This was the question put to me by [UK] Treasury Committee Chairman Andrew Tyrie MP when I appeared before the Committee on January 6th to give evidence on the Bank of England's latest Financial Stability Report.
This is a question to which many of us on our side have given much thought and I believe it to be the single most important question in the whole field of bank regulatory policy.
I was nonetheless caught off-guard when Mr. Tyrie asked it at the beginning of the session – I was expecting questions on the Bank's latest nonsense, the results of its new stress tests – and my initial response was less than it should have been. But no excuse: it was a perfectly reasonable and entirely foreseeable question – the obvious question, even – and I still didn't see it coming. Reminds me of the blunders I would occasionally make when I played competitive chess: I obviously haven't improved much.
Thankfully, he asked me the same question again at the close of the session, and his doing so allowed me to give the correct answer clearly, an emphatic 'No'. However, by this point there was no time to elaborate on the reasons why a bank in difficulties should be denied assistance.
These reasons go straight to the whole can of worms and my follow-up letter to Mr. Tyrie should, I hope, help to set the record straight.
My message to other advocates of free markets is that leaving aside the usual bailouts-are-bad stuff, we really should give more thought to what an Armageddon Plan B might look like: Yes, no bailouts would be best, even in our intervention-infested system, but in that case why do we humour lender-of-last-resort and, more to the point, if the government is even considering intervention in what it (rightly or wrongly) sees as an emergency in which something-really-ought-to-be-done-NOW, then what should we advise it to do – other than 'Don't'?
Mark my words: if we don't give the government constructive advice, it will do what it always does when a crisis breaks out: it will panic and the chances of any sensible policy response will be zero.
So here is the text of the letter, dated January 12th:
"Dear Mr. Tyrie,
I would like to thank you for the opportunity to give evidence to the Treasury Committee at its meeting on January 6th.
At that meeting you asked me if the authorities should assist a bank that gets into difficulties.
My answer is 'No' but I should like to elaborate.
Consider first a free or laissez-faire banking system in which there is no central bank, no financial regulation and no other state interventions such as deposit insurance. In such a system, competitive pressures would force the banks to be financially strong; bankers who ran down their banks' capital ratios or took excessive risks would eventually lose their depositors' confidence and be run out of business, so losing their market share to more conservative and better-run competitors. Bankers themselves would have serious skin in the game and therefore have strong incentives to keep their banks sound: for them, bank failure would be personally costly. Banks would then be tightly governed and conservatively risk-managed, and the banking system as a whole would be highly stable.
There would still be occasional failures due to the incompetence of individual bankers, but these would be few and far between, and not pose systemic threats.
These claims from free-banking theory are broadly confirmed by the historical experiences of the many free or loosely regulated banking systems of the past, most notably the experiences of Scotland pre-1845 and 19th century Canada.
In such a system, there is no good case for official assistance to any bank in difficulties. A bank failure would be painful to those involved, but the possibility of bankruptcy is unavoidable in any industry in a healthy capitalist economy, and this includes the banking industry. Letting a badly run bank fail also sends out the right signals – it encourages other bankers to avoid the same mistakes, it encourages depositors to be careful with the banks they choose and it avoids the moral hazards inevitably created by any policy of assistance.
Modern banking systems differ from these systems because of the presence of extensive systems of state intervention, including a central bank, a central bank lender of last resort function, deposit insurance, capital adequacy regulation and other forms of financial regulation. In different ways, each of these interventions makes the banking system less stable: central banks through erratic and usually loose monetary policies, which create inflation and fuel asset price cycles, and generally destabilise the macroeconomy; the lender of last resort and deposit insurance by creating moral hazards that lead to excessive risk-taking by bankers; capital regulation by creating short-termist incentives for banks to reduce their capital (e.g., by playing games with risk models and risk weights); and financial regulation generally by its large compliance costs and its stifling of innovation. Over time, these interventions have made the banking system weaker and weaker, even though their usual stated intention was to strengthen the banking system rather than to weaken it.
However, even with the banking system already seriously weakened by a long history of misguided government interventions, the best policy response is still to refuse assistance to banks in difficulties. I say this for two main reasons:
• the systemic effects of bank difficulties tend to be exaggerated even in a systemic crisis, sometimes grossly so; and
• interventionist policy responses tend to make matters even worse.
The ideal response by policymakers is to refuse assistance point-blank – and to announce such a policy in advance so the bankers know where they stand.
Policymakers should follow the advice of Lord Liverpool, who was PM at the time of the last systemic banking crisis pre-2007, that of December 1825. In May that year, he foresaw the looming crisis and warned the House of Lords about the "general spirit of speculation, which was going beyond all bounds and was likely to bring about the greatest mischief on numerous individuals." He wished it to be "clearly understood" that those involved "entered on their speculations at their own peril and risk" and he thought it his duty to declare that he would "never advise the introduction of any bill for their relief; on the contrary, if any such measure were proposed, he would oppose it" and he hoped Parliament would reject it.
