General Accountability Office Struggles with Dodd-Frank’s Volcker Rule

Bradley Sabel is a partner at Shearman & Sterling LLP. This post is based on a Shearman & Sterling client publication by Mr. Sabel and Donald Lamson. Other posts about the Volcker Rule are available here.

The General Accountability Office (“GAO”) recently completed a study mandated by the Dodd-Frank Wall Street Reform and Consumer Protection Act of the risks and conflicts of interest associated with proprietary trading by U.S. banking institutions (the “Study”). Proprietary trading will be prohibited by Dodd-Frank’s Volcker Rule beginning in July 2012, and the Study was intended to provide input for the U.S. bank regulatory agencies in their proposed regulations implementing its requirements, which are expected in the very near future. The Study addresses (1) what is known about the risks and conflicts of interest associated with proprietary trading and the potential effects of the Act’s restrictions and (2) how regulators have overseen such trading and the challenges they face in implementing the restrictions. This is a more limited approach than the scope of the Study that Congress mandated, and already has garnered criticism by two Senators.

GAO completed its Study [1] with a recommendation that regulators should collect and review more comprehensive information on the nature and volume of activities potentially covered by the Act as a predicate to adopting final rules implementing the Volcker Rule. [2] The recommendation is not surprising, given that GAO encountered difficulty in collecting information on proprietary trading to complete its Study. The Treasury Department and the financial regulators have implicitly declined or at most agreed only to consider this as part of their rulemaking, as they apparently have concerns with the scope and direction of the recommendation.

I. Risks of Proprietary Trading


One of the basic problems in studying proprietary trading lies in defining the term. The Act defines “proprietary trading” in Section 619, which contains the Volcker Rule, to include engaging as principal for the trading account of the banking entity in any transaction to purchase or sell or otherwise acquire or dispose of securities, derivatives, contracts for sale of a commodity for future delivery, and an option on the foregoing. A number of instruments, such as U.S. government securities, are excepted from the prohibition on proprietary trading. Regulators have discretion to add other financial instruments to the list of prohibited instruments, including, for example, foreign exchange. A separate definition of “trading account” refers to profiting from short term price movements. [3]

GAO referred to the definition of proprietary trading in Section 619 but, due to data collection problems, limited its Study to stand-alone proprietary trading, which it defined as “trading at stand-alone proprietary trading desks, which are those organized by a banking entity with the specific purpose of conducting trading using the firm’s own capital.” GAO apparently ignored that there is a second definition of proprietary trading that was to inform its Study. Section 989 of the Act, which requires the Study, defines proprietary trading more broadly to include investing as principal in securities, commodities, derivatives, hedge funds, private equity firms, or other financial products that the GAO might add. [4] The definition in Section 989 has no exceptions and comprehends long-term investments. This omission by GAO to consider a broader array of financial instruments has led to criticism, as will be examined below.

GAO recognized that proprietary trading can take several forms. Banks conduct proprietary trading in separate desks, or in conjunction with their market making activities, by accumulating positions at levels exceeding the firm’s typical or necessary inventory for facilitating customer trades. FDIC representatives informed GAO that in 2005 the regulators tried to define proprietary trading as part of an effort to better oversee such activity but ultimately could not. They noted that preventing proprietary trading required a subjective, case-by-case evaluation.

The Study also found that, notwithstanding the Volcker Rule’s focus on short-term trading, some think the definition of proprietary trading should be expanded to include other activities conducted as principal, such as long-term investments, which in some cases have caused significant losses or failures. Part of the failure of Lehman Brothers was attributed to investment in commercial real estate and private equity investments. Some suggest that proprietary trading should be defined to include holding risky tranches of mortgage-backed securities, CDOs, or other securities that can be held by an underwriter as part of a securitization. Conversely, some argue for a narrower definition, out of concern that the Volcker Rule, if read literally, could reach conduct that never was intended to be covered.

Quantitative Results

Banks conduct proprietary trading both at proprietary-trading desks and elsewhere within their firms. GAO determined that collecting information on activities other than at stand-alone proprietary trading desks was not feasible because the firms did not separately maintain records on such activities. As a result, GAO limited its analysis to data on 26 stand-alone proprietary-trading desks at the six largest U.S. bank holding companies from June 2006 through December 2010. GAO considered the firms’ combined revenues or losses from stand-alone proprietary trading for each of the 18 quarters in that time period, and compared the results to revenue from all trading activities (including stand-alone proprietary trading) and total bank holding company revenue.

GAO made a number of quantitative findings, primarily that the risks of stand-alone proprietary trading are disproportionately large compared to any benefits.

