and Senator
Crapo's question has raised a question in my mind, a real
question not a rhetorical one, and I will just see if anyone
would like to respond to it. Is this flickering phenomenon
related to entities that do not have market-making
responsibilities so that they can come in and just ping the
system without any responsibility to actually follow through?
Is that related and we should be thinking about that?
Mr. Gasser. Yes, I would not say that I know for certain,
but I think it stands to reason that it is probably--and I
certainly do not want to tar the HFT community with a brush
here, but it is probably a high-frequency strategy, as Joe
alluded, that has been deployed into every market center
liquidity pool. We have seen, you know, some pretty significant
anomalies from time to time, and, you know, 10 o'clock in the
morning, 10:30 in the morning, where you have got these just
huge spikes. They tend to be focused on large cap liquid names,
and to get back to the point of investor perception, if folks
had, you know, more insight into this, I cannot see that it
would be a positive from that perspective. In fact, it would
probably be perceived as just another example of a participant
that had the ability to affect other participants in orders of
magnitude larger than the capital they have deployed or their
desire to actually execute trades.
Chairman Reed. So I think that unless someone objects, this
issue of--one of the categories we have to look at in market
formulation is the responsibilities of market makers, who
should be a market maker and what rights, what responsibilities
they have, and that is an open question now given the new
technology. I assume that that would be a point of agreement by
the panel.
I want to thank you all again. Senator Crapo was absolutely
correct. Your testimony has been both thoughtful and thought-
provoking, so expect lots of written questions.
We appreciate the time that you have made. The preparation
that you have undertaken has been obvious in this testimony.
I will ask my colleagues, all the colleagues on the
Subcommittee, if they have additional questions or written
statements, please get them to us by December 28th, and then we
will get them to you as quickly as possible and ask for your
prompt response.
The only thing, as we have been going back and forth about
these dark pools, and I think you, Mr. Gasser, made the point
that it is probably not the best term, I think there might be a
rule of thumb that anything that could be used as a title for a
Batman movie is probably not something that is good in the
financial markets. But that is just a quip.
With no further questions, I will adjourn the hearing.
Thank you, gentlemen.
[Whereupon, at 10:51 a.m., the hearing was adjourned.]
[Prepared statements and responses to written questions
supplied for the record follow:]
PREPARED STATEMENT OF JOSEPH MECANE
Executive Vice President and Head of U.S. Equities, NYSE Euronext
December 18, 2012
Introduction
Chairman Reed, Ranking Member Crapo, and Members of the
Subcommittee, my name is Joe Mecane and I am EVP and Head of U.S.
Equities at NYSE Euronext (NYX)--a leading global operator of financial
markets and provider of trading technologies. NYX's exchanges in Europe
and the U.S. trade equities, futures, options, fixed-income, and
exchange-traded products. In the U.S., we operate three equities
exchanges, two options exchanges, one futures exchange, and a
technology business that provides comprehensive commercial technology,
connectivity, and market data products and services.
While the U.S. continues to have the most liquid markets in the
world and remains at the forefront of innovative technology used to
conduct electronic trading, the infrastructure used to operate the
markets each day has grown so sophisticated that few fully appreciate
how well our markets actually operate in a highly competitive,
fragmented, and complex environment. This has made it difficult for
market participants, regulators, and Congress to determine the extent
to which the growth in the number of trading venues, the speed at which
trading platforms operate, and use of automated trading are beneficial.
However, in light of the market events that have occurred in recent
years, I'd like to focus on how technology and our market structure
have created unnecessary complexity and mistrust of markets; and,
relatedly, what NYX believes the industry, regulators, and Congress
should be doing to address it.
Market Structure Drivers Toward Computerized Trading
Decimalization. Electronic trading has added tremendous benefit to
the capital markets, including lower costs of execution, faster speed
of execution and, in some cases, greater transparency. However, the
trend toward computerized trading was accelerated and fostered by
several significant regulatory changes that drove the market to become
more electronic. One important factor was decimalization of the markets
in 2001, which had an effect of decreasing average spreads by roughly
38 percent in NYSE- and NASDAQ-listed securities, directly benefiting
end investors. \1\ At the same time institutional commissions, borne
directly by end investors, were declining and decreased 33 percent \2\
in the years leading up to Regulation NMS (Reg. NMS) implementation. In
fact, almost all reductions in spreads and commissions occurred prior
to the implementation of Reg. NMS and led to a huge expansion of
electronic trading because human traders could no longer effectively
make markets in this environment, and because institutions and brokers
began relying more on algorithmic trading to access the market and
reduce their costs of trading. This began a steady progression to have
the most sophisticated algorithms and technology, since the smartest,
the fastest, and the first prevailed--well before the implementation of
Reg. NMS in 2007.
---------------------------------------------------------------------------
\1\ Data is calculated based on decrease in dollar value of
spreads between 2001 and 2007, when the next major market structure
changes were implemented through Reg. NMS. Consolidated Tape
Association and NYX.
\2\ Tabb Group: U.S. Long-Only Institutional Average Commission
Rates, 2005-2012.
---------------------------------------------------------------------------
Regulation NMS. In 2007, just as the technology among the trading
community was becoming more sophisticated, the Securities and Exchange
Commission (SEC) adopted Reg. NMS. This regulation gave brokers the
freedom to trade around markets such as the New York Stock Exchange
(NYSE) when the NYSE was in ``slow'' mode, \3\ and at the same time
forced participants to access the national best bid or offer (NBBO) in
the market. Because exchanges competed by establishing the NBBO, speed
among markets became the competitive differentiator based on one
exchange's ability to set the NBBO faster than a competing market.
While Reg. NMS also established the Order Protection Rule to protect
visible orders and encourage displaying quotes, today more than 3,000
securities have over 40 percent \4\ of their volume occurring off-
exchange in dark markets. In the NYSE MKT listed market, which
represents 709 securities, off-exchange trading accounted for 42
percent of the volume in November. This level of off-exchange activity
erodes the incentive for market makers to continue to trade the less
active securities, has a negative effect on price discovery \5\ and
threatens to further decrease the incentives for companies to go
public.
---------------------------------------------------------------------------
\3\ Reg. NMS: http://www.sec.gov/rules/final/34-51808.pdf.
\4\ Consolidated Tape Association.
\5\ CFA Institute: ``Dark Pools, Internalization and Equity Market
Quality'', October 2012, Weaver: ``Off Exchange Trading and Market
Quality in a Fragmented Market'' (May 2011), Tabb Group: ``A Spotlight
in the Dark: An Inevitable Debate'', November 2012.
---------------------------------------------------------------------------
ATSs and internalization. Today, there are around 63 execution
venues in the U.S. markets, including 13 exchanges and 50 dark pools.
Exchanges find themselves competing more directly with Alternative
Trading Systems (ATSs or dark pools) and broker internalization, which
are able to employ different practices than exchanges with far less
oversight and disclosure. Some of this competition is through cost,
some through order handling practices, and much of it is through client
segmentation whereby nonexchange venues are able to incentivize their
own or third party liquidity provisions based on the nature of the
person they are trading against. As a result of this advantage, large
broker-dealers continue to move more order flow into their own private
trading venues for a ``first look'' before routing on to the lit public
markets. Since the implementation of Reg. NMS, we've seen two markets
evolve--the lit public, regulated and accessible market versus the
dark, selective and private nontransparent market.
As you can see, technology and the rules that govern the U.S.
equity markets have resulted in the creation of a trading
infrastructure primarily focused on speed and resulting complexity
through which professional traders can identify and access liquidity--
too often at the expense of retail investors and market integrity. To
accomplish this, exchanges, brokers, and vendors have had to build
expensive networks with the capacity to keep up with the growth of
messages delivered each day to market participants seeking liquidity,
as well as learn how to interact in a very complex ecosystem.
In response to this new flow of orders, exchanges have developed
new order types. Order types have different purposes, such as giving
cost certainty or competing with the client segmentation that exists
off-exchange. Regardless of the reason for the specific order type,
most are premised on the goal of attracting liquidity back to the
public markets for the purpose of enhancing transparency and price
discovery. Moreover, all order types must be pre-approved by the SEC
and published for public comment, something that is unique to exchanges
and which does not exist for ATSs or brokers who internalize.
Recommendations
The bottom line is that our market structure incentivized these
various levels of increased complexity. Our main message is that if we
want to reduce the complexity of technology and our markets, we should
simplify the overall market structure. Doing so would certainly prove
beneficial for the future of our national market system, for investors
and issuers, and to the growth and well-being of our economy--including
efficient access to capital to fund innovation, new business and job
creation.
In this regard, key questions include determining who should lead
the change process, and what should be done to correct course while
ensuring that we continue to have the most transparent and liquid
markets in the world.
NYX believes that the SEC is best suited to propose meaningful
market structure changes--and, in fact, regulators in other global
markets, including Canada, Australia, and Europe, are already taking
action. With Congressional oversight, the SEC should continue with the
holistic review it began in 2010 with the Concept Release on Equity
Market Structure \6\ by proposing changes that will promote additional
transparency, fairness, and long-term capital formation. This
unfinished initiative needs to be completed and made a 2013 priority.
---------------------------------------------------------------------------
\6\ http://www.sec.gov/rules/concept/2010/34-61358.pdf
---------------------------------------------------------------------------
We believe that changes to be considered should include a review of
market maker obligations, the Sub-Penny Rule, the Order Protection
Rule, tick sizes for illiquid securities, and addressing the conflicts
and overlap between broker-dealers and exchanges, including the
obligations and responsibilities of each when providing like services.
The Consolidated Audit Trail, proposed by the SEC, also is a vital
component to ensuring effective surveillance in a highly fragmented
marketplace. Such surveillance should include better identification and
reporting on high frequency trading, similar to that being discussed by
the Commodity Futures Trading Commission, to increase the transparency
of this practice.
NYX also believes that, in light of the existing complexity of the
markets and the technology and trading glitches that have occurred this
year, all trading venues should ensure a robust set of policies and
procedures around their systems development life cycle. Although
testing may not be the most exciting part of our markets, the hyper-
competition that exists in this industry lends itself to excessive
levels of change rates just to remain competitive and compliant with
new regulatory requirements. The industry has been faced with
implementing new back stops such as single-stock circuit breakers,
market-wide circuit breakers, limit up-limit down, and possibly kill
switches. These regulatory mechanisms have cost the industry tens of
millions of dollars to implement over the past several years and have
been developed in response to some of the negative effects of highly
complex markets, in an effort to protect against those inevitable
situations when the unforeseen occurs.
Conclusion
In closing, I want to reiterate our belief that although our
capital markets are the best in the world, there remains room for
improvement. Technology and innovation should not be the cause of
crisis and fear in our markets. Under the right conditions and
structure, they are assets and produce opportunity for all market
participants. Our recommendations have a simple premise: implement
market structure changes that enhance transparency, fairness and price
discovery for investors and level the playing field for trading venues.
Thank you for inviting me to testify and I look forward to your
questions.
______
PREPARED STATEMENT OF DANIEL MATHISSON
Head of U.S. Equity Trading, Credit Suisse Securities LLC
December 18, 2012
Introduction
Good morning, and thank you for giving me the opportunity to share
my views on the best structure for our Nation's stock markets. My name
is Dan Mathisson, and I am the Head of U.S. Equity Trading for Credit
Suisse. \1\
---------------------------------------------------------------------------
\1\ Credit Suisse provides its clients with private banking,
investment banking, and asset management services worldwide. Credit
Suisse offers advisory services, comprehensive solutions and innovative
products to companies, institutional clients and high-net-worth private
clients globally, as well as retail clients in Switzerland. Credit
Suisse is active in over 50 countries and employs approximately 48,400
people. Credit Suisse is comprised of a number of legal entities around
the world and is headquartered in Zurich. The registered shares (CSGN)
of Credit Suisse's parent company, Credit Suisse Group AG, are listed
in Switzerland and, in the form of American Depositary Shares (CS), in
New York. Further information about Credit Suisse can be found at
www.credit-suisse.com.
---------------------------------------------------------------------------
The U.S. broker-dealer subsidiary of Credit Suisse Group has been
operating continuously in the United States since 1932, when the First
Boston Corporation was founded. Today, Credit Suisse employs
approximately 9,200 people in the United States, and 48,400 people
globally. We are one of the largest U.S broker-dealers, executing 12.4
percent of all U.S. equity volume in 2012. Most of that volume derives
from our 1,600 institutional clients, which include the largest mutual
funds and pension funds in America, representing the savings of tens of
millions of Americans.
I have been working in the U.S. equity markets for more than 20
years, the last 12 of which have been at Credit Suisse in New York.
This is the second time I have been given the privilege of appearing
before this Committee, and I appreciate the chance to appear here
today.
Summary
Credit Suisse believes that equity market quality has improved
markedly over the past two decades, and that the competition spurred by
the adoption of Regulation ATS and Regulation NMS has benefited the
average investor. However, there is still plenty of room for
improvement in the market structure. Within the past decade, our
Nation's exchanges have transitioned to a for-profit model, after more
than 200 years as not-for-profit, member-owned organizations. Despite
their new for-profit status, exchanges have retained quasi-governmental
status as SROs (Self-Regulatory Organizations), and exchanges still
receive significant public funding through the market data revenue
plans. We believe that this new model for the markets has proven itself
to be costly to investors, unfair to broker-dealers, and rife with
conflicts for the exchanges themselves. We suggest that ending
exchanges' status as SROs and transferring those regulatory
responsibilities to FINRA or the SEC would put all market players on a
level playing field and would benefit the average investor by creating
markets that would be simpler, less vulnerable to disruptions, and less
expensive to operate.
1. Are the U.S. markets working effectively?
Although the markets are not perfect, Credit Suisse believes that
the market structure changes of the past 20 years have been successful
in their goal of creating equity markets that are better than in the
prior era. The empirical evidence shows that Regulation ATS and
Regulation NMS have led to an increase in liquidity and a decrease in
the total number of market disruptions. We have found this holds true
for both large and small issuers.
Credit Suisse recently completed a broad survey of market quality
in the U.S. equity market, and found that in every empirical measure,
the U.S. markets are functioning better than ever. \2\ The study found:
---------------------------------------------------------------------------
\2\ Ana Avramovic, ``Who Let the Bots Out?'' Credit Suisse Trading
Strategy, May 2012. Also see ``June 2012 Chartbook'', Credit Suisse
Trading Strategy.
---------------------------------------------------------------------------
Positives:
Overnight market volatility in 2012 is at a 15-year low.
Intraday market volatility has been steadily decreasing
since 2005.
Bid-Ask spreads in the U.S. are the tightest in the
developed world.
Bid-Ask spreads have been clearly and steadily declining
since Reg NMS was introduced, controlling for volatility.
Average size of bids and offers has increased since 2004.
The number of market disruptions, a.k.a. ``mini flash
crashes'', has been decreasing since 2000.
Negatives:
Quote flickering has increased, with the number of daily
changes in the NBBO (National Best Bid Offer) per million
shares traded at an all-time high in 2011.
