Read Manias, Panics and Crashes text version

Manias, Panics and Crashes

1. Definitions

The three words in the title of this section - 'manias', 'panics' and 'crashes' - make up the title of Charles P Kindleberger's book on financial crises1. Let us begin with some definitions. One of the dictionary meanings of 'mania' is 'excessive or unreasonable desire: a craze'2. Applied to financial markets, we are talking about a frenzied burst of buying of a financial asset, which causes the price of that asset to rise very sharply. In Kindleberger's phrase, a mania is 'speculative excess'. The implication is that people buy on the assumption that the price will continue to rise and thus they will be able to sell at a higher price later, making a speculative profit. People behaving in this way appear to forget the possibility that the price might fall. They dismiss potential losses from their minds. This collective refusal to acknowledge, at least in the case of one particular asset, that prices in markets might fall as well as rise is against all market experience. It can, thus, be held to be 'unreasonable', as in the dictionary definition or 'irrational'. The use of the words 'desire', 'craze' and 'frenzied' here suggests that other motives, notably greed and rapaciousness, take over and dominate rationality, although people may also be sucked into the craze by a defensive fear that they are missing out on something. The sharp rise in price that results from 'mania' is known as a 'bubble'. This implies that this big increase in demand has no solidity: that there is no logical basis for the high price. In modern economics language, we would say that the price had become detached from its 'market fundamentals'. That is, in rational markets, prices should be determined by real economic variables. There will always be differences of opinion as to what those real economic variables are, but the idea is that prices must have an objective basis. It is possible that during a 'mania', people do not entirely forget the 'fundamental' price of the asset and they are aware of the losses they stand to make if the price were to fall to that level. Thus, as they continue to buy at higher and higher prices, they are aware that they are taking ever greater risks. They can only justify these increasing risks by the prospects of ever-greater profits. In these circumstances, it can be argued, it is rational to go on buying and, indeed, to buy in ever greater quantities, causing the price to rise even faster. According to this view, the situation has become unreal but people behave rationally within it. We shall return to this idea later. The use of the word 'bubble' also implies that the bubble might easily burst - that the price might collapse just as quickly as it rose. The obvious reason for this is that the price has risen simply because of an unsupported belief that the price will continue to rise. Any doubt that this will continue will cause some people to stop buying. As soon as the rate of increase in demand slows down, the price rise slows and people begin to sell in order to take their profit. The price starts to fall. The mania has ended. This leads people to ask how far the price might fall and, temporarily, leads to an assessment of the fundamental basis of the value of the asset. They then become aware of the extent of the risk they are facing. Alternatively, if they were always aware of the degree of risk, they now realize the extent of the danger to them as individual investors. It becomes imperative to sell before the price falls very far. Everyone rushes to sell, the price starts to fall very sharply and 'panic' sets in. The dictionary3 gives one meaning of 'panic' as 'a state of terror about investments, impelling men to rush and sell what they Charles P Kindleberger (3rd edn., 1996), Manias, Panics and Crashes. A History of Financial Crises, New York: John Wiley and Sons 2 R M Kirkpatrick (ed.) (1983), Chambers 20th Century Dictionary, Edinburgh: WR Chambers Ltd. p. 765 3 ibid., p. 916


