Wednesday, May 16, 2012

No, None of This Makes Any Sense

Frederick J. Sheehan is the author of Panderer to Power: The Untold Story of How Alan Greenspan Enriched Wall Street and Left a Legacy of Recession  (McGraw-Hill, 2009) and "The Coming Collapse of the Municipal Bond Market" (Aucontrarian.com, 2009)



Frederick Sheehan will speak at the Committee for Monetary Research and Education (CMRE) dinner on Thursday, May 17, 2012. It will be held at The Union League Club in New York. He will discuss "How We Got Here."
 
            After the financial crisis in 2008, "Too-Big-To-Fail" banks had to go. In 2006, the four largest banks - J.P. Morgan Chase, Bank of America, Citigroup, and Wells Fargo - held 33% of U.S. bank assets. Now, they hold 41% of U.S. bank assets and grow by the minute.

The Federal Reserve is, at least on paper, the country's leading bank regulator. Instead, it behaves like the TBTF banks' turbocharger. Federal Reserve Chairman Ben S. Bernanke is full of talk, and nothing else:

"First, is that 'viewed too big to fail' is a very, very serious problem, and one that was much bigger than expected. And I think that if there is only one thing we do in financial reform, it is to get rid of that problem."

-Federal Reserve Chairman Ben S. Bernanke, November 17, 2009, testifying before the Financial Crisis Inquiry Commission.


            A cause of the 2008 financial crisis was the failure of bank-risk models. Those who understood the Value at Risk model (VaR: the standard) knew it would fail. It is designed to fail in a financial crisis. The same model failed at Long-Term Capital Management and Enron. Yet, the Value at Risk model is still the primary model used to limit risk at financial institutions.

A financial crisis develops once in a blue moon. Therefore, there is less than a one percent chance of a meltdown, as defined by the model. The VaR model captures 95% (or 99%) of possible scenarios, as defined by banks and, supposedly, in conjunction with regulators and rating agencies. ("House prices never go down nationally.") J.P. Morgan invests within the 99% of scenarios as modeled by VaR.

If the VaR model were to include that one percent ("tail risk," in the argot) in its measurement of likely losses, J.P. Morgan would only hold Treasury bills. That assumes J.P. Morgan thinks the risk-free rate is defined by Treasuries. The Bank of Bernanke is doing its all to terminate this academic benchmark.

Ben Bernanke did not discuss VaR models before the Financial Crisis Inquiry Commission in 2009. He is a vague sort of fellow, so et cetera-ed himself from the burden of learning anything about banking before his appearance:

"To avoid another financial crisis, we need to identify "the macroeconomic context, evolution in the types of businesses, and their risk management, et cetera."

-Federal Reserve Chairman Ben S. Bernanke, November 17, 2009, testifying before the Financial Crisis Inquiry Commission.


On the very same day when J.P. Morgan Chairman Jamie Dimon announced his bank had lost a few billion dollars due to a haywire VaR model, Simple Ben told a congregation of central banking enthusiasts in Chicago what a swell job he is doing as the United States' leading bank regulator. The speech is a piƱata of false claims poised to scatter around the global village. The final blow could strike at any time. Possibly, at a bank with a $71 trillion derivative book (i.e., J.P. Morgan):    

 

"A number of key systemic risk measures that evaluate the potential performance of firms during times of financial market stress have improved in recent months. These indicators of systemic risk are now well below their levels in the crisis, and, overall, they present a picture of a banking system that has become healthier and more resilient. ....Such measures include the conditional value at risk, or CoVaR, which is an estimate of the extent to which a bank's distress would be associated with an increase in the downside risk to the financial system."

 


-Federal Reserve Chairman Ben S. Bernanke, "Banks and Bank Lending: The State of Play," conference on Bank Structure and Competition, Chicago, Illinois, May 10, 2012

The Fed chairman probably thought he would impress the audience when one of his footmen wrote"CoVaR" rather than "VaR" in his speech. At least, Investopedia.com does not rate CoVaR any better than VaR at controlling that good-for-nothing tail risk:

 

CoVaR: Conditional Value at Risk was created to be an extension of Value at Risk (VaR). The VaR model does allow managers to limit the likelihood of incurring losses caused by certain types of risk - but not all risks. The problem with relying solely on the VaR model is that the scope of risk assessed is limited, since the tail end of the distribution of loss is not typically assessed. Therefore, if losses are incurred, the amount of the losses will be substantial in value. 


            Possibly reading more into the story below than is true, it appears J.P. Morgan announced it was junking CoVaR, and readopting its plain, old VaR model, at the moment (one hopes) Ben S. Bernanke was extolling CoVaR's qualities in Chicago:

Front-page headline story in the following day's Wall Street Journal, May 11, 2012:

J.P. MORGAN'S $2 BILLION BLUNDER: BANK ADMITS LOSSES ON MASSIVE TRADING BET GONE WRONG; DIMON'S MEA CULPA

"Fears Deepen Over Risk Model" Financial Times, May 14 2012:  

"It is one more failing in the history of shortcomings for the model. Last week, JP Morgan Chase revealed a major defect in one of its key risk management tools. Instead of helping to predict the surprise $2 billion trading loss announced by the bank, Value-at-Risk had helped disguise the riskiness of JP Morgan's portfolio."

 

"Trading Desks Face Tighter Regulations," Financial Times, May 14, 2012:

 


J.P. Morgan "said it was reverting to an older version of its VAR metric after having switched to a new model earlier in the year."

From the same story: "'How can a hedging strategy turn into a huge trading loss? It doesn't make any sense,' the regulator said."

 


This is not going to end well.