January 16, 2008
Why regulators should intervene in bankers’ pay
By Martin Wolf
"You really don’t like bankers, do you?” The question, asked by a former banker I met last week, set me back. “Not at all,” I replied. “Some of my best friends are bankers.” While true, it was not the whole truth. I may like many bankers, but I rather dislike banks. I recognise their necessity, but fear their irresponsibility. Worse, they are irresponsible partly because they know they are necessary.
My attitude to the banking industry is not a prejudice. It is a “postjudice”. My first experience with out-of-control banking was when I watched the irresponsible lending that led to the devastating developing-country debt crises of the 1980s.
The world has witnessed well over 100 significant banking crises over the past three decades. The authorities have even had to rescue important parts of the US financial system – on most counts, the world’s most sophisticated – four times during the same period: from the developing country debt and “savings and loan” crises of the 1980s to the commercial property crisis of the early 1990s and now the subprime and securitised-credit crisis of 2007-08.
The remainder of this column can be read here. Debate from our panel of economists appears below.
John Willman (FT): My colleague Martin Wolf is, as always, persuasive on the performance of bankers and the contribution their remuneration system makes in precipitating periodic banking crises.
However, to justify regulation, one has to be sure that the consequences are not worse than a failure to regulate.
After all, G7 economies have grown powerfully over the decades - despite the periodic crises. Regulation of bankers’ pay might have avoided or mitigated those crises, but it might also have dampened growth.
Martin’s argument is basically that on this aspect of the market economy, governments or regulators know better than firms what is good for business and the economy.
That is debatable. There are of course plenty of economies where bankers are not free to act as they do in the West, and none of them has such a strong economy.
Posted by: John Willman | January 16th, 2008 at 2:07 pm | Report this commentMartin Wolf: I agree entirely with John that one needs to be confident that regulation would be better than no regulation. It may well not be. But we cannot pretend that the alternative is, in fact, no regulation. Banks are heavily regulated already. Unfortunately, those regulations do not deal with core incentive problems effectively.
I think the relatively simple intervention in the structure of remuneration I propose would allow us to get rid of regulation of risk-management systems altogether (which is, I suspect, a complete waste of time). My bigger point, however, is that clever regulation must focus on the alignment of incentives between decision-makers and the wider interests at stake, which are not, in this case, only those of shareholders. I am certainly not suggesting that we should regulate the activities of bankers even more heavily. I want to influence their incentives, instead.
I do fear, however, as I said, that these crises are dangerous to the legitimacy of the entire capitalist system. There is more at stake here than the bankers’ annual bonuses.
Posted by: Martin Wolf | January 16th, 2008 at 5:48 pm | Report this commentGerard Caprio: I applaud your article “Why regulators should intervene in bankers’ pay”, and agree with your reluctance to do so, as one always should be careful of imposing new regulation, especially in finance, a field in which regulatory avoidance is especially easy and in which the rewards for such behavior can be large. You are also correct that the market will not likely arrive at a more sensible compensation structure on its own, though it is useful to remember that it did so in the distant past. It used to be the case that banker officers in the nineteenth century had to post a bond, as it was recognized that information problems made it difficult to monitor them well. The Rajan piece that you so rightly commend was not proposing to limit salaries but rather to limit their structure – compel banks to defer a certain and presumably large portion of compensation for a number of years until the outcome of their risky decisions are known, at least with a higher degree of certainty than at year-end. As bankers are compensated, so they will behave; if they can make a quick return before the costs of their strategy materialize, we should not be surprised when they do so.
A route to getting to this outcome could be to go a bit further down a road that regulators already are taking and compel greater disclosure of and some penalty for compensation systems that reward imprudent risktaking. The US system of CAMELS – not just capital, assets, management, earnings, and liquidity, but also systems (or sensitivity to risk) – are judged in the process of supervision. Yet the most important component of risk taking is how management rewards risk taking, without which knowledge it is hard to understand what a bank might be doing. As I and co-authors James Barth and Ross Levine recommended in Rethinking Bank Regulation: Till Angels Govern (CUP, 2006), there is no evidence that capital requirements or supervision work in improving the development, efficiency, safety, integrity, and governance of banks, but good evidence that market discipline can help, this from an analysis of a large cross-country database. Yet market discipline does not happen in a vacuum, and we recommended that supervision be used to support the market monitoring process, rather than as in the current Basel initiative, in which market discipline has widely been acknowledged to be an afterthought. In this context, markets need to see sufficiently detailed results of the CAMELS assessments, and regulators need to mark down banks that do not defer a sufficiently large percentage of their compensation. This might be a more attractive path to the type of regulatory outcome we would like to see, rather than to have bankers’ compensation debated in congresses and parliaments around the world, which surely accounts for the shudder I detected in your article.
