Banking StandardsWritten evidence from Professor Charles A.E. Goodhart
A. Property Finance has been the Main Common Factor
There have been three financial crises in the UK since the 1970s. All of them have centred around property finance, especially bank lending for commercial property, but also mortgage lending for residential property. This was the case in 1973–76 (the Fringe Bank crisis), 1989–91 and 2007–09 (Northern Rock, HBoS). Similarly housing finance was at the centre of the crisis in the USA, Ireland and Spain.
Houses, and commercial property, vary a lot in value, whereas mortgage finance is fixed in value. Hence, during up-turns the leveraged value of the owners rises dramatically, while in downturns the owners get under water. When the latter happens, foreclosure just amplifies the price decline, whereas forebearance makes it difficult to value the bank mortgage asset and leads to criticisms about lack of transparency, etc. It is difficult to reform housing finance in the middle of a crisis, and the American authorities have tried and largely failed. If half of the effort expended on bank reform had gone into housing finance reform, we might be in a better condition. The Danish covered bond mechanism is worth exploring.
The normal perception of bank intermediation is that this involves taking deposits from households to lend to non-financial corporates. This is invalid. As Adair Turner has frequently documented the vast majority of bank assets now take the form of loans to households, mostly house mortgages, and much of the remaining assets are property related. The separation of retail banking, as in Vickers and Liikanen, will reinforce their focus on housing/property finance. Far from making the retail bank safer, by reducing such banks’ diversification the present trend of bank “reform” is helping to guarantee another bank/housing/property crisis some 20/25 years ahead. The present crisis is driving house building well below the demographic underlying build-up of demand. This will lead to a long period, in the medium term future, of rising housing prices. As memories of 2007–09 fade, and they will, people (and bankers and politicians and even regulators) will extrapolate that upwards housing price “trend” into the future, perhaps gaining extra confidence from current banking “reforms”, as the introduction of Basel II, inflation targetry and the great moderation helped to instil confidence in 2003–07.
B. Competition and Macro-prudential Counter-cyclical Policy
When housing/property prices have been rising, it is fairly obvious why borrowers want to increase their leverage, to ascend the rungs of the property ladder. But why do banks let them do so, since such assets are in fixed debt form, and hence subject to tail risk? An important cause of such excessive leverage in booms is competition. Challenger banks, such as Northern Rock and Anglo Irish, ease terms (down payments, etc) aggressively in order to make larger short-term profits and grow into “too-large-to-fail” status. Other, bigger and more staid, banks have to ease terms alongside, or lose market share. Countries that did best in the crisis, eg Australia, Canada, India, maintained a protected, cartelised domestic market. The concept that competition is inimical to financial stability is somewhat unpalatable, but nevertheless true. The great crash and depression of 1929–33 was at the time ascribed to “excessive” competition and the financial reforms subsequently introduced were largely aimed at reducing competition within the financial system, eg by constraining banks’ ability to vary their interest rates.
Everyone enjoys a housing boom. If it was realised that it was “unsustainable”, it would stop on its own accord. Whereas in hindsight it may seem obvious that it was a “bubble”, at the time most people, including most borrowers, most bankers, most commentators and most politicians—and even perhaps most regulators—believed, either in the boom’s continuation, or, at worst, a “soft landing”; “this time it’s different”. Consequently the use of counter-cyclical macro-prudential policy will normally be contrary to majority, conventional wisdom, as it would have been in 2005–06 for example. Not only does it take great courage to defy majority opinion, but, if successful in preventing a crisis, this will be used as evidence that the policy was not needed in the first place! The Bank of England’s FPC has not even asked for the power to vary LTV ratios; it is thought that this is because of concern that public opinion is not ready for its use as a counter-cyclical mechanism.
In this context I have proposed the identification, and use, of presumptive indicators, which, if triggered, require the FPC either to take countervailing action, or to explain in public, why not. Such a “comply or explain” for regulators may stiffen their backbones in a boom. Dealing with financial busts is even harder. Almost by definition a financial bust means that the prior safety barriers had proven insufficient. The inevitable human response to that is that “This must never happen again”, and to reinforce the safety barriers, ie to tighten and toughen such regulatory safety barriers. But to do so in a bust is pro-cyclical, as now. It would be counter-cyclical, post 2008, to lower capital adequacy requirements; instead, of course, they have been ratcheted sharply upwards, with the predictable consequence of greater deleveraging. Can regulatory policy ever be other than strongly pro-cyclical, (as it is now) immediately after a serious financial crisis?
