Wednesday, August 24, 2011

The Endogeneity of Risk

I was at a risk manager conference and met someone involved in risk capital allocations at a large bank. Interestingly, she said that mortgages now had the same allocation as credit cards. When I was doing this in 1999, credit cards had the highest risk capital allocation within the bank, and mortgages were just above US Treasuries among 'safe' asset classes. Back then, historical loss rates were near zero, and 'underwriting innovations' were just a gleam in Bill Syron's eye.

So, in only 10 years a major asset class transmogrifies from safest to riskiest. The underwriting (credit risk minimums and money down payment), the loss-in-event-of-default assumption (new legal risks), the 'willingness to pay', and collateral price volatility, have all changed significantly. Perhaps this is all endogenous, that a 'safe' asset like mortgages must become risky because everyone--investors, mortgage issuers, home builders, legislators, non-profits--all see it as a vehicle to achieve various ends.

1 comment:

B. A. said...

It sure is, I reckon. The more people rely on the assumption that something is safe, the riskier it eventually becomes (after a period when it all looks like a self fulfilling prophecy, Soros reflexivity thing). I bet gold will be considered a very risky asset soon. I do long dated exotics. Back in the day they were considered much safer than what the prop guys were doing. May well have been so if the books stayed small. People somehow seem to think risk is to do with the particular structure, product, asset class etc. It's more to do with how many people are relying on the same liquidity assumptions as you. The same derivative can be safe if you are 0.1% of the daily volume traded in your underlying, but in some cases we ended up being 50%. The pricing model is useless in this case.