This is the accessible text file for GAO report number GAO-12-991 
entitled 'Securities Investor Protection Corporation: Customer 
Outcomes in the Madoff Liquidation Proceeding' which was released on 
September 13, 2012. 

This text file was formatted by the U.S. Government Accountability 
Office (GAO) to be accessible to users with visual impairments, as 
part of a longer term project to improve GAO products' accessibility. 
Every attempt has been made to maintain the structural and data 
integrity of the original printed product. Accessibility features, 
such as text descriptions of tables, consecutively numbered footnotes 
placed at the end of the file, and the text of agency comment letters, 
are provided but may not exactly duplicate the presentation or format 
of the printed version. The portable document format (PDF) file is an 
exact electronic replica of the printed version. We welcome your 
feedback. Please E-mail your comments regarding the contents or 
accessibility features of this document to Webmaster@gao.gov. 

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. 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: 
GAO: 

Report to Congressional Requesters: 

September 2012: 

Securities Investor Protection Corporation: 

Customer Outcomes in the Madoff Liquidation Proceeding: 

GAO-12-991: 

GAO Highlights: 

Highlights of GAO-12-991, a report to congressional requesters. 

Why GAO Did This Study: 

After the collapse of Bernard L. Madoff Investment Securities, LLC—a 
broker-dealer and investment advisory firm with thousands of 
individual and institutional clients—the Securities Investor 
Protection Corporation (SIPC), which oversees a fund providing up to 
$500,000 of protection to qualifying individual customers of failed 
securities firms, selected a trustee to liquidate the Madoff firm and 
recover assets for its customers. In March 2012, GAO issued GAO-12-
414, which examined selection of the Trustee, his method for 
determining customer claims, and expenses of the liquidation, among 
other things. This report discusses (1) the extent to which account 
activity varied by type of Madoff customer, (2) the nature of claims 
filed, and rejected or approved, with the Trustee for reimbursement of 
losses, (3) litigation and settlement activity the Trustee has pursued 
in seeking to recover assets for distribution to customers, and (4) 
the effect of the fraud on customers’ federal income tax liabilities. 
GAO reviewed transaction and claims data from the Trustee, lawsuits 
filed by the Trustee, IRS rules and guidance, and interviewed the 
Trustee, private sector tax experts, and officials from IRS, SIPC, and 
the Securities and Exchange Commission. 

What GAO Found: 

GAO’s analysis of Madoff account data shows that more than three-
fourths of the firm’s customers were individuals and families 
(individuals). The remaining accounts were held by institutions, such 
as pension funds and charities. A higher proportion of accounts held 
by an individual (60 percent) were “net winners” based on their net 
equity position-—meaning they had withdrawn more from their accounts 
than they had deposited-—compared to accounts held by institutions (50 
percent). Correspondingly, 40 percent of institutional accounts were “
net losers” that had deposited more into their accounts than they had 
withdrawn, compared to 29 percent of individuals’ accounts that were 
net losers. However, individual and institutional accounts had similar 
deposit and withdrawal activity from 1981 through 2008, including 
increased withdrawals immediately before the firm’s failure in 
December 2008. 

GAO’s analysis shows that the Trustee’s decisions to accept or reject 
claims were similar for individual and institutional account holders. 
Of the more than 16,000 claims, about 66 percent were denied because 
the customers were not direct account holders of the Madoff firm, but 
instead had invested in funds or other vehicles that held accounts 
directly with the firm. For the remaining claimants who were directly 
invested, the Trustee generally used the customers’ net investment 
positions—-that is, whether they were net winners or net losers—-to 
determine claims. In examining claims decisions by customer type, GAO 
found the Trustee denied claims filed by individuals and institutions 
determined to be net winners in similar proportions. Similarly, most 
claims filed by individuals or institutions determined to be net 
losers were allowed. 

The Trustee has been pursuing litigation to recover, or “claw back,” 
assets from net winner customers and others that can be used to 
reimburse customers that did not withdraw all of their principal 
investments. For those customers that withdrew fictitious profits—-net 
winners-—the Trustee has been pursuing more than 1,000 lawsuits to 
recover funds, as allowed under federal bankruptcy law and state law. 
In about 60 suits, the Trustee has sought more than fictitious 
profits, to include principal or other funds received, arguing the 
parties knew or should have known of the fraud. Thus far, the Trustee 
said he has recovered about $9.1 billion of the $17.3 billion in 
principal investments lost by customers who filed claims, including 
$8.4 billion from settlement agreements. 

Because the Madoff fraud affects customers’ taxable income, it also 
affects tax collections by the Department of the Treasury. Under 
Internal Revenue Service (IRS) rules, Madoff customers can deduct lost 
principal and fictitious profits on which they paid taxes while 
holding their accounts. However, IRS does not maintain statistics on 
specific frauds or their impacts on tax collections, and the tax 
impact may be reduced because some taxpayers may not be able to fully 
use this tax relief, such as those that lack other income that can be 
offset by these deductions. Tax experts expressed concerns about the 
lack of clarity over how payments stemming from fraud-related 
avoidance actions filed by the Trustee will be treated for tax 
purposes. In response to a recommendation in a draft report that IRS 
provide guidance to help limit taxpayer errors resulting in over- or 
underpayment of taxes, the agency issued such guidance on September 5, 
2012, in the form of “frequently asked questions” posted to its 
website.  

View [hyperlink, http://www.gao.gov/products/GAO-12-991]. For more 
information, contact A. Nicole Clowers at (202) 512-8678 or 
clowersa@gao.gov. 

[End of section] 

Contents: 

Letter: 

Background: 

Individual Customers Tended to Be Net Winners, While Transaction 
Patterns for Individuals and Institutions Were Generally Similar: 

Trustee Denied Feeder Fund Investor Claims, and Decided Individual and 
Institutional Claims Similarly: 

Trustee Has Been Pursuing Hundreds of Lawsuits to Recover Assets: 

Effects of the Fraud on Income Tax Revenues Cannot Be Determined, and 
New Guidance on Clawbacks Could Reduce Taxpayer Errors: 

Agency and Third Party Comments and Our Evaluation: 

Appendix I: Objectives, Scope, and Methodology: 

Appendix II: Trustee Avoidance Actions and Other Claims: 

Appendix III: Largest Madoff Accounts by Transaction Volume, Total 
Withdrawals, and Net Winnings: 

Appendix IV: Details of Settlements by the Madoff Trustee: 

Appendix V: GAO Contact and Staff Acknowledgments: 

Tables: 

Table 1: Legal Bases Trustee Can Use to Recover Funds from Customers 
and Others Involved in Ponzi Schemes: 

Table 2: Madoff Account Holder Types and Net Positions: 

Table 3: Status of Good Faith Litigation by Madoff Trustee, as of May 
2012: 

Table 4: Ten Largest Good Faith Complaints by Madoff Trustee among GAO 
Sample, by Amount Sought: 

Table 5: Most Prevalent Bases on Which the Madoff Trustee Has Alleged 
Customers or Others Acted in Bad Faith: 

Table 6: Ten Largest Bad Faith Complaints by Madoff Trustee, by Amount 
Sought: 

Table 7: Key Allegations in Six Largest Feeder Fund Complaints by 
Madoff Trustee, by Amount Sought: 

Table 8: Settlement Agreements Reached by Madoff Trustee, by Category, 
as of July 2012: 

Table 9: Examples of Customer Circumstances among Hardship 
Applications Accepted and Denied: 

Table 10: Ten Largest Madoff Accounts by Transaction Volume: 

Table 11: Ten Largest Madoff Accounts by Total Withdrawals: 

Table 12: Ten Largest Madoff Accounts by Net Winnings: 

Table 13: Madoff Settlement Agreements of at Least $20 Million: 

Table 14: Key Provisions among Largest Madoff Settlements: 

Figures: 

Figure 1: Deposit and Withdrawal Activity for Madoff Customer 
Accounts, by Year and Account Holder Type, 1981 to 2008: 

Figure 2: Madoff Customers' Withdrawals of Principal, by Year and 
Account Holder Type, 5 Years before 2008 Failure: 

Figure 3: Trustee's Disposition of Claims Filed in Madoff Liquidation, 
as of April 2012: 

Figure 4: Trustee's Disposition of Claims Filed in Madoff Liquidation, 
by Net Loser and Net Winner Status, as of April 2012: 

Figure 5: Trustee's Disposition of Madoff Customer Claims, by Customer 
Type and Net Investment Status, as of April 2012: 

Figure 6: Number and Amount of Allowed Madoff Customer Claims, as of 
July 2012: 

Figure 7: Madoff Trustee Hardship Program Process and Outcomes, as of 
May 2012: 

Figure 8: Results of the Clawback Element of the Madoff Trustee's 
Hardship Program, as of June 2012: 

Abbreviations: 

AMT: Alternative Minimum Tax: 

CIO: chief information officer: 

CPLR: Civil Practice Law and Rules: 

DCL: Debtor and Creditor Law: 

FSM: final statement method: 

IRS Internal Revenue Service: 

NIM: net investment method: 

PACER: U.S. Courts' Public Access to Court Electronic Records: 

SEC: Securities and Exchange Commission: 

SIPA: Securities Investor Protection Act of 1970: 

SIPC: Securities Investor Protection Corporation: 

[End of section] 

United States Government Accountability Office: 
Washington, DC 20548: 

September 13, 2012: 

The Honorable Scott Garrett:
Chairman:
Subcommittee on Capital Markets and Government Sponsored Enterprises:
Committee on Financial Services:
House of Representatives: 

The Honorable Peter King:
House of Representatives:
The Honorable Carolyn McCarthy:
House of Representatives: 

The Honorable Ileana Ros-Lehtinen:
House of Representatives: 

After the collapse of Bernard L. Madoff Investment Securities, LLC--a 
broker-dealer and investment advisory firm with thousands of clients--
in December 2008, thousands of customers found they had lost billions 
of dollars to a Ponzi scheme Madoff had run for years.[Footnote 1] 
Within days of the collapse, the Securities Investor Protection 
Corporation (SIPC), a nonprofit, nongovernmental membership 
corporation responsible for providing financial protection to 
customers of failed securities firms, designated a trustee--attorney 
Irving H. Picard (referred to as the Trustee throughout this report)--
to oversee the liquidation of the Madoff firm and recover assets for 
the benefit of former customers. 

The Securities Investor Protection Act of 1970 (SIPA) established 
procedures for liquidating failed broker-dealers. In a liquidation 
under SIPA, the trustee establishes a fund of customer property 
consisting of the cash and securities held by the broker-dealer on 
behalf of customers, plus any assets recovered by the trustee, for 
distribution among customers. The actions of the trustee in carrying 
out a liquidation have direct effects for former customers who lost 
money, such as how much of their losses the trustee can recover. There 
are also indirect effects--for example, losses and potential 
recoveries can affect customers' tax liabilities, which in turn affect 
tax collections by the U.S. Treasury. 

Because the Trustee's activities affect thousands of customers, you 
asked us to examine a series of questions about his efforts to resolve 
the Madoff liquidation. We agreed to examine these issues in two 
reports. The first, issued in March 2012, examined selection of the 
Trustee, his method for determining customer claims, and expenses of 
the liquidation, among other things.[Footnote 2] This second and final 
report discusses (1) the extent to which account activity varied by 
type of Madoff customer; (2) the nature of claims filed, and rejected 
or approved, with the Trustee for reimbursement of losses; (3) 
litigation and settlement activity the Trustee has pursued in seeking 
to recover assets for distribution to customers; and (4) the effect of 
the fraud on customers' federal income tax liabilities, including the 
effect on amounts that would have been due if investor losses had been 
based on customers' reported final statement holdings. 

For this report, we obtained and analyzed data from the Trustee, 
including information on deposits to and withdrawals from Madoff 
customer accounts from April 1981 to December 2008. We also analyzed 
Trustee data on claims customers made for reimbursement from proceeds 
of the firm's liquidation and the determinations he reached on these 
claims. We examined lawsuits the Trustee has been pursuing and 
settlements he has reached with customers and others, including 
reviewing a randomly selected sample of 50 of the more than 1,000 
cases filed seeking the return of fictitious profits, as allowed under 
federal and state law. We reviewed 29 publicly available cases from 
among 30 actions in which the Trustee is arguing defendants knew or 
should have known about the fraud, referred to as "bad faith" cases. 
In addition, we reviewed the largest cases among 27 actions focusing 
on funds that invested customer money in Madoff accounts. We 
interviewed the Trustee and members of his law firm, as well as SIPC 
executives, and officials of the Securities and Exchange Commission 
(SEC), which has oversight authority for SIPC. We also interviewed 
four individuals with income tax expertise, including a law professor 
and three tax advisors who we selected based on their tax experience 
or publications related to tax issues. We reviewed Internal Revenue 
Service (IRS) guidance and interviewed IRS officials about tax 
treatment of fraudulent schemes like the Madoff case. We conducted 
this performance audit from March to September 2012 in accordance with 
generally accepted government auditing standards. Those standards 
require that we plan and perform the audit to obtain sufficient, 
appropriate evidence to provide a reasonable basis for our findings 
and conclusions based on our audit objectives. We believe that the 
evidence obtained provides a reasonable basis for our findings and 
conclusions based on our audit objectives. 

Background: 

SIPC's mission is to promote confidence in securities markets by 
seeking to return customers' cash and securities when a broker-dealer 
fails. SIPC provides advances for these customers up to the SIPA 
protection limits--$500,000 per customer, except that claims for cash 
are limited to $250,000 per customer.[Footnote 3] SIPA established a 
fund (SIPC fund) to pay for SIPC's operations and activities. SIPC 
uses the fund to make advances to satisfy customer claims for missing 
cash and securities, including notes, stocks, bonds, and certificates 
of deposit. The SIPC fund also covers the administrative expenses of a 
liquidation proceeding (including costs incurred by a trustee, 
trustee's counsel, and other advisors) when the general estate of the 
failed firm is insufficient. 

SIPC finances the fund through annual assessments it sets for member 
firms, plus interest generated from its investments in Treasury notes. 
If the SIPC fund becomes, or appears to be, insufficient to carry out 
the purposes of SIPA, SIPC can borrow up to $2.5 billion from Treasury 
through SEC. That is, SEC would borrow the funds from Treasury and 
relend them to SIPC. According to SIPC senior management, recent 
demands on the fund, including from the Madoff case, together with a 
change in SIPC bylaws that increased the target size of the fund from 
$1 billion to $2.5 billion, led SIPC to impose new industry 
assessments totaling about $400 million annually. The assessments, 
equal to one-quarter of 1 percent of net operating revenue, will 
continue until the $2.5 billion target is reached, according to SIPC 
senior management. The new assessments replaced a flat annual 
assessment of $150 per member firm.[Footnote 4] Under the new levies, 
the average assessment for 2010 was $91,755 per firm, with a median of 
$2,095, according to SIPC. 

Liquidations under SIPA: 

SIPA authorizes SIPC to begin a liquidation action by applying for a 
protective order from an appropriate federal district court if it 
determines that one of SIPC's member broker-dealers has failed or is 
in danger of failing to meet its obligations to customers and one or 
more additional statutory conditions are met.[Footnote 5] The broker-
dealer can contest the protective order application. If the court 
issues the order, the court appoints a trustee selected by SIPC, or, 
in certain cases, SIPC itself, to liquidate the firm. While SIPC 
designates the trustee, that person, once judicially appointed, 
becomes an officer of the court. As such, the trustee exercises 
independent judgment and does not serve as an agent of SIPC. 

Under SIPA, the trustee must investigate facts and circumstances 
relating to the liquidation; report to the court facts indicating 
fraud, misconduct, mismanagement, or irregularities; and submit a 
final report to SIPC and others designated by the court. Also, the 
trustee is to periodically report to the court and SIPC on his or her 
progress in distributing cash and securities to customers. To the 
extent that it is consistent with SIPA, the proceeding is conducted 
pursuant to provisions of the Bankruptcy Code. 

Promptly after being appointed, the trustee must publish a notice of 
the proceeding in one or more major newspapers, in a form and manner 
determined by the court. The trustee also must see that a copy of the 
notice is mailed to existing and recent customers listed on the broker-
dealer's books and records, and provide notice to creditors in the 
manner SIPA prescribes. Customers must file written statements of 
claims. The trustee's notice includes a claim form, and also informs 
customers how to file claims and explains deadlines. 

Determination of Madoff Customer Claims: 

Once filed, the claims undergo various reviews, according to the 
Trustee. First, the Trustee's claims agent reviews claims for 
completeness; if information is found to be missing, the claims agent 
sends a request for additional information. Second, the Trustee's 
forensic accountants review each claim form, information from the 
Madoff firm's records about the account at issue, and information 
submitted directly by the claimant. The Trustee uses the results of 
this review in assessing his determination of the claim. Finally, 
claims move to SIPC, where a claims review specialist provides a 
recommendation to the Trustee on how each claim should be determined. 
Once that recommendation has been made, the Trustee and trustee's 
counsel review it, as well as legal or other issues raised previously. 
When the Trustee has decided on resolution of a claim, he issues a 
determination letter to the claimant.[Footnote 6] 

As of the start of 2012, the Trustee had received 16,519 customer 
claims in the Madoff proceeding, and reached determinations on all but 
two of them. According to SIPC, many Madoff customers were older. For 
example, according to Trustee information we reviewed on his hardship 
program (described later in this report), more than half of applicants 
were age 71 or older.[Footnote 7] 

In a liquidation under SIPA, amounts in the customer property fund 
generally are distributed to the failed firm's customers according to 
the value of their account holdings, or "net equity." SIPA generally 
provides the net equity amount is what would have been owed to the 
customer if the broker-dealer had liquidated all their "securities 
positions," less any obligations of the customer to the firm.[Footnote 
8] In the Madoff case, if the Trustee recovers less than the total 
amount of allowed claims, some claimants likely will receive only a 
portion of their allowed claims. The Trustee told us his goal is to 
recover the full amount, but that is not likely, given developments in 
litigation and decisions to settle cases. 

In SIPA liquidations not involving fraud, trustees typically determine 
that the amounts owed to customers match the amounts shown on their 
final statements, in what is known as the "final statement method" 
(FSM). However, in cases involving fraud, amounts in customer accounts 
may not correspond to statement amounts. In the Madoff case, the 
Trustee determined that the securities positions shown on customer 
statements were fictitious. As a result, supported by SIPC and SEC, he 
decided to value each customer's net equity according to the amount of 
cash deposited less any amounts withdrawn--a method known as the "net 
investment method" (NIM). Under NIM, Madoff claimants generally divide 
into two categories: "net winners," who have withdrawn more than the 
amount they invested with the Madoff firm, and "net losers," who have 
withdrawn less than they invested. 

Some customers challenged the Trustee's decision on valuing customer 
net equity, but two courts have considered the issue--the U.S. 
Bankruptcy Court for the Southern District of New York and the U.S. 
Court of Appeals for the Second Circuit--and each has affirmed the 
Trustee's decision to use NIM. In June 2012, the U.S. Supreme Court 
declined to hear an appeal on the issue, thus concluding legal 
challenges to the Trustee's decision.[Footnote 9] 

Efforts by the Trustee to Recover Assets for the Benefit of Customers: 

The Trustee has taken various steps to recover assets for distribution 
to former Madoff customers, including recovery of bank account 
balances and sale of the firm's assets. In addition, the Trustee has 
filed hundreds of lawsuits known as "avoidance actions" or 
"clawbacks." Avoidance powers enable a trustee to "avoid," or set 
aside, certain transfers made by a debtor--here, the Madoff firm--
prior to the bankruptcy filing, in order to recover transferred funds 
for the benefit of the estate. In pursuing these actions, a trustee 
can generally seek return of fictitious profits paid to investors, and 
in some cases, principal amounts withdrawn, for specified periods of 
time--90 days, 2 years, and 6 years preceding the filing. In doing so, 
the Trustee has available state statutes, common law claims, and 
federal bankruptcy law upon which to build his cases. Actions also can 
vary according to whether the Trustee alleges a customer acted in good 
or bad faith. Under the Bankruptcy Code, a recipient's good faith or 
bad faith is not relevant to whether the transfer is avoidable but 
does affect the extent of the recipient's liability. Table 1 
summarizes the legal avenues available. A fuller discussion of legal 
remedies available can be found in appendix II. 

Table 1: Legal Bases Trustee Can Use to Recover Funds from Customers 
and Others Involved in Ponzi Schemes: 

Legal action and authority: Preferential transfer avoidance; 
Bankruptcy Code § 547; 
Payments and periods covered: Transfers 90 days prior to bankruptcy 
filing, except 1 year from filing for transfers to "insiders"; 
Potential liability: Good faith recipients: Principal amounts invested; 
Potential liability: Bad faith recipients: Principal amounts invested. 

Legal action and authority: Fraudulent transfer avoidance: 
constructive fraud; Bankruptcy Code § 548(a)(1)(B); 
Payments and periods covered: Transfers 2 years prior to filing; 
Potential liability: Good faith recipients: Fictitious profits; 
Potential liability: Bad faith recipients: Fictitious profits and 
principal. 

Legal action and authority: Fraudulent transfer avoidance: actual 
fraud; Bankruptcy Code § 548(a)(1)(A); 
Payments and periods covered: Transfers 2 years prior to filing; 
Potential liability: Good faith recipients: Fictitious profits; 
Potential liability: Bad faith recipients: Fictitious profits and 
principal. 

Legal action and authority: Fraudulent transfer avoidance: 
constructive fraud; Bankruptcy Code § 544(b); 
New York Debtor and Creditor Law (DCL) §§ 273-275; 
New York Civil Practice Law and Rules (CPLR) § 213(1); 
Payments and periods covered: Transfers 6 years prior to filing; 
Potential liability: Good faith recipients: Fictitious profits; 
Potential liability: Bad faith recipients: Fictitious profits and 
principal. 