possess'. The dictionary also refers to 'contagious fear'. This again suggests that people stop behaving as individuals and adopt 'herd' behaviour. Even those who have no idea of the fundamental value of the asset take part in the frenzied selling. The price might be driven below any possible idea of a 'fundamental value' of the asset. The market 'crashes'. Kindleberger refers to 'revulsion' from speculative excess as leading to panics and crashes. This introduces almost a moral element into the picture - that people are prepared to accept a great degree of unreality but that, at some point, an attachment to real values reasserts itself. Alternatively, we might argue, that people are ultimately repelled by the greed which becomes obvious in bursts of manic buying. This would seem to imply that manias are inevitably followed by crises4. 2. Are manias, panics and crashes possible? Modern economic theory is constructed on the assumption that economic behaviour is always rational. That is, it is assumed that people act in their own self-interest and that they always act to maximize their own utility, within the constraints imposed on them. These constraints might take many forms. Most obviously, people's economic behaviour is constrained by the stock of capital they possess and by their natural abilities, education and training, which strongly influence the income they can obtain. They are constrained also by the actions of governments and by the behaviour of other market agents, which help to determine the wage rates they can obtain for their work, the prices they must pay for goods and services and the rate of interest they receive on their capital or must pay in order to borrow. The application of this conventional economic theory to the analysis of market behaviour produces the notion of market efficiency, which we discussed briefly earlier in the course, where we distinguished between operational, allocational and informational efficiency. The financial markets literature is largely concerned with informational efficiency - the idea that market prices are based on the best information available. In a fully efficient market, everyone has access to the best available information and speculation is not possible. It follows that, if financial markets are fully efficient, manias and panics cannot occur. Economists who support the efficient markets hypothesis reject, therefore, the historical approach of writers such as Kindleberger. Instead, they spend their time on complex econometric tests of the efficient markets hypothesis. In his introduction to the third edition of his book, Kindleberger notes, '...the appearance of the book by Robert Flood and Peter Garber5, with a forceful presentation in a number of papers of the conclusion that the null hypothesis of no bubbles could not be rejected by econometric tests ..., and in particular in Garber's paper, "Tulipmania," that the high price of tulips in the Dutch Republic, which soared from the fall of 1636 to February 16376, was not a mania, for the higherpriced bulbs at least, but a rational market response to fundamentals.' Kindleberger, op. cit. p. ix Thus, we have a conflict between those who see financial crises as being a common feature of financial market behaviour and as indicating weaknesses in the behaviour of markets and those who argue that markets always behave efficiently. This has important implications for economic policy. Kindleberger and other writers, who recognize the existence of irrational behaviour within markets, see the need for some form of government intervention to overcome market failure or to correct the undesirable results Kindleberger contents himself, however, with saying that they are 'if not inevitable, at least historically common', op.cit. p. 2 5 Robert D Flood and Peter M Garber (1994), Speculative Bubbles, Speculative Attacks and Policy Switching, Cambridge, Mass.: MIT Press. 6 See Appendix 1


that markets sometimes produce. The supporters of market efficiency argue, on the other hand, that markets should be left to operate freely and that any 'undesirable results' must stem from the nature of the constraints placed upon markets, notably by governments. Thus, for market efficiency supporters, governments cause market problems, rather than counteract market failures. Our next step, therefore, is to look more closely at market efficiency theory and at the testing of it. We shall then look at arguments produced within conventional economic theory, which attempt to explain 'bubbles'. We shall do this here in relation to foreign exchange markets, although the principles apply equally to all financial markets. 3. The application of market efficiency to foreign exchange markets According to supporters of efficient markets, foreign exchange markets should follow a number of rules. These rules are known as interest rate parity, purchasing power parity and the international Fisher effect (also known as Fisher open). Let us consider each of these in turn. 3.1 Interest rate parity We have seen in seminars that the foreign exchange market consists of two distinct markets - the spot and forward markets. In the spot market, currencies are bought and sold for delivery within two working days. In the forward market, exchange rates are agreed for the delivery of currency later, say one month or three months later. Forward foreign exchange rates are quoted as being at either a premium or discount to the spot exchange rate. If a currency is at a forward premium, it is more expensive to buy forward than to buy spot. If it is at a discount, it is less expensive forward than spot. What determines the relationship between the spot and forward rates of exchange? That is, why are some currencies at a forward discount and others at a forward premium? Consider the position of two investors, X and Y. Each of them has one million Euros to invest in a secure form for three months. X buys German government securities at current Euro interest rates. Y sells her euros for sterling in the spot foreign exchange market and buys sterling securities, the interest rate on which are higher than on equivalent euro securities. At the end of the three months, the British securities mature and Y sells the sterling for euros in the spot market. Who finishes with more euros at the end of three months? Clearly, there are only two factors involved: (a) the difference in interest rates on UK and German securities; and (b) the difference in the £/Euro exchange rate at the beginning and at the end of the three months. Since interest rates in the UK are higher than in Germany, if there were no change in exchange rates during the period, Y would finish up with more euros. However, X could finish with more euros, if the value of sterling against the Euro fell sufficiently during the three months to more than outweigh the higher interest payments received by Y. If the strategies of X and Y are thought likely to produce the same result, we have uncovered interest parity. The word 'uncovered' simply indicates that Y would be taking a risk in following her strategy, because she does not know what is going to happen to the exchange rate over the ensuing three months. The expected future spot rate of exchange is crucial to her decision. ______________________________________________________________________________ Uncovered interest parity when the gains from investing in a country with a higher interest rate are equal to the expected losses from switching into that country's currency and back into the original currency ________________________________________________________________________________

