FRSGlobal comments on the problem with the US stress tests

11 May 2009

Andrew Liegel, risk and regulatory specialist at FRSGlobal, commented on the US banking stress tests and how historical comparisons fail to address the problems of transparency and leverage in today's market:

The results from last week's stress tests (Supervisory Capital Assessment Program – SCAP) were largely unsurprising, but the regulators did release what baseline loss estimates had been given to banks.

Perhaps the biggest criticism is that the two-year loss projections may not be as relevant as the regulators think they are. The losses range from a low of 3-4 per cent for a baseline of consumer and industrial loans to a high of 21-28 per cent for the more adverse projection of subprime loans. The regulators use the baseline of a 9.1 per cent loss rate, implying that this is adequate given that even in the Great Depression of the 1930s, the highest two-year cumulative loss rate was less than 9 per cent.

The main problem with the SCAP analysis is that is focuses too much on aggregated historical loss data and not enough on the historical loss data relative to the changes in the composition of the financial contracts on which the losses were based.

If most of the loans and mortgages in today's market had the same characteristics of the 1930s (interest rates, no collateralisation, similar Loan-to-Value ratios), then better historical comparisons could be drawn. However, by doing a simple historical comparison, regulators are failing to address the problems in the leverage and transparency in today's contracts. A quick historical comparison will not fix this anytime soon.

Become a bobsguide member to access the following

1. Unrestricted access to bobsguide
2. Send a proposal request
3. Insights delivered daily to your inbox
4. Career development