INTERNATIONAL MONETARY CONFERENCE
HOTEL ADLON, BERLIN – 3 JUNE 2003
Howard Davies
Chairman, Financial Services Authority

The corporate governance industry is a growth business, and has been so for the last decade or so. At present, that distinguishes it from most parts of the financial sector, of course.

There are corporate governance codes a go-go. At global level, we have an OECD code. The World Bank has been active in producing advice for developing countries. The international regulatory bodies have all said their piece. Even the Commonwealth has developed its own best-practice guide.

In individual countries there have been a rash of home-grown initiatives. The Sarbanes-Oxley Act in the US is the best known, if not the best-loved. But a recent review for the European Commission revealed the existence of some forty different corporate governance codes across the EU.

That does not surprise me, because in the United Kingdom we have been especially fertile in generating such codes. First there was Cadbury, then Greenbury. And, predictably, once there were two codes, they began to breed. Since then, in quick succession, there have been Hampel, Turnbull and most recently two new ones, Smith and Higgs, appeared on the same day. This is one area in which we can proudly say that the United Kingdom leads the world in productivity.

But in this role of honour let us reserve a special place for the European Commission. Just the other day, on the 21st of last month in fact, the Commission sent a communication to the Council and the Parliament on modernising company law and enhancing corporate governance in the European Union. In that report the Commission concludes that "there is no need for an EU corporate governance code". That observation alone must be worth some kind of prize. Though the deregulatory effect was somewhat diminished as what followed was an EU action plan and the construction of a European corporate governance forum.

There is, of course, much good sense in many of these codes. I am in little doubt that the pioneering work done by Adrian Cadbury in the UK has been influential, and beneficially influential, in developing the role of corporate Boards in Britain. More recently, too, I find a lot to support in Robert Smith’s work on audit committees, which I commend to you if you haven’t seen it – and not only because Robert Smith was for a while a member of the FSA Board.

But I am often struck by the fact that many of these recommendations are very "by and large". They devise templates for use across a wide range of businesses, large and small, financial and non-financial, whose requirements are often very different. And, indeed, they sometimes look like solutions in search of a problem. Rarely are they founded in a careful analysis of just what has gone wrong in large corporations, and on the changes that might be needed to prevent those problems emerging again.

So, today, I propose – unusually – to proceed the other way round. In other words, to look at what problems we have, in practice, seen in financial institutions and how far those problems can be sourced back to corporate governance failures. In doing so I will, following my brief, focus exclusively on financial institutions. Though, given the FSA’s broad coverage, I will not just talk about banks, but insurers and securities firms too.

Before I do so, I should make two preliminary points. First, problems as perceived by regulators are not always the same as problems perceived by firms. We are bound to take compliance failures particularly seriously and tend to see them as indicative of a larger issue. I am aware that some firms do not take that view. So I have tended to concentrate on those problems which both of us would certainly accept as serious – in other words corporate collapses and major losses of shareholders funds, often accompanied by the disappearance of senior management. That latter phenomenon is an example of something which we would not necessarily consider a problem, but many of you would.

The second preliminary point is that corporate governance, in the sense of the structures of control and accountability in an institution, is not always the same as management. Sometimes there are management failures which are simply the result of poor day to day judgements, and which do not necessarily indicate a structural weakness, or a failure of oversight. The two are often unhelpfully confused. I will try my best to keep them apart, though it isn’t always easy to do so. Some recent corporate failures in the UK have been lumped in with Enron as failures of governance, which in my view were largely a result of a wrong strategy. And no corporate governance code will protect you fully against strategic error, or paying too much for an acquisition.

So what can we learn from experience?

There is not as much systematic and useable analysis of the reasons for failure in financial institutions as one would like. But there are some helpful sources.

European banking supervisors studied a number of banking problems across the continent between 1988 and 1998. Their overall conclusion was that management and control weaknesses were underlying, fundamental and contributory in almost all of the cases they considered.