Overall the study concluded there was no empirical evidence of
negative market performance other than the increased cost of message
traffic. Many other academic studies have found similarly positive
results. A Rutgers University study released in September 2012 that
examined data back to 1993 concluded that market quality breakdowns are
41 percent less frequent post-Reg NMS than prior to the rule. \3\ A
Feb. 2010 broad review of the equities markets by three well-respected
professors concluded, ``Virtually every dimension of U.S. equity market
quality is now better than ever.'' \4\
---------------------------------------------------------------------------
\3\ See, Cheng Gao and Bruce Mizrach, Rutgers University, ``Market
Quality Breakdowns in Equities'', Sept. 2012.
\4\ James Angel, Lawrence Harris, Chester Spratt, ``Equity Trading
in the 21st Century'', Feb. 23, 2010.
---------------------------------------------------------------------------
2. Have liquidity and price discovery been impacted by the flow of
stock trading volume to off-exchange venues?
Implicit in this question is a statement that volume has been
moving to off-exchange venues, which is factually incorrect. Over the
past 5 years, volume has not shifted to off-exchange venues. Figure 1
shows the percentage of U.S. volume executed on exchanges from January
2008 through November 2012. As can be seen, the percentage of volume
executed off-exchange has been remarkably constant over the past 5
years.
While the level of on-exchange vs. off-exchange volume has stayed
remarkably constant, there has been a shift in volume from NYSE and
Nasdaq to two new exchanges: BATS and DirectEdge. As seen in Figure 2,
from October 2007 to October 2012, Nasdaq and the NYSE floor lost a
combined 24.6 points of market share. New exchanges BATS and DirectEdge
gained 22.4 points over that same period. Virtually all of the loss in
the traditional exchanges' market share is explained by the rise of
these two well-managed and efficient exchange competitors.
3. How does the current market structure impact market integrity or
investor confidence?
Credit Suisse believes that the current market structure is not
optimal for investor protection and market integrity. The aftermath of
the Facebook IPO on May 18, 2012, revealed a significant flaw in the
existing market structure. As has been widely reported, the Nasdaq
exchange experienced extensive system failures during the initial
public offering of Facebook Inc., causing others to suffer losses
estimated to exceed $500 million. \5\ Nonetheless, Nasdaq has only
offered a $62 million settlement to those that suffered losses due to
Nasdaq's failures, an offer it views as an ``accommodation,'' given its
view that it is legally immune from liability. \6\
---------------------------------------------------------------------------
\5\ Ashley Lau, ``Nasdaq Defends Facebook Compensation Plan-Letter
to SEC'', Reuters, Sep. 19, 2012, available at http://www.reuters.com/
article/2012/09/19/us-nasdaq-facebook-letter-idUSBRE88I18V20120919.
\6\ See, Self-Regulatory Organizations; The NASDAQ Stock Market
LLC; Notice of Filing of Proposed Rule Change to Amend Rule 4626--
Limitation of Liability, Exchange Act Release No. 67507 at 4 (July 26,
2012).
---------------------------------------------------------------------------
Because the Exchange Act of 1934 vests the exchanges with self-
regulatory authority, courts have traditionally afforded exchanges
``absolute immunity'' from civil liability for damages arising in
connection with their regulatory operations. \7\ The basis for this is
the common law doctrine of sovereign immunity, under which a Government
entity may not be held liable for acts taken in its official capacity.
Because an exchange is empowered to perform a ``quasi-governmental''
regulatory function, courts have found that exchanges ``stand in the
shoes'' of the SEC, and they receive the same immunity that the SEC
would be granted. \8\
---------------------------------------------------------------------------
\7\ See, e.g., Barbara v. NYSE, 99 F.3d 49 (2d Cir. 1996); DL
Capital Group, LLC v. Nasdaq Stock Market, Inc., 409 F.3d 93 (2d Cir.
2005).
\8\ See, e.g., D'Alessio v. NYSE, 258 F.3d 93 (2d Cir. 2001).
---------------------------------------------------------------------------
It is a dangerous situation when a for-profit enterprise can cause
half a billion dollars of losses for others, and not have the risk of
being held legally liable. Instead of bearing the cost of its own
failures, Nasdaq believes that ``the risks associated with system
malfunctions should be allocated among all exchange members, rather
than being borne solely by the exchange.'' \9\ In other words, Nasdaq
asserts that the costs should be borne by the shareholders of the for-
profit broker-dealers that suffered the harm, rather than the
shareholders of the for-profit exchange that caused the harm.
---------------------------------------------------------------------------
\9\ Id. at 28.
---------------------------------------------------------------------------
The contrast to recent events involving Knight Capital Group is
striking. On August 1, 2012, a system error caused a Knight Capital
Group broker-dealer subsidiary to send a slew of erroneous orders,
resulting in $440 million in losses. Since Knight is not an exchange,
and therefore does not benefit from sovereign immunity, rather than
externalizing these costs on the market as a whole, Knight's
shareholders suffered the losses. \10\
---------------------------------------------------------------------------
\10\ See, e.g., Scott Patterson, ``SEC Nixed Knight's Plea for a
Do-Over'', Wall St. J., Aug. 6, 2012, available at http://
online.wsj.com/article/
SB10000872396390444246904577571113923528168.html.
---------------------------------------------------------------------------
A fundamental principle of the law is that if a private enterprise
wrongfully causes harm to others, it may be held liable to pay for the
financial damages that ensue. In conducting its operations and
implementing new systems, a broker-dealer must consider the risks and
potential costs of potential liability and act accordingly. An
exchange, on the other hand, may operate in a reckless manner.
Sovereign immunity may have made sense when exchanges were not-for-
profit, member-owned regulatory organizations that existed for the good
of their members. But today, the NYSE and all exchanges are for-profit
enterprises that are not particularly different from broker-dealers.
While they still have a few vestigial regulatory functions, they
outsource the vast majority of their regulatory responsibilities to
FINRA.
Exchanges now function as broker-dealers in many ways. For example,
Nasdaq announced in May of 2012 they would compete with broker-dealers
by selling execution algorithms, which involve significantly more
complex technology than simply crossing stock like the Facebook IPO.
\11\ Complex trading technology like algorithms should go through
rigorous quality assurance testing, and maximum caution should be
exercised when rolling out these types of programs. We believe that
providers of trading technology will naturally exercise greater caution
if they have material liability when their technology fails.
---------------------------------------------------------------------------
\11\ ``Nasdaq to Offer Algorithms, Competing with Brokers'', by
Nina Mehta, Bloomberg News, May 14, 2012. Article quotes Professor
Bruce Weber saying, ``Before electronic trading really took off, it was
clear where the exchange function ended and the brokerage function
began. That line is getting blurred.''
---------------------------------------------------------------------------
Repairing this problem in the current market structure is
especially important since Regulation NMS does not allow broker-dealers
to ignore an exchange's bids or offers, essentially compelling brokers
to trade with every exchange, whether or not they find an exchange's
technology to be reliable, and whether or not they find the exchange's
liability policy to be fair and equitable. \12\ Policy makers should
examine whether it still makes sense for exchanges to be considered
quasi-governmental entities, given that they are no longer member-
owned, no longer not-for-profit, and no longer have much of a direct
regulatory function.
---------------------------------------------------------------------------
\12\ See, Regulation NMS Rule 611(a).
---------------------------------------------------------------------------
4. Are exchanges and dark pools on a ``level playing field''?
Regulation ATS was specifically passed to allow broker-dealers to
create electronic crossing networks that automated their traditional
job of crossing client orders. ATSs, a subset of which are known as
``dark pools'', operate under a very different regulatory structure
than exchanges. Nasdaq and NYSE have claimed that regulators need to
ensure that exchanges and dark pools are on a ``level playing field''
to protect the for-profit exchanges from losing further market share.
\13\ However, their ``level the playing field'' argument has the
situation backwards, because there is a clear and massive economic
advantage to being an exchange.
---------------------------------------------------------------------------
\13\ See, ``U.S. Market Structure Overview: Briefing for House
Staff'', Nasdaq/NYSE, June 12, 2012.
---------------------------------------------------------------------------
Within the past 5 years, two major ATSs, BATS, and DirectEdge, both
voluntarily chose to become exchanges, \14\ spending millions of
dollars and devoting years of effort to make the switch. In describing
its history, the parent company of BATS Exchange explained that it
converted from an ATS to an exchange in order ``to participate in and
earn market data fees from the U.S. tape plans [and] reduce our
clearing costs . . . .'' \15\ While many ATSs have applied to the SEC
to convert to exchange status, and all were willing to accept exchange
responsibilities, \16\ we are not aware of a single exchange that has
tried to convert to ATS status. While exchange status does come with
some burdens, clearly market participants are happy to accept those
costs in return for the five significant advantages of being an
exchange.
---------------------------------------------------------------------------
\14\ See, In re the Applications of EDGX Exchange, Inc., and EDGA
Exchange, Inc. for Registration as National Securities Exchanges,
Exchange Act Release No. 61698 (Mar. 12, 2010); In the Application of
BATS Exchange, Inc. for Registration as a National Securities Exchange,
Exchange Act Release No. 58375 (Aug. 18, 2008) [hereinafter BATS
Exchange Registration Order].
\15\ See, Amendment No. 5 to Form S-1, BATS Global Markets, Inc.
(Mar. 21, 2012) at 2, available at http://sec.gov/Archives/edgar/data/
1519917/000119312512125661/d179347ds1a.htm [hereinafter BATS Form S-1].
\16\ See, e.g., ``SEC Special Study: Electronic Communication
Networks and After-Hours Trading'', at n. 27 (referring to the
applications of Island ECN, NextTrade, and Archipelago), available at
http://www.sec.gov/news/studies/ecnafter.htm.
---------------------------------------------------------------------------
The 5 big advantages exchanges have over ATSs:
1. Exchanges have absolute immunity on errors, having historically
been considered quasi-governmental entities. \17\ Courts have
typically ruled that exchange immunity holds even in cases of
gross negligence or willful misconduct. An ATS is a regular
business that has liability for its actions.
---------------------------------------------------------------------------
\17\ ``Nasdaq Exchange Immunity May Limit Losses From Facebook
Claims'', by Nina Mehta, Bloomberg News, June 13, 2012.
2. Exchanges receive ``tape revenue.'' The CTA (Consolidated Tape
Association) has a legal monopoly on providing a consolidated
stream of real-time data from our Nation's stock markets. The
CTA makes a profit of approximately $400 million per year,
which is then distributed to its participant exchanges based on
---------------------------------------------------------------------------
a complex formula. ATSs do not receive tape revenue.
3. Exchanges pay no clearing fees. An ATS is a party to both sides
of each transaction that passes through it, while an exchange
merely facilitates the transaction. Therefore ATSs pay
significant clearing fees, whereas exchanges pay no clearing
fees.
4. Exchanges have no net capital requirements. An ATS operator must
meet stringent net capital requirements. Exchanges face no such
requirement.
5. Exchanges can display bids and offers directly into the National
Market System. An ATS cannot display a bid or offer directly
into the National Market System. Instead, an ATS must pay an
exchange to display bids and offers on their behalf. \18\
---------------------------------------------------------------------------
\18\ FINRA hosts an ``Alternate Display Facility'' to allow ATSs
to display their bids and offers, but due to outdated technology, this
service is not operational as of December 2012.
---------------------------------------------------------------------------
5. How has the operating model of exchanges been influenced by their
change from not-for-profit organizations to for-profit
companies?
Since becoming for-profit companies, exchanges have a fiduciary
responsibility to their shareholders to maximize profits. A major
source of revenue and profit for the exchanges comes from the sale of
market data.
The exchanges, together with FINRA, have a Government-granted
monopoly over the sale of market data to the public--including the fees
from market data generated by off-exchange trading. The Consolidated
Tape Association (the ``CTA''), which administers the consolidated tape
on behalf of the exchanges and FINRA, charges high fees to the
investing public for real-time market data. While there is no
systematic transparency into the CTA's finances, some information can
be gleaned from the exchanges' parent companies' public financial
disclosures:
NASDAQ: The NASDAQ OMX Group, Inc. reported earning $115
million of net U.S. tape revenue from the CTA during 2011. \19\
This amount is separate from, and in addition to, the $135
million NASDAQ earned from the sale of proprietary U.S. market
data products.
---------------------------------------------------------------------------
\19\ NASDAQ OMX Group, Inc., 2011 Form 10-K at 57, available at
http://sec.gov/Archives/edgar/data/1120193/000119312512077518/
d259668d10k.htm.
NYSE: NYSE Euronext earned $193 million from market data
relating to U.S. equity trading in 2011, although it is not
entirely clear from their disclosure whether this includes
revenue unrelated to the U.S. consolidated tape plans. \20\
---------------------------------------------------------------------------
\20\ NYSE Euronext, 2011 Form 10-K at 60, available at http://
sec.gov/Archives/edgar/data/1368007/000119312512086538/d275617d10k.htm
[hereinafter NYSE 2011 Form 10-K].
BATS: BATS Global Markets, Inc. earned $55.4 million from
its share of revenue from the U.S. tape plans in 2011. \21\
---------------------------------------------------------------------------
\21\ See, BATS Form S-1, supra note 15 at 18, 39.
The estimated $400 million in market data revenues that the CTA
distributes to the exchanges are after operational and administrative
expenses have been paid. Given that real-time data is a Government-
granted monopoly, and market data prices are not set by the market and
are not subject to competition, the investing public is arguably being
overcharged for market data by approximately $400 million a year.
Historically, the SEC has justified granting exchanges the
exclusive right to sell market data as a form of user tax to fund the
exchanges' regulatory expenses. In 1999, the SEC stated that exchanges
are entitled to market data revenues to offset the cost of regulating
their markets. \22\ However, the amounts earned by the exchanges today
far exceed their regulatory expenses and act as a major profit center
for exchanges.
---------------------------------------------------------------------------
\22\ See, ``Concept Release: Regulation of Market Information Fees
and Revenue, Exchange Act Release No. 42208'' (Dec. 9, 1999)
[hereinafter SEC 1999 Market Data Concept Release].
---------------------------------------------------------------------------
In Congressional testimony, the CEO of NYSE Euronext estimated that
it would ``spend nearly $85 million for U.S. equity market surveillance
in 2012.'' \23\ This is far outpaced by the market data revenue NYSE
Euronext earns, which appears to have totaled $193 million in 2011 from
market data relating to U.S. equities trading. \24\ NASDAQ OMX Group
earned $115 million of net U.S. tape revenue during 2011, but spent
only $35 million on regulatory expenses across the entire holding
company--apparently even including regulatory expenses relating to
their non-U.S. exchanges. BATS Global Markets, Inc. earned $55.4
million from its share of the U.S. tape plans in 2011. But it spent
only one tenth of this amount on regulatory expenses of $5.5 million in
2011, including costs under its outsourcing agreements for regulatory
services to be provided by other SROs. \25\
---------------------------------------------------------------------------
\23\ 2012 Market Structure Hearings, supra note Error! Bookmark
not defined. (statement of Duncan Niederauer, CEO, NYSE Euronext),
available at http://financialservices.house.gov/uploadedfiles/hhrg-112-
ba16-wstate-dniederauer-20120620.pdf.
\24\ NYSE Euronext, 2011 Form 10-K at 60, available at http://
sec.gov/Archives/edgar/data/1368007/000119312512086538/d275617d10k.htm
[hereinafter NYSE 2011 Form 10-K].