  • Compared to overall revenues, stand-alone proprietary trading generally produced small revenues over several years as opposed to large losses during the financial crisis.
  • Over 13 quarters, stand-alone proprietary trading produced combined revenues of $15.6 billion in six firms, but losses over five quarters of $15.8 billion, resulting in an overall loss of about $221 million.
  • In 108 individual firm-quarters, stand-alone proprietary trading did not significantly increase quarterly trading revenues during positive quarters; in quarters when both stand-alone proprietary trading and total trading resulted in losses, proprietary trading comprised a substantial portion of total losses.
  • Stand-alone proprietary trading involved greater risk than total trading activities on average to generate the same amount of revenue; the firms’ Value at Risk (“VaR”) risk models were less capable of predicting the actual risks associated with proprietary trading.
  • These firms’ hedge and private equity fund investments also experienced small revenues in most quarters but somewhat larger losses during the crisis compared to total firm revenues.

In reaching these conclusions, GAO acknowledged but apparently did not take into account a number of weaknesses in its data gathering. Even assuming it is appropriate to study proprietary trading with a population of only six holding companies, there were significant problems with the data that GAO was able to collect. In certain instances, the holding companies identified proprietary trading activity for which they were unable to provide revenue data. Further, some of the activity within the firms’ stand-alone desks, such as trading in government securities, might not be subject to the Volcker Rule’s prohibition. The firms also may have included losses from CDO underwriting activities in the data on stand-alone proprietary trading.

Moreover, GAO may have misinterpreted available data. Most basic, four of the firms made money, and two lost money, from stand-alone proprietary trading over the 4.5 year time period, but GAO portrayed the activity in the aggregate as money-losing. One of the six firms was responsible for both the largest quarterly revenue at any single firm of $1.2 billion and two of the largest quarterly losses of $8.7 billion and $1.9 billion, but no apparent effort was made to consider the group’s income or losses without including such outlying data.

Finally, GAO acknowledged, but still ignored in its analysis, the fact that proprietary trading losses during the recent crisis were dwarfed by other problems. Losses associated with lending and other risky activities during the recent financial crisis were far greater than those associated with stand-alone proprietary trading. For example, one of the firms increased its loan loss reserve by more than $14 billion in 2008 and another increased its reserves by almost $22 billion in 2009.

Qualitative Evidence

After analyzing the available quantitative evidence, GAO considered the qualitative arguments for and against implementing restrictions on proprietary trading. While GAO identified the concerns of financial institutions, policymakers, and researchers about the potential negative consequences of the restrictions, it generally discounted those concerns and presented counterarguments. These concerns included that the restrictions could:

  • reduce the ability of banks to offset risks in other areas,
  • reduce the competitiveness of U.S. firms by limiting their activities compared to their international competitors, [5]
  • reduce the amount of liquidity in financial markets, and
  • push risky trading activities to less-regulated institutions, such as hedge funds.

On the funds side, market participants argued against the Volcker Rule’s one year limit for a sponsoring bank to limit its holdings of a fund to 3% of the fund’s assets. Participants claimed that the 3% test ignores the market’s demand that a fund generally must have a three year track record before its interests can be successfully marketed. [6]

II. Supervision

GAO reviewed what it portrayed as a history of weak supervision by financial regulators leading to the recent crisis. It also set out its apparent disagreement with regulators on the quantitative financial data necessary to issue final regulations implementing the Volcker Rule’s restrictions.

“Challenged” Supervision

GAO began by reviewing its own record of criticizing financial regulators, especially the Federal Reserve, Securities and Exchange Commission (“SEC”), and Office of Thrift Supervision (“OTS”), for being “challenged” in overseeing large financial institutions’ risk management efforts. In particular, GAO noted that the supervisory-related failures of SEC and OTS resulted in changes that eliminated their role in overseeing holding companies. With respect to trading, before Dodd-Frank, oversight of trading activities generally did not distinguish between proprietary trading and trading conducted on behalf of customers, because regulators examined both activities when assessing risk management. There generally were not separate procedures for examining proprietary trading activities.

The Study notes improvements in the quality of bank supervision. Since the crisis, various regulatory changes have been made to reduce the risks that trading and other activities pose to the safety and soundness of large institutions. Regulators are overseeing changes that financial institutions are making to their risk management models, including improvements to their stress testing. Several holding companies now use VaR measures with longer time horizons that include a fuller range of economic cycles to increase their models’ accuracy and consistency. Changes to capital requirements, broadly and with respect to the trading books at financial institutions, should also mitigate risks to the financial system.

FSOC Report

GAO reviewed at length the Financial Stability Oversight Council’s (“FSOC’s”) January 2011 report making recommendations for implementing the Act’s restrictions on proprietary trading and hedge fund and private equity investments. GAO largely endorsed these recommendations.

FSOC had found that the most basic challenge for regulators was to distinguish prohibited proprietary trading from market making and appropriately define terms associated with the restrictions on hedge fund and private equity fund investments. The FSOC study approached the issue by recommending that firms monitor certain metrics that could indicate when impermissible proprietary trading is occurring within permitted market-making activities. FSOC suggested and GAO agreed that the use of quantitative metrics related to revenue, risk, inventory, and customer flow could guard against future impermissible activities. Revenue-to-risk measures can help distinguish market-making from proprietary trading, because the lower VaR measures associated with market-making result in higher revenue-to-risk data than with proprietary trading. In addition, in light of the Volcker Rule’s focus on short-term activity, inventory turnover and aging metrics can track the length of time assets remain on balance sheet.

GAO warned against too light an approach to supervising trading. Without appropriate monitoring, financial institutions could abuse permissible activities, using them to conduct prohibited proprietary trading activities. GAO concluded that some financial institutions have pursued strategies that were a combination of client-focused transactions and proprietary positioning. These combined activities could be considered impermissible but go unnoticed if they were not monitored appropriately.

Implementing Rules

The Act requires the Federal Reserve, Office of the Comptroller of the Currency (“OCC”), Federal Deposit Insurance Corporation, SEC, and the Commodity Futures Trading Commission (“CFTC”) to issue final rules to implement the restrictions on proprietary trading and hedge fund and private equity fund investments by October 2011. GAO criticized the agencies’ efforts to collect data that may be useful in adopting these regulations. In response to GAO’s specific suggestion that the Federal Reserve and OCC gather qualitative and quantitative data on the trading activities at each of the six firms, the agencies agreed only to collect some general information.

These regulators initially considered collecting specific data on the nature and volume of proprietary trading and investment activity at the largest firms as part of the FSOC study, but chose instead to gather only qualitative information about how the entities monitor and manage the risks of trading and investment activities. The regulators have not collected more comprehensive information on the grounds that they have not yet written the final rules to define the types of trading and investment activities that will be prohibited. As a result, the regulators have not compiled specific data on the nature and volume of trading at stand-alone proprietary trading desks or reviewed the extent to which the firms are taking proprietary positions as part of conducting other trading or investment activities.

GAO believes more comprehensive information, including more complete data on the amounts of revenue and VaR levels of these firms’ market-making desks, would provide regulators a comparative baseline for observing trading inventories and revenue changes. The final rules would be more effective if regulators first identified and collected information on a broader set of activities than may be prohibited, to help ensure they are aware of all trading and funds that could be covered. Completing a more in-depth review of activities that may be covered by the Act could provide information on the potential impact of the restrictions, how firms are preparing for them, whether there are efforts to evade the restrictions, and how to improve monitoring and enforcement. Because the regulators have not collected more detailed information on the number and nature of trading desks where proprietary trading could be occurring, or firms’ hedge fund and private equity fund investment activities, they risk not being able to most effectively implement the restrictions.

Thus, in seeming frustration at the refusal of the bank regulators, GAO suggested that the FSOC direct the Office of Financial Research (“OFR”) to collect such information and share it with regulators as authorized under the Act, a step that goes beyond the recommendations in the FSOC’s report. [7]

III. Reaction

GAO customarily gives agencies that are the subject of a study or report the opportunity to respond to findings and recommendations before it issues a final document. Not a single regulator outright endorsed the GAO’s recommendation to use the OFR to gather information as a predicate to a final rulemaking implementing the Volcker Rule. The agencies replied with virtually identical letters, thanking the GAO for its efforts. The Treasury Department and the bank regulators also tactfully responded that it would be preferable for the agencies to collect and review necessary information, without mentioning the OFR. The SEC and CFTC simply promised to consider the recommendation.

GAO also provided a draft of the Study to Senators Merkley and Levin, the authors of the Volcker Rule, who provided a more direct reaction. [8] The Senators were disappointed that the Study was based on a subset of the narrower version of proprietary trading in Section 619 of the Act, stand-alone trading in the six largest U.S. bank holding companies, and thus did not capture the scope of the term that the Senators intended when Congress mandated the Study. As a result, by focusing only on stand-alone proprietary trading conducted in six bank holding companies, the GAO did not study proprietary trading wherever it occurs. Nor did the Study explain how proprietary trading losses can contribute to losses in bank capital. Accordingly, the Senators asked the GAO to revise its Study. It is not clear how the GAO revised its study in response to this reaction.


It is unclear how significant an effect the Study will have on financial regulators as they hurry to complete their implementing rules by October 2011, without yet having proposed any rules. From the reaction gained thus far, it is clear that GAO has not satisfied critics who oppose or support the Volcker Rule. Moreover, there are clear divisions among government agencies on how best to adopt final rules implementing the Volcker Rule. Worse still, financial institutions subject to the Volcker Rule have received little additional guidance on how best to comply with the rule. As regulators prepare to issue proposed rules, they may be chastened by the criticism directed at the GAO. As supporters and critics of the Volcker Rule continue to debate these points, it seems questionable whether the final decisions on the scope of the Volcker Rule will satisfy anyone.


[1] GAO-11-529, “Proprietary Trading: Regulators Will Need More Comprehensive Information to Fully Monitor Compliance with New Restrictions When Implemented” (July 2011) (“Study”), mandated by the Dodd-Frank Wall Street Reform and Consumer Protection Act, Public Law No. 111-203, 124 Stat. 1376 (2010) (“Dodd Frank” or the “Act”).
(go back)

[2] There are two basic features of the Volcker Rule that apply to banking groups (including any subsidiary in the group) operating in the United States: a prohibition on “proprietary trading” unrelated to customer-driven business, and a prohibition on private equity and hedge fund “sponsorship” and investment, subject to exceptions, and restrictions on certain relationships (so-called “covered transactions”) with “advised” or “managed” private equity and hedge funds. If you wish to review further information regarding the Volcker Rule, you may refer to our Client Publications entitled “FSOC Study on Implementing the Volcker Rule — A Series of Missed Opportunities and Some Surprises” (January 24, 2011) at–a-series-of-missed-opportunities-and-some-surprises-01-24-2011/ and “The Volcker Rule Continues to Garner Outsized Attention in the Wake of Passage of Financial Reform Legislation” (October 19, 2010) at–the-volcker-rule-continues-to-garner-outsized-attention-in-the-wake-of-passage-of-financial-reform-legislation-10-19-2010.
(go back)

[3] See Dodd-Frank Section 619(h)(4) and (6), 12 U.S.C. 1851(h)(4) and (6).
(go back)

[4] Dodd-Frank Section 989(b)(2). Congress directed the GAO to study whether: (A) proprietary trading presents a material systemic risk to the stability of the U.S. financial system; (B) proprietary trading presents material risks to the safety and soundness of the entities that engage in such activities; (C) proprietary trading presents material conflicts of interest between entities that engage in trading and their clients; (D) clients of those entities receive adequate disclosure regarding the risks and conflicts of proprietary trading; and (E) regulators are able to monitor and contain any risks and conflicts of interest related to proprietary trading. In conducting the study, the GAO was directed to consider (A) current practice relating to proprietary trading; (B) the advisability of a complete ban on proprietary trading; (C) limitations on the scope of activities that entities may engage in with respect to proprietary trading; (D) the advisability of additional capital requirements for covered entities that engage in proprietary trading; (E) the need for enhanced restrictions on transactions between affiliates related to proprietary trading; and (F) the need for enhanced public and accounting disclosures relating to proprietary trading. As explained below, the GAO study touches only on some of these topics.
(go back)

[5] GAO admitted that foreign regulators indicated that the Volcker Rule could cause U.S. banks to lose business to their competitors in Europe and elsewhere.
(go back)

[6] GAO downplayed this concern by noting that losses occurring at hedge funds and other nonbank entities are less likely to pose risks to the U.S. banking system than those occurring within bank holding companies. To the extent that proprietary trading migrates to nonbanks, no actual reductions in the level of market liquidity may occur. Yet GAO separately acknowledged that some failures at nonbanks illustrate the potential for significant financial losses and systemic risks: Long-Term Capital Management (1998) (Fed-facilitated recapitalization to avoid a rapid liquidation of trading positions might cause adverse effects on some markets) and Bear Stearns (2007) (infusion of cash into hedge funds that experienced losses from holding CDOs containing subprime mortgages).
(go back)

[7] This suggestion ignores that the bank regulators are voting members of the FSOC, making it unlikely that the FSOC would vote to direct the OFR to collect data.
(go back)

[8] See letter dated June 23, 2011 from Sens. Jeff Merkey and Carl Levin to Hon. Gene Dodaro, Comptroller General.
(go back)

Both comments and trackbacks are currently closed.

One Comment

  1. Wayne Isaacks
    Posted Monday, August 29, 2011 at 11:53 am | Permalink

    merging investment banks into commercial banks to save the failed investment basnks poisons the commercial banks with the incipient disease of the investment banks. That disease is letting the “house” gamble. Like any broker or any hedge, what was a lucative and market-making activity gives over to the wild risk of specualtion and outright gambling. Unrestrained bonuses on “profits” measured without the inextricably linked losses and the break-even at best record of proprietary trading, yields huge compensation to bankers (bookies) gambling with the House money, and eventually with the customers'(beters’) money. Failure is certain as this is little reomved from Madhoff’s scheme.