In our current system such a response would require political leadership with uncommon vision and nerves of steel. When the next crisis occurs, it will explode unexpectedly, taking policymakers off guard. They will be under extreme pressure to respond quickly – probably within hours – on the basis of inadequate information, whilst bankers lobby intensely for immediate assistance: if we don't get bailed out, the world will end, etc., the usual scare mongering. Under such circumstances, it would be extremely difficult for even the best political leadership to avoid being dragged into making the same mistakes made repeatedly in previous crises.
These mistakes include:
• panicky rescues, which are later shown to be unnecessary, ill-judged and in some cases illegal;
• the abandonment of previous 'commitments' to let badly run institutions fail;
• bankers being rewarded for their failures by being made personally better off than they would have been had their banks been allowed to fail; and
• more regulation or regulatory reshuffles accompanied by the usual empty promises that 'it' won't happen again, made by the very people who had no idea what they were doing when they were in charge the last time round.
So how can we avert such outcomes? A good start would be an Act to prohibit future assistance: as much as possible within the confines of our constitution, we should seek to tie the government to the mast. "Much as I would like to help you", the PM can say, "my hands are tied."
But even with this Act in place, there is still the difficult question: if the government does respond to the next crisis, then what should it do?
To that question I would propose a publicly disclosed Plan B, whose main features would include:
• a programme to keep the banking system as a whole operating at a basic level to prevent widespread economic collapse;
• fast-track bankruptcy processes to resolve problem banks and, where possible, return them to operation as quickly as possible;
• a prohibition of cronyist sweetheart deals for individual banks or bankers;
• provisions to ensure that senior managers of any failed banks are made strictly liable to severe personal financial penalties;
• a holding-to-account of senior bankers, regulators and policymakers, including the opening of criminal investigations into the activities of any banks that fail;
• the establishment of a legal regime that imposes high standards of personal liability on senior bankers;
• the restoration of sound accountancy standards; and
• a radical programme to deregulate the banking industry.
This programme would include the abolition of the current regulatory structure including the PRA and FCA, the ending of deposit insurance, the UK's withdrawal from the Basel system of capital regulation, and the reform (and preferably, abolition) of the Bank of England. These reforms would rein-in the out-of-control moral hazards that permeate our current banking system and restore the personal responsibility, tight governance and conservative risk-taking that are the keys to a sound banking system.
Contingency planning for the next crisis should also provide for only two possible responses by the authorities: either Plan A (i.e., do nothing) or Plan B as just set out. Any intermediate response should be prohibited, as that would merely open the door to the usual mistakes that the authorities are prone to make in such circumstances.
In short, in response to your question about whether a bank should receive assistance, my answer would be 'No', but if we are to avoid another bungled policy response when the next crisis occurs it would be wise to have a credible Plan B in place to address upfront the Armegeddon scenario of a possible systemic collapse. And if it does intervene, the government should use the opportunity to clean up banksterism once and for all and restore a sound banking system based on the principles of personal responsibility and laissez-faire.
Durham University/Cobden Partners [etc.]"
There is a lot more to say on this subject, but one of the points that emerges most clearly for me is the pressing need for free-market narratives of the financial crisis, blow-by-blow accounts of how it should and might have been. In this context – and off the top of my head – I would particularly recommend the following (with apologies to those whose work I have overlooked):
John A. Allison, The Financial Crisis and the Free Market Cure, McGraw-Hill 2013, esp. chapters 14-17.
Richard Kovacevich, "The Financial Crisis: Why the Conventional Wisdom has it All Wrong", Cato Journal Vol. 34, No. 3 (Fall 2014): 541-556.
Vern McKinley, "Run, Run, Run: Was the Financial Crisis Panic over Institution Runs Justified?" Cato Policy Analysis 747, April 10, 2014
George Selgin, "Operation Twist-the-Truth: How the Federal Reserve Misrepresents its History and Performance", Cato Journal Vol. 34, No. 2 (Spring/Summer 2014): 229-263.
These are all US-oriented of course and we badly need to work on similar narratives for the UK, Ireland and Europe.
But going back to the Treasury Committee, most of the discussion was on the regulatory risk models – or more precisely, on what is wrong with regulatory risk modelling and in particular, the Bank's stress tests. I have to say, too, that I was greatly heartened to see the skepticism of the MPs towards the models and their openness towards our ideas, much of which is obviously down to the pathbreaking work that Steve Baker is doing on the Committee. But let me come to all that in another posting.
[Slightly adapted from a blog published on the Cobden Centre website, January 20 2015.]