Credit Derivatives: Confirmation Backlogs Increased Dealers'	 
Operational Risks, but Were Successfully Addressed after Joint	 
Regulatory Action (13-JUN-07, GAO-07-716).			 
                                                                 
Over-the-counter (OTC) credit derivatives are privately 	 
negotiated contracts that allow a party to transfer the risk of  
default on a bond or loan to another party without transferring  
ownership. After trading in these products grew dramatically in  
recent years, backlogs of thousands of trades developed for which
dealers had yet to formally confirm the trade terms with	 
end-users--such as hedge funds, pension funds, and insurance	 
companies--and other dealers. Not confirming these trades raised 
the risk that losses could arise. GAO was asked to review (1)	 
what caused the trade confirmation backlogs and how they were	 
being addressed and (2) how U.S. financial regulators were	 
overseeing dealers' credit derivative operations, including the  
security and resiliency of the information technology systems	 
used for these products. GAO analyzed data on credit derivatives 
operations that dealers submitted to regulators, reviewed	 
regulatory examination reports and work papers, and interviewed  
regulators, dealers, end-users, and industry organizations.	 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-07-716 					        
    ACCNO:   A70666						        
  TITLE:     Credit Derivatives: Confirmation Backlogs Increased      
Dealers' Operational Risks, but Were Successfully Addressed after
Joint Regulatory Action 					 
     DATE:   06/13/2007 
  SUBJECT:   Banking regulation 				 
	     Bonds (securities) 				 
	     Credit bureaus					 
	     Derivative securities				 
	     Federal regulations				 
	     Lending institutions				 
	     Policy evaluation					 
	     Risk management					 
	     Securities regulation				 
	     Trade agreements					 
	     Trade regulation					 
	     Policies and procedures				 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO 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.                                         **
**                                                              **
******************************************************************
GAO-07-716

   

     * [1]Results in Brief
     * [2]Background

          * [3]Processing Credit Derivatives Trades
          * [4]Regulation of OTC Credit Derivatives

     * [5]Manual Processes and Trade Assignments Led to Backlogs of Un

          * [6]Two Factors Largely Caused the Confirmation Backlogs at Deal
          * [7]Confirmation Backlogs and Unilateral Assignments Increased D
          * [8]Under the Direction of FRBNY as well as Other Regulators, De
          * [9]Dealers Reduced Backlogs through Various Steps
          * [10]Market Participants Agreed to End Unilateral Assignments and
          * [11]Dealers Found Benefits in the Joint Regulatory Initiative
          * [12]Dealers and Other Market Participants Continue to Work to Re

     * [13]The Joint Regulatory Initiative Enhanced U.S. Regulators' Ov

          * [14]Federal Bank Regulators Were Monitoring Efforts Taken by Ban
          * [15]SEC Also Conducted Oversight of Credit Derivatives Confirmat
          * [16]Through the Joint Regulatory Initiative, Regulators Are Obta

     * [17]Agency Comments
     * [18]GAO Contact
     * [19]Staff Acknowledgments
     * [20]GAO's Mission
     * [21]Obtaining Copies of GAO Reports and Testimony

          * [22]Order by Mail or Phone

     * [23]To Report Fraud, Waste, and Abuse in Federal Programs
     * [24]Congressional Relations
     * [25]Public Affairs

Report to Congressional Requesters

United States Government Accountability Office

GAO

June 2007

CREDIT DERIVATIVES

Confirmation Backlogs Increased Dealers' Operational Risks, but Were
Successfully Addressed after Joint Regulatory Action

GAO-07-716

Contents

Letter 1

Results in Brief 3
Background 5
Manual Processes and Trade Assignments Led to Backlogs of Unconfirmed
Trades at Dealers, but Industry Efforts Have Significantly Reduced the
Backlogs 11
The Joint Regulatory Initiative Enhanced U.S. Regulators' Oversight of
Dealers' Efforts to Reduce Their Backlogs 30
Agency Comments 38
Appendix I Scope and Methodology 40
Appendix II GAO Contact and Staff Acknowledgments 42

Tables

Table 1: Share of Market by Credit Derivative Product, 2006 6
Table 2: Number of Credit Derivatives Trade Confirmations Outstanding for
14 Major Dealers, September 2005 11
Table 3: Outstanding Confirmation Reduction Goals and Totals for the 14
Credit Derivatives Dealers 22
Table 4: Outstanding Confirmations and Related Data Provided by Dealers
from October 2006 to March 2007 23

Figures

Figure 1: Total Notional Amount of Credit Default Swaps Market, 2001 to
2006 7
Figure 2: Steps for Processing a Credit Derivative Trade 9
Figure 3: Total Outstanding Confirmations at 14 Major Dealers from
September 2005 to October 2006 21
Figure 4: Share of Total Monthly Credit Derivatives Trades of the 14
Dealers Confirmed Electronically, September 2005 to October 2006 24
Figure 5: Total Monthly Volume of Assigned Trades and Number of
Unconfirmed Assignments Outstanding for More Than 30 Days for 14 Major
Dealers, September 2005 and October 2006 27

Abbreviations

BIS Bank for International Settlements (Basel, Switzerland)
CSE Consolidated Supervised Entity
FSA Financial Services Authority (United Kingdom)
FRBNY Federal Reserve Bank of New York
DTCC Depository Trust and Clearing Corporation
ISDA International Swaps and Derivatives Association
LIBOR London Interbank Offered Rate
OCC Office of the Comptroller of the Currency
OTC over the counter
SEC Securities and Exchange Commission

This is a work of the U.S. government and is not subject to copyright
protection in the United States. It may be reproduced and distributed in
its entirety without further permission from GAO. However, because this
work may contain copyrighted images or other material, permission from the
copyright holder may be necessary if you wish to reproduce this material
separately.

United States Government Accountability Office
Washington, DC 20548

June 13, 2007

Congressional Requesters

Until late 2005, the growth in trading volume of over-the-counter (OTC)
credit derivatives had greatly outpaced the processing capabilities of the
financial firms offering these products--heightening the operational risk
that such firms could incur losses from human errors or system failures.
OTC credit derivatives are privately negotiated contracts that allow a
party to transfer the risk of default on a bond or loan to another party
without transferring ownership. In a credit default swap, for example, a
bond investor agrees to pay a periodic premium to a financial firm in
exchange for the firm's agreement to compensate the bond investor for any
losses if the bond issuer defaults on the bonds. Like other OTC
derivatives, credit derivatives are typically bought and sold through
dealers, namely banks and securities broker-dealers, that stand ready to
buy or sell credit derivatives to end-users, such as hedge funds, pension
funds, and insurance companies. Although OTC trading in credit derivatives
is not regulated in the United States, the dealers are subject to
supervision by their respective regulators, including U.S. banking and
securities regulators.1

Introduced in the early 1990s, credit derivatives surpassed a total
notional amount of $34 trillion at year-end 2006.2 As trading volume grew
exponentially in recent years, major dealers developed backlogs of
thousands of trades for which the trade terms had not been formally
confirmed with their counterparties, which included end-users and other
dealers. Having unconfirmed trades could allow errors to go undetected at
dealers and later result in losses, a situation that an official from the
United Kingdom's regulator of credit derivatives dealers characterized as
"an accident waiting to happen."

1The Securities and Exchange Commission has antifraud authority over
credit derivatives that are "security-based swap agreements," as defined
in Section 206B of the Gramm-Leach-Bliley Act (15 U.S.C. 78c note). See,
e.g., Section 10(b) of the Securities Exchange Act of 1934 (15 U.S.C.
78j(b)).

2The notional amount is the amount upon which payments between
counterparties to certain types of derivatives contracts are based. For
credit derivatives, the notional amount serves as the basis for
determining the periodic premium payment made by one party to another in
return for compensation in the event of loss from a default. In this
regard, the credit derivatives market's notional amount is an indicator of
the market's volume but does not necessarily represent the credit and
market risks to which counterparties are exposed from their credit
derivatives contracts.

Given the concerns about the inability of the credit derivatives market's
infrastructure to keep up with the growth in trading volume, you asked us
to review the causes of the confirmation backlogs and the steps U.S.
financial regulators were taking to address the issue. This report
discusses (1) what caused the backlogs and the steps being taken to
address them and (2) U.S. financial regulators' oversight of the
operational risk that dealers faced from the backlog in credit derivatives
confirmations, including the security and resiliency of related
information technology systems.3

To determine the causes of the backlogs and the steps that are being taken
to address these issues, we analyzed the trading volume of credit
derivatives, confirmation backlogs, and other transactional data provided
by major dealers of credit derivatives to U.S. and foreign regulators
through Markit Group, a provider of independent data, portfolio
valuations, and trade processing for OTC derivatives. We examined the
procedures that this firm employs to collect and analyze the data and
determined that the data were sufficiently reliable for our purposes. We
also reviewed and analyzed reports on the credit derivatives market by
industry associations, international organizations, firms, and academics.
We interviewed eight dealers of credit derivatives, a hedge fund, and
various industry trade organizations, including organizations representing
OTC derivatives dealers and derivatives end-users, including the
International Swaps and Derivatives Association (ISDA). To determine how
dealers' exposures to operational risks associated with credit derivatives
were being supervised, we interviewed staff from the Federal Reserve,
including its examiners for two banks; the Office of the Comptroller of
the Currency (OCC), including examiners for three banks; the Securities
and Exchange Commission (SEC); and the United Kingdom's Financial Services
Authority (FSA). We also reviewed and analyzed examinations conducted
between 2004 and 2006 by the Federal Reserve and the OCC on the activities
in credit derivatives of five banks and by the SEC covering the holding
companies of five securities broker-dealers that engage in credit
derivatives activities. We conducted our work in Charlotte, North
Carolina; Chicago; New York; and Washington, D.C., from August 2006 to
March 2007 in accordance with generally accepted government auditing
standards. Appendix I provides a detailed description of our scope and
methodology.

3While organized futures exchanges recently announced their plans to offer
credit derivatives, our report discusses only credit derivatives traded in
the OTC derivatives market. In addition, although concerns have been
raised about the potential for credit derivatives to raise systemic risk
and be used to trade on insider information, our report addresses only the
operational risks raised by credit derivatives.