\25\ See, BATS Form S-1, supra note 15 at 73-74.
---------------------------------------------------------------------------
The enormous revenues from market data are way out of proportion
with the costs of exchanges' self-regulatory responsibilities. Market
data revenue has simply become a Government-granted windfall at the
expense of the investing public.
Furthermore, the current tape revenue system potentially encourages
odd distortions in the markets. Because the CTA allocates revenue to
the exchanges based on a complex formula involving variables such as
each exchange's number of quotations, for-profit exchanges try to set
policies and services that will increase the level of quoting activity.
We believe this is a major factor in why quote flickering has markedly
increased, with the number of daily changes in the NBBO (National Best
Bid Offer) per million shares traded recently climbing to an all-time
high. \26\ It is logical to assume that if the tape revenue system were
reformed, quote flickering would be seen as a wasteful expense rather
than a lucrative source of revenue.
---------------------------------------------------------------------------
\26\ Ana Avramovic, ``Who Let the Bots Out?'' Credit Suisse
Trading Strategy, May 2012.
---------------------------------------------------------------------------
6. What regulatory or legislative changes should be considered by
regulators or Congress?
Credit Suisse suggests three policy changes, each of which is
designed to make markets more reliable than they are today or reduce
investor cost:
1. Remove the SRO status of the for-profit exchanges. For-profit
entities should not be shielded from liability for damages that
arise as a result of their own actions. For-profit entities
should not be able to audit and regulate their competitors.
Exchanges have already transferred most of their regulatory
tasks to FINRA. It is time for Congress to revoke their special
quasi-governmental status and Government privileges.
2. Perform a review of the pricing and rebate system operated by the
consolidated tape plans. The CTA plans collect over $400
million a year from the investing public, most of which then
gets rebated to the for-profit exchanges that collectively run
the plans. Forty years after these plans were established, \27\
we believe the tape revenue model is obsolete. In the current
system, the investing public overpays for market data, and the
exchanges receive a Government-granted windfall.
---------------------------------------------------------------------------
\27\ The SEC adopted Rule 17a-15 in November 1972, establishing
the current tape plans.
3. Lift the restrictions that limit broker-dealers to 20 percent
ownership in exchanges. Although there is no rule or law
limiting broker-dealer ownership in exchanges, there is a
precedent set by the regulators to cap broker-dealer ownership
of an exchange at 20 percent. Allowing broker-owned ATSs to
follow in the footsteps of BATS and DirectEdge and become
exchanges would level the playing field between exchanges and
---------------------------------------------------------------------------
ATSs, ultimately resulting in lower costs for investors.
Thank you for the opportunity to appear today and I will be happy
to answer any questions that you may have.
Witness Background Statement
Dan Mathisson is the Head of U.S. Equity Trading for Credit Suisse.
He is responsible for block trading, program trading, and electronic
trading at Credit Suisse.
Mr. Mathisson joined Credit Suisse in 2000 as a trader in the Index
Arbitrage department, shortly after which he founded the Advanced
Execution Services (AES) group, which executes trades on behalf of
institutional clients using algorithmic techniques. Prior to joining
Credit Suisse, he was the head equity trader at D.E. Shaw Securities,
where he worked from 1992-2000.
Mr. Mathisson writes a column about trading and markets for Traders
Magazine. In 2011 he was named one of the ``Top Ten Innovators of the
Decade'' by Advanced Trading magazine, which cited him for creating the
modern algorithmic trading desk. Mr. Mathisson received a B.A. in
Economics from the University of Michigan, and he is a Chartered
Financial Analyst.
______
PREPARED STATEMENT OF ERIC NOLL
Executive Vice President and Head, Nasdaq OMX Transaction Services
December 18, 2012
Thank you Chairman Reed and Ranking Member Crapo for the
opportunity to testify today on computer trading in U.S. securities
markets.
Computer trading is a fact of life and has been the default method
of trading for billions of trades over the past several years--Billions
of trades that happen on our market and others without any concern or
problem. While there are issues to review, computer trading has a
proven track record of delivering benefits for investors and market
participants that includes bringing new investors to the markets,
equalizing the information advantage that used to be the staple of
manual markets, lowering trading costs and giving the market expanded
abilities to handle trade and message traffic growth that would freeze
manual markets. As we saw during the financial crisis of 2008, the U.S.
equities markets did not freeze-up and billions of trades that
investors needed were handled by our computers. And of course, this was
done without any contribution from the equity asset class to the list
of problems that had to be managed. So, while we have experienced some
trading anomalies like the Flash Crash of May 6, 2010, and a number of
computer-trading events. I believe these are isolated technology
incidents and not symptoms of deeper market structure concerns. At
NASDAQ OMX we are laser-focused, every day, on how do we improve our
market and make it more resilient and robust?
This question is critical because a well-functioning equity market
is needed for efficient capital formation, innovative competition, and
job creation. Companies like Microsoft, Cisco, and Intel, used capital
raised from listing on NASDAQ to make cutting edge products that have
transformed our lives. Along the way, these companies created millions
of jobs and strengthened many communities. Innovative high-growth
companies attract new talent and that talent pool then demands new
goods and services. This virtuous cycle has played out in dozens of
venture zones, from Silicon Valley to the Northern Virginia high tech
corridor. And they have created enormous wealth opportunities, allowing
millions of average investors to share in that wealth--enabling them to
buy homes, put children through college, and retire with financial
security.
In light of recent events, some may forget the unique and central
role exchanges have played and continue to play in U.S. equities
markets. All that your average constituent associates with ``the
market'' starts with an exchange. The iconic public companies they
recognize--Apple, Google, eBay, and Amazon--must satisfy exchanges
listing standards, and they remain subject to exchange regulations
against corporate fraud and abuse. The exchange listing process and
regulatory program culminate in the IPO process that provides
entrepreneurs with efficient platforms for capital formation and job
creation. Only equities exchanges such as NASDAQ are entrusted with the
important responsibility to be a catalyst for growth and wealth
creation.
After the IPO, exchanges have a unique and continuing duty to
foster price discovery and transparency. Exchanges like NASDAQ create
and disseminate the ticker symbols and prices that your constituents
see on television stations like CNBC, in newspapers like the Wall
Street Journal, and at Internet portals like Yahoo Finance. Exchange
quotes then create the reference price for all other trading, not only
in equities but in other asset classes as well. Dark pools and other
competitors use exchange quotes as a reference price for trading
equities. Markets, such as the Chicago Mercantile Exchange and the
Chicago Board Options Exchange, use equities exchange quotes to trade
options, futures, and other derivatives. Vanguard, Fidelity, and Schwab
use exchange prices for mutual funds, ETFs, and other instruments.
Those ticker symbols are a byproduct of the rules and sophisticated
regulatory systems that equities exchanges develop and enforce to
protect investors and to provide orderly markets. They are the result
of a system that is by law fully transparent, and that publicly
discloses all rules and prices to all customers and treats all
customers equally.
Only exchanges have the authority and responsibility to oversee
broker-dealers as they interact with the market. That authority is the
result of a rigorous public process of qualifying to be an exchange
conducted by the SEC--in the case of NASDAQ it took 6 years. Exchanges
alone adopt member and market regulation rules, develop automated
surveillance systems to detect rule violations, and discipline broker-
dealers that violate rules and harm investors. Congress recognized that
enforcing rules in U.S. securities markets is so important that two
regulators rather than one are needed to enforce them. Congress
codified the authority of exchanges to act as self-regulatory
organizations (SROs), to set and enforce trading rules and to halt
trading during extraordinary national or international events. SROs
supply the SEC and other regulators vital information about the trends
and performance of U.S. capital markets. The SEC is our partner in
protecting investors.
In fact, exchanges have heavy responsibilities to create a safe
market for investors, characterized by fair access, transparency, and
efficiency. No other market participant is charged with or even
permitted to undertake this burden. Alternative trading systems (ATSs)
are not entrusted to regulate and discipline their users in this
manner. An ATS can choose to regulate its users, but it must then
register as an exchange and accept SRO responsibilities. Today,
virtually every ATS has the option to register as an exchange. One need
only look at the list of SRO responsibilities and obligations that
registration triggers to understand why so few ATSs voluntarily take
that step.
While we often discuss the importance of capital formation, our
regulatory responsibilities and the IPO process, let me add another
important, but sometimes overlooked branch of our role in the markets--
our role in the daily trading dynamics of the market. Trading and
trading behaviors like price discovery, best bid and best offer and
visible liquidity are very important to companies as they might seek
secondary offering cash injections to their businesses and use their
stock as currency in the market to achieve strategic goals like
acquisitions. Price discovery and transparent liquidity are also very
important to investors as they make informed decisions about which
stocks to buy and at what price and when to sell. All the buying and
selling and active trading in the equities market is not a grand game
of speculators--it has real job creators and investors looking for the
market's best information to make rational business and investing
decisions. Exchanges maximize transparency, strive for fairness and
support that price discovery engine and it is our unique market role to
perform that function. We are not in business just to see trading for
trading's sake. NASDAQ OMX is an exchange to produce transparent
quoting and trading that helps price discovery, helps add liquidity,
tightens spreads and benefits the continuous market is what we strive
to support.
Cooperation
The role of exchanges is more important than ever in today's
challenging environment. U.S. markets are complex, fragmented, and
interconnected. Markets and traders leverage new technologies to trade
near the speed of light. We at NASDAQ are working tirelessly to ensure
that markets are strong and fair, and that as the pace of trading
accelerates, so too does the pace of regulation and investor
protection.
When computerized trading appears as a threat to investors, the SEC
naturally turns first to exchanges for assistance. Regardless of where
the problems began, regardless of where the damage was felt, the
exchanges are always on the front lines partnering with the SEC and we
work closely with the SEC to fix and improve the equities markets. In
the aftermath of May 6th, the SEC and the exchanges worked quickly and
cooperatively to devise new protections to keep computer trading errors
from spreading too rapidly or inflicting unacceptable harm on the
overall market. The exchanges reformed their rules for breaking trades,
instituted single stock circuit breakers, updated market-wide circuit
breakers, and we will implement the Limit Up/Limit Down mechanism in
February. NASDAQ has also developed tools to help broker-dealers manage
their obligations under the Market Access Rule.
In the wake of several highly publicized computer malfunctions, the
exchanges are again leading the industry in a collaborative working
group. A key and challenging initiative being discussed by this
Industry Working Group, and one that NASDAQ fully supports and is
helping to lead and define, is the implementation of ``Peak Net
Notional Exposure'' levels, or ``kill switches,'' that would
automatically trigger a cessation of trading when an individual firm
exceeds predetermined risk thresholds. The Industry Working Group is
considering various approaches to both the SRO-level and broker-dealer
level requirements, as well as a means for coordinating cross-market
checks to create the market-wide check needed to combat the effects of
market fragmentation and interconnectedness.
Testing
One important area of focus is testing and industry preparedness.
NASDAQ is partnering with Carnegie Mellon University to form the
Carnegie Mellon Software Engineering Institute dedicated to help bring
the industry together to improve the resilience of financial services
technology. We hope to form and lead a group of market participants,
regulators, technology providers, and academic institutions with the
goal of driving resilience in the large scale software engineering and
technology arena and being recognized globally as a leader in helping
the financial markets become more resilient and robust.
The industry has learned through experience that it must change the
way we test. In the past, industry-wide system changes have utilized a
testing methodology that tested for system design integrity. For
example we might test a software update by having our members send us
test orders to ensure the software does what we are asking it to do.
Or, we might ask members to challenge our systems with high volumes.
Instead, we should be testing each other's systems to try to break
them. A more robust testing environment would assume breakdowns by all
testing participants to visualize the impact on a system's integrity.
Such ``destructive'' testing will spot troubles that the kinder-gentler
testing of the past would not uncover.
High Frequency Traders (HFT) firms have attracted much media
attention, but they are not the only ``fast'' players in the
marketplace. Exchanges, dark pools and broker systems are all connected
and all use sophisticated technology. These systems communicate in
slices of time that approach the speed of light. This is a great
achievement, but it means that previously minor events now represent
profound risks that can tangibly affect investor confidence. NASDAQ OMX
is not immune to this issue, and we are committed to answering this
challenge.
High Frequency Trading
NASDAQ believes that technological developments must be implemented
in a manner that ensures all investors a ``fair deal.'' Average
investors must not be placed at a disadvantage to professional traders
by rules that permit selective disclosure of information, preferential
access to trading interest, or the appearance of a two-tiered market.
All markets that trade the same securities should be equally
transparent about their operations, including the rules governing their
trading systems, the criteria for admission and the prices of
comparable services. The Commission must regularly examine whether the
application of new technologies contribute to regulatory arbitrage.
For example, exchanges and regulators around the world are
analyzing the pros and cons and overall impact on markets of HFT. The
International Foresight Project was commissioned by the British
Government's Department for Business, Innovation, and Skills (BIS) to
investigate the effects of HFT. This definitive and independent study,
led by Sir John Beddington, Chief Science Officer for the British
Government, found that HFT is likely positive for markets. Similarly,
the Swedish Financial Services Agency released its own report finding
that HFT in that country also had a positive impact on liquidity, and
that regulators and exchanges continue to refine their tools for
ensuring proper surveillance.
Many in the public arena vilify HFT as a business model issue. It
is our view to always caution against such sweeping criticism. When,
like the Beddington study, HFT is studied in depth, you find benefits
to several metrics from the broad participation in our markets by firms
that we consider to be high frequency traders. Like the British and
other studies, we find that HFT trading tightens spreads and adds very
valuable liquidity--certainly positive for our markets. We know that
everyone in the markets has a profit motive and that generally
incentivizes innovation and competition among participants. What we
know from experience is that our industry, no matter the business
model, will always attract individual players who cross the line and
NASDAQ OMX, the other exchanges, FINRA (the Financial Industry
Regulatory Authority) and the SEC work to expose those individual bad
actors. It seems the tenor of the debate about HFT has become too
broadly negative towards the business model. The academic evidence
about HFT supports the fact that they generally add value to the
market.
It is not enough simply to vilify fast trading. Regulators and
exchanges are working to identify and address specific bad actors and
specific bad outcomes such as false, misleading or deceptive practices.
NASDAQ has worked diligently to ensure that the pace of its regulation
matches the pace of trading. NASDAQ has partnered with FINRA to develop
special HFT inspections. For example, in December of 2010, NASDAQ OMX
retained outside experts to assist in assessing and improving our
internal training program on HFT strategies. Through focus and effort,
NASDAQ's Market Watch staff has developed 11 new alerts (algorithms
specifically designed to spot certain trading behaviors) in addition to
the 21 surveillances FINRA utilized for HFT related reviews.
To improve our own regulatory program and the regulatory programs
of exchanges around the world, NASDAQ invested in state-of-the-art
technology. In 2010, NASDAQ acquired The SMARTS Group, the world's
leading provider of software for automated surveillance for exchanges,
regulators, and brokers. With SMARTS, NASDAQ literally can deploy high
speed surveillance to match high speed and any other kind of trading.
We have held demonstrations for many Members and staff of this
Committee to demonstrate the power of the SMARTS system. The feedback
from those demonstrations has been positive.
