Friday, May 29, 2009

main one is that the UK does not disclose results, nor the detailed assumptions in terms of loss rates

TO BE NOTED: From Alphaville:

“The US and the UK stress tests have two very different objectives”

BNP Paribas has come out with an analysis of the UK’s methodology for stress-testing its banks. Conclusion? Meh.

Analysts Olivia Frieser and Andrea Cicone concede the test is a “real one”, in so far as the parameters - 12 per cent unemployment, 50 per cent fall in house prices and a six per cent decline in GDP from peak to trough - are reasonably demanding:

The assumptions seem severe enough to us and therefore the stress test seems real

However, they note:

if we were picky, we would say that they remain static tests, and that these assumptions are merely in line with the base case of our admittedly bearish economists.


In blunt terms, some disclosure is better than none, but we still feel we do not have enough disclosure to feel that UK banks will not need any additional capital in coming years.

Frieser and Cicone currently have a ‘reduce’ recommendation on RBS and Lloyds Banking Group.

As for how the UK’s methodology compares with the US, they highlight two main differences:

The main one is that the UK does not disclose results, nor the detailed assumptions in terms of loss rates, while the US has provided plenty of information in this regard, down to the individual bank holding company level.

The UK stress scenarios look sufficiently “stressing”, whereas even the more adverse scenario used by the Fed was fairly generous for the banks, being more benign than our own forecasts.

These differences, they say, confirm the fact that the US and the UK have two profoundly different objectives (emphasis ours):
the UK one is a tool used by the FSA and Treasury to truly assess the possible cost of the APS programme (hence the more stressing scenario, the “confidentiality” of the results, and the fact that is an ongoing practice);

on the other hand, the US one was fundamentally a confidence-building exercise for the benefit of investors (hence the more accommodative scenarios, the disclosure of detailed information about the results, and the one-off nature)

In other words, one’s a model and one’s PR - but that may just be FT Alphaville’s typically jaded take on the matter.

Related links:
Stress testing, a retrospective view - FT Alphaville


Don the libertarian Democrat May 29 15:48
You're correct about what the Stress Tests became in the US, but they should have done what the UK has done. The confidence in CAP should have followed from clear government guarantees.

No comments: