Financial instability and the banking sector
In the course of an analysis of the relationship between competition and regulation in banking, Revell (1986) briefly surveys some of the literature on financial crises associated with the work of Minsky and Kindleberger. He argues that aggressive competition in the financial sector tends to lead to financial crises and provides a motivation, beyond that of the desire to form a cartel to secure supernormal profits, for banks to form agreements to avoid taking on unduly risky business and adopting risky practices. Such agreements to suspend interbank competition can be regarded as a primitive form of self-regulation to preserve the safety of the financial system and the continued existence of established institutions. He acknowledges the paucity of empirical backing for Minsky's statements.25 The main evidence presented by Minsky (1977), for example, is the deterioration of a few financial ratios between 1950 and 1974. However, despite the lack of evidence, Revell declares a 'gut feeling' that Minsky is right and goes on to suggest a method that might be used to verify the FIH (see below).
Revell draws attention to a study by Barclay (1978) in which nineteenth century financial crises in Britain are compared with the 1973-4 secondary banking crisis. Barclay develops a sort of cobweb model of crises in which surplus profits in one sector of the economy, perhaps resulting from technological innovation or government regulatory changes, attract a large entry which is dangerous to the banking sector. The economic system is regarded as unstable and highly competitive so that a large entry is attracted and profits are forced below normal levels. The new entrants do not all know about each other's intentions, and there is a time lag before they can start production. Assuming that all firms do not have identical cost structures, some will leave the industry as profits fall, some will stay and some will go to the wall.
The model can be applied directly to the banking sector, where self-imposed or government-imposed regulations lead to supernormal profits. Established banks will have qualities of reputation and reliability. New entrants, so called fringe or secondary banks, must offer differentiated products, leading to a tendency towards innovation, or price more cheaply. They may therefore attract more risky business without fully compensating by adding appropriate risk premiums to their charges. The profits derived from risky business will be higher as a rule, in the absence of failure, so that imprudent bankers will do better than prudent ones. Competition leads to a lowering of the quality of banking products and this increases the potential for crises since depositors are interested primarily in the security of their deposits rather than the profitability of the bank. However, at the first sign that deposits are not secure, withdrawals will occur, and these can have domino effects which reverberate around the system. Influxes of producers attracted by high profits are, therefore, potentially more serious for the banking sector than for other sectors, while the inherent instability creates regulatory conditions that are likely to result in excess profits and encourage established banks to innovate.26 When a shock hits the system profitability declines and capital, which is commonly run down relative to assets and risks taken in the euphoria of boom markets, proves inadequate. This is because past failures come to be regarded as anomalous as memory of them decays and regulators adopt more liberal attitudes.
Fringe or secondary banks are commonly allowed to fail by the regulators. The profitability of the remaining 'core' banks will eventually be restored, and the process could be cyclical. Recent banking history furnishes numerous examples, including the Latin American debt crisis, of established banks tapping into new and profitable lines of business only to find that competition from smaller banks gradually erodes margins and encourages mispricing of risks.
Barclay argues that the preconditions for attracting new entrants, namely surplus profitability, existed prior to the UK secondary banking crisis of 1973-4 because banking was organised as a cartel supported by the Bank of England. He suggests that the banking cycle could be grafted on to the general economic cycle as innovations in the nonbank sector, leading to supernormal profits, attract new entrants which require bank finance for their ventures in the manner described by Schumpeter (1939) - whose work is discussed in section 4.2. The banks then have a tendency to become over-exposed to the sector with supernormal profits and can face bad debts in the future as a result. The involvement of the UK secondary banks in the commercial property sector can be seen as a combination of both these themes. Government regulations created pockets of excess profitability in the property markets and attracted entry. This led to a property boom and stimulated the growth of the fringe banking sector, which became over-exposed and provided speculative finance to companies in the property sector.27 The 1987-8 Texas property market collapse and its effects on the banking sector leading to the bail-out of the state's largest bank holding company, First Republic Bank, is another example of the potentially damaging effects that the tendency towards over-exposure to excessively profitable sectors can have.
The 2007-9 Global Financial Crisis had its origin in the US home loan market, rather than the commercial property sector, but also involved the development of a large fringe, or 'shadow banking' sector, as it became known, Rajan (2010).
The Barclay model, Revell claimed, could be developed to show that increases in competition in banking lead in almost all cases to growth of bad banking practices involving excessive risk-taking. Such practices are likely because retribution only follows in the next crisis, which might be caused by an adverse shock but is unlikely to occur until the banking system is in a state of what Minsky would call fragility. Losses and failures resulting from bad banking are, therefore, likely to be bunched. Before the crisis, the profit advantage lies with the aggressive banks that indulge in unsound and speculative practices, but established banks are likely to be drawn into bad banking in order to boost their profitability. A sort of Gresham's Law of banking holds in the sense that bad banking drives out good.
