Securities Market Operations: The Effects of SOES on the Nasdaq Market
(Letter Report, 08/31/98, GAO/GGD-98-194).

Pursuant to a congressional request, GAO provided information on the
Small Order Execution System (SOES) and its effects on the Nasdaq Stock
Market, focusing on the: (1) extent to which SOES is being used for its
intended purpose; (2) effects of SOES on the marketplace; (3) results of
attempts to limit trading on SOES or replace SOES; and (4) effects of
recent market developments and proposals affecting SOES.

GAO noted that: (1) the National Association of Securities Dealers
(NASD) developed SOES to provide automatic access to the Nasdaq Market
and market makers for the securities orders of small, retail investors;
(2) however, 1995 NASD data show that the primary users of SOES are SOES
day trading firms; (3) these firms generally have provided access to
SOES and proprietary computer software to assist the trading of day
traders; (4) market makers told GAO that many retail investors' orders
are being filled by market makers using their own internal automated
systems, at prices that are to match the best Nasdaq prices; (5) the day
traders have used a trading advantage provided by the automatic
execution feature of SOES, proprietary software the SOES day trading
firms designed to benefit from this feature, and Nasdaq rules that
limited market maker access to SOES, to profit at the expense of market
makers and their customers; (6) this advantage has diminished over time
because of market maker software improvements and changes in trading
rules; (7) market makers and SOES day trading firm officials disagree
about the effects of SOES day trading on the market; (8) NASD has
proposed various rule changes to reduce or eliminate the advantage of
SOES day traders; (9) however, a court decision in favor of a SOES day
trader strengthened the position of the SOES day trading firms and their
day traders; (10) Securities and Exchange Commission (SEC) officials
told GAO that they had become more skeptical of NASD rule proposals to
change SOES, and that NASD had failed in two attempts to eliminate SOES'
automatic execution feature in favor of systems that attempted to limit
the availability of automatic executions exclusively to smaller
investors and that routed orders to market makers for execution; (11)
despite NASD's lack of success in controlling SOES traders through rule
changes, the volume of SOES trading as a percentage of total Nasdaq
share volume has been declining; (12) to make automatic trade executions
more widely available, Nasdaq has proposed an Integrated Order Delivery
and Execution System that would replace two existing systems, including
SOES, and provide an integrated order routing and execution system and a
voluntary limit order file, in which customer orders at specified prices
could be matched and executed; and (13) Nasdaq officials told GAO that
this system is intended to eliminate the unintended double liability
incurred by market makers as a result of SOES day trading, without
unduly benefiting market makers.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  GGD-98-194
     TITLE:  Securities Market Operations: The Effects of SOES on the 
             Nasdaq Market
      DATE:  08/31/98
   SUBJECT:  Securities regulation
             Antitrust law
             Brokerage industry
             Stock exchanges
             Information systems
             Trade regulation
             Restrictive trade practices
IDENTIFIER:  National Association of Securities Dealers Small Order 
             Execution System
             Nasdaq Integrated Order Delivery and Execution System
             
******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO report.  Delineations within the text indicating chapter **
** titles, headings, and bullets are preserved.  Major          **
** divisions and subdivisions of the text, such as Chapters,    **
** Sections, and Appendixes, are identified by double and       **
** single lines.  The numbers on the right end of these lines   **
** indicate the position of each of the subsections in the      **
** document outline.  These numbers do NOT correspond with the  **
** page numbers of the printed product.                         **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
** A printed copy of this report may be obtained from the GAO   **
** Document Distribution Center.  For further details, please   **
** send an e-mail message to:                                   **
**                                                              **
**                                            **
**                                                              **
** with the message 'info' in the body.                         **
******************************************************************


Cover
================================================================ COVER


Report to Congressional Requesters

August 1998

SECURITIES MARKET OPERATIONS - THE
EFFECTS OF SOES ON THE NASDAQ
MARKET

GAO/GGD-98-194

The Effects of SOES on the Nasdaq Market

(233540)


Abbreviations
=============================================================== ABBREV

  DJIA - Dow Jones Industrial Average
  DOJ - Department of Justice
  ECN - Electronic Communication Network
  NASD - National Association of Securities Dealers
  PTA - Professional Trading Account
  SEC - Securities and Exchange Commission
  SOES - Small Order Execution System

Letter
=============================================================== LETTER


B-278412

August 31, 1998

The Honorable Charles E.  Schumer
House of Representatives

The Honorable Gary L.  Ackerman
House of Representatives

The Honorable Sam Gejdenson
House of Representatives

This report responds to your request that we study the Small Order
Execution System (SOES) and its effects on the Nasdaq Stock Market.\1
SOES is a securities trading system that provides immediate automated
executions of trades in the Nasdaq Market.  Our objectives were to
determine (1) the extent to which SOES is being used for its intended
purpose, (2) the effects of SOES on the marketplace, (3) the results
of attempts to limit trading on SOES or replace SOES, and (4) the
effects of recent market developments and proposals affecting SOES. 


--------------------
\1 The Nasdaq Stock Market, Inc., is a wholly-owned subsidiary of the
National Association of Securities Dealers (NASD), a registered
securities association.  Nasdaq began operation in 1971 as the first
screen-based stock market for non-exchange listed securities. 


   BACKGROUND
------------------------------------------------------------ Letter :1


      HOW THE NASDAQ STOCK MARKET
      WORKS
---------------------------------------------------------- Letter :1.1

The Nasdaq Market is a market for securities traded
"over-the-counter" through a network of computers and telephones,
rather than on a stock exchange floor.  Nasdaq is an electronic
communications system in which certain NASD member broker-dealers act
as market makers by quoting prices at which they are willing to buy
("bid quote") or sell ("ask quote") securities for their own accounts
or for their customers.  For the securities that are listed on the
Nasdaq system, market makers enter into the system their bid and ask
quotes and the number of shares they are willing to buy or sell
(referred to as "depth" or "size").  Market makers are obligated to
trade at the depth that they are publicly displaying under Nasdaq's
firm quote rule.\2 Market makers often choose, or under best
execution obligations are required, to provide executions at least as
favorable as that being displayed publicly by other participants,
regardless of whether the market maker itself is publicly displaying
that price or depth. 

The Nasdaq computer system brings together the bid, ask, and depth
quotations of the numerous market makers for all Nasdaq-listed
securities for display to Nasdaq member broker-dealers and other
subscribers.  The Nasdaq workstation, through which all Nasdaq market
makers and other member broker-dealers can view the market, displays
and continuously updates bid and ask prices and the quote depth in a
montage of quotes for each listed security.  The Nasdaq Market has
thus been referred to as a "dealer" market in which prices are set by
the interaction of dealer quotes, although, as discussed later, this
has changed under SEC's Order Handling Rules to include the
interaction of limit orders.\3

In such a dealer market, market makers have the responsibility to
provide liquidity by being continuously willing to buy and sell the
security or securities in which they are making a market (or risking
their capital), at least at their publicly quoted prices.  Liquidity
is the ease with which the market can accommodate large volumes of
securities trading without significant price changes.  As a result,
an individual who wants to buy or sell a security at the quoted price
would not have to wait until someone is found who wishes to take the
opposite side in the desired transaction.  The difference between the
listed bid and ask prices is the "spread" that market makers retain
as compensation for their effort and risk. 

Investors who want to buy or sell Nasdaq-listed securities place
orders with broker-dealers.  The brokers have several ways to handle
the orders.  If the broker makes a market in the stock, it may
execute the order itself, generally through its own internal
automated execution system.  If the broker does not make a market in
the stock, it may send the order to market makers through Nasdaq's
automated order entry and delivery system, called SelectNet. 
SelectNet orders may either be directed to a specific market maker or
broadcast generally to all market makers.  In conformance with SEC
and NASD rules, market makers must execute directed SelectNet orders,
commonly referred to as "preferenced" orders,\4 up to their displayed
quotation size, if the orders are priced equal to their quote, unless
the market maker has just traded or moved its quote.  Directed orders
that must be executed are commonly referred to as "liability" orders. 
Other nonliability SelectNet orders can be accepted, negotiated, or
declined by one or more market makers.  Brokers may enter small
orders of up to 1,000 shares for most stocks into SOES, where they
are automatically executed without negotiation.  The broker may
negotiate large orders or those requiring special handling directly
with one or more market makers. 


--------------------
\2 The firm quote rule requires that a market maker execute a
transaction at its displayed quotations for at least a normal unit of
trading, 100 shares, or for the actual size of the quotation if it is
greater than the normal unit. 

\3 A limit order is a customer order to buy or sell a security at a
specific price. 

\4 Preferenced orders are executed at the best quoted price, even
though that price may not be that market makers quoted price. 
However, market maker's can designate the securities for which they
are willing to accept preferenced orders.  Unpreferenced orders are
routed on a rotational basis to the SOES market makers quoting the
best bid or ask prices and are executed at those prices. 


      HOW SOES WORKS
---------------------------------------------------------- Letter :1.2

SOES is a system that allows small orders placed through it to be
automatically executed against Nasdaq market makers at the best bid
or ask prices displayed on the Nasdaq system.  SOES trading has
generally represented less than 10 percent of total Nasdaq shares
traded.  SOES can be used by NASD member firms that register as
participating SOES order entry firms.  These firms may only enter
orders that have been received from their public customers.  Some of
these firms' primary, if not exclusive, business line is SOES
trading.  We call these firms SOES day trading firms.  All Nasdaq
market makers in Nasdaq National Market securities (Nasdaq's largest
stocks, accounting for over 90 percent of Nasdaq volume) must be SOES
market makers and receive SOES executions from SOES order entry
firms.  Participation in SOES by market makers in Nasdaq's smaller
securities (called Small Cap securities) is voluntary. 

Market makers can also use SOES to execute their customer orders. 
However, Nasdaq rules do not allow market makers to enter orders into
SOES for stocks for their proprietary trading (trading for their own
account). 

