An article in the New York Times quoted by Barry Ritholtz on risk management tells some very salutary lessons.? Value-at-Risk, and other systems that rely on liquid tradable markets fail when bid-ask spreads widen.
One of my main lessons on risk comes from the concept of “bicycle stability versus table stability.” As I said at RealMoney: “This emphasis on the size of monthly payments to the consumer reminds me of the 1920s. We have traded table stability for bicycle stability. A bicycle is stable if it continues to move forward; a table is stable regardless. In an effort to ‘lock in’ housing prices in markets that are rising rapidly, many people are doing things that are rational in the short run, but not necessarily in the long run.”
I’ve talked about the the difference between bicycle and table stability before at this blog, notably:
- Bicycle Stability Versus Table Stability
- Update on Indicators (Bearish point #1)
- On Information Cascades and Lemmings
If your risk control methods require liquidity, then they won’t work when you need reduction of risk the most.? There are no free lunches in risk management.? In order to get “table stability” leverage has to be reduced, and in some cases, cash balances carried in order to assure that an enterprise can survive in all circumstances.? It implies a lower ROE in good times, in order to be sustainable in the worst times.
My (and others) current hypothesis, which still needs vetting, is that during this crisis banks are buying treasuries (including with TARP monies) trying to recapitalize, and put more credit worthy assets on their balance sheets. If treasuries are indeed “in a bubble”, then if that market drops, do we have another round of marking to market? I guess at least there is readily available pricing for treasuries.
David-
I absolutely love the Bicycle versus Table stability concept. It is elegant in its simplicity.
Manhattan Real Estate has wide bid ask spreads.
One final note here — financial risk models correspond to bicycle stability, while cash flow driven actuarial models correspond to table stability.