Risk management for fixed-income trading

As the fallout and after-shocks of this Summer's mini-crisis in credit markets continue to unfold, there is one term that keeps popping up: risk management. It is easy to understand how coping with risk can be very tricky with inherently complex investments such as venture capital, distressed debt, or corporate bailouts, but how on earth can risk management for fixed-income trading be so difficult and how badly can you really screw it up even if you wanted to? Somehow, even the biggest and "best" banks on Wall Street were unable to manage the risks of their own fixed-income trading.

This morning I was reading an article in The New York Times by Eric Dash and Landon Thomas Jr. entitled "The Man in Citi's Hot Seat" which tells us that:

But the $600 million loss in fixed-income trading still raises questions about the bank's controls. Risk models did not malfunction, Citigroup managers say, and Mr. Prince has called the trading losses an aberration. He has commissioned a report to learn what mistakes were made.

Hmmm... they are saying that the risk models failed to predict or mitigate huge losses but in the same breath they are saying that those models did not "malfunction." Exactly what distinction are they trying to draw between failure and malfunction? A loss is a loss, right? One would expect that all failures are aberrations, right? Well, apparently not. I guess the idea is that any given model only factors in a specific subset of risks and "malfunction" does not apply to "risks" which were not explicitly included in the models and in fact may have been explicitly excluded. Sounds like a typical insurnace policy.

I read this amusing piece of logic:

Citigroup executives concede that $5.9 billion is a substantial number but say they are certain that they properly managed risk...

What??? I wonder how they are defining "risk" that they can claim with a straight face that a $6 BILLION loss is well within the range of expected "proper risk management." Wow, what is wrong with this picture!

Now, in truth, I myself cannot even imagine what kind of believable "risk model" could ever be developed which would realistically encompass all forms and ranges of risk which an investment manager needs to be prepared to encounter over say a fifty-year time horizon. I'm no Einstein, but even a person of relatively average intellect should be able to recognize when a "model" is promising to deliver results beyond comprehension.

There is clearly a level of intellectual dishonesty here, or maybe the lawyers in fact have some hidden fine print that actually says that the risk models are for "information only" and should not be construed as actual "investment advice."

On top of all of this are two basic questions:

1. How can fixed-income trading be so risky?
2. Why are big Wall Street banks engaged in fixed-income trading in the first place?

The latter is the elephant standing in the middle of the room that nobody wants to talk about.

Just to be clear, when we (or the banks) talk about "fixed-income trading", we are not talking about the banks acting as brokers for customers wishing to buy or sell bonds. Rather, we are talking about proprietary, inhouse trading desks that are using the banks own money to place bets. The bank itself is betting on the movements of interest rates and supply and demand for debt securities. Whereas a true investor customer will be making investment decisions on the yield and risk of the debt security, the bank is seeking to buy and sell debt securities on a short-term basis based on their perception of change of market prices with very little regard to the actual yield or even the actual credit risk (since the security is not likely to be held long enough to be exposed to the actually credit risk events.) In most cases, they are simply trading off of the normal volatility of prices, with volatility further exacerbated by all of the short-term trading of the banks themselves. Like hedge funds, they will also engage in heavily-leveraged trades. For all intents and purposes, these inhouse trading desks are effectively hedge funds.

The real problem is not that there is a high risk of loss on the average fixed-income trade, but that so many of them are low-risk and are very profitable, so that the banks are lulled into a sense of complacency by their own great past success.

One technical problem traders encounter is that as everybody runs up the same learning curve and profit opportunities with short-term trades get skimpier, traders begin "reaching for yield" and gradually turn into medium-term speculators or even longer-term investors and more highly-leveraged where they less-skilled and are now exposed to things like actual credit risk events (e.g., bankruptcies and foreclosures) and illiquid markets. The beauty of short-term trades is that they can be quickly unwound for only modest losses. The ugliness of medium and longer-term "trades" is that they cannot be unwound quickly if at all and losses can be very large. The lesson here, as it has always been, is that "reaching for yield" really sucks as a strategy from a risk management perspective.

In truth, maybe the big banks are simply not being honest and upfront about risks and those risk models. I suspect that there is more than enough statistical data available that would have told bank executives that a blowup and loss of this magnitude can be expected every once in a while and that the proper attitude is to amortize such losses over the vast number of trades that were very profitable over a number of years. If that is their thinking, they certainly haven't been upfront about it.

My point is not that banks are unable to cope with these risks, but simply that it is highly inappropriate for such "pillars of the community" to be taking such risks in the first place. Sure, allow banks to engage in limited fixed-income trading to the extent that it is necessary to facilitate the transactions of the customers of the bank, but this concept of proprietary inhouse trading desks is a wild distraction from the true goal of providing banking services. If executives want to run a hedge fund, let them leave do so out side of the bank, not under the umbrella of "banking."

-- Jack Krupansky