All foreign exchange transactions in the above example take place in the spot market. However, the existence of forward exchange markets allows a third strategy, one that overcomes the risk associated with uncovered interest arbitrage. A third investor, Z, buys sterling securities and, at the same time, sells sterling three months forward at an agreed rate of exchange. He is, thus, able to calculate exactly how many euros he will receive when the sterling securities mature in three months' time. Hence, he can make a precise comparison between the number of euros he would receive from buying UK securities and from leaving his money invested at home in German securities. This comparison will be influenced by: (a) the difference in interest rates on UK and German securities; and (b) the difference between the spot and 3-month forward exchange rates for sterling against the Euro. It follows that, for Z's strategy to produce the same results as Y's strategy, these two must be equal. In other words, the forward discount or premium must be equal to the difference in interest rates on the two currencies. We can show that if sterling interest rates are higher than euro interest rates, this difference must be balanced by sterling being at a forward discount against the euro. If the discount on three months forward sterling were less than the difference between the two interest rates, investors would sell German securities (forcing their price down and pushing the yield on them up); buy sterling spot (forcing up the spot exchange rate of sterling; buy UK securities (forcing their price up and the yield on them down). They would then cover their exchange rate risk by sell sterling three months forward (forcing down the three-month forward exchange rate of sterling). Thus, the interest rate differential between UK and German securities would be reduced and, at the same time, the discount on forward sterling would increase. This would continue until the interest rate differential equalled the forward discount on sterling. This position is known as covered interest parity. ______________________________________________________________________________ Covered interest parity when the gains from investing in a country with a higher interest rate are equal to the forward discount on that country's currency ________________________________________________________________________________ We have thus established a relationship between spot and forward exchange rates. Several other questions follow, however: (a) Why do interest rates differ among countries? (b) What determines the existing spot rates of exchange? (c) What causes spot rates of exchange to change over time? In fully efficient markets, each of these may also be answered by a rule based upon arbitrage operations. 3.2 Differences in interest rates among countries - the Fisher effect Nominal rates of interest consist of two elements: (a) the real rate of interest; and (b) the expected rate of inflation. Thus, differences in expected inflation rates provide one cause of differences in international interest rates. This still leaves us to explain differences in real interest rates. However, if we ignore exchange rate risk and assume perfect capital mobility and perfect information, market theory tells us that capital should move from capital-rich countries in which the real rate of return on capital is low to capital-scarce countries with high real rats of return on capital. This movement should continue until real rates of interest are equal across countries. In this case, nominal interest rates would differ from one country to