These efforts have paid off. NASDAQ surveillance and referrals to
FINRA and the SEC have improved compliance. While we cannot go into
great detail, we have a full plate of pending Investigations on issues
related to High Order/Low Execution Ratios, Wash Trading, Layering,
Manipulation of the closing auction, Manipulation through master-sub
relationship, Supervision, Order Entry controls. NASDAQ has detected
violations by high frequency traders resulting in fines as high as $3.5
million and in the expulsion of firms and individuals from the
securities industry. NASDAQ is protecting investors from people that
use technology to prey on them. Our goal is always to constantly
evaluate and improve our market to make it as robust and fair as
possible using technology and the wisdom and experience of our
industry-best employees.
Complexity
Any evaluation of the health of our markets and the ecosystem of
computer trading must include a discussion of complexity. There are 13
registered exchanges active in the U.S. equities markets. The SEC also
allows trading on 40+ venues in the U.S. where a broker can send one or
more of their orders. Each of these venues has its own systems and
procedures and each competes for orders from brokers and ultimately
investors. Each venue has its own order types and each is continually
talking to investors to develop new order types that satisfy their
needs. The result is dozens or even hundreds of different order types
for members to understand and program. Is the explosion of order types
helpful or harmful for the market?
While some order types have come under intense media and regulatory
scrutiny, let me be clear, NASDAQ OMX order types do not provide
advantages to certain users allowing them to jump ahead in line at a
given price level. NASDAQ believes that each order type it creates
should be designed to make our markets better, and to improve
transparency and price discovery. Fairness and equal access are key SRO
responsibilities and we will always adhere to those principles. NASDAQ
goes through a rigorous process to get order types approved by the SEC.
As an exchange we have to expose innovative ideas to the market through
the notice and comment process, often allowing our competitors time to
mimic our idea and beat us to market. That is part of our SRO burden.
For the sake of transparency and to help members understand our order
types, we recently posted on our Web site a list and a plain-language
description of NASDAQ's order types.
Computer trading and some of the concerns that have been outlined
to Congress are in many respects the direct result of market structure
decisions. Many problems with our markets stem from well-intentioned
regulations like Regulation ATS and Regulation NMS, which sought to
promote competition and to resolve tensions between electronic and
floor-based trading. Regulation NMS has led to an increase in dark
trading, which denies market participants a clear view of trading
interest in a given stock. Dark trading is a concern in many countries;
Canada recently modified its market structure to limit dark trading and
to maximize price discovery. The Commission has similar market
structure proposals pending since 2009.
Market-Based Approaches
In addition to regulatory enhancements, NASDAQ has also developed
several market-based approaches to improve the trading experience, and
help reestablish the prominence of the public company model. For
example, NASDAQ launched the first ``price/size'' market to create
incentives for quotes that offer deep liquidity rather for quotes that
are fast. Also, NASDAQ voluntarily eliminated flash orders from its
equity markets. NASDAQ also introduced the ``MinLife'' order to
incentivize a longer quote life. Finally, NASDAQ is the only exchange
to recently institute a charge for excessive messages to discourage a
trading technique used primarily by high frequency traders.
NASDAQ OMX is also working to improve the market structure for
small public companies that are job creation dynamos when given a
supportive ecosystem. This past year, the JOBs Act recognized the
importance of special rules for these emerging companies. However,
Congress did not go far enough and consider how these companies were
treated once they actually go public. Regulation NMS subjects these
smaller stocks to a one-size-fits-all market structure. Apple,
Microsoft, GE, and other large cap stocks trade relatively well,
despite a highly fragmented marketplace. Small companies however are
not best suited for a fragmented liquidity pool and dark trading.
Smaller stocks do not perform well in the fragmented marketplace no
matter their listing venue. This can compromise the momentum for
smaller public companies and capital formation within this class of
stocks. There are innovative ideas to empower small companies to help
their stocks trade more often and more efficiently:
Tick Size Pilot Program: Allows smaller companies to opt-in
for a wider tick size for their stock to allow more spread for
market-makers to be incentivized. Multiple tick size regimes
are already used already in numerous other countries
successfully.
Market-Maker Support Pilot Programs: Allows the company to
opt-in to a program to provide economic support for more
aggressive quoting and trading in their stocks. These programs,
common around the world, allow the exchange to stand between
the broker and the listed company to improve the trading of a
stock.
Conclusion
NASDAQ OMX is passionate about the critical role we play in capital
formation, investor protection and job creation. While it presents
challenges to everyone, ultimately we believe that technology is an
important part of the solution for ensuring orderly and fair markets.
We view efforts to slow-down our markets as counterproductive. Building
robust and dependable markets requires legislators, regulators and
market participants to continue to come together to drive positive
evolution. NASDAQ OMX is committed to working with Congress, the SEC,
our fellow exchanges and all market participants to make the U.S.
equity market the best in the world. NASDAQ OMX appreciates the
opportunity to testify. I look forward to your questions.
______
PREPARED STATEMENT OF ROBERT C. GASSER
Chief Executive Officer and President, ITG
December 18, 2012
Introduction
Chairman Reed, Ranking Member Crapo, and other Members of the
Subcommittee, thank you for the opportunity to testify this morning on
the topic of ``rules of the road'' for computerized trading venues. On
behalf of a leading agency broker, my goal is to offer an unbiased,
fact-based view on the current state of U.S. equity market structure.
ITG is a NYSE-listed company with 17 offices across 10 countries and
nearly 1,100 employees. As an agency broker, ITG provides trading
services, technology, analytics and research to a wide array of leading
asset managers. Throughout our 25-year history, we have worked in
partnership with major mutual funds, pension funds and other
institutional investors, innovating to improve trading and investment
performance. In my testimony today I would like to offer a brief
overview of current market structure, discuss some recent events which
have impacted investor confidence and look at some ways to restore this
confidence. There has been much written of late about the quality of
our equity markets. This morning we hope we can infuse some data and
analysis into the debate.
Market Structure
Competition amongst market centers and broker dealers spawned by
the passage of Regulation ATS in December 1998 has led to intense
competition for liquidity and ultimately to fragmentation. This
fragmentation has undoubtedly introduced complexity into our
marketplace but has been a positive force in reducing execution costs.
Technology has provided market participants, including retail investors
and mutual funds, competition for order flow.
Global asset managers, as fiduciaries, have an obligation to
achieve best execution. The global market standard requires all asset
managers of size to measure the quality of their execution and its
effect on the investment process. ITG is the world's largest provider
of TCA, or Transaction Cost Analysis. We measure millions of trades
executed on behalf of hundreds of global asset managers. Our TCA data
clearly demonstrates that institutional investors have benefited
greatly from the evolution of U.S. market structure. Over the past 12
years, there has been a 70 percent decrease in average total equity
trading costs in the U.S. As the data indicates, U.S. market structure
is not broken. The current ecosystem of displayed and dark markets has
resulted in significantly reduced costs that in almost all cases have
been distributed back to investors. There is no evidence to suggest
that competition and fragmentation have damaged price discovery or
harmed capital formation.
ITG is not a market maker, and we do not take on proprietary
positions. In other words, we do not have ``skin in the game'' when it
comes to the debates around broker internalization, as our system
provides ``meaningful price improvement'' to buyside investors as
described in Regulation NMS. Based on our data, we would conclude that
Broker-Dealer internalizers, or broker-dealer dark pools as they are
sometimes known, provide a useful permeable layer between the client
and the displayed markets. Brokers have a fiduciary responsibility to
their clients while exchanges do not, and these liquidity pools would
not exist unless benefit was derived by the customer. Most recently,
Australia and Canada have imposed regulations around internalization
that will provide similar action here in the U.S. Early returns do not
look promising in terms of the effects on liquidity and trading costs.
Regressing to an oligopoly of U.S. exchanges is clearly not the answer.
Investor Confidence
Unfortunately, the evidence also suggests that the investing public
has become disenchanted with equities. According to the Investment
Company Institute, over half a trillion dollars has been pulled from
U.S. equity mutual funds since the start of 2008. Much of this can be
attributed to the reduced risk appetite of baby boomers and the
relative safety of bonds supported by easy monetary policy.
The May 2010 Flash Crash, the Facebook IPO, and Knight Capital's
trading debacle this past summer provide little comfort that U.S.
equity markets are a safe place to trade or invest. Add in the
suspicions that the investing public has about high frequency trading
and its perceived impact on the quality of markets, and you have a
recipe for anecdote and conjecture overcoming facts and reason.
Where speed is concerned, it is clear that the law of diminishing
returns must be applied to further dramatic shifts in the foundations
of our equity marketplace. Microseconds versus milliseconds do not
matter to the wider, more important, audience. We need to restore
investor confidence, but not at the cost of disturbing the progress
that has been made.
Recommendations
The SEC's Consolidated Audit Trail, if implemented properly
and cost effectively, will give investors confidence that
regulators can police bad actors and predatory strategies.
The consistent application of the Market Wide Circuit
Breakers and the Limit-Up Limit-Down Plan to all market centers
would likely prevent a market disruption of ``Flash Crash''
proportions.
Costs should be borne by market participants who create
excessive quote traffic without executing order flow.
Market data should be distributed to all market
participants equally.
Marketwide risk should be monitored at a central clearing
house that would have the ability to terminate a broker-
dealer's connectivity to the national market system in the
event of a rogue program released to the market.
Conclusion
These five measures would give the investing public the protections
they need to confidently invest in the world's strongest and most
resilient market while still deriving all of the cost savings and
liquidity benefits which have been achieved over the past decade.
Lastly, as the regulations called for by the Dodd-Frank Act begin to
take hold across other asset classes, the lessons we have learned in
equities will be applied to those markets.
Price discovery, central clearing, transaction cost analysis, and
pre- and post-trade transparency will become as deeply integrated into
foreign exchange and fixed income markets as they are in equity
markets. And innovation will come more quickly to those markets because
of the lessons learned in equities. For this reason, our equity market
structure is all the more important to our broader financial system.
Thank you again for the opportunity to share our views on these
important questions. I would be happy to answer any questions at the
appropriate time.
RESPONSES TO WRITTEN QUESTIONS OF CHAIRMAN REED
FROM JOSEPH MECANE
Q.1. There was testimony that ``dark trading'' is a concern in
many countries and that Canada recently modified its market
structure to limit dark trading and to maximize price
discovery. Canadian regulators have imposed a new framework
governing how dark pools and undisplayed orders are allowed to
operate including priority of lit over dark flow and a minimum
price improvement requirement for dark orders. What is your
view of these reform proposals? Would these measures make the
U.S. markets more or less fair and transparent? Please explain.
Would these measures be feasible in the U.S. markets? Why or
why not?
A.1. To be clear, we are not adverse to the concepts of
internalization or dark pools. We believe they are a valuable
part of the market in terms of facilitating the execution of
institutional block orders and providing significant amounts of
liquidity to retail orders. However, there is a balance that
must be struck between internalization and public price
discovery in order to maintain a healthy public market. We
believe the new Canadian rules--which are designed to encourage
transparency, support the price discovery process, reward
displayed orders with increased execution opportunities, and
increase liquidity for all--can be incorporated into the U.S.
markets to help achieve this difficult balance. Additionally,
the Canadian rules recognize the value that dark trading can
offer to minimize the market impact of trading blocks and
provide meaningful price improvement--again, useful paradigms
for preserving the value of internalization while creating
mechanisms to facilitate publicly displayed liquidity.
A key objective of the Canadian rules was to encourage the
posting of visible orders and expose liquidity to the widest
variety of participants by ensuring that visible orders execute
before dark orders at the same price across the entire market.
\1\ In stark contrast to Canada, in the U.S., market
participants who take an added risk by displaying their orders
are being traded ahead of, normally at the same price, on dark
markets. This has the effect of reducing the execution
experience on ``lit'' markets and further encourages liquidity
providers to move into the dark markets. We believe this effect
is borne out by the statistics we noted in our testimony that
show that U.S. market structure has led to widespread and rapid
growth in dark trading. \2\
---------------------------------------------------------------------------
\1\ After significant analysis and industry commentary, the
Investment Industry Regulatory Organization of Canada (IIROC) and the
Canadian Securities Administrators (CSA) took the position that
limiting the use of off-exchange trading was critical to maintaining
the quality of the price discovery process and adopted new rules which
went into effect on October 15, 2012. http://www.bcsc.bc.ca/
uploadedFiles/securitieslaw/policy2/23-
405%20Dark%20Liquidity%20in%20the%20Canadian%20Market.pdf. Rosenblatt
Securities indicated the portion of Canadian equity trading done by
dark pools was 2.06 percent in November 2012, down from 5.67 percent in
September 2012. Rosenblatt Securities, ``Let There Be Light'', December
19, 2012.
\2\ Off-exchange dark trading grew from 19.5 percent in October
2007 to a record level of 36.6 percent as of January 2013, including
38.5 percent in Nasdaq-listed securities. Off-exchange trading has
increased 300 basis points in only two months since October 2012.
Moreover, as of January 2013, approximately 3800 securities (49 percent
of total securities) have over 40 percent of volume traded off-exchange
(CTA).
---------------------------------------------------------------------------
We are concerned about the aggregate effect of dark or
internalized activity on the overall U.S. marketplace, mainly
the effect on spreads and volatility, as noted in recent
research by the CFA Institute and Professor Daniel Weavers of
Rutgers Business School \3\--especially in the thousands of
less actively traded securities, as well as the selection bias
in orders that do make it to public markets. Moreover, the
Joint CFTC-SEC Advisory Committee urged the SEC to review the
issue of internalization/dark trading in 2010. \4\ Changes such
as these are also in line with a key objective of Regulation
NMS (Reg. NMS), namely displayed order protection. \5\
---------------------------------------------------------------------------
\3\ http://www.cfapubs.org/doi/abs/10.2469/ccb.v2012.n5.1; http://
papers.ssrn.com/sol3/papers.cfm?abstract_id=1846470
\4\ In a summary report on Recommendations Regarding Regulatory
Responses the Market Events of May 6, 2010, the Joint CFTC-SEC Advisory
Committee noted that ``the impact of the substantial growth of
internalizing and preferencing activity on the incentives to submit
priced order flow to public exchange limit order books deserves further
examination'' and recommended the SEC review whether to ``adopt its
rule proposal requiring that internalized or preferenced orders only be
executed at a price materially superior (e.g., 50 mils for most
securities) to the quoted best bid or offer. http://www.sec.gov/
spotlight/sec-cftcjointcommittee/021811-report.pdf
\5\ In the rule the Commission stated: ``The Commission agrees
that strengthened protection of displayed limit orders would help
reward market participants for displaying their trading interest and
thereby promote fairer and more vigorous competition among orders
seeking to supply liquidity.'' http://www.sec.gov/rules/final/34-
51808.pdf
Q.2. What is an example of an order type that the NYSE has
created to make our markets better, and improve transparency
and price discovery? How does this order type provide equal
access? Does the NYSE allow dark or undisplayed orders or
---------------------------------------------------------------------------
provide dark execution? Why or why not?