Results in Brief

Two factors largely led to the substantial backlogs of unconfirmed trades
that dealers had amassed by 2005, though regulators, dealers, and others
have since made considerable progress in reducing these backlogs. From
2002, trading volume in credit derivatives was expanding exponentially,
with particularly rapid growth from 2004 to 2005, as the average number of
trades done weekly at large dealers increased from 644 to 1,450. As a
result, by the end of September 2005, 14 of the largest credit derivatives
dealers had, in aggregate, over 150,000 unconfirmed trades, with nearly
two-thirds of these remaining unconfirmed for more than 30 days. The
delays in confirming trades largely resulted from (1) dealers and
end-users relying on inefficient manual processes that could not
adequately keep up with the rapidly growing volume and (2) the difficulty
of confirming trade information after some end-users began frequently
assigning their side of existing trades to new parties without notifying
the original dealer. The backlog of unconfirmed trades created operational
risk by potentially allowing trade errors to go undetected that could lead
to losses and other problems. For example, undetected errors could result
in legal disputes over contract terms and cause dealers to incorrectly
measure and manage their market or credit risk.4 To mitigate these risks,
some dealers were informally verifying the key economic terms of trades
with counterparties to ensure that trades were accurately recorded and
risks were accurately measured, but the extent to which these practices
were followed varied. According to regulators, the trade assignment
practice posed a "collective action" problem because dealers could not
individually stop the practice. As a result, in September 2005, U.S. and
foreign financial regulators participated in a joint regulatory initiative
organized by the Federal Reserve Bank of New York (FRBNY) and prompted the
14 major credit derivatives dealers to work together to reduce the number
of unconfirmed trades and address the underlying causes of these backlogs.
Using automated systems to confirm trades and adopting a protocol
requiring end-users to obtain dealers' consent before assigning trades,
the 14 dealers reduced the number of confirmations outstanding for more
than 30 days by 94 percent (to around 5,500 trades) by the end of October
2006. Dealers and others are continuing to work to reduce operational
risks, in part by further automating the market's infrastructure--for
example, by developing a central depository to store virtually all trades
and automate other processes.

4Market risk is the potential for loss because of a decrease in value of a
credit derivative contract resulting from a change in market conditions.
Credit risk is the potential for loss from the failure of the counterparty
to perform on its credit derivative contract.

U.S. bank and securities regulators had been overseeing the exposure of
credit derivatives dealers to operational and other risks, but were better
able to monitor the resolution of the backlog problem once they began
receiving industrywide data under FRBNY's joint regulatory initiative.
Through supervision and examinations, the pertinent U.S. federal bank
regulators--the Federal Reserve and the OCC--became aware of the backlogs
at U.S. banks engaged in credit derivatives activities as early as late
2003 and were monitoring banks' efforts to reduce their backlogs before
the joint regulatory initiative. The securities regulator, SEC, was
generally aware of the backlogs since late 2004, but SEC staff became more
concerned about them through periodic discussions with broker-dealers
subject to the SEC's Consolidated Supervised Entity (CSE) program during
the summer of 2005. As part of their examinations, the bank and securities
regulators have also been reviewing how these dealers maintained the
security and resiliency of the information technology systems used for
credit derivatives. Although U.S. and foreign regulators were aware of
confirmation backlogs at individual dealers, none of the regulators
oversaw all the dealers or had data on the size of the backlog
industrywide. However, under the joint regulatory initiative begun in
September 2005, the dealers have been providing the regulators with
aggregate data on their backlogs and other operational measures, giving
regulators an effective means for monitoring the industry's progress in
reducing the backlogs. In recognition of the potential for similar
operational problems to arise in other OTC derivatives markets, including
OTC equity derivatives, U.S. and foreign regulators have begun to collect
similar data for other OTC derivative products.

We provided a draft of this report to the Federal Reserve, OCC, and SEC
for their review and comment. The Federal Reserve and SEC provided
technical comments, which we incorporated as appropriate.

Background

Introduced in the early 1990s, credit derivatives have been widely adopted
as a tool for allowing market participants to take on or reduce their
exposure to credit risk. First used primarily by banks to reduce credit
exposures stemming from loans made to clients, credit derivatives have
evolved to include an array of different products (table 1). According to
regulators and others, credit derivatives have the potential to improve
the overall efficiency and resiliency of the financial markets by
spreading credit risk more widely across a large and diverse pool of
investors. According to the British Bankers' Association, single-name
credit default swaps remain the most common type of credit derivative,
comprising about 33 percent of the market in 2006, though their share of
the market has decreased since 2004. These swaps allow the buyer of
protection to transfer the credit risk associated with default on debt
issued by a single corporation or sovereign entity--called the reference
entity.5 With a standard credit default swap, the buyer of credit risk
protection pays a quarterly premium payment to the seller of credit risk
protection over the life of the contract, typically 5 or more years.
Should a defined credit event occur, such as a default by the specified
corporation on the referenced debt, the protection seller would assume the
losses.6 As table 1 shows, other commonly traded products include full
index trades, synthetic collateralized debt obligations, and tranched
index trades.

5Market participants can buy or sell credit derivatives for the purposes
of speculating, arbitraging, or hedging, even if they do not have a direct
exposure to the referenced entity.

6Credit events include, for example, failure to pay, restructuring, and
bankruptcy.

Table 1: Share of Market by Credit Derivative Product, 2006

Credit derivative   Market                                                 
product             share(percentage)  Product description                 
Single-name credit                33%  Credit derivatives based on bonds   
default swaps                          from a single issuer, such as a     
                                          corporation or a sovereign entity.  
Full Index trades                  30  Credit derivatives referencing      
                                          multiple corporations or sovereign  
                                          entities that are gathered into a   
                                          standardized portfolio and offered  
                                          to investors as one unit. Indexes   
                                          are usually categorized by          
                                          characteristics such as industry,   
                                          geographic region, or credit        
                                          quality.                            
Synthetic                          16  Credit derivatives referencing      
collateralized debt                    multiple corporations or sovereign  
obligations                            entities and gathered into a        
                                          standardized portfolio--customized  
                                          for investors--and separated into   
                                          various risk categories (or         
                                          tranches) that vary by the          
                                          likelihood of incurring losses.     
                                          Obligations from the tranches are   
                                          then sold to investors according to 
                                          the desired risk/return profile.    
Tranched index                      8  Index trades that are divided into  
trades                                 various risk tranches, with         
                                          investors selecting the risk and    
                                          return profile they prefer among    
                                          the various risk categories in the  
                                          standardized index.                 
Othersa                            13  A variety of special-purpose        
                                          products that collectively          
                                          represent a relatively small share  
                                          of the credit derivatives market.   
Total                            100%                                      

Source: Percentages from British Bankers' Association and definitions from
GAO analysis of multiple sources.

a "Others" include products that each account for less than a 6 percent
share of the market.

In the credit derivatives market, banks and securities broker-dealers
generally serve as the product dealers, acting as the buyer or seller in
credit derivative trades with end-users or other dealers. The top five
dealers in 2005, ranked by total trading volumes as estimated by Fitch
Ratings, were Morgan Stanley, Deutsche Bank, Goldman Sachs, JP Morgan
Chase, and UBS. End-users of credit derivatives include hedge funds,7
insurance companies, pension funds, and mutual funds.8

According to ISDA, which conducts periodic surveys of market participants,
the credit derivatives market has grown dramatically in recent years,
increasing from an estimated total notional amount of nearly $1 trillion
outstanding at year-end 2001 to over $34 trillion at year-end 2006 (see
fig. 1).9 Part of this rapid growth has been attributed to product
innovation and an increasing number of market participants, particularly
hedge funds. Despite its expansion, the credit derivative market is still
much smaller than the OTC interest rate derivatives market, which had a
total notional amount outstanding of around $286 trillion at year-end
2006.10

7Hedge funds are generally considered private investment funds that are
not required to register with SEC because of the limited number or
sophisticated nature of their investors. Hedge funds commonly seek to
achieve a positive, absolute return and invest in a wide variety of
financial instruments, such as equity and fixed income securities,
currencies, over-the-counter derivatives, and futures contracts.

8The top five end-users of credit derivatives are banks and broker-dealers
(44 percent), hedge funds (32 percent), insurers (17 percent), pension
funds (4 percent), and mutual funds (3 percent). Ross Barrett and John
Ewan, BBA Credit Derivatives Report 2006 (London: British Bankers'
Association, September 2006).

Figure 1: Total Notional Amount of Credit Default Swaps Market, 2001 to
2006

9ISDA is a global trade association representing market participants in
privately negotiated derivative transactions, which are commonly called
OTC derivatives. Since 2000, ISDA has conducted an annual operations
benchmarking survey of its members to collect performance data on
operations processing of OTC derivatives.

10The market for OTC interest-rate derivatives includes interest-rate
swaps and options as well as cross-currency interest rate swaps. For
example, an interest-rate swap is a transaction in which one party pays
periodic amounts based on a specified fixed rate and the other party pays
periodic amounts based on a specified floating rate that is reset
periodically, such as the London Interbank Offered Rate, or LIBOR (the
interest rate paid on interbank deposits in the international money
markets).

Processing Credit Derivatives Trades

Traders and sales staff at dealers who interact with customers represent
the dealer's "front office." The staff in the front office generally use
electronic systems to capture the trade data and transmit it to the
systems used to manage market and credit risk. Dealers also have "back
offices," which include staff that record, verify, and confirm trades
executed by the front office. As shown in figure 2, the steps for entering
into and processing an OTC trade for credit derivatives include
negotiation, capture, verification, and confirmation. These processes have
been increasingly automated over time, but some remain manual. For
example, a relatively small percentage of credit derivative
products--generally those with more customized and complex terms--cannot
be confirmed electronically. In addition, various post-trade processes
occur during the life of a credit derivatives contract, including making
or receiving premium payments, exchanging collateral, and settling
contracts after a credit event occurs, such as a bond default.

Figure 2: Steps for Processing a Credit Derivative Trade

Segregating these various duties into front and back office
responsibilities serves to maintain operational integrity, such as by
identifying data entry errors and to minimize fraud and other violations.
Management responsibilities performed by the back office vary by
institution, but they may include evaluating transactional exposure
against established market and credit limits and risk management
reporting. Some dealers have combined a number of the functions performed
by the back office, such as risk management, into a middle office, and
some use a separate risk management group.

Regulation of OTC Credit Derivatives

Because OTC credit derivative transactions occur between private parties
and are not traded on regulated exchanges, they are not subject to
regulation in the United States, provided that the parties and other
aspects of the transaction satisfy requirements of the Commodity Exchange
Act.11 For credit derivatives that would otherwise be securities, the
transactions fall within the definition of "swap agreement" in the
Gramm-Leach-Bliley Act.12 The Commodity Exchange Act allows unregulated
derivatives trading in certain types of commodities by eligible parties
under limited circumstances.13 Similarly, the Commodity Exchange Act and
the Securities Act of 1933 allow unregulated derivatives trading by
eligible parties under limited circumstances.14 Although the OTC credit
derivatives products themselves are not regulated, certain market
participants are. If the dealer is a U.S. bank federally chartered as a
national bank, it is supervised by OCC. If a bank is owned by a bank
holding company, its holding company is regulated by the Federal
Reserve.15 These bank regulators oversee these entities to ensure the
safety and soundness of the banking system and the stability of the
financial markets. If the credit derivatives dealer is a securities
broker-dealer, it is overseen by SEC. According to U.S. regulators, some
of the U.S. banks and securities broker-dealers also conduct credit
derivatives trades in foreign affiliates subject to foreign regulation.
Similarly, other participants in the credit derivatives market include
foreign banks that are supervised by foreign regulators and, in some
cases, also by U.S. regulators if operating in the United States.