Before January 1997, SOES rules generally required market makers to
quote a minimum of 1,000 shares to buy or sell (for some less
actively traded shares, the minimums may be 500 or 200 shares). 
Under a pilot program begun in January 1997, called the actual size
pilot, these regulatory minimums were removed for 150 test stocks,
and market makers were allowed to enter quotations at any size in 100
share increments (and multiples of 100).  SEC approved the expansion
of actual size rules to all Nasdaq stocks on July 15, 1998, effective
July 20, 1998. 


   RESULTS IN BRIEF
------------------------------------------------------------ Letter :2

NASD developed SOES to provide automatic access to the Nasdaq Market
and market makers for the securities orders of small, retail
investors.  However, 1995 NASD data show that the primary users of
SOES, accounting for over 80 percent of SOES trading volume, are SOES
day trading firms.  These firms generally have provided access to
SOES and proprietary computer software to assist the trading of day
traders.  Day traders usually (1) trade the maximum number of shares
allowed on SOES, (2) trade frequently during the day, and (3) rarely
carry an inventory of securities overnight.  Market makers told us
that many retail investors' orders are being filled by market makers
using their own internal automated systems, at prices that are to
match the best Nasdaq prices.  The day traders have used a trading
advantage provided by the automatic execution feature of SOES,
proprietary software the SOES day trading firms designed to benefit
from this feature, and Nasdaq rules that limited market maker access
to SOES, to profit at the expense of market makers and their
customers.  This advantage has diminished over time because of market
maker software improvements and changes in trading rules. 

Not surprisingly, market makers and SOES day trading firm officials
disagree about the effects of SOES day trading on the market.  For
example, market makers told us that day traders are market
professionals, who should not have an advantage in the marketplace. 
They said that the day traders not only profit at the expense of
market makers, but also cause losses for market maker customers, who
may have to buy at higher prices and sell at lower prices than they
otherwise would.  In contrast, SOES day trading firm officials said
that day trader activity has forced market makers to monitor and
update their quotes, or prices, and has ensured that orders from
individual investors are automatically executed at the best prices. 
The various studies that have attempted to assess SOES effects on the
market have not effectively isolated the effects of SOES from those
of many other changes in the market, such as revised trading rules,
that have occurred during the same time SOES has been operating. 

NASD has proposed various rule changes to reduce or eliminate the
advantage of SOES day traders.  SEC approved some of the earlier rule
changes.  However, a court decision in favor of a SOES day trader, as
well as two separate investigations by the Department of Justice and
SEC of market maker collusion to fix prices on the Nasdaq Market,
which market makers eventually settled without admitting or denying
guilt, strengthened the position of the SOES day trading firms and
their day traders.  SEC officials told us that they had become more
skeptical of NASD rule proposals to change SOES, and that NASD had
failed in two attempts to eliminate SOES' automatic execution feature
in favor of systems that attempted to limit the availability of
automatic executions exclusively to smaller investors and that routed
orders to market makers for execution. 

Despite NASD's lack of success in controlling SOES traders through
rule changes, the volume of SOES trading as a percentage of total
Nasdaq share volume has been declining, from almost 8 percent in
January 1997 to less than 5 percent in 1998.  Some of the factors
that may have influenced this decline include changes in SEC and NASD
rules, increased use of ECNs by SOES day traders, and market makers'
use of their internal automated trading systems.  To make automatic
trade executions more widely available, Nasdaq has proposed an
Integrated Order Delivery and Execution System that would replace two
existing systems, including SOES, and provide an integrated order
routing and execution system and a voluntary limit order file, in
which customer orders at specified prices could be matched and
executed.  This system would allow market makers, as well as SOES day
traders, access to automatic execution of all trades.  Nasdaq
officials told us that this system, among other things, is intended
to eliminate the unintended double liability incurred by market
makers as a result of SOES day trading, without unduly benefiting
market makers.  This double liability arises because market makers
are required to execute trades against their quotes through both
SelectNet and SOES However, some features of the proposed system are
controversial, and SEC has received about 2,100 public comment
letters on the system. 


      SCOPE AND METHODOLOGY
---------------------------------------------------------- Letter :2.1

To determine the extent to which SOES was being used for its intended
purpose and how it has performed, we interviewed officials from NASD
and SEC and reviewed agency documents.  We also interviewed selected
market participants about their use of SOES, including officials from
seven market making firms, five of which were among the largest
Nasdaq trading firms; three SOES day trading firms; one mutual fund;
and two industry trade associations, one representing market makers
and the other SOES day trading firms. 

To determine past NASD attempts to limit SOES day trading, we
reviewed NASD proposed rule changes and SEC approval orders.  We
interviewed current and former SEC officials to obtain information on
the reasons for which changes were or were not approved.  We also
reviewed pertinent SEC reports, primarily its 21(a) report,\5

public comment letters, and other documents. 

To determine the effects of SOES on the marketplace, we interviewed
the previously listed selected market participants.  We also reviewed
and analyzed the following six studies on the market effects of SOES: 
(1) The Importance of Firm Quotes and Rapid Executions:  Evidence
From the January 1994 SOES Rule Changes by Jeffrey Harris and Paul
Schultz; (2) SOES Trading and Market Volatility by Robert H. 
Battalio, Brian Hatch, and Robert Jennings; (3) Day Trading on
Nasdaq's Automatic Small Order Execution System (SOES):  Adverse
Selection and Spread by George Benston and Robert Wood; (4) Nasdaq
Market Reform:  New Evidence That Competition From the Public Lowers
Trading Costs by Michael J.  Barclay, William G.  Christie, Jeffrey
Harris, Eugene Kandel, and Paul H.  Shultz; (5) Effects of the
Removal of Minimum Sizes for Proprietary Quotes in the Nasdaq Stock
Market by the staff of NASD Economic Research; and (6) SOES and the
Nasdaq Stock Market by Nasdaq Economic Research (Dean Furbush, D. 
Timothy McCormick, Kathleen L.  Mitchel, Lorraine M.  Reilly, James
P.  Selway III, and Michael P.  Watson).  In addition, we interviewed
other academicians about SOES, as well as the Chief Economist at NASD
and officials from SEC's Division of Market Regulation.  We also
interviewed the selected market participants about the various
studies. 

To determine the recent market developments and proposals affecting
SOES, we reviewed SEC and NASD rule changes since December 1996
concerning SOES and the latest NASD proposal to replace SOES.\6 We
interviewed the selected market participants to obtain their views on
the proposed SOES replacement system and reviewed public comment
letters. 

We did our work in accordance with generally accepted government
auditing standards between October 1997 and July 1998 in Washington,
D.C.; Baltimore, MD; New York, NY, and vicinity; and Houston, TX. 


--------------------
\5 Report Pursuant to Section 21(a) of the Securities Exchange Act of
1934 Regarding the NASD and the Nasdaq Market, U.S.  Securities and
Exchange Commission (August 8, 1996). 

\6 Release No.  34-39718; File No.  FN SR-NASD-98-17, March 4, 1998. 


   NASD ESTABLISHED SOES TO
   PROVIDE AUTOMATIC EXECUTION OF
   SMALL RETAIL ORDERS, BUT IT HAS
   BEEN USED PRIMARILY BY DAY
   TRADERS
------------------------------------------------------------ Letter :3

NASD established SOES in 1984 as a means for market makers to provide
timely and efficient execution and processing of small retail orders. 
When first established, SOES was a voluntary system for market makers
that allowed them to lower their processing costs by providing
automatic execution for these orders at the best available prices. 
After the 1987 stock market crash, during which many market makers
withdrew from SOES and small investors had difficulty executing their
orders, NASD made market maker participation in SOES mandatory for
Nasdaq National Market securities.  Since then, SOES day trading
firms have developed trading strategies to take advantage of the
automatic execution feature of SOES, and day traders have become the
primary users of SOES.  The day traders trade frequently, usually
trade the maximum number of shares allowed on SOES, and rarely carry
an inventory of securities from one day to the next.  SOES has
provided the day traders and its other users access to market makers
during the periods of high trading volume that have occurred since
the 1987 market crash. 


      NASD ESTABLISHED SOES TO
      PROVIDE AUTOMATIC EXECUTION
      OF SMALL RETAIL ORDERS
---------------------------------------------------------- Letter :3.1

Nasdaq market makers recognized the need for increased automation of
the over-the-counter market when, during a surge of trading volume
that began in the fall of 1982, the handling of routine small orders
became an increasing burden for them.  In November 1984, NASD
submitted a proposed rule change to SEC to reduce this burden.  The
submission announced that NASD had developed a new, automated, small
order execution system for routing and executing limited sized orders
in over-the-counter securities.  Its purpose was twofold:  to improve
the efficiency of these securities transactions by allowing the
orders of brokers' small retail customers to be sent electronically
to Nasdaq market makers, and to receive automatic executions at the
best bid or ask price.  The orders, initially of 500 shares or less,
were to be sent on a rotating basis to competing market makers, who
could participate on a voluntary basis.  In general, such small
orders are considered to be of a size less likely to cause a movement
in a stock's price or significantly reduce available liquidity. 
Member firms could participate in SOES as either market makers or
order entry firms.  SOES was approved by SEC in February 1985. 

The automatic execution feature of SOES is consistent with the
objectives of the 1975 Securities Acts Amendments.  These amendments
set forth five objectives for the securities markets to function in
the public interest, provide appropriate protection of investors, and
maintain fair and orderly markets.  The objectives are (1) efficient
execution of securities transactions; (2) fair competition among
market participants; (3) widespread availability of quotation and
trade information; (4) the practicability of orders executed in the
best market; and (5) the opportunity, consistent with the foregoing
provisions, for investors' orders to be executed without dealer
participation.  SOES provides for efficient execution of investors'
orders at the best bid or ask price without negotiating with a
dealer. 


      MARKET CRISIS WAS BASIS FOR
      MAKING SOES PARTICIPATION
      MANDATORY
---------------------------------------------------------- Letter :3.2

From October 14 through 21, 1987, the U.S.  stock markets experienced
an accelerated decline and record trading volumes.  The Dow Jones
Industrial Average (DJIA) fell by over 30 percent--the largest
decline since the market crash of October 1929.  The Nasdaq composite
index declined 27.2 percent during October 1987, while Nasdaq trading
volume reached the unprecedented level of 288 million shares on
October 21, 1987. 

