The strange logic behind bank stress tests.

 
 

Bank regulators and indeed economists in general think they face a dilemma, namely that if bank regulations are too stringent, that cuts into bank profits to an unnecessary extent, and consequently reduces GDP. On the other hand, if regulations are too lax, then there is likely to be a repeat of the 2008 bank crises which had very severe economic consequences including loss of GDP.
 
Thus the purpose of bank stress tests is supposedly to determine the minimum level of “stringency” that is compatible with a high level of certainty there is no repeat of the 2008 episode.

Well the first major question mark over that whole idea is that Sir John Vickers, chairman of the main UK  investigation into banks in the wake of the 2008 crisis said that  bank regulations are still nowhere good enough and that it is only a matter of time before we have another 2008. The title of his article was “Storm is coming” and it was published by the Express.

However a more fundamental and theoretical flaw is the assumption that a very high level of stringency, say enough to reduce the chance of a repeat of 2008 from one in fifty in any given year to one in five hundred actually involves any cut in GDP. One reason for saying that is that it is false logic to think that because the activities of one particular type of lender (banks) declines, that therefor GDP declines. Obviously GDP will decline ALL ELSE EQUAL. But the all else equal assumption is badly flawed first to the extent that lending by banks is replaced by lending by other lenders, e.g. mutual funds (“unit trusts” in the UK) and second to the extent that loan based (i.e. debt based) economic activity is replaced by non loan / non debt based activity.

So the crucial question here is: what sort of set up maximises GDP? Is it one where the risk of a repeat of 2008 is possible, but low (e.g. the above mentioned one in five hundred chance), or is it a set up where there is absolutely no chance of such a repeat?

Well it’s widely accepted in economics that GDP is maximised where market forces prevail, unless there are very specific and clear reasons for thinking those forces should not prevail (which is often the case). Now the big problem with the “significant risk of repeating 2008, e.g. the above one in five hundred chance” is that it is quite clearly not a free market set up and for the following reason.

Any depositor who wants a bank to earn interest for them is into commerce just as much as if they deposit money with a stockbroker or private pension fund for the same reason, and it is widely accepted that in a free market those who are into commerce should carry relevant risks. But letting bank depositors earn interest while being protected by a state run deposit insurance system and the possibility of bank bail outs, however remote, is not a free market set up!

So the most genuine free market (i.e. the scenario where GDP is maximised) is one where those who are into commerce get no protection WHATEVER from state run deposit insurance systems or bank bail outs. I.e. the most genuine free market set up is (shock horror) a full reserve banking system where those who fund loans carry all relevant risks.

Do depositors fund bank loans?

Some readers may object to the above paragraphs by claiming they  get too near to saying depositors fund bank loans and by citing a view that has become fashionable of late, namely that depositors do not fund loans.  I actually dealt with that point on this blog recently here.