Legal action and authority: Fraudulent transfer avoidance: actual 
fraud; Bankruptcy Code § 544(b); DCL § 276; CPLR §§ 203(g), 213(8); 
Payments and periods covered: Transfers 6 years prior to filing, or 
potentially earlier depending on timing of discovery of fraud; 
Potential liability: Good faith recipients: Fictitious profits; 
Potential liability: Bad faith recipients: Fictitious profits and 
principal. 

Legal action and authority: Common law claims: e.g., conversion, 
unjust enrichment; Bankruptcy Code §§ 323, 541; 
Common law principles as applicable; 
Payments and periods covered: Period covered depends on nature of 
claim asserted; 
Potential liability: Good faith recipients: N/A; 
Potential liability: Bad faith recipients: Damages, depending on 
nature of claim. 

Source: GAO summary of federal bankruptcy, New York state law. 

[End of table] 

Generally, transfers are avoidable as actually fraudulent if the 
debtor-transferor had intent to defraud; or, as constructively 
fraudulent if they were made without fraudulent intent but for less 
than equivalent or fair value, while the debtor was insolvent. Courts 
generally presume Ponzi scheme payouts to be actually fraudulent. As 
for constructive fraud, Ponzi scheme transfers in excess of principal 
are not made for value--meaning they are not paid to satisfy any 
legitimate obligation the debtor owed to the recipient--and thus fall 
within the constructive fraud provisions.[Footnote 10] 

Individual Customers Tended to Be Net Winners, While Transaction 
Patterns for Individuals and Institutions Were Generally Similar: 

In our analysis of information obtained from the Trustee, we 
identified 7,994 accounts with at least one transaction. We examined 
names on each of these accounts, to determine whether individuals and 
families (to which we refer jointly as individuals) or institutions 
held the accounts.[Footnote 11] We found that individuals held more 
than three-fourths (77 percent) of accounts, while almost one-quarter 
(23 percent) of accounts were held by institutions, such as charities, 
pension funds, and feeder funds.[Footnote 12] Using these groupings, 
we further examined the account holders' claims outcomes--whether they 
were net winners or net losers--and the pattern of their transactions 
leading up to the Madoff firm's collapse.[Footnote 13] The Trustee, 
however, noted that accounts held by institutions generally 
represented funds from individuals as well, because with the exception 
of funds from nonprofit organizations, the institutions were investing 
money on behalf of individuals. The Trustee said that based on his 
examination, he was not aware of any direct corporate investment in 
the Madoff firm. 

Accounts Held by Individuals Tended to Be Net Winners: 

As shown in table 2, our analysis indicates that a higher proportion 
of accounts held by individuals (60 percent) were net winners that had 
withdrawn more than they had deposited over the lifetime of their 
accounts, compared to accounts held by institutions (50 percent). We 
also found that more institutional accounts (40 percent) were net 
losers than were individual accounts (29 percent). Among individual 
accounts, net winners outnumbered net losers by 2-to-1, with outcomes 
split more evenly for institutional accounts. Overall, for both 
categories combined, we found that 57 percent of account holders were 
net winners. About one-third (32 percent) were net losers, whose total 
withdrawals were less than their total investment deposits. The 
remaining 11 percent of accounts had zero balances from a net 
investment perspective, with account holders having withdrawn exactly 
what they had invested. 

Table 2: Madoff Account Holder Types and Net Positions: 

Dollars in millions: 

Net losers:
Account holder type: Individual/family; 
Number of accounts: 1,810; 
Percentage of all accounts of this type: 29%; 
Net investment position: $5,135; 
Account holder type: Institution; 
Number of accounts: 725; 
Institution: Percentage of all accounts of this type: 40%; 
Institution: Net investment position: $14,790. 

Net winners: 
Account holder type: Individual/family; 
Number of accounts: 3,667; 
Percentage of all accounts of this type: 60$; 
Net investment position: ($5,902); 
Account holder type: Institution; 
Number of accounts: 910; 
Percentage of all accounts of this type: 50%; 
Institution: Net investment position: ($11,824). 

Zero balance: 
Account holder type: Individual/family; 
Number of accounts: 675; 
Percentage of all accounts of this type: 11$; 
Net investment position: $0; 
Account holder type: Institution; 
Number of accounts: 177; 
Percentage of all accounts of this type: 10$; 
Institution: Net investment position: $0. 

Total: 
Account holder type: Individual/family; 
Number of accounts: 6,152; 
Percentage of all accounts of this type: 100$; 
Net investment position: ($767); 
Account holder type: Institution; 
Number of accounts: 1,812; 
Percentage of all accounts of this type: 100$; 
Institution: Net investment position: $2,965. 

Source: GAO analysis of Madoff Trustee data. 

Notes: Positive net investment positions represent deposits in excess 
of withdrawals; negative net investment positions indicate withdrawals 
in excess of deposits. Numbers may not sum to totals shown due to 
rounding. 

[End of table] 

Table 2 also shows that as a group, institutional accounts lost 
principal amounts they invested in the fraud, while accounts owned by 
individuals in the aggregate withdrew more than they invested. For 
individuals, the total net investment position at the time of the 
Madoff firm's failure was ($767 million), meaning they had withdrawn 
more money than they originally invested. As a result, individual 
account holders as a group were net winners. By contrast, the total 
net investment position for institutional accounts as a group showed 
that they were net losers, having made nearly $3.0 billion more in 
deposits than withdrawals.[Footnote 14] We note that while we divided 
the population into account types for analytical purposes, the Trustee 
and SIPC executives stressed that each customer claim was determined 
on a case-by-case basis without regard to whether the account holder 
was an individual or institution.[Footnote 15] In addition, SIPC 
executives noted that notwithstanding the overall totals for the two 
customer categories, there were nevertheless both net winners and net 
losers in each grouping. 

Individual and Institutional Accounts Had Similar Patterns of Deposits 
and Withdrawals: 

Our analysis found that individual and institutional accounts had 
similar deposit and withdrawal activity throughout the 27-year period 
we examined, including during the period immediately before the 
failure in 2008. For both groups, total annual deposits and total 
annual withdrawals increased steadily since 1981, growing by greater 
amounts in more recent years.[Footnote 16] Overall, deposit volume 
more than doubled between 2005 and 2007, rising from $4.5 billion to 
$9.4 billion annually. Withdrawal volume grew more slowly from 2005 to 
2007, rising only about 14 percent, from $5.7 billion to $6.4 billion. 
However, withdrawal volume in 2008, at $12.6 billion, was almost 
double the volume in 2007. Figure 1 breaks out total deposit and 
withdrawal activity by individual and institutional accounts. As 
noted, the pattern of activity is similar for each group, except that 
deposit volume for individual accounts was about the same in 2007 and 
2008, at $3.4 billion annually, while deposit volume from 
institutional accounts fell between 2007 and 2008, from $6.0 billion 
to $5.1 billion. 

Figure 1: Deposit and Withdrawal Activity for Madoff Customer 
Accounts, by Year and Account Holder Type, 1981 to 2008: 

[Refer to PDF for image: multiple line graph] 

Year: 1981; 
Individual deposits: $68 million; 
Individual withdrawals: $10 million; 
Institutional deposits: $76 million; 
Institutional withdrawals: $17 million. 

Year: 1982; 
Individual deposits: $35 million; 
Individual withdrawals: $28 million; 
Institutional deposits: $18 million; 
Institutional withdrawals: $17 million. 

Year: 1983; 
Individual deposits: $32 million; 
Individual withdrawals: $24 million; 
Institutional deposits: $28 million; 
Institutional withdrawals: $29 million. 

Year: 1984; 
Individual deposits: $18 million; 
Individual withdrawals: $32 million; 
Institutional deposits: $21 million; 
Institutional withdrawals: $29 million. 

Year: 1985; 
Individual deposits: $23 million; 
Individual withdrawals: $40 million; 
Institutional deposits: $64 million; 
Institutional withdrawals: $52 million. 

Year: 1986; 
Individual deposits: $54 million; 
Individual withdrawals: $34 million; 
Institutional deposits: $50 million; 
Institutional withdrawals: $56 million. 

Year: 1987; 
Individual deposits: $76 million; 
Individual withdrawals: $68 million; 
Institutional deposits: $76 million; 
Institutional withdrawals: $72 million. 

Year: 1988; 
Individual deposits: $74 million; 
Individual withdrawals: $73 million; 
Institutional deposits: $77 million; 
Institutional withdrawals: $76 million. 

Year: 1989; 
Individual deposits: $95 million; 
Individual withdrawals: $97 million; 
Institutional deposits: $114 million; 
Institutional withdrawals: $93 million. 

Year: 1990; 
Individual deposits: $150 million; 
Individual withdrawals: $133 million; 
Institutional deposits: $156 million; 
Institutional withdrawals: $120 million. 

Year: 1991; 
Individual deposits: $228 million; 
Individual withdrawals: $186 million; 
Institutional deposits: $162 million; 
Institutional withdrawals: $208 million. 

Year: 1992; 
Individual deposits: $577 million; 
Individual withdrawals: $284 million; 
Institutional deposits: $496 million; 
Institutional withdrawals: $898 million. 

Year: 1993; 
Individual deposits: $531 million; 
Individual withdrawals: $554 million; 
Institutional deposits: $461 million; 
Institutional withdrawals: $373 million. 

Year: 1994; 
Individual deposits: $375 million; 
Individual withdrawals: $356 million; 
Institutional deposits: $341 million; 
Institutional withdrawals: $413 million. 

Year: 1995; 
Individual deposits: $426 million; 
Individual withdrawals: $49 million6; 
Institutional deposits: $656 million; 
Institutional withdrawals: $636 million. 

Year: 1996; 
Individual deposits: $701 million; 
Individual withdrawals: $489 million; 
Institutional deposits: $782 million; 
Institutional withdrawals: $761 million. 

Year: 1997; 
Individual deposits: $1.153 billion; 
Individual withdrawals: $766 million; 
Institutional deposits: $1.389 billion; 
Institutional withdrawals: $947 million. 

Year: 1998; 
Individual deposits: $1.294 billion; 
Individual withdrawals: $1.056 billion; 
Institutional deposits: $2.101 billion; 
Institutional withdrawals: $.1263 billion. 

Year: 1999; 
Individual deposits: $1.413 billion; 
Individual withdrawals: $1.000 billion; 
Institutional deposits: $1.984 billion; 
Institutional withdrawals: $1.552 billion. 

Year: 2000; 
Individual deposits: $1.342 billion; 
Individual withdrawals: $1.076 billion; 
Institutional deposits: $2.001 billion; 
Institutional withdrawals: $2.266 billion. 

Year: 2001; 
Individual deposits: $846 million; 
Individual withdrawals: $1.438 billion; 
Institutional deposits: $2.499 billion; 
Institutional withdrawals: $2.032 billion. 

Year: 2002; 
Individual deposits: $1.180 billion; 
Individual withdrawals: $1.721 billion; 
Institutional deposits: $.2867 billion; 
Institutional withdrawals: $2.532 billion. 

Year: 2003; 
Individual deposits: $1.359 billion; 
Individual withdrawals: $1.807 billion; 
Institutional deposits: $3.160 billion; 
Institutional withdrawals: $3.124 billion. 

Year: 2004; 
Individual deposits: $2.291 billion; 
Individual withdrawals: $1.578 billion; 
Institutional deposits: $2.723 billion; 
Institutional withdrawals: $2.838 billion. 

Year: 2005; 
Individual deposits: $1.495 billion; 
Individual withdrawals: $2.654 billion; 
Institutional deposits: $3.054 billion; 
Institutional withdrawals: $.3000 billion. 

Year: 2006; 
Individual deposits: $2.311 billion; 
Individual withdrawals: $2.458 billion; 
Institutional deposits: $5.064 billion; 
Institutional withdrawals: $3.269 billion. 

Year: 2007; 
Individual deposits: $3.414 billion; 
Individual withdrawals: $2.729 billion; 
Institutional deposits: $5.957 billion; 
Institutional withdrawals: $3.719 billion. 

Year: 2008; 
Individual deposits: $3.398 billion; 
Individual withdrawals: $4.476 billion; 
Institutional deposits: $5.083 billion; 
Institutional withdrawals: $8.097 billion. 

Source: GAO analysis of Trustee data. 

[End of figure] 

Our analysis also showed that in the final year of the Madoff firm's 
existence (2008), both individuals and institutional account holders 
withdrew large amounts of principal they had invested. Figure 2 shows 
such principal withdrawals, excluding any fictitious profits, during 
quarterly periods leading up to the failure. Withdrawals of principal 
began to increase three quarters before the failure, with most of the 
increase in withdrawals occurring in the 90 days before the Madoff 
firm's failure. Under his clawback authority, described earlier, the 
Trustee can, among other things, seek to recover principal withdrawals 
during this period. 

Figure 2: Madoff Customers' Withdrawals of Principal, by Year and 
Account Holder Type, 5 Years before 2008 Failure: 

[Refer to PDF for image: multiple line graph] 

 Total Withdrawals of Principal Individual Withdrawals of Principal 
Institution Withdrawals of Principal 

Quarters before failure[A]: 20; 
Total: $519 million; 
Individual: $150 million; 
Institutional: $368 million. 

Quarters before failure[A]: 19; 
Total: $677 million; 
Individual: $204 million; 
Institutional: $472 million. 

Quarters before failure[A]: 18; 
Total: $419 million; 
Individual: $172 million; 
Institutional: $248 million. 

Quarters before failure[A]: 17; 
Total: $410 million; 
Individual: $227 million; 
Institutional: $183 million. 

Quarters before failure[A]: 16; 
Total: $557 million; 
Individual: $213 million; 
Institutional: $344 million. 

Quarters before failure[A]: 15; 
Total: $534 million; 
Individual: $299 million; 
Institutional: $235 million. 

Quarters before failure[A]: 14; 
Total: $538 million; 
Individual: $206 million; 
Institutional: $333 million. 

Quarters before failure[A]: 13; 
Total: $1.566 billion; 
Individual: $837 million; 
Institutional: $729 million. 

Quarters before failure[A]: 12; 
Total: $970 million; 
Individual: $280 million; 
Institutional: $689 million. 

Quarters before failure[A]: 11; 
Total: $1.058 billion; 
Individual: $322 million; 
Institutional: $736 million. 

Quarters before failure[A]: 10; 
Total: $654 million; 
Individual: $216 million; 
Institutional: $439 million. 

Quarters before failure[A]: 9; 
Total: $517 million; 
Individual: $176 million; 
Institutional: $340 million. 

Quarters before failure[A]: 8; 
Total: $793 million; 
Individual: $196 million; 
Institutional: $598 million. 

Quarters before failure[A]: 7; 
Total: $518 million; 
Individual: $285 million; 
Institutional: $233 million. 

Quarters before failure[A]: 6; 
Total: $865 million; 
Individual: $470 million; 
Institutional: $395 million. 

Quarters before failure[A]: 5; 
Total: $1.072 billion; 
Individual: $633 million; 
Institutional: $439 million. 

Quarters before failure[A]: 4; 
Total: $823 million; 
Individual: $280 million; 
Institutional: $543 million. 

Quarters before failure[A]: 3; 
Total: $1.041 billion; 
Individual: $377 million; 
Institutional: $663 million. 

Quarters before failure[A]: 2; 
Total: $1.915 billion; 
Individual: $646 million; 
Institutional: $1.269 billion. 

Quarters before failure[A]: 1; 
Total: $5.182 billion; 
Individual: $1.575 billion; 
Institutional: $3.606 billion. 

Source: GAO analysis of Trustee data. 

[A] "Quarters" refers to successive 90-day periods preceding failure. 

[End of figure] 

Accounts that had large withdrawals, particularly those made just 
before the Madoff firm's collapse, could suggest that such customers 
knew of the fraudulent scheme and were attempting to avoid suffering 
losses when the firm failed. Although the increase in withdrawals we 
identified during the period preceding collapse could suggest some 
customers anticipated the firm's failure, this activity may also have 
been due to investor reactions to the financial crisis, which was 
peaking at the time, or to other factors affecting where investors 
place their money. As we describe later, the Trustee has filed 
extensive litigation to recover withdrawals that many customers made. 
In appendix III, we show other results of our analysis, listing the 
largest Madoff accounts by transaction volume, total withdrawals, and 
net winnings. 

Trustee Denied Feeder Fund Investor Claims, and Decided Individual and 
Institutional Claims Similarly: 

As required by SIPA, the Trustee solicited claims from Madoff 
customers for reimbursement of their losses, with approved claimants 
eligible to receive a share of any cash or securities Madoff held on 
behalf of his customers, plus assets recovered by the Trustee during 
the liquidation proceeding. According to claims data provided by the 
Trustee, a total of 16,519 claims were filed. As shown in figure 3, 
most of the claims were denied. Sixty-six percent of the claims were 
denied because the filers had not invested directly with the Madoff 
firm themselves (referred to as "third party" denials). Instead, they 
had invested in feeder funds or other vehicles that owned the accounts 
at the firm. The Trustee determined that under SIPA, only those who 
had invested directly with the Madoff firm were customers for claims 
purposes.[Footnote 17] 

Figure 3: Trustee's Disposition of Claims Filed in Madoff Liquidation, 
as of April 2012: 

[Refer to PDF for image: pie-chart] 

Denied--third party: 66.45 (10,976); 
Denied: 16.4% (2,703); 
Deemed determined: 1.5% (248); 
Allowed: 14.7% (2,425); 
Other: 1.0% (165); 
Total: 16,519. 

Source: GAO analysis of Trustee data. 

Notes: "Other" category consists of: claims withdrawn; claims not yet 
determined; and claims determined to be no claim, because they lacked 
necessary information or did not claim an amount. "Deemed determined" 
means subject to litigation with the Trustee. 

[End of figure] 

Among 5,543 claims from direct investors remaining after denial of the 
third party claims, the Trustee denied 2,703 claims (16 percent of all 
claims). Almost all of the denied claims were from net winners, 
meaning that they had withdrawn more money than they invested. The 
Trustee allowed 2,425 claims (15 percent of all claims), totaling $7.3 
billion.[Footnote 18] Of the allowed claims, the majority were filed 
by net losers, meaning they had withdrawn less from their account than 
they had invested. Figure 4 details the Trustee's disposition of 
claims, by net winner and loser status. However, there were exceptions 
to the general pattern of approvals and denials by net investment 
status. For example, 10 (less than 1 percent of all claims) net winner 
claims were allowed. According to the Trustee, in some cases, these 
claims were allowed because the account holders repaid certain 
withdrawals made from their accounts, and when they returned those 
funds, they became eligible for an allowed claim.[Footnote 19] In 
other instances, net winner claims were allowed when combined with net 
loser account(s) held by the same party. The Trustee also denied seven 
net loser claims, most often because the account was combined with net 
winner account(s) held by the same party. 

Figure 4: Trustee's Disposition of Claims Filed in Madoff Liquidation, 
by Net Loser and Net Winner Status, as of April 2012: 

[Refer to PDF for image: 2 pie-charts] 

Net losers: 
Denied: 0.3% (7); 
Deemed determined: 8.8% (243); 
Allowed: 87.3% (2,414); 
Other: 3.6% (100); 
Total: 2,764. 

Net winners: 
Denied: 97.1% (2,543); 
Deemed determined: 0.2% (5); 
Allowed: 0.3% (9); 
Other: 2.4% (63); 
Total: 2,620. 

Source: GAO analysis of Trustee data. 

Notes: "Other" category consists of: claims withdrawn; claims not yet 
determined; and claims determined to be no claim, because they lacked 
necessary information or did not claim an amount. "Deemed determined" 
means subject to litigation with the Trustee. 

[End of figure] 

Our analysis of the Trustee's data shows that his acceptance or denial 
of claims was similar for both individual and institutional accounts, 
given account holders' net investment position.[Footnote 20] Figure 5 
shows a breakdown of claims outcomes, by account type and net 
investment status. As the figure shows, 98 percent and 94 percent of 
claims filed by individuals and institutions, respectively, who were 
determined to be net winners were denied--with 5 percent of the 
institutional net winner claims being withdrawn before final 
determination--so that disposition of both types was therefore almost 
identical. Similarly, most claims filed by individuals or institutions 
that were determined to be net losers were allowed or deemed 
determined (that is, subject to litigation with the Trustee).[Footnote 
21] 

Figure 5: Trustee's Disposition of Madoff Customer Claims, by Customer 
Type and Net Investment Status, as of April 2012: 

[Refer to PDF for image: 4 pie-charts] 

Net losers: Individuals; 
Denied: 0; 
Deemed determined: 7% (132); 
Allowed: 92% (1,748); 
Other: 1% (24); 
Total: 1,909[A]. 

Net losers: Institutions; 
Denied: 0; 
Deemed determined: 13% (109); 
Allowed: 78% (663); 
Other: 9% (76); 
Total: 850[A]. 

Net winners: Individuals; 
Denied: 98% (2,118); 
Deemed determined: 0; 
Allowed: 0; 
Other: 2% (42); 
Total: 2,169[A]. 

Net winners: Institutions; 
Denied: 94% (423; 
Deemed determined: 0; 
Allowed: 1% (4); 
Other: 5% (21); 
Total: 449[A]. 

Source: GAO analysis of Trustee data. 

[A] Percentages are rounded; categories with values less than 1 
percent are not shown. 

Notes: "Other" category consists of: claims withdrawn; claims not yet 
determined; and claims determined to be no claim, because they lacked 
necessary information or did not claim an amount. "Deemed determined" 
means subject to litigation with the Trustee. 

[End of figure] 

We also found that some net losers who would otherwise appear to be 
eligible for a claim approval did not file claims. In particular, we 
identified 197 net loser accounts (less than 3 percent of all 
accounts) that did not file. The Trustee told us such customers may 
not have filed claims for several reasons: 

* The accounts may be owned by a person or entity that also has 
accounts that were larger net winners. 

