“Banking on the State” by Andrew Haldane and Piergiorgio Alessandri is making waves in official circles. Haldane, Executive Director for Financial Stability at the Bank of England, is widely regarded as both a technical expert and as someone who can communicate his points effectively to policymakers. He is obviously closely in line – although not in complete agreement – with the thinking of Mervyn King, governor of the Bank of England.
Haldane and Alessandri offer a tough, perhaps bleak assessment. Our boom-bust-bailout cycle is, in their view, a “doom loop”. Banks have an incentive to take excessive risk and every time they and their creditors are bailed out, we create the conditions for the next crisis.
Any banker who denies this is the case lacks self-awareness or any sense of history, or perhaps just wants to do it again.
The Haldane-Alessandri “doom loop” is fast becoming the new baseline view, i.e., if you want to explain what happened or – more interestingly – what can happen going forward, you need to position your arguments relative to the structure and data in their paper.
For example, at Mr. Bernanke’s reconfirmation hearing, these issues will come up in some fashion. The contrast between the hard-hitting language of the “doom loop” and Ben Bernanke’s odd statements on the dollar yesterday could not be more striking. Still, there is no reason to regard the Haldane-Alessandri version of the doom loop as the final word; in fact, this where the debate now heads. (This link gives as useful introduction to relevant aspects of banking theory, as well as Eric Maskin’s insightful personal take.)
To help move the discussion forward, here are some issues for Banking on the State raised in discussions with top experts (who prefer to remain anonymous):
- The authors say that it is clear, in retrospect, that banks were excessively leveraged. But how did regulators/supervisors miss the implications of this at the time? Banks’ balance sheets started expanding from 1970 onwards (page 3) and by 2000 “balance sheets were more than five times annual UK GDP.” This was not an overnight development – see the last sentence on page 8 which says “Higher leverage fully accounts for the rise in UK banks’ return on equity up until 2007”. It may be difficult for a central banker to come clean on who convinced whom that modern banking in this form is safe – but at a minimum the authors should draw lessons from earlier failures of regulators/supervisors when discussing prospective changes in the framework of regulation. Could some of the changes being proposed suffer the same fate as all previous attempts to regulate big banks? It seems the authors answer is that just moving things to Pillar I (from Pillar II) will help. This sounds like wishful thinking.
- The author are right that US banks faced a leverage ratio constraint, which European banks did not. But US banks circumvented this by setting up SIVs – see the damage at Citi for details. Again, what were the regulators/supervisors thinking when they allowed this?
- The authors assume that the equity owners of banks are almost always protected and therefore “the rational response by market participants is to double their bets”. This does not seem to have been true in practice. For example, why was it so difficult for banks to raise capital after the initial flurry of new capital from Sovereign Wealth Funds (SWFs)? Why did some banks share prices fall so much (Citi, Merrill Lynch, Morgan Stanley, etc)? This cannot not be characterized as a rational response by markets if equity holders were implicitly protected. In fact, new capital (either from the state, or even in some cases from SWFs) came in the form of (expensive) preferred stock and diluted existing holders. The doom loop is surely more about what happens to insiders (rich and powerful bank executives, with strong political connections) and creditors (investment funds run by rich and powerful nonbank executives, with strong political connections).
- Part of the (relatively) reasonable performance of hedge funds was due to them being forced quite early on to reduce leverage and asset holdings because banks were short of capital and tightened lending conditions. This fortuitously allowed hedge funds to reduce exposure before the crisis became most acute. Haldane and Alessandri seem a little too inclined to believe the hedge funds’ own rhetoric at this stage. This is worrying – the intellectual origins of our last crisis lie with central bankers believing that the private financial sector has evolved into a safer form.
- To be clear, and a little contrary to what the authors imply: Most hedge funds do not operate with unlimited liability. Often they have “watermark” provisions, limiting their fees while the fund shows losses. But it is a simple matter to close down a failing fund and, a week or so later, open another (how many funds has John Meriwether closed?). This will feed the next doom loop.
- The private sector is unlikely to be able to self insure (e.g., various proposals discussed on page 18) because of the potential size of losses in a systemic event. We know there was private insurance for a large portion of the assets (CDOs insured through monolines, for example) but these insurers did not have credible resources. Similarly, implicit state guarantees may also not be sufficient (e.g., Iceland). This suggests strict controls on size of the financial system relative to the economy (and the tax base) may be necessary.
- The paper is also relatively weak on the role of monetary policy in fuelling the doom loop. But that is relatively easy to add on.
The overall conclusion of the paper follows uneasily from the main analytical thrust. How can we believe that for the regulators, “next time is different“? Most likely, next time will be exactly the same, with different terminology: the financial sector “innovates”, regulators buy their story that risks are now properly managed, and the ensuing bailout (again) breaks all records.
It’s all politics. Unless and until you break the political power of our largest banks, broadly construed, we are going nowhere (or, rather, we are looping around the same doom).
Barney Frank points out that small banks have political clout also, and of course he’s correct that this drives some issues. But how many small banks spend their time (and lobbying dollars) on Capitol Hill insisting that large banks must not be broken up?
Our core problem is that we now have banks that are Too Big To Fail; if you don’t agree, read and publicly refute Haldane. In theory, these big banks could be effectively regulated, but this is a leap of faith that experienced policymakers (e.g., Mervyn King and Paul Volcker) are increasingly unwilling to make.
The biggest banks must be broken up. This is not sufficient to end the doom loop, but it is necessary.
By Simon Johnson