Market makers withdrew from voluntary SOES participation during the
decline, making it necessary for trades that would have been executed
automatically by SOES to be handled by contacting a market maker by
telephone.  This added to the already high volume of telephone calls
to market makers, making them unusually difficult to contact. 
Consequently, many investors could not contact market makers and
obtain executions of trades.  The result was a severe reduction of
liquidity in the Nasdaq market, particularly for small retail
investors. 

Despite the withdrawals by some market makers, SOES remained open and
in operation throughout the October 1987 market crash and, according
to SEC, continued to provide an effective means for the execution of
small orders.  However, the difficulty experienced by small investors
in accessing the market during the break caused NASD, with SEC
encouragement, to conclude that enhancements to SOES and Nasdaq were
needed.  In March 1988, NASD proposed rule changes that SEC approved
in June 1988.  The rule changes (1) prohibited a firm that withdraws,
on an unexcused basis, as a Nasdaq market maker in a security from
reentering Nasdaq as a market maker in that security for 20 business
days; (2) made SOES participation mandatory for all market makers in
Nasdaq National Market securities; (3) enabled NASD to establish
different levels of maximum order size limits of 200, 500, and 1,000
shares for SOES orders in national market securities, depending on
the trading characteristics of the different securities; and (4)
required a market maker to execute a minimum of 5 trades at a
security's maximum order size. 

The various public commentators on the proposal to make SOES
mandatory were about equally divided between favoring and opposing
the proposal.  Some commentators believed that mandatory SOES
participation might result in a widening of spreads by market makers
to handle their increased risk from the automatic SOES executions. 
However, neither the commentators, NASD, nor SEC expected that SOES
day trading firms would develop and that day traders would become the
primary users of SOES. 


      PRIMARY USERS OF SOES ARE
      DAY TRADERS
---------------------------------------------------------- Letter :3.3

Shortly after market maker participation became mandatory, a few
firms realized that they could profit from using SOES.  NASD data
show that these SOES day trading firms and their day traders became
the primary users of SOES.  NASD officials have stated that the
characteristics of SOES day traders are distinct from those of retail
investors.  They said that SOES day traders trade at the maximum
allowable SOES transaction size, and that their trading volume is
much higher than that of a typical retail investor.  NASD data show
that SOES day traders are also usually position-neutral--having
neither a long (owning stock) nor a short (borrowing stock) position
at the end of the day.  Nasdaq reported that, as of September 1995,
day traders accounted for about 83 percent of SOES share volume, 89
percent of the dollar volume, and about 69 percent of all SOES
trades. 

SOES day trading firm officials told us that day traders have several
methods they use to buy and sell stocks.  In addition to SOES, they
may also use electronic communication networks (ECN), such as
Instinet, or other Nasdaq order delivery systems such as SelectNet.\7
These officials said that day traders used ECNs in an attempt to
negotiate prices that would offer them the highest profit when they
closed out their positions.  Until SEC instituted new trading rules
in 1997, as discussed later, the prices in ECNs were not generally
available to the investing public.  Day traders accessed these
systems through the SOES day trading firms.  These firms also
provided customized, proprietary software and various services, such
as facilities and training, to assist the day traders that
constituted their customers.  The software systems included such
features as real time quotes, news information, custom tickers that
display market maker updates, charts of stock prices, and lists of
the most active stocks. 


--------------------
\7 ECNs are generally privately operated screen-based electronic
trading systems that allow participants to enter priced orders that
then are widely disseminated to third parties.  Participants may
include retail investors; mutual funds, pension funds, and other
institutional investors; broker-dealers; and market makers.  The
sponsors of these systems are currently regulated by SEC as
broker-dealers, even though the systems may operate differently from
the activities of traditional broker-dealers.  Instinet is a private
trading system that is linked as an ECN to Nasdaq.  SOES day trading
firms, and their day traders, have access to Selectnet and Instinet
because the firms, as NASD members, are allowed access to ECNs
through their Nasdaq workstations. 


      SOES PROVIDED A TRADING
      ADVANTAGE FOR DAY TRADERS
---------------------------------------------------------- Letter :3.4

SOES day traders have had a trading advantage in the Nasdaq market
over market makers and their customers because day traders can
execute trades faster.  Day traders execute trades immediately,
through SOES, that market makers, with similar orders, would have to
manually enter and possibly negotiate through SelectNet.  Day
traders' software systems have enabled them to take advantage of
their immediate, automatic access to SOES by entering as many trades
as they wanted, quickly cancel orders or enter multiple orders with
an individual keystroke, and be notified about securities that
reached predetermined prices.  Conversely, market makers have not
been able to use the day traders' techniques on SOES because Nasdaq
rules until recently restricted market makers from entering SOES
orders on behalf of their customers for stocks in which they make a
market.  In addition, market makers still cannot use SOES for their
proprietary trades.  The ability to execute trades faster than market
makers has allowed SOES day traders to profit from short term price
movements in stocks. 

In addition, market makers told us that SOES day traders have an
advantage over them because the day traders can get executions
through SOES faster than the market makers can update their quotes. 
Specifically, market makers told us that the day traders initially
used their advantage to profit at the expense of market makers that
did not closely monitor their stocks.  The SOES day trading firms
provided day traders with software to monitor and identify market
maker quotes that were "stale"--that is, as the prices of securities
moved up and down, the market maker's quote may have been out of line
with other market maker quotes for a short period of time.  For
example, one market maker who was not closely monitoring a particular
stock may have left his bid and ask prices (the prices at which he is
willing to buy or sell stock respectively) at $19.75 and $20, at the
same time that other market makers moved their bid and ask quotes,
because of some news on the stock, to $20.25 and $20.50.  By taking
advantage of their software that allowed them to identify this
discrepancy, SOES traders could profit from the opportunity presented
because they could automatically buy at $20 and then sell at $20.25. 
However, these opportunities became less frequent as market makers
began to automate their quote monitoring. 

At the time of our review, day traders generally used software
provided by the SOES day trading firms to monitor a handful of stocks
that had numerous intraday price movements, or to monitor many
actively traded stocks, looking for trends to develop in order to buy
in "up-trending" markets and sell in "down-trending" markets.  They
tried to buy stocks when they anticipated that prices would be
increasing, selling later at higher prices; or similarly, selling
stocks when they anticipated prices would fall and buying back later
at lower prices--in either case profiting on the price momentum. 
According to market makers, one way day traders do this is to
identify when institutional traders are buying or selling large
amounts of a particular stock, and then buy or sell ahead of them. 
For example, assume an institutional trader wanted to buy 10,000
shares of a particular stock.  If 10 market makers are each quoting
prices for 1,000 shares, the trader's broker begins to purchase all
the available shares at the quoted price through SelectNet.  However,
after purchasing 1,000 shares from each of two market makers, the
broker finds that SOES day traders have purchased the rest of the
shares through SOES, using the advantage provided by the automatic
execution feature of SOES and their software that allows them to
execute many orders at the same time.  The SOES traders then sell the
8,000 shares to the broker at a higher price to the broker. 
According to market makers, this happens hundreds of times a day and
causes institutional traders, as well as retail investors, to have to
buy at higher prices and sell at lower prices because they do not
have the same speed of access to the market as SOES day traders. 

Market makers told us that they have established their own internal
automated trading systems, which they use to fill retail investors'
orders, and made software improvements in order to minimize their
exposure to day traders.  As a result of the improvements and changes
in the trading rules, which we discuss later, the trading advantage
that SOES day trading firms have had over market makers and their
customers has decreased. 


      SOES HAS PROVIDED ACCESS TO
      MARKET MAKERS DURING PERIODS
      OF HIGH TRADING VOLUME
---------------------------------------------------------- Letter :3.5

Although regulators have had only two opportunities to observe SOES
performance under periods of high trading volume, their analyses have
shown that SOES provided investors placing small orders continued
access to market makers throughout those periods.  Only two periods
of exceptionally high trading volume and volatility have occurred
since June 1988 when SOES became mandatory--in October 1989 and
October 1997. 

On October 13 and 16, 1989, the securities markets experienced
unusual price volatility and trading volume.  On the 13th, the DJIA
experienced a decline of 190 points--the second largest single-day
point decline to that time.  On the next trading day, the 16th, the
DJIA closed up 88 points, the fourth largest point gain to that
time.\8 SEC reported that market maker withdrawal from SOES
participation during this period was significantly lower than in
October 1987.  SEC attributed this to the rule changes adopted the
previous June, which made SOES participation mandatory and penalized
withdrawals with a 20-day ban before reentry.  SEC also found that
SOES operated well and succeeded in handling its share of the third
highest volume to that time on the Nasdaq Market.  SEC concluded
that, in part because SOES was able to handle the increased retail
order flow, the Nasdaq Market operated much more effectively than in
October 1987. 

On October 27, 1997, the DJIA fell 550 points, 7.2 percent, the
largest 1-day point drop in its history.  The Nasdaq composite index
fell by 7 percent, and trading volume was 906.4 million shares, well
above the 1997 average daily volume of about 647.4 million shares. 
The next day, October 28, 1997, after trading resumed, the market
rallied, and the Nasdaq index gained back much of what had been lost
the previous day.  That day, Nasdaq recorded trading volume of 1.3
billion shares--its first ever billion share day. 

NASD's study of trading on October 27 and 28, 1997, found that SOES
continued to provide investors efficient access to market makers on
both days.  SOES accounted for 6.9 percent of total Nasdaq share
volume on Monday the 27th, which was similar to its use on previous
days.  The number of SOES orders entered on the 27th was about
114,000, which was well above the daily average for the previous week
of about 83,000 orders.  On Tuesday the 28th, SOES trading accounted
for 6.4 percent of total Nasdaq share volume.  On the 28th, the
nearly 172,000 orders entered into SOES were more than twice the
number entered on the average day.  An NASD official told us that,
while the SOES proportion of total Nasdaq volume did not rise,
additional retail investor trading volume could have been executed
through market makers' internal automated systems.  Market makers
established such in-house trading systems during the years between
the 1989 and 1997 market breaks. 