* The accounts may have been subject to settlements that affected 
claims. 

* Some account holders may simply have not thought it worthwhile to 
file a claim. 

* Some account holders thought filing a claim would call attention to 
themselves, which they wanted to avoid. Similarly, some foreign 
account holders might not have wanted their government to know they 
had an account in the United States for tax or other reasons, and may 
have believed that filing a claim would increase the likelihood that 
their government would find out about the account. 

Finally, our analysis of the Trustee data shows that while most 
allowed claims were for less than $1 million, a small number of large 
claims account for more than half of the total amount of all approved 
claims. As shown in figure 6, 1,127 allowed claims--each for less than 
$1 million--account for 54 percent of claims allowed. The total value 
of these claims is about $380 million, or about 5 percent of all 
allowed claim amounts. In contrast, the Trustee allowed 93 claims for 
$10 million or more each. Although these claims represent about 4 
percent of all allowed claims, the total value of these claims is $4.5 
billion, or 61 percent of all allowed claim amounts. The accounts with 
the largest allowed claims were $1.57 billion for Optimal 
Multiadvisors, a Geneva-based hedge fund of Spanish banking 
corporation Banco Santander, and $741 million for M-Invest Limited, a 
feeder fund created by Swiss bank Union Bancaire Privee.[Footnote 22] 
The remaining allowed claims were all for less than $300 million each. 
We also examined the distribution of claim size by customer type--
individual and institutional accounts--and found generally similar 
results. Although the distributions were not identical to the overall 
pattern, the same results generally held for each customer type, with 
each group having relatively more lower-value claims and relatively 
fewer higher-value claims. 

Figure 6: Number and Amount of Allowed Madoff Customer Claims, as of 
July 2012: 

[Refer to PDF for image: vertical bar graph] 

Size of allowed claim: Less than $500,000: 
Number of allowed claims: 789; 
Total allowed claims: $380 million. 

Size of allowed claim: $500,000 - less than $1 million; 
Number of allowed claims: 338; 
Total allowed claims: approximately $400 million. 

Size of allowed claim: $1 million - less than $2 million;  
Number of allowed claims: 364; 
Total allowed claims: approximately $500 million. 

Size of allowed claim: $2 million - less than $10 million;  
Number of allowed claims: 508; 
Total allowed claims: approximately $2 billion. 

Size of allowed claim: greater than or equal to $10 million;  
Number of allowed claims: 93; 
Total allowed claims: $4.5 billion. 

Source: GAO analysis of Trustee data. 

Note: Amounts shown reflect allowed claims plus amounts resulting from 
claims deemed determined through litigation. 

[End of figure] 

Trustee Has Been Pursuing Hundreds of Lawsuits to Recover Assets: 

The Madoff Trustee is pursuing various litigation to recover assets 
from customers and others that can be used to reimburse those 
customers that have allowable claims under SIPA.[Footnote 23] For 
those customers that withdrew fictitious profits in excess of their 
investments--net winners--the Trustee is pursuing more than a thousand 
lawsuits to recover these funds as allowed under federal bankruptcy 
law and state law. The Trustee is also suing some individuals and 
entities that he argues knew or should have known about the fraud and 
from whom he is seeking to recover more than just fictitious profits. 
In addition, the Trustee has filed other actions involving feeder 
funds. Through such efforts, the Trustee has obtained billions of 
dollars in settlement agreements with customers that either faced the 
possibility of litigation or had been sued already.[Footnote 24] 
However, the Trustee established a hardship program, in which he 
expedited claims processing or declined to pursue litigation for 
individuals who could demonstrate financial distress. 

Trustee's Authority and Approach to Litigation: 

As discussed earlier, various laws grant authority to the Trustee to 
seek return of funds paid out by the Madoff firm. This includes 
federal bankruptcy and state laws, which allow actions to return 
transfers from the failed entity--the debtor--made in different 
periods of time, including within 90 days, 2 years, or 6 years prior 
to the bankruptcy filing. In deciding whether to bring an action, the 
Trustee told us that he has generally considered the same factors as 
would apply in a typical bankruptcy case or those that a private 
plaintiff would also likely consider. In particular, the Trustee said 
he has considered the costs and benefits of taking the action--that 
is, how much could be recovered and at what cost--as well as prospects 
for success and potential legal barriers, such as the statute of 
limitations. 

While the overall goal of an action is to recover a reasonable amount 
for the customer property fund, the Trustee also told us that when 
filing an action, he might not seek all possible assets. Many former 
account holders were older and renting their homes, he said. At the 
same time, some customers have assets that are protected against 
judgments, such as homes or pension assets, which reduces the amount 
of assets he can pursue. The Trustee told us that while the amount of 
potential recoveries is important, he has not established any minimum 
amount in deciding whether to sue. The underlying decision on whether 
to sue, he said, is the acknowledgment that each time he declines to 
pursue an action, he allows a customer who withdrew more than their 
principal invested to keep money taken from others who had not 
recovered as much as they had invested. 

According to information the Trustee provided, his litigation has 
targeted a high portion of amounts withdrawn from the firm during the 
periods preceding failure for which he is legally authorized to pursue 
recoveries. For example, the total amount withdrawn from the Madoff 
firm during the 90-day period under which federal bankruptcy law 
allows recoveries was $5.5 billion, he said. Of that, his lawsuits 
have sought $5.2 billion, or 95 percent. The total amount withdrawn 
under the 2-and 6-year recovery periods allowed by federal and state 
law was $9.7 billion. Of that, the Trustee has sought $8.4 billion, or 
87 percent, he said. According to the Trustee, the differences between 
amounts withdrawn and sought are based on consideration of the factors 
whether to sue as noted earlier, as well as having reached settlements 
to return funds prior to filing of an action.[Footnote 25] 

The Trustee Has Filed Hundreds of Actions to Recover Excess 
Withdrawals from Customers Not Claimed to Have Known of the Fraud: 

According to our review, the Trustee has filed 1,002 actions seeking 
to recover $3.5 billion from Madoff customers that were net winners 
but that are not alleged to have had knowledge of the fraud or been in 
a position to know about it--referred to by the Trustee as "good 
faith" defendants. According to the Trustee, the good faith 
designation indicates that while defendants profited from the fraud, 
he did not have evidence they had knowledge of the scheme or were in a 
position to know about it. When bad faith is not alleged, the laws 
providing the Trustee with the ability to recover funds allow him to 
seek principal withdrawn during the 90 days preceding the Madoff 
firm's failure--the preference period--plus fictitious profits 
withdrawn in the 6 years prior to the failure.[Footnote 26] The 
Trustee told us that although he anticipated negative publicity from 
filing suits against customers who did not know about the fraud, the 
$3.5 billion at issue in the good faith cases was an amount too large 
to ignore.[Footnote 27] In these good faith cases, the amounts sought 
by the Trustee range from a low of $33,000 to a high of $152 million. 
The average amount sought was $3.5 million, with the median at $1.4 
million. 

Some of the good faith cases have settled, but about 88 percent of 
complaints remained in litigation or were on appeal as of May 2012, 
according to information we obtained from the Trustee.[Footnote 28] 
Table 3 summarizes the status of the good faith cases. In general, 
most good faith cases are proceeding slowly, with the Trustee seeking 
to mediate and settle them, he told us. 

Table 3: Status of Good Faith Litigation by Madoff Trustee, as of May 
2012: 

Status of case: In litigation; 
Number of cases: 709; 
Percentage of cases: 70.8%. 

Status of case: Some defendants dismissed and case not dismissed; 
Number of cases: 176; 
Percentage of cases: 17.6%. 

Status of case: Other[A]; 
Number of cases: 59; 
Percentage of cases: 5.9%. 

Status of case: All defendants dismissed and case closed; 
Number of cases: 57; 
Percentage of cases: 5.7%. 

Status of case: In litigation/on appeal; 
Number of cases: 1; 
Percentage of cases: 0.1%. 

Status of case: Total; 
Number of cases: 1,002; 
Percentage of cases: 100%. 

Source: GAO analysis of Madoff Trustee data. 

[A] "Other" is comprised of these categories: "All defendants 
dismissed and case not closed," "All defendants settled and dismissed 
and case closed," "All defendants settled but not dismissed and case 
not closed," and "Some defendants settled and dismissed and case not 
closed." 

[End of table] 

We examined a random sample of 50 good faith cases and found them to 
be similarly structured and generally citing the same federal and 
state laws as grounds for recovery. The good faith actions generally 
include the name of the defendants, their account numbers, the amounts 
sought, and the legal basis on which the Trustee relied to seek 
recovery. In several filings, the Trustee also included an accounting 
of deposits and withdrawals. In contrast to a bad faith action, the 
good faith complaints we reviewed generally do not include narrative 
details of the defendant's history or relation to the Madoff firm. 
Table 4 summarizes the 10 largest actions in our sample, by amount 
sought. 

Table 4: Ten Largest Good Faith Complaints by Madoff Trustee among GAO 
Sample, by Amount Sought: 

Case: 1096-1100 River Road Associates, LLC; 
Amount sought: $52,907,751; 
Type of defendant(s): Trust, individual, limited liability company; 
Status: Some defendants dismissed and case not closed. 

Case: The 1995 Jack Parker Descendant Trust No. 1; 
Amount sought: $48,808,407; 
Type of defendant(s): Trust, individual; 
Status: In litigation. 

Case: Robert L. Silverman; 
Amount sought: $30,490,003; 
Type of defendant(s): Individual, corporation, other; 
Status: In litigation. 

Case: RAR Entrepreneurial Fund LTD; 
Amount sought: $17,000,000; 
Type of defendant(s): Partnership, corporation, individual; 
Status: In litigation. 

Case: GMR S.A.; 
Amount sought: $13,172,507; 
Type of defendant(s): Corporation; 
Status: In litigation. 

Case: The Mosaic Fund, L.P.; 
Amount sought: $12,696,347; 
Type of defendant(s): Partnership, individual; 
Status: In litigation. 

Case: Shum Family Partnership III, LP; 
Amount sought: $10,000,000; 
Type of defendant(s): Partnership, corporation; 
Status: In litigation. 

Case: Bell Ventures Limited; 
Amount sought: $7,400,064; 
Type of defendant(s): Individual, other; 
Status: In litigation. 

Case: MBE Preferred Limited Partnership; 
Amount sought: $4,860,000; 
Type of defendant(s): Partnership, trust, individual, limited 
liability company; 
Status: In litigation. 

Case: Trust created under agreement 1/9/90 by Leonard Litwin; 
Amount sought: $4,588,818; 
Type of defendant(s): Trust, individual; 
Status: In litigation. 

Source: GAO review of random sample of 50 Madoff Trustee case filings. 

[End of table] 

The Trustee has pursued a variety of legal approaches in the cases we 
sampled. Thirty-eight cases cited the 2-year period provided in the 
federal bankruptcy statute, while 48 cases cited the 6-year period. In 
six of the cases, the Trustee sought to recover transfers from the 90-
day preference period. In 20 of the 50 cases, the Trustee also sued to 
recover funds he alleged were transferred from defendants to third 
parties, known as "subsequent transferees." For instance, in one case, 
the Trustee alleged that an individual account holder transferred to a 
relative some or all of $1.2 million in fictitious profits withdrawn. 
Additionally, in five cases, the Trustee has also sought to 
temporarily disallow customer claims filed by defendants. For example, 
one limited liability company filed a claim for SIPC coverage in May 
2009, and the Trustee filed suit to recover funds in December 2010. As 
part of the action, the Trustee sought to disallow the SIPC claim 
until $4 million was recovered for the Madoff estate. 

In one good faith case in our sample, the defendant was Madoff's 
nephew, considered an insider, and had received fraudulent transfers, 
but was not deemed to have acted in bad faith. Specifically, according 
to the Trustee, the nephew worked full-time at the Madoff firm 
beginning in 1980, most recently as director of administration. The 
Trustee alleged he received preference transfers, fictitious profits, 
and improperly used Madoff firm funds to pay for personal expenses. 
However, the Trustee told us his investigation indicated the relative 
was not likely aware of the fraud. 

In Some Asset Recovery Actions, the Trustee Has Alleged That 
Defendants Knew or Should Have Known of the Fraud: 

In addition to the good faith lawsuits, the trustee has also filed 30 
actions against individuals or entities in which he has alleged the 
defendants acted in bad faith because they either knew, or should have 
known, of Madoff's fraudulent investment scheme.[Footnote 29] 
According to the Trustee, asserting that a defendant acted in bad 
faith allows him to potentially recover greater amounts than in other 
actions. For example, in a bad faith action, a trustee can seek to 
recover not only fictitious profits, but also seek principal amounts 
invested as well. Additionally, while New York debtor and creditor law 
generally limits to 6 years the period for which a trustee can seek 
recovery from parties that received funds from a bankrupt entity, a 
trustee can sue a customer that acted in bad faith in earlier years, 
if it can be shown the defendant knew or should have known of the 
fraud and certain other conditions are met. The trustee can also 
pursue common law claims such as conversion and unjust enrichment 
against bad faith defendants that received other funds from the 
bankrupt entity, such as receiving cash or purchases of goods on their 
behalf.[Footnote 30] Common law claims allow the Trustee to seek 
compensatory damages to recover sums allegedly received improperly by 
the defendant. However, although the Trustee has the authority and 
standing to enforce common law claims, two recent judicial decisions 
dealing specifically with the Madoff liquidation limited his ability 
to pursue common law claims.[Footnote 31] 

Our review of the 30 bad faith actions showed that the complaints list 
several hundred defendants and seek $11 billion overall, with demand 
amounts ranging from more than $500,000 to approximately $6.7 billion. 
[Footnote 32] Defendants include Madoff family members, employees of 
the Madoff firm, individuals who identified investors for Madoff, and 
other business associates of Madoff. In addition to suing defendants 
as individuals, the complaints also seek to recover funds from 
vehicles these individuals used to invest in the Madoff scheme, 
including corporations, limited partnerships, trusts, estates, 
partnerships, foundations, and profit-sharing plans. Additionally, the 
Trustee is suing individual retirement accounts, to recover funds 
allegedly received from another defendant. According to the Trustee, 
his objective in alleging bad faith, even for some net loser 
defendants, is to maximize recoveries for distribution to other harmed 
customers.[Footnote 33] 

The Trustee has cited various factors in arguing that the defendants 
in the bad faith cases knew or should have known of the Madoff fraud. 
These factors include being an employee or officer of the Madoff firm, 
and acting to perpetuate the fraud or being in a position that should 
have allowed them to identify it. For example, among bad faith 
defendants is the manager of the Madoff computer system that generated 
falsified transactions and customer statements. The Trustee is also 
suing Madoff's brother, who was the firm's chief compliance officer, 
whose duties were to ensure the firm was complying with applicable 
laws. Other individuals being sued for bad faith include those who had 
close enough access to either Madoff or his firm that they could or 
should have made inquiries that could have revealed the Ponzi scheme. 
In some instances, the Trustee alleges that defendants were also 
sophisticated investors with sufficient education or specialized work 
experience in securities law, accounting, or finance that they should 
have known their returns were unusual. Table 5 shows the most 
frequently cited bases on which the Trustee alleged bad faith, 
according to our analysis of available cases. 

Table 5: Most Prevalent Bases on Which the Madoff Trustee Has Alleged 
Customers or Others Acted in Bad Faith: 

Basis for allegation defendant(s) knew or should have known of the 
fraud: Defendant enjoyed unique access to Madoff firm, or had close 
personal relationship with Madoff; 
Number (percentage) of bad faith cases (N=29): 29; (100%). 

Basis for allegation defendant(s) knew or should have known of the 
fraud: Defendant received funds or financial benefits from Madoff firm 
in manner that was clearly fraudulent or suspicious in nature; 
Number (percentage) of bad faith cases (N=29): 24; (83%). 

Basis for allegation defendant(s) knew or should have known of the 
fraud: Defendant was Madoff employee or other associate who actively 
conducted or concealed fraud, or was Madoff officer who breached 
fiduciary duty of due diligence; 
Number (percentage) of bad faith cases (N=29): 21; (72%). 

Basis for allegation defendant(s) knew or should have known of the 
fraud: Madoff accounts held by defendant experienced unusual purported 
transactions such as backdating or reversing or canceling of trades; 
or included guaranteed rates of return; 
Number (percentage) of bad faith cases (N=29): 19; (66%). 

Basis for allegation defendant(s) knew or should have known of the 
fraud: Defendant was either reasonably prudent or sophisticated 
investor, based on education or work experience; 
Number (percentage) of bad faith cases (N=29): 16; (55%). 

Source: GAO analysis of bad faith complaints filed by Madoff Trustee. 

Notes: A single bad faith case may cite multiple bases for acting in 
bad faith. Percentage figures based on 29 complaints publicly 
available, and excluding 1 complaint filed under seal. 

[End of table] 

The Trustee told us that initially, he believed that anyone working in 
Madoff's investment advisory unit would be a bad faith participant. 
However, through additional investigation, including document 
examination, taking of depositions, and reconstruction of computer 
records, the Trustee said he determined not all such employees may 
have been aware of the fraud. For example, the Trustee initially 
believed Madoff's secretary was involved in the fraud, given her 
position close to Madoff, but later determined she did not have 
knowledge of the fraudulent activities.[Footnote 34] 

In several cases, the Trustee is alleging bad faith against Madoff 
relatives who worked at the firm, or other, nonemployee relatives who 
received funds. For example, the Trustee sued Madoff's wife to recover 
funds transferred from the firm that were used to buy a yacht for 
approximately $2.8 million, to pay off a $1.1 million credit card 
balance, or to fund other investments. The Trustee sued the wife of 
Madoff's brother, who worked at the firm, for receiving $1.5 million 
in purported salary from 1996 to 2008, after the Trustee's 
investigation concluded she never performed any work for the position. 
Among defendants who were not employees or relatives are individuals 
who allegedly identified and recruited investors for the Madoff firm. 
For example, the Trustee alleges an accounting firm pooled hundreds of 
millions of dollars for investment while keeping tens of millions of 
dollars itself. 

Although many Madoff customers typically experienced outsized returns 
based on market developments, some bad faith defendants received 
exceptionally high returns, the Trustee alleges. In one case, for 
example, the Trustee cites returns as high as 175 percent. Moreover, 
in some cases, the bad faith defendants include large investors for 
whom the trustee alleges their financial or business sophistication 
provided them with the ability to realize that they were benefiting 
from a fraud. For example, the Trustee alleges one set of defendants 
is a closely held family business that was also an investor in another 
investment fraud that was so similar to the Madoff firm it should have 
been clear Madoff was also running a fraud. In his complaint, the 
Trustee quotes an employee of the business as saying shortly after 
Madoff's arrest, "Our CIO [chief information officer] always said it 
was a scam, 'too good to be true'[.]" Another bad faith defendant, 
according to the Trustee, had been closely associated with Madoff 
professionally and socially for decades, investing in the firm through 
more than 60 entity and personal accounts. According to the Trustee, 
some of the accounts reported consistently high annual returns between 
20 percent and 24 percent, with only 3 months of negative returns over 
12 years. Other accounts of the defendant sometimes experienced 
returns greater than 100 percent or even 300 percent, he alleges. 
According to the Trustee, the defendant acted as an investment 
advisor, and thus should have known such returns were not likely 
possible without fraud. 

As shown in table 6, the 10 largest bad faith complaints, measured by 
amount sought, have sought a total of $10.7 billion, or more than 97 
percent of the $11.0 billion the Trustee is seeking in all 30 bad 
faith litigations. As of April 2012, the Trustee had obtained 
settlements, or a partial settlement, in four of these cases. 

Table 6: Ten Largest Bad Faith Complaints by Madoff Trustee, by Amount 
Sought: 

Bad faith complaint: Picower; 
Type/description: Individual with long ties to Madoff; 
Amount sought: $6,746,066,538; 
Amount sought as percentage of total bad faith actions: 61.3%. 

Bad faith complaint: Chais; 
Type/description: Individual with long ties to Madoff; 
oversaw funds invested with Madoff firm; 
Amount sought: $1,152,597,632; 
Amount sought as percentage of total bad faith actions: 10.5%. 

Bad faith complaint: Katz; 
Type/description: Sophisticated investors with long ties to Madoff; 
Amount sought: $1,018,098,199; 
Amount sought as percentage of total bad faith actions: 9.3%. 

Bad faith complaint: Avellino & Bienes; 
Type/description: Individuals and entities that helped perpetuate and 
sustain the fraud; 
Amount sought: $904,515,689; 
Amount sought as percentage of total bad faith actions: 8.2%. 

Bad faith complaint: Peter B. Madoff; 
Type/description: Madoff family members, sophisticated financial 
professionals, and Madoff firm corporate officers; 
Amount sought: $226,374,622; 
Amount sought as percentage of total bad faith actions: 2.1%. 

Bad faith complaint: Cohmad Securities Inc; 
Type/description: Individuals and entities that helped perpetuate and 
sustain the fraud; 
Amount sought: $213,080,310; 
Amount sought as percentage of total bad faith actions: 1.9%. 

Bad faith complaint: Magnify Inc; 
Type/description: Individuals and entities that were sophisticated 
investors with unique access to Madoff that channeled money to 
international entities; 
Amount sought: $154,403,464; 
Amount sought as percentage of total bad faith actions: 1.4%. 

Bad faith complaint: Glantz; 
Type/description: Individuals who were sophisticated investors and had 
long ties to Madoff; 
Amount sought: $113,385,537; 
Amount sought as percentage of total bad faith actions: 1.0%. 

Bad faith complaint: Blumenfeld; 
Type/description: Individual with close ties to Madoff; 
Amount sought: $100,033,990; 
Amount sought as percentage of total bad faith actions: 0.9%. 

Bad faith complaint: American Securities Management LP; 
Type/description: Sophisticated investors and entities with close ties 
to Madoff; 
Amount sought: $91,896,296; 
Amount sought as percentage of total bad faith actions: 0.8%. 

