Recent Items

Toothless Fed, Part 2 (Risk Management Shortcomings)

Posted on by

Forgive us if we seem to be picking on New York Fed president Timothy Geithner. Actually, not that we know him, but he has a reputation (by Fed standards) for candor. So the problems we have with his speech should not be seen as an attack on him, but on the increasing difficulty of the Fed’s position.

We are beginning to suspect the Fed has no clothes, and other people are starting to think that too.

First we have Mohamed El-Erian, president of Harvard Management Corporation, saying in a Financial Times article that the Fed’s last 17 rate hikes have been ineffective because “liquidity factories” have created more money than the Fed has withdrawn, and the only way for the Fed to put the brakes on the leapfrogging of asset price inflation and leverage is by raising rates enough to slow growth.

Then we have the unfortunate confluence of Geithner’s speech last Friday, in which he described the efforts of the Fed to encourage better risk management procedures, with an unfavorable assessment of their effectiveness in a study released Monday by Deloitte (cited further below):

Even the most sophisticated participants in the markets for these instruments find the risk management challenges associated with these instruments daunting….The dramatic growth in the volume of over-the-counter derivatives and the growth in the number and size of leveraged funds inevitably complicate the resolution of the failure of a large financial institution….

These issues are the principal focus of day-to-day supervision and market oversight in the major financial centers around the world. The Federal Reserve is actively involved in a range of efforts, working closely with the primary supervisors of the major global financial institutions and the critical parts of the financial infrastructure, to encourage further progress….We are encouraging more sophisticated and more conservative management of credit exposures in over-the-counter derivatives and structured financial products, as well as of exposures to hedge funds.

Unfortunately, the Finacial Times reports today in its story, “Risk of derivatives ‘not fully evaluated’” cites a Deloitte study that in effect gives the Fed and other regulators bad marks in their campaign to get financial institutions to manage their own risks:

Fewer than half of global financial institutions account sufficiently for complex financial and commodity exposures in assessing the riskiness of their holdings, according to a survey by Deloitte.

The results suggest that much of the industry may “lag behind the explosive growth of credit derivatives and their attendant risks”, the firm said. Risks in energy and other commodity derivatives were also not fully evaluated by most institutions, the survey found.

“The bar on what constitutes effective risk management is constantly being raised. Most institutions have an unfinished agenda,” said Owen Ryan, head of Deloitte’s capital markets practice.

The firm’s risk management survey involved 130 mostly global institutions, primarily commercial and retail banks and diversified groups, with total assets of $21,000bn….

The Deloitte survey showed that only 41 per cent of executives reported using value-at-risk (VaR) models to cover credit derivatives, with fewer than half of them using VaR analysis extensively. VaR is a measure of the probable losses on a portfolio if historically big – but not extreme – market moves occur.

Fewer than half of respondents regularly used “stress-testing”, a technique recommended in reports on systemic risk. Stress-testing aims to estimate losses in a severe stock market crash.

If you know anything about risk management, this report is even worse than it sounds. VAR is a relatively old risk management approach (at least 10 years old, which is ancient). In crude terms, it looks at historical price movements of instruments, then subjects the portfolio to price movements that fall within the 99% probability range. Now there are problems with VAR. The first is that securities prices aren’t normally distributed, they are subject to “fat tails,” meaning extreme movements occur with more frequency than in a normal distribution (you can presumably adjust for that by pushing your VAR out a little further than 99%). A bigger problem in this context is how do you even know what price movements to assume for instruments that have been trading for less than 5 years? You don’t have a large enough data sample to make informed judgments.

But the scary thing is that VAR is pretty basic. For nearly half the firms not even to be using VAR says they are operating by the seat of their pants.

Stress testing is another broadly used technique (well in theory broadly used, the Deloitte findings suggest otherwise) that is very valuable, but less standard in its application (there is more art and customization in how you define the inputs). But again, it is very worrisome that a large proportion of firms aren’t using this method.

Now if they were using more sophisticated methodologies they had developed on their own, that would be very good news. But there is no indication from Deloitte that they are doing that.

And if the big financial firms that are subject to regulatory oversight are operating like this, one can only imagine how things are at hedge funds.

This outcome is an indictment of the posture the Fed has taken towards new financial instruments (and I wish I could point to something in print, but I voiced my worries privately at the time). The Fed, rather than prescribing any mechanisms relative to derivatives (operational requirements, reporting requirements) that could have forced implementation of a minimum level of risk management standards, instead took the “let a thousand flowers bloom” approach. They did stick their noses in what member banks were doing, but took the posture of observers. I had the sense at the time that they didn’t fully understand the instruments or exposures, but weren’t willing to admit that, nor to hold up “innovation.” And remember, they only regulate banks, so their subjects could have argued that a heavy regulatory hand by the Fed was ceding the derivatives business to securities firms (of course, the Fed could have tried to reach some sort of rapprochement with the SEC, but that would have taken time, and both regulators would have faced intense lobbying). Yet there was no change in posture even after the LTCM near-disaster, which was seen as a unique event (we’ve given some thought to that…)

But even this last bit gets back to the basics: financial regulators have been having trouble regulating for some time. If you have any doubts, read this article, “How the SEC Was Handcuffed,” a review of Arthur Levitt’s book “Take On the Street.” If Levitt, someone who knew a good deal about the securities industry (he had been president of Shearson Hayden Stone, then head of the American Stock Exchange) and a bit about Washington, was stymied under the Clinton Administration, you can imagine how hard it is for regulators to do their job now.

Print Friendly


  1. Anonymous

    Absolutely agree that the VAR is not sufficient, here is a nice example I saw on Bloomberg.

    On Tuesday night, Bloomberg Television, the London-based financial channel, said that Merrill Lynch found that using the Value at Risk (VAR) model, required by the supervisors, had calculated its maximum exposure equal to $65 million.

    Instead, it lost $8 billion.

    Yesterday, the president of Merrill Lynch gave a lecture, saying that VAR is not able to calculate market risk exposure, and it is incapable of estimating credit risk.

    This risk has been identified in this book from 2004 book Economic Capital Allocation (Chorafas). This book also gave advice on how to develop and use a better model than VAR.

Comments are closed.