SEC has not completed its analysis of trading on October 27-28, 1997. 
However, SEC officials told us that they found no problems with small
retail investors accessing the market during this 2-day period. 
According to the officials, SOES maintained its share of total
trading over the period, which generally indicates that it allowed
small orders to execute automatically against market maker quotes. 

The volume of SOES trading has increased significantly since SOES
began.  According to NASD data, when SOES was introduced the system
accounted for about 4,500 trades per day.  In 1995, SOES accounted
for over 40,000 trades a day, and by mid-October 1997, it accounted
for about 83,000 trades a day.  SOES share of the total Nasdaq market
before 1995 was generally between 1 and 4 percent of total share
volume.  In 1995, SOES share of total volume increased to over 6
percent and remained at that level until, as discussed later, it
declined after SEC and NASD rules' changes and increased use of ECNs. 


--------------------
\8 The percentage decline, 6.9 percent, was only the 20th largest
from 1900 through 1989, and the following day's percentage gain, 3.4
percent, was not among the top 20 largest gains during those years. 


   OPINIONS VARIED ABOUT THE
   EFFECTS OF SOES DAY TRADERS ON
   THE MARKET
------------------------------------------------------------ Letter :4

Opinions varied about the effect of SOES trading on the Nasdaq
Market.  Market makers and SOES day trading firms disagreed about
their respective roles in the market.  They also disagreed about the
effects of SOES day trading on spreads, liquidity, and volatility. 
Various studies did not help to resolve these disagreements because
they did not clearly and precisely isolate the effects of SOES day
trading from those of many other changes in the market, such as
revised trading rules, that have occurred since SOES has been
operating.  Further, some market participants alleged that the nature
of SOES day trading has allowed SOES to be used for price
manipulation.  Although regulators said they have found no indication
that SOES has been put to such a use, they continue to examine for
such issues. 


      MARKET MAKERS AND SOES DAY
      TRADING FIRMS DISAGREED
      ABOUT THEIR ROLES IN THE
      MARKET
---------------------------------------------------------- Letter :4.1

Market makers told us that they viewed day traders as sophisticated
professional traders who are detrimental to the market because they
profit at market makers' and others' expense.  They said that SOES
day trading not only has cost them money, but also has resulted in
their institutional customers, such as pension funds and mutual
funds, having to pay higher prices for securities because the day
traders have an advantage in the market as described previously. 

SOES day trading firm officials told us that day traders are small,
individual investors trading for their own accounts.  They said that
day trading has benefited the market by forcing market makers to
monitor and update their quotes and by increasing price competition. 
They said that, by taking advantage of "stale" market maker quotes,
SOES day traders have forced market makers to improve their quote
monitoring and revise their quotes immediately as the market changes,
thereby improving the efficiency with which prices are determined. 
Further, they said that using SOES is the least expensive way for day
traders to execute trades.  For example, the maximum charge to
execute a trade through SOES is $.50 for market makers and $.50 for
order entry firms, whereas the charge to execute a trade through
other trading systems can be as much as $15. 


      DISAGREEMENTS ABOUT SOES
      EFFECTS ON SPREADS
---------------------------------------------------------- Letter :4.2

Market makers have indicated that SOES day trading has caused widened
spreads, but SOES day trading firms disagreed.  Since SOES was made
mandatory, market makers stated their cost of doing business has
increased as a result of SOES day trading, which has caused them to
increase their spreads.\9 Market makers reported that they were
initially unable to adequately monitor all the stocks in which they
made a market, which increased the probability of stale quotes and
thus the likelihood that they could lose money to SOES day traders. 
Additionally, they reported that the automatic execution feature had
enabled SOES day traders to execute their orders before market makers
could execute their customers' orders.  Market makers have said that
day traders' strategies have increased their costs of doing business,
and that in order to cover these costs, they have widened their
spreads.  SOES day trading firm officials, however, said that day
trader SOES activity causes market makers to monitor and update their
quotes, benefiting everyone in the market.  They also said that they
often quote better prices than market makers when they do the trades
to close out their positions, which results in a narrowing of
spreads. 

Studies that examined SOES effects on spreads have focused on whether
SOES day trading actually widens spreads.  The studies recognized
that many factors, such as rule changes, affected spreads.  However,
during the periods covered by the studies, they were unable to
isolate the effects of SOES from these other factors.  One study
found that the 1997 rule changes implemented in the Nasdaq Market had
caused spreads to narrow while SOES day trading was occurring.\10
However, this study did not focus on measuring the effects of SOES. 
Another study found that adverse selection\11 attributed to SOES day
trading had contributed to wider spreads.\12 Specifically, this study
reported that market makers were, in essence, being forced to trade
with SOES traders, who may have been better informed than other
traders about which way prices would move.  This may put market
makers at a disadvantage and cause them to suffer an opportunity
loss.  Because market makers have an obligation to both buy and sell
stock, they may be buying at higher prices when they need to buy, and
selling at lower prices when they need to sell, because of SOES
traders' activities.  Their opportunity loss is the difference
between the prices at which they actually bought and sold and those
at which they could have bought and sold without the temporary price
effects of the SOES activity.  The study attempted to measure the
market makers' losses caused by this adverse selection, and concluded
that market makers increased their spreads to offset these losses. 
Measuring adverse selection is always very difficult because it
relies on measuring differences in information that are not
observable.  As a result, it is not clear that the study accurately
measured the opportunity losses resulting from adverse selection.  In
addition, the study did not isolate or control for the influences of
other factors that could have affected prices, such as rule changes. 
Because the study did not control for these other factors, the study
may not have accurately measured the effects of adverse selection. 


--------------------
\9 Market makers earn their income off the spread between their
buying and selling prices for stocks.  Market makers have said that,
to the extent that their costs rise, they raise their prices and
widen their spreads. 

\10 Michael J.  Barclay, William G.  Christie, Jeffrey H.  Harris,
Eugene Kandel and Paul H.  Schultz, Nasdaq Market Reform:  New
Evidence that Competition from the Public Lowers Trading Costs,
Working Paper 97-10, July 1, 1997. 

\11 Adverse selection is a term taken from the field of insurance
analysis.  It describes a tendency for poorer risks to seek or
continue coverage to a greater extent than do better risks.  Adverse
selection leads to higher costs for insurers and to attempts by
insurers to put restrictions on policies.  In securities markets, the
parallel is that some participants are better informed than others. 
Market makers prefer to negotiate prices as a way of determining how
well informed traders are and to reduce adverse selection costs. 

\12 George Benston and Robert A.  Wood, Day Trading on Nasdaq's
Automatic Small Order Execution System (SOES):  Adverse Selection and
Spread, Unpublished, 1997. 


      DISAGREEMENTS ABOUT SOES
      EFFECTS ON LIQUIDITY
---------------------------------------------------------- Letter :4.3

Market makers said that they have reduced the number of stocks in
which they make markets, in part because of SOES.  They said that
they did this because their costs increased due to SOES day trader
activity.  The more stocks they covered, the more likely they were to
have stale quotes, which in turn attracted SOES day traders.  They
said that the reduction has resulted in less liquidity in the market
because they are putting up less capital in total than formerly, and
because they are offering to buy or sell smaller amounts of the
stocks in which they make a market.  Furthermore, they stated that
the cost of liquidity may have risen because their retail and
institutional customers cannot purchase as much of a stock at a given
price as they could before SOES. 

SOES day trading firms said that, for the most part, day trading has
not caused a reduction in liquidity.  They stated that, because they
are purchasing and selling securities at the market maker's spread,
and doing so frequently, they are not only providing the market maker
with increased profits, but are also putting their capital at risk,
thereby providing liquidity to the market.  Market participants have
also said that other market making firms have started making markets
in those stocks that had experienced an initial reduction of market
maker activity, thereby continuing to provide liquidity to the
market. 

Some studies concluded that SOES day trading had caused market makers
to make markets in fewer stocks.  For instance, NASD looked at the
stocks whose number of market makers had declined since SOES was made
mandatory, and found that the stocks with higher levels of SOES
activity were the ones with the largest decline in market maker
participation.\13 While acknowledging that SOES day trading has
affected market maker activity in some stocks, the approaches taken
by the studies did not often take into account other factors, such as
rule changes, profitability, or market conditions, that may
contribute to fewer market makers in stocks.  Additionally, it was
difficult to determine from the studies whether market makers would
continue to make markets in fewer stocks in the long run. 


--------------------
\13 Nasdaq Economic Research (Dean Furbush, D.  Timothy McCormick,
Kathleen L.  Mitchel, Lorraine M.  Reilly, James P.  Selway III, and
Michael P.  Watson), "SOES and the Nasdaq Stock Market", Unpublished,
November 29, 1995. 


      DISAGREEMENTS ABOUT SOES
      EFFECTS ON VOLATILITY
---------------------------------------------------------- Letter :4.4

Market makers and other market participants said that SOES day
trading has caused increased volatility in the market.  They stated
that the trading strategies of SOES day traders, such as the momentum
based strategy of buying in "up trending" markets and selling in
"down trending" markets, has led to increased volatility. 
Conversely, SOES day trading firms said that volatility has been the
result of market forces, not SOES trading. 

Some of the studies examining volatility showed that SOES day trading
and volatility were correlated.  However, the studies did not clearly
establish whether SOES day trading caused volatility or whether other
market factors caused volatility, which then attracted SOES day
trading.  Some studies found that significant SOES trading occurred
on days with large price changes in stocks, and concluded that SOES
day trading causes volatility.  One study, which assessed the
behavior of quotes where clusters of maximum-sized SOES trades
occurred, found that price volatility and SOES activity were
correlated.  Another study, which attempted to determine whether
increased volatility had attracted SOES trading or whether SOES
trading caused volatility, found a correlation between instances of
high volatility and periods in which a large portion of trades were
being SOES trades.  The study concluded that large volumes of SOES
trading caused increased volatility in the short run, but contributed
to lower volatility in the long run.\14

Other studies also indicated that SOES day trading actually
concentrated the price discovery process by moving prices to their
equilibrium faster than would have happened in the absence of such
trading. 