Bad faith complaint: Total; 
Type/description: [Empty]; 
Amount sought: $10,720,452,277; 
Amount sought as percentage of total bad faith actions: 97.4%. 

Source: GAO analysis of Madoff Trustee data and bad faith complaints 
filed by the Trustee. 

[End of table] 

Trustee Seeking Nearly $100 Billion in Cases Involving Entities That 
Channeled Investor Funds to the Madoff Firm: 

As noted, feeder funds are investment vehicles that collected funds 
from investors and then channeled the money to the Madoff firm. The 
Trustee has filed 27 bad faith actions against feeder fund defendants, 
seeking nearly $100 billion. The amounts sought range from a low of 
$10.4 million to a high of $58.5 billion. The average amount sought is 
$3.7 billion, with a median of $182.4 million. The amounts the Trustee 
is seeking include fictitious profits, principal amounts invested, 
fees, interest, and in some cases, punitive damages. Typically, the 
Madoff feeder funds raised money from high-net-worth individuals in 
operations that spanned the globe. In addition to the feeder funds 
themselves, banks and other financial institutions are among the 
defendants in these cases. For example, some were custodians of the 
feeder funds, some were administrative agents, and some were involved 
in marketing the funds to prospective investors. Some banks and 
financial institutions had more than one role in their involvement 
with feeder funds. 

According to the Trustee, feeder funds account for about $14.2 billion 
of the approximately $19.6 billion in total principal lost by all 
customers. In suing for bad faith, the Trustee has alleged these 
defendants either knew or should have known of the Madoff fraud. In 
some of the feeder fund cases, the Trustee is also alleging 
participation in, and concealment of, the fraud. With the bad faith 
allegations, as with bad faith cases filed against individuals, the 
Trustee can sue to recover funds beyond principal withdrawn in the 
preference period and fictitious profits withdrawn in the 6-year 
period. In addition, he can use common law grounds to seek additional 
damages based on alleged harm caused by parties aiding the fraud. As a 
result, the total he is seeking in the six largest feeder fund actions 
is $94 billion, which greatly exceeds the amount of principal these 
entities invested into the scheme. Table 7 summarizes these cases, 
which represent nearly 95 percent of the total amount sought in feeder 
fund cases. 

Table 7: Key Allegations in Six Largest Feeder Fund Complaints by 
Madoff Trustee, by Amount Sought: 

Case: Kohn; 
Total sought: $58.5 billion; 
Key Trustee allegations: 
* Defendant, who operated as Madoff insider, knew Madoff was a fraud 
but fed more than $9.1 billion to the firm over more than 20 years, 
often from wealthy Europeans; 
* Madoff secretly paid Kohn at least $62 million for steering funds to 
his firm; 
Status: Some defendants dismissed and case not closed. 

Case: JP Morgan Chase; 
Total sought: $19 billion; 
Key Trustee allegations: 
* Madoff firm's primary banker for more than 20 years. Its due 
diligence revealed the fraud, but the bank continued business as usual; 
* Defendant received at least $500 million in fees and profits from 
Madoff business and also selling structured financial products tied to 
Madoff feeder funds; 
Status: In litigation. 

Case: HSBC; 
Total sought: $9 billion; 
Key Trustee allegations: 
* Enabled the fraud by encouraging more than $8.9 billion in 
investments in an international network of feeder funds through its 
role as marketer, custodian, and administrator of the funds; 
* Defendants ignored numerous indicators of fraud and surrendered all 
custodial duties to the Madoff firm without any disclosure to 
investors; 
* Aided and abetted the fraud in order to reap more than $400 million 
in fees; 
Status: In litigation/on appeal. 

Case: Fairfield Sentry Limited; 
Total sought: $3.5 billion; 
Key Trustee allegations: 
* Madoff firm's largest feeder fund operator, with nearly $14 billion 
invested in 2008. It failed to perform due diligence after signs of 
fraud and misled regulators and investors; 
* Defendants had actual knowledge of the fraud and received more than 
$1 billion in fees; 
Status: Some defendants settled but not dismissed/on appeal. 

Case: Tremont Group Holdings; 
Total sought: $2.1 billion; 
Key Trustee allegations: 
* Madoff firm's second largest feeder fund operator channeled more 
than $4 billion into the fraud; 
* Defendants failed to perform due diligence after signs of fraud, 
despite repeated warnings and opportunities; 
* Defendants received more than $240 million in fees over 15 years; 
Status: Some defendants settled but not dismissed on appeal. 

Case: UBS; 
Total sought: $2 billion; 
Key Trustee allegations: 
* UBS partnered with Access International Advisors to enable the fraud 
through international feeder funds; 
* UBS served as custodian and administrator for the feeder funds, only 
to surrender all custodial duties to the Madoff firm without 
disclosure to investors; 
also relied upon the firm to verify existence of trades; 
* Based on its own due diligence, UBS knew the Madoff firm was likely 
a fraud, but continued to do business as usual and earned at least $80 
million in fees; 
Status: In litigation. 

Case: Total; 
Total sought: $94.1 billion. 

Source: GAO review of individual case filings. 

[End of table] 

The Trustee told us that compared to other litigation he is pursuing, 
feeder fund actions require different types of evidence, due to the 
nature of their operations. For example, feeder fund managers 
benefited through investment and management fees charged by the funds, 
which were subsequently paid to individuals as salaries and bonuses. 
The Trustee considers these payments to be fraudulent. In the Tremont 
case, for instance, the firm managed five funds that invested directly 
with the Madoff firm, as well as more than a dozen other funds that 
were indirect Madoff investors, via investments in the directly 
invested funds. The Trustee alleged that in the 6 years before the 
Madoff firm failed, Tremont defendants received more than $180 million 
in management, administration, and other fees; bonuses; profits; 
compensation; dividends; and partnership distributions. Additionally, 
a number of the feeder fund defendants were net losers. In such cases, 
the Trustee also sought to defer SIPA claims they filed pending 
resolution of the Trustee's actions against them. In general, given 
the complexities of feeder fund relationships, the Trustee determined 
amounts sought by looking at principal, not fictitious profits, he 
told us.[Footnote 35] 

As shown in table 7, the Trustee has alleged that banks, financial 
services entities, and related individuals facilitated the Madoff 
fraud. For example, HSBC served as marketer, custodian, and 
administrator for numerous feeder funds. The Trustee alleged that HSBC 
surrendered all custodial duties to the Madoff firm, while continuing 
to collect fees, and without any disclosure to investors. The Trustee 
stated that this surrender removed a system of checks and balances and 
allowed the Madoff firm to assert the existence of assets and trades 
that never existed. As for marketing feeder funds, the Trustee 
maintains that HSBC acquiesced to Madoff's demands to keep his name 
out of offering documents, despite the bank's own concerns about its 
inability to conduct proper due diligence on Greenwich Sentry, a 
Madoff feeder fund. Feeder funds also worked with banks to create 
derivative products based on feeder fund returns, according to the 
Trustee. For example, investors in these "leveraged notes" would be 
entitled to receive returns based on a feeder fund's returns, while a 
financial institution, usually a bank, would receive fees for 
structuring the notes and interest for lending funds used as part of 
the investment. Concurrently, the bank would purchase shares in the 
feeder fund in order to hedge its exposure in the leveraged notes. The 
end result, according to the Trustee, was that hundreds of millions of 
additional dollars were invested into the Madoff operation. 

In bringing the feeder fund actions, the Trustee has made a number of 
specific allegations to support bad faith and complaints of illicit 
funds received. While particular allegations vary across cases, we 
found the six cases we examined shared three major elements. First, 
defendants are alleged to have profited from the fraud primarily 
through fees received. For example, UBS is alleged to have received 
fees for purportedly serving in custodial and asset management 
functions for the feeder funds, Luxalpha and Groupement Financier. UBS 
sponsored the formation of Luxalpha and served as prime banker for 
Groupement Financier. The Trustee considers the fees UBS derived to be 
customer property that should be recovered. Second, the Trustee 
alleges defendants breached their duty of due diligence to their 
customers. For example, the Trustee alleged that Fairfield Greenwich 
Group, Madoff's largest feeder fund group, did not "properly, 
independently, and reasonably perform due diligence into the many red 
flags strongly indicating Madoff was a fraud." Third, the defendants 
are alleged to have either aided and abetted or actively participated 
in the fraud. In the JP Morgan Chase case, the Trustee alleged that 
through its interactions in different capacities--as banker, lender, 
and investor--with the Madoff firm over 20 years, the bank was 
uniquely positioned to see the fraud and put a stop to it. Instead, 
the trustee alleges that this institution continued to conduct 
business as usual, which allowed the firm to profit and the fraud to 
continue unabated. 

As with other actions, the Trustee has sought feeder fund recoveries 
based on the 90-day preference period and the 2-and 6-year periods, 
relying on both the federal bankruptcy statute and New York state law. 
In instances where the defendant was considered an insider, the 
Trustee sought to extend the preference period to 1 year. The Trustee 
has also sought to recover transfers to subsequent transferees. In a 
number of cases, for example, the Trustee has pursued this course 
against individuals paid a salary or a bonus by the banks or feeder 
funds involved. 

Trustee Settlements to Date Have Recovered Nearly Half of Customer 
Investment Losses: 

In hundreds of cases, the Madoff Trustee has reached settlements with 
former customers and others, either before or after filing clawback 
actions, and these agreements have produced recovery of a significant 
amount of assets. As of April 2012, the Trustee had entered into 441 
settlement agreements in which the opposing parties agreed to return 
about $8.4 billion--an amount equal to about 49 percent of the 
approximately $17.3 billion in principal investments lost by customers 
who filed claims. According to our review, the settlement amounts 
range from a low of $36 to a high of $5 billion, with an average of 
$19 million and a median of $66,000. Through July 2012, the Trustee 
had collected 85 percent of total settlement amounts, or about $7.1 
billion. Many settlement terms are complex. For example, settlements 
with feeder funds that were net losers require these entities to 
return certain funds they received before the Trustee will consider 
accepting their loss claims. 

The Trustee groups settlements into three major categories: 

* Prelitigation: Settlements reached before the Trustee filed a 
clawback action. 

* Litigation: Settlements reached in cases where the Trustee had 
already filed a clawback action. 

* Customer avoidances: Recoveries based on the 90-day preference 
period, but where no clawback action was filed. 

A fourth category, "Funds not yet received," is a temporary accounting 
of amounts that will be allocated to the three main categories upon 
receipt of the first settlement payment. Table 8 shows a summary of 
the Trustee's settlement agreements by category as of July 2012. As 
the table shows, for example, the customer avoidance category 
represents 85 percent of all settlement agreements reached, although 
these cases have the smallest total dollar value among the categories. 
Although, as the table shows, the settlements total $8.4 billion, the 
Trustee through July 2012 had yet to receive $1.2 billion in 
settlement funds, due to pending appeals and specific provisions in 
settlement agreements. 

Table 8: Settlement Agreements Reached by Madoff Trustee, by Category, 
as of July 2012: 

Customer avoidances; 
Number: 375; 
Percentage of settlements: 85%; 
Total settlement amount: $117,173,238; 
Amount collected: $114,642,776; 
Percentage collected: 97.8%. 

Litigation settlements; 
Number: 44; 
Percentage of settlements: 10; 
Total settlement amount: $5,185,109,046; 
Amount collected: $5,099,292,433; 
Percentage collected: 98.3%. 

Prelitigation settlements; 
Number: 13; 
Percentage of settlements: 3; 
Total settlement amount: $1,892,758,633; 
Amount collected: $1,892,758,598; 
Percentage collected: 100%. 

Funds not yet received[A]; 
Number: 9; 
Percentage of settlements: 2; 
Total settlement amount: $1,199,150,000; 
Amount collected: N/A; 
Percentage collected: N/A. 

Total; 
Number: 441; 
Percentage of settlements: 100; 
Total settlement amount: $8,394,190,917; 
Amount collected: $7,106,693,807; 
Percentage collected: 84.7%. 

Source: GAO analysis of Madoff Trustee data. 

[A] Includes settlements of all types for which funds have yet to be 
received. Amounts from settlements under appeal and settlements with 
provisions for future payments fall into this category. 

[End of table] 

The Trustee told us he and his counsel consider a number of factors 
when deciding to enter into a settlement agreement. There are no 
formal criteria, he said, but factors such as location of the 
defendants, litigation risk, and timing of the clawback can influence 
the settlement decision. Two of the most important factors, the 
Trustee told us, are the defendant's ability to pay and whether the 
settlement will produce proceeds that enhance the customer fund he is 
building during the liquidation. The bankruptcy court must approve 
settlement agreements worth $20 million or more. In addition to 
whether a settlement is in the best interest of the estate, the court 
also considers whether a proposed settlement is fair and equitable, 
and above "the lowest point in the range of reasonableness," the 
Trustee told us. In deciding whether a particular agreement falls 
within that range, the court considers: 

* probability of success in litigation; 

* difficulties of collection; 

* litigation complexity, and expense, inconvenience, and potential for 
delay; and: 

* creditor interests. 

On the key issue of ability to pay, the Trustee said his approach is 
that he would rather agree to a settlement for less money and be able 
to collect it, than win a judgment for a greater sum that he is 
ultimately unable to collect. He cited the Katz-Wilpon settlement as 
an example.[Footnote 36] Originally, the Trustee sued for more than $1 
billion, but reached a settlement in April 2012 for $162 million, 
based on his conclusion of what the defendants could pay. In other 
areas, the Trustee also told us he is sensitive about clawing back 
funds from charitable institutions, saying that although they have 
received other investors' funds, he did not want to put them in the 
position of raising money in order to pay a settlement. We examined 
all but one settlement agreement among those worth at least $20 
million.[Footnote 37] Appendix IV provides details of these 
settlements, including amounts sought and obtained, the extent to 
which settlements have been paid, strategies behind settlements, and 
key provisions. 

A number of factors motivate counterparties to settle with him, the 
Trustee told us. One is that defendants have fiduciary duties to 
clients or customers, and amounts at issue in a case might be so 
significant that, for example, feeder fund managers might conclude 
their duty is to settle in order to recover assets for customers. Some 
parties, to protect client confidentiality, may settle in order to 
prevent disclosure of client information that could become public in 
litigation. Other parties might seek closure. Overall, the Trustee 
told us he thinks he has built a solid record in his settlements. The 
market itself has validated his efforts, he said, which can be seen in 
increases in the price of Madoff claims being traded following 
announcement of recent settlement agreements.[Footnote 38] 

As part of our review of records provided by the Trustee, we noted 
some customer accounts having a negative balance. For example, in the 
Picower case, the records showed a negative balance of $6.3 billion. 
In theory, this reflected some kind of margin account or debit 
account, the Trustee told us, even though such an amount would not 
have been in keeping with standard industry practices. Such negative 
balances raised questions whether the reported amounts represented 
debt owed by customers to the Madoff firm, and if so, whether the 
presence of such debt diminished the value of settlements obtained 
from such customers. To obtain the $7.2 billion Picower settlement, 
for example, the Trustee told us that he agreed to characterize the 
amount being returned as a loan repayment. However, both the Trustee 
and SIPC executives told us that the use of this terminology, or 
presence of negative balances, does not change the effect of 
underlying actions, which was to extract fictitious profits from the 
Madoff firm. In particular, the Trustee told us, he determined claims 
for all accounts based on the money invested less money withdrawn 
method. Purported indebtedness or negative balances did not affect 
those calculations, he said, and had no effect on settlement efforts. 

Under a Hardship Program, the Trustee Expedited Claims and Avoided 
Lawsuits for Individuals in Financial Distress: 

The Trustee created a hardship program in which he considered the 
degree of financial distress of Madoff customers in processing claims 
and deciding whether to pursue clawback litigation. He told us that 
the program was to recognize harm the Madoff fraud caused to former 
customers. No other SIPC liquidation has had such an option, the 
Trustee told us. This program, which was not open to institutional 
customers, had two elements for individuals who could demonstrate 
financial hardship: 

* Claims. The Trustee provided expedited consideration of claim 
applications. This did not provide applicants with any more favorable 
treatment, as claims were still determined on the NIM basis. But it 
accelerated consideration of claim applications, which, if approved, 
could result in customers receiving any SIPC advances due them more 
quickly. According to the Trustee, if complete account information was 
available, a determination on qualification for expedited claim review 
was made in about 20 to 30 days, with another 20 to 30 days to 
determine the actual claim. This compares to a typical claim taking 3 
months or more. 

* Clawbacks. The Trustee would not sue for clawbacks, or would drop 
suits already filed.[Footnote 39] 

The Trustee told us he invited applications to the hardship program, 
but also included some claimants in the program on his own initiative. 
Figure 7 shows a breakdown of the Trustee's consideration of hardship 
cases. 

Figure 7: Madoff Trustee Hardship Program Process and Outcomes, as of 
May 2012: 

[Refer to PDF for image: illustration] 

Hardship program: 
To recognize financial hardship the Madoff fraud created for 
customers, the Trustee said he created a hardship program. It has two 
elements—expedited consideration of claims applications, and declining 
to pursue "clawback" actions to recover funds withdrawn from the firm. 
This figure shows outcomes of each element. 

Claim processing: 
Approved: 214: 
* Net losers: 93 (43%); 
* Net winners or zero balance: 121 (57%); 
Denied: 127: 
* Lack of hardship: 22 (17%); 
* No account or not individual: 45 (35%); 
* Net winners: 60 (47%). 

Clawback actions: 

No action: 
Based on information received in claims hardship application: 167; 
Based on other information obtained by Trustee: 93. 

Applications received: 350: 
* Approved: 125 (36%); 
* Denied: 40 (11%); 
* Referred to settlement: 82 (23%); 
* Pending: 103 (29%). 

Source: GAO analysis of Trustee data. 

Note: Percentages are rounded. 

[End of figure] 

According to the Trustee, he assessed general factors in considering 
hardship applications, but there were no formal criteria or decision 
rules. Instead, the Trustee told us, he applied his judgment after 
applications were reviewed at the Trustee's counsel law firm. The 
general factors, applicable to both the claims and clawback elements 
of the program, included whether, due to lost investments or the 
possibility that funds withdrawn must be returned, the customer: 

* needed to return to work, 

* had declared bankruptcy, 

* was unable to pay for living expenses, 

* was unable to pay for dependents, or: 

* suffered from health problems. 

Additionally, the Trustee said he took into account whether former 
customers used any fictitious profits to pay taxes. If so, the Trustee 
told us he considered such payments in his decisions. 

The process for considering hardship applications was similar for both 
the claims and clawback elements, the Trustee told us. For claims, an 
attorney reviewed an application and made a recommendation to the 
Trustee, who then decided the matter. For clawbacks, the Trustee's 
counsel team responsible for the litigation reviewed the application, 
and, if necessary, would seek additional information. The team would 
make a recommendation to a review committee, comprised of five 
lawyers, for its evaluation. The committee would make a recommendation 
to the Trustee, who then decided the matter. 

According to information we reviewed, hardship program applicants were 
predominantly older, and the most commonly cited reasons for financial 
distress, for either element, were the inability to pay living 
expenses and health problems. The Trustee said most cases were not 
difficult to decide, based on the evidence presented. He told us that 
while he and others reviewed applications carefully, requesting 
additional information when necessary, they attempted to not be overly 
intrusive into private affairs. We reviewed a number of applications 
provided by the Trustee, illustrating acceptance or rejection of 
customers' hardship applications. Table 9 summarizes some of the cases 
we reviewed. 

Table 9: Examples of Customer Circumstances among Hardship 
Applications Accepted and Denied: 

Hardship program option: Claims; Application approved: Former customer 
said he suffered catastrophic injury, requiring multiple surgeries, 
forcing him out of work and jeopardizing financial stability. "I am 
scared of becoming homeless overnight. Please do not allow me to become…
a virtual ward of the state"; 
Application denied: Couple with primary residence and two vacation 
properties said they were forced to borrow to meet annual expenses of 
$450,000 plus living costs, and that SIPC payment would allow them to 
hold on until real estate market hopefully improved. 

Hardship program option: Clawbacks; Application approved: Older woman 
said she lost husband, and his disability payments; had only Social 
Security income; and turned to food stamps, Medicaid, and grants from 
local charity. Approving hardship request "might allow me to sleep at 
night without the fear of being homeless"; 
Application denied: Older man with annual income of about $350,000, 
assets of about $7.5 million, and no liabilities said higher expected 
costs for medicine and care would force use of savings. "I would 
greatly appreciate your revoking the lawsuit that you have started 
against me." 

Source: GAO review of Madoff Trustee information. 

[End of table] 

As noted, the claims hardship part of the program did not alter 
outcomes, as it provided only expedited review. For the clawback 
portion, results varied by type of action, according to information 
the Trustee supplied in response to our queries. Figure 8 summarizes 
results. 

Figure 8: Results of the Clawback Element of the Madoff Trustee's 
Hardship Program, as of June 2012: 

[Refer to PDF for image: 2 pie-charts] 

Defendants dismissed or not sued: 9%; 
Dismissed: 191; 
Not sued: 259. 
About 5,000 defendants in all clawback actions. 

Actions dismissed or not filed: 28%; 
Dismissed: 69; 
Not filed: 259; 
1,133 clawback actions filed.  

Source: Madoff Trustee. 

[End of figure] 

SEC and SIPC Involved in Trustee's Litigation and Consider It 
Successful Thus Far: 

Within weeks of the Madoff firm's failure, SEC officials were studying 
whether clawback actions were permissible, and concluded they were. 
SEC officials told us the issue was a difficult one, because innocent 
investors would become the target of lawsuits; at the same time, if 
the Trustee did not pursue the recovery actions, others would be hurt. 
From the beginning of the case, SEC thought that clawbacks would 
produce the bulk of assets available for distribution to customers, 
because billions of dollars had been withdrawn from the firm shortly 
before its failure. 