Particular features of boom periods that draw banks into unsound practices can be identified, Revell argues. The general price level will be rising, and prices of equities and properties will rise with it so that capital gains will be an important component of financial calculations. Banks not only speculate on rising price levels themselves but also issue loans that can only be serviced if the price of borrowers' output, and that of their collateral, rises continuously. During the boom, Revell argues, the yield curve usually implies that financing long-term assets with short-term deposits and liabilities yields high profits. The importance of these factors is magnified when, as in the post-war period and particularly the 1970s, accelerating inflation is superimposed on the business cycle. He further argues that risk varies over time and that his analysis implies that the probability of loss increases with each year that has passed since the previous crisis. The behaviour of bank managements, however, ignores this factor.
Provisions for loan losses and assessments of capital adequacy by banks are normally based on the averaging of losses over the immediate past; consequently, what might be regarded as in-house bank insurance28 declines once the cyclical upswing is under way. This tendency is reinforced, Revell argues, by the fast growth of asset values, which usually outstrip the provisions for bad debts and additions to capital, and by the growing feeling of euphoria as speculation continues to be rewarded with profits. He observes that the next step in the argument has not been documented satisfactorily. It consists of the hypothesis that one of the reactions of the established banks, which have by definition survived one or more crises, is to face the continuation of the boom (in which increased risk and competition went hand in hand) by banding together to restrain bank competition. It is necessary, he notes, to examine the history of interbank agreements or 'cartels' in a number of countries to see if wider support for the hypothesis, which relies heavily on UK experience in the way that Minsky's FIH does on US experience, can be found.
The shift from primitive self-regulation to regulatory involvement by the authorities becomes necessary as soon as it is demonstrated that the former can break down (Goodhart 1985, Ch. 4). Revell suggests that this point was not reached in most countries until the 1930s. The 'cartels' proved inadequate because they lacked the means to enforce behaviour (Goodhart 1985, Ch. 4), especially once threatened by a competitive fringe. It is significant, Revell feels, that many of the agreements were continued with official blessing after the 1930s, and it was not until the mid-1960s that there were significant developments leading to increased competition in many banking systems around the world that involved disbandment or modification of the cartel agreements. This process is particularly evident in the United Kingdom and has continued in the 1970s and 1980s (Llewellyn 1985, 1986). Competition in the US banking system also began to increase in the 1960s and accelerated in the 1970s and early 1980s (Mullineux 1987c,e). The latter development was led by the broadly defined banking industry, which successfully exploited regulatory loopholes, and permissive legislation followed later. Also in the United States, and then elsewhere, financial innovation was stimulated in the 1970s and 1980s by advances in information technology, improved telecommunications networks, high inflation and volatile exchange and interest rates.
Revell draws attention to the fact that changes in regulatory regimes in Europe since the late 1960s were designed to increase competition in the broadly defined banking sector and were followed in a very few years by the most significant financial crisis since the War. He infers that the authorities in many countries thought competition would lead established banks to drive out fringe operators. This idea seemed to lie behind the Competition and Credit Control arrangements introduced in the United Kingdom in 1971 (see Bank of England 1971). The clearing banks were instructed to abandon their interest-setting cartel and in return were released from certain controls, allowing them to compete more effectively with the fringe banks. The result, Revell argues, was the secondary banking crisis which erupted in the United Kingdom in 1973 and, as a result of similar structural changes elsewhere, particularly in Europe, failures in other countries. Revell attributes the failures to bad banking. A number of failures in the early 1970s were associated with losses on foreign exchange markets, but closer inspection, Revell concludes, demonstrates that these banks were tempted to speculate on the foreign exchange markets in order to recoup losses made on domestic banking operations.
The UK secondary banking crisis is a classic example of bad banking. The fringe banks raised short-term wholesale funds and invested them in long-term loans to finance speculative ventures, particularly in the property market. These ventures yielded no immediate income so that interest due had constantly to be added to the principal. In Minsky's terminology, they therefore participated in Ponzi finance. In addition to speculative banking, there were many cases of fraud and embezzlement. A worrying feature of the buildup to the crisis was that the fringe banks were able to raise finance from the established banks, which were prevented by the regulatory authorities from participating in the binge directly. The euphoria of the period was such that few saw the dangers -possibly because bank managers had no personal experience of the previous crisis in the 1930s, regarding it as an anachronism, Revell postulates. The period of euphoria came to an abrupt end with the acceleration of inflation, the freezing of the property market and the drop in equity prices. After the first hint of bank failure, the supply of wholesale funds to the fringe banks, via the interbank market, dried up, and the Bank of England's 'lifeboat' operation was launched.30 Subsequent examples of widespread bad banking can be found in the United States, where many savings and loan associations got into trouble in the late 1970s, for example; and, on an international scale, in connection with the Latin American debt problem. When the Mexican crisis broke in 1982 it was realised that the country risks involved in sovereign lending to developing countries, which needed to refinance their debts continuously to meet interest payments, had been severely underestimated.