--------------------
\14 Robert H.Battalio, Brian Hatch, and Robert Jennings, "SOES
Trading and Market Volatility", Journal of Financial and Quantitative
Analysis, Vol.  32, No.  2 (June 1997). 


      SOME MARKET PARTICIPANTS
      SAID THAT, BY ITS NATURE,
      DAY TRADING HAS RESULTED IN
      PRICE MANIPULATION
---------------------------------------------------------- Letter :4.5

Some market participants said that, by its nature, day trading has
resulted in the manipulation of some stocks' prices.  These
participants stated that, although SOES may not be the cause of price
manipulation, the nature of day trading, especially momentum-based
trading, can assist those who want to manipulate stock prices.  For
example, they said that stock manipulators can momentarily quote
prices for stocks that trigger SOES momentum trading by putting
prices into an ECN, and then cancelling the quote before a trade
occurs.  They have pointed to specific stocks that have incurred
certain "spikes" in their prices over short periods of time
concurrent with periods of large volumes of SOES day trading.  They
have further stated that no logical reason existed for the rapid
increases in the prices of the stocks other than the likelihood that
the stock prices were being manipulated. 

Both NASD and SEC officials that we talked to said that they have
investigated the allegations of price manipulation and have not found
that SOES was being used for such purposes, but that their
investigations are continuing.  NASD officials said that, in most of
the cases evaluated thus far, the increase in SOES volume was
attributable to investors' natural reactions to news and other
information in the marketplace that would affect price levels.  They
further stated that, while SOES volume and price volatility were
correlated in certain stocks, the correlation does not support
allegations that the stocks were manipulated through SOES.  We plan
to review this issue in more detail during future work. 


   ATTEMPTS TO LIMIT SOES DAY
   TRADING OR REPLACE SOES
------------------------------------------------------------ Letter :5

After NASD and Nasdaq market makers became concerned about day
traders use of SOES, NASD proposed various SOES rule changes to limit
the activity of the day traders.  SEC approved these rule changes,
but SOES day traders contested one of these changes in federal court. 
The court's decision and the outcomes of two federal investigations
of the competitiveness of the Nasdaq Market served to bolster the
case for automated trade execution in that market.  SEC did not act
on two later NASD attempts to replace SOES with other systems that
did not feature immediate automatic execution of small orders because
of (1) technical concerns about the systems' operations, (2) market
participant opposition, and (3) lack of provisions for the execution
of small orders without market maker intervention.  A chronology of
SOES and SOES-related events can be found in appendix I. 


      NASD CHANGED ITS RULES TO
      LIMIT SOES DAY TRADING
---------------------------------------------------------- Letter :5.1

Not long after SOES became mandatory, market makers complained to
NASD that SOES day trading firms were using SOES to execute orders
for day traders, which was not its purpose.  In response, NASD
proposed several rule changes to limit SOES day trading activity. 
These changes are summarized below and discussed in more detail in
appendix I. 

  -- An August 1988 rule change prohibited SOES users from "order
     splitting"--that is, dividing orders they received that were
     larger than the SOES maximum order size into smaller parts so
     that they could be entered into SOES. 

  -- A December 1988 rule change prohibited entering orders into SOES
     on behalf of a professional trading account (PTA).  A PTA was
     defined to include any account in which five or more day trades
     had been executed through SOES during any trading day or where a
     professional trading pattern in SOES was demonstrated.  The rule
     defined "day trade" or "day trading" to mean the execution of
     offsetting trades in the same security for generally the same
     size during the same trading day. 

  -- An October 1991 rule change set criteria for NASD to designate a
     trading account a PTA.  The criteria included excessive
     frequency of short-term trading, excessive frequency of
     short-sale transactions, existence of discretion, and direct or
     physical access to SOES execution capability.\15 The rule change
     expanded the definition of a "day trade" by eliminating the
     restriction that both sides of a trade must be executed through
     SOES to be considered a day trade. 

  -- A second October 1991 rule change allowed market makers a
     15-second delay to update a quotation before being obliged to
     execute a second similar order in the same security through
     SOES.  This rule change also allowed market makers to decline
     preferencing by SOES order entry firms.\16

  -- The December 1993 rule changes, which are known as "Interim SOES
     Rules," further attempted to restrict day traders' use of
     SOES.\17

Approved on a 1-year pilot basis, the rule changes (1) reduced the
maximum size order eligible for SOES execution from 1,000 shares to
500 shares, (2) reduced the minimum exposure limit from 5 times to 2
times the maximum order size, (3) allowed market makers to
voluntarily use a new Nasdaq system feature that automatically
updates quotations once the market makers' exposure limit has been
exhausted, and (4) prohibited short-sale transactions through SOES. 

SEC approved all these rule changes.  As a former SEC official told
us, SEC's sympathy toward the market makers' position was
understandable in view of the fact that SEC encouraged NASD,
following the 1987 market crash, to make SOES participation mandatory
for market makers.  SEC officials told us that SEC approved NASD's
rule filings until SEC's investigation of NASD showed NASD's abusive
behavior toward day traders and the illegal pricing convention among
Nasdaq market makers that SEC reported in its 21(a) report. 


--------------------
\15 A short sale is the sale of a security that is not owned by the
seller to take advantage of an anticipated decline in the price of
the security. 

\16 SOES allows preferencing of orders in which the orders are
designated to be sent to a particular market maker. 

\17 The rules were considered interim in anticipation of NASD's
designing a system to replace SOES. 


      OTHER EVENTS FURTHER
      BOLSTERED SOES DAY TRADING
---------------------------------------------------------- Letter :5.2

While NASD and SEC were attempting to limit the activities of SOES
day traders, a series of events occurred during the period 1993
through 1996 that helped further legitimize and bolster the position
of SOES day traders in the market.  A lawsuit filed against SEC by
SOES day traders contesting its approval of the PTA definition, a
Department of Justice investigation of the practices of certain
Nasdaq market makers and Nasdaq's market structure, and SEC's own
investigation of NASD and the Nasdaq Market--together caused SEC to
become more skeptical of Nasdaq market makers' views and increased
SEC scrutiny of NASD rule proposals affecting SOES. 


         COURT DECISION FAVORED
         DAY TRADERS
-------------------------------------------------------- Letter :5.2.1

In 1991, shortly after SEC's approval of NASD's rule expanding the
definition of a PTA, William Timpinaro, a SOES trader, and others
filed a petition with the U.S.  Court of Appeals for the District of
Columbia Circuit.  The lawsuit sought the court's review of SEC's
rule approval decision.\18

In August 1993, the court remanded the case to SEC to provide further
support for its assertions that failure to restrict professional
traders from SOES would cause market makers to cease making markets
or widen spreads.  In addition, the court found the definition of a
PTA unconstitutionally vague and subject, in part, to open
interpretation, and therefore asked SEC to address the vagueness of
the definition. 

In response to the decision, NASD withdrew the PTA rules.  According
to SEC officials, the Timpinaro case showed that SEC decisions on
SOES rule changes were likely to be litigated, which caused SEC to be
more cautious about the actions it considered to resolve problems
surrounding SOES. 


--------------------
\18 Timpinaro v.  SEC, 2 F.3d 453 (D.C.  Cir.  1993). 


         ANTITRUST INVESTIGATION
         FOUND MARKET MAKERS'
         CONDUCT ANTICOMPETITIVE
-------------------------------------------------------- Letter :5.2.2

In October 1994, the Department of Justice's (DOJ) Antitrust Division
started a broad review of Nasdaq's market structure.  DOJ started its
review following (1) reports in the media about an academic study on
spreads on the Nasdaq Market and related market maker behavior\19 and
(2) a class action antitrust complaint filed by Nasdaq investors
against various Nasdaq market makers.\20

DOJ's review resulted in a July 17, 1996, civil action.  DOJ alleged
that the market makers adhered to and enforced a pricing convention
that was designed to and did deter price competition among the market
makers in their trading of stocks with the public.  It also stated
that the illegal pricing convention resulted in investors incurring
higher transactions costs than if the market makers had not acted to
restrain competition. 

On that same day, the defendants and DOJ filed a proposed order to
settle the complaint and resolve the allegations.  The order sought
to eliminate the anticompetitive conduct and establish procedures to
ensure that such conduct did not recur.  Specifically, the market
makers were to refrain from adhering to the pricing convention
regarding their actions to fix, raise, lower, or maintain prices or
quotes for Nasdaq securities.  The order also required each defendant
firm to adopt an antitrust compliance program and designate an
antitrust compliance officer.  The officer is required to monitor and
tape record certain telephone conversations between stock traders and
report any instances of noncompliance with the order to DOJ. 
Although approved by the District Court, the settlement has been
stayed pending appeal.  On August 6, 1998, the appeals court affirmed
the district court's approval of the settlement. 


--------------------
\19 The study was actually published in December 1994.  William G. 
Christie and Paul H.  Schultz, "Why Do Nasdaq Market Makers Avoid
Odd-Eighth Quotes?" Journal of Finance, 49 (Dec.  1994), pp. 
1813-1840. 

\20 In December 1997, the parties reached a proposed settlement,
which the court has yet to approve.  The 30 market makers involved
agreed to pay a total settlement of $1.01 billion. 


         SEC'S INVESTIGATION
         REVEALED MARKET MAKER
         ABUSES
-------------------------------------------------------- Letter :5.2.3

In August 1996, SEC settled its investigation with NASD.  In the
21(a) report on the investigation, SEC found that Nasdaq market
makers adhered to a pricing convention that often increased
transaction costs paid by customers.  The pricing convention
particularly affected small orders, such as those transacted over
SOES.  SEC also found failure to honor price quotations and trade
reporting violations among Nasdaq market makers.  NASD was also found
lax in prosecuting rule violations by market makers, while it focused
enforcement efforts on the activities of SOES day trading firms. 

According to SEC officials, the investigation caused SEC to change
its views regarding SOES.  As noted in SEC's report, the evidence of
the pricing convention and other anticompetitive behavior of market
makers provided SEC ample reason to doubt that SOES traders were to
blame for the width of the spreads in the Nasdaq Market.  In settling
the matter, NASD consented to SEC's findings and sanctions and agreed
to undertake actions to eliminate any pricing conventions and detect
and punish anticompetitive behavior. 