SEC officials told us they have had discussions with the Trustee (and 
SIPC) on his clawback litigation, on such issues as legal theories 
being employed, risks presented, legal costs, and expected future 
developments. But they have not had day-to-day involvement in the 
litigation, nor been involved in developing the legal strategies 
employed, the officials told us. SEC officials told us an examination 
will come later, but based on their experience, the Trustee appears to 
be conducting the litigation in an acceptable manner and has applied 
the law properly and fairly. They said they were not aware of any 
particular problems or issues with the litigation, although courts 
have not always adopted the Trustee's positions. SEC officials told us 
they were not concerned about runaway or needless litigation, and the 
main outcome of the litigation is that the Trustee has recovered large 
sums that are considerably more than initially expected.[Footnote 40] 

SIPC executives likewise told us they always thought clawbacks would 
be a critical part of recovering assets for customers. SIPC did not 
have any discussions with the Trustee about clawbacks prior to his 
appointment, they said, other than to discuss whether there were 
adequate resources available at his law firm to bring the expected 
cases. The officials characterized their involvement with the Trustee 
as providing institutional knowledge and high-level advice and 
strategy, albeit with the understanding that the independent Trustee 
need not accept it. SIPC's most significant contribution is payment of 
litigation expenses, SIPC executives told us, because otherwise, the 
Trustee could not finance his litigation. The SIPC executives told us 
they consider the Trustee's litigation a success, but that his efforts 
are thus far incomplete, as many cases remain outstanding. Although 
total settlement amounts to date reflect mostly one case ($5 billion 
for Picower), other settlements still have produced significant sums. 
The cases the Trustee is pursuing are complex, involving difficult and 
time-consuming legal work to resolve, the executives said. Costs of 
the litigation are favorable when compared to what attorneys would 
collect in contingent lawsuits, they said. 

Effects of the Fraud on Income Tax Revenues Cannot Be Determined, and 
New Guidance on Clawbacks Could Reduce Taxpayer Errors: 

Because the Madoff fraud affects customers' taxable income, it also 
affects federal tax collections by the U.S. Treasury. Madoff customers 
can seek tax relief for fraud-related losses in several ways, 
including one special procedure announced by IRS in the wake of the 
firm's failure. However, IRS officials were unable to quantify the 
overall impact of the fraud on tax collections, and the impact may be 
reduced by various factors that could limit taxpayers' ability to take 
full advantage of the tax relief available. In addition, while the use 
of either NIM or FSM to determine customer net equity could lead to 
different outcomes for account holders, either method likely reduces 
tax revenues. Tax experts expressed concerns about the lack of clarity 
over how payments stemming from fraud-related avoidance actions, or 
clawbacks, filed by the Trustee will be treated for tax purposes. 
After we identified concerns to IRS that lack of guidance could lead 
to taxpayer errors resulting in over-or underpayment of taxes, the 
agency issued such guidance. 

Madoff Customers Can Seek Tax Relief Several Ways: 

The Madoff fraud affected the federal income tax liabilities of former 
customers in two primary ways, according to our review.[Footnote 41] 
First, customers likely paid federal income taxes on fictitious 
profits reported to them in each year they held their account. 
[Footnote 42] Second, they likely suffered theft of funds invested, 
which under tax law would be considered an investment theft loss. 

Typically, there are two ways to address these effects upon discovery 
of such fraud, according to IRS officials.[Footnote 43] Taxpayers can 
file amended returns, in which they remove fictitious profits from 
previously reported income. In addition, they can claim a theft loss 
deduction against income, to reflect principal amounts stolen and 
fictitious profits reflected on account statements that were not 
removed on any amended returns.[Footnote 44] Under the Internal 
Revenue Code, an investment theft occurs when a taxpayer loses 
property to theft in connection with a transaction entered into for 
profit. Taxpayers can use amended returns or claim theft loss 
deductions, singly or in tandem, depending on their situation, 
according to IRS officials.[Footnote 45] But each of these approaches 
has limitations, according to the officials. For instance, taxpayers 
can generally file an amended return to claim a refund of taxes paid 
within 3 years of the original filing date, or 2 years from the date 
the tax was paid, whichever is later. In the case of the theft loss 
deduction, taxpayers cannot take the deduction to the extent they have 
been reimbursed for the loss, and if they have a claim for 
reimbursement, they cannot deduct their loss until the amount of 
recovery to be received can be "ascertained with reasonable 
certainty." In the Madoff case, the Trustee is working to recover 
assets to benefit former customers, but how much he will ultimately 
recover, and by when, remains unknown. 

In addition to these typical remedies, IRS has also provided another 
option, referred to as a "safe harbor" approach. This allows taxpayers 
to deduct a percentage of lost principal including all previously 
reported profits in a single year--the year of a criminal charge 
against the perpetrator, or 2008 for the Madoff case.[Footnote 46] 
According to IRS, the purpose of the safe harbor is to ease the 
compliance burden for both the agency and taxpayers, avoiding what can 
be complicated questions on size and timing of a theft loss deduction. 
Under the safe harbor approach, taxpayers can deduct 95 percent of 
their losses in the year of discovery (when the lead figure is 
criminally charged). The loss is calculated by adding principal 
invested plus profits (whether fictitious or real), less cash 
withdrawals and recoveries from SIPC or other sources. For taxpayers 
seeking recovery from third parties, such as through lawsuits, the 
figure is reduced to 75 percent. If the taxpayer follows the safe 
harbor requirements, IRS agrees not to challenge the taxpayer's 
treatment of a qualified loss as a theft loss, and taxpayers waive 
their right to other remedies that might have been available.[Footnote 
47] The safe harbor approach deems the reasonable prospect of recovery 
condition of the standard theft loss deduction to be satisfied in the 
year of discovery. Thus, Madoff customers electing to use the safe 
harbor approach may be able to recognize their losses earlier than 
under the normal method for deducting a theft loss.[Footnote 48] 
However, if customers using the safe harbor approach later receive 
distributions of recovered assets from the Trustee that cause their 
claimed deduction to exceed their actual losses, they must report the 
excess amounts as income in the year received.[Footnote 49] 

IRS officials told us they could not say which method of claiming tax 
relief is best for taxpayers, because individual tax situations can 
vary widely.[Footnote 50] Tax practitioners to whom we spoke said the 
safe harbor is attractive for its certainty and ease of use, but some 
taxpayers may be better off using the traditional methods. However, 
one practitioner told us that anyone who can use the safe harbor 
should do so. 

The Fraud's Impact on Tax Collections Cannot Be Determined, but It May 
Be Reduced by Limits on Taxpayers' Ability to Fully Realize Tax Relief: 

According to IRS officials, the agency cannot determine the tax 
revenue loss to the U.S. Treasury that will result from Madoff 
customers seeking relief for their fraud-related losses. IRS cannot 
identify Madoff taxpayers, and even if it could, it does not collect 
necessary information to conduct a post-Madoff analysis of the fraud's 
impact on tax revenues. IRS officials told us they generally do not 
maintain statistics on any particular Ponzi scheme or identified 
investment fraud. They also told us they cannot identify which Madoff 
customers are using which tax relief method, further complicating any 
effort to assess the impact of the fraud on tax revenues. 

In any case, although IRS cannot determine the amount of any revenue 
loss, the Madoff fraud's effect on tax collections could be reduced by 
various factors that can limit taxpayers' ability to take full 
advantage of their losses. These factors generally affect the ability 
of taxpayers to claim an investment theft loss deduction, rather than 
limit the ability to file amended returns. According to IRS officials, 
SIPC executives, and tax practitioners, factors affecting the ability 
to make use of the theft loss deduction include: 

* Deductions need income. The theft loss deduction is a deduction 
against income, not a tax credit. Therefore, to use the deduction, 
taxpayers must have income to apply it against. If they do not have 
sufficient income, they cannot use all or part of the deduction. This 
could be a common situation, because many Madoff customers are older 
and without income. If taxpayers have insufficient income to make use 
of the deduction in a particular year, IRS rules allow theft loss 
deductions to be carried over to other years--generally, backwards for 
3 years, and forward for 20. But taxpayers must still have income, and 
even if they do, it could take a number of years to fully apply their 
deductions on that income, meaning that benefits could be delayed. One 
tax practitioner told us that even with extended carryback and 
carryforward periods, he would expect many people--especially smaller 
investors--will not be able to use their deductions, for lack of 
income against which to apply it. 

* "Leakage." There can be considerable "leakage" when using the 
carryforward and carryback options for the theft loss deduction--that 
is, loss of other deductions when taking the theft loss deduction. 
Individuals claiming the theft loss deduction might also have other 
personal deductions, such as interest, taxes, or charitable 
contributions. With application of the theft loss carryback or 
carryforward amounts, income is reduced, with the result being that 
income against which to claim the personal deductions can be lost. 
With insufficient income against which to claim the personal 
deductions, they are lost as well, offsetting benefits of the theft 
loss deduction. This is because the personal deductions are not 
themselves subject to carryback or carryforward. Generalizing about 
the effects of leakage is difficult because such calculations are 
taxpayer-specific, but the effect can be substantial, according to one 
tax practitioner. 

* Rate differences. A taxpayer may have initially paid taxes on 
fictitious profits at a relatively high marginal rate, but later 
realize a theft loss deduction at a lower rate. This can mean the 
actual value of fraud-related tax relief received is less than initial 
amounts of tax paid. For example, someone may have paid taxes at a 35 
percent rate, but be subject to a 15 percent rate when claiming 
deductions, because, for example, they lost investment income or 
retired and their income has fallen. As a result, their deduction 
reduces their income by a smaller amount than the amount of taxes that 
they paid in the past when their income tax bracket was 
higher.[Footnote 51] Such a rate difference could be significant if a 
taxpayer uses the IRS safe harbor approach.[Footnote 52] A taxpayer 
may have received reported profits for a number of years, but be 
required, under the safe harbor approach, to deduct all losses in a 
single year. Such a large deduction in one year could reduce the 
taxpayer's marginal tax rate from a high rate to a low rate. For 
example, a taxpayer may have paid taxes at the 35 percent rate, but by 
taking a deduction for all losses in a single year, find their tax 
rate averages out significantly lower. Further, the manner in which 
the Alternative Minimum Tax is calculated could also cause customers 
to realize the benefit of their theft loss deductions at a rate lower 
than when they initially paid taxes.[Footnote 53] 

* Death. Should Madoff customers have died, estate and trust taxation 
issues could prevent full utilization of tax relief arising from the 
Madoff losses. For example, losses can offset estate income, but any 
losses remaining may not transfer with the property in subsequent tax 
considerations, according to one tax practitioner. 

In addition, other factors also stand to affect tax collections, 
either providing additional revenues or increasing revenue loss, 
according to our review. These factors include: 

* Future tax liability. Taxpayers using the safe harbor approach may 
owe additional taxes in the future. By allowing taxpayers to claim 95 
percent (or 75 percent) of their losses, the approach assumes a 5 
percent (or 25 percent) recovery of assets by the Trustee or in other 
recovery proceedings. According to IRS officials, if actual recoveries 
exceed those amounts, taxpayers must declare the excess as income and 
pay taxes on that income. Currently, the Trustee expects recoveries to 
be at least 50 percent, meaning losses taxpayers have claimed under 
the safe harbor could be overstated, triggering the future tax 
liability.[Footnote 54] 

* Other deductions. Investors can generally deduct expenses incurred 
in the production of income, IRS officials noted. That means that over 
the course of the Madoff fraud, customers would have been able to 
reduce taxable income based on any expenses the Madoff firm charged 
them. 

* Madoff insiders. The Trustee told us that payments received by 
Madoff insiders raise tax issues. In some cases, Madoff made loans to 
immediate family, other relatives, and close associates. He often 
forgave such loans later, but the forgiven amounts were not reported 
to IRS as income, the Trustee said. In other cases, some people 
received large cash payments from Madoff that were not reported as 
income. Additionally, according to the Trustee, some insiders 
periodically asked Madoff to produce gains or losses on their 
accounts, presumably in order to offset income from non-Madoff sources 
for tax purposes. 

* Timing. Even if the government surrenders tax revenue as Madoff 
customers realize tax relief, the U.S. government collected and had 
use of tax receipts for multiple years. Meanwhile, as discussed 
earlier, taxpayers may have difficulty making full use of available 
benefits today. Given time value of money, and difficulty in 
capitalizing on benefits, this is advantageous to the government, tax 
practitioners and others told us. 

Either Claims Valuation Method Is Likely to Reduce Tax Collections: 

While the use of NIM or FSM to determine net equity produce different 
outcomes for customers, both would likely reduce tax collections for 
the U.S. Treasury. Under NIM, net winners generally have their claims 
denied and are not eligible for reimbursement from the SIPC fund. 
Because SIPC payments reduce the amount that a taxpayer would claim as 
a loss, these customers would then likely have correspondingly larger 
theft loss deductions. In turn, those higher deductions could cause 
revenue losses for the Treasury that would not have been experienced 
if the Trustee had used FSM, which would have provided higher SIPC 
reimbursements.[Footnote 55] This can be seen in SIPC estimates for 
coverage under the two methods. Under NIM, SIPC estimates an outlay of 
$889 million for payment of SIPC advances to Madoff customers. If FSM 
had been used to value claims, SIPC executives estimate that SIPC 
reimbursements would have increased by an additional $1.2 billion to 
about $2.1 billion. According to IRS officials, an increase in SIPC 
coverage amounts--or any other coverage of losses--will 
correspondingly lower theft loss deductions.[Footnote 56] 

Ultimately, though, the choice of either claims determination method 
creates the potential for loss in tax revenues, because both NIM and 
FSM would create deductions against income by parties affected by the 
Madoff liquidation. While using FSM might have lowered theft loss 
deductions, owing to the greater SIPC reimbursements, it also would 
have caused greater demand on the SIPC fund, according to SIPC 
executives. As a result, SIPC's broker-dealer members would have had 
to pay additional amounts to keep the fund at the level targeted by 
the SIPC board, the executives said. These greater amounts would have 
been either in the form of higher member annual assessments, or 
maintaining SIPC's recently increased assessment for a longer period, 
they said. At this new assessment rate, SIPC members are currently 
paying about $400 million into the fund annually. 

Under SIPA, the assessments are an ordinary business expense, SIPC 
executives told us. Thus, they are deductible as business expenses for 
tax purposes by member broker-dealers, which would have the effect of 
lowering members' taxable income (or increasing losses). As a result, 
rather than the U.S. Treasury facing lower tax collections from Madoff 
customers due to use of NIM, it would experience lower revenues from 
broker-dealers under FSM. Although this tax trade-off effect is 
straightforward to describe, estimating how, if at all, tax revenues 
would change under one method compared to the other is not possible, 
due to taxpayer-specific reasons described earlier. 

The Trustee told us that the effect on Madoff customers' tax 
liabilities was not a consideration in his determination of how to 
calculate investor net equity.[Footnote 57] There is no statutory 
support for any such consideration, he said, and even if there was, 
considering tax implications would have created a substantial burden. 
To consider any tax implications, the Trustee said it would have been 
necessary to examine details of each account, which would have 
significantly increased the cost and amount of time to consider 
claims. Further, the Trustee told us he did not research or compile 
data on tax implications. However, he said he did provide information 
to IRS and the U.S. Department of Labor.[Footnote 58] 

Tax Treatment of Avoidance Actions Could Affect Accuracy of Income Tax 
Filings: 

As described earlier, a significant part of the Trustee's efforts to 
recover assets for distribution to Madoff customers is his avoidance 
action, or clawback, litigation, in which he seeks to recover funds 
paid to certain customers. In general, if taxpayers, due to a 
clawback, return money previously paid to them, they are entitled to 
some reduction in tax liability as a result, IRS officials told us. 
However, application of relevant law, which deals with issues such as 
timing and nature of income, can be very taxpayer-specific, they said. 
IRS officials also told us initially that the agency did not have 
generally applicable guidance on the treatment of those payments. They 
said the agency had been seeking to formulate the right answer for 
dealing with clawbacks, but that it has otherwise provided only 
"factually specific guidance" on a case-by-case basis. In cases in 
which IRS has not issued such specific guidance, taxpayers must rely 
on the Internal Revenue Code, regulations, court cases, and relevant 
revenue rulings by the agency. Tax practitioners to whom we spoke 
noted uncertainties in determining how clawbacks should be treated for 
tax purposes, and that this makes completing income tax returns 
challenging and could contribute to errors. 

IRS officials to whom we spoke said they had not issued guidance on 
this topic because their general approach was to initially focus on 
issuing guidance in areas with more widespread effect, such as the 
safe harbor procedures. A part of IRS's mission is to help taxpayers 
understand and meet their tax responsibilities, and more than a 
thousand Madoff account holders and others face the possibility of 
having to return funds to the Trustee as a result of clawbacks. Future 
financial fraud cases could involve clawbacks in their resolutions as 
well. Without additional guidance to taxpayers for such situations, 
the potential for taxpayer error is increased, which could lead to 
either over-or underpayment of taxes to the U.S. Treasury. A recent 
audit by the Department of the Treasury Inspector General for Tax 
Administration illustrated tax compliance issues surrounding 
investment theft losses. After reviewing what it said was a 
statistically valid sample of 140 returns claiming investment theft 
loss deductions for 2008, the inspector general's audit estimated that 
82 percent of 2,177 tax returns may have erroneously claimed 
deductions totaling more than $697 million and resulting in revenue 
losses of approximately $41 million. Three percent of the tax returns 
the inspector general sampled included taxpayers who claimed more than 
$215,000 in investment theft losses resulting from the Madoff scheme. 
[Footnote 59] This audit did not specifically investigate treatment of 
clawbacks. 

Given the number of taxpayers that could be affected by clawbacks, in 
the Madoff case or others, the lack of guidance could affect the 
accuracy of many tax returns and potentially involve billions of 
dollars in returned funds. We identified this concern to IRS and 
recommended in a draft report that the Commissioner of Internal 
Revenue ensure that the agency provide taxpayer guidance on a timely 
basis on the proper tax treatment of funds returned through avoidance 
actions or settlements arising from cases of investment fraud. 
Subsequently, IRS on September 5, 2012, issued such guidance, in the 
form of "frequently asked questions" on how to treat clawbacks, which 
were posted to the agency's website.[Footnote 60] 

Agency and Third Party Comments and Our Evaluation: 

We provided a draft of this report to SEC, SIPC, IRS, and the Trustee 
for their review and comment, and each provided technical comments, 
which we have incorporated as appropriate. 

We are sending copies of this report to the appropriate congressional 
committees, the SEC Chairman, the SIPC President, the Commissioner of 
Internal Revenue, and the Trustee for the Madoff liquidation. In 
addition, this report is available at no charge on the GAO website at 
[hyperlink, http://www.gao.gov]. If you or your staff have any 
questions regarding this report, please contact me at (202)-512-8678 
or clowersa@gao.gov. Contact points for our Offices of Congressional 
Relations and Public Affairs may be found on the last page of this 
report. GAO staff who made major contributions to this report are 
listed in appendix V. 

Signed by: 

A. Nicole Clowers: 
Director: 
Financial Markets and Community Investment: 

[End of section] 

Appendix I: Objectives, Scope, and Methodology: 

This report discusses (1) the extent to which account activity varied 
by type of customer of the failed Bernard L. Madoff Investment 
Securities, LLC firm; (2) the nature of claims filed, and rejected or 
approved, with the Trustee for reimbursement of losses following the 
firm's failure due to fraud; (3) litigation and settlement activity 
the Trustee has pursued during the subsequent liquidation of the firm, 
in seeking to recover assets for distribution to customers; and (4) 
the effect of the Madoff fraud on customers' federal income tax 
liabilities, including the effect on amounts that would have been due 
if investor losses had been based on customers' reported final 
statement holdings.[Footnote 61] 

To examine the extent to which account activity varied by Madoff 
customer type, we obtained data files from the Madoff Trustee 
containing account holder names and transaction activity, including 
deposits and withdrawals from April 1981 to December 2008. We 
calculated customers' net investment status--whether accounts were 
"net winners," in which customers withdrew more than they had 
deposited, or "net losers," in which they deposited more than they 
withdrew--by summing transactions for each account.[Footnote 62] 
Because customer statements are not available for earlier dates, the 
transactions data provided by the Trustee contains only transactions 
on or after April 1, 1981. For accounts opened before April 1, 1981, 
the data lists a deposit for amounts shown on customer statements as 
of April 1, 1981, and we considered these entries as deposits for our 
analysis as well. These deposits totaled about 0.18% of all deposits. 
Based on account names, we classified account holders as either 
individual or family--which we jointly considered as individual--or 
institutional, such as charities, pension funds, or investment 
vehicles known as "feeder funds" that collected funds from individual 
investors for deposit with the Madoff firm. Using the account names, 
we classified as individual or institutional more than 99.6 percent of 
7,994 accounts we identified with at least one transaction.[Footnote 
63] We could not classify 30 accounts because the account holder names 
did not provide sufficient information to make a determination. These 
30 accounts had a total of $82 million in deposits and $117 million in 
withdrawals, each of which is less than 0.1 percent of all deposits 
and withdrawals. Using the account classifications and net investment 
status, we analyzed subgroups of the overall customer population, to 
examine potential differences in account activity. Specifically, we 
reviewed deposits, withdrawals, and timing of these transactions. In 
reviewing withdrawals, we also focused specifically on withdrawal of 
principal amounts, as distinct from fictitious profits. This is 
because under his authority to sue for return of assets, the Trustee 
can, among other things, presumptively seek to recover principal 
withdrawals during the 90-day period immediately before the Madoff 
firm failed. In our analysis, we relied on our own examination of data 
provided by the Trustee. The Trustee reviewed our results, noting some 
small differences between our review and his work, but generally 
confirming our findings. Upon further review of our work, we 
determined that the variances arose from differing methodologies and 
particular algorithms used to conduct the analysis. For example, we 
excluded from our analysis accounts that had no transactions, and we 
counted multiple allowed claims for one account as a single allowed 
claim. Further, in determining whether accounts were net winners or 
net losers, we summed all applicable deposits and withdrawals to 
determine net positions. In a small number of cases, this produced 
very small negative or positive account balances, which we considered 
to be different than zero. We also interviewed the Trustee and members 
of his law firm on customer-type and transaction-related issues. To 
assess the reliability of the account and transaction data provided by 
the Trustee, we interviewed members of the trustee's counsel law firm 
and a contractor that manages the data, reviewed reports of the 
forensic accountants that assembled the data from records of the 
Madoff firm, and examined the data for invalid or missing data points. 
We concluded the data were sufficiently reliable for our purposes. 