Revell also examines other bank failures in Europe and the United States in the early 1970s and is struck by the similarity of the business undertaken by the failed fringe banks, most of which were linked in some way with the property market. The fact that there were incidences of bad banking in many countries suggests the possibility of a common cause. Revell believes this to be the relaxation of structural controls over banking systems with a view to making them more competitive and the association of these developments with accelerating inflation and the general euphoria that accompanied it. He postulates that there might have been no ill effects if liberalisation had taken place at a time of financial stability, but the combination of greater competition with inflation and euphoria proved lethal. This begs the question of why inflation increased in the first place. In terms of Minsky's FIH, it could have been a result of a rise in government deficit financing to head off a perceived increase in the risk of crises as the economic expansion in the early 1970s built towards a euphoric boom and financial instability increased.
Revell is troubled by the fact that failures occurred even in countries with the strictest prudential regulatory regimes, since this implies that competition may not be able to ensure sound banking and that increased prudential regulation cannot be relied upon to contain the tendency of increased competition to generate bad banking practices. He feels that the prevalence of bad banking might, however, be attributable to widespread euphoria. All regulatory systems, he observes, even the most formal statutorily backed ones, leave a lot of discretion to the supervisor. If supervisors are also influenced by the euphoria, they will not act to prevent the development of unusual practices which do not contravene the rules. This implies that regulators and supervisors should not relax their vigilance in times of euphoria and that the regulatory and supervisory system must be sufficiently flexible to deal with rapid structural changes.
Metcalfe (1982) shares Revell's view that crises are a recurrent phenomenon in banking systems and feels that the generic characteristics have been succinctly described in Kindleberger (1978). He argues that crises originate with events that significantly increase profit expectations in one sector of the economy and stimulate an increased demand for finance. Innovations could play an important role here, as Barclay (1978) observed.31 The 2007-9 Global Financial Crisis (GFC) with its subprime mortgage lending boom are the associated increased issuance of mortgage backed securities and the innovation of credit default swaps and collateralized debt obligations (CDOs), bears this out, Tett (2009).
The extension of bank credit then increases the money supply and self-exciting euphoria develops in the manner described by Minsky and Kindleberger. More and more firms and households are tempted into speculative finance and are led away from rational behaviour, and manias or bubbles result. As Kindleberger observed (1978, p. 17), the term mania emphasises the irrationality and the term bubble foreshadows the bursting. Even the suspicion that the tissue of expectations is weakening can spark a crisis. Only a small incident is needed to transform manic behaviour into panic behaviour, which inflicts widespread damage. Metcalfe argues that banks exemplify a general characteristic of all organised social systems, which is that of the management of a system of expectations under conditions of risk and uncertainty.32 Like other social systems, the effective functioning of the banking system depends on a fabric of co-ordinated expectations. Mutual expectations guide the behaviour of participants and are modified in a process of interaction. To form their own expectations, individuals must draw on assumptions and beliefs about the rules and values of others.33 Value consensus cannot be assumed but values do coalesce into more or less coherent doctrines, myths or ideologies, which provide the intellectual rationale and moral basis for legitimate actions. This feature of social organisation, he argues, is particularly prevalent in banking where the extension of credit creates webs of rights and obligations spread over time.
The fulfillment of expectations depends on confidence in the continuing viability of the whole system and vice versa. As Hicks (1967) observed, however, stability of the banking system cannot be taken for granted because it is potentially unstable in two directions. Performance will fall short of potential if there is a lack of confidence; but overconfidence, as in the buildup to a crisis, will produce unrealistically high expectations. Because expectations are never precisely fulfilled in any complex social organisation, there is a continuous management problem of avoiding the extremes of mistrust and overconfidence. The maintenance of balance requires an institutional framework or regulatory system that encourages confidence while restraining constituent organisations.
Before leaving the topic of instability in the banking sector, Diamond and Dybvig's (1983) contribution should be considered. In their model, the illiquidity of assets provides the rationale for both the existence of banks and their vulnerability to runs. It is assumed that even if assets are traded on competitive markets with no transactions costs, low returns are received by agents forced to 'liquidate' early. Agents are, therefore, concerned about the cost of being forced into early liquidation and will write contracts reflecting these costs. Investors face private risks which are not directly insurable because they are not publicly verifiable, due to the existence of information deficiencies and asymmetries. The onset of the GFC in August/September 2007, involved a North Atlantic Liquident Squeeze, Mullineux (2008), in which banks, which had become increasingly reliant on wholesale financing through the interbank market, found that they could not renew that funding due to increased perception of the liquidity risk associated with lending to bank counterparties.
Banks serve the function of indirectly transforming illiquid assets by offering highly liquid liabilities, such as demand deposits which are redeemable at their nominal value; they thereby provide indirect insurance that allows agents to make payments when they want or need to and, in so doing, share the risks. Efficient risk-sharing can result if confidence is maintained but if agents lose confidence and panic, incentives are distorted and a bank run results. To prevent loss of confidence, Diamond and Dybvig advocate comprehensive, government-backed deposit insurance. Its provision is, however, likely to create moral hazard problems, which must be dealt with if an optimal bank regulatory regime is to be developed. The GFC indeed prompted widespread extension of deposit insurance cover. Risk-related deposit insurance premiums or capital adequacy requirements would appear to be needed to contain the moral hazard, but such issues are beyond the scope of this book.