RBA Annual Conference – 1991 General Discussion
Ferguson and Thompson
The main discussion centred on three areas:
- the question of whether the RBA should supervise only banks or seek to supervise a much wider range of financial institutions; this raised the question of the objectives of supervision;
- the extent or nature of bank supervision; in particular, how intrusive the ideal supervisory system should be;
- the issue of whether more appropriate, or more detailed, supervision might have gone a long way to overcome some of the problems (increases in bad loans, falls in profitability, etc.) which emerged within the banking sector in the early 1990s.
Who Should be Supervised?
The current approach, which focusses on banks, could be traced back to the Royal Commission on Money and Banking of the mid 1930s, which equated system stability with the protection of bank depositors (subsequently reflected in the terms of the Banking Act). However, there had been substantial changes in the financial system since then. In Australia, as in many other countries, other types of financial institutions were performing tasks which were often difficult to distinguish from those performed by banks. Also, banks themselves had moved into a much wider ranges of activities than had been envisaged at the time of the Royal Commission or the drawing up of the Banking Act.
Was it still appropriate that RBA prudential policies focussed only on the protection of bank depositors? Most other financial institutions are covered by some supervisory arrangements. Could it, however, be argued that a range of other financial institutions and their depositors be given the same “protections” as bank depositors? Most participants felt there would be little merit in this approach. Banks remained the dominant group of financial intermediaries. In addition, there was value for the economy in having a spectrum of risk available to savers. It was important, however, that savers be well informed about the relative riskiness of different classes of investment.
A second question concerned problems caused by the diversification of banks' activities. Areas discussed included, not only the non-bank subsidiaries traditionally associated with banks (such as finance companies, etc.), but also the burgeoning areas of funds management, life insurance/superannuation and similar activities in which banks were becoming involved. How could the RBA be sure that problems within such associated activities would not spread to the banks themselves?
One response was that prudential policies already took into account the business undertaken by subsidiary companies insofar as supervision was possible on a consolidated basis. This aimed to ensure that minimum capital requirements were adequate for the activities of the “bank” broadly defined, and limited large credit exposures of the consolidated group. However, other activities of banks, such as the newer ones mentioned above, were not readily amenable to this approach. Here the Bank's policy was one of seeking to ensure that a bank was sufficiently separated from that associated institution or activity so that it was not weakened by any perceived difficulties in this related activity. In practice, this would mean that it would have to be clear that banks did not stand behind the businesses with which they were associated. Some participants argued that it would be extremely difficult to ensure such a separation in practice.
There was some passing discussion, in this context, of deposit insurance. If prudential policy had as a major aim protection of bank depositors, would it not be sensible to formalise that through deposit insurance? One suggestion was that existing non-callable deposit arrangements (NCDs) could be the basis of funding such increase. Others noted the serious problems with deposit insurance – such as moral hazard and the difficulty of determining appropriate funding. Further discussion of this issue was resumed in the next session (see discussion of following papers).
The Extent of Supervision
What should be the extent or nature of supervision? Current supervisory practices in Australia tended to be relatively informal, relying on setting broad prudential standards (e.g. for capital), analysis of data provided by banks, discussions with banks, etc. It did not extend to on-site inspection or examination.
Some participants suggested that there was a strong case for some form of on-site supervision by the Bank. However, most recognised that such an approach also had disadvantages. The balance of costs and benefits could depend very much on the degree of such on-site supervision – in particular, how intrusive it was. At the one extreme was the US approach, which involved extensive on-site examination. One participant pointed out that the US system was extremely labour intensive; the inspection of Australian operations of US banks (which were very small relative to their parents) had recently involved 48 US bank inspectors being in Australia for a month. (That sort of supervision had not prevented the Australian subsidiaries or their foreign parents incurring bad loans.)
Most participants felt that such an approach, while no doubt providing some supervisory insights, carried some very definite disadvantages. First, there was no evidence that such an approach was effective in preventing problems. The US domestic experience of recent years was cited as illustrating that point. A second and more subtle disadvantage was the possibility that the central bank itself would, by virtue of its in-depth inspection, become more implicated in the inspected bank's operations. In some sense, banks being supervised through detailed inspection processes could be perceived to have gained the central bank's imprimatur. Consequently, they could begin to take less responsibility for their own actions; and markets might become less inclined to apply critical judgments to the performance of banks.
At the same time, many saw a strong case for greater contact between supervisors and the banks. Some spoke of the benefits which would accrue from the central bank spending more time in the banks, obtaining a feel for the way a bank conducts its operations and the personnel involved. Such an approach was likely to yield more information than greater reliance on providing more written information.