A very similar conclusion was reached by a group of European insurance supervisors who looked at 21 cases of failure or near failure in European insurance companies between 1996 and 2001. (One astonishing feature of the work was that there were, during that 5-year period, a total population of 270 cases to consider, from which that selection of 21 were made.)

That report shows that there were usually a number of contributory causes to the collapse of an insurer. Poor underwriting practice, or inadequate reserving, was often the proximate cause of failure. But in all the case studies underlying management or governance causes were identified, in many cases relating to significant systems and controls issues. The widespread underwriting or asset problems were able to arise because of these fundamental weaknesses, and the combination of poorly managed risks made firms particularly vulnerable to adverse external events.

  • When the group compared problem cases with other firms who weathered similar circumstances better, a pattern emerged of the following four forms of management malfunction:
  • Incompetence, with firms straying outside their field of expertise or uncritically following the herd instinct.
  • Excessive risk appetite, or objectives that were at odds with the prudent management of the business;
  • Lack of integrity, or
  • Lack of autonomy and inappropriate pressure for short-term results from, perhaps, the parent company.

Following my distinction between management and governance problems, I see governance issues in at least 3 of these 4 cases. Certainly a Board should identify the risks involved in companies straying outside their field of expertise, should ensure that there are no incentive structures in place which promote excessive risk-taking, and should ensure that the business does not come under inappropriate pressure to maintain earnings or market share.

A third interesting source can be found in the New York Fed’s most recent economic policy review, published in April of this year. It was a special issue entitled "Corporate governance: what do we know and what is different about banks?"

The volume includes a series of interesting analytical pieces, by different hands. They all, from different perspectives, try to assess just how much relationship there is between good governance and corporate success. As that, after all, is what ought to interest shareholders, not elegant governance per se.

One of the papers concludes that Board composition does not seem to be a useful predictor of firm performance. That is an interesting conclusion given the focus on Board composition in some of the codes. On the other hand, they found that in the US, at least, Board size does have a negative relationship to performance. In other words, the bigger the Board, the poorer the results.

But what is more striking about this study is what it says we do not know. The Fed’s researchers conclude that "research on the corporate governance of public institutions has raised more questions than answers. In particular, the causes of problems and the consequences of governance structures remain elusive".

Furthermore, they note that there are different corporate governance structures in different industries, which have grown up in response to the different needs of companies in each sector. So, they conclude, "reforms that do not take into account industry differences may not have the same intended effect across industries".

One conclusion we might draw from that review is that we should be wary of ‘one size fits all’ codes, which propose the same rules for companies of different sizes and in different sectors.

Some of the bankers among you might welcome this conclusion. But I ought to add that one of the pieces in the Fed’s report argues that "a clear case can be made for bank directors being held to a broader, if not higher standard of care than other directors". Essentially, their point is that there are important stakeholders in banks other than shareholders – including major creditors and other institutions, given the safety net. The same authors go on to argue that, as a result, they support "a hybrid approach to corporate governance in which most firms are governed according to the US model, while banks are governed according to a variant of the Franco-German paradigm".

I have made it a rule in my working life that, when I hear the word paradigm, I switch off and move on. So let me move, now, from academic analysis to anecdote. But not, I should emphasise, just one anecdote. I have sought to reflect on what the FSA’s experience - of financial regulation - and that of its predecessor bodies has told us about the role of poor corporate governance in corporate failure, and what might usefully be done to minimise the risk of such failures in the future.

One cannot thereby construct a comprehensive theory of the sources of success, or indeed failure, in financial institutions. But one can draw out some features which seem likely to increase the risks which institutions run.

Our experience shows, for example, that a dominant chief executive, or indeed business head, who is not effectively challenged by the Board or his colleagues, is a danger sign. Similarly, a Board lacking in relevant experience is unlikely to act effectively as a constraint on excessive risk taking.

