Winner of the New Statesman SPERI Prize in Political Economy 2016

Saturday 12 April 2014

The IMF on Bank Subsidies

If a bank is too important to fail (TITF), it in effect gets a subsidy from the public. That subsidy is like an insurance contract for those who lend to these banks: if the bank looks like it will fail, it will be bailed out by the government and depositors will get their money back. This in turn means that TITF banks can borrow more cheaply, so they get the benefit of this subsidy every year. TITF banks could do various things with this subsidy: they could make their loans to firms or consumers cheaper (thereby undercutting competition from smaller banks), they could make higher profits that go to either shareholders or as bonuses to bankers themselves, or they could take excessive risks. They will probably do some combination of all of them.

In 2009 the Bank of England calculated the value of this subsidy at £109 billion: that is about £1750 for each person in the UK. The TITF banks of course dispute this figure. (Donald MacKenzie has a very readable account of one example in the London Review of Books.) A week ago the IMF published their own study (pdf), using two different market based methods to measure this subsidy. (The IMF chapter is very readable, but Simon Johnson also has a good summary here.) This is a very imprecise science, but the IMF confirm that subsidies to TIMF banks are very large, although the £109 billion figure quoted above is probably at the upper end of the range of estimates (as the Bank also acknowledged in a later study). However, if we described this number as each member of the public’s contribution to help pay bankers bonuses (which it could well be), I think everyone would agree even a more modest figure is unacceptable.

There are two particularly interesting features of the IMF analysis: it calculates numbers across countries and across time. On the first, some might have assumed that TITF subsidies would be largest in the US, but this is not the case. In dollar terms subsidies in the UK and Japan are of a similar size to the US, and of course the UK is a smaller country, so per capita subsidies are larger in the UK. In dollar terms subsidies appear largest in the Euro area. The IMF also calculate subsidies before the crisis (2006-7), during the crisis (2008-10) and after the crisis (2011-12). The worrying aspect of these calculations is that the subsidies do not seem to have fallen substantially in the post crisis period compared to pre-crisis.

Worrying, but hardly surprising. In principle the TITF problem is fairly easy to solve: as Admati and Hellwig convincingly argue the proportion of the bank’s balance sheet that is backed by equity should be much much higher. (In simple terms, if a bank gets into trouble there are many more shareholders able to absorb losses before a government bailout is required.) The problem of TITF banks is political. As I discussed here, the lobbying power of the TITF banks is enormous. This is not just a matter of bribing campaign contributions to politicians. In the UK there is some evidence that the depth of the recession is partly down to lack of lending by banks, and the bank’s response to any proposals to tighten regulation is to imply that this will ‘force’ them to lend even less. If it is suggested that additional capital could come from reducing bank bonuses, they say all the talent will migrate to overseas banks. Quite simply, the TITF banks have immense power. Until the political will to take on the banks is found, we will each continue to subsidise bank bonuses.

And there will be further financial crises. For those in the UK who think the Vickers Commission put this problem to bed [1], it is essential to read this article by one of its members, Martin Wolf. In reviewing the Admati and Hellwig book, he writes: “Once you have [understood the economics], you will also appreciate that we have failed to remove the causes of the crisis. Further such crises will come.”

Postscript: for more, see this discussion via Mark Thoma.

[1] Because the IMF study tracks estimates of the subsidy to TITF banks through time, it can look at how the subsidy changed when the Vickers report was published (Table 3.2 and Figure 3.8). Publication is associated with a significant fall in the subsidy, but it was not nearly enough to eliminate it.   


  1. Re lobbying, the City spends £90m a year on lobbying. See:

    Re Simon’s claim that “There will be another financial crisis”, Gordon Brown had an article in the New York Times at the end of last year making that point. His article was entitled “Stumbling towards the next crash”. See:

    As to how to dispense with bank subsidies and make credit crunches impossible, that’s easy: implement full reserve banking. Under that system, money which depositors want to be 100% safe is not loaned on: it’s just lodged at the central bank. So no failure is possible from that source, ergo no subsidies are needed.

    As to lending, that is funded by entities or “banks” which are 100% funded by share holders. If those “banks” make silly loans, the banks don’t collapse. All that happens is that the shares lose value. So no subsidies are needed there, plus sudden bank failures and credit crunches don’t happen.

    As George Selgin put it in his book on banking, “For a balance sheet without debt liabilities, insolvency is ruled out”.

  2. We need a global democracy, much bigger capital reserves and greater limits on capital flows especially in regards to secret finance and low tax durestrictions. This will NEVER happen before most of use becone a

  3. The problem is essentially political. Clement Attlee would have made short work of the criminal bankers (and yes, they're largely actual criminals -- look up the fraudulent land transfers where they failed to register the land transfers in England, or failed to file them at the county courthouses in the US). Nationalize the banks, replace the management, arrest the criminals.

    But where is our Clement Attlee?


Unfortunately because of spam with embedded links (which then flag up warnings about the whole site on some browsers), I have to personally moderate all comments. As a result, your comment may not appear for some time. In addition, I cannot publish comments with links to websites because it takes too much time to check whether these sites are legitimate.