A.2. The number and characteristics of order types offered by
exchanges has been of recent discussion in the press and within
the industry. It is important to note the overarching market
structure context within which new order types are created, not
the least of which is the competitive and complex dynamic among
Exchanges and non-Exchange participants that is fostered by
inequitable regulation. Among other things, we develop order
types to allow clients to control how their orders interact
with others in a complicated market environment. Many order
types are developed to help participants comply with
requirements of Reg. NMS, while others enable a participant to
control their execution costs. More recently and significantly,
Exchanges have developed order types to attempt to compete with
practices that are allowed by non-Exchange venues, some of
which are undisplayed. One of these order types, and perhaps
the newest and most innovative in our suite of order types, is
the RLP order.
The Retail Liquidity Program (RLP), which was approved in
July 2012, gives retail investors the ability to receive price
improvement at a sub-penny increment of at least $0.001 in an
exchange environment. In comparison, off-exchange venues have
been permitted to segment customer order flow, and trade in
sub-penny increments, without limits since the inception of
Reg. ATS. Approval of this relatively simple and beneficial
order type took several months to develop and discuss with the
SEC before the 240 day public process that took place once we
publicly filed the proposal with the Commission. Prior to the
existence of RLP, exchanges were not permitted to target
specific customer segments, even where there could be
significant benefits to the retail investing public. While
arguments can be made about whether sub-penny executions and
segmentation are the ``right'' market structure, in general we
firmly believe that exchanges should be able to compete on
equal footing with other venues when the investing public can
benefit from price improvement in a competitive, liquid price
discovery process.
The development, testing, and approval process for a new
exchange order type is extensive and lengthy--sometimes taking
a year or more to receive SEC approval. Exchange orders types
are required to be available to members on a fair and
reasonable basis and are detailed in SEC rule filings which are
published for public comment and subject to the scrutiny of our
direct competitors. \6\ By contrast, ATSs and brokers offering
``dark'' or internalized trade execution services do not
publicly disclose details about how their trading functionality
works, including order ranking and execution rules, and the
algorithms and order types offered to clients are not subject
to prior public comment, SEC approval, or even SEC review.
These participants use their regulatory advantage as a
competitive edge to develop order functionality and this often
drives client demands for Exchange order types.
---------------------------------------------------------------------------
\6\ NYSE Euronext operates three equity exchanges in the U.S.,
NYSE, NYSE MKT (MKT), and NYSE Arca (Arca), each of which publicly
discloses all order types in SEC-approved rule filings and on our
publicly available Web site. http://usequities.nyx.com/markets/nyse-
equities/order-types; http://usequities.nyx.com/markets/nyse-arca-
equities/order-types
---------------------------------------------------------------------------
Overall, to the extent an initiative develops to
``streamline'' the number of order types, the most effective
way to accomplish this would be through a simplification across
all market venues of the underlying market structure that gave
rise to the proliferation of order types.
Q.3. Rule 612 of Regulation NMS prevents sub-penny quoting.
However, under an exemption in Rule 612 the SEC actually allows
broker-dealers to execute orders in sub-penny increments.
Because there are no quoting obligations for broker-dealer
internalization, broker-dealers can provide price improvement
to their customers in the form of sub-penny executions. What
are advantages or disadvantages of sub-penny quoting? Is
special pricing good for retail investors? Why or why not?
A.3. Sub-penny quoting and trading raise two separate but
interrelated issues. First is the ability for certain venues to
trade in sub-penny increments. The second is whether quoting
should similarly be allowed down to the sub-penny level.
Our primary viewpoint is that the rules applying to sub-
penny trading and quoting should be consistent across venues,
regardless of the market participant utilizing it. From a
public policy perspective, if sub-penny trading is allowed in
non-Exchange venues, we believe similar conventions should be
allowed in Exchange venues also. On a related note, as we've
highlighted elsewhere, we worry about the rising level of
activity that trades in front of visible, displayed liquidity
and would suggest that a minimum amount of price improvement be
required for this convention. Until the RLP program, Exchanges
were only permitted to offer midpoint executions to its
members. However, in August 2012, under the RLP program, for
the first time an Exchange was allowed on a pilot basis to
permit the execution of retail orders in sub-penny increments
at a minimum of $.001 to mirror some of what occurs in
internalization. \7\ However, the RLP program remains far more
restrictive than broker internalization, and is currently
limited to a small segment of the marketplace. Despite the
positive impact it may have on the retail investor, the RLP
program only accounted for 0.14 percent of consolidated volume
in December 2012.
---------------------------------------------------------------------------
\7\ SEC Release No. 34-67347: http://www.sec.gov/rules/sro/nyse/
2012/34-67347.pdf.
---------------------------------------------------------------------------
The related question is whether sub-penny quoting should be
allowed alongside sub-penny trading. This is also the topic of
an upcoming SEC Roundtable, and should also be considered in
the context of our previous point. However our basic view is
that to the extent sub-penny trading continues to be allowed,
then a more fair market structure would be to similarly provide
for sub-penny quoting. While we agree with many of the negative
implications of sub-penny quoting--such as investor confusion,
a noneconomic denomination, quote traffic implications--we
believe the current structure puts the public markets are at a
significant competitive disadvantage that should be addressed.
The chart below shows sub-penny trading (excluding mid-
point executions) in securities priced above $1 rose from 2.7
percent in January 2006 to 10.5 percent of volume in January
2013. In securities priced between $1 to $5, sub-penny trading
(excluding mid-point executions) rose from 3.3 percent in
January 2006 to 14.0 percent in January 2013. Nearly all sub-
penny trading (excluding mid-point executions) occurs off-
exchange.
Q.4. Does the larger percentage spread in low-price stocks lead
to greater internalization by OTC market makers or more trading
volume in dark pools? If so, why? Should the Commission
consider reducing the minimum pricing increment in Rule 612 for
lower priced stocks?
A.4. We believe the current penny quoting requirement has
created artificially wide spreads in many liquid low-priced
stocks where there is significant internalization. \8\ There
are two primary ways to potentially alter this dynamic--one by
decreasing the minimum tick increment, and the other by
incentivizing displayed public liquidity over that traded
privately, as we have highlighted in our previous answers.
Reducing the minimum pricing increment in liquid low-priced
securities would lower investor trading costs and improve
market transparency. In contrast to the ``one size fits all''
approach of the U.S., many countries have adopted tiered tick
sizes based on the price level of a stock, with tick increments
less than a penny.
---------------------------------------------------------------------------
\8\ http://www.sec.gov/spotlight/regnms/
jointnmsexemptionrequest043010.pdf
---------------------------------------------------------------------------
The decision to internalize a trade is influenced by the
trade-off of profit opportunity versus ``risk'' assumed. In
general, the wider the spread in a lower-priced stock, the more
profit opportunity and incentive to internalize. This incentive
is further distorted to the extent there are differing
increment rules on different types of venues, and also to the
extent venues are allowed to trade in front of displayed
liquidity in sub-penny increments.
The other contributing factor to greater internalization in
low-price stocks is the uniform access-fee cap of $.0030 per
share that applies across all stocks as defined in Regulation
NMS. Some thought should be given to reducing the access-fee
cap in these stocks commensurate with either price or spread.
Q.5. Some commentators have suggested that the Securities and
Exchange Commission (SEC) should lift the ban on locked
markets. Locked markets occur when a trader attempts to place a
bid on one exchange at the same price as an offer on a
different exchange. They argue that bids and offers at the same
price but different venues should be forced to interact, and
this could reduce fragmentation in the marketplace and perhaps
reduce the prevalence of trading in dark pools and
internalization venues. What is your view of this proposal?
A.5. Prior to Reg. NMS market centers were allowed to display
quotes that locked or crossed other markets for NASDAQ-listed
securities. Commenters at that time noted that locked/crossed
markets could cause investor confusion and detract from market
efficiency and were prohibited within Regulation NMS. Recently,
some market participants have suggested that locked markets
should potentially be allowed as a way to reduce fragmentation
and internalization.
We agree that the current market structure has incentivized
too much fragmentation and executions purely based on the
public market quotes. There are several ways to potentially
address this issue, some of which have been addressed in our
previous responses. Locked markets are another way to
potentially reduce the fragmentation incentive by decreasing
the potential profit opportunity from internalizing, and/or
reducing the need for a similar number of protected quotes.
As with any proposed changes, it is important to evaluate
both the positives and negatives of any changes. On the
positive side, locked markets would make transparent more
trading interest--nondisplayed buy and sell orders on market
centers that are executable but not currently eligible for
display. Locked markets would also reduce trading costs by
eliminating the spread and would potentially simplify some
order types that were developed to comply with Reg. NMS rules.
However, on the negative side, we specifically worry that
locked markets would become common in many active securities
that continually quote at a penny spread today, potentially
creating investor confusion.
The benefits and concerns resulting from locked markets
need to be carefully considered as part of a holistic review of
U.S. equity market structure. Other potential alternatives to
allowing locked markets exist and should be considered. These
include depth-of-book protection as well as other suggestions
highlighted in other responses such as incentivizing displayed
liquidity and/or introducing consistent trade and quote
increments in the industry.
Q.6. Internalizing broker-dealers often pay retail brokers to
direct customer orders to their trading venues. Does this
practice advantage or disadvantage retail investors? Why or why
not? Why is retail flow valuable to trade against?
A.6. Trading with retail flow is regarded as advantageous for
two reasons--on average, the ``informational'' content tends to
be lower than other types of order flow encountered in the
market, and secondly, retail flow tends to be smaller and have
less liquidity impact in a given stock. When a small order is
sent by an institutional client, it is often a part of a larger
order, and the counterpart interacting with the first order
does not know that there will be significantly more stock
behind it.
Payment for order flow is an issue that has been widely
debated in the industry and also warrants further ongoing
review due to the inherent conflicts that exist when the
payment is not directly accruing to the end investor. It is
also important that payment for flow not be viewed in
isolation. Payment for order flow, price improvement, and
access fees are interrelated topics that should be dealt with
concurrently.
The traditional retail internalization model that exists
today is primarily done through the interaction of retail
market orders with a single wholesaler. Internalization is
beneficial to the individual retail counterparty to the trade
if the execution results in meaningful price improvement.
However, internalization deprives both retail and institutional
limit orders displayed on exchanges from interacting with those
orders. The markets and price discovery optimally function
through interactions with multiple, diverse market
participants. When there is a proper mix of market
participants, there is a greater incentive for market makers to
quote more aggressively than a single participant. This
incentive affords investors an opportunity to receive a
superior fill on an exchange, one of the primary drivers behind
the development of our RLP program. The Exchange environment
brings competition from multiple liquidity providers, which
offers retail investors greater price improvement potential,
while also continuing to protect displayed quotes, which is
fundamental to preserving market quality for all investors.
Q.7. Broker-dealers that operate their own dark pools can get a
fee advantage when they route customer orders through their own
dark pools rather than through a lit exchange. Does this create
a conflict of interest between investors, who want to buy or
sell stocks at the best prices, and brokers, who want to avoid
exchange fees? Why or why not?
A.7. There are two related topics that this question raises.
The first is one of best execution and whether the interests of
the customer are being put first. The second is a broader
question about the cumulative impact of these practices on
overall market quality.
There are 13 equity exchanges offering a wide variety of
fee choices for customers. Some exchanges offer high rebates to
encourage liquidity providing and charge corresponding fees to
remove liquidity. Under this model, investors who take
additional risk by posting liquidity and showing their
investment interest may be rewarded as price makers through a
rebate. Investors who do not want to set prices but rather only
act as takers of liquidity are charged a fee. This model was
created by exchanges as a way to incentivize market
participants to make public markets and assist in the price
discovery process. This mechanism has become increasingly
depended upon by exchanges as the more highly desired order
flow from retail and institutional investors is executed in
dark trading venues. Several exchanges also offer low take fees
or even rebates to remove liquidity--minimizing the incentive
for brokers to avoid exchanges simply because of fees.
As noted in the previous response, investors trading
through a single private broker venue may not be optimizing
their execution if the reason is to avoid exchange take fees or
to maximize business in the broker-dealer owned dark pool. In
an exchange environment, however, the order would be exposed to
competitive participants within a transparent environment with
a real chance for execution optimization.
As was expressed by Invesco's Chief Investment Officer
during the June 2012 House Financial Services Committee's
hearing, internalization whereby the broker-dealers garner
information advantages creates a conflict of interest that they
believe does not advantage the investor. \9\ Similar conflict
and disclosure issues were also recently raised by FINRA. \10\
---------------------------------------------------------------------------
\9\ http://financialservices.house.gov/uploadedfiles/hhrg-112-
ba16-wstate-kcronin-20120620.pdf
\10\ 2013 Regulatory and Examination Priorities Letter: http://
www.finra.org/web/groups/industry/@ip/@reg/@guide/documents/industry/
p197649.pdf.
---------------------------------------------------------------------------
More significantly, however, is the effect this aggregate
activity has on overall market quality. As outlined in recent
research by Tabb Group and the CFA Institute, \11\ the
incentive to segment markets and reduce transparency
jeopardizes the price discovery process and can adversely
impact costs for all investors. Contrary to the stated goals of
Reg. NMS, today's market structure incentives result in
increasingly higher levels of dark trading in broker owned
venues, resulting in increased conflicts between investors and
their executing brokers.
---------------------------------------------------------------------------
\11\ Tabb Group, ``A Spotlight in the Dark: An Inevitable
Debate'', p. 8, Exhibit 4, November 2012; http://www.cfapubs.org/doi/
abs/10.2469/ccb.v2012.n5.1.
Q.8. The U.S. exchanges are required to submit data to a
centralized network where it is aggregated before being
disseminated to the public. However, the Consolidated Tape
Association/Securities Information Processor or SIP now lags
behind direct proprietary market data feeds. As a result, high
frequency trading firms and other market participants now
usually colocate their computer servers at every exchange and
subscribe to proprietary data feeds offered by the exchanges in
order to capitalize on latency arbitrage opportunities. Why
should anyone receive market data flow from other than the
consolidated tape? Should market data be distributed to all
market participants equally? How do you think this can best be
---------------------------------------------------------------------------
accomplished?
A.8. The consolidated tape data feeds include the best-priced
quotations of all exchanges and all reported trades, as well as
the calculation of the NBBO, short sales restriction
indications, single stock circuit breaker indicators, and other
data. Proprietary data feeds of individual exchanges include
those exchanges' own best-priced quotations and trades as well
as other information not available through the consolidated
tape, such as depth-of-book prices, which are required for
trading larger or more sophisticated orders, particularly in a
Regulation NMS environment. Because of these differences,
market participants may choose to take market data from the
consolidated tape and/or directly from exchanges or other
vendors, depending on their individual needs. In any event, the
SEC has to approve exchanges' proprietary data products before
they are sold to the public, and one of the general
requirements for approval is that the data is indeed fairly and
equally accessible to those that want to subscribe to the data
feeds. Moreover, speed differences from a user perspective are
not unfair or unusual. Because the data that exchanges send to
the securities information processors (SIPs) is consolidated by
the SIPs and then redistributed, the information in the
proprietary data feeds of exchanges is permitted to reach
market participants faster than the information sent from the
SIP to the same market participants. Indeed, this was expressly
contemplated in Regulation NMS where the SEC noted:
Commenters were concerned about the statement in the
Proposing Release that the distribution standards would
prohibit a market from distributing its data
independently on a more timely basis than it makes
available the ``core data'' that is required to be
disseminated through a Network processor. Instinet, for
example, requested that the Commission clarify that the
proposal would not require a market center to
artificially slow the independent delivery of its data
in order to synchronize its delivery with the data
disseminated by the Network. Adopted Rule 603(a) will
not require a market center to synchronize the delivery
of its data to end-users with delivery of data by a
Network processor to end-users. Rather, independently
distributed data could not be made available on a more
timely basis than core data is made available to a
Network processor. Stated another way, adopted Rule
603(a) prohibits an SRO or broker-dealer from
transmitting data to a vendor or user any sooner than
it transmits the data to a Network processor. \12\
---------------------------------------------------------------------------
\12\ See, SEC Release No. 34-51808; File No. S7-10-04 at p. 271
(August 29, 2005).