Counter-cyclical macro-prudential policy is likely, for such reasons, to be a weak reed. Instead, there is a longer-run dynamic to financial regulation. Immediately after a crisis it is reinforced into a much tougher straitjacket. This restricts financial intermediation, as it is meant to do. A combination of such restriction, and the effect of the memory of the crisis on risk aversion, keep bankers conservative and crises at bay. So the restrictions come to seem unnecessary and get relaxed. Such liberalisation leads to greater credit expansion, growth and higher asset prices, until the subsequent crisis. It will happen again. This is almost inevitable.
C. Incentives
Let me revert to the question of why bankers allowed leverage to increase so rapidly in the period before 2007. In large part this was because they shared the conventional wisdom of the time which was that lending based on real estate was not risky. There are several papers by Stulz, and co-authors, documenting that bankers, like most others (such as Greenspan) shared a common fallacy. It was, of course, such lending (mortgage based assets) that brought down Lehman Bros; their derivative trading had remained profitable. If one should define residential mortgages, and mortgage-based derivatives, and property loans as part of the natural “utility” functions of a bank, then it was the “utility”, not the “casino” that sank the system.
The logic of this has been accepted by some of the keener structuralists, such as Kotlikoff and Kumhoff. They would divide the financial system into narrow banks, which could hold only perfectly safe (public sector) near cash assets, and be the sole providers of deposits and payments services on the one hand, and fully equity, (or long-dated fixed time deposit), financed investment trusts on the other hand. The reduction in maturity mismatch, and the resultant increase in funding costs, would raise spreads over the interest rates on public sector debt, and lower credit expansion, savings and investment. In a system without exchange controls on capital flows much financial intermediation would move abroad. Even with exchange controls, there would be disintermediation into other channels, eg peer to peer lending. Whether such radical changes would be beneficial or publicly acceptable is far from clear.
A second reason why bankers allowed leverage to increase so rapidly, at the expense of enhanced tail risk, was that they have had incentives to behave in this way. Bankers are responsive to, and largely remunerated in the same way as, shareholders. They generally have bonuses in equity form; most senior bankers have, and are expected to have, a large equity shareholding in their own bank. Equities have limited liability status. The down-side is limited; the up-side is not. This convex pay-off, equivalent to a call option on the bank’s assets, makes the pursuit of risk an attractive way of enhancing one’s own welfare for a banker. To follow such incentives is natural. Given the incentives to raise the return on equity by increasing leverage, the surprise is perhaps that it took them so long to do so.
There are three main strands of proposals for reducing (tail) risks in banking. These are:
(1)
(2)
(3)
Earlier in this note it was argued why counter-cyclical macro-prudential regulation was unlikely to work. The kind of structural limitations on banking necessary to prevent future crises, which the Vickers/Liikanen proposals will not achieve, are (probably) too radical to have any chance of successful acceptance. This suggests that more thought/effort needs to be applied to a reconsideration of bankers’ incentive structures.
This is beginning to happen. Liikanen suggested that bankers’ bonuses be paid in the form of bail-inable bonds, not equities. Sir Martin Jacomb proposed, in an FT op-ed, that the most senior management of a bank be required to accept unlimited liability on their (inalienable) equity position. In a similar vein, one could require any bank employee whose remuneration was beyond some limit to take some minimum percentage of their total remuneration in the form of equity shares, whose particular feature would be that they would be subject to double (or treble, or choose the multiple) liability in the event of default. These could not be sold, until n years after leaving the company, except in the event of death or of inability to meet essential payments (eg divorce) without such sales.
Most of the public’s fury with bankers has occurred because they seem to have continued to receive huge payment despite widespread failure. But given their incentive structure bankers have behaved entirely rationally; indeed given that structure one could argue that most of them have been surprisingly responsible. Now would be a good time to review the remuneration arrangements of senior bankers. No doubt any such measures would lead to an exodus of “stars”. Would that matter?
24 October 2012