117 U.S.C. SS 1 - 25, as amended.

1215 U.S.C. 78c note.

13Under the Commodity Exchange Act, credit derivatives transactions
generally are not subject to regulation if the transactions are between
"eligible contract participants" (as defined in the act) and either do not
take place on a "trading facility" or occur only on an electronic trading
facility and are conducted on a principal-to-principal basis or are
subject to individual negotiation by the parties. See 7 U.S.C. SS 1a(12),
1a(13), 2(d), 2(g).

14The Securities Exchange Act of 1934 and the Securities Act of 1933 each
exclude from the definition of security both "security-based swap
agreements" and "non-security-based swap agreements," as defined in
Sections 206B & C of the Gramm-Leach-Bliley Act (15 U.S.C. 78c note). See
15 U.S.C. 78c-1 and 15 U.S.C. 77b-1. However, SEC has antifraud authority
over credit derivatives that are security-based swap agreements. See note
1.

15For a more detailed description of the regulation of banks and
securities broker-dealers, see GAO, Financial Regulation: Industry Changes
Prompt Need to Reconsider U.S. Regulatory Structure, [26]GAO-05-61 ,
(Washington, D.C.: October 2004).

Manual Processes and Trade Assignments Led to Backlogs of Unconfirmed Trades at
Dealers, but Industry Efforts Have Significantly Reduced the Backlogs

As the credit derivatives market grew, lack of automation and other
factors led to large backlogs of unconfirmed trades at dealers. The eight
dealers we interviewed told us that they began to experience a significant
increase in their backlogs of unconfirmed trades ranging from the middle
of 2003 to the first half of 2005. According to ISDA's survey data,
trading volume in credit derivatives more than doubled around this period,
with the average number of trades conducted at large firms increasing from
644 trades a week in 2004 to 1,450 trades a week in 2005. According to
data provided to regulators by 14 of the largest credit derivatives
dealers--which include U.S. and foreign banks and securities
broker-dealers--these dealers collectively executed around 130,000 trades
in September 2005, and dealers' backlogs of confirmations outstanding had
risen to over 150,000 (table 2).16 Of these, 63 percent had been
outstanding for more than 30 days, and 41 percent had been outstanding for
more than 90 days.

Table 2: Number of Credit Derivatives Trade Confirmations Outstanding for
14 Major Dealers, September 2005

                            Number of confirmations   Percentage of the total 
                                        outstanding confirmations outstanding 
Confirmations                             56,224                       37% 
outstanding 30 days or                                                     
less                                                                       
Confirmations                             97,650                       63% 
outstanding more than 30                                                   
days                                                                       
Confirmations                             63,322                      41%a 
outstanding more than 90                                                   
days                                                                       
Total confirmations                      153,860                      100% 
outstanding                                                                

Source: GAO analysis of Markit Group data.

aThe number of confirmations outstanding more than 90 days is included in
the total of confirmations outstanding more than 30 days. As a result, the
percentages do not add up to 100 percent.

16As identified in an attachment to the Federal Reserve Bank of New York's
September 15, 2005, press release, the 14 dealers are Bank of America;
Barclays Capital; Bear, Stearns & Co.; Citigroup; Credit Suisse; Deutsche
Bank; Goldman Sachs Group; HSBC; JP Morgan Chase; Lehman Brothers; Merrill
Lynch & Co.; Morgan Stanley; UBS; and Wachovia Bank.

Two Factors Largely Caused the Confirmation Backlogs at Dealers

A major factor contributing to the backlogs was dealer and end-user
reliance on largely manual processes for confirming credit derivative
trades that could not keep up with the rapidly growing trade volume.
Unlike highly automated processes for confirming trades in corporate
stocks, the processes that dealers were generally using to confirm their
credit derivative trades relied on inefficient manual procedures. For
example, a dealer would manually prepare a confirmation and fax it to the
counterparty; in turn, the counterparty would manually compare its trade
record against the confirmation and, if the terms matched, fax the signed
confirmation to the dealer. Such manual processes were resource intensive
and generally lacked the scalability required to efficiently confirm the
rapidly growing volume of trades.

Recognizing the need to improve the efficiency of the confirmation process
for credit derivatives, dealers had been working with the Depository Trust
and Clearing Corporation (DTCC) to increase the use of an automated
confirmation system.17 DTCC staff said that they started to work with
several dealers in 2002 to create an automated system to electronically
compare, match, and confirm credit derivative trades. The initial strategy
was to have the system confirm only single-name credit default swaps and
then to expand the system's capabilities to confirm other credit
derivative products and provide other services. DTCC launched its
automated system, Deriv/SERV, in late 2003, and 15 dealers and 7 end-users
had signed up to use the system by around mid-2004. DTCC staff explained
that obtaining wider use of Deriv/SERV took time, in part because of the
need to publicize the system and because users needed to train their staff
and revise their systems to use Deriv/SERV. According to staff at one
hedge fund, many end-users did not initially use Deriv/ SERV because they
lacked the necessary technology. Consequently, up to 85 percent of the
credit derivative trades were being confirmed manually during 2004,
according to market participants. However, DTCC has expanded Deriv/SERV's
capabilities to confirm a broader range of credit default swaps and as
much as 46 percent of trades were being confirmed electronically by
September 2005, according to data provided to regulators by 14 major
dealers.

17Through subsidiaries, DTCC provides clearance, settlement, and
information services for equities, corporate and municipal bonds,
government and mortgage-backed securities, and OTC credit derivatives.

The second major factor contributing to the backlogs was the increasing
incidence of end-users transferring their positions to other
counterparties. Although the length of the contract for the most popular
credit derivatives typically spans 5 years, some end-users, particularly
hedge funds, engaged in frequent "in and out" trading of these products or
had other incentives to liquidate their positions earlier. To do so, the
end-users assigned their sides of trades to third parties. Although the
agreements accompanying the trades did not permit assignments without the
dealer's prior consent, the dealers agreed to assignments after the fact
because of competitive pressures and because the new counterparties (the
assignees) tended to be other dealers. In effect, these assignments (also
called novations) ultimately resulted in a new contract between the
original dealer and the new counterparty, which would not be reflected on
the dealer's records until the original dealer accepted the assignment. A
hedge fund official told us that when his firm wanted to terminate trades
early, it initially returned to the original dealers, but the dealers
charged termination fees that made this method more costly than assigning
the trades.18 Assignments have provided greater market liquidity and price
discovery,19 but according to dealers and regulators, they complicated the
confirmation process. Without prior knowledge of an assignment, the
original dealer could not readily confirm the details of the new trade
until the dealer became aware of the assignment. Some end-users said that
they obtained consent from the original dealers but that the dealers were
not communicating the information internally to the appropriate staff for
the purpose of confirming the trade.

Although assigned trades were a small share of total trading volumes, they
represented a disproportionately large share of unconfirmed trades because
of the time required to identify the correct counterparty. According to
data provided by the 14 major dealers to regulators, trade assignments
accounted for 13 percent of dealers' trading volume in September 2005 but
40 percent of their total confirmations outstanding for more than 30 days
at the end of September 2005. Dealers told us that they typically detected
unilateral assignments through payment errors. For example, a dealer would
receive a premium payment from a party other than the party with which it
had entered into the trade. Importantly, market participants had agreed to
settle premium payments due under credit default swaps on a quarterly
basis in order to provide greater market liquidity. Bank examiners told us
that because of this settlement cycle, it could take a dealer as many as
90 days or more to detect a unilateral assignment through a payment error.

18To economically terminate a trade, an end-user could enter into an equal
but opposite trade that offsets the original trade. This approach has the
disadvantage of building up large netted positions between market
participants. Hedge funds prefer to avoid this outcome because it creates
additional operational costs.

19Liquidity is the extent to which market participants can buy and sell
contracts in a timely manner without changing the market price, and price
discovery is the process of determining price on the basis of supply and
demand factors.

Dealers we spoke with identified several other factors that hampered their
efforts to confirm trades in a timely manner. First, some dealers told us
that as the volume of trading in credit derivatives grew, they faced
challenges hiring experienced back-office staff and that training new
staff took months. Second, other dealers said that the lack of
standardized documentation, particularly for new products, led to disputes
over the trade terms or the need to negotiate them, further delaying
confirmation.20 Compounding matters, there was a shortage of derivatives
attorneys available for such negotiations, according to a bank examiner.
Finally, two dealers said that the industry lacked standardized reference
data to identify the specific entities referenced in credit derivative
contracts. One of the dealers told us that the lack of such data led to
mistakes in recording trades and hampered electronic confirmations.
Mistakes in documenting the correct reference entity prompted a group of
dealers to develop a database of reference entities and obligations in
2003 that has become an industry standard.

Confirmation Backlogs and Unilateral Assignments Increased Dealers' Operational
Risks

Although dealers were capturing their credit derivatives trades in their
risk management systems to manage the associated market and credit risks,
the substantial backlog of unconfirmed trades heightened dealers'
operational risk, potentially hampering their ability to effectively
manage other risks. As with any trading activity, dealers engaging in
credit derivative trades are exposed to market, credit, and other risks
that they must adequately measure, monitor, and control. According to
dealers and their regulators, the major credit derivatives dealers
generally were entering their credit derivatives trades promptly into
their trade capture systems and, in turn, measuring, monitoring, and
managing the credit and market risks associated with those trades.21
Dealers, for example, measure and manage market risk by estimating the
potential losses that a portfolio of positions may suffer and then impose
limits that restrict the estimated losses to an acceptable level.
Similarly, dealers manage counterparty credit risk--which can produce
losses if the dealers fail to receive payments owed to them--generally by
measuring the total credit exposure to, and creditworthiness of,
individual counterparties, and not allowing these exposures to exceed
pre-established limits.

20ISDA issued a standard confirmation form for credit derivatives in 1998,
a set of definitions for credit derivatives in 1999, and a set of revised
definitions in 2003 to reflect industry changes.

Although the credit and market risks were being managed, the large
backlogs of unconfirmed trades increased dealers' operational risks.
Confirmations serve as an internal control to verify that both parties
agree to the trade terms and have accurately recorded the trade in their
systems. For this reason, trades should be confirmed as soon as possible.
Having unconfirmed trades could allow errors to go undetected that might
subsequently lead to losses and other problems. Errors could be made at
any time--for example, counterparties could miscommunicate when making a
trade or dealers could enter the wrong trade data into their systems. If
such errors go undetected, a dealer could make an incorrect premium
payment to a counterparty or inaccurately measure and manage risk
exposures, notably market and counterparty credit risks. Similarly, errors
could lead to legal disputes between a dealer and a counterparty if a
credit event triggered a contract settlement.

