Why Does Goldman Need A Fed Exemption For VaR Calculations?

[Joanne98]

Posted by Tyler Durden at 8:27 PM
Lately the topic of Goldman's VaR has taken on significant prominence, not least because as Zero Hedge disclosed yesterday, it hit a record high. The implications for this were large enough that even Bloomberg picked up on this story. Many readers raised questions of how is it even remotely possible for the company to have a VaR in the low-mid $200 MM ballpark, yet to post a record number of $100MM+ trading days in Q1; Zero Hedge is willing to wager that the upcoming 10-Q release will demonstrate another record number of $100MM+ days in the just closed quarter as well.How is that possible?

The clue may come from a February 5 letter by the Federal Reserve to Goldman CAO Sarah Smith. The letter had come in response to GS requests for "temporary exemptions from the application of certain aspects of the Board's Market Risk Rules for state member banks and bank holding companies and the Board's general risk-based capital rules for bank holding companies." Basically through the end of 2009 Goldman is basically using non-traditional. SEC approved VaR models as can be seen here:

GSGI has requested that (1) through December 31, 2009, GSGI and Bank be permitted to use certain Value-at-Risk ("VaR") models approved by the SEC... to determine their capital requirements for specific risk under the Market Risk Rules; (2) through December 31, 2009, GSGI and Bank be permitted to use methods approved by the SEC to determine their capital requirements under the Market Risk Rules for those trading assets, including distressed debt and restricted stock investments that the SEC did not require to be included in the VaR-based models of GSGI and Bank; and (3) GSGI be allowed to use methods approved by the SEC to calculate risk-based capital requirements for its nonfinancial equity investments that are subject to the Board's Credit Risk Capital Rules.
The letter goes into detail explaining why a bank needs to follow a MRR VaR methodology. Yet what is not made clear is i) why does Goldman need almost a full year of alternative VaR calculation and MRR exemption and ii) what is the protocol for the SEC to enforce VaR compliance when Goldman's ultimate regulator is the Federal Reserve. The exemption raises critical questions not only with regard to the validity of the company's indicate VaR, but also downstreaming capital requirement reports. Zero Hedge would be remiss to point out that a very close relationship between the most critical financial company in the world and the most discredited regulator (SEC) does not bode well for confidence in this critical risk indicator, which as many have pointed out, is clearly the main metric by which to measure not only the performance, but the risk capacity of the world's largest government-backstopped hedge fund. Mr. Canaday, Mr. Van Praag - the floor is, again, yours. In your absence, Zero Hedge will, and encourages it readers to, contact Mr. Homer Hill at the Federal Reserve Bank Of New York at [phone] to provide additional clarity on the matter

http://zerohedge.blogspot.com/2009/07/why-does-goldman-need-fed-exemption-for.html

[leveymg]

1. What would be "normal" VaR and minimum reserves for GS?

You mean Goldman is playing fast and loose with its trading risks? Again? Do they expect to be bailed-out again?

[econoclast]

2. It could be that VaR is "very precise ... but precisely wrong"

I urge DUrs to google Andrew Haldane Bank of England

Find and read his Feb 09 speech callers Why Banks Failed the StressTest

VaR's failure is a major reason

he says in part. ... VaR is very precise but it is better to be roughly right than precisely wrong.

[Joanne98]

3. Welcome to DU econoclast. Here's the link to that.........

http://www.bankofengland.co.uk/publications/speeches/2009/speech374.pdf