To examine the nature of claims filed and then rejected or approved 
following the Madoff firm's failure, we obtained claims data from the 
Trustee. We tallied claims received and claims dispositions, while 
also examining the total claims population, and claims outcomes. We 
matched claims information with account information as described above 
to examine claim information by customer type and net investment 
status. As described above, we relied on our own examination of data 
provided by the Trustee. We also interviewed the Trustee and members 
of his law firm on claims-related issues. To assess the reliability of 
claims data provided by the Trustee, we interviewed members of the 
trustee's counsel law firm and a contractor that manages the data and 
examined the data for invalid or missing data points. We concluded the 
data were sufficiently reliable for our purposes. 

To examine litigation and settlement activity the Trustee has pursued 
during liquidation of the Madoff firm, we obtained and analyzed court 
documents covering a range of legal activity. These included: lawsuits 
against net winners that are not alleged to have had knowledge of the 
fraud or been in a position to know about it--referred to by the 
Trustee as "good faith" defendants; lawsuits against individuals and 
entities the Trustee argues knew or should have known about the fraud--
referred to by the Trustee as "bad faith" defendants; lawsuits against 
investment vehicles that collected funds from investors and invested 
them with the Madoff firm; and agreements the Trustee has reached to 
settle a number of the actions he has filed as part of his asset 
recovery efforts. We selected a random sample of 50 good faith actions 
for examination, from among more than 1,000 cases filed; we reviewed 
29 publicly available bad faith complaints from among 30 such actions 
filed, analyzing the most frequently cited bases for the Trustee's 
allegations of bad faith; we examined the largest complaints among 27 
actions filed against feeder funds; and we reviewed the largest 
settlements the Trustee had reached during his litigation efforts. In 
addition, we examined Trustee records associated with the "hardship 
program," in which the Trustee expedited claims processing or declined 
to pursue litigation for customers that could demonstrate financial 
distress. We interviewed the Trustee and members of the trustee's 
counsel law firm on litigation and settlement issues, and we conducted 
legal research on the Trustee's legal basis for pursuing asset 
recovery actions. We also interviewed Securities and Exchange 
Commission (SEC) officials and executives of the Securities Investor 
Protection Corp. (SIPC) for their views on conduct of Madoff-related 
litigation. 

To examine the effect of the Madoff fraud on customers' federal income 
tax liabilities, including potential differences based on how investor 
losses were calculated, we examined relevant portions of the Internal 
Revenue Code and Internal Revenue Service (IRS) revenue rulings and 
revenue procedures.[Footnote 64] In addition, we reviewed a September 
2011 audit by the Department of the Treasury Inspector General for Tax 
Administration. We also interviewed SEC and IRS officials, SIPC 
executives, and four individuals with income tax expertise, including 
a law professor and three tax advisors who we selected based on their 
tax experience or publications related to tax issues. 

We conducted this performance audit from March to September 2012 in 
accordance with generally accepted government auditing standards. 
Those standards require that we plan and perform our audit to obtain 
sufficient, appropriate evidence to provide a reasonable basis for our 
findings and conclusions based on our audit objectives. We believe 
that the evidence obtained provides a reasonable basis for our 
findings and conclusions based on our audit objectives. 

[End of section] 

Appendix II: Trustee Avoidance Actions and Other Claims: 

Avoidance actions, often called "clawbacks," enable a bankruptcy 
trustee to set aside certain transfers of property made by the debtor 
within specified periods preceding a bankruptcy filing, in order to 
recover the property for the benefit of the estate and its creditors. 
These actions are authorized by the federal Bankruptcy Code (title 11, 
United States Code) as well as state laws, which section 544(b) of the 
Bankruptcy Code makes available to bankruptcy trustees. Avoidance 
powers apply fully to trustees conducting liquidations under the 
Securities Investor Protection Act (SIPA).[Footnote 65] 

As discussed below, two types of avoidance actions authorized by the 
Bankruptcy Code are preferential transfers (or preferences) and 
fraudulent transfers. Fraudulent transfers are further subdivided into 
two types: actual fraud and constructive fraud. Individual lawsuits 
brought by trustees can (and in the Madoff liquidation, often do) 
include preference avoidance counts, avoidance counts alleging both 
constructive and actual fraud under the Bankruptcy Code, and avoidance 
counts arising under state law. Indeed, the bankruptcy court has 
affirmed the Madoff Trustee's right to bring a wide range of avoidance 
claims under both the Bankruptcy Code and New York law.[Footnote 66] 

A trustee's avoidance powers are especially strong when the 
liquidation involves a Ponzi scheme. This is because courts have 
developed a series of specific interpretive rules, including the 
"Ponzi scheme presumption," which work to the advantage of the trustee 
and to the disadvantage of the recipients of money paid out by Ponzi 
schemers--particularly when it comes to recovering payments that 
represent fictitious profits.[Footnote 67] 

Preferential Transfers: 

Avoidance actions based on preferential transfers, which are governed 
by section 547 of the Bankruptcy Code, enable the trustee to avoid and 
recover payments the debtor made to creditors within 90 days preceding 
the bankruptcy filing, or up to 1 year prior to filing in the case of 
transfers to "insiders."[Footnote 68] Under section 547(b) of the 
Code, a trustee may avoid any transfer of a property interest of the 
debtor to or for the benefit of a creditor made within the preference 
period "for or on account of an antecedent debt" while the debtor was 
insolvent, if the transfer enabled the creditor to receive more than 
the creditor would have received without it under the bankruptcy 
distribution. The U.S. Supreme Court has observed that preference 
avoidance is a "mechanism [that] prevents the debtor from favoring one 
creditor over others by transferring property shortly before filing 
for bankruptcy."[Footnote 69] 

There are several statutory exceptions to avoidance under section 
547(c). For example, preference avoidance does not extend to payments 
of a debt incurred by the debtor "in the ordinary course of business." 
However, the courts consistently hold that the concept of "ordinary 
course of business" has no application in a Ponzi scheme setting and, 
therefore, cannot provide a defense to preference avoidance. 
Furthermore, preference avoidance does not take account of the conduct 
of the recipient. Thus, it applies to all "transferees," or recipients 
of transfers, including innocent victims of Ponzi schemes. 

For the above reasons, a trustee usually has no difficulty avoiding as 
preferential transfers Ponzi scheme payments made during the 
preference period. At the same time, the prevailing view is that 
preferential transfer actions can only reach payments that represent 
customer investments of principal in a Ponzi scheme--not fictitious 
profits. In the leading precedent on this subject, the court reasoned 
that since a Ponzi investor does not have a valid claim to fictitious 
profits, payouts based on them are not made on account of an 
"antecedent debt" as required for preference avoidance.[Footnote 70] 
This limitation has little if any practical effect, however, since 
fictitious profits paid during the preference period are recoverable 
through the fraudulent transfer avoidance actions described below. 

Fraudulent Transfers: 

Avoidance actions based on fraudulent transfers are governed by 
section 548 of the Bankruptcy Code. Section 548(a)(1) provides that a 
trustee may avoid any transfer of an interest of the debtor in 
property made within 2 years before the filing date--: 

if the debtor made the transfer "with actual intent to hinder, delay, 
or defraud any entity" to which the debtor was or became indebted; or: 

if the debtor received "less than equivalent value in exchange for" 
the transfer and was insolvent at the time of the transfer, became 
insolvent as a result of the transfer, was engaged in business for 
which the debtor's remaining property provided unreasonably small 
capital, or met one of several other conditions specified in section 
548(a)(1)(B). 

Section 548 thus provides for two types of avoidance actions. An 
action based on section 548(a)(1)(A) is one for actual fraud. As the 
language indicates, it applies to transfers made by a debtor with 
actual intent to defraud. By contrast, an action based on section 
548(a)(1)(B) is one for constructive fraud and does not require 
fraudulent intent on the part of the debtor. Instead, the key 
consideration is whether the transfer was made for less than 
equivalent value while the debtor was insolvent. 

As a practical matter, the distinction between actual and constructive 
fraud under section 548 has little significance in the unique context 
of a Ponzi scheme, where the trustee is seeking to recover fictitious 
profits. By virtue of the Ponzi scheme presumption, Ponzi scheme 
payouts are generally considered to be actually fraudulent. Case law 
has held that transfers beyond the principal investment lack value, 
making those transfers recoverable under the constructive fraud 
provisions. Thus, transfers of fictitious profits are subject to 
avoidance as both actual and constructive fraudulent transfers. 

The concept of fraud under section 548(a) focuses on the debtor-
transferor rather than the transferees. Thus, all recipients of Ponzi 
scheme payouts, innocent or otherwise, are potentially subject to 
avoidance actions based on both actual and constructive fraud. 
However, the recipients' status is important in determining the nature 
and extent of their liability. Section 548(c) of the Bankruptcy Code 
provides that the transferee in a fraudulent transfer avoidance action 
may retain any interest transferred upon showing that the interest was 
taken "for value and in good faith." Principal invested in a Ponzi 
scheme is considered value with respect to a good faith transferee in 
this context, while fictitious profits are not. In general, therefore, 
the liability of good faith recipients in a Ponzi scheme-related 
fraudulent transfer action will be limited to fictitious profits paid 
to "net winners"--that is, investors who withdrew more than they 
invested in the scheme. Recipients who cannot establish their good 
faith are liable for the return of their principal investment as well 
as any fictitious profits paid out during the 2-year avoidance period. 

Good faith in an avoidance action is an affirmative defense, so the 
transferee has the burden of proof.[Footnote 71] In this regard: 

" ... If the investor knew or should have known that the debtor's 
investment scheme was too good to be true, then the investor fails to 
carry his burden of proving that he accepted sums from the debtor in 
good faith, and the trustee is entitled to recover all amounts the 
investor received from the debtor."[Footnote 72] 

Judicial decisions vary on how to apply the "knew or should have 
known" standard. One line of precedent follows an "objective" or 
"reasonable person" test, whereby an investor has a duty to inquire 
into the legitimacy of transactions in the face of "red flags" that 
would make a reasonable investor suspicious.[Footnote 73] Under this 
approach, an investor who fails to inquire into suspicious 
circumstances, or even one who inquires but conducts an inadequate 
inquiry, cannot establish a good faith defense. By contrast, other 
courts have applied a so-called "subjective" test, whereby investors 
are found to have acted in good faith so long as they lacked actual 
knowledge of the fraud or did not "turn a blind eye" in the face of 
obvious signs of fraud.[Footnote 74] One recent U.S. District Court 
opinion in a Madoff-related case endorsed the subjective test of good 
faith.[Footnote 75] 

Avoidance Actions under New York Law: 

As noted, a trustee, including a SIPA liquidator, can bring avoidance 
actions under state law as well as the federal Bankruptcy Code. Since 
the Madoff firm was formed as a New York limited liability company 
with its principal place of business there, the Madoff Trustee's 
avoidance actions include claims under New York's fraudulent 
conveyance law, codified in sections 270 through 281 of the New York 
Debtor and Creditor Law (DCL) (McKinney 2012). 

New York law does not provide for preference avoidance actions. 
However, it does contain fraudulent conveyance provisions that are 
similar to section 548 of the Bankruptcy Code. Sections 273 through 
275 of the DCL are its constructive fraud provisions, authorizing the 
avoidance of transfers made without "fair consideration."[Footnote 76] 
Section 276 of the DCL provides for the avoidance of transfers based 
on actual fraud, using language similar to section 548(a)(1)(A) of the 
Bankruptcy Code: 

"Every conveyance made and every obligation incurred with actual 
intent, as distinguished from intent presumed in law, to hinder, 
delay, or defraud either present or future creditors, is fraudulent as 
to both present and future creditors."[Footnote 77] 

Additionally, DCL section 276-a provides for the award of attorney 
fees in a successful avoidance action based on actual fraud where 
fraud on the part of the transferee is demonstrated. 

The main difference between avoidance actions under the federal 
Bankruptcy Code and the DCL relates to the reach-back period. In 
contrast to the 2-year Bankruptcy Code reach-back, DCL constructive 
fraud avoidance actions are subject to the general 6-year statute of 
limitations in section 213(1) of the New York Civil Practice Law and 
Rules (CPLR) (McKinney 2012).[Footnote 78] Thus, these actions can 
reach transfers made up to 6 years before the filing date. 

Under CPLR sections 203(g) and 213(8), the reach-back period for 
actions based on actual fraud is 6 years before the filing date or 2 
years from the time the fraud was discovered or could have been 
discovered with reasonable diligence, whichever date is earlier. The 
Madoff Trustee has used this authority to seek avoidance and recovery 
of fraudulent transfers occurring more than 6 years before the filing 
date where the recipients received the transfers in bad faith. 

Other Considerations: 

Claims by recipients of avoidable payments. Section 502(d) of the 
Bankruptcy Code generally requires disallowance of any claim against 
the bankruptcy estate by the recipient of an avoidable transfer, 
unless the recipient has paid over the avoidable amount. The Madoff 
Trustee has invoked this authority in seeking to temporarily disallow 
SIPA claims and other claims brought by some defendants in avoidance 
actions. 

Subsequent transferees. Under section 550 of the Bankruptcy Code, the 
trustee can seek recovery of the proceeds of an avoided transaction 
from not just the initial transferee but also from "subsequent 
transferees"--third parties who obtained funds from those receiving 
funds directly from the bankrupt entity. Thus, for example, the Madoff 
trustee can maintain avoidance actions against Madoff investors who 
received payouts through feeder funds or other intermediaries. 
However, there are limits to this authority. For example, a trustee 
cannot recover from a subsequent transferee who can show receipt of 
the proceeds for value, in good faith, and without knowledge that the 
transfer was avoidable. Also, an action under section 550 generally 
must be commenced within 1 year after the avoidance of the transfer 
for which recovery is sought against. 

Like the Bankruptcy Code provisions, sections 278 and 279 of the DCL 
authorize avoidance and recovery against subsequent transferees, 
except purchasers for fair consideration without knowledge of the 
fraud. 

Stockbroker safe harbor. One important area of uncertainty concerning 
securities-related Ponzi scheme avoidance actions is the applicability 
of the so-called "stockbroker safe harbor" or "stockbroker defense" 
provided in section 546(e) of the Bankruptcy Code. This provision 
exempts certain securities transactions from avoidance actions except 
those alleging actual fraud.[Footnote 79] Thus far, Madoff-related 
judicial decisions have reached different conclusions regarding the 
applicability of section 546(e). The bankruptcy court held that 
section 546(e) did not apply to the Madoff Ponzi scheme,[Footnote 80] 
while a district court held that it did.[Footnote 81] 

Other Claims: 

A number of lawsuits the Madoff trustee has filed against bad faith 
transferees contain common law claims in addition to avoidance counts 
under the Bankruptcy Code and the New York DCL. These claims include, 
for example, counts for conversion, unjust enrichment, and money had 
and received. 

A trustee has the authority and standing to sue to enforce claims that 
the debtor had prior to bankruptcy that represent property of the 
estate, including common law claims.[Footnote 82] However, two recent 
U.S. District Court decisions dealing with the Madoff liquidation 
emphasized limits on the Madoff Trustee's ability to pursue common law 
claims.[Footnote 83] In both cases, the courts concluded that the 
Trustee could not pursue certain common law claims that, in the 
courts' view, more appropriately belonged to creditors rather than the 
debtor and the estate. Furthermore, with respect to debtor claims, the 
courts noted that Madoff and his firm were wrongdoers, and that the 
Trustee stands in their shoes for purposes of pursuing common law 
claims. Therefore, the courts concluded, such claims would be subject 
to the legal principle that bars resolution of disputes between one 
wrongdoer and another.[Footnote 84] 

[End of section] 

Appendix III: Largest Madoff Accounts by Transaction Volume, Total 
Withdrawals, and Net Winnings: 

We obtained and analyzed transaction data for customers of the failed 
Bernard L. Madoff Investment Securities, LLC firm. Table 10 shows 
account information and status for the 10 largest Madoff accounts by 
transaction volume. 

Table 10: Ten Largest Madoff Accounts by Transaction Volume: 

Dollars in millions: 

Account name[A]: NORMAN F LEVY C/O KONIGSBERG WOLF & CO; 
Transaction volume[B]: $222,322; 
Net investment position[C]: -$61; 
Final statement account balance[D]: $0; 
Status: As part of settlement with Trustee that included 39 Levy-
related accounts, Levy heirs will repay $220 million in withdrawals of 
fictitious profits. Settlement appealed by other Madoff customers. 

Account name[A]: CITCO GLOBAL CUSTODY N V FBO FAIRFIELD SENTRY LTD; 
Transaction volume[B]: $8,487; 
Net investment position[C]: $509; 
Final statement account balance[D]: $3,226; 
Status: Settlement reached covering four accounts, in which the funds' 
claims were reduced by almost $1 billion, while providing $70 million 
to customer property fund. The funds retained allowed claims of $230 
million. 

Account name[A]: CITCO GLOBAL CUSTODY N V FBO FAIRFIELD SENTRY LTD; 
Transaction volume[B]: $8,453; 
Net investment position[C]: $554; 
Final statement account balance[D]: $3,248; 
Status: Settlement reached covering four accounts, in which the funds' 
claims were reduced by almost $1 billion, while providing $70 million 
to customer property fund. The funds retained allowed claims of $230 
million. 

Account name[A]: DECISIONS INCORPORATED; 
Transaction volume[B]: $6,397; 
Net investment position[C]: -$6,162; 
Final statement account balance[D]: -$5,388; 
Status: Account owned by Jeffry Picower. Under settlement with 
Trustee, Picower estate agreed to repay all $7.2 billion in net 
withdrawals from 34 accounts, $5 billion to Trustee and $2.2 billion 
to U.S. Department of Justice. 

Account name[A]: FAIRFIELD SENTRY LIMITED C/O FAIRFIELD GREENWICH 
GROUP; 
Transaction volume[B]: $4,771; 
Net investment position[C]: $14; 
Final statement account balance[D]: -$95; 
Status: See CITCO GLOBAL CUSTODY N V FBO FAIRFIELD SENTRY LTD, above. 

Account name[A]: FAIRFIELD SENTRY LIMITED C/O FAIRFIELD GREENWICH 
GROUP; 
Transaction volume[B]: $4,745; 
Net investment position[C]: $0; 
Final statement account balance[D]: -$95; 
Status: See CITCO GLOBAL CUSTODY N V FBO FAIRFIELD SENTRY LTD, above. 

Account name[A]: HARLEY INTERNATIONAL FUND LTD C/O FORTIS PRIME FUND 
SOLUTION; 
Transaction volume[B]: $3,500; 
Net investment position[C]: $1,279; 
Final statement account balance[D]: $2,578; 
Status: Did not file claim to recover losses. Not involved in Madoff-
related litigation or settlements. 

Account name[A]: RYE SELECT BROAD MKT FUND LP C/O TREMONT PARTNERS; 
Transaction volume[B]: $2,442; 
Net investment position[C]: $1,648; 
Final statement account balance[D]: $2,346; 
Status: One of accounts of Tremont Group Holdings, Inc. Trustee sought 
to recover $2.1 billion in withdrawals by Tremont Group, alleging it 
knew or should have known of warning signs about Madoff firm. Under 
settlement with Trustee, Tremont Group will pay $1 billion and have 
allowed claims of $3 billion. 

Account name[A]: UBS (LUXEMBOURG) SA FBO LUXALPHA SICAV; 
Transaction volume[B]: $2,346; 
Net investment position[C]: $762; 
Final statement account balance[D]: $1,299; 
Status: Trustee filed complaint seeking recovery of about $2 billion 
in withdrawals. Complaint partially dismissed by U.S District Court 
judge. 

Account name[A]: HSBC SECURITIES SERVICES (LUXEMBOURG) SA SPEC CUST 
ACCT; 
Transaction volume[B]: $2,313; 
Net investment position[C]: $1,178; 
Final statement account balance[D]: $1,894; 
Status: Trustee filed complaint seeking $2 billion in withdrawals and 
$6.6 billion in damages. Complaint partially dismissed by U.S. 
District Court judge, and Trustee has appealed. 

Source: GAO analysis of Madoff Trustee data and court filings. 

[A] The name listed on the account. Some entities had multiple 
accounts under the same or similar names. Each row in this table 
represents a single account. 

[B] The sum of all deposits and withdrawals. 

[C] The net investment position of the account at the time of the 
Madoff firm failure. Positive net investment positions represent 
deposits in excess of withdrawals; negative net investment positions 
indicate withdrawals in excess of deposits. 

[D] The account balance listed on the final statement for the account. 

[End of table] 

Table 11 shows account information and status for the 10 largest 
Madoff accounts by total withdrawals. 

Table 11: Ten Largest Madoff Accounts by Total Withdrawals: 

Dollars in millions: 

Account name[A]: NORMAN F LEVY C/O KONIGSBERG WOLF & CO; 
Total withdrawals[B]: $111,126; 
Total deposits[C]: $111,192; 
Net investment position[D]: -$61; 
Final statement account balance[E]: $0; 
Status: See NORMAN F LEVY C/O KONIGSBERG WOLF & CO in table 10 above. 

