stress tests

Falling Portuguese and Irish bond prices of late will have been hurting whoever is holding them. But who?

If the stress test result data is anything to go by, German banks are the most exposed of Europe’s larger banks. Caution is required here, as the stress test sample excluded medium and small banks, plus this is trading book data and positions will have changed. Caveats in mind, nine of the 14 German banks to be stress tested had exposure to Irish and Portuguese sovereign debt, totalling €4.1bn.

WestLB, Espirito Santo, Santander and RBS were easily the most exposed to the sovereign debt of both countries combined; each had more than €1bn in exposure, mostly for Portugal. Caixa Geral, Credit Agricole, HSBC and SocGen each had more than $750m.

These data should be seen in context: Read more

Basel attendees have effectively raised the tier one capital ratio from 2 to 7 per cent. The package, popularly known as Basel III, sets a new ratio of 4.5 per cent, but also sets a buffer of 2.5 per cent, setting an effective buffer of 7 per cent. (Banks with capital ratios falling between 4.5 and 7 per cent will face restrictions on paying dividends and discretionary bonuses.) Changes will not be entirely phased in till 2019.

Had the stress tests required 7 per cent, rather than 6 per cent, 17 more European banks would have failed under the adverse scenario with sovereign shock (see column T1_AdvS, below). These are almost exclusively PIIGS banks – except for one Slovenian bank, and two German banks (Norddeutsche and Deutsche Postbank). Read more

Breathe easy: Luxembourg’s banks have performed well in a national stress test. The two larger banks, Dexia and KBC, performed well in Europe-wide stress tests earlier in the year, so you’ll be forgiven for having been quite unconcerned about the small state’s banking sector.

The scenarios were concocted a while back, it seems. Of the four shocks, falling property prices or falling EU GDP have the greatest negative impact on the banks’ capital ratios. The good news is that the ratios remain comfortably above 4 per cent in each case. The bad news is that the shocks are independent, and it is more than plausible that house prices would fall and growth reverse at the same time. After all, we’ve seen that before, quite recently. Read more

If a sovereign default had been factored into the recent stress tests, which banks would have failed and how severe would the contagion have been?

Not too many and not too bad, says the Peterson Institute. They used the sovereign holdings provided by 84 of the 91 banks to model an additional shock: namely, a Greek default. Researchers find bank collapses to be relatively limited – Greek banks collapse, predictably, as do those in Cyprus, but: Read more

*Updated to included the originally missing bank, Sunday 10.21am

The results are in. Complaints have already begun that the stress tests weren’t stringent enough, but with the handy table below, you can fashion your own stress test… just pick your preferred Tier 1 ratio and click the column heading to sort. Read more

Confirmation that PIGS banks are the riskiest of the stress test batch, according to credit default swap data – just as banks from two of these countries reveal capital-raising plans.

Sort the stress test banks on the column headings below to see for yourself. Markit has kindly provided five-year CDS data for as many banks as possible, plus the year-to-date change (in basis points). (Quick reminder: the higher a CDS spread, the greater the cost to insure against default, i.e. the higher the market-perceived risk. A 300bp spread means it would cost $30,000 to insure $1m.) Read more

Stress testing could become an institution in itself. Adair Turner has told reporters he expects authorities to move toward annual stress tests, which would be made public, and behind them tests of specific institutions running on a rolling basis. The head of the FSA also repeated his assertion that British banks would fare well under European stress tests, and that those tests would be sufficiently stringent. His reasoning, though, will not comfort markets.

First, Lord Turner describes the European tests as “adequate” rather than, say, “stringent”. And second, his confidence in British banks is founded upon private scenarios run by the FSA last year, which he describes as “more extreme” than those planned by Cebs, Reuters reports. Read more

July 23: stress test result day, right? Wrong. Consolidated results are now to be released on that date, with individual banks’ results delayed by two weeks. Cue a fortnight of frenzied speculation in the markets, as investors try to work out which banks are at the vulnerable end of anonymous bar charts and scatter plots. Presumably this has been done to allow banks more time to raise equity. But then, after two weeks of volatility, even some banks that have fared well will be needing it!

This decision seems to miss the point that stress tests are all about shoring up market confidence. Even bad results can be taken well by markets if investors believe the scenarios are accurate, transparent and really do consider the worst possible situation. Indeed, equity markets rose after the SCAP results from US stress tests last year. But then US markets could have the ultimate confidence in their banks, because the US government had pledged to bail-out any banks subsequently unable to raise funds through the markets. European banks – at this point – come with no such guarantee.

We’ve got details on purpose and sample, but methodology is thin on the ground. See the table: there are two scenarios and at least five variables, but only one of the ten resulting methods is supplied. Sigh.

The aim of the tests are described in the statement:
1) to assess the banks’ ability to absorb further shocks on credit / market risks, including sovereign risks;
2) to assess current dependence on public support (read: can we exit safely yet?) Read more

The data’s finally out. There are 91 banks. More to follow on methodology.