      ATTEMPTS TO REPLACE SOES
      HAVE FAILED TO WIN SEC
      APPROVAL
---------------------------------------------------------- Letter :5.3

NASD has twice tried to replace SOES with alternative systems.  In
March 1994, NASD proposed a system called N*Prove.  N*Prove was
designed to replace SOES, and its immediate automatic trade
executions, with a system that would give market makers 15 seconds to
decline an incoming order before the order would be executed. 
N*Prove included a limit order file that provided the opportunity for
some customer orders to interact with each other without market maker
intervention.  SEC was concerned that such a system could result in a
queuing of orders, insufficient opportunities for customer
interaction without the intervention of a market maker, and
inadequate enforcement of NASD's firm quote rule.  NASD withdrew the
N*Prove proposal without formal action by SEC. 

Late in 1995, NASD proposed to replace SOES with NAqcess.  Like
N*Prove, NAqcess was to be an order delivery system and provided a
limit order file, but it was modified to resolve problems posed by
N*Prove.  NAqcess would provide for automatic execution of a
customer's order after it was displayed for 20 seconds, and if the
market maker did not decline the order.  To lessen the queuing
problem associated with N*Prove, NAqcess would have routed orders to
market makers with the best quotes in priority order, at 1-second
intervals.  Orders declined would be re-presented to available market
makers. 

SEC received more than 1,000 comment letters, and NAqcess was never
approved.  SEC officials told us that many commentators, including
institutional investors, opposed NAqcess.  Day traders opposed
NAqcess because it would have eliminated the SOES immediate automated
execution feature and given market makers the opportunity to decline
trades.  The commentators noted the anticompetitive behavior and
backing-away violations of market makers revealed by SEC's 21(a)
investigation.  SEC officials said that SEC asked NASD to address the
commentators' concerns.  They said NASD could not resolve the
concerns and withdrew the proposal. 


   RECENT MARKET DEVELOPMENTS AND
   PROPOSALS AFFECTING SOES
------------------------------------------------------------ Letter :6

In January 1997, NASD implemented SEC's order handling rules and
began an actual size pilot program.  On July 15, 1998, SEC approved a
rule allowing market makers to quote actual size for all stocks. 
Also, in June 1997, NASD implemented a rule change that decreased the
minimum quotation increment allowed from 1/8th to 1/16th of a dollar. 
Analyses by NASD and others show that these changes have decreased
spreads and have been associated with a reduction in SOES' share of
Nasdaq trading volume.  In March 1998, NASD proposed replacing SOES
with a new system that would make automatic trade executions more
widely available to market participants and that would limit the
unintended double execution liability incurred by market makers. 
While most market participants agree with this proposal, some of its
provisions are controversial. 


      NEW RULES HAVE LESSENED SOES
      ACTIVITY
---------------------------------------------------------- Letter :6.1

SEC's order handling rules became effective first for actively
traded, then later for all, Nasdaq securities.  The order handling
rules required Nasdaq market makers to display customer limit orders
and to disseminate the best prices placed by market makers in ECNs,
which previously were not included in the Nasdaq Market.  SEC
intended these rules to make the Nasdaq Market a more competitive,
customer order-driven market and thus reduce bid-ask spreads. 

A follow-up NASD study showed that spreads for the first 150 stocks
that became subject to the rules narrowed by about 33 percent.\21

Another study found that, for the first 100 stocks phased in between
January 20, 1997, and February 10, 1997, spreads narrowed by over 35
percent.\22 The study also showed that the proportion of share volume
traded through SOES fell by 6 percentage points for the first 50
stocks subject to the rule changes.  As spreads become narrower, the
ability of SOES day traders to earn a profit on a trade of a given
size, such as 1,000 shares, may be lessened. 

Starting in January 1997, NASD also implemented an actual size pilot
program.  This program allowed market makers to display quotes in
minimum sizes of 100 shares for 50 stocks, then later for another 100
stocks.  NASD indicated that allowing actual size quotes in
conjunction with the requirement to display limit orders would narrow
spreads.  NASD's initial, June 1997, study showed a reduction in the
average spread of about 33 percent when the same stocks were compared
before and after the pilot began.  However, the reduction in spreads
was about the same for both pilot and nonpilot stocks, indicating
that the pilot had no significant effect on spreads.  A second study
of 103 stocks brought into the pilot in November 1997 also showed the
same result.  As discussed previously, on July 15, 1998, SEC approved
a rule to permanently allow market makers to display actual size
quotations for all Nasdaq securities. 

Data in the NASD studies indicated that actual size quotations
reduced SOES market share of Nasdaq volume.  Data in the first study
showed that SOES market share declined by only about 2.9 percent for
nonpilot stocks, but declined by about 17.3 percent for pilot stocks. 
Data in the second study showed SOES market share to have declined by
19.5 percent for nonpilot stocks and 26.7 percent for pilot stocks. 
However, NASD further analyzed these data to control for differences
in price, volume, and interday volatility, and concluded that actual
size quotations had little effect on SOES market share. 

In May 1997, SEC approved a NASD proposed rule, which became
operational early in June.  The rule decreased the minimum quotation
increment from 1/8th to 1/16th of a dollar for securities with a bid
price greater than or equal to $10.  The change was intended, among
other things, to provide investors with a greater opportunity to
receive better prices when they buy or sell stocks.  NASD's
preliminary analysis showed that the decrease in the quotation
increment resulted in a decrease in the quoted spread for stocks
subject to SEC's order handling rules. 

Before these changes were made, in January 1997, SOES trading
accounted for 12.9 percent of the total dollar volume of Nasdaq
trading and 7.7 percent of total share volume.  About 1 year later,
in February 1998, SOES trading accounted for 6.8 percent of total
dollar volume and 4.6 percent of total share volume.  According to a
NASD official, part of this reduction could also be attributed to
increased use of ECNs; market makers' internal automated trading
systems; and software improvements for SelectNet, Nasdaq's automated
order entry and delivery system.  Officials of SOES day trading firms
told us that they have developed ECNs to allow their day traders, and
other investors, more direct access to the market than that provided
by SOES. 


--------------------
\21 NASD Economic Research Department, Effects of the Removal of
Minimum Sizes for Proprietary Quotes in The Nasdaq Stock Market,
Inc., unpublished, June 4, 1997. 

\22 Michael J.  Barclay, William G.  Christie, Jeffrey H.  Harris,
Eugene Kandel, and Paul H.  Schultz, The Costs of Trading Nasdaq
Issues:  The Impact of Limit Orders and ECN Quotes, The Charles A. 
Dice Center for Research in Financial Economics, Fisher College of
Business, The Ohio State University (July 1, 1997). 


      PROPOSED SYSTEM WOULD MAKE
      AUTOMATIC EXECUTION
      AVAILABLE TO ALL TRADERS
---------------------------------------------------------- Letter :6.2

In March 1998, SEC published for comment a NASD proposal to establish
a new Integrated Order Delivery and Execution System.  The system
would replace SOES and SelectNet and would provide an integrated
order routing and execution system and a voluntary limit order file. 
The proposed system would provide investors and traders, including
market makers, immediate and automatic executions for their orders in
strict time priority (on a first-come, first-served basis).  As
proposed, with the approval of the actual size rule, the new system
permits market participants to enter orders up to 999,999 shares. 

The proposed system would allow market makers to turn the system off
momentarily to prevent having trades executed against them while they
are handling telephone orders.  Orders of greater than 1,000, but
less than 5,000, shares are to be presented to market makers or ECNs
for 17 seconds, with the options of accepting, declining, or doing
nothing.  If nothing is done, the order will default to an execution
at the displayed quote amount.  Orders of 5,000 shares or more are to
be presented to market makers for review for 32 seconds and then, if
not acted on, will also default to an execution.  Market makers and
ECNs would only have to honor one execution at the quoted size and
then, depending on the order size, would have 17 or 32 seconds to
update quotes.  The proposed system, with the approval of the actual
size rule, would allow proprietary orders of any NASD member, both
market makers and non-market makers, to be entered. 

SEC has received about 2,100 comment letters on this proposal.  The
commentators generally have favored adoption of the new system. 
However, commentators representing SOES day trading firms and day
traders have opposed the feature that allowed market makers to shut
off the system while taking telephone orders.  They also opposed the
17- and 32- second delays in which market makers could choose to
decline orders of more than 1,000 and 5,000 shares, respectively. 
They said that the shutoff feature and time delays were inconsistent
with NASD's Firm Quote Rule.  Other commentators, representing
investment bankers and broker-dealers, opposed the system's limit
order file.  They said that the limit order file, which would allow
orders to be executed without market maker intervention, would
provide NASD an unfair advantage over its member firms that also
provide limit order services.\23 Finally, some large market makers
and Nasdaq-listed companies favored adoption of the system without
any changes. 

As published, the system proposal does not contain any analytical
justification of the share sizes at which the time delays would
become operational.  Time delays allow market makers to delay the
execution of trades, which reduces market efficiency.  However, time
delays may be appropriate to protect market makers from unlimited
automatic trade executions at a given price.  If time delays are
appropriate, the actual experience of market makers regarding share
sizes that are typically negotiated may provide a basis for
determining the share size limits.  For example, the six market
makers we interviewed about their operations told us that they had
in-house automated execution systems.  One market maker said that 95
percent of its trades are executed using its internal automated
system.  Another said that its customers can obtain automatic
executions through its internal automated system for orders of up to
2,000 shares. 


--------------------
\23 The limit order file would result in Nasdaq's competing with
market makers to execute these orders. 


   CONCLUSIONS
------------------------------------------------------------ Letter :7

The use of SOES primarily by day traders was an unintended
consequence of mandatory market maker participation in SOES.  Small
retail trades, which were expected to be done through SOES, instead
are being done to a large extent through market makers' internal
automated systems developed after SOES day traders became active. 
These systems have provided market makers' small retail customers the
same kind of immediate, automatic trade executions available through
SOES, even during periods of unusually high trading volume and
volatility such as occurred during October 1989 and October 1997. 
However, they are not mandatory, and no major market disruptions on
the order of October 1987 have occurred during this period to test
whether market makers' internal systems would continue to provide
immediate, automatic executions during such a crisis. 