Account name[A]: DECISIONS INCORPORATED; 
Total withdrawals[B]: $6,279; 
Total deposits[C]: $117; 
Net investment position[D]: -$6,162; 
Final statement account balance[E]: -$5,388; 
Status: See DECISIONS INCORPORATED in table 10 above. 

Account name[A]: CITCO GLOBAL CUSTODY N V FBO FAIRFIELD SENTRY LTD; 
Total withdrawals[B]: $3,989; 
Total deposits[C]: $4,497; 
Net investment position[D]: $509; 
Final statement account balance[E]: $3,226; 
Status: See CITCO GLOBAL CUSTODY N V FBO FAIRFIELD SENTRY LTD in table 
10 above. 

Account name[A]: CITCO GLOBAL CUSTODY N V FBO FAIRFIELD SENTRY LTD; 
Total withdrawals[B]: $3,949; 
Total deposits[C]: $4,504; 
Net investment position[D]: $554; 
Final statement account balance[E]: $3,248; 
Status: See CITCO GLOBAL CUSTODY N V FBO FAIRFIELD SENTRY LTD in table 
10 above. 

Account name[A]: FAIRFIELD SENTRY LIMITED C/O FAIRFIELD GREENWICH 
GROUP; 
Total withdrawals[B]: $2,378; 
Total deposits[C]: $2,392; 
Net investment position[D]: $14; 
Final statement account balance[E]: -$95; 
Status: See CITCO GLOBAL CUSTODY N V FBO FAIRFIELD SENTRY LTD in table 
10 above. 

Account name[A]: FAIRFIELD SENTRY LIMITED C/O FAIRFIELD GREENWICH 
GROUP; 
Total withdrawals[B]: $2,373; 
Total deposits[C]: $2,373; 
Net investment position[D]: -$0; 
Final statement account balance[E]: -$95; 
Status: See CITCO GLOBAL CUSTODY N V FBO FAIRFIELD SENTRY LTD in table 
10 above. 

Account name[A]: HARLEY INTERNATIONAL FUND LTD C/O FORTIS PRIME FUND 
SOLUTION; 
Total withdrawals[B]: $1,111; 
Total deposits[C]: $2,389; 
Net investment position[D]: $1,279; 
Final statement account balance[E]: $2,578; 
Status: See HARLEY INTERNATIONAL FUND LTD C/O FORTIS PRIME FUND 
SOLUTION in table 10 above. 

Account name[A]: RYE SELECT BROAD MARKET PRIME FUND, LP; 
Total withdrawals[B]: $1,010; 
Total deposits[C]: $799; 
Net investment position[D]: -$211; 
Final statement account balance[E]: $596; 
Status: See RYE SELECT BROAD MKT FUND LP C/O TREMONT PARTNERS in table 
10 above. 

Account name[A]: UBS (LUXEMBOURG) SA FBO LUXALPHA SICAV; 
Total withdrawals[B]: $752; 
Total deposits[C]: $1,554; 
Net investment position[D]: $802; 
Final statement account balance[E]: $1,299; 
Status: See UBS (LUXEMBOURG) SA FBO LUXALPHA SICAV in table 10 above. 

Account name[A]: ASCOT PARTNERS LP; 
Total withdrawals[B]: $748; 
Total deposits[C]: $974; 
Net investment position[D]: $226; 
Final statement account balance[E]: $1,862; 
Status: One of the feeder funds created by J. Ezra Merkin. As part of 
settlement with New York Attorney General, Merkin agreed to pay $405 
million to certain customers. Trustee filed separate complaint, 
seeking $500 million in withdrawals, and is seeking injunction against 
the settlement between Merkin and clients, arguing it violates 
automatic stay provisions of Bankruptcy Code and stay orders of 
district court. 

Source: GAO analysis of Madoff Trustee data and court filings. 

[A] The name listed on the account. Some entities had multiple 
accounts under the same or similar names. Each row in this table 
represents a single account. 

[B] The sum of all withdrawals made from the account. 

[C] The sum of all deposits made into the account. 

[D] The net investment position of the account at the time of the 
Madoff firm failure. Positive net investment positions represent 
deposits in excess of withdrawals; negative net investment positions 
indicate withdrawals in excess of deposits. 

[E] The account balance listed on the final statement for the account. 

[End of table] 

Table 12 shows account information and status for the 10 largest 
Madoff accounts by "net winnings," that is, the amount by which an 
account holder withdrew more than they invested in the Madoff firm. 

Table 12: Ten Largest Madoff Accounts by Net Winnings: 

Dollars in millions: 

Account name[A]: DECISIONS INCORPORATED; 
Transaction volume[B]: $6,397; 
Net investment position[C]: -$6,162; 
Final statement account balance[D]: -$5,388; 
Status: See DECISIONS INCORPORATED in table 10 above. 

Account name[A]: JEANNE LEVY-CHURCH; 
Transaction volume[B]: $742; 
Net investment position[C]: -$573; 
Final statement account balance[D]: -$1,699; 
Status: Account associated with Norman F. Levy. See NORMAN F LEVY C/O 
KONIGSBERG WOLF & CO in table 10 above. 

Account name[A]: THE LAMBETH CO; 
Transaction volume[B]: $649; 
Net investment position[C]: -$378; 
Final statement account balance[D]: $402; 
Status: Among accounts owned or associated with Stanley Chais. Trustee 
seeking repayment of $1.2 billion in withdrawals. 

Account name[A]: JA PRIMARY LTD PARTNERSHIP; 
Transaction volume[B]: $287; 
Net investment position[C]: -$250; 
Final statement account balance[D]: $0; 
Status: Most recent transaction was 1996, and account not involved in 
litigation. 

Account name[A]: THE PICOWER FOUNDATION; 
Transaction volume[B]: $367; 
Net investment position[C]: -$242; 
Final statement account balance[D]: $598; 
Status: See DECISIONS INCORPORATED in table 10 above. 

Account name[A]: THE BRIGHTON COMPANY; 
Transaction volume[B]: $367; 
Net investment position[C]: -$233; 
Final statement account balance[D]: $384; 
Status: See THE LAMBETH CO above. 

Account name[A]: AVELLINO & BIENES #5 C/O FRANK AVELLINO; 
Transaction volume[B]: $299; 
Net investment position[C]: -$213; 
Final statement account balance[D]: $0; 
Status: Account used by feeder fund created by Frank Avellino and 
Michael Bienes; 
the first Madoff feeder fund, according to Trustee. Trustee is seeking 
repayment of $905 million in withdrawals from accounts associated with 
Avellino and Bienes. 

Account name[A]: RYE SELECT BROAD MARKET PRIME FUND, LP; 
Transaction volume[B]: $1,809; 
Net investment position[C]: -$211; 
Final statement account balance[D]: $596; 
Status: See RYE SELECT BROAD MKT FUND LP C/O TREMONT PARTNERS in table 
10 above. 

Account name[A]: CARL SHAPIRO TRUST U/D/T 4/9/03; 
Transaction volume[B]: $547; 
Net investment position[C]: -$201; 
Final statement account balance[D]: $56; 
Status: Among accounts owned by Carl Shapiro or family. Family agreed 
to $625 million settlement, which is lower than $1 billion Trustee 
sought to recover withdrawals of fictitious profits. Shapiro family 
contended Trustee did not properly account for $500 million in 
deposits. 

Account name[A]: FRANCIS N LEVY C/O KONIGSBERG WOLF & CO PC; 
Transaction volume[B]: $378; 
Net investment position[C]: -$193; 
Final statement account balance[D]: -$1,149; 
Status: Account associated with Norman F. Levy. See NORMAN F LEVY C/O 
KONIGSBERG WOLF & CO in table 10 above. 

Source: GAO analysis of Madoff Trustee data and court filings. 

[A] The name listed on the account. Some entities had multiple 
accounts under the same or similar names. Each row in this table 
represents a single account. 

[B] The sum of all deposits and withdrawals. 

[C] The net investment position of the account at the time of the 
Madoff firm failure. Positive net investment positions represent 
deposits in excess of withdrawals; negative net investment positions 
indicate withdrawals in excess of deposits. 

[D] The account balance listed on the final statement for the account. 

[End of table] 

[End of section] 

Appendix IV: Details of Settlements by the Madoff Trustee: 

In hundreds of cases, the Madoff Trustee has reached settlements with 
former customers and others, either before or after filing clawback 
actions, and these agreements have produced recovery of a significant 
amount of assets. This appendix provides details of these settlements, 
including amounts sought and obtained, the extent to which settlements 
have been paid, strategies behind settlements, and key provisions. 

As of April 2012, the Trustee had entered into 441 settlement 
agreements in which the opposing parties agreed to return about $8.4 
billion--an amount equal to about 49 percent of the approximately 
$17.3 billion in principal investments lost by customers who filed 
claims. Based on a bankruptcy court review threshold, we examined all 
but one settlement agreement worth at least $20 million.[Footnote 85] 
Table 13 provides a summary of the 13 settlements we identified. These 
settlements account for more than 97 percent of all settlement 
amounts. They include settlements from family estates, such as 
Picower, Shapiro, and Levy, as well as feeder funds, in the case of 
UBP, Optimal, Fairfield, Mount Capital, and Trotanoy Investment. 
[Footnote 86] The group also includes settlements with charitable 
foundations and the federal government. As the table shows, the 
amounts collected by individual case vary widely. 

Table 13: Madoff Settlement Agreements of at Least $20 Million: 

Dollars in millions: 

Case: Picower[A]; 
Case type: Investor; 
Amount initially sought: $7,200; 
Settlement amount: $5,000; 
Settlement amount as percentage of all settlements: 59.6%; 
Payments received as of July 2012: $5,000; 
Percentage of settlement received: 100%. 

Case: Tremont; 
Case type: Feeder fund; 
Amount initially sought: $2,100; 
Settlement amount: $1,025; 
Settlement amount as percentage of all settlements: 12.2%; 
Payments received as of July 2012: [Empty]; 
Percentage of settlement received: 0%. 

Case: Shapiro; 
Case type: Investor; 
Amount initially sought: $1,045; 
Settlement amount: $550; 
Settlement amount as percentage of all settlements: 6.6%; 
Payments received as of July 2012: $550; 
Percentage of settlement received: 100%. 

Case: UBP; 
Case type: Feeder fund; 
Amount initially sought: $1,000; 
Settlement amount: $470; 
Settlement amount as percentage of all settlements: 5.6%; 
Payments received as of July 2012: $470; 
Percentage of settlement received: 100%. 

Case: IRS[B]; 
Case type: Government; 
Amount initially sought: $326; 
Settlement amount: $326; 
Settlement amount as percentage of all settlements: 3.9%; 
Payments received as of July 2012: $326; 
Percentage of settlement received: 100%. 

Case: Optimal; 
Case type: Feeder fund; 
Amount initially sought: N/A[C]; 
Settlement amount: $233.8; 
Settlement amount as percentage of all settlements: 2.8%; 
Payments received as of July 2012: $233.8; 
Percentage of settlement received: 100%. 

Case: Levy; 
Case type: Investor; 
Amount initially sought: $221; 
Settlement amount: $220; 
Settlement amount as percentage of all settlements: 2.6%; 
Payments received as of July 2012: $220; 
Percentage of settlement received: 100%. 

Case: Katz-Wilpon; 
Case type: Investor; 
Amount initially sought: $1,018; 
Settlement amount: $162; 
Settlement amount as percentage of all settlements: 1.9%; 
Payments received as of July 2012: [Empty]; 
Percentage of settlement received: 0%. 

Case: Fairfield; 
Case type: Feeder fund; 
Amount initially sought: $3,054; 
Settlement amount: $70; 
Settlement amount as percentage of all settlements: 0.8%; 
Payments received as of July 2012: $18.3; 
Percentage of settlement received: 26.2%. 

Case: Hadassah[D]; 
Case type: Charitable foundation; 
Amount initially sought: $77; 
Settlement amount: $45; 
Settlement amount as percentage of all settlements: 0.5%; 
Payments received as of July 2012: $45; 
Percentage of settlement received: 100%. 

Case: Mount Capital; 
Case type: Feeder fund; 
Amount initially sought: $46; 
Settlement amount: $43.5; 
Settlement amount as percentage of all settlements: 0.5%; 
Payments received as of July 2012: $43.5; 
Percentage of settlement received: 100%. 

Case: Trotanoy Investment; 
Case type: Feeder fund; 
Amount initially sought: $182.4 plus damages at trial; 
Settlement amount: $29; 
Settlement amount as percentage of all settlements: 0.3%; 
Payments received as of July 2012: [Empty]; 
Percentage of settlement received: 0%. 

Case: Hynes; 
Case type: N/A; 
Amount initially sought: N/A; 
Settlement amount: $20; 
Settlement amount as percentage of all settlements: 0.2%; 
Payments received as of July 2012: $20; 
Percentage of settlement received: 100%. 

Case: Total; 
Settlement amount: $8,194; 
Settlement amount as percentage of all settlements: 97.6%; 
Payments received as of July 2012: $6,927; 
Percentage of settlement received: 84.5%. 

Source: Madoff Trustee, GAO analysis. 

[A] The Picower settlement totals $7.2 billion and has two parts: $5 
billion for the Trustee and $2.2 billion for the U.S. Department of 
Justice civil forfeiture program. The Trustee told us that in the case 
of such dual settlements, the Trustee first obtained his desired 
settlement, and then any other amounts were added on. Thus, amounts 
recovered by other parties did not reduce amounts obtained by the 
Trustee, he said. 

[B] To recover purported tax withholdings Madoff made for foreign 
investors. The Trustee said Madoff collected and paid more than $330 
million to IRS, but some foreign account holders settled separately 
with the IRS before the Trustee reached his agreement. The Trustee's 
settlement with IRS reflects the balance. 

[C] Clawback action not filed; no amount formally sought. 

[D] As discussed earlier, the Trustee alleged Hadassah received at 
least $77 million in avoidable transfers, but settled at the lower 
amount in consideration of the organization's charitable activities. 

[End of table] 

Among these settlements, the Optimal case has played an influential 
role, the Trustee told us. Optimal was the sponsor of two feeder 
funds, and it was an early case with potentially risky legal issues 
present, such as jurisdiction and foreign entities--it was not clear 
the U.S. Bankruptcy Code or SIPA apply outside the United States, the 
Trustee told us. The Trustee secured an agreement requiring Optimal to 
return 85 percent of amounts received during the 90-day preference 
period. Included was a term providing that if the Trustee settled with 
another party on more advantageous terms, Optimal would also receive 
those terms. 

The 85 percent figure effectively became the benchmark for settlements 
of similar claims, the Trustee told us. In turn, the Trustee also set 
a complementary benchmark for good faith claims at 95 percent, in 
consideration of the comparative ease in making such cases compared to 
bad faith actions. The Trustee told us that in the Optimal settlement, 
he considered it important to secure the agreement in order to build 
the customer fund. However, looking back, he said the "most favored 
nation" provision has been an issue. Optimal recently attempted to 
invoke the clause following the Trustee's recent settlement with 
Hadassah. Due to the complexity of these cases, it is difficult to 
discern what constitutes a settlement to which the clause would apply, 
the Trustee said. 

Based on our examination of the largest settlement agreements, table 
14 provides a summary of key settlement provisions. 

Table 14: Key Provisions among Largest Madoff Settlements: 

Settlement: Optimal; 
Key provisions: "Most favored nation" clause where Optimal agreed to 
pay 85 percent of amounts received during 90-day preference period, 
but if another party gets more favorable agreement, Optimal gets those 
terms. 

Settlement: Fairfield; 
Key provisions: $1.2 billion in customer claims reduced to $230 
million. Trustee and liquidator of feeder funds involved agreed to 
share in recoveries from feeder funds' officers and management 
companies involved. Trustee said partnership aligns interests of both 
parties and enhances prospects of recovery. 

Settlement: Tremont; 
Key provisions: Defendants do not share in settlement proceeds, as 
they forego profits through fees, earnings, management expenses. 
Trustee says refusal to "settle cheap"--initial settlement offer was 
$176 million, versus $1.025 billion eventually obtained--established 
credibility. 

Settlement: IRS; 
Key provisions: IRS agreed to return to Trustee funds Madoff had 
collected from foreign account holders as income tax withholdings. 
Foreign account holders cannot bring actions against IRS, United 
States, or the Trustee in the future. 

Settlement: Katz-Wilpon; 
Key provisions: Claims defendants had following collapse of Madoff 
firm become assets to fund settlement. Katz and Wilpon personally 
guarantee up to $29 million if other defendants do not pay. 

Settlement: Hadassah; 
Key provisions: Settled $77 million case for $45 million, to avoid 
causing failure of Israeli hospital funded by defendant. 

Settlement: Picower, Levy; 
Key provisions: Recovered all, or nearly all, of eligible fictitious 
profits; 
parties' claims for losses also withdrawn. 

Source: Madoff Trustee and GAO review of settlement agreements filed 
by the Trustee. 

[End of table] 

The Trustee's settlement agreements also go beyond simply obtaining 
cash agreements, we found, as allowance of customer claims and SIPC 
advances have also been key components of settlements of feeder fund 
cases. For example, the Tremont settlement, in addition to its $1.025 
billion settlement amount, also included allowing more than $3 billion 
in customer claims and granting SIPC advances for eligible accounts. 
With the exception of the Katz-Wilpon case, all seven settlements that 
included allowed customer claims as part of the agreement were feeder 
fund cases. Because these feeder funds were net losers, the settlement 
agreements granted SIPC advances to each of the funds that directly 
held Madoff accounts. In the Katz-Wilpon agreement, the allowed 
customer claim was assigned to the Trustee, so that customer 
distributions will fund the $160 million settlement payment. 

For both the Picower and Levy settlements, the Trustee told us he 
believed he obtained the largest sums possible. The $5 billion 
settlement in the Picower case represents 100 percent repayment of 
funds received by the Picower estate and related investors named in 
the complaint. Similarly, the Levy settlement, reached before 
litigation was filed, represents nearly 100 percent of the amounts the 
Levy account holders withdrew during the 6 years prior to the Madoff 
firm's collapse.[Footnote 87] Additionally, both Picower and Levy 
withdrew their customer claims as part of the settlements, the Trustee 
said. 

[End of section] 

Appendix V: GAO Contact and Staff Acknowledgments: 

GAO Contact: 

A. Nicole Clowers, (202) 512-8678 or clowersa@gao.gov: 

Staff Acknowledgments: 

In addition to the contact named above, Cody J. Goebel, Assistant 
Director; Donald W. Brown; Daniel S. Kaneshiro; Jonathan M. Kucskar; 
David J. Lin; Marc W. Molino, Barbara M. Roesmann; Christopher H. 
Schmitt; Andrew J. Stephens; Ethan S. Wozniak; and Henry R. Wray made 
major contributions to this report. 

[End of section] 

Footnotes: 

[1] A Ponzi scheme is an investment fraud that involves the payment of 
purported returns to existing investors from funds contributed by new 
investors. 

[2] See GAO, Securities Investor Protection Corporation: Interim 
Report on the Madoff Liquidation Proceeding, [hyperlink, 
http://www.gao.gov/products/GAO-12-414] (Washington, D.C.: Mar. 7, 
2012). 

[3] The cash limitation amount is subject to potential adjustment for 
inflation every 5 years. According to SIPC, the $500,000 limit for 
securities, rather than the limit for cash, applied in the Madoff 
liquidation. At the start of the Madoff case, the cash protection 
limit was $100,000 per customer. 

[4] The Dodd-Frank Wall Street Reform and Consumer Protection Act 
raised the potential minimum assessment that SIPC can generally charge 
member firms, basing it on a percentage of gross revenues from the 
firm's securities business. Pub. L. No. 111-203, § 929V(a), 124 Stat. 
1868 (2010). 

[5] For SIPC to initiate a proceeding, at least one of the following 
other factors must exist: (1) the firm must be insolvent under the 
Bankruptcy Code or unable to meet its obligations as they become due; 
(2) the firm is subject to a court or agency proceeding in which a 
receiver, liquidator, or trustee has been appointed; (3) the firm is 
not compliant with applicable requirements under the Securities 
Exchange Act of 1934 or financial responsibility rules of SEC or 
financial self-regulatory organizations; or (4) the firm is unable to 
show compliance with such rules. In the smallest proceedings (in 
which, among other factors, the claims of all customers are less than 
$250,000), SIPC may directly pay customer claims without filing an 
application for a protective decree with a court and without the 
appointment of a trustee. 

[6] The letter also informs claimants of their right to object to the 
determination and how to do so. The bankruptcy court judge overseeing 
the liquidation rules on a customer's objections after holding a 
hearing on the matter. Decisions of the bankruptcy court may be 
appealed to the appropriate federal district court, and then upward 
through the federal appellate process. 

[7] This may not be representative of the entire Madoff customer 
population, but illustrates the age issue. Information is for 
individual applicants with reported ages. 

[8] 15 U.S.C. § 78lll(11). 

[9] For a fuller discussion of legal issues surrounding the Trustee's 
decision to use NIM, and the appeals that followed the decision, see 
the legal appendix in [hyperlink, 
http://www.gao.gov/products/GAO-12-414]. For the Supreme Court's 
denial, which came after our first report, see 654 F.3d 229 (2nd Cir. 
2011), cert. dismissed, 132 S. Ct. 2712 (U.S. June 4, 2012) (No. 11-
968), cert. denied 2012 WL 396489 (U.S. June 25, 2012) (11-969), and 
cert. denied 2012 WL 425188 (U.S. June 25, 2012) (11-986). 

[10] In addition to the facts of each case, actual liability will 
depend on legal issues such as the applicability of the "stockbroker 
safe harbor" (Bankruptcy Code § 546(e)), the standards used to 
determine good faith, and the extent of the trustee's standing to 
bring common law claims. 

