Behaviour consistent with bubbles, namely a shock followed by a rapid fall, a slow change and then a rapid recovery can be described by catastrophe theory. (See discussion at end of section 2.3.) Cusp catastrophe models, Varian (1979) demonstrates, can be used to explain cycles containing recessions, and also cycles containing depressions, in which the return from the crash is gradual and drawn-out. The path followed is dependent on the severity of the shock. The shock examined by Varian is a wealth shock which might, for example, emanate from a stock market crash. The links between financial market crashes and economic recessions and depressions remain uncertain, however, as Galbraith's (1954) analysis of the possible links between the 1929 New York Stock Exchange crash and the subsequent depression demonstrates. In contrast, the October 1987 crash appears to have had relatively minor effects, which suggests that the impact of financial crashes depends on the underlying condition of the economy at the time and the distribution and proportion of wealth-holding in the form of stocks and shares at the time.
If bubbles exist, then plans to save and invest may be subject to erratic change because, in a world of uncertainty, the expectations on which decisions are based are liable to change rapidly and over-optimism or euphoria can build up. As a result, Keynes argued (1936, p.322), misguided over-optimistic expectations can lead to investment that would normally be deterred at the prevailing rate of interest.
The identification of the cause of the buildup to euphoria, which encourages the speculation that underlies bubbles, remains a problem. Even in the most notorious case of the speculative orgy of 1928-9, Galbraith (1954) finds it difficult to pin down the exact cause. He notes that easy money was not the cause because interest rates were relatively high. He observes, however, that speculation on a large scale requires a sense of confidence and optimism and a feeling of trust in the intentions of others. Savings must be plentiful in aggregate, but speculation can rely on borrowed funds. If people feel well off, then they may be willing to 'have a flutter' and take more risks in the hope of enhanced return. He postulates that speculation is, therefore, likely to follow a long period of prosperity which leads to rising expectations of future incomes. He cites the South Sea Bubble of 1720 as an example. Savings had grown rapidly and domestic returns had declined. The collapse of the bubble destroys the mood that fosters speculation but, with time and the dimming of memory, immunity to speculation wears off and a recurrence becomes possible, he argues.
Minsky's FIH and related work imply that bubbles entail elements of irrationality, but bubbles can apparently also occur under rational behaviour. The rational bubble models, however, add little to our understanding of the phenomena, and some of those who have demonstrated the possibility of their existence believe that irrationality is displayed in historical accounts of bubbles. The irrational bubble theories imply that financial markets provide vehicles for speculative excesses. Minsky views the financial system itself as basically unstable. Others, such as Eichengreen and Portes (1987), see the financial system as potentially sound, once appropriately regulated, but as playing a significant role in generalising the effects of shocks emanating from the real sector.
If the financial system is fundamentally unstable, then regulation is clearly required. Minsky (1986), however, argues that it will not be possible to assure financial stability through regulation since instability laid to rest by one set of reforms will eventually emerge in a new guise. Those that do not share Minsky's views have a greater belief in the potential for regulation to prevent speculative excesses and the resultant crises. Nevertheless, the rapid financial innovation in the 1970s and 1980s ensures that regulators and supervisors must be ever-vigilant and develop flexible systems to contain the speculative urge and associated risks. In this vein the Brady Report50 recommended that margin requirements in the US futures markets should be tightened following the October 1987 crash. In addition, should a financial crisis occur, the LOLR and BOLR should act swiftly to prevent it leading to a multiple credit contraction and recession or depression.
Eichengreen and Portes (1987) postulate that the banking system serves as a buffer between the real sector and the wider financial system. As a result of deregulation, however, banks are becoming increasingly integrated with the wider financial system (Mullineux 1987b, c). This has a number of regulatory and supervisory implications. The risk of 'contagion' is increased, in the sense that banks are more likely to be adversely affected by depositors' reactions to losses in securities markets in which banks are known or believed to be participating. To the extent that securities markets, as well as the real sector, are a source of shocks, then the banking system's buffer role will be expanded. The LOLR cover provided by central banks to support the banks' buffer function, and thereby to protect the payments system and the credit relationships it entails, will implicitly have to be extended to cover the wider financial system in order to protect banks from 'contagion'. (See Dale 1988 for further discussion.)
The discussion of the role of money and finance in pre-Keynesian business cycle theories in section 3.2, and of the FIH in section 3.3, revealed a tendency to concentrate on the behaviour of banks rather than the wider financial system. The implication appeared to be that there is something special about banks. It has been asked, however, if banks differ from other financial intermediaries because they are more heavily regulated or if the regulatory strictures are applied because they are different. We have no space to explore such issues here but, to the extent that the financial system increases the ability of the economy to bear risks and in so doing allows more investment and faster growth to take place,52 then it has a key role and its ability to absorb and manage risk must be assured. This provides a rationale for regulation. In most financial systems banks continue to play a key role, despite the apparent shift away from traditional bank loan finance towards funding through marketable securities. By virtue of their risk-bearing or insurance53 role, they probably remain the most important shock absorbers or buffers in the financial system. This implies that in order to ensure that their buffer role is not impaired, it is particularly important to regulate banks.
If the banks do indeed perform a key buffer function, then the role of the banking sector in the economy is more complex than envisaged by Schumpeter (1939) and Shackle (1938), whose work will be discussed in the next chapter, and Minsky. Schumpeter and Shackle assumed that banks provided finance for innovatory investment and a secondary wave of expansion, while Minsky argues that banks engage in speculative and Ponzi finance. It may well be that financial fragility develops as the expansionary phase is sustained, as Minsky argues, and that this creates conditions that ensure a more severe downswing than would otherwise have occurred, as Schumpeter (1939) argued. If this is the case, then bank regulation, aimed at curbing speculatory excesses and maintaining the banking sector's ability to perform its buffer function, would be an essential component of anticyclical policy.