another only because of differences in expected inflation rates. This is sometimes known as the Fisher closed hypothesis. 3.3 The determinants of spot exchange rates - purchasing power parity If we do assume that real interest rates are equal across all countries and that differences in nominal interest rates simply reflect differences in expected inflation rates, we can assume that there will be no movements of capital internationally. The balance of payments balance will depend solely on the export and import of goods and services. If we further ignore differences in the quality of traded goods, and continue to assume perfect information, the only basis for choosing between foreign goods and home goods, and hence the only cause of flows of currency from one country to another will be differences in price. However, if, at the existing exchange rate, goods are cheaper in the USA than in the UK, UK citizens will switch to US goods. To do this they will sell sterling and acquire dollars, forcing down the value of £ relative to the $. This process of goods arbitrage should continue until prices in the two countries expressed in a common currency are equal. This is the essential principle of purchasing power parity (PPP). In this form - absolute PPP spot exchange rates in equilibrium are a reflection of differences in price levels in different countries. More importantly, changes in spot exchange rates will reflect differences in inflation rates among countries. This is known as relative purchasing power parity. Notice that differences in expected inflation rates are now equal to two things: the expected change in spot exchange rates and the difference in interest rates. It follows that in equilibrium, differences in interest rates must equal the expected changes in the spot rates of exchange. This equality is sometimes known as the international Fisher effect or the Fisher open hypothesis. 3.4 Market efficiency and the foreign exchange markets Next, we must distinguish between forward rates of exchange and the future spot rate of exchange. The one-month forward rate is a rate agreed now for a delivery of currency in one month's time and thus is known. The future spot rate one month ahead is what the spot rate will be in one month's time and is unknown. However, having made our assumption of perfect markets, we have established a link between the two. Covered interest parity establishes a link between interest rate differentials and forward rates of exchange. The combination of purchasing power parity and the Fisher effect establishes a link between interest rate differentials and expected future spot rates of exchange. It follows that if people behave rationally and are perfectly informed and all the other assumptions of perfect markets hold, forward rates of exchange will accurately predict movements in spot rates of exchange. Of course, we cannot (even with all of the assumptions above) claim that forward rates of exchange will predict future spot rates perfectly accurately. After all, news will come to the market, taking all market participants by surprise. Nonetheless, if the ideas of market efficiency are correct, we should be able to say: a. News will occur randomly, with an equal chance of causing the value of a currency to rise or fall. Thus, errors in our forecasts of future exchange rates will occur but will cancel out over time. Forecasters should never be systematically wrong. b. Random and unpredictable news will be the only factor influencing changes in spot rates. Existing spot rates of exchange and changes in them will be of no relevance. The volume of trading in the market is also of no interest. c. News, when it does come to the market, will be equally available to everyone and will immediately be taken into account. Thus, no one will be able to make excess profits from the market.

d. Volatile exchange rates must be the result of rational revisions of the market's perception of the equilibrium exchange rate because of news relating to market fundamentals, such as changes in tastes and technology. 3.5 Efficient markets and bubbles If the efficient markets hypothesis holds, however, how can we explain the existence of bubbles in the foreign exchange and other markets? Economists developed a set of models that attempted to explain sudden and apparently inexplicable jumps in the value of a currency through the phenomenon of rational bubbles. Such models typically start in a disequilibrium position. The models then set out to explain why rational decisions may cause the market to move further away from equilibrium rather than returning to it. For example, in trying to explain the inexorable rise in value of the US dollar between 1981 and 1985, Dornbusch7 started with an overvalued exchange rate of the dollar. Investors were assumed to be risk neutral - that is, they are concerned only with the expected value of their profits not with their dispersion around the mean. In other words, a strategy which has a risk of high losses if things go wrong but a potential for high profits if they go right is equivalent to one in which potential losses and profits are both low. Thus, in Dornbusch's example, investors had to compare two probabilities: · that the exchange rate would return to equilibrium; and · that it would go on rising. The further the exchange rate was currently above the equilibrium rate, the greater was the potential loss for investors if it fell back to equilibrium; and the greater the required profits would need to be if the rate kept on rising. To put it another way, the greater was the risk of a crash, the faster the rate of appreciation needed to be to compensate for potential losses. Investors were thus obliged to go on buying the currency pushing the rate up further and further, although there was no economic justification for it. Another common approach is to is to allow for the existence of two kinds of forecasters in the market. In Goodhart's model8, for example, dealers based their decisions on a weighted average of the forecasts of market efficiency theorists and modellers of fundamentals, with the weights determined by the relative past success of the two forecasts. Again, the model assumed an overvalued exchange rate to begin with. In the absence of news, the market efficient forecast was for no change; while the fundamentalists predicted that the exchange rate would fall to equilibrium. Goodhart assume next that a random shock forced the exchange rate further away from equilibrium. Both forecasts were thus wrong, but the market efficiency forecast was closer to being correct. Therefore, in the next period, the weights were changed to reflect this, causing the predicted fall in the exchange rate to be smaller. Dealers were then required to buy more dollars, forcing the rate up yet further. This leaves us to explain why the market moved away from equilibrium in the first place, but this can be easily done. As we mentioned earlier, this could result from the intervention of governments or central banks. 3.6 The testing of the efficient markets hypothesis


Dornbusch R (1984), `The overvalued dollar', Lloyds Bank Review, 152, April


Goodhart C A E (1988), 'The foreign exchange market: the random walk with a dragging anchor', Economica, 55, 437-60

Clearly, to understand the operation of foreign exchange markets, it is important to know whether market efficiency holds. Thus, econometric tests are carried out to see whether the forward exchange premium is an unbiased predictor of the corresponding exchange rate change. This implies that people behave rationally and are risk neutral. If, however, they are risk averse, the forward rate of exchange will be a biased predictor - the future spot rate will not be equal to the forward rate because speculators will demand a risk premium to allow for the risk associated with assuming an uncovered position in the market. What have been the results of these many tests? They have been overwhelmingly against the proposition that the forward premium on foreign exchange is an unbiased predictor of the future exchange rate change. That is, they have been against the joint hypothesis of rationality and risk neutrality. The simplest explanation of this failure would be to accept the proposition that investors are risk averse and that there is thus a risk premium. For example, Isard's (1987) study9 suggested that, the forward premium on non-dollar currencies against the dollar was consistently less than the expected depreciation of the dollar against other major currencies, implying the requirement of a risk premium on non-dollar currencies. This was as high as 10 per cent in the case of the French franc. In other words, French interest rates needed to be significantly higher than US interest rates in order to attract capital into the franc, despite the fact that the franc was expected to gain in value against the dollar. In general, however, the empirical literature is not strongly supportive of the risk premium argument10. This has led to the questioning of the assumption of rationality. That is, people might behave irrationally in markets or, in the jargon, they may be suboptimal processors of information. This view has been strengthened by the rejection of the efficient market hypothesis also in the stock market literature.

P. Isard (1987), 'Lessons from empirical models of exchange rates', IMF Staff Papers, Vol. 34 no. 1, pp. 1-28 10 see C. Paul Hallwood and Ronald MacDonald (2nd edn. 1994), International Money and Finance, Oxford: Blackwell, Ch. 11.


Appendix One Tulipomania11 Tulip bulbs were introduced into the Netherlands from Turkey in 1594. From the early years of the seventeenth century, they became a status symbol and the subject of a collecting craze. This applied particularly to bulbs with a rare virus infection, which produced fabulous, often unique streaking patterns on the petals of the flowers. A buying craze developed. Their prices rose sharply and, by the 1630s, they were being bought for speculation. The inexperienced middle classes flocked into the market, just before it reached its peak. Travelling salesmen spread the frenzy to remote villages. 'Colleges' were established in the taverns of provincial towns to allow people to bid at evening auctions. Futures contracts and, later, options in tulip bulbs were introduced. Stories tell of the very high prices, often in kind, paid for single bulbs. The mania reached its peak in 1637, when prices were almost doubling every day. One great rarity (a Semper Augustus bulb) fetched 6,000 florins (equal to £500,000 in today's money). The crash followed and, by 1639, a bulb that sold for 5,000 florins in 1637 was worth only one-tenth of a guilder. The case for rational behaviour in the tulip market rests on the propagating ability of the rare bulbs. One bulb may, over a hundred years, produce a million bulbs. Each of these bulbs would be worth only a fraction of the value of the original bulb. None the less, the future earnings from the bulb could be expected to be very high. This might explain part of the sharp increase in prices, but it is difficult to believe that it could explain what was happening at the peak of the boom. The crash was triggered by actions taken on 2nd February 1637 by officials in Haarlem, who feared that the tulip hysteria might destabilise the local economy. The bubble burst and almost overnight bulbs became unsaleable. Ruin followed for many. World-renowned merchant banks in Amsterdam were crippled. Many lawsuits were entered into. The judges decreed that debts on futures and options contracts fell into the category of gambling and were not recoverable in law. The Netherlands slid into depression.


This account is taken entirely from: C P Kindleberger (1996), op. cit. pp. 100-101 and Bernice Cohen (1997), The Edge of Chaos. Financial Booms, Bubbles, Crashes and Chaos, Chichester: John Wiley and Sons


Manias, Panics and Crashes

8 pages

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate


You might also be interested in

Steer cover PM
Microsoft Word - 06 Homa Note - Marketing Performance Mgnt r030708.doc
Engleski rjecnik Strucni engleski za ekonomske definicije