The Bank of England’s incredible stress tests

The Bank of England’s 2016 stress tests are literally incredible. Of the seven big financial institutions covered in the exercise, only one (RBS) failed and that only by a narrow margin.

So what is wrong with the stress tests? Well, one problem is that the Bank used book values instead of market values. It should have used market values (for reasons explained here) and if one uses market values instead of book values, then four of the biggest five banks fail the test.

Another problem is the existence of a lot of hidden leverage associated with positions that do not appear on banks’ balance sheets, such as derivatives positions. The result is that no one can tell from banks’ published financial statements how leveraged the banks really are.

But perhaps the biggest problem is inadequate accounting standards. The weaknesses of IFRS accounting standards have been well-documented: they include the overvaluation of retained earnings, asset values and profits; and inadequate provisions for expected losses. To quote a recent letter in the Financial Times:

better forecasts and better weatherproofing both depend on a deeper problem being resolved: the poor quality of the numbers we are relying on to tell us what banks’ capital actually is. Is the stated “capital” in fact capable of absorbing lending or trading losses that inevitably come in a downturn?

At the heart of the crisis would appear to sit faulty accounts and unreliable audits. In the EU alone, between September 2008 and the end of 2010, more than 300 banks went cap in hand to governments for support—in the form of capital injections, asset relief, liquidity aid or debt guarantees. Few banks [had been] identified as having insufficient capital [prior to September 2008].

All of these banks had previously been signed off as capital adequate by their regulators. That is some regulatory failure.

Nor is it just reported asset values that are the problem. If retained earnings or profits are inflated – and the IFRS rules give bank management give plenty of scope and incentive to game these figures – then inappropriate distributions of dividends and bonuses will be made, which will have the effect of secretly depleting bank capital and inflating reported capital figures – and once again, you cannot tell from the reported figures what the true situation actually is. Indeed, one cannot even tell from the reported figures whether a bank is even solvent.

For more details, see my longer article here.