Further, these operational risks have the potential to contribute to
broader market problems. For example, in its July 2005 report on
strengthening the stability of the global financial system, the
Counterparty Risk Management Policy Group II, composed of representatives
of dealers and end-users, noted that as the number of outstanding credit
derivatives trades continues to grow, a credit event involving a popular
reference entity could materially strain the ability of market
participants to settle transactions in a timely and efficient manner.22
However, these operational risks did not result in such broader market
problems, in part because of favorable market conditions when the
confirmation backlog arose and because only seven referenced entities in
the United States have defaulted since 2005--with market participants able
to effectively settle trades referencing these entities.

21Dealers also told us that their trade capture systems automatically feed
the data on credit derivatives trades to their accounting systems. Thus,
the dealers captured their trades in their books and records, which are
used to prepare their financial statements.

22See, Counterparty Risk Management Policy Group II, Toward Greater
Financial Stability: A Private Sector Perspective (July 27, 2005).

Although unconfirmed trade backlogs were growing, dealers had been taking
steps to reduce the operational risks associated with these trades. To
ensure that the trade data being captured and used to manage risks were
accurate, dealers were informally contacting their counterparties before
sending out confirmations to verify the key economic terms of the trades,
but this practice varied among the nine dealers reviewed. Specifically,
five dealers generally followed this practice for their credit derivative
trades, according to their staff or examiners. In contrast, two dealers
generally had been informally verifying trade terms for only those trades
considered higher risk, according to their staff or examiners. Finally,
staff at two other dealers said that they generally were not verifying
trade terms before confirmation because their counterparties preferred not
to do so.23 The dealers also were monitoring their confirmation backlogs
based on risk, such as by the number of days an unconfirmed trade was
outstanding. Moreover, two dealers curtailed business with clients that
had a large number of outstanding confirmations. In addition, the dealers
had reviewed their confirmation processes and were improving them by,
among other things, upgrading technology, reorganizing operations, and
hiring staff. While dealers found some errors after confirming their
trades, only two of the dealers interviewed told us that they had suffered
a $1-million-or-more loss as a result of an error stemming from their
confirmation backlog but characterized the losses as immaterial.

Like unconfirmed trades, unilateral assignments increased operational,
credit, and legal risks. First, unilateral assignments led to operational
risk by creating new trades that were not being confirmed promptly to
detect errors. Second, to effectively manage credit risk, dealers must
know, at a minimum, the correct identities of the counterparties to their
credit derivative contracts. Unconfirmed trades arising from unilateral
assignments meant that dealers did not always know the exact counterparty
to which they were exposed. As a result, their ability to accurately
measure their credit exposure and enforce their pre-established limits on
it was hampered. Moreover, because dealers did not always know the correct
counterparties for each of their trades, they often made premium payments
to, or received payments from, the wrong entity. Third, unconfirmed
assigned trades also raised dealers' legal risk because of the potential
for counterparties to later dispute the terms of the trade or the
enforceability of the contract. For example, a court may deem an assigned
trade as legally invalid if the original dealer did not provide its
written consent. As the Counterparty Risk Management Policy Group II
reported in 2005, some assignments occurred before the original trades
were confirmed, increasing the risk of potential disputes over the status
and the terms of the trade.

23In March 2006, the 14 dealers committed to verify the key economic terms
of (1) standardized trades whose terms remained unmatched and thus
unconfirmed for 5 or more business days after trade date and (2)
nonstandardized trades within 3 business days after trade date.

Several factors helped to mitigate the risks arising from unilateral
assignments. According to dealers and regulators, the assignments did not
increase market risk for dealers because dealers generally were capturing
the key economic terms of the trades in their risk management systems
accurately, and these terms remained the same when a trade was assigned.
Further, although unilaterally assigned trades impaired the ability of
dealers to measure and manage their counterparty credit risk, dealers and
examiners told us that hedge funds and other end-users assigned nearly all
of their trades to dealers, given their role as intermediaries to
end-users. Because dealers were typically more creditworthy than the
end-users assigning the trades, the original dealers ended up with more
creditworthy counterparties after an assignment, according to dealers and
examiners. Situations could arise, however, where this factor would not
necessarily mitigate the original dealer's counterparty credit exposure.24
In addition, dealers told us that they had collateral arrangements with
their counterparties to manage their credit risk. For example, dealers
required hedge funds to post a negotiated amount of initial collateral,
such as cash or securities, for each trade they entered into with
dealers.25 As a risk management practice, two dealers told us that they
would not release collateral to their counterparties until they verified
that a trade was assigned. In addition, a provision of the standard
contract that counterparties enter into as part of conducting derivatives
transactions--known as the ISDA Master Agreement--required counterparties
to obtain the written consent of their counterparty before assigning a
trade.26 Some dealers told us that they could have relied on this
provision, if needed, to reject a unilateral assignment. Finally, none of
the dealers said that their counterparties tried to nullify an assigned
trade.

24The original dealer would not necessarily benefit from the trade being
assigned to a more creditworthy dealer. As an example, if a hedge fund
unilaterally assigned a contract to a new dealer to realize a gain from
the contract, the new dealer would now be exposed to credit risk relative
to the original dealer. However, the original dealer's credit risk
exposure to the hedge fund could increase after the assignment if the
trade had been offsetting other trades that the dealer had with the hedge
fund.

25The dealers also required hedge funds and other counterparties to
periodically post additional collateral to cover their credit risk
exposure resulting from changes in the value of the contracts.

Under the Direction of FRBNY as well as Other Regulators, Dealers and Others
Have Worked Collaboratively to Considerably Reduce the Backlog and Address Its
Causes

The unilateral assignments and the increasing backlogs raised regulatory
concerns that prompted U.S. and foreign regulators and the major credit
derivative dealers to seek a collective solution. FSA, which oversees
financial activities in the United Kingdom, took one of the first actions
to address the backlogs by sending dealers a letter in February 2005.27
The letter expressed FSA's concern about dealers' level of unsigned
confirmations and asked them to consider the robustness of their
operational processes and risk management frameworks in the rapidly
evolving credit derivatives market. U.S. regulatory staff told us that
they had been aware of the backlogs since at least 2004 through their
oversight activities and discussions with other regulators. For example,
in 2004, U.S. bank examiners began to identify the growing backlogs of
unconfirmed trades at dealers, including how unilateral assignments were
contributing to such backlogs. Although they began monitoring dealers'
efforts to resolve these issues, the regulators recognized in spring of
2005 that individual dealer efforts to address the practice of unilateral
assignments were proving unsuccessful and that greater automation was
needed. Regulatory staff told us that these unilateral assignments posed a
"collective action" problem, in that dealers could not individually stop
the practice for fear of losing business to other dealers that did not
require counterparties to notify them prior to assigning a trade.
According to regulatory staff, the prevalence of unilateral assignments
was especially troubling because dealers did not always know the
counterparties to their trades, raising questions about dealers' ability
to accurately manage the risks of these activities. In addition,
regulatory staff said that the fact that dealers did not always know their
counterparty's identity raised operational concerns about the ability of
market participants to settle trades, should a large reference entity
default. Finally, regulators noted that resolving the causes of the
backlogs required multilateral regulatory involvement, because no single
regulator oversaw all the dealers.

26The ISDA Master Agreement sets forth standardized terms regulating
general obligations of the parties, events of default, netting, early
termination, transfer, currency provisions, and definitions. The Master
Agreement and its related documentation are designed, among other things,
to allow parties to establish under a single agreement all the
"non-economic" terms--such as representations and warranties, events of
default, and termination events--that will govern each individual
derivative transaction between the parties. The specific "economic" terms
of the individual derivatives contracts--such as the rate or price,
notional amount, maturity, and collateral--are then negotiated and
documented on a transaction-by-transaction basis. The Master Agreement
contained a provision requiring the written consent of the other party
prior to a party's assignment of its rights and obligations under the
transaction to a third party. The Master Agreement, together with any
amendments by the parties, is given effect in confirmations.

27Gay Huey Evans, Director, Markets Division, Capital Markets Sector
Leader, Financial Services Authority, to Chief Executive Officers
(regarding operations and risk management in the credit derivatives
market), February 2005,
http://www.fsa.gov.uk/pages/library/communication/pr/2005/022.shtml
(accessed May 2007).

To address these problems with confirmation backlogs and unilateral trade
assignments, FRBNY convened a meeting in September 2005 with the 14 major
credit derivative dealers and their regulators--referred to as the joint
regulatory initiative. Regulatory representatives from around the
world--including OCC, SEC, FSA, the German Financial Supervisory
Authority, and the Swiss Federal Banking Commission--attended the meeting
as supervisors of at least one of the major dealers involved in the
initiative. At this meeting, the U.S. and foreign regulators discussed how
the dealers would improve assignment practices and resolve the
confirmation backlogs. In October, the dealers sent FRBNY a letter that
outlined the steps to be taken to improve the credit derivatives
industry's practices and confirmation backlogs.28 The plan included

           o establishing target dates and levels by which to reduce the
           confirmation backlogs,
           o increasing the use of electronic confirmations systems,
           o supporting the implementation of a protocol to end unilateral
           assignments,
           o improving the process for settling credit derivatives contracts
           after a credit event, and
           o providing regulators with monthly data for measuring dealers'
           progress.

28Senior management of Bank of America; Barclays Capital; Bear, Stearns &
Co.; Citigroup; Credit Suisse First Boston; Deutsche Bank AG; Goldman,
Sachs & Co.; HSBC Group; JP Morgan Chase; Lehman Brothers; Merrill Lynch &
Co.; Morgan Stanley; UBS AG; and Wachovia Bank; letter to Mr. Timothy
Geithner, President, Federal Reserve Bank of New York, October 4, 2005,
http://www.newyorkfed.org/newsevents/news/markets/2005/an051005.html
(accessed May 2007).

To enable the regulators to monitor the dealers' progress as part of the
joint regulatory initiative, the 14 dealers agreed to collect data on
their credit derivatives activities, including trading volume, unconfirmed
trades, and trades confirmed using automated systems. Under the agreement,
the dealers provide their individual data to Markit Group, a provider of
independent data, portfolio valuations, and OTC derivatives trade
processing. In turn, Markit Group aggregates the data across the dealers
to protect the confidentiality of each dealer's data and then provides the
regulators with aggregate data in a monthly report.

In February 2006, FRBNY hosted a follow-up meeting with the dealers and
their regulators to discuss the progress and stated that it was encouraged
by the progress that had been made. Following the meeting, the dealers
sent FRBNY a letter committing to further improvements in market practices
to "achieve a stronger steady state position for the industry."29 Among
the commitments the dealers made were (1) to ensure that all trades with
standardized terms that were eligible for automated processing would be
processed electronically, and (2) to work with DTCC to create a central
depository to store electronically the details of all credit derivatives
contract terms. In September 2006, FRBNY hosted a third follow-up meeting
with the dealers and regulators to discuss the dealers' progress.

Since the initial meeting in September 2005, the 14 dealers have
significantly reduced the number of outstanding confirmations. As shown in
figure 3, the aggregated data that has been provided to regulators showed
that the 14 dealers had reduced their total number of confirmations
outstanding from 153,860 in September 2005 to 37,306, or by about 76
percent, by the end of October 2006.

29Senior management of Bank of America; Barclays Capital; Bear, Stearns &
Co.; Citigroup; Credit Suisse; Deutsche Bank AG; Goldman, Sachs & Co.;
HSBC Group; JP Morgan Chase; Lehman Brothers; Merrill Lynch & Co.; Morgan
Stanley; UBS AG; and Wachovia Bank; letter to Mr. Timothy Geithner,
President, Federal Reserve Bank of New York, March 10, 2006,
http://www.newyorkfed.org/newsevents/news/markets/2006/an060313.html
(accessed May 2007).

Figure 3: Total Outstanding Confirmations at 14 Major Dealers from
September 2005 to October 2006

Under the joint regulatory initiative organized by FRBNY, each dealer
committed to incrementally reducing its number of confirmations
outstanding more than 30 days by various amounts over the course of the
following 9 months. The data that the dealers have been providing to
regulators showed that they collectively exceeded each of the reduction
goals they had agreed to meet and had reduced by 94 percent the total
number of confirmations outstanding over 30 days from the September 2005
level to the October 2006 level (table 3). The dealers were able to
achieve this reduction even though their monthly trading volume in credit
derivatives generally increased during this period.

Table 3: Outstanding Confirmation Reduction Goals and Totals for the 14
Credit Derivatives Dealers

                     Outstanding                                              
                    confirmation   Trades unconfirmed for more than   Percent 
As of date     reduction goal                            30 days reduction 
September 30,                                             97,650           
2005                                                                       
January 31,               30%                             45,288       54% 
2006                                                                       
April 30, 2006            50%                             27,405       72% 
June 30, 2006             70%                             15,997       84% 
October 31,                                                5,558       94% 
2006                                                                       

Source: GAO analysis of Markit Group data.

After October 2006, four additional foreign dealers have joined the
original 14 dealers in providing monthly confirmation backlog and related
data to Markit Group for aggregation and distribution to the regulators.
As shown in table 4, with the inclusion of the additional dealer data, the
total number of outstanding confirmations over 30 days has increased in
comparison to the level at the end of October 2006, especially in March
2007. At the same time, table 4 shows that monthly trading volume has
increased beginning in January 2007, and the number of confirmations
outstanding more than 30 days as a share of the total number of
outstanding confirmations has decreased slightly during this period. U.S.
regulatory staff characterized the rise in the confirmation backlog as
modest and attributed it generally to the increase in trading volume and
noted that the automation of the confirmation process has helped dealers
handle the increased volume.

Table 4: Outstanding Confirmations and Related Data Provided by Dealers
from October 2006 to March 2007

               Number of   Total   Total number of  Confirmations outstanding 
                 dealers monthly       outstanding     more than 30 days as a 
               providing trading     confirmations share of total outstanding 
Time period      data  volume more than 30 days              confirmations 
October            14 190,849             5,558                        15% 
2006                                                                       
November           16 185,352             5,802                        16% 
2006                                                                       
December           17 139,649             8,282                        25% 
2006                                                                       
January            18 216,850             6,784                        16% 
2007                                                                       
February           18 234,155             7,380                        13% 
2007                                                                       
March 2007         18 347,061            11,940                        14% 

Source: GAO analysis of Markit Group data.

Dealers Reduced Backlogs through Various Steps

To achieve these reductions in their unconfirmed trade backlogs, dealers
took various steps. For example, dealers engaged in events called "lock
ins" with other dealers and, to a lesser extent, end-users. Under a lock
in, operations staff from either two dealers or staff from one dealer and
one of their key end-user customers convened in a room and compared the
trades they had conducted together until all or almost all were reconciled
and confirmed. Dealers and end-users also used "tear-up services" to
reduce the total number of open trades and thus eliminate the number of
trades that needed to be confirmed. In a tear-up process, an automated
system matches up offsetting positions across many market participants,
allowing those trades to be, in effect, terminated and thereby removing
the need to confirm such trades.

To prevent new trades from adding to the backlog, the dealers also
increased their use of automated confirmation systems and set deadlines
for confirming trades. First, as part of the joint regulatory initiative,
the 14 major dealers committed to use DTCC's automated system, Deriv/SERV,
to confirm trades made with other dealers by the end of October 2005 and
to require their active clients to use it or a comparable automated
system, such as SwapsWire, by mid-January 2006.30 As shown in figure 4,
the share of the total monthly trades confirmed electronically increased
from 46 percent to 85 percent between the end of September 2005 and the
end of October 2006.31 Moreover, at the end of October 2006, the dealers
collectively had 3,900 active clients--of which 98 percent, on average,
were using an automated confirmation system or were in the process of
subscribing to one.

Figure 4: Share of Total Monthly Credit Derivatives Trades of the 14
Dealers Confirmed Electronically, September 2005 to October 2006

Second, the 14 dealers committed to electronically confirming all trades
that could be confirmed electronically (i.e., contracts with standardized
terms) within 5 business days of the trade date, by the end of October
2006. Deriv/SERV has continually expanded its capabilities to
electronically confirm not only a wider range of products but also changes
to existing contracts, including assignments. At the end of October 2006,
about 90 percent of the total trades were eligible to be confirmed
electronically, and about 94 percent of those eligible trades were
electronically confirmed, according to the data provided by the 14
dealers. Of the trades confirmed electronically, 84 percent, on average,
were confirmed within the stipulated 5 business days.32 In addition, the
industry has taken steps to help ensure that new products do not create
backlogs. Officials at DTCC and ISDA said that they have formed industry
working groups and revised certain procedures to reduce the time it takes
to standardize the legal documentation for new credit derivatives, in turn
enabling these products to be confirmed electronically by Deriv/SERV.

30The term "active client" was initially defined as a client that executed
five or more Deriv/SERV-eligible trades a week, on average, for the past 3
months with an individual dealer. The standards were changed at the end of
March 2006 to mean a client that executed one Deriv/SERV-eligible trade or
more a week, on average, for the past 3 months with an individual dealer.

31At the end of March 2007, 86 percent of the total credit derivatives
trades were confirmed electronically, based on data provided by 18 dealers
to Markit Group.

Market Participants Agreed to End Unilateral Assignments and Thereby Addressed a
Key Factor Contributing to the Backlogs

An additional step taken to prevent further confirmation backlogs was to
end the practice of unilateral assignments, which ISDA and market
participants had been attempting to address since at least 2002. For
example, ISDA published a novation agreement to document assignments in
2002, issued provisions governing credit derivatives assignments as part
of its 2003 documentation standards for credit derivatives, and issued
novation definitions and guidance on best practices for assignments in
2004. Despite such efforts, an ISDA working group found that market
participants were using different practices to process assignments,
increasing risks to counterparties and creating operational inefficiency
and backlogs in processing trades.

ISDA officials told us that in early 2005 they had the working group start
(1) to develop a protocol to streamline practices for dealers and
end-users to follow to assign a trade and (2) to reach out to end-users as
part of the effort. In the summer of 2005, the working group began
circulating a draft protocol for comment. Shortly before the regulators
initiated their joint action in September 2005, ISDA issued its voluntary
Novation Protocol, and major dealers signed up for it. In their October
2005 letter to FRBNY, the dealers committed to finalizing a guide to
support the protocol's implementation. By signing the protocol, a party
seeking to assign a trade agrees to obtain the consent of its original
counterparty through e-mail or other electronic means. Although the major
dealers signed the protocol in September, some end-users were initially
reluctant to sign, in part because they were concerned that dealers would
not be able to consent to assignments promptly. In response, all the major
dealers agreed to reply to assignment requests within 2 hours. By November
30, 2005, 2,000 market participants had signed the protocol.33 Most of the
major hedge funds have signed the protocol, according to officials from
the Managed Funds Association, which represents the majority of the
largest hedge funds.

32At the end of March 2007, about 92 percent of the total trades were
eligible to be confirmed electronically, and about 93 percent of these
eligible trades were confirmed electronically based on data provided by 18
dealers to Markit Group. Of the trades confirmed electronically, about 86
percent on average were confirmed within 5 business days after the trade
date.

According to some dealers and U.S. financial regulators, the widespread
adoption of the ISDA Novation Protocol has effectively ended the practice
of unilateral assignments, eliminating a key factor that had contributed
to the backlogs. Because the vast majority of assignments become
dealer-to-dealer trades, the protocol enables dealers to monitor each
other to ensure that clients are complying with it. If a dealer were to
allow its client to assign a trade without obtaining the original dealer's
consent, the original dealer would discover the compliance failure when it
discovered the assignment. To facilitate the confirmation of assignments,
Deriv/SERV also expanded its system in mid-2005 to electronically confirm
assignments. As a result of the ISDA protocol and automation of the
assignment process, the number of unconfirmed assigned trades outstanding
for more than 30 days declined from around 39,500 at the end of September
2005 to around 940 at the end of October 2006 (fig. 5), even though the
number of trades being assigned during this period generally increased.34
From the end of September 2005 to the end of October 2006, the share of
assignments confirmed electronically has increased from 24 percent to 82
percent.

33ISDA subsequently issued the Novation Protocol II to allow new
participants to the credit derivatives market to obtain the benefits of
the original protocol. As of May 9, 2007, 268 market participants have
adhered to the new protocol.

34After October 2006, Markit Group stopped providing data on outstanding
confirmations for assigned trades.

Figure 5: Total Monthly Volume of Assigned Trades and Number of
Unconfirmed Assignments Outstanding for More Than 30 Days for 14 Major
Dealers, September 2005 and October 2006

Dealers Found Benefits in the Joint Regulatory Initiative

The dealers and other market participants we interviewed uniformly viewed
the joint regulatory initiative as instrumental in reducing the backlog,
automating the credit derivatives market's infrastructure, and bringing
the industry together to address the confirmation backlog problem. The
market participants noted that regulatory support was crucial in
encouraging cooperation among dealers and end-users to address problems
related to the confirmation backlog. Specifically, they said that
regulators' involvement helped to persuade certain end-users to agree to
adhere to ISDA's Novation Protocol. In addition, they told us that the
joint regulatory initiative catalyzed industry efforts to move to
automated confirmation matching services such as DTCC's
Deriv/SERV--bringing about automation sooner than it otherwise would have
occurred. Such intervention expedited the adoption of automated tools by
end-users, enhancing dealers' efforts to implement such tools as
Deriv/SERV. Additionally, the joint initiative led to the formation of a
group composed of dealers that meets weekly to discuss, among other
things, operational issues. Two dealers told us that this group has helped
to resolve problems in processing confirmations and also allowed the
dealers to hear the views of end-users.