It is also clear, to me at least, that a Board which does not display sufficient interest in strategic issues is putting the institution at risk. That is one aspect of the current corporate governance debate which concerns me. The codes of practice often seem to me to encourage Boards to focus attention on risk management and control processes, in a way which can leave strategic thinking as an optional extra. Yet incoherent strategies can be just as potent a source of instability as poor risk management itself.

A related point, which emerges strongly from our experience, is that Boards must be open to external stimulus. It is dangerous for a Board to accept all its information and advice from within the company, and be unprepared to listen to dissenting views.

From this necessarily brief review of the evidence, and particularly of the sources of failure in financial firms, I draw some tentative conclusions. It is important to recognise, however, as I said at the outset, that the evidence base for firm recommendations on corporate governance in financial institutions is thinner than one would like, and certainly not robust enough to offer a standardised set of recommendations valid at all times and in all places.

My own principal conclusions are:

  • First, that people are more important than processes. Many of the failed firms, or near failed firms which we have encountered, had Boards with the prescribed mix of executives and non-executives, with socially acceptable levels of diversity, with directors appointed through impeccably independent processes, yet where the individuals concerned were either not skilled enough for, or not temperamentally suited to, the challenge role that came to be required when the business ran into difficulty.
  • Secondly, and in spite of my first conclusion, there are some good practice processes worth having. Properly constituted audit committees, and Board risk committees can play an important role, as long as they are prepared to listen carefully to sources of advice from outside the firm.
  • Third, and this is a foundation stone of the FSA’s approach, a regulatory regime built on senior management responsibilities is absolutely essential. In some of the cases we have wrestled with, senior management did not consider themselves to be responsible for the control environment and indeed, in the old pre FSA regime, were able successfully to claim that they were not responsible even if the business failed. So our regulation is built on a carefully articulated set of responsibilities up and down the business. It is important that they are not unrealistic. We do not expect the CEO to check in the bottom drawers of each of his traders for unbooked deal tickets. But we do expect the CEO to ensure that there is a risk management structure and a control framework throughout the business which ought to identify aberrant behaviour, or at least prevent it going on unchecked for any length of time.
  • One consequence of this senior management regime, my fourth point, is that regulators must focus attention on the top level of management in the firm. For the major firms we regulate we insist that our supervisors have direct access to the Board, and that they present to the Board their own unvarnished view of the risks the firm is running, and of how good the control systems are by comparison with the best of breed in their sector. Unfortunately, we find some resistance to this approach. The management of some of our firms want to negotiate the regulators assessment, so that when it reaches the Board it is an agreed paper and sufficiently bland to cause no debate. I think a well-structured Board, and a confident management, should welcome an independent view, even expressed at the Board level, which they may challenge and contest if they wish. And non-executive directors should find it helpful to see a knowledgeable view of the institution which does not come from or through its own senior management.
  • My fifth and penultimate point may not be a popular one. Boards should take more interest in the nature of the incentive structure within the organisation. I am not talking solely about the pay of the CEO, important though that is to get right - as some firms in Britain have recently discovered. I am talking about ensuring that the incentives within the firm, and pay is a very powerful one, are aligned with its risk appetite. A number of our most problematic cases have their roots in a misalignment of incentives.
  • Lastly, no corporate governance system will work well unless there is some engagement on the part of shareholders. Boards are responsible to shareholders. That is the received wisdom in Anglo-American capitalism, at least. But if those shareholders are not prepared to vote their shares, and show little interest in business strategy, then that accountability is somewhat notional, and unlikely to be effective. Certainly regulators cannot hope to substitute for concerned and challenging shareholders, though in some senses they may complement them.
  • These six points do not, I should say, amount to another, "Davies Code" of corporate governance in the financial sector. The one thing the world, certainly as seen from London, does not need is another code. But I hope there are some points here which may provide raw material for the discussion and debate which follows.

More Speeches: