The Real Price of Proprietary Trading

Co-authored by James Crotty, Gerald Epstein & Iren Levina.

President Obama's endorsement of the "Volcker Rule" -- a set of proposals designed to reduce public financial support for risky proprietary trading and hedge/private equity fund ownership by commercial banks and bank holding companies -- has elicited an intense chorus of criticism from bankers, economists and policy makers. These critics make three main claims. First, proprietary trading had little to do with the current financial crisis and therefore restricting it would do little to prevent a replay; and second, proprietary trading provides a very small percentage of bank revenues and therefore is not significant. Third, Volcker's very narrow limits would make it easy for bankers to evade the restrictions.

We argue that conventional banker wisdom is incorrect: proprietary trading had a great deal to do with the crisis and it also contributes significantly to major bank revenue. Yet, critics are right to argue that the Volcker rule has to be strengthened and broadened, specifically to large investment banks and the shadow banking system, in order to significantly reduce the risk in the system to acceptable levels and to limit the likelihood of the need for future taxpayer bail-outs.

Proprietary Investment and Trading Helped Crash the System

Risky proprietary investments by investment banks, along with trading for clients whose decisions were influenced by these banks, was one of the main forces that sustained upward pressure on security prices in the bubble. Indeed, by running large trading books, banks had inside information on client trading patterns and could use that information to front-run, and thereby help sustain market trends.

Banks maintain large inventories of the securities ostensibly to facilitate trading, but these inventories in fact include substantial quantities of proprietary investments hidden within market-maker inventories. By 2008, bank trading books held hundreds of billions of disguised proprietary investments. By mid-April of 2008, banks had lost roughly $230 billion dollars on their super-senior CDO (collateralized debt obligation) proprietary holdings that regulators and other interested parties believed were simply inventories of assets held to facilitate client trading (Tett, 2008). These losses were probably created from about three quarters of a trillion dollars worth of risky assets. Clearly, proprietary trading was a major cause of the recent crisis.

Despite these massive losses, within a day or two of the Volcker Rule announcement, the press was full of stories quoting data from bank analysts that proprietary trading was very small. There is strong evidence that these widely cited numbers are much too low. A closer look at the data suggests where the low-ball estimates may be coming from. It also provides a good benchmark for us to provide estimates of the size of this activity on banks' revenues. First, as we see below, the widely reported data are likely to be taken from the crisis years, 2008 or 2009, rather than from the height of the bubble years. But if one is trying to assess the impact of strict rules in preventing future problems, surely the pre-crisis years are more relevant. Second, the quoted shares are of gross revenue rather than net revenue. Gross revenue is, of course, a much bigger number and therefore will reduce the size of the ratio. Net revenue is a much more relevant figure because it is that, which is divided into salaries, bonuses and profits. In other words, it is a much better measure of the bottom line. If one looks at the ratio or trading income as a share of net revenue while it was 5.1 % in the year of the crash, it was more than 45% at the height of the boom in 2006. Third, the widely cited estimates are almost certainly trying to estimate proprietary trading in the most narrow way possible (trading for own account) and likely do not take into account the often arbitrary lines that are drawn between own account, client account and market making aspects of trading, position taking and investment.

To illustrate these points we undertook some rough calculations of trading for three banks: Goldman Sachs, Morgan Stanley and Citi. These estimates must be seen as very rough because the data is so difficult to break down, but we believe they illustrate our points. First, consider Morgan Stanley. The data shows that in 2008, Morgan Stanley's trading and investment revenues were about 2% of total revenue as was widely reported in the press. Though we do not claim that this is precisely "proprietary trading" as defined narrowly, the fact that this figure for Morgan Stanley for 2008,was so widely cited by bank analysts does suggest that our data is in line with quoted estimates. Now, note that this 2% figure is from 2008, the year the system crashed. But in 2006, at the height of the bubble, trading income as a share of total revenue was more than 19%.

A similar story holds for Goldman Sachs. In 2008, trading income as a share of gross revenue was re-ported in the media to be around 10% and according to our figures, about 15%. But if one goes back to the boom years of 2006, it was more than a third of the gross revenue, almost 35%. As a percentage of net revenue, trading income was much higher, 36% in 2008; in 2006 and 2007, it was a whopping 64% or more of net revenue. For Citigroup, our numbers are even rougher than for Morgan Stanley and Goldman, but they tell an interesting tale. Trading and investment revenue as a share of gross revenue in 2006, at the height of the bubble, was only about 5% of gross revenue, the number cited by many in the press. If one uses the more appropriate net revenue figure then this share jumps to over 9%.

Interestingly, if one looks at the contributions to Citi's revenue losses during the crash, according to these admittedly crude estimates trading and principle investments played a significant role. In 2008, for example, trading and principle investment losses amounted to 20% of gross revenue and over 40% of net revenue. If one counts these trading losses as a percentage of the declines of total and net revenue, these numbers become much higher. For example, between 2007 and 2008, Citigroup's total revenues fell by almost $50 billion and net revenues fell by almost $26 billion. In 2008, Citigroup lost $22 billion which amounts to 44% of total revenue losses and more than 80% of net revenue losses. Contrary to the bankers and pundits that claim that "proprietary trading" did not cause the crisis, these losses led to a tax payer bailout and constitute, in fact, one of the main components of what most of us mean by "the financial crisis."


1. The Volcker Rule must utilize a broad definition of proprietary trading, investment and position taking if it is going to succeed. Trading and proprietary investments made a much bigger contribution to bank revenues (and losses) than bankers and the press have suggested. It is virtually impossible to distinguish proprietary trading and investments from market making and trading for clients, which often involve proprietary investments and position taking by the banks. Trying to make highly restrictive definitions of proprietary trading will make it very easy for banks to use accounting gimmicks to move investments, position taking and proprietary income to where it is not restricted by the rule.

2. The Volcker Rule must be expanded to significantly help reduce systemic danger. Most important is to expand restrictions to investment banks, as well as to the "shadow" banking and financial world.

3. There needs to be a revolution in accounting standards and reporting. Regulators and the public do not have adequate information on the quantity and quality of these investments. Accounting standards and reporting requirements need to be complemented by close, hands-on financial examination by regulators. Without this, banks will be able to manipulate trading, investments and positions around virtually any set of rules that are created.

A more detailed version of this post can be found at

Crotty, James; Gerald Epstein and Iren Levina. 2010. "Proprietary Trading is a Bigger Deal than Many Bankers and Pundits Claim." Political Economy Research Institute, January.
Tett, Gillian (2008) 'Super-senior' CDO investors flex their muscles," Financial Times, April 14.