Further, in the SEC's 2010 Concept Release on Equity Market
Structure, the Commission commented the average latency for the
consolidated data feeds was generally less than 10 milliseconds
at that time. This latency captures the difference in time
between receipt of data by the SIP from the exchanges and
distribution of the data by the SIP to the public. Since that
time, continued improvements have been made to the SIP's
processing and today the average quote latency is less than 1
millisecond for Tape A&B securities and less than 2
milliseconds for Tape C securities (such data is made publicly
available on a quarterly basis). [http://www.nyxdata.com/CTA]
We would note that the amount of data sent via proprietary
feeds is far greater than on the consolidated feed, so the cost
of telecommunications and the time to process this data when
received is higher as well. Given these differences, we believe
the provision of consolidated and proprietary data is largely
---------------------------------------------------------------------------
about choice for customers.
Q.9. The October closure of the markets for 2 days due to
Hurricane Sandy raised questions about the financial sector's
preparedness for the next natural or man-made disaster. What
challenges did the markets face in reopening after Hurricane
Sandy? What should have been done differently after Hurricane
Sandy to keep providing trading services to customers and
maintain market integrity? What can the stock exchanges do to
prepare the market for another disaster? What changes need to
be made by market participants and regulators?
A.9. The effects of Hurricane Sandy on the northeast were
devastating. Lives were lost and thousands of families were
displaced. The markets faced several challenges leading up to
and during Hurricane Sandy. Reports called for a severe storm
with massive flooding, widespread power outages, the shutdown
of major transportation methods and dangerous travel
conditions. In retrospect, all of these reports came true.
Further, many of the largest financial service companies,
exchanges, the securities clearinghouse (DTCC), and countless
related smaller firms are headquartered in areas that expected
a severe impact. The storm was forecast to affect such a wide
area that many firms had both their primary and backup
facilities within the affected region.
NYSE senior management was in contact with the SEC,
industry trade associations, member firms, and other exchanges
throughout the weekend of October 27th and October 28th to
consider the challenges posed by the forecasted hurricane. In
the end, this group collectively decided that the risk/reward
of opening the market while all industry participants with a
local footprint would operate in contingency seemed
inconsistent with providing a stable, liquid market.
Following the hurricane, the challenges faced by the
markets were primarily borne by industry personnel. While the
markets re-opened on October 31st, industry personnel faced
significant obstacles common to all residents of the
surrounding area. Power and communication outages hampered
people's ability to work remotely. Access and transportation to
certain areas like lower Manhattan were severely restricted.
Basic necessities in these areas like lodging and food
presented some challenges initially. There was significant
damage to communications around lower Manhattan and surrounding
areas. While there were many challenges to overcome, we note
that the markets re-opened without incident on October 31st.
There are several lessons to be learned from this event. At
a minimum, businesses learned the importance of a well-prepared
and tested business continuity plan. At NYSE Euronext, we are
actively considering changes to our current disaster recovery
model for NYSE and NYSE MKT. The exchanges can take steps to
both review BCPs and regularly test backup trading locations.
Regular testing should both include a test of connectivity with
exchange backup facilities and cause minimal disruption to the
financial markets or its' participants. The industry should
consider whether to make these tests mandatory for all
participants. Industry conversations are underway to assess
lessons learned and prepare for similar events.
Q.10. It has recently been reported that the SEC's Interim
Inspector General found that some staff in the SEC's Trading
and Markets Division did not encrypt computers containing
confidential data from the exchanges and clearing agencies they
were overseeing. What measures are being taken now or should be
taken by market participants and regulators to better secure
data in today's high tech markets?
NYSE Euronext is committed to providing our customers with
a secure network delivering the highest levels of reliability
in the industry. The reliability and availability of NYSE
Euronext's Secure Financial Transaction Infrastructure (SFTI)
is dependent upon many dynamics including the prevention of
security breaches and cyber attacks. The security problems
uncovered through the SEC's Inspector General Report were
unfortunate and deeply concerning to NYSE Euronext, however we
believe the important outcome is that the SEC appears to be
taking steps to prevent a similar situation from occurring
again the future. We also support a more targeted approach to
regulator accumulation of entity data to ensure that any
security issues have more limited impact.
As witnessed by several high profile cyber security
breaches during the past several years, the Federal Government
has developed a notification process that includes a number of
agencies including the SEC, CIA, FBI, and DHS. NYSE Euronext is
supportive of these coordinated efforts by Federal agencies and
will continue to be partner with them in securing our financial
markets.
As offered during the hearing, we would also like to take
the opportunity to clarify the record on some of the
representations and data provided by the Credit Suisse witness.
We believe the below clarifications will provide a more
accurate depiction of the facts.
Point 1--Levels of ``Off-Exchange'' Trading
Mr. Daniel Mathisson of Credit Suisse offered in his
testimony that the ``statement that volume has been moving to
off-exchange venues . . . is factually incorrect. Over the past
5 years, volume has not shifted to off-Exchange venues.'' \13\
We strongly disagree.
---------------------------------------------------------------------------
\13\ Written Testimony of Daniel Mathisson of behalf of Credit
Suisse, Before the Senate Banking Committee Subcommittee on Securities,
Insurance, and Investment ``Computerized Trading Venues: What Should
the Rules of the Road Be?'', December 18, 2012.
---------------------------------------------------------------------------
Credit Suisse's conclusions rely on data that incorporates
a misleading and narrowly focused definition of ``off-
exchange'' and ``on-exchange'' trading: literally, whether a
market was registered as an exchange or not during the relevant
time frame. In particular, Mr. Mathisson includes BATS and
Direct Edge, which were ``Electronic Communication Networks''
or ECNs in the ``off-Exchange'' category. These ECNs displayed
public quotes and participated in price discovery in a way that
was very similar to exchanges, and subsequently became
registered exchanges. \14\
---------------------------------------------------------------------------
\14\ BATS became a national securities exchange in October 2008
and Direct Edge spun off the EDGA/EDGX exchanges in July 2010.
---------------------------------------------------------------------------
Under a more accurate and accepted measurement methodology
for ``off-Exchange'' trading, such ``off-Exchange'' activity
would not include ECNs such as BATS and Direct Edge. When
measured correctly, ``off-Exchange'' trading, which would only
include dark pools and internalization desks, rose from 19.5
percent in October 2007 to 35 percent in December 2012 and 37
percent in January 2013, as shown in Figures 1 and 2. Similar
classifications were highlighted in recent reports by Tabb
Group and Rosenblatt Securities. \15\
---------------------------------------------------------------------------
\15\ Tabb Group, ``A Spotlight in the Dark: An Inevitable
Debate'', p. 8, Exhibit 4, November 2012; Rosenblatt Securities, U.S.
Securities Volume: Year-In-Review, January 4, 2013.
---------------------------------------------------------------------------
Mr. Mathisson also commented that dark pool activity has
increased, but ``other internalization'' \16\ has declined.
This statement also is misleading. Dark pool activity has
indeed tripled from 4 percent to 13 percent of total volume
over the cited period. However, what Mr. Mathisson refers to as
``other internalization'' has risen from 12 percent to 21
percent of total U.S. trading activity from between January
2008 and December 2012, as shown in Figure 2.
---------------------------------------------------------------------------
\16\ Other internalization generally includes trades matched
internally by broker-dealers based on the public quote. While these
market-makers maintain various quoting and pricing obligations to their
own client's orders, the desk may step in front of exchange-displayed
orders by simply matching the displayed price. This can occur either on
their client's behalf or for their own account (Tabb Group, ``A
Spotlight in the Dark: An Inevitable Debate'', November 2012).
Source: Rosenblatt Securities, Consolidated Tape Association.
Note: Total Off-Exchange Trading excludes BATS and DirectEdge
as they operated as ECNs at that time.
Point 2--SRO Immunity
In his written testimony, Mr. Mathisson stated that (1)
``courts have traditionally afforded exchanges `absolute
immunity' from civil liability for damages arising in
connection with their regulatory operations''; (2) ``It is a
dangerous situation when a for-profit enterprise can cause half
a billion dollars of losses for others, and not have the risk
of being held legally liable''; and (3) that ``Exchanges have
absolute immunity on errors . . . '' [emphasis added]. \17\
---------------------------------------------------------------------------
\17\ Written Testimony of Daniel Mathisson of behalf of Credit
Suisse, Before the Senate Banking Committee Subcommittee on Securities,
Insurance, and Investment ``Computerized Trading Venues: What Should
the Rules of the Road Be?'', December 18, 2012.
---------------------------------------------------------------------------
These statements are flatly inaccurate and warrant
clarification.
Exchanges do have immunity from private lawsuits for
damages when engaging in conduct consistent with the quasi-
governmental powers delegated to them pursuant to the Exchange
Act, and we believe that this well-developed doctrine is
critical to ensuring that exchanges are protected when
fulfilling those statutorily delegated powers. However,
exchanges still answer to the SEC for their conduct, and can
and have been fined by their primary regulator. This is true
regardless of whether immunity applies in particular cases.
Separately, we have commercial limits of liability in place
with our members, which are governed by rules and membership
agreements approved by the SEC. These rules and agreements
enable members to be reimbursed for certain types of errors
made by the exchange. Moreover, these rules are submitted to
the normal notice-and-comment process required by the Exchange
Act, which gives the public--including our members--the
opportunity to voice concerns about the exchange's rules.
Point 3--Tape Revenue for Consolidated Market Data
In his written and oral testimony, Mr. Mathisson/Credit
Suisse asserts that, ``The CTA (Consolidated Tape Association)
has a legal monopoly on providing a consolidated stream of
real-time data from our Nation's stock markets. The CTA makes a
profit of approximately $400m per year which is then
distributed to its participant exchanges based on a complex
formula. ATSs do not receive tape revenue'' [emphasis added].
\18\ Mr. Mathisson later alleges that `` . . . the exchanges
are entitled to market data revenues to offset the costs of
regulating their markets . . . [and] revenues from market data
are way out of proportion with the costs of exchanges' self-
regulatory responsibilities.'' The citation for this statement
is the SEC's 1999 Concept Release: Regulation of Market
Information Fees and Revenue, Exchange Act Release No. 42208
[hereinafter SEC 1999 Market Data Concept Release]. Finally Mr.
Mathisson/Credit Suisse alleges that ``[b]ecause the CTA
allocated revenue to the exchanges is based on a complex
formula involving variables such as each exchange's number of
quotations, for-profit exchanges try to set policies and
services that will increase the level of quoting activity.''
\19\ Mr. Mathisson/Credit Suisse claims that a by-product of
these alleged policies is quote flickering. These claims
require clarification:
---------------------------------------------------------------------------
\18\ Written Testimony of Daniel Mathisson of behalf of Credit
Suisse, Before the Senate Banking Committee Subcommittee on Securities,
Insurance, and Investment ``Computerized Trading Venues: What Should
the Rules of the Road Be?'', December 18, 2012.
\19\ Written Testimony of Daniel Mathisson of behalf of Credit
Suisse, Before the Senate Banking Committee Subcommittee on Securities,
Insurance, and Investment ``Computerized Trading Venues: What Should
the Rules of the Road Be?'', December 18, 2012.
1. While the CTA plan does not directly pass tape revenue to
non-Exchange participants, the two FINRA TRFs rebate an
estimated $50 to $60 million in tape revenue annually
back to brokers, including ATSs and broker-dealer
---------------------------------------------------------------------------
internalizers.
2. The SEC 1999 Market Data Concept Release was not a
rulemaking or adjudication by the SEC--it was an
invitation for comment on matters the SEC was
considering, and it did not result in rulemaking. In
addition, the SEC 1999 Market Data Concept Release
repeatedly refers to the market data pool as being used
to fund ``the market's operation and regulation''
[emphasis added]. As such, it is not accurate to
represent that the SEC expressed a view, even in the
context of the SEC 1999 Market Data Concept Release,
that market data fees are exclusively meant to fund
regulation.
3. The market data allocation formula is not based on an
``exchange's number of quotations,'' but instead is
partly based on dollar size and the amount of time at
the NBBO (i.e., quoting share), which accounts for 50
percent of the allocation formula. The other 50 percent
allocation is based on trading share. This allocation
formula was established by the SEC in Reg NMS to
incentivize both price discovery and market quality.
Importantly, the formula requires that a quotation must
be displayed for a minimum of one full second to earn
quote credits. \20\ This minimum time period was put in
place by the SEC specifically to prevent quote
flickering. In this regard, the SEC stated in its
explanation to commenters that opposed the need for a
market data plan formula, `` . . . the Commission's
primary objective is to correct an existing flaw in the
current formulas by allocating revenues to those SROs
that, even now, benefit investors by contributing
useful quotations to the consolidated data stream.''
\21\ As such, we believe it is incorrect to assert that
the SEC market data allocation formula incentivizes
quote flickering.
---------------------------------------------------------------------------
\20\ http://www.sec.gov/rules/final/34-51808.pdf
\21\ http://www.sec.gov/rules/final/34-51808.pdf, p. 264.
More broadly, as noted in our oral testimony, registered
exchanges operate under a regime of obligations and benefits.
These obligations are significant, and include: rule filing
obligations, restrictions on business operations, the inability
to transact with public customers, member ownership
limitations, fair access and self-regulatory obligations. Mr.
Mathisson and Credit Suisse seem to argue that any economic
incentives associated with the fulfillment of these important
obligations should be allocated to the broker-dealers and
investment banks and not to the exchanges that bear these
significant regulatory obligations that benefit the industry
and subsequently, the public.
As stated in our testimony, we would welcome a holistic
review of our current U.S. market structure, including the
benefits and obligations of registered exchanges and broker-
dealers; however we would caution that the investment banks'
and broker-dealers' economic self-interests should not be the
linchpin upon which investor protection rests.
Thank you again for the opportunity to testify and provide
our thoughts. We would be happy to answer any additional
questions that you may have.
------
RESPONSES TO WRITTEN QUESTIONS OF CHAIRMAN REED
FROM DANIEL MATHISSON
Q.1. There was testimony that ``dark trading'' is a concern in
many countries and that Canada recently modified its market
structure to limit dark trading and to maximize price
discovery. Canadian regulators have imposed a new framework
governing how dark pools and undisplayed orders are allowed to
operate including priority of lit over dark flow and a minimum
price improvement requirement for dark orders. What is your
view of these reform proposals? Would these measures make the
U.S. markets more or less fair and transparent? Please explain.
Would these measures be feasible in the U.S. markets? Why or
why not?
A.1. Canada rolled out new rules in October of 2012 that were
intended to limit broker-dealer crossing platforms. The new
rules forced several market participants to either exit
Canadian markets (e.g., Goldman Sachs Sigma X) or drastically
alter their business models (e.g., TSX Alpha Intraspread).
According to the Rosenblatt Report released on December 19th
2012, combined dark volumes in Canada fell significantly to
2.06 percent in November, from 3.54 percent in October and an
original 5.67 percent of equity trading volume in September
before the new framework went into effect. While it appears the
new rules have skewed the competitive environment, hurting
Canadian ATS operators and helping exchanges, we have seen no
empirical evidence that this change has been beneficial for
investors.
We believe that the U.S. regulators should carefully review
any future academic studies that demonstrate how the Canadian
change has impacted liquidity, price stability, bid/ask
spreads, and other empirical measures of market health. In the
absence of evidence that the rule change has helped investors,
it would not make sense for the U.S. to blindly follow Canada's
lead. We believe that a rule change like this would have a
major impact on the existing market structure, and this should
only be considered as part of a comprehensive market reform
that reviews the entire competitive landscape of the for-profit
exchanges and broker-dealers.
Q.2. Rule 612 of Regulation NMS prevents sub-penny quoting.
However, under an exemption in Rule 612 the SEC actually allows
broker-dealers to execute orders in sub-penny increments.
Because there are no quoting obligations for broker-dealer
internalization, broker-dealers can provide price improvement
to their customers in the form of sub-penny executions. What
are advantages or disadvantages of sub-penny quoting? Is
special pricing good for retail investors? Why or why not?
A.2. We believe that sub-penny trading is net beneficial for
retail clients, as well as for institutional clients. In the
current market structure, retail clients typically receive an
immediate fill at a price slightly better than the bid or
offer. For example, if Ford Motors is showing a 13.83 offer, a
retail client buying 2,000 shares would often receive a fill
of, say, 13.829, saving a small amount of money on the
transaction vs. the offer price. Known as ``price
improvement,'' those small amounts can add up to significant
sums over a lifetime of investing.
The SEC has recently relaxed Rule 612, by approving pilot
programs by the NYSE and BATS Exchange which offer price
improvement and allow sub-penny trading (``Retail Liquidity
Program'' approved in July 2012). Since exchanges are now
permitted to accept and rank orders in sub-penny increments in
the RLP program, we believe broker-dealer operated Alternative
Trading Systems should be able to do so as well.
Q.3. Some commentators have suggested that the Securities and
Exchange Commission (SEC) should lift the ban on locked
markets. Locked markets occur when a trader attempts to place a
bid on one exchange at the same price as an offer on a
different exchange. They argue that bids and offers at the same
price but different venues should be forced to interact, and
this could reduce fragmentation in the marketplace and perhaps
reduce the prevalence of trading in dark pools and
internalization venues. What is your view of this proposal?
A.3. While the rationale for banning locked markets under Rule
611 of Regulation NMS Rule is to prevent trading at inferior
prices, one of the unintended effects of Rule 611 has been the
proliferation of exchange order types with the sole purpose of
allowing sophisticated users to unfairly gain queue priority
over simple limit orders. These order types have contributed to
an increase in pinging and cancellation rates. We believe that
lifting the ban on locked markets would greatly simplify the
U.S. equity markets and possibly lead to better execution
results for investors. We recommend that the SEC implement a
pilot program lifting the ban on locked markets in several NMS
symbols to conduct analysis and gather empirical data to
measure the effects and impact to the markets.
Q.4. Internalizing broker-dealers often pay retail brokers to
direct customer orders to their trading venues. Does this
practice advantage or disadvantage retail investors? Why or why
not? Why is retail flow valuable to trade against?
A.4. Retail investors are advantaged by the current system,
where brokers are allowed to engage in market-making. Market-
makers typically pay rebates to retail brokers for order flow,
which allows the retail brokers to charge very low commissions
to the end customer. Market-makers typically provide better
quality of execution than exchanges, as execution quality
statistics are made public under SEC Rule 605, resulting in
savings for the investor.
Retail flow is considered valuable to market-makers because
it typically consists of a steady stream of small market
orders. A market-maker will try to buy near the bid from client
A and sell near the offer to client B, capturing a spread in
the process. Market-makers prize small orders because they are
less risky than large institutional orders.
Q.5. Broker-dealers that operate their own dark pools can get a
fee advantage when they route customer orders through their own
dark pools rather than through a lit exchange. Does this create
a conflict of interest between investors, who want to buy or
sell stocks at the best prices, and brokers, who want to avoid
exchange fees? Why or why not?
A.5. This practice does not create a conflict of interest.
Under Regulation NMS, orders in the U.S. may not ``trade
through'' the National Best Bid Offer (NBBO), meaning that if a
stock is offered at 18.50 on an exchange, a broker-dealer may
not route a buy order to a venue where the order would be
filled at 18.51 or higher. The client is guaranteed to do no
worse than the best current price on the national market
system. Within that significant constraint, as part of their
best execution obligations, broker-dealers take into account
multiple factors when deciding where to route client orders.
These factors include the potential for price improvement, the
speed of execution, and the likelihood that an order will be
filled.
Q.6. The U.S. exchanges are required to submit data to a
centralized network where it is aggregated before being
disseminated to the public. However, the Consolidated Tape
Association/Securities Information Processor or SIP now lags
behind direct proprietary market data feeds. As a result, high
frequency trading firms and other market participants now
usually colocate their computer servers at every exchange and
subscribe to proprietary data feeds offered by the exchanges in
order to capitalize on latency arbitrage opportunities. Why
should anyone receive market data flow from other than the
consolidated tape? Should market data be distributed to all
market participants equally? How do you think this can best be
accomplished?
A.6. In an ideal system, we believe that all intraday market
data would be supplied by the SIP. However, we recognize that
this would create a significant disruption to the existing
business models of the for-profit exchanges as well as many
trading firms, and therefore regulators should carefully
analyze and review the costs and benefits prior to changing the
current data distribution model.
Q.7. The October closure of the markets for 2 days due to
Hurricane Sandy raised questions about the financial sector's
preparedness for the next natural or man-made disaster. What
challenges did the markets face in reopening after Hurricane
Sandy? What should have been done differently after Hurricane
Sandy to keep providing trading services to customers and
maintain market integrity? What can the stock exchanges do to
prepare the market for another disaster? What changes need to
be made by market participants and regulators?
A.7. The decision to close the U.S. equities markets on October
29th and October 30th was made after serious consideration and
with industry-wide consensus on Sunday, October 28th, as
Hurricane Sandy was bearing down on East Coast. Several
contingency plans were discussed, but the primary concern was
the safety of employees. All contingency plans involved a
number of key employees having to commute and work in areas in
immediate danger of being flooded or destroyed.
Most U.S. financial institutions, including Credit Suisse,
have backup power generators and disaster recovery sites to
maintain uninterrupted trading and customer service during
fire, flooding, or any other event impacting the firms' primary
location. However, given the rarity of events as severe as
Hurricane Sandy, it would be impractical and tremendously
costly for the entire industry to maintain fully operational,
fully staffed alternative sites and facilities in various
locations throughout the country. According to our research,
during the past 100 years, the U.S. markets have been closed
due to inclement weather only five times, including the two
days of Hurricane Sandy. We agree that a continued focus on
disaster recovery is prudent and necessary, but believe it is
acceptable for the markets to close during extreme region-wide
weather events, since these occur so rarely.
------
RESPONSES TO WRITTEN QUESTIONS OF CHAIRMAN REED
FROM ERIC NOLL
Q.1. There was testimony that ``dark trading'' is a concern in
many countries and that Canada recently modified its market
structure to limit dark trading and to maximize price
discovery. Canadian regulators have imposed a new framework
governing how dark pools and undisplayed orders are allowed to
operate including priority of lit over dark flow and a minimum
price improvement requirement for dark orders. What is your
view of these reform proposals? Would these measures make the
U.S. markets more or less fair and transparent? Please explain.
Would these measures be feasible in the U.S. markets? Why or
why not?
A.1. NASDAQ OMX does not oppose all dark trading, but rather
focuses on the need to make our markets as efficiently
transparent as possible to ensure that the price discovery
process is robust and that capital formation is maximized.
NASDAQ OMX recognizes that large institutional trades
between natural buyers and sellers should be able to occur at
prevailing market prices outside the confines of an exchange or
ECN in order to minimize disruption to the market. In executing
all other types of transactions the emphasis should be on open,
efficient, and transparent order interaction in order to ensure
that the price discovery process is robust, broker-client
conflicts of interest are minimized, and investor confidence
and capital formation are maximized. These nonblock
transactions can occur on exchanges or on ECNs. The Canadian
policy decision to limit the use of closed alternative systems
shows that regulators increasingly view it as important that
open order interaction be strongly encouraged. We support
innovative solutions like the Canadian Reforms and would like
to see the SEC fully explore like-minded reforms here in the
United States. Although the U.S. and Canadian markets are very
different in scale there are common economic issues underlying
the trading process in both markets. We view the SEC as
appropriately positioned to weigh and balance the ramifications
of investor protection, competition, liquidity, and regulatory
complexity and look forward to working with them to understand
the U.S. capital market's needs in this area.
Q.2. What is an example of an order type that the NYSE has
created to make our markets better, and improve transparency
and price discovery? How does this order type provide equal
access? Does the NYSE allow dark or undisplayed orders or
provide dark execution? Why or why not?
A.2. In comparison with some markets, NASDAQ adheres to a
principle of maintaining relative simplicity in its roster of
order types, and ensuring that the operation of all of its
order types is well understood by market participants. To that
end, NASDAQ recently posted a webinar describing its order
types, freely available to market participants and other
members of the public at http://www.brainshark.com/nasdaqomx/
vu?pi=zF7zJ6aUZzoG0z0. Further descriptions of order type
operation are available at http://www.nasdaqomxtrader.com/
content/ProductsServices/Trading/Workstation/rash--
strategy.pdf. Notably, NASDAQ does not offer any order types
that allow an order to jump ahead of previously posted orders
in execution priority. All order types are available to members
on equal terms.
NASDAQ believes that it is necessary to offer market
participants the option of posting nondisplayed orders to
enable NASDAQ to compete with the multiplicity of exchanges and
alternative trading venues that offer this option, and because
market participants representing large orders may have a
legitimate need to use nondisplayed orders as a means to guard
against adverse price movements. Moreover, nondisplayed orders
that are designed to price at the midpoint between the national
best bid and offer provide a means for market participants to
offer price improvement to their counterparties.
Of course NASDAQ believes that its exchange should be the
preferred venue for price discovery of the securities that it
lists and trades. To that end, our schedule of fees and rebates
for order execution offers a rebate for liquidity provided
through displayed orders that is higher, and in many instances
significantly higher, than the rebate for liquidity provided
through nondisplayed orders. The effect of these financial
incentives is borne out through actual patterns of trading,
with approximately 90 percent of executions on NASDAQ based on
the use of displayed orders. In addition, order types that
NASDAQ offers that are designed to reprice aggressively to
increase the order's chances for execution are generally
required, by their terms, to be displayed orders.
Q.3. Rule 612 of Regulation NMS prevents sub-penny quoting.
However, under an exemption in Rule 612 the SEC actually allows
broker-dealers to execute orders in sub-penny increments.
Because there are no quoting obligations for broker-dealer
internalization, broker-dealers can provide price improvement
to their customers in the form of sub-penny executions. What
are advantages or disadvantages of sub-penny quoting? Is
special pricing good for retail investors? Why or why not?
A.3. NASDAQ believes that quoting increments must be uniform
across all trading venues and market participants, including
over-the-counter trading and broker-dealer internalization.
Additionally, a one-size-fits-all increment regime is
inefficient and ineffective; increments must be flexible and
tailored to the characteristics of varying classes of
securities. In fact, NASDAQ has several times urged the
Commission to add tick increments both larger and smaller than
those set forth in SEC Rule 612, most recently in its April
2010 joint letter with the NYSE and other exchanges seeking an
SEC exemption from Rule 612 for certain actively traded
securities. No action has been taken on that request for
exemptive relief.
These two principles dictate a reversal of the above-
referenced exemption from Rule 612. That exemption has
contributed significantly to the migration of trading away from
lit markets and into the dark markets. While broker-dealers
claim to provide price improvement to retail orders executed in
the dark, that price improvement may be financially immaterial
and outweighed by the negative impact of dark trading on price
formation and market fairness.
Q.4. Some commentators have suggested that the Securities and
Exchange Commission (SEC) should lift the ban on locked
markets. Locked markets occur when a trader attempts to place a
bid on one exchange at the same price as an offer on a
different exchange. They argue that bids and offers at the same
price but different venues should be forced to interact, and
this could reduce fragmentation in the marketplace and perhaps
reduce the prevalence of trading in dark pools and
internalization venues. What is your view of this proposal?
A.4. This is an interesting theoretical concept that we think
should be considered along with other approaches as the SEC
evaluates U.S. market structure. The spread between bid and
offer historically existed to compensate providers of
liquidity, typically trading professionals, for the risk and
expense associated with posting orders to buy and sell. In a
series of well-conceived reforms, the SEC has opened the
process of posting orders to buy and sell to all investors. As
investors are motivated to trade for reasons other than short
term trading profits motivating the professional there is no ex
ante reason that a positive bid and offer spread should exist
in situations where investors are separated by their preferred
locus of trading. Should such locked markets occur, as they do
frequently under Regulation NMS today, we expect that these
locked markets would quickly clear as brokers and traders act
to buy or sell at a price more favorable than would otherwise
exist. We recognize that the SEC's position on locked markets
was driven in part by the market access fees charged by certain
trading platforms during the time Regulation NMS was under
consideration and that a review of the SEC's ban on locked
markets might reasonably also include a review of the market
access fee caps created by Regulation NMS.
Q.5. Internalizing broker-dealers often pay retail brokers to
direct customer orders to their trading venues. Does this
practice advantage or disadvantage retail investors? Why or why
not? Why is retail flow valuable to trade against?
A.5. NASDAQ believes that the public would view negatively the
practice of retail brokers ``shopping'' their orders to the
highest bidder, particularly where it results almost
exclusively in their orders being routed to and executed in
dark markets. Notably, the overwhelming majority of retail
investors' orders, whether entered by individuals, mutual
funds, or pensions, are handled by sophisticated trading firms
that are obligated to deliver best execution to such orders and
to disclose payment arrangements. The question becomes whether
those duties place sufficient constraints on potential
conflicts of interest to overcome the negative perception of
payment for order flow practices.
Q.6. Broker-dealers that operate their own dark pools can get a
fee advantage when they route customer orders through their own
dark pools rather than through a lit exchange. Does this create
a conflict of interest between investors, who want to buy or
sell stocks at the best prices, and brokers, who want to avoid
exchange fees? Why or why not?
A.6. Fragmentation and darkness are eroding the quality and
perceived fairness of U.S. markets. NASDAQ supported the
Commission's November 2009 proposal to modify the regulation of
nonpublic trading interest. That proposal would have attempted
to address the problem of so-called ``dark'' pools that use
indications of interest to, effectively, display quotes to only
a small subset of the national market system. It also would
have begun to address the proliferation of dark pools that has
fostered the two-tiered market which is anathema to the public
interest.
Q.7. The U.S. exchanges are required to submit data to a
centralized network where it is aggregated before being
disseminated to the public. However, the Consolidated Tape
Association/Securities Information Processor or SIP now lags
behind direct proprietary market data feeds. As a result, high
frequency trading firms and other market participants now
usually colocate their computer servers at every exchange and
subscribe to proprietary data feeds offered by the exchanges in
order to capitalize on latency arbitrage opportunities. Why
should anyone receive market data flow from other than the
consolidated tape? Should market data be distributed to all
market participants equally? How do you think this can best be
accomplished?
A.7. The U.S. national market system creates the fastest,
deepest, most transparent and most reliable market data of any
jurisdiction. No other market system creates aggregate data
capturing a dollar or share volume of trading from as large a
number of trading venues as the United States. In addition to
aggregate data from the network processors, the flexibility
granted by the Commission in Regulation NMS has sparked
innovation and competition by exchanges, ATSs, and broker-
dealers to create a previously unimagined variety of optional
data products. This broad range of options supports an equally
broad range of competing business models that require different
amounts, types, and speeds of data. The coexistence of multiple
competing business models, supported by a variety of data
products, contributes positively to price formation and
liquidity. In NASDAQ's view, eliminating proprietary data
products that support competing business models would harm the
U.S. market and investors with little or no off-setting
benefit. Additionally, it is worth noting that market
participants that collocate in exchange facilities serve a wide
variety of business models, including many highly sophisticated
firms that serve retail investors.
Q.8. The October closure of the markets for 2 days due to
Hurricane Sandy raised questions about the financial sector's
preparedness for the next natural or man-made disaster. What
challenges did the markets face in reopening after Hurricane
Sandy? What should have been done differently after Hurricane
Sandy to keep providing trading services to customers and
maintain market integrity? What can the stock exchanges do to
prepare the market for another disaster? What changes need to
be made by market participants and regulators?
A.8. Hurricane Sandy was a devastating storm that hit at both
the physical assets and human resource center of our industry.
It was the first time since 1885 that two trading days were
lost to weather-related issues. We prepare for emergencies such
as Sandy and other unexpected market events in several ways.
With respect to physical infrastructure, NASDAQ OMX and most of
our partners in the trading community maintain geographic and
systems diversity that would have allowed contingency trading
and regulation of our markets in the aftermath of Hurricane
Sandy.
NASDAQ OMX implemented our own emergency plans before,
during and after Sandy made landfall. Our key employees and
systems were in place and ready to operate. From a legal
perspective, the SEC granted our board of directors and persons
designated by the board the authority, memorialized in our by-
laws, to take ``any action'' regarding the operation of our
market or the trading of any and all securities in the event of
an emergency or extraordinary market conditions. This gives us
the ability to adapt to unexpected events to protect the
integrity of the market, to protect investors and the public
interest, and otherwise to ensure that we are fulfilling our
mandates under the Securities Exchange Act of 1934.
Market integrity was clearly threatened by the moral
dilemma presented by the prospect of operating the markets
during Hurricane Sandy. Operating the market would have
demanded that professionals from across our industry venture
out into dangerous conditions, potentially imposing additional
burdens on already stressed public safety personnel. Owing to
the risks to human life in New York and New Jersey and for
employees of the markets, our customers and regulators, the
markets made the best decision under the circumstances. NASDAQ
supported this collective decision, and exercised its emergency
authority to close its market.
The industry learned valuable lessons from the experience.
We have already looked at how our contingency plans worked in a
real-life situation. We are pleased that, without hesitation,
we can say that NASDAQ's testing and readiness plans, thanks to
our dedicated employees, had our market ready to operate and
trade all of our listed stocks in a normal and uninterrupted
manner.
Q.9. It has recently been reported that the SEC's Interim
Inspector General found that some staff in the SEC's Trading
and Markets Division did not encrypt computers containing
confidential data from the exchanges and clearing agencies they
were overseeing. What measures are being taken now or should be
taken by market participants and regulators to better secure
data in today's high tech markets?
A.9. NASDAQ is fully cognizant that attaining perfection in the
operation and implementation of technology is impossible, even
when it is developed and monitored by the most diligent
personnel and rigorous systems. At the SEC's own October 2,
2012, Technology and Trading Roundtable the SEC's academic
expert stated as much. Dr. Nancy Leveson, Professor of
Aeronautics and Astronautics and Engineering Systems, at the
Massachusetts Institute of Technology, describing her 47 years
in computer science, at IBM, MIT, and in her own business
stated:
Let me tell you a little bit of what I've learned in the
last 47 years. The first lesson is that all software contains
errors. I have not in all of that time ever come across any
software that was nontrivial in which no errors were found
during operations. The errors may not surface for a long time
but they are lurking there and waiting for just the right
conditions to occur. There are also some myths about certain
industries being able to create perfect software but
unfortunately this is patently untrue. No industry creates
perfect software.
NASDAQ fully agrees with this statement and, as such,
understands the difficulty that all organizations, including
the SEC, experience in implementing technology in a secure,
effective manner.
Moreover, the SEC Inspector General findings with respect
to computer security at the SEC are another reminder that
security is a shared responsibility between the markets,
participants, and Government. The SEC seems fully committed to
resolving the issues highlighted in the IG report, which raised
concerns about one isolated area of the SEC that can be
addressed appropriately through training and understanding
about security protocols and best practices for the treatment
of all sensitive information. NASDAQ OMX would welcome the
opportunity to work with the SEC to ensure that the data we
share with them is communicated in a secure manner and
protected from unauthorized disclosure in the future. NASDAQ
OMX is investing heavily to mitigate risks in the computer
security area--we have hired nationally respected experts,
upgraded our own systems, are in the process of enhancing our
systems integrity testing and are demanding similar actions by
those with whom we interact. NASDAQ OMX supports Congressional
action to foster information sharing between national
infrastructure operators and Government agencies that have
access to information about cyber security threats, technology
best practices and other resources that can help us protect
vital infrastructures like our security markets.
------
RESPONSES TO WRITTEN QUESTIONS OF CHAIRMAN REED
FROM ROBERT C. GASSER
Q.1. There was testimony that ``dark trading'' is a concern in
many countries and that Canada recently modified its market
structure to limit dark trading and to maximize price
discovery. Canadian regulators have imposed a new framework
governing how dark pools and undisplayed orders are allowed to
operate including priority of lit over dark flow and a minimum
price improvement requirement for dark orders. What is your
view of these reform proposals? Would these measures make the
U.S. markets more or less fair and transparent? Please explain.
Would these measures be feasible in the U.S. markets? Why or
why not?
A.1. The Canadian regulators have clearly stated they made
these changes despite a total lack of evidence that dark
trading was negatively impacting Canadian market quality. Most
academic evidence suggests that not only does dark not harm
price discovery, it generally improves it. \1\ ITG's own
analysis of the Canadian markets has shown a continued decline
in institutional trading costs as dark pools have gained
acceptance and grown in that market. While it is too early to
fully discern the impact of the October 2012 rule change, there
is certainly no glaring evidence these rules have improved the
marketplace or added to the availability of displayed
liquidity. Early indications in the Canadian equity markets are
that there has been a decline in order flow at the dark venues
and an increase in trading flow posted as dark order types on
the displayed markets.
---------------------------------------------------------------------------
\1\ See, Ian Domowitz, ITG, ``Are We Missing the Evidence in the
Global Dark Pool Debate?'', December 2010 http://is.gd/qpumXJ. See
also: Haoxiang Zhu, MIT, ``Do Dark Pools Harm Price Discovery?'',
November 2012. http://www.mit.edu/zhuh/Zhu_darkpool.pdf
---------------------------------------------------------------------------
The U.S. equity market is much larger and more complex than
the Canadian market. With 13 exchanges and over 40 dark pools,
there is healthy competition for order flow, which has resulted
in a dramatic decrease in both implicit and explicit trading
costs over the past decade. It is a generally accepted
principle in the U.S. that regulators should not make changes
that simply favor one set of competitors over another without a
resultant improvement in overall market structure. In our
assessment, imposing rules similar to the Canadian measures in
the U.S. would neither be feasible nor advisable, as they would
not result in improved liquidity or better outcomes for
institutional or retail investors.
Q.2. Rule 612 of Regulation NMS prevents sub-penny quoting.
However, under an exemption in Rule 612 the SEC actually allows
broker-dealers to execute orders in sub-penny increments.
Because there are no quoting obligations for broker-dealer
internalization, broker-dealers can provide price improvement
to their customers in the form of sub-penny executions. What
are advantages or disadvantages of sub-penny quoting? Is
special pricing good for retail investors? Why or why not?
A.2. There is clear value in price improvement through sub-
penny executions, but sub-penny quoting is unnecessarily
complex and of dubious benefit to investors. Sub-penny
executions, in the aggregate, have lowered execution costs for
retail investors by millions of dollars. It is appropriate for
retail investors to receive this type of special pricing
because retail order flow carries little risk of adverse
selection; specifically, interaction with retail order flow
does not involve many of the concerns that arise when
interacting with orders from certain high frequency trading
strategies, such as information leakage and price
deterioration. In contrast, sub-penny quoting does not offer a
clear benefit. It would add to the already onerous burden of
rapidly growing message traffic and would impede market making,
as market makers would not be able to offer meaningful price
improvement. Sub-penny price quotes would make stock auctions
difficult, if not impossible, eroding overall liquidity in the
marketplace.
Q.3. Some commentators have suggested that the Securities and
Exchange Commission (SEC) should lift the ban on locked
markets. Locked markets occur when a trader attempts to place a
bid on one exchange at the same price as an offer on a
different exchange. They argue that bids and offers at the same
price but different venues should be forced to interact, and
this could reduce fragmentation in the marketplace and perhaps
reduce the prevalence of trading in dark pools and
internalization venues. What is your view of this proposal?
A.3. We believe the ban on locked markets causes more problems
than it solves, and as such it should be lifted. There is much
unnecessary complexity in current equity market structure
stemming from efforts to prevent locking and crossing of
markets. Locked markets are not that difficult to address from
a market structure perspective. Market participants are capable
of providing best execution in locked market situations when
handling customer orders on an agency or riskless principal
basis. Accordingly, we would support efforts to lift the ban.
Q.4. Internalizing broker-dealers often pay retail brokers to
direct customer orders to their trading venues. Does this
practice advantage or disadvantage retail investors? Why or why
not? Why is retail flow valuable to trade against?
A.4. As discussed earlier, retail order flow is valuable
because it is not informationally ``heavy'' and interacting
with it does not involve significant information leakage. As a
result, broker dealers are often willing to pay for this order
flow, which yields clear benefits to retail investors in the
form of lower trading commissions. We would argue that the
practice of online brokers offering trading to retail customers
for less than $10 per execution would not be possible without
trading rebates or payment for order flows. In general, we do
not believe that this practice disadvantages retail investors
because broker-dealers have fiduciary responsibilities to their
clients to achieve best execution. It should also be noted that
broker-dealers that receive payment for order flow are required
to disclose such arrangements to the public pursuant to Rule
611 of Regulation NMS, and to their customers in confirmations
required under Rule 10b-10 of the Securities and Exchange Act
of 1934.
Q.5. Broker-dealers that operate their own dark pools can get a
fee advantage when they route customer orders through their own
dark pools rather than through a lit exchange. Does this create
a conflict of interest between investors, who want to buy or
sell stocks at the best prices, and brokers, who want to avoid
exchange fees? Why or why not?
A.5. While the routing of customer orders through internal dark
pools creates the potential for a conflict of interest, we
believe it is a manageable conflict. Broker-dealers have a
fiduciary duty to provide best execution when handling client
orders. This duty should be the controlling factor when making
order routing decisions Fulfillment of this duty can, and in
most cases is, accomplished through the use of detailed
transaction cost analysis and extensive evaluations of
execution quality, (e.g., opportunities for price improvement,
speed of execution, number of venues checked, character and
volatility of the market, etc.). Such reviews and analyses can
demonstrate whether internal crossing provides equivalent or
better execution costs and superior transaction prices compared
with routing orders to an external market center. In most
cases, using an internal dark pool yields three clear
advantages over exchanges: a lack of information leakage, lower
latency and potential opportunities for price improvement.
We would also note that there is much less of a conflict of
interest with an agency dark pool as opposed to a pool which
contains broker-dealer principal or proprietary order flow. In
the case of an agency pool such as ITG's POSIT, the broker-
dealer is generally unconflicted as it does not have a vested
interest in terms of profiting from one side of the trade. In
addition, internal crossing also protects clients from
incurring costs for lost opportunities. For example, a dark
pool operator could forego an internal crossing opportunity by
routing a client order to an exchange for the purpose of
avoiding a potential conflict of interest. However, by the time
the order arrives at the exchange, the targeted liquidity may
no longer be accessible. In such a situation, the client would
have been deprived of an opportunity to execute at the same
displayed price (or even better) in the dark pool.
Q.6. The U.S. exchanges are required to submit data to a
centralized network where it is aggregated before being
disseminated to the public. However, the Consolidated Tape
Association/Securities Information Processor or SIP now lags
behind direct proprietary market data feeds. As a result, high
frequency trading firms and other market participants now
usually colocate their computer servers at every exchange and
subscribe to proprietary data feeds offered by the exchanges in
order to capitalize on latency arbitrage opportunities. Why
should anyone receive market data flow from other than the
consolidated tape? Should market data be distributed to all
market participants equally? How do you think this can best be
accomplished?
A.6. We strongly believe that there should be equal
opportunities for accessing market data. All market
participants should have the same access to market data feeds,
either through the SIP or direct proprietary feeds. As a matter
of practice not all participants will elect to invest in the
technology, subscription fees, and/or colocation costs to be
able to take advantage of all proprietary data feeds. While we
believe it is important to ensure equality of market data
access, it would not be appropriate to attempt to ensure
equality of outcomes through regulatory means.
Q.7. The October closure of the markets for 2 days due to
Hurricane Sandy raised questions about the financial sector's
preparedness for the next natural or man-made disaster. What
challenges did the markets face in reopening after Hurricane
Sandy? What should have been done differently after Hurricane
Sandy to keep providing trading services to customers and
maintain market integrity? What can the stock exchanges do to
prepare the market for another disaster? What changes need to
be made by market participants and regulators?
A.7. We believe the decision to shut the market for 2 days in
the wake of Hurricane Sandy was the correct one. The damage to
lower Manhattan and the loss of life in affected areas supports
this contention. The 2-day hiatus ensured the safety of broker
dealer and exchange staff and enabled market participants to
test and fully activate their business continuity programs.
When markets re-opened they did so in an orderly and efficient
manner. In hindsight, better coordination between exchanges and
market participants would have resulted in a smoother re-
opening process. Of note, the decision to shut the equity
markets on the Monday after Sandy hit did not come down until
late Sunday night. There was some concern among broker-dealers
that the exchanges' disaster recovery plans had not been
sufficiently vetted. The lessons for preparation for future
disasters are to increase communication among all market
participants and to increase awareness regarding the specifics
of exchange disaster recovery plans.