Despite its unintended consequence, SOES has benefited the market by
providing efficient execution of securities transactions at the best
bid and ask prices and allowing investor orders to be executed
without manual market maker intervention.  These results are
consistent with national market system goals.  In addition, the SOES
automatic execution feature provides market discipline for market
makers and ensures access to the market for small investors in times
of market stress.  We did not come to any definitive conclusions
about SOES day trading's effects on spreads, liquidity, or price
volatility in the Nasdaq Market because the various studies and our
discussions with market makers and SOES day trading firm officials
could not separate the effects of SOES trading from those of other
market changes. 

SOES day traders' ability to access Nasdaq quotes and trade faster
through SOES than market makers and their customers could trade
through other systems has provided them an advantage that they have
used to profit at the expense of market makers and their customers. 
Allowing one market segment such a continuing advantage over others
is contrary to the goal of fair competition among market
participants.  However, the day traders' advantage from their use of
SOES has become less of a factor in the market over time, as market
makers have improved systems to address the effects of automated
trading and regulators have adjusted market rules.  SOES day trading
may be further affected by (1) the 1997 order handling rules that
have provided all market participants, including day traders, more
direct access to the best available prices through ECNs than is
available through SOES, and (2) the 1998 actual size rule that
reduces market maker exposure to day traders if the market makers
choose to reduce their quotation sizes. 

Nasdaq, SEC, and affected market participants have been unable to
agree on a new system to replace SOES that would retain the benefits
of automatic execution.  The latest Nasdaq proposal to replace SOES
and SelectNet with an integrated system is intended to do both. 
However, it could establish delays for the execution of some trades
that may hinder market efficiency.  Such delays could also disrupt
the market discipline provided by the immediate, automatic execution
of retail orders.  Determining a share size at which time delays may
be appropriate for market makers could be done by analyzing their
experience regarding share sizes that are typically negotiated, and
by better documenting the size of orders most often submitted by
retail investors. 


   RECOMMENDATIONS
------------------------------------------------------------ Letter :8

We recommend that the Chairman, SEC, ensure that any trading system
approved to replace the current Nasdaq trading systems be designed to
correct the trading advantage available to SOES day traders and
provide for the immediate, automatic execution of investors' small
orders at the best possible prices.  We also recommend that the
Chairman, SEC, require Nasdaq to provide adequate data analysis to
support aspects of the system that may affect market efficiency and
discipline.  For example, such analysis should be used to determine
the appropriate share size at which time delays would be allowed in
the proposed Integrated Order Delivery and Execution System. 


   AGENCY COMMENTS AND OUR
   EVALUATION
------------------------------------------------------------ Letter :9

NASD and SEC provided written comments on a draft of this report. 
(See apps.  II and III.) Both NASD and SEC recognized that the issues
the report addresses are complex and that the parties involved have a
vested interest and often widely divergent views on how SOES should
operate.  NASD commented that the report provides a balanced and
accurate description of the issues.  While SEC generally agreed with
the report's conclusions and recommendations, it took exception to
our conclusion that SOES day traders currently enjoy an advantage
over other market participants, which SEC referred to as an "unfair
advantage".  We continue to believe that SOES day trader's ability to
access Nasdaq quotes and trade faster through SOES than market makers
and their customers could trade through other systems has provided
day traders an advantage that they have used to profit at the expense
of market makers and their customers.  We also found that while this
advantage exists, it has diminished over time and become less of a
factor in the market as market makers have improved systems to
address the effects of automated trading and regulators have adjusted
market rules. 

Separately, we made technical corrections to the draft as suggested
by both agencies. 


---------------------------------------------------------- Letter :9.1

We will provide copies of this report to the Chairman and Ranking
Minority Members of the House Commerce Committee, SEC, NASD, and
other interested committees and organizations.  Copies will be made
available to others on request. 

Major contributors to this report are listed in appendix IV.  Please
call me on (202) 512-8678 if you or your staff have any questions
about the report. 

Richard J.  Hillman
Associate Director, Financial Institutions
 and Markets Issues


CHRONOLOGY OF SOES AND
SOES-RELATED EVENTS
=========================================================== Appendix I


      FEBRUARY 1985
------------------------------------------------------- Appendix I:0.1

SEC approves proposed NASD rule change for implementation of SOES. 
NASD created SOES to provide an economical and efficient system to
route and execute small retail agency orders, initially 500 shares,
of Nasdaq securities.  Use of SOES by NASD members was to be
voluntary. 


      JUNE 1988
------------------------------------------------------- Appendix I:0.2

SEC approves NASD proposed rule change to make SOES participation
mandatory for all market makers in Nasdaq national market securities. 
The change was intended to ensure investors access to the Nasdaq
market after investors experienced problems in attempting to access
the market during the high volume trading of the week of October 19,
1987.  (We found that market makers had withdrawn from SOES and
withdrew from making markets.\24 ) The change was expected to
facilitate the automatic execution of customers' small orders without
the need for telephone contact between order entry and executing
firms.  The proposal established maximum order size limits of 200,
500, and 1,000 shares, depending on the trading characteristics of
different securities. 


--------------------
\24 Financial Markets:  Preliminary Observations on the October 1987
Crash, (GAO/GGD-88-38, Jan.  26,1988). 


      AUGUST 1988
------------------------------------------------------- Appendix I:0.3

SOES rules are amended so that agency orders no larger than the
maximum order size received from public customers may be entered by a
SOES order entry firm and orders in excess of the maximum size may
not be divided into smaller parts to meet the size requirements
("order splitting").  The change was made in response to market maker
complaints of rapid-fire executions against their quotes by SOES day
traders. 


      DECEMBER 1988
------------------------------------------------------- Appendix I:0.4

SEC approves NASD SOES rules changes to prohibit members from
entering SOES orders on behalf of Professional Trading Accounts
(PTA).  A PTA was defined to include any account in which five or
more day trades had been executed through SOES during any trading day
or where a professional trading pattern in SOES is demonstrated.  The
rule defined "day trade" or "day trading" to mean the execution of
offsetting trades in the same security for generally the same size
during the same trading day.  Professional day traders were using
SOES automatic executions to take advantage of slight disparities in
market maker quotes arising from market makers' being slow to update
quotes in response to market news.  SEC favored the rule change
because it was concerned that the SOES participation requirements,
combined with the presence of professional day traders in the market,
posed risks that would cause market makers to eliminate their market
maker positions in stocks, thereby reducing liquidity in the market. 


      AUGUST 1990
------------------------------------------------------- Appendix I:0.5

SEC approves SOES rules change to prohibit market makers from
entering agency orders into SOES in securities in which they make
markets.  This change was intended to prevent market makers from
executing customer orders they have received against their
competitors, using SOES. 


      NOVEMBER 1990
------------------------------------------------------- Appendix I:0.6

SEC approves SOES rules change to provide SOES capability for
entering and storing limit orders. 


      OCTOBER 1991
------------------------------------------------------- Appendix I:0.7

SEC approves SOES rules change to allow market makers to (1) use a
15-second delay to update quotes after receiving a SOES execution and
(2) indicate order entry firms from which they will accept
preferenced orders. 

SEC approves SOES rules change to expand the definition of a PTA by
eliminating a requirement that both sides of a trade must be executed
through SOES for it to be considered a day trade.  The change set
forth criteria for NASD to designate a trading account a PTA,
including excessive frequency of short-term trading, excessive
frequency of short-sale transactions, existence of discretion, and
direct or physical access to SOES execution capability.  The rule
change also expanded the definition of a "day trade" by eliminating
the requirement that both sides of a trade must be executed through
SOES for it to be considered a day trade.  NASD believed these
changes were necessary to curtail professional day traders' use of
SOES to "pick off" market makers with quote discrepancies.  NASD
regarded the day trading as an abuse of SOES, for which the system
was not created.  NASD and certain members thought that SOES day
trading exacerbated market volatility, created wider spreads, and
resulted in a loss of liquidity for individual and institutional
investor orders.  SEC expressed concern that the presence of day
traders could result in a widespread reduction in market making and
have a significant negative impact on market liquidity. 

William Timpinaro et al.  file a petition for review of the SEC
approval orders with the U.S.  Court of Appeals for the District of
Columbia District. 


      APRIL 1993
------------------------------------------------------- Appendix I:0.8

SEC responds to questions submitted to it by the Court of Appeals
concerning the interrelationship of the Firm Quote Rule with the PTA
and 15-Second Rules. 


      NOVEMBER 1993
------------------------------------------------------- Appendix I:0.9

The Court of Appeals remands the PTA Rules back to SEC for it to
provide further explanations of the professional trader rule and to
consider whether the rules were unacceptably vague. 


      DECEMBER 1993
------------------------------------------------------ Appendix I:0.10

NASD withdraws the PTA rules because they did not prove effective in
limiting the use of SOES, thus eliminating restrictions enacted by
those rules and the associated definitions of day trading and day
traders. 

SEC approves rule change, called "Interim SOES Rules," to allow a
1-year pilot that provided for (1) a reduction in the SOES maximum
order size from 1,000 to 500 shares, (2) a reduction in the exposure
limit from 5 times to 2 times the maximum order size, (3) an
automated function to update market makers quotations once the
exposure limit is exhausted, and (4) prohibition of short sales
through SOES.  The changes were expected to reduce the effects that
active trading through SOES had on market maker risk and SOES
participation.  NASD stated that it proposed these rules to take the
place of the withdrawn PTA rules in an attempt to "limit the types of
transactions that may be executed through SOES and thereby limit the
use of SOES by those who use it in a manner inconsistent with the
intended use of SOES." SEC believed that SOES day trading imposed
increased risks and costs on Nasdaq market makers, which caused them
to widen spreads and might cause them to withdraw from making
markets.  SEC further reasoned that these costs outweighed any
momentary increases in liquidity and marginal increases in price
discovery that day trading contributed to the market.  Since December
1993, items (1) and (3) of the Interim Rules have been periodically
extended pending the time when a system to replace SOES is adopted. 
Item (2) was extended through January 1997 when the exposure limit
went down one times the maximum order size.  Item (4) was extended
until February 1995. 