One participant felt that there were improvements that could be made to current practices rather than to policies as such. For example, he suggested that the Reserve Bank could encourage a common language and common standards between banks – areas mentioned included that of credit-quality guidelines, when to classify loans as performing or non-performing, etc. One of the problems at present was the varying standards adopted by different banks. There was a call for better information on the direction of lending, e.g. how much was being lent for property.
More direct contact between the RBA and external auditors of banks was also called for. It was recognised that there were a number of complicating factors in this; how realistic was it to expect auditors appointed by banks to form an important part of the central bank supervision process? Auditors' allegiance might, in theory at least, be thought to be with those paying the bills; that is, with the banks.
Could Supervision have Prevented the Problems which Emerged?
The most extensive discussion of the session focussed on the question of whether better supervisory practices could have forestalled the banks' problems which showed up as loan losses in the early 1990s.
The general conclusion which emerged was that, while there were areas in which supervisory policy could have been improved (see Section 2 above), to have prevented all the excesses would only have been possible with the help of a very heavy dose of hindsight. It was also noted that the Australian banking system had emerged from the 1980s in relatively sound condition. The losses incurred had fallen mainly on the owners of banks; lost capital had been replaced.
There were a number of factors at work which contributed to the losses of banks and which made supervision difficult, e.g.:
- competition was extremely strong as a result of new bank entry and deregulation of the established banks. The domestic banks took the threat from foreign banks very seriously (perhaps too seriously). There was a reduction in “spreads” on banking business, particularly in the wholesale markets, which meant there was not enough fat to cover the loan losses when they eventually appeared;
- the “culture” of banking altered – from transactions based on customer relationships as in the days of regulation to transactional, non-relationship banking in the 1980s. The resulting gap which developed between the banks and the customers over this period proved costly in many instances;
- a related factor falling under this heading was the tendency for banks to focus on the acquisition/creation of new assets rather than the maintenance of existing assets. One participant felt that a neglected aspect of this period was change in the market for loan officers; from being a low-key market in the 1970s, it became overheated in the 1980s. It was a good idea for loan officers to make more loans. In the short run, more loans meant more profits. If things went wrong, there was a fair chance that the loan officer would have changed jobs by then. With the market buoyant, there were always new job offers;
- banks were slow to improve their risk assessment and management procedures. This led to concentrations of risk which ultimately proved damaging (see below);
- the type of lending sometimes proved to be inappropriate. An example cited was the use of negative pledge lending. While this was a sensible lending strategy in some specific instances, particularly where blue-chip companies were involved, it became far too widely used in relation to weak borrowers. Lending focussed on assets rather than sustainable cash flow;
- a related point was the observation that capital markets could not be said to have restrained the activities of banks, or the sectors of the economy that boomed and subsequently declined. How realistic is it to criticise bankers for making “overstated” valuations of assets or a particular business venture when capital markets themselves – especially the share markets – were giving signals which suggested valuations even greater than those adopted by the banks? Once again, the consensus on this seemed to be that the latter half of the 1980s was a unique time in many respects – marking the beginning of the transition from a largely regulated to highly deregulated financial system.
The combination of these factors meant that formal supervisory policies were being introduced in a particularly difficult time. It is unlikely that any set of prudential policies – short of outright prohibitions on certain types of lending and/or restrictions on volumes of lending (which are really old direct controls rather than prudential supervision) – could have effectively come to grips with the circumstances. Even if this were to have occurred, lending could have shifted elsewhere within the system or borrowers gone offshore – as many did anyway.
There was a lot of discussion on why lending risks came to be heavily concentrated. Why was it that so many banks chose to lend heavily in areas that ultimately turned sour, e.g. property?
It was pointed out that, to some extent, concentrations of risk occur naturally and are hard for banks to avoid. For example, the most profitable areas of the economy will tend to be expanding most rapidly and seeking the assistance of financial institutions to facilitate their business activities. This will often mean that particular sectors of the economy will be taking on loans more rapidly than others at various times in the cycle. Banks, for competitive reasons, will often find this difficult to resist, particularly if failure to follow trends leads to losses in market share. By definition, this can lead to concentration of risk.
Some participants asked why banks could not diversify their lending portfolio by selling off loans in areas they were overexposed to; i.e. use portfolio theory to manage credit risk. At this point, the discussion again returned to the question of what was the essence of banking. If, as agreed (see Section 2), it was the making of loans to firms on whose credit risk the bank had good information, then it would be very difficult to sell these loans. The purchaser would have no special information, and would therefore not be in a position to judge whether they were worth buying. They could well assume that the seller would keep the good loans and sell the bad ones. In other words, credit risk could not be diversified by securitisation. In some limited cases where information was not crucial, e.g. in relatively homogenous loans such as house mortgages, securitisation was possible, but not for most types of business loans.