Gerard Caprio is professor of economics and chair of the Center for Development Economics at Williams College
Posted by: Gerard Caprio | January 16th, 2008 at 7:07 pm | Report this commentMartin Wolf: I thank Gerard Caprio for his comments. My idea, too, is only to affect the structure of remuneration, not its level. We agree that the big issue here is incentives. I would be happy with any system that guides markets towards more effective and appropriate incentives for bankers. We cannot easily regulate what bankers do, but we may be able to influence the incentives they face.
Posted by: Martin Wolf | January 16th, 2008 at 9:52 pm | Report this commentHans-Joachim Duebel (guest): Based on the evidence from the current mortgage crisis, I think that Martin should be less scrupulous in proposing regulatory incentives favouring trailing or performance-related pay for bank management. Unfortunately, the crisis provides little reason to believe that the alternative, greater market control, could provide better results than regulation. Simply compare mortgage or capital market performance in the heavily regulated German with the partly unregulated (i.e. market-controlled) or weakly regulated US markets. Consider moreover that de-facto unregulated and owner-managed, i.e. at arm’s length market-controlled, companies such as Countrywide or New Century were dragged down the abyss first, and you must conclude that the markets itself, and not just its principal-agent problems, are a large part of the problem.
Problems as deeply entrenched as the deplorable short-termism in the financial markets usually root in mistakes made during education. Business schools have indoctrinated today’s financial managers now for decades with what I would call the Black-Scholes ideology: a degenerated view of financial markets as being primarily subjected to random price volatility in which the only successful financial strategy is trading, not long-term investment. The oil price, interest rate and exchange rate volatility of the 1980s, for which economists often had no explanations but which filled the coffers of traders, entrenched such views. However, if the current financial crisis, in many ways an aftershock of the 1980s, shows one thing, than it is the relevance of the possession of sound economic models of the primary markets underlying finance, e.g. of the housing market, for successful financial strategies.
Could government therefore intervene also in private business school curricula to ensure a balance between the long- and the short-term view, between real and financial sphere? Where I was educated, at University of Bonn, taking simultaneous courses in economics and business was mandatory (and this in fact also benefited the economists). One could only hope that the next round of financial regulation we are facing will not just end in tightening the operational corset further, but also will succeed in raising the educational and intellectual standards of those who operate.
Hans-Joachim Duebel is an international consultant in mortgage finance and founder of Finpolconsult, Berlin
Posted by: Hans-Joachim Duebel | January 17th, 2008 at 9:14 am | Report this commentJim O’Neill: In my opinion, Martin needs to spend a 6 month secondment in a bank!
Posted by: Jim O'Neill | January 17th, 2008 at 1:33 pm | Report this commentAnton Kuzmanov (guest): I am a regular reader of Mr Wolf’s columns and rather respect his point of view. Here, though, I must disagree, for the following reasons:
(1) Bankers’ pay, in and of itself is not the issue in the current crisis - if it were, the financial results at Goldman Sachs and Morgan Stanley, for example, would be the same, as bankers have generally the same pay packages in both places. I give those two banks as an example simply beqacuse it is clear that difference in risk management approach explains much of the difference in the quarterly results between the two banks.
(2) Pay regulation, in general, is not a positive stimulus for progress in any area - if that were the case, the US government, for example, will be among the best-functioning organizations, as civil servants’ pay is governed by their respective GS grade level.
(3) Market-driven pay is a driver of financial innovation - in a free-market environment pay is one of the mechanisms to attract the best individuals and allow them to drive innovation. Yes, sometimes innovation is subverted for short-term gains, but the market does eventually corrects itself, as recently “retired” CEOs of poorly performing banks can attest. While the consequences of misused financial technology are dire, I respectfully submit that securitization has been a major financial innovation that has greately increased the ability of individuals to borrow for home purchases at significantly lower rates (and, mind you, nobody was complaining about this financial technology 2-3 years ago when the homeownership rate in the U.S. reached 67%).