      MARCH 1994
------------------------------------------------------ Appendix I:0.11

NASD files with SEC a proposal for a system called N*Prove.  N*Prove
was designed to replace the SOES immediate automatic execution system
with an order delivery system that would have given Nasdaq market
makers 15 seconds to decline incoming small orders, rather than have
the orders automatically executed against them.  The N*Prove proposal
also included a limit order file that would have provided the
opportunity for some customer orders to interact with each other. 
N*Prove was to route orders to market makers on a rotating basis,
depending on their priority ranking in the Nasdaq quote montage. 


      OCTOBER 1994
------------------------------------------------------ Appendix I:0.12

DOJ Antitrust Division announces that it is undertaking a broad
review of a number of aspects of Nasdaq's market structure. 


      DECEMBER 1994
------------------------------------------------------ Appendix I:0.13

NASD responds to an SEC staff request to provide an assessment of the
potential for order queues to develop under a system proposed to
replace SOES, called N*Prove.  N*Prove would route orders, on a
rotating and priority basis, to individual Nasdaq market makers for
automated execution, unless the market maker declined the order. 


      FEBRUARY 1995
------------------------------------------------------ Appendix I:0.14

SEC approves extension of the Interim SOES Rules through March 27,
1995, except that the prohibition on short sales was not extended
because of the acceptance of a broader Nasdaq short-sale rule. 


      MARCH 1995
------------------------------------------------------ Appendix I:0.15

SEC approves extension of Interim SOES Rules through October 2, 1995,
except that the change in the maximum order from 1,000 to 500 shares
was not extended because the available data demonstrated no
significant improvement or detriment at either share amount. 


      OCTOBER 1995
------------------------------------------------------ Appendix I:0.16

SEC approves extension of Interim SOES Rules through January 31,
1996. 


      DECEMBER 1995
------------------------------------------------------ Appendix I:0.17

SEC publishes, to solicit comments, a proposed NASD rule change that
would replace SOES with NAqcess.  NAqcess would provide for automatic
execution of customers' orders after they are displayed for 20
seconds and the market maker does not decline the order.  Orders
could only be declined if another order was just executed and the
market maker is updating quotations.  NAqcess also provided for a
limit order file against which order would be automatically executed. 
According to SEC officials, SEC received over 1,000 comment letters,
and NASD withdrew the NAqcess proposal. 


      FEBRUARY 1996
------------------------------------------------------ Appendix I:0.18

SEC approves extension of Interim SOES Rules through July 31, 1996. 


      JULY 1996
------------------------------------------------------ Appendix I:0.19

SEC approves extension of Interim SOES Rules through July 31, 1997. 

DOJ files a civil action complaint re:  United States v.  Alex Brown
& Sons, et al.  (S.D.N.Y.) pursuant to Section 4 of the Sherman Act,
as amended, 15 U.S.C.  section 4 seeking equitable and other relief
to prevent and restrain violations.  DOJ alleged that the defendants
and others adhered to and enforced a "quoting convention" that was
designed to, and did, deter price competition. 

Defendants of United States v.  Alex Brown & Sons, et al.  and DOJ
file a proposed Stipulation and Order to resolve the allegations of
the DOJ complaint.  The order would eliminate the anticompetitive
conduct and establish procedures to ensure that such conduct did not
recur. 


      AUGUST 1996
------------------------------------------------------ Appendix I:0.20

SEC completes its investigation of the Nasdaq Market.  In the
investigation report, the 21(a) report, SEC found that Nasdaq Market
makers adhered to a pricing convention that often increased
transaction costs paid by customers.  The pricing convention
particularly affected small orders, such as those transacted over
SOES, which are executed at the best available bid and ask prices. 
SEC also found that NASD had harassed SOES day trading firms. 
Further, it found that NASD had been lax in its enforcement of
extensive trade reporting and failure to honor quotation (backing
away) violations by Nasdaq market makers.  In its settlement, NASD
agreed to take actions to eliminate any pricing conventions and
detect and punish anticompetitive behavior. 


      JANUARY 1997
------------------------------------------------------ Appendix I:0.21

SEC approves extension of Interim SOES Rules through July 31, 1997. 

SEC approves SOES rule change that would establish a 5-second grace
period between SOES executions in locked and crossed markets.  A
locked market occurs when one market maker's quoted bid price for a
security equals another market maker's ask price.  A crossed market
occurs when one market maker's quoted bid price is greater than
another market maker's ask price.  The change was intended to
continue to provide incentive to market makers to update quotes and
enhance market makers' ability to react to SOES transactions in
locked and crossed markets. 

SEC's order handling rules become effective for actively traded, then
all, Nasdaq securities.  The order handling rules required Nasdaq
market makers to display customer limit orders and to disseminate the
best prices placed by market makers in ECNs.  A follow-up NASD study
showed that spreads had narrowed by 33 percent.  Another study by
economists Barclay, Christie, Harris, Kandel, and Schultz found that,
for the affected stocks, spreads narrowed by over 35 percent, and the
proportion of share volume traded through SOES fell by 6 percentage
points, as a result of the rule changes. 

SOES rule changes become effective with the order handling rules. 
Notable among the changes are (1) a market maker's displayed
quotation size is to constitute the minimum exposure limit that the
market maker is obligated to execute and (2) market makers are
allowed to enter all customer orders into SOES, including those for
stocks in which they make a market. 

NASD implements an actual size pilot program that allowed market
makers to display quotes in minimum sizes of 100 shares for 50 stocks
(then later another 100 stocks) being phased in under the order
handling rules.  NASD stated that allowing actual size quotes in
conjunction with the requirement to display limit orders would narrow
spreads. 


      APRIL 1997
------------------------------------------------------ Appendix I:0.22

The federal district court granted the Stipulation and Order filed in
United States v.  Alex Brown & Sons et al., 963 F.  Supp.  235
(S.D.N.Y.  1997). 


      MAY 1997
------------------------------------------------------ Appendix I:0.23

SEC approves a NASD rule change to decrease the minimum quotation
increment for securities with a bid price greater than or equal to
$10 that are listed and traded on the Nasdaq Stock Market to 1/16th
of $1.00.  The change was intended to enhance the transparency of the
Nasdaq Market, provide investors with a greater opportunity to
receive better execution prices, facilitate greater quote
competition, promote the price discovery process, contribute to
narrower spreads, and enhance the capital formation process.  NASD's
preliminary analysis showed that the decrease in the quotation
increment resulted in a decrease in quoted spread for stocks subject
to SEC's order handling rules. 


      DECEMBER 1997
------------------------------------------------------ Appendix I:0.24

SEC approves a NASD rule change regarding (1)excused withdrawals of
market makers from participation as market makers in Nasdaq national
market securities and (2)reinstatements of market makers who fail to
reenter quotes after their quote sizes have been decremented through
SOES trading.  The rule change listed specific factors for which
market makers can be excused or reinstated. 

SEC approves a NASD rule change to indicate that, once SOES executes
an unpreferenced market order or marketable limit order against a
SOES market maker, that market maker is not required to execute
another unpreferenced SOES order at the same bid or offer in the same
security until 17 seconds have elapsed, absent a quotation update by
the market maker within such 17-second period.  The additional 2
seconds were added to the 15-second delay to allow time for market
makers to receive reports of SOES executions. 


      MARCH 1998
------------------------------------------------------ Appendix I:0.25

SEC publishes, for public comment, a proposed NASD rule change to
establish a new Integrated Order Delivery and Execution System.  The
system would replace existing SOES and SelectNet systems and provide
an integrated order routing and execution system featuring: 

1.  Automatic trade executions for all. 

The system is expected to provide investors and traders immediate and
automatic executions in strict time priority.  It would provide a
limit order file through which both limit orders and market maker
quotes would be subject to automatic execution.  However, it would
allow market makers to turn the system off momentarily to prevent
themselves from getting hit while handling telephone orders. 

2.  No size limits and no ephemeral quotes. 

The system would allow orders of up to 999,999 shares to be entered,
with a minimum life of 10 seconds. 

3.  No multiple executions. 

A market maker would have to honor only one execution at its quoted
price and would have 17 seconds to update its quote.  For trades of
more than 1,000 shares the market maker would have the option of
declining the trade if a trade had already been executed at the quote
or if the market maker was updating its quote. 

NASD submits proposed rule change to permanently allow market makers
to quote actual size by reducing the minimum quotation size
requirement for market makers in all securities listed on Nasdaq to
one normal unit of trading (at least 100 shares or multiples
thereof).  NASD staff study of test period found that the actual size
rule had no effect on market quality or SOES access.  One commentator
opposed to the rule change disputed NASD's finding, contending that
the actual size rule resulted in substantial reduction in market
depth and liquidity, especially during times of market stress. 


      JULY 1998
------------------------------------------------------ Appendix I:0.26

SEC approves a NASD proposed rule change to allow market makers to
quote permanently in actual size by reducing the minimum quotation
size requirement for all Nasdaq securities to one normal unit of
trading of 100 shares and multiples of 100.  In its study of the
effects of the actual size rule, NASD stated that artificial minimums
were no longer necessary and that the rule allows market makers to
better manage risks with no harm to market quality or SOES access. 




(See figure in printed edition.)Appendix II
COMMENTS FROM NASD
=========================================================== Appendix I




(See figure in printed edition.)Appendix III
COMMENTS FROM THE SECURITIES AND
EXCHANGE COMMISSION
=========================================================== Appendix I



(See figure in printed edition.)


MAJOR CONTRIBUTORS TO THIS REPORT
========================================================== Appendix IV

GENERAL GOVERNMENT DIVISION

Michael A.  Burnett, Assistant Director
Tamara E.  Cross, Evaluator-in-Charge
David P.  Tarosky, Senior Evaluator
Mitchell B.  Rachlis, Senior Economist

OFFICE OF THE GENERAL COUNSEL

Rosemary Healy, Senior Attorney


*** End of document. ***