[11] We classified more than 99 percent of accounts based on account 
names. See appendix I for more information about our methodology and 
the accounts we could not classify. SEC officials told us that Madoff 
urged smaller investors to pool funds with others, such as friends or 
relatives, to increase account size. They said this may help explain 
the number of entities such as trusts or partnerships that were among 
the firm's customers. We use "family" to include family partnerships 
and trusts, of which there were many, according to the Trustee. 

[12] Feeder funds are investment vehicles that collected money from 
investors and channeled the funds to the Madoff firm. We describe 
Trustee litigation against feeder funds later in this report. 

[13] In the analysis presented in this section and elsewhere, we 
relied on our own examination of data provided by the Trustee. The 
Trustee reviewed our results, noting some small differences between 
our review and his work, but generally confirming our findings. Upon 
further review of our work, we determined that the variances arose 
from differing methodologies and particular algorithms used to conduct 
the analysis. For example, we excluded from our analysis accounts that 
had no transactions, and we counted multiple allowed claims for one 
account as a single allowed claim. Further, in determining whether 
accounts were net winners or net losers, we summed all applicable 
deposits and withdrawals to determine net positions. In a small number 
of cases, this produced very small negative or positive account 
balances, which we considered to be different than zero. 

[14] We note the effects of one large account--Decisions, 
Incorporated, an institutional account used by Jeffry Picower. As part 
of a settlement, the Trustee is recovering all of the about $6.2 
billion in net winnings for this account. If that account is excluded 
from the analysis, then the total net investment position for 
institutional accounts would increase from about $3.0 billion to about 
$9.1 billion. The settlement recovers a total of $7.2 billion, 
covering multiple Picower accounts. 

[15] According to SIPC executives, criteria for evaluation included 
whether the claimant: had an account; withdrew more than the amount 
deposited; and had knowledge of the fraud, or should have had 
knowledge of the fraud. 

[16] We excluded one account (among the accounts owned by Norman F. 
Levy) from our analysis in this section because its transaction 
history varied significantly from other accounts. Trustee data show 
this account had 64,000 transactions and $222 billion in transaction 
volume (the sum of deposits and withdrawals). The next largest account 
by volume had volume of $9 billion. At the time of the failure, this 
Levy account had a net investment position of only ($61 million), 
however, and a final statement amount of $0. As discussed later in 
this report, the Trustee reached a settlement to recover fictitious 
profits. 

[17] The Trustee denied claims from feeder fund investors, determining 
that although the feeder funds themselves qualified as Madoff 
customers under SIPA, the feeder fund investors did not. The investors 
contested that determination, and in June 2010, the Trustee filed a 
motion with the Bankruptcy Court requesting an order upholding his 
denial of the investors' claims. In June 2011, the Bankruptcy Court 
granted the Trustee's motion, Securities Investor Protection Corp. v. 
Bernard L. Madoff Inv. Securities LLC, 454 B.R. 285 (Bankr. S.D.N.Y. 
2011), after which the investors appealed. In January 2012, the U.S. 
District Court for the Southern District of New York affirmed the 
Bankruptcy Court's ruling. In re Aozora Bank Ltd., 2012 WL 28468 
(S.D.N.Y. 2012). The case is on appeal to the U.S. Court of Appeals 
for the Second Circuit. SEC and SIPC supported the Trustee's 
determination. 

[18] SIPC is paying $889 million in advances on these claims, 
according to SIPC. These advances will be counted against allowed 
claims when claims are paid. 

[19] Under the formula the Trustee is using, claim eligibility is 
based on a customer's net investment (total investments less 
withdrawals) plus any repayments of withdrawals of principal the 
Trustee has obtained as part of his efforts to recover assets. 

[20] As SIPC executives told us, in a SIPA liquidation, each account 
holder with an approved claim receives the same pro-rated share of 
customer property, regardless of whether they are an individual or an 
institution. Institutions generally receive more in total payments, 
but only because they invested, and lost, more principal. 

[21] SIPC executives told us they confirmed all of the Trustee's claim 
determinations, in order that subsequent distributions based on the 
claims decisions would be accurate. 

[22] Both of these accounts were large net losers. The Optimal account 
related to the claim lost $1.44 billion of principal, and M-Invest 
lost $541 million of principal. The difference between the allowed 
claim amount and the lost principal is due to repayment of certain 
withdrawals through settlements, which subsequently increased the 
allowable claim amount. As noted earlier, the Trustee generally denied 
claims from feeder fund customers, but feeder funds themselves were 
eligible to file claims as direct Madoff investors. 

[23] According to the Trustee, through August 2012, $1.1 billion had 
been distributed to customers, with an additional $2.4 billion 
expected by September 2012. 

[24] In addition to seeking recoveries through litigation efforts 
described here, the Trustee has also recovered other assets for the 
benefit of customers, including approximately $500 million in bank 
accounts and about $300 million in securities, he told us. The Trustee 
has also sold Madoff corporate assets. Thus far, the Trustee said 
total recoveries are about $9.1 billion. 

[25] In the descriptions of Trustee litigation that follow, we have 
summarized the Trustee's complaints, but not sought to independently 
verify allegations or obtain the views of any defendants. 

[26] The failure date is December 11, 2008. SIPC filed its application 
for liquidation on December 15, 2008, and the district court issued a 
protective decree that began the SIPA liquidation proceeding against 
Madoff and his firm. 

[27] In some cases, the Trustee's first contact with account holders 
has been the filing of these complaints. This followed attempts, early 
on in the case, to reach out to some account holders, to learn more 
about their situations, the Trustee told us. However, the reaction was 
negative, with complaints of invasion of privacy. Thus, the Trustee 
said he ceased those efforts. 

[28] The latest available data from the Trustee at the time of our 
review were as of May 2012. 

[29] In addition to these cases, the Trustee is also pursuing bad 
faith litigation involving feeder funds, which we describe later in 
this report. 

[30] Common law is the body of law derived from judicial decisions, 
rather than from statutes or constitutions. It is sometimes also 
referred to as case law. Conversion is the wrongful possession or 
disposition of another's property. Unjust enrichment is the retention 
of a benefit conferred by another, without appropriate compensation. 

[31] The cases are Picard v. JP Morgan Chase & Co., 460 B.R. 84 
(S.D.N.Y. 2011), and Picard v. HSBC Bank PLC, 454 B.R. 25 (S.D.N.Y. 
2011). See appendix II for details. 

[32] Twenty-nine of the 30 bad faith complaints are available 
publicly, either through the U.S. Courts' Public Access to Court 
Electronic Records (PACER) system or the trustee's website, 
[hyperlink, http://www.madofftrustee.com]. One bad faith complaint is 
under seal and unavailable. According to the trustee, this complaint 
involves a husband and wife as defendants and is sealed because it 
includes allegations about the defendants' federal income taxes. 

[33] In addition to the bad faith actions described here, the Trustee 
has also filed separate bad faith actions to recover funds received by 
"subsequent transferees"--third parties who received funds paid to 
them by Madoff customers. 

[34] Bankruptcy Rule 2004 provides the ability to develop information 
through examinations, depositions, and other methods, the Trustee told 
us. He said he used this ability to develop information leading to 
good faith/bad faith determinations. 

[35] Many feeder funds themselves are in liquidation, with their own 
trustees, the Trustee told us. Feeder fund customers can bring a claim 
within their own liquidations, he said. 

[36] Saul Katz and Fred Wilpon are co-owners of the New York Mets 
Major League Baseball franchise. Katz, Wilpon, members of their 
families, and business associates were named as defendants in the 
Trustee's clawback action. 

[37] The excluded case was sealed for confidentiality, with the 
Trustee having filed no clawback action. Thus, details were 
unavailable. 

[38] After the Trustee made determinations on customer claims for 
reimbursement, a market emerged for trading of approved claims in 
which the right to receive the distribution of customer property from 
the Madoff liquidation can be sold by the claimant to other parties. 

[39] As noted, the hardship program was open to individuals, but the 
Trustee told us that in some clawback cases, a family trust might have 
been named in an action. In such instances, if all matters tied back 
to individuals, the cases could nevertheless be considered under the 
clawback option of the hardship program. 

[40] SEC officials also told us that before the Trustee created the 
hardship program, there was significant agency concern about 
clawbacks, and whether the Trustee would file actions against small 
net winners. The specific concerns were high cost of pursuing such 
cases, and the relatively small amounts of money involved, the 
officials told us. They said they expressed these concerns to the 
Trustee, who thereafter created the program, which satisfactorily 
addressed SEC's concerns. 

[41] We confine our discussion here to federal income taxes. 
Additional tax effects are to be expected to the extent state-or local-
level income tax filings rely on or otherwise make use of federal 
income tax calculations. 

[42] Ordinarily, according to IRS officials, investors must pay taxes 
on dividends and interest, as well as on capital gains from sales or 
liquidations. They do not pay taxes on unrealized capital gains. 
However, customer statements from the Madoff firm indicated it was 
regularly liquidating customer positions; thus, capital gains and 
losses were reported to investors as realized gains or losses. SIPC 
executives told us positions were reported sold monthly, as Madoff 
purported to move customer holdings into cash. 

[43] We underscore these are general descriptions only. According to 
IRS officials and tax practitioners, individual cases can vary 
significantly, depending on specific taxpayer circumstances. 

[44] According to IRS, an investment theft loss deduction is allowed 
only to the extent an investor has not received any reimbursement for 
the loss. 

[45] Generally, according to IRS officials, the ability to amend prior 
returns or to take a theft loss deduction does not vary by taxpayer 
type, such as individual or corporation. However, when an individual 
invests in a Ponzi scheme through another entity, such as a 
partnership, the rules can be different. An individual who invests in 
a Ponzi scheme by means of a "flow-through" entity, like a partnership 
or a limited liability company treated as a partnership for tax 
purposes, can amend returns or take a theft loss deduction, based on 
the proportionate share of income or deductions that flow through to 
partners or members of the entity. By contrast, an individual 
investing through a C-corporation would not be able to take a theft 
loss deduction, as that option would lie with the corporation itself. 
In general, IRS officials said, the tax code treats the partnership as 
an entity, but doesn't tax at the entity level; instead, only 
participants are taxed. 

[46] See Internal Revenue Service, Rev. Proc. 2009-20. 

[47] Generally, an investor that had actual knowledge of the fraud 
before it became public would not be eligible to use the safe harbor, 
according to IRS officials. However, they could still amend returns to 
remove fictitious income or take a theft loss deduction, as 
applicable. This is because the tax code requires only that money was 
lost and that it was stolen from the investor. Even if an investor had 
knowledge of the fraud, as long as they were victims of theft, they 
are still eligible for amended returns or the theft loss deduction, 
according to the officials. However, depending on the investor's 
actual involvement with the fraud, IRS could deny the theft loss 
deduction or not allow removal of fictitious income, according to the 
officials. 

[48] In addition, deductions under the safe harbor approach take into 
account all previously taxed income from a scheme, whether it was 
fictitious or real, which is more similar to the FSM method than the 
NIM method. 

[49] Likewise, customers using the safe harbor approach may have an 
additional deduction of the remaining 5 percent or 25 percent of the 
loss, depending on the amount of recovered assets from the Trustee or 
from other claims. 

[50] They also noted that confidentiality provisions generally bar IRS 
from releasing taxpayer-specific tax return information. See Internal 
Revenue Code, sec. 6103, "Confidentiality and disclosure of returns 
and return information." 

[51] In simplified numerical terms, a taxpayer who paid tax at the 35 
percent rate on $100,000 in income would pay taxes of $35,000. But if 
the taxpayer later took a $100,000 theft loss deduction, and their tax 
rate at the time was no longer 35 percent, but had fallen to 15 
percent instead, the tax relief realized would be $15,000--$20,000 
less than the amount of tax they initially paid. IRS officials told us 
that owing to individual circumstances, the reverse could also be 
true: A taxpayer could have paid a lower rate at the time of the 
fraud, and a higher rate later, making a deduction more valuable. 

[52] The discussion here and in the "Future tax liability" item 
following focus on the safe harbor approach, but we note the same 
considerations also apply to theft loss deductions generally. 

[53] The Alternative Minimum Tax (AMT) is a parallel tax calculation 
that attempts to ensure that those benefiting from certain tax 
advantages still pay a minimum amount of tax. The AMT provides an 
alternative set of rules for calculating taxes, and if a taxpayer's 
tax as calculated in the normal manner falls below that amount, the 
taxpayer is required to make up the difference by paying the 
alternative minimum tax. 

[54] IRS officials noted that the Trustee's calculation of losses for 
claims purposes is different from losses calculated for tax purposes, 
and the difference in methods could affect whether additional taxes 
are triggered. 

[55] Although SIPC, SEC, and the courts have supported the Trustee's 
decision to use NIM, his selection of this method remains of interest 
and concern to some. Thus, we discuss the tax implications of both NIM 
and FSM for comparison purposes only. 

[56] However, once a customer's SIPC payment, under either NIM or FSM, 
reaches the SIPC coverage limit of $500,000 per customer, additional 
losses no longer have this tax effect, according to IRS officials and 
the trustee's counsel. That is, once the SIPC coverage limit is 
reached, taxpayers under either method would deduct similar amounts of 
losses, because there would be no additional SIPC payments to offset 
losses. 

[57] As discussed earlier, the Trustee did consider the tax issue with 
respect to hardship cases, whereby he took into account whether 
customers made withdrawals in order to pay income taxes. SIPC 
executives and SEC officials likewise told us tax liabilities were not 
a factor in their considerations. 

[58] According to the Trustee, IRS was interested in the NIM 
calculation of net equity; the length of time customer accounts were 
open, perhaps to assess the impact of potential tax benefits; and 
information on types of customer accounts, such as individuals, 
corporations, or nonprofits. The Department of Labor received account 
information. 

[59] See Treasury Inspector General for Tax Administration, Many 
Investment Theft Loss Deductions Appear to be Erroneous, 2011-40-124, 
September 27, 2011. 

[60] Given the timing of IRS's release of the guidance, we had 
insufficient time to evaluate its content. See the guidance at 
[hyperlink, http://www.irs.gov/uac/FAQs-Related-to-Ponzi-Scenarios-for-
Clawback-Treatment]. 

[61] The Trustee in the Madoff liquidation is Irving H. Picard; the 
trustee's counsel is the law firm of Baker & Hostetler LLP, of which 
the Trustee is a member. 

[62] We excluded from our analysis transactions we identified as 
instances of the Madoff firm debiting foreign account holders for 
purported U.S. federal income tax withholding and paying those amounts 
to the Internal Revenue Service (IRS). These amounts, totaling about 
$330 million, were later credited to the accounts through a December 
2011 settlement between the Trustee and IRS. 

[63] Specifically, we used a computer program to classify accounts 
based on keywords we identified that were associated with particular 
types of accounts, such as "JT" (an abbreviation of "joint trust," 
which indicates an individual or family account) or "pension" 
(denoting a pension plan, which indicates a institutional account). We 
reviewed the automatically classified accounts to identify cases for 
which the classification was not appropriate. For other accounts not 
containing the keywords we identified, we classified them by type in 
two independent examinations, and then reconciled any differences 
between the two reviews. 

[64] A revenue ruling represents IRS conclusions on the application of 
law to the facts of a situation. A revenue procedure is a statement of 
a procedure that affects the rights or duties of taxpayers or other 
members of the public, or otherwise should be a matter of public 
knowledge. 

[65] Section 6(b) of SIPA, 15 U.S.C. § 78fff(b), makes the Bankruptcy 
Code generally applicable to SIPA liquidations. Section 7(a), 15 
U.S.C. § 78fff-1(a), vests in SIPA trustees "the same powers and title 
with respect to the debtor and the property of the debtor, including 
the same rights to avoid preferences, as a trustee" under the 
Bankruptcy Code. In addition, section 8(c), 15 U.S.C. § 78fff-2(c), 
provides that whenever customer property is not sufficient to fully 
pay all applicable claims, "the trustee may recover any property 
transferred by the debtor which, except for such transfer, would have 
been customer property if and to the extent that such transfer is 
voidable or void" under provisions of the Bankruptcy Code. 

[66] Picard v. Merkin (In re Bernard L. Madoff Investment Securities 
LLC), 440 B.R. 243 (Bankr. S.D.N.Y. 2010), leave to appeal denied, 
2011 W.L. 3897970 (S.D.N.Y. 2011). 

[67] For a comprehensive description of the legal principles 
applicable to avoidance actions in Ponzi scheme liquidations and how 
they differ from other bankruptcies, see Mark A. McDermott, Ponzi 
Schemes and the Law of Fraudulent and Preferential Transfers, 73 Am. 
Bankr. L.J. 157 (Spring 1998). 

[68] Section 101(31) of the Bankruptcy Code defines "insider" in the 
case of a company such as the Madoff firm to include directors, 
officers, partners, other persons in control of the company, and their 
relatives. The courts recognize the term "insider" also may extend to 
anyone with a sufficiently close relationship to the debtor that they 
do not deal at arm's length. See, e.g., In re Longview Aluminum, LLC, 
657 F.3d 507, 509-510 (7th Cir. 2011). 

[69] Begier v. IRS, 496 U.S. 53, 58 (1990). 

[70] Wooton v. Barge (In re Ronald Cohen), 875 F.2d 508 (5th Cir. 
1989). 

[71] An affirmative defense is an assertion by a defendant, which, if 
proven, constitutes a defense to a charge or complaint. 

[72] McDermott, note 3, 176-177 (footnotes omitted). 

[73] See, e.g., In re Bayou Group, LLC, 439 B.R. 284 (S.D.N.Y. 2010). 

[74] See, e.g., In re Teleservices Group, Inc., 444 B.R. 767 (Bankr. 
W.D.Mich. 2011). 

[75] Picard v. Katz, 462 B.R. 447 (S.D.N.Y. 2011). 

[76] "Fair consideration" is defined in DCL § 272. 

[77] One potential difference between actual fraud avoidance under the 
Bankruptcy Code and New York law is that DCL § 276 may be interpreted 
to require a showing of fraudulent intent on the part of the 
transferee as well as the transferor. Court decisions applying § 276 
are inconsistent on this point. See precedents described in Picard v. 
Merkin, note 2, 440 B.R. at 257. 

[78] See Dowlings, Inc. v. Homestead Dairies, Inc., 88 A.D.3d 1226, 
1229-30, 932 N.Y.S.2d 192, 196-97 (2011), and cases cited. 

[79] Section 546(e) provides in part: " Notwithstanding sections 544, 
545, 547,548(a)(1)(B) and 548(b) of this title, the trustee may not 
avoid a transfer that is a margin payment . . . or settlement payment 
. . . made by or to (or for the benefit of) a . . . stockbroker . . . 
or that is a transfer made by or to (or for the benefit of) a . . . 
stockbroker . . . in connection with a securities contract, as defined 
in section 741(7), . . .that is made before the commencement of the 
case, except under section 548(a)(1)(A) of this title." 

[80] Picard v. Merkin, note 2. 

[81] Picard v. Katz, note 11. 

[82] See, e.g., Bankruptcy Code §§ 321, 541(a); Securities Investor 
Protection Corp. v. Bernard L. Madoff Investment Securities LLC (In re 
Bernard L. Madoff), 429 B.R. 423 (Bankr. S.D.N.Y. 2010), aff'd, 848 F. 
Supp.2d (S.D.N.Y. 2012). 

[83] Picard v. JP Morgan Chase & Co., 460 B.R. 84 (S.D.N.Y. 2011), and 
Picard v. HSBC Bank PLC, 454 B.R. 25 (S.D.N.Y. 2011). 

[84] This legal principle against adjudicating wrongdoer versus 
wrongdoer claims is called "in pari delicto." 

[85] The $20 million Hynes settlement (see table) was sealed for 
confidentiality, with the Trustee having filed no adversary action. 
Thus, details were unavailable to us. 

[86] Feeder funds are investment vehicles that collected money from 
investors and channeled the funds to the Madoff firm. 

[87] This does not include $84 million paid to the Levy foundation, 
which the Trustee did not seek to recover, because the funds had been 
donated to charitable causes. 

[End of section] 

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 website [hyperlink, http://www.gao.gov]. Each 
weekday afternoon, GAO posts on its website newly released reports, 
testimony, and correspondence. To have GAO e-mail you a list of newly 
posted products, go to [hyperlink, http://www.gao.gov] and select 
“E-mail Updates.” 

Order by Phone: 

The price of each GAO publication reflects GAO’s actual cost of 
production and distribution and depends on the number of pages in the 
publication and whether the publication is printed in color or black 
and white. Pricing and ordering information is posted on GAO’s 
website, [hyperlink, http://www.gao.gov/ordering.htm]. 

Place orders by calling (202) 512-6000, toll free (866) 801-7077, or 
TDD (202) 512-2537. 

Orders may be paid for using American Express, Discover Card, 
MasterCard, Visa, check, or money order. Call for additional 
information. 

Connect with GAO: 

Connect with GAO on facebook, flickr, twitter, and YouTube.
Subscribe to our RSS Feeds or E mail Updates. Listen to our Podcasts.
Visit GAO on the web at [hyperlink, http://www.gao.gov]. 

To Report Fraud, Waste, and Abuse in Federal Programs: 

Contact: 
Website: [hyperlink, http://www.gao.gov/fraudnet/fraudnet.htm]; 
E-mail: fraudnet@gao.gov; 
Automated answering system: (800) 424-5454 or (202) 512-7470. 

[End of document] 

Congressional Relations: 

Katherine Siggerud, Managing Director, siggerudk@gao.gov, (202) 512-4400
U.S. Government Accountability Office, 441 G Street NW, Room 7125
Washington, DC 20548. 

Public Affairs: 
Chuck Young, Managing Director, youngc1@gao.gov, (202) 512-4800
U.S. Government Accountability Office, 441 G Street NW, Room 7149 
Washington, DC 20548.