Should Value at Risk Be Abandoned Entirely?

Many banks are still using the risk measure, known as Value at Risk (VaR), despite its failure leading up to the 2008 financial meltdown.

var-large.jpg

Wall Street banks are continuing to use a widely discredited measure of risk, which could prove to be a “very treacherous guide to the future,” according to the author of a new book on the financial crisis.

Pablo Triana, a professor of finance at ESADE Business School in Barcelona, says that many banks are still using the risk measure, known as Value at Risk (VaR), to determine leverage on their trading portfolios and how much capital they should set aside against possible losses. Even though VaR was widely discredited after the crisis, its use has not been prohibited. The Basel Committee on Banking Supervision has adopted new regulations for banks that continue to use VaR, creating a series of add-ons designed to better reflect the risk involved and to determine how much regulatory capital is necessary for banks to hold when VaR is used to determine leverage.

“I think it can be argued that the Basel committee’s measures don’t go far enough and create a kind of moral hazard,” says Triana, whose book is entitled The Number That Killed Us. “If they openly recognize that VaR messed up then why do they still embrace it? I think it would be much better to get rid of it.”

VaR is a mathematical and analytical method that tries to estimate future losses for a trading portfolio with a statistical degree of confidence. It forecasts, for example, that a bank will lose a maximum of X amount with 95 percent or 99 percent probability. Of course, as we now know, in 2008 an event outside the so-called statistical “fat tail” in the other 1 percent took place and caused huge losses in the trading portfolios of banks that were leveraged up to 1000 percent on subprime bonds.

Triana argues that regulators ought to require that banks go back to the systems they used before VaR became popular in the 1990s. “If you stop listening to flawed mathematical models and you begin to think and use common sense, I think that can be an improvement,” Triana says.

Sponsored

Before becoming an academic, Triana worked for Credit Suisse in London. He received a master’s degree in economics from New York University’s Stern School of Business.

Triana argues that banks could continue to use VaR for internal decisions about investing their portfolios, but that regulators should halt the practice entirely when it comes to determining how much capital needs to be set aside to offset a trading risk. He prefers instead the net capital rule, which was used by the Securities and Exchange Commission starting in the 1920s, as superior to using VaR. Under this system, a human regulator rather than a computer decided that assets that are obviously more risky than other assets should carry a higher capital charge.

“VaR doesn’t differentiate between the fundamental characteristics of assets,” Triana argues. “All it does is look at the past historical data and infer some future risks for the assets.” Triana argues persuasively that the use of VaR was a major cause of the financial crisis. “All of the Wall Street banks had a lot of toxic trading assets like credit default obligations and subprime bonds and they had almost no capital to back that up,” he says. “VaR allowed banks to leverage up to one thousand to one on trading portfolios that were very toxic.” He quotes Lord Turner, chairman of Britain’s Financial Services Authority, as saying that “it is clear in retrospect that VaR measures of risk were faulty.”

Among the rule changes, the Basel Committee will now insist that banks use the worst historical data when making their VaR predictions of future losses, not just average periods. But Triana says the new rules still come up short. “They are still based on probability models. They can still be imperfect and they still yield very funny outcomes,” he says.

Another shortcoming of VaR is that it treats assets alike regardless of risk. So it would measure Treasuries and subprime bonds in the same way, even though their risk profiles are completely different. All it does is look at historical data for those assets. Triana argues in his book that banks should have put their portfolios of subprime bonds in their banking books rather than their trading books. But they knew that with VaR they could leverage those portfolios dramatically, which they would not be allowed to do in their banking books, he claims.

“Any regulatory system that allows a bank almost unlimited leverage on what probably were the most toxic securities ever conceived by humanity is very flawed and a very dangerous system of regulation,” Triana says.

Related