Dealers and Other Market Participants Continue to Work to Reduce Backlogs,
Diminish Operational Risks, and Improve Market Infrastructure

While the dealers have made significant progress since 2005, they have
continued their efforts to reduce backlogs and improve the infrastructure
of the credit derivatives market. First, in addition to committing to
confirm virtually all standardized trades electronically within 5 business
days of the trade date, the dealers have committed to confirming all
nonstandardized trades within 30 days after trade date. Because
nonstandardized trades are complex and customized, such trades must be
confirmed manually, according to ISDA officials. According to data
provided by the dealers to regulators over the last 3 months, these trades
have accounted for less than 10 percent of the total credit derivatives
trading volume. According to regulators and dealers, these trades are
generally complex and involve issues that require time to be legally
negotiated before the trades can be confirmed. However, Federal Reserve
and OCC staff have expressed concern that taking 30 days to confirm
nonstandardized trades is too long and are continuing to work with dealers
to reduce the confirmation time. The 14 dealers have made considerable
progress in promptly confirming their nonstandardized trades, reducing the
number of such trades remaining unconfirmed for more than 30 days from
around 5,600 at the end of September 2005 to fewer than 440 by the end of
October 2006. However, as the four additional dealers began providing
their data after October 2006, the number of unconfirmed nonstandardized
trades rose, reaching around 6,800 at the end of March 2007. To mitigate
the risk of any unconfirmed nonstandardized trades, the dealers have
committed to verifying the key economic terms of such trades informally
within 3 business days of the trade date. As of the end of March 2007,
dealers were meeting this commitment, on average, for around 54 percent of
their nonstandardized trades.35

Second, under the joint regulatory initiative, the dealers have worked to
reduce operational risks by committing to improvements in the settlement
process for credit default swaps. For example, credit default swaps
generally require that the purchaser of credit protection under a credit
default swap deliver the bonds (or loans) referenced in the contract to
the counterparty if the bond issuer goes bankrupt. In exchange, the
counterparty pays the par, or face, value of the bonds to the protection
purchaser. This settlement method avoids difficulties that could arise
when the bonds are valued after a bankruptcy.36 However, situations can
arise in which the amount of bonds needing to be delivered exceeds the
amount of outstanding securities.37 For example, when the auto parts maker
Delphi filed for bankruptcy in 2005, credit derivatives on its bonds and
loans totaled an estimated $28 billion in notional amount, but Delphi had
only $5.2 billion in bonds and loans outstanding. In addition to
increasing the difficulty of meeting the delivery obligation under a
credit derivatives contract, a temporary shortage of bonds also could
cause the price of the needed securities to increase immediately following
a default. To facilitate settlement in this type of situation, ISDA has
developed protocols to allow contracts to be settled in cash rather than
by delivery of the debt. Under this process, the bond's price is
established through an auction, and the counterparties providing credit
risk protection pay their counterparties in cash based on the difference
between the bond's auction price and par value. Since 2005, ISDA has used
its protocols to facilitate cash settlement in seven credit events
involving U.S. firms. The protocols covering the first six credit events
enabled cash settlement of only credit default swap indexes. In its most
recent form, the protocol permits cash settlement in index, single-name,
and certain other credit derivatives. ISDA plans to include the cash
settlement mechanism in its revised documentation standards for credit
derivatives in 2007.

35Although 18 dealers provided data to Markit Group on the number of their
nonstandardized trades not confirmed within 30 days, only 12 of these
dealers provided data on the extent to which they were verifying the
economic terms of such trades within 3 business days after the trade.

Finally, DTCC is working with dealers and end-users to implement a central
trade depository to automate trade processes other than confirmation and
thus reduce operational risks. Under the joint regulatory initiative, the
dealers committed to work with DTCC to create (1) a database to
electronically store the official legal record of all credit derivative
contracts eligible for automated confirmation, taking into account
subsequent changes made to the contracts, such as assignments, and (2) a
central infrastructure supporting the warehouse to standardize and
automate post-trade processes over the life of each contract.38
Specifically, the warehouse will support premium payment calculations and
facilitate not only the bilateral payment settlement of electronically
confirmed trades but also reconciliations for collateral management and
credit event processing. DTCC launched the warehouse in November 2006,
with all new trades electronically confirmed through Deriv/SERV
automatically loaded into the warehouse. In addition, dealers are
inputting their trade data for their existing credit derivatives contracts
into the warehouse to create a complete database, and this effort is
expected to continue through 2007. Two dealers told us that the trade
input process is an extensive project, because around a million trades
need to be inputted. DTCC also plans to expand the warehouse to support
central payment calculation and settlement capability in 2007.

36Since 1998, ISDA documentation for credit derivatives has provided
counterparties with the option of settling their contracts in cash rather
than by physical delivery. Under the cash-settlement option, price
quotations are obtained for the referenced debt and used to determine the
amount of the settlement payment. Most market participants have preferred
the physical delivery option because of concerns about being able to
obtain accurate quotations after a credit event.

37Because counterparties need not own the debt referenced in a credit
derivative, they can enter into an essentially unlimited number of credit
derivative contracts referencing such debt. In comparison, the amount of
outstanding debt issued by a firm is fixed at any particular time.

The Joint Regulatory Initiative Enhanced U.S. Regulators' Oversight of Dealers'
Efforts to Reduce Their Backlogs

U.S. financial regulators were overseeing the operational and other risks
at individual credit derivatives dealers through their continuous
supervision and examinations. After the joint regulatory initiative, the
industrywide data from the major dealers provided the regulators with more
effective means of monitoring the resolution of the backlog and related
problems.

38The potential expansion to other OTC derivative products will depend on
market demand and input from the senior group working with DTCC.

Federal Bank Regulators Were Monitoring Efforts Taken by Banks to Address
Confirmation Backlogs through Continuous Supervision and Examinations

The Federal Reserve and OCC were aware of the confirmation backlogs at
banks and were monitoring efforts to address them before the joint
regulatory initiative. Of the 14 dealers participating in the joint
regulatory initiative, nine are U.S. or foreign banks. Five of the banks
are chartered as national banks and individually supervised and examined
by OCC through its teams of examiners. Each of these banks is a subsidiary
of a bank holding company or financial holding company supervised by the
Federal Reserve.39 OCC staff told us that the five U.S. banks conduct
around 90 percent of their credit derivatives activities within the bank,
not in their holding companies or other subsidiaries. As a result, OCC
bank examiners are primarily responsible for overseeing the banks' credit
derivatives activities but coordinate their oversight with their Federal
Reserve counterparts. Federal Reserve officials told us that their
examiners also oversee the U.S. operations of the foreign banks that are
major dealers participating in the joint initiative.

At the four national banks and one foreign bank we reviewed, management
had become aware of the confirmation processing and backlog problems at
their own banks primarily through internal audit reports or management
information reports tracking outstanding confirmations. The timeframes in
which the problems surfaced at the banks and were brought to management's
attention varied, with one bank's management learning about the problems
in late 2003 and another bank's management not until the summer of 2005.
Nonetheless, according to examiners of these banks, as bank management
became aware of these problems, they provided these audit or management
reports or had discussions with the bank examiners supervising their
institutions. Bank examiners told us that they continually supervise how
the banks are identifying, monitoring, and managing their operational,
credit, market, and other risks posed by credit derivatives and other
products through reviews of internal audit and management reports,
meetings with key bank officials, and examinations.

39The Bank Holding Company Act of 1956, as amended, generally requires
that holding companies with bank subsidiaries register with the Federal
Reserve as bank holding companies. Among other things, the act restricts
the activities of the bank holding companies to those the Federal Reserve
determined, as of November 11, 1999, to be closely related to banking.
Under amendments to the act made by the Gramm-Leach-Bliley Act, a bank
holding company can qualify as a financial holding company and may engage
in a broad range of additional financial activities, such as securities
and insurance underwriting.

After learning of the backlog problems at the banks, the examiners said
that they monitored each bank's efforts to address the processing problems
and reduce the backlog, such as by periodically reviewing reports tracking
the backlog, meeting with bank management and staff, or conducting
examinations. Through their supervision, for example, the examiners
reviewed the level of resources the banks were devoting to processing
their credit derivatives trades. They also examined to varying degrees the
credit derivatives confirmation process of four of the five banks between
2004 and 2006. For example, in 2004, examiners reviewed the progress that
two banks were making in reducing their confirmation backlogs and in
addressing the causes of the backlogs, including assignments of credit
derivative trades. Based on examinations done in 2005, examiners directed
two banks to develop plans to ensure that their infrastructures were
capable of supporting the trading volume of credit derivatives, and the
examiners said that the banks had developed such plans. In addition to
focusing on the confirmation backlogs, examiners generally examined how
well each of the five banks was managing its market, credit, and other
risks associated with its credit derivatives activities. The examiners did
not examine one bank's confirmation process because the bank was in the
process of implementing a plan to address its backlog, but examiners
monitored the bank's progress through informal reviews.

Bank regulators were also reviewing the banks' efforts to ensure the
security and resiliency of their information technology systems. Managers
at the four national banks we interviewed described taking various steps
to ensure the security of their credit derivatives systems. For example,
the systems used at these banks included restrictions on who could input
or access data in the systems. Managers at these banks also were
responsible for periodically reviewing and testing their staff's access
rights to the systems to ensure that they were appropriate. According to
bank staff, the security controls were reviewed or tested regularly by
internal and external auditors--for example by conducting penetration
tests in which auditors would attempt to obtain unauthorized access to the
systems. In addition, the bank officials told us that they have taken
steps to ensure the resiliency of their systems, including processing
their credit derivatives in several different locations, creating off-site
backup facilities, and developing disaster recovery plans. The examiners
of these banks told us that they had tested or reviewed whether the banks
were complying with controls designed to protect the security and
resiliency of their information technology systems. For example, examiners
told us that they reviewed managers' oversight of their staff's access
rights to the systems and checked for testing of business continuity
plans.

SEC Also Conducted Oversight of Credit Derivatives Confirmation Backlogs at
Major Broker-Dealers

Unlike the bank regulators, SEC only recently began providing oversight of
the credit derivatives activities of broker-dealers because such
activities have generally been conducted in affiliates not subject to SEC
regulation. According to SEC staff, the five U.S. broker-dealers that are
active in the credit derivatives market generally book their trades in
unregulated affiliates that are not subject to SEC supervision because
they are not registered, nor required to be registered, with SEC.40
However, in June 2004 SEC instituted its Consolidated Supervised Entity
(CSE) program, under which large broker-dealers may qualify for
alternative net capital rules in exchange for consenting to supervision on
a consolidated basis by SEC.41 The five U.S. broker-dealers engaged in
credit derivatives trading applied for and were granted CSE status. Under
the SEC's CSE program, SEC supervises the broker-dealers on a consolidated
basis, with its prudential supervision extending beyond the broker-dealers
to their unregulated affiliates and holding companies.42 The five
broker-dealers participating in the CSE program are also participating in
the joint regulatory initiative.

