Following on Kevin’s post, here are a few more thoughts for possible discussion in advance of the stress test results.
1. I think the criticism of the insufficiently adverse assumptions for 2010 has been overdone. While it does raise questions about our ability to forecast the future when we have such a hard time estimating the past, getting 2010 wrong really shouldn’t matter. If BlackRock are taking as hard a look at the balance sheets as we are led to believe, current conditions should already be incorporated in the loan loss estimates. The value of the baseline and adverse scenarios is in understanding how things might evolve from here. In other words, it is the delta from 2010 that matters.
2. A related criticism is that the ESRI/TSB house price index is underestimating the true decline in house prices (the index has house prices down by 38 percent from peak by Q4 of 2010). But again the current state of the housing market should be reflected in the current state of the loan book. If house prices are really down 50 to 60 percent, then the declines under the adverse — or even the baseline — scenario truly get us into Morgan country.
3. There appears to be a “my (preferred) stress tests are more stressful than yours” attitude to the choice of the adverse scenario in the stress-testing exercise. But I think this sometimes reflects a misplaced view of the purpose of the adverse scenario. In addition to seeing what bank losses would be under an adverse (but somewhat arbitrary) scenario, the adverse scenario plays an important role in triggering recapitalisation. In the current exercise, my understanding is that recapitalisation will be triggered if the Core Tier 1 falls below 10.5 percent (please do correct me if this is not correct as the documentation is not that clear). Thus the toughness of the test must be judged by looking at the combination of the adverse scenario and the target ratio under that scenario. A 10.5 percent target is an extremely tough target by any measure. Under the last round of tests, the baseline target was 8 percent and the stress (adverse) scenario was 4 percent. The former has been raised by 4 percentage points to 12 percent; the latter has been raised by 6.5 percentage points to 10.5 percent.
It is noteworthy that IL&P was the only one of the tested banks to fall foul of the 4 percent stress scenario last time round (see here). It is not really surprising that they are in deep trouble with a stressed target of 10.5 percent. This sensitivity to the stress scenario must reflect the importance of mortgages (and in particular buy-to-let mortgages) on the balance sheet of IL&P, and also the large additional assumed declines in house prices under this scenario.