SEC enshrining Value at Risk models despite the models' history of failure

Despite the failure of Value at Risk models, the SEC is enshrining these models in the rules for how much capital banks have to hold against their derivatives.

As reported by Market Watch,

A proposal issued by the Securities and Exchange Commission this week requiring banks to hold more capital for their securities derivatives enshrines an approach that some regulatory observers believe has failed..... 

based on the proposal, big banks can be approved by the SEC to model their own risk internally, using so-called “value-at-risk” modelling, to help calculate how much capital they must hold for their securities derivatives. 

Bank risk officers employ value-at-risk modeling, a complex calculation, to estimate the maximum amount they expect the institution can lose in a certain period of time. 

Former Goldman Sachs (US:GS) banker Wallace Turbeville argues that big banks will be able to produce the results they want by inputting their own assumptions into VAR models. 

He added that regulators don’t seem to be able to employ enough people to understand how risks are measured using these models, and banks’ risk officers are often too timid to raise concerns when they see them. 

“You have to ask the question: If the regulators can’t figure out the value-at-risk modelling what is the point of all this?”” Turbeville said. “What benefit are these derivatives providing people?”...

Regular readers know that banks should be required to provide ultra transparency and disclose on an ongoing basis their current global asset, liability and off-balance sheet exposure details.

With this information, market participants could assess the riskiness of each bank's securities derivatives and the adequacy of its capital.

Unlike the regulators, market participants are not constrained by having too few people who understand how to measure risk using these models or by being too timid to raise concerns.  In fact, since there is money to be made, one can expect that the market participants will act on the basis of how they assess the risk of the banks.

Clifford Rossi, professor at University of Maryland’s business school, said the SEC and other regulators don’t have enough resources to keep up. 

“Models in the wrong hands without sufficient understanding is a problem,” he said. 

“Applying the models in a way where we come away with a false sense of security is a problem.” 

Rossi said each bank applies different assumptions, a situation that adds complexity for regulators.

This is why market participants need banks to provide ultra transparency so the market participants can do their own analysis.

Meanwhile, he added, risk officers are spending more time dealing with regulatory requests, leaving them with less time to monitor the firm’s own internal risks. 

“How do we really know that Wells Fargo (US:WFC)  has a less risky profile than another big bank?” Rossi said. “Unless you are sitting next to them as they work on their terminal it is extraordinary difficult to keep tabs.”

 The way to keep tabs is to have banks provide ultra transparency.

He said that regulators and bank boards need to oversee that the risk managers are keying in the assumptions on the model — not the traders. 

Michael Greenberger, a former Commodity Futures Trading Commission official and a professor at the University of Maryland School of Law, said all the value-at-risk models are unsatisfactory in terms of finding out how much capital a bank should hold. “This is all pie-in-the-sky mathematics that has no correlation to the real world,” he said.

Actually, Value at Risk is the bank's answer to the regulators when the regulators ask if they have their risk under control.

Turbeville said observers need to step back and think about how the U.S. financial system functioned before derivatives. 

“If you can’t figure out what the risk and there is no way to actually figure out what the projection is and these things can be highly explosive, then don’t we really need to say we should not do certain things and quit trying to measure what is impossible to understand,” he said.