Posted by: Anton Kuzmanov | January 17th, 2008 at 4:15 pm | Report this commentIn response to Jim, is this an offer?
Posted by: Martin Wolf | January 18th, 2008 at 6:49 am | Report this commentOn a more serious note, I would love to know what Jim thinks is wrong not with my proposal (I could write the critique quite easily myself) but with my analysis of the incentives at work in contemporary banking. I would also love to know why he things Goldman has avoided the mistakes made by others. Is it because the culture of the partnership, with its very different incentives, still operates? Is it because people at Goldman are just smarter? Is it “there but for the grace of God go I?”
Posted by: Martin Wolf | January 18th, 2008 at 6:55 am | Report this commentEdward Golosov (guest): “Is this really the cause of the problem?…”
Martin has written a very timely article and I applaud the courage it took: surely, Martin has a few friends who are bankers, and they are likely to be displeased. The issue of assymetrical, option-like pay structure in banks is incentivisng the bankers to make short-term decisions and to take disproportinately huge risks - yet I see two areas in Martin’s article where his analysis did not quite go the full depth:
1) Why regulating compensation for risk-taking will be more effective than regulating the risk taking itself, which is in place already but isn’t quite working? Banking industry does attract very bright people, both academically and practically, and if they are so successful at arbitraging the regulation of risk, can you imagine what degree of ingenuity will go into arbitraging the rules that limit their own compensation?!
2) We know that the best regulator is the free market itself yet we believe that we cannot leave it to the market to deside on the faith of financial institutions because a fall of some of them would have too much of a social impact, take e.g. Nothern Rock.
Let’s not mix up too different issues here. First, Nothern Rock is a deposit-taking institution, i.e. a retail, not an investment bank. Retail banks is where people keep their money thinking it is safe, “money in the bank” - if they are not completely risk-free, then the system is misleading and risk-free financial institutions for retail market should be made available. One can argue that people can always buy government-issued paper - but is it that easy? Retail banks are on every high street while, have you actually tried to buy say gilts from your securities broker? Is it as convenient as every-day banking? Give people real risk-free options and explain that other institutions are risky. And, most importantly, should those risky ones fail, let them fail! There will be short-term political pain but the long-term educational benefit will certainly outweight.
As for the investment banks, nobody seriously expects them to be bailed-out Nothern Rock style (when did this happen last time?), and these are the places where the payoff structure is the most skewed, so buy introducing a new regulation, are we really addressing the problem or just creating another set of rules that won’t work?
Edward Golosov is a banker
Posted by: Edward Golosov | January 20th, 2008 at 1:06 pm | Report this commentLawrence Summers: I am uncertain as to the merits of Martin’s idea. But in at least two respects his argument is flawed.
First, he asserts that the change he suggests is not one that any one bank can make unilaterally. Why not? If he is serious about caring only about the structure of compensation not the level, then a bank that wants to change to a system that better aligns incentives can do this and can provide a level of compensation sufficient to compensate for the deferral element. Presumably if the whole industry is forced to make this change something similar will happen.
Second, there is a distinction under recognized by Martin between compensating people based on annual profits and compensating them based on stock market performance which recognizes forward looking risks. There is a reason why trading houses and especially those with a reputation for risk taking have much lower p/e’s than other institutions. Managers compensated in stock or options have strong incentives to run their firms in ways where they are credibly around for a long time.
Third, there are all kinds of technical problems - what about managers ability to hedge their stock? What about their excess incentive to sell out their institution so as to realize gains early?
Is it only proposed that this be done for insured depositories or all financial institutions, including those that don’t have stock? I doubt we want to encourage more people to leave major institutions and start their own hedge funds.
There is the meta question of why deferred stock will deter the behavior Martin doesn’t, when public humiliation and a foreshortened tenure of earning 30 million a year does not?
I think these are all serious problems, but it is easier to be critical than constructive. I am inclined to think that leaving stock prices out of it and causing traders’ compensation to be more subject to clawback after bad years may have merit.
Posted by: Larry Summers | January 21st, 2008 at 5:17 am | Report this comment