Although aware that backlogs for OTC derivatives were an issue, SEC staff
became aware of the extent of the credit derivatives backlogs at U.S.
broker-dealers through continuous supervision and examinations conducted
after these firms applied to the CSE program. SEC officials noted that
although they were generally aware of the backlog in confirming credit
derivatives through a study of the credit derivatives market conducted by
the Joint Forum in 2004, they were surprised at the extent of the problem
by the summer of 2005.43 According to SEC staff, in 2005 risk managers and
internal auditors at the CSE broker-dealers told SEC staff about the
confirmations backlog and its potential impact on the credit derivatives
market. Broker-dealer staff, during the summer of 2005, were periodically
discussing with SEC the resources they were devoting to reducing their
backlogs and the associated risks. For example, one firm devoted about 30
full-time staff and 20 consultants to reducing its backlog, according to
SEC staff. Supplementing this information about the confirmations backlog,
examinations conducted as part of the CSE application process also
assisted SEC in learning more about the nature of the backlog and related
concerns. As part of the application process, SEC examined the firms'
internal risk management systems and controls, issuing examination reports
from November 2004 through January 2006. SEC targeted credit derivatives
products within the scope of all but one of its five application
examinations, choosing products that posed the greatest risks and
represented the highest volume in the firms.44 Examination findings
related to the credit derivatives confirmation backlog included delays in
issuing confirmations promptly after the trade date and discrepancies
between confirmation documentation and output data from systems used to
input trades. Broader examination findings included concerns that internal
audits at some firms did not always document processes or sufficiently
follow up on recommendations and that some firms did not accurately
compute counterparty credit ratings, in some cases for hedge fund
counterparties. The findings were shared with the firms, and SEC has
monitored the firms' implementation of its recommendations.45

40Affiliates of broker-dealers that do not engage in the securities
business within the United States are not required to register with SEC as
broker-dealers.

4169 Fed. Reg., 34428 (June 21, 2004). The rule release states that the
rule amendments respond, in part, to international developments.
Affiliates of certain U.S. broker-dealers that conduct business in the
European Union (EU) have stated that they must demonstrate that they are
subject to consolidated supervision at the ultimate holding company level
that is "equivalent" to EU consolidated supervision. SEC supervision
incorporated into these rule amendments is intended to meet this standard.

42According to the SEC, under this program, SEC staff conduct various
supervisory activities with respect to firms subject to the program,
including reviewing monthly, quarterly, and annual filings; holding
monthly meetings with senior management at the holding company; and
conducting examinations of the holding company, the broker-dealer, and
material affiliates not subject to supervision by a principal regulator.

43The Joint Forum--comprising a group of international bank,
broker-dealer, and insurance company supervisors that includes SEC--issued
a 2004 study of the credit derivatives market. The study reviews financial
stability issues associated with credit derivatives and recommends, among
other things, that confirmations be promptly executed after completing a
transaction and that the industry issue clear guidance on the time
required to issue and receive confirmations.

44According to an SEC examination official, credit derivatives were not
covered for one firm for a number of reasons. First, the selection of
products to review for this particular examination came before the
widespread concern about the backlog in credit derivative confirmations
was known. Second, other products were also high volume and high risk at
the firm. Third, most of the credit derivative trades were booked in this
firm's London subsidiary and therefore regulated by FSA. The goal of the
examinations was to capture information about unregulated affiliates of
the firm that had not previously been subject to review by any regulator.

In addition to overseeing firms' credit derivatives backlogs, SEC staff
told us that their CSE broker-dealer examinations conducted at the time of
application also addressed the security and resiliency of these firms'
information technology systems, including those that are used for credit
derivatives activities. For example, at the broker-dealers, SEC staff
reviewed reports by firms' internal audit departments on security and
resiliency of information technology systems in general, as some of these
systems handled credit derivatives transactions. In addition, SEC
examinations included business continuity planning reviews based on draft
interagency standards on protecting the resiliency of the U.S. financial
system.

Through the Joint Regulatory Initiative, Regulators Are Obtaining Data to More
Effectively Track Industrywide Progress on Reducing Confirmation Backlogs

Although both U.S. banking and securities regulators were individually
overseeing aspects of U.S. dealers' credit derivatives activities, the
joint regulatory initiative provided U.S. and foreign regulators with
information that enabled them to better oversee the progress being made by
the major dealers to address the backlog issue. While the individual
regulators had data on the backlogs at the dealers under their
supervision, no one regulator supervised all 14 major dealers and thus had
data on the size of the problem across all dealers. Under the joint
regulatory initiative, U.S. financial regulatory staff said that they told
the 14 dealers what information they needed in order to track dealers'
progress in addressing the backlog and related problems. Based on the
capabilities of their management information systems, the dealers
collectively developed a template to collect standardized metrics. The
data include information on trading volume, trade assignments, trades
confirmed electronically and manually, and confirmations outstanding based
on length of time the trades remained unconfirmed. Under the arrangement,
each dealer provides the standardized data to its primary regulator at the
end of the month. For example, the U.S. broker-dealers provide their data
to SEC, and the national banks provide their data to OCC. In addition,
each dealer provides its data to Markit Group, which aggregates the data
across all the dealers to preserve the confidentiality of each dealer's
data and computes averages for the metrics, such as the average number of
outstanding confirmations for the dealers. In turn, Markit Group provides
the U.S. and foreign regulators and dealers participating in the joint
regulatory initiative with a set of the aggregate data.

45While the SEC examinations did not initially focus on the confirmation
backlog specifically, this was included as part of the examination process
when SEC became aware that the backlog in credit derivatives was an
industrywide concern.

According to U.S. and foreign financial regulators, the aggregate and
individual dealer data have provided regulators with an effective tool for
tracking overall and individual dealer progress. According to U.S.
regulators, using a template to standardize data collection has helped to
ensure the comparability of the data across the dealers. The regulators
also told us that the data are critical to the joint regulatory
initiative, because the combined data provide transparency, enabling the
regulators to track the progress of individual dealers under their
supervision and helping each dealer to see how well it is doing relative
to the average. Similarly, FSA officials told us that the aggregate data
has provided regulators with a simple way to monitor the backlog level for
the entire market and to compare individual dealers' backlog levels
against the average. The officials also said that the common set of
measures has helped to instill discipline among the dealers.

Under the joint regulatory initiative organized by FRBNY, the U.S. and
foreign regulators are continuing to monitor the credit derivatives market
and have expanded their efforts in September 2006 to address confirmation
backlogs in the market for OTC derivatives based on equities.46 According
to dealers, it takes longer to confirm an OTC equity derivative trade than
any other type of OTC derivative trade, because such trades are processed
largely through manual rather than automated means because of the lack of
standardized trade documentation. Based on data provided by the major
dealers to the regulators, they had over 81,000 unconfirmed trades at the
end of November 2006, with around 31,000, or 54 percent, of these trades
remaining unconfirmed for over 30 days. In a November 2006 letter to
FRBNY, 17 dealers committed to working with industry organizations to
improve the efficiency of the equity derivatives market, in part through
the greater adoption of automation.47 Among other things, the dealers
committed (1) to reduce by 25 percent the number of unconfirmed trades
outstanding more than 30 days by the end of January 2007 based on dealers'
highest level of outstanding confirmations from July to September 2006 and
(2) to use at least one industry-accepted electronic confirmation service
and one other such platform by the end of March 2007. U.S. regulatory
staff told us that dealers met the first goal but that, as of April 2007,
one dealer had not yet fully met the second goal. At the end of March
2007, the number of unconfirmed equity derivative trades outstanding more
than 30 days rose to around 43,000 trades. U.S. regulatory staff said that
the increase generally resulted from the inability of the manual processes
used by dealers and end-users to confirm trades to keep pace with the
increase in trading volume. The dealers also agreed to continue to provide
the U.S. and foreign regulators with standardized data not only on credit
derivatives but also on OTC equities, interest rate, foreign exchange, and
commodity derivatives.48 U.S. regulators said that they wanted to track
data across the major OTC derivatives products to ensure that work done in
connection with equity derivatives does not hamper the ability of the
dealers to process their other OTC derivative trades in a timely manner.
Such data will assist regulators in monitoring the operational and other
risks raised by OTC derivative products.

46OTC equity derivatives are financial contracts whose value is derived
from an underlying equity or equity index. For example, an equity or
equity index swap is a transaction in which one party pays periodic
amounts based on a fixed price or rate and the other party pays periodic
amounts based on the performance of an individual equity or equity index,
such as the Standard and Poor's 500 Index.

Given that individual efforts could not fully resolve the backlog problem,
U.S. and foreign regulators we interviewed said that the joint regulatory
initiative proved instrumental in ensuring that the problem was addressed.
According to representatives of FSA, bringing together the various
financial regulators from throughout the world was an approach that worked
very well to ensure collaboration among regulatory bodies. Similarly, U.S.
bank examiners told us that the joint regulatory initiative served an
important role in getting the dealers to work collectively and by
providing a level regulatory playing field. U.S. and foreign regulators
and dealers are already applying this model to address similar issues in
the OTC equity derivatives market.

47As identified in an attachment to the Federal Reserve Bank of New York's
November 21, 2006, press release, the 17 dealers are Bank of America,
N.A.; Barclays Capital; Bear, Stearns & Co.; BNP Paribas; Citigroup;
Credit Suisse; Deutsche Bank AG; Dresdner Kleinwort; Goldman, Sachs & Co.;
HSBC Group; JP Morgan Chase; Lehman Brothers; Merrill Lynch & Co.; Morgan
Stanley; Societe Generale; UBS AG; and Wachovia Bank, N.A.

48The dealers are providing standardized data on their OTC credit, equity,
and interest rate derivative trades to regulators through Markit Group. A
foreign exchange committee sponsored by the Federal Reserve Bank of New
York is separately monitoring the foreign exchange derivatives market.
Interest rate, foreign exchange, and commodity derivatives are financial
contracts whose value is respectively derived from an underlying (1)
interest rate, such as the London Interbank Offered Rate (the interest
rate paid on interbank deposits in the international money markets); (2)
currencies, such as the U.S. dollar and Canadian dollar; and (3)
commodities, such as natural gas or gold.

Agency Comments

We provided a draft of this report to the Federal Reserve, OCC, and SEC
for their review and comment. The Federal Reserve and SEC provided
technical comments, which we incorporated as appropriate.

