Do UK Banks Need More Capital?

Yes.  The Bank of England’s November 2012 Financial Stability Report has made the case for UK banks to shore up their levels of capital, perhaps to the tune of up to £20-50bn according to the FT.  Three arguments seem to have made: (i) banks may have to pay fines because of “misbehaviour”;  (ii) provisions for expected losses need to be made and (iii) risk weights may not be appropriate, in the sense that the weights applied may under predict risk.  Rather predictably commercial banks have complained about the new requirements arguing that they want more certainty about capital targets.  I am not quite sure that financial institutions, whose basic role is to price payoffs in uncertain future states of nature and allow people to write contracts for smoothing consumption and investment decisions over those uncertain futures, should be calling for certainty otherwise their very rationale seems to me to disappear in a puff of full information.

But more importantly, in a recent talk, I drew upon a chart published in the June 2012 Financial Stability Report (see below, p 6), that showed a pleasing movements towards some reduction in funding costs for more well capitalised banks.  That the market might now be thought to be punishing more risky banks – measured by their level of capital – means that market scrutiny of bank balance sheets may be starting to re-enforce plans for macro-prudential instruments.   So rather than an unwanted deadweight loss, or simple compliance with regulatory demands, there may be direct market benefit for financial institutions that move to more sensible levels of capital.     

One clear reason why the risk held by bank might be under-reported is that under European convention, sovereign bonds are given a zero risk weight in any calculation of risk-weighted assets.  This means that commercial banks do not have to hold any capital against these assets but which nevertheless, in reality, do carry significant risk.   This means that there is a fundamental shortfall in capital held in proportion both to the quantity and actual risk of sovereign debt. 

In fact, I examined the capital shortfall of the banking system in two ways.  Firstly, using the Eurozone Stress tests of late 2011 (Source: EBA, Eurostat, European Commission and Reuters Breaking Views), I examined the losses of all EMU and non-EMU banks in the EU from a 50% haircut (debt forgiveness, essentially) in Greek and also all Portuguese, Italian, Irish, Greek and Spanish Debt under scenarios for a 7% or 10% capital ratio target.  The capital shortfall in the case of only a 50% Greek haircut and a 10% capital target was some Euro348bn, for the UK alone this was a shortfall of Euro72bn.   Obviously some of the required adjustments may have already been made in the last year but such a calculation may not deal with further losses induced in the system, as banks may incur losses to other parties under a generalised haircut.  (These results are reported in The Euro in Danger.)  Secondly, and more recently, I then looked at the loss in terms of GDP from a 75% haircut on Greek debt and 30% on Ireland, Italy, Portugal and Spain.  With the same 10% capital target, we get a shortfall of Euro75bn in the UK, which rises to Euro221bn if the capital target is 15%.  Now while I think it is unlikely that increasing bank capital to extraordinarily high levels will be necessary or sufficient and much work remains to work out what of the optimal amount for each bank and in the system as a whole, this simple calculation seems to imply that the risk of continuing problems in the Euro Area would seem sufficient to justify the extra call for capital of this magnitude.