As agreed with your offices, unless you publicly announce its contents
earlier, we plan no further distribution of this report until 30 days
after the date of this report. At that time, we will send copies of this
report to other interested congressional committees and the Chairman,
Federal Reserve; the Comptroller of the Currency; and the Chairman, SEC.
We will also make copies available to others upon request. The report will
be available at no charge on the GAO Web site at [27]http://www.gao.gov .

If you or your staff have any questions regarding this report, please
contact me at (202) 512-6878 or [28][email protected] . Contact points for
our Offices of Congressional Relations and Public Affairs may be found on
the last page of this report. Key contributors to this report are listed
in appendix II.

Yvonne Jones
Director, Financial Markets and Community Investment

Congressional Requesters

The Honorable John D. Dingell, Chairman
The Honorable Joe Barton, Ranking Member
Committee on Energy and Commerce
House of Representatives

The Honorable Edward J. Markey, Chairman
The Honorable Fred Upton, Ranking Member
Subcommittee on Telecommunications and the Internet
Committee on Energy and Commerce
House of Representatives

The Honorable Bobby L. Rush, Chairman
The Honorable Cliff Stearns, Ranking Member
Subcommittee on Commerce, Trade, and Consumer Protection
Committee on Energy and Commerce
House of Representatives

The Honorable Jan Schakowsky
House of Representatives

Appendix I: Scope and Methodology

To identify what caused the credit derivatives dealers' trade confirmation
backlogs and how the backlogs are being addressed, we analyzed credit
derivatives trading volume, confirmation backlog, and other transaction
data provided by the major dealers to Markit Group, a provider of
independent data, portfolio valuations, and over-the-counter derivatives
trade processing. We also analyzed operations and other data that dealers
provided to the International Swaps and Derivatives Association (ISDA), a
global over-the-counter derivatives trade association. We conducted data
reliability assessments for the Markit Group and ISDA data and determined
that the data were sufficiently reliable for our purposes. We also
reviewed reports and relevant publications from industry associations,
industry working groups, international organizations, companies, and
academics on the credit derivatives market. Of the 14 dealers
participating in the joint regulatory initiative, we interviewed
operations and other staff from three U.S. banks, one foreign bank, and
four U.S. securities broker-dealers.1 We selected these dealers to ensure
that we included a range of characteristics, based on type of regulator
(bank or broker-dealer), trading volume (high or low), and headquarters
location (United States or foreign). We also interviewed staff from the
Federal Reserve, including its examiners for two banks; the Office of the
Comptroller of the Currency (OCC), including its examiners for three
banks; the Securities and Exchange Commission (SEC); and the U.K.'s
Financial Services Authority (FSA). We reviewed examinations conducted
between 2004 and 2006 and other supervisory materials covering the eight
dealers we interviewed and two other dealers that also participated in the
joint regulatory initiative. In addition, we interviewed representatives
from industry associations, including ISDA, the Managed Funds Association
(representing hedge funds) and the Securities Industry and Financial
Markets Association (representing securities firms, banks, and asset
managers).2 Finally, we interviewed officials from the Depository Trust
and Clearing Corporation about its automated services for processing
credit derivative trades and an official from a hedge fund.

1As identified in an attachment to the Federal Reserve Bank of New York's
September 15, 2005, press release, the 14 dealers are Bank of America;
Barclays Capital; Bear, Stearns & Co.; Citigroup; Credit Suisse First
Boston; Deutsche Bank; Goldman Sachs Group; HSBC; JP Morgan Chase; Lehman
Brothers; Merrill Lynch & Co.; Morgan Stanley; UBS; and Wachovia Bank.

2The Securities Industry and Financial Markets Association is the result
of a merger between the Securities Industry Association and the Bond
Market Association.

To determine how U.S. financial regulators were overseeing the dealers'
operational risk, including related information technology systems
associated with credit derivatives activities, we reviewed examination
manuals and other supervisory or regulatory guidance prepared by the
Federal Reserve, the OCC, and the SEC. We also reviewed and analyzed
supervisory strategies prepared and examinations conducted between 2004
and 2006 by the Federal Reserve and the OCC on the credit derivatives
activities of five banks participating in the joint regulatory initiative.
In addition, we reviewed and analyzed examinations conducted between 2004
and 2006 by SEC covering the holding companies of the five securities
broker-dealers participating in the joint regulatory initiative. Also, we
interviewed staff at the Federal Reserve, OCC, and SEC participating in
the joint regulatory initiative. We also interviewed Federal Reserve or
OCC examiners assigned to supervise and examine five of the banks
participating in the joint regulatory initiative and SEC staff who
examined the holding companies of the five securities broker-dealers
participating in the joint regulatory initiative. Finally, we interviewed
FSA officials to understand their efforts in identifying the confirmation
backlogs and in participating in the joint regulatory initiative.

We conducted our work in Charlotte, North Carolina; Chicago; New York; and
Washington, D.C., from August 2006 to March 2007 in accordance with
generally accepted government auditing standards.

Appendix II: GAO Contact and Staff Acknowledgments

GAO Contact

Yvonne D. Jones (202) 512-8678 or [email protected]

Staff Acknowledgments

In addition to the contact named above, Cody Goebel, Assistant Director;
Robert Lee; Paul Thompson; Marc Molino; Emily Chalmers; and Richard
Tsuhara made key contributions to this report.

This is a work of the U.S. government and is not subject to copyright
protection in the United States. It may be reproduced and distributed in
its entirety without further permission from GAO. However, because this
work may contain copyrighted images or other material, permission from the
copyright holder may be necessary if you wish to reproduce this material
separately.

GAO's Mission

The Government Accountability Office, the audit, evaluation and
investigative arm of Congress, exists to support Congress in meeting its
constitutional responsibilities and to help improve the performance and
accountability of the federal government for the American people. GAO
examines the use of public funds; evaluates federal programs and policies;
and provides analyses, recommendations, and other assistance to help
Congress make informed oversight, policy, and funding decisions. GAO's
commitment to good government is reflected in its core values of
accountability, integrity, and reliability.

Obtaining Copies of GAO Reports and Testimony

The fastest and easiest way to obtain copies of GAO documents at no cost
is through GAO's Web site ( [29]www.gao.gov ). Each weekday, GAO posts
newly released reports, testimony, and correspondence on its Web site. To
have GAO e-mail you a list of newly posted products every afternoon, go to
[30]www.gao.gov and select "Subscribe to Updates."

Order by Mail or Phone

The first copy of each printed report is free. Additional copies are $2
each. A check or money order should be made out to the Superintendent of
Documents. GAO also accepts VISA and Mastercard. Orders for 100 or more
copies mailed to a single address are discounted 25 percent. Orders should
be sent to:

U.S. Government Accountability Office 441 G Street NW, Room LM Washington,
D.C. 20548

To order by Phone: Voice: (202) 512-6000
TDD: (202) 512-2537
Fax: (202) 512-6061

To Report Fraud, Waste, and Abuse in Federal Programs

Contact:

Web site: [31]www.gao.gov/fraudnet/fraudnet.htm
E-mail: [32][email protected]
Automated answering system: (800) 424-5454 or (202) 512-7470

Congressional Relations

Gloria Jarmon, Managing Director, [33][email protected] (202) 512-4400 U.S.
Government Accountability Office, 441 G Street NW, Room 7125 Washington,
D.C. 20548

Public Affairs

Paul Anderson, Managing Director, [34][email protected] (202) 512-4800
U.S. Government Accountability Office, 441 G Street NW, Room 7149
Washington, D.C. 20548

[35]www.gao.gov/cgi-bin/getrpt?GAO-07-716 .

To view the full product, including the scope
and methodology, click on the link above.

For more information, contact Yvonne Jones at (202) 512-8678 or
[email protected].

Highlights of [36]GAO-07-716 , a report to congressional requesters

June 2007

CREDIT DERIVATIVES

Confirmation Backlogs Increased Dealers' Operational Risks, but Were
Successfully Addressed after Joint Regulatory Action

Over-the-counter (OTC) credit derivatives are privately negotiated
contracts that allow a party to transfer the risk of default on a bond or
loan to another party without transferring ownership. After trading in
these products grew dramatically in recent years, backlogs of thousands of
trades developed for which dealers had yet to formally confirm the trade
terms with end-users--such as hedge funds, pension funds, and insurance
companies--and other dealers. Not confirming these trades raised the risk
that losses could arise.

GAO was asked to review (1) what caused the trade confirmation backlogs
and how they were being addressed and (2) how U.S. financial regulators
were overseeing dealers' credit derivative operations, including the
security and resiliency of the information technology systems used for
these products. GAO analyzed data on credit derivatives operations that
dealers submitted to regulators, reviewed regulatory examination reports
and work papers, and interviewed regulators, dealers, end-users, and
industry organizations.

After trading volumes grew exponentially between 2002 and 2005, the 14
largest credit derivatives dealers--including U.S. and foreign banks and
securities broker-dealers--accumulated backlogs of unconfirmed trades
totaling over 150,000 in September 2005. These backlogs resulted from
reliance on inefficient manual confirmation processes that failed to keep
up with the rapidly growing volume and because of difficulties in
confirming information for trades that end-users transferred to other
parties without notifying the original dealer. Although these trades were
being entered into the systems that dealers used to manage the risk of
loss arising from price changes (market risk) and counterparty defaults
(credit risk), the credit derivatives backlogs increased dealers'
operational risk by potentially allowing errors that could lead to losses
or other problems to go undetected. In response, a joint regulatory
initiative involving U.S. and foreign regulators directed the 14 major
dealers to work together to reduce the backlogs and address the underlying
causes. By increasing automation and requiring end-users to obtain
counterparty consent before assigning trades, the 14 dealers reduced their
total confirmations outstanding more than 30 days by 94 percent to 5,500
trades by October 2006, as shown in the figure below.

Outstanding Confirmations at 14 Major Dealers, September 2005 to October
2006

Through ongoing supervision and examinations, U.S. banking and securities
regulators became aware of the credit derivatives backlogs as early as
late 2003 and had been monitoring efforts taken by each dealer to reduce
its backlog. Under the joint regulatory initiative, regulators obtained
aggregate data from the dealers that allowed regulators to better monitor
how backlogs were being resolved. Recognizing the potential for similar
problems to arise in other OTC derivatives markets, regulators began
obtaining similar data for other OTC derivative products in November 2006.

References

Visible links
  26. http://www.gao.gov/cgi-bin/getrpt?GAO-05-61
  27. http://www.gao.gov/
  28. mailto:[email protected]
  29. http://www.gao.gov/
  30. http://www.gao.gov/
  31. http://www.gao.gov/fraudnet/fraudnet.htm
  32. mailto:[email protected]
  33. mailto:[email protected]
  34. mailto:[email protected]
  35. http://www.gao.gov/cgi-bin/getrpt?GAO-07-716
  36. http://www.gao.gov/cgi-bin/getrpt?GAO-07-716
*** End of document. ***