Stock Market Volatility - A Psychological Phenomenon?

David Barrett - Senior Sophister

The volatility of stock prices is a well known phenomenon to all investors. Why, though, is this volatility so pronounced? Can the Efficient Market Hypothesis account for such major market realignments as the stock market crash of October 1987? David Barrett assesses these questions and concludes that some 'herd-like instinct' seems to be undeniable and that self-fulfilling prophesies are not altogether figments of the imagination. These psychological factors can then live alongside standard economic influences in determining stock prices.
This essay examines the whole issue of stock market volatility. It is apparent that there are extremely wide day-to-day changes in the prices quoted on most stock exchanges. Many people have tried to put forward theories to explain this phenomenon and more still have tried to use these theories in order to predict future changes in prices. However, most economic theorists have ignored the fact that there is no universally accepted body of work explaining what is behind these day-to-day price changes. Instead they have concentrated on market details in the mistaken belief that the question has already been answered. This is not the case. In this essay the author will attempt to critically examine the two main schools of thought on the subject and, hopefully, will find evidence to support one of them. As it stands theorists cannot agree on whether or not it is economic or psychological realities which are the major cause of price fluctuations in the stock market. This is an important issue in the study of investment analysis as it brings into question the whole realm of fundamental and technical analysis, something on which millions of pounds are spent every year. If it could be proved that there are no sound economic reasons for price changes in the stock market then these two forms of analysis could be rendered worthless. Initially, the author will discuss the readings that have been found relevant to the topic of the essay. Then the author will explore his own thoughts on the subject with the help of an analysis of the greatest period of market volatility in recent times - 'Meltdown Monday' or the stock market crash of 1987.

Amongst the literature of most relevance to the whole volatility issue is Robert Shiller's 'Market Volatility'. Shiller is a firm advocate of the popular model explanation of stock market volatility. Popular models are a qualitative explanation of price fluctuations. In short, it proposes that investor reactions, due to psychological or sociological beliefs, exert a greater influence on the market than good economic sense arguments. It should be noted however, that Shiller does not totally disregard the work of economists before him who proposed the Efficient Market Hypothesis (EMH). In fact, he admits that the EMH can be substantiated by statistical data but he believes that investor attitudes are of great importance in determining price levels(Schiller)His book provides statistical evidence that excess volatility exists in the stock market and therefore volatility cannot be totally explained by the EMH. Excess volatility is the name given to that level of volatility over and above that which is predicted by efficient market theorists. In Shiller's eyes this excess volatility can be attributed to investors' psychological behaviour. He claims that substantial price changes can be explained by a collective change of mind by the investing public which can only be explained by its thoughts and beliefs on future events, i.e. its psychology.

The popular models theory, according to Shiller, proposes that people act inappropriately to information that they receive. Thus freely available information is not necessarily already incorporated into a stock market price as the EMH would have one believe. Shiller examines in some detail the underlying assumptions which form the basis of the EMH and attempts to discredit them. He says that investor behaviour depends on ex-post values, which is the value of an asset taking into account the future dividends. By definition though, ex-post values cannot be known ahead of the payment of dividends and so if future dividends are expected to be high then the ex-post value today will also be high. So if investors knew the future dividend then forecasting the future price (Pt) would present no problem, according to the EMH, using ex-post values (Pt*):

Pt = EtPt*

In other words price equals the best possible forecast or expectation of ex-post value. It should, however, be noted that that capital gains or losses in a share (i.e. price fluctuations) have no effect on ex-post values as true ex-post values only reflect the payoffs that the investment itself produces. If the efficient market is a reality these gains or losses are just related to changes in information about ex-post values. If the EMH does not hold, then these gains or losses have nothing to do with ex-post values and may simply be a reaction to other investors' actions.

The question mentioned earlier of excess volatility is addressed fully in an article by Shiller in the 1981 June issue of 'The American Economic Review' entitled, simply 'Do Stock Prices Move Too Much To Be Justified By Subsequent Changes In Dividends?'. The efficient market theorists claim that the EMH can be used to explain sudden movements in price. For example, new information about dividends could be released. Shiller's argument is that the fluctuations are far too big to be accounted for by mere changes in information. He provides statistical evidence to suggest that fluctuations in dividends, (which determine ex-post value) due to their nature of being calculated on a moving average, would have to be quite substantial both in terms of size and length of a trend to resemble the fluctuations in price. Even during the great depression of the 1930s dividends were only significantly below their IR growth path for four years(Schiller)The mathematics of a moving average would smooth this out.

In the article Shiller also addresses the debate as to the use of dividends or earnings in calculating ex-post values. It has been argued that the use of dividends in this calculation does not accurately reflect the simple efficient market model. These theorists claim that the price should be the discounted present value of expected future earnings. Shiller claims that earnings are only relevant to the share price insofar as earnings are indicators of future dividends. He says that earnings are, in reality, only an accounting concept and, as has been proved, can thus be easily manipulated to reflect whatever the accountant wants to show about a particular stock. Earnings bear no relationship to revenue which is what the average investor is concerned with. By contrast this is exactly what dividends do. Another argument which has been put forward to discredit the use of dividends in ex-post calculations is that they are an arbitrary figure. A firm simply decides, for whatever reason, to pay out a certain level of dividend. Thus dividends are only a fraction of earnings and are not representative of the value of a stock. Shiller admits that this argument is a valid one but only in the case where no dividends are paid. If any level of dividend is paid one can support their use. Over time, according to Shiller, earnings are so heavily discounted that they contribute very little to the value of a stock. As such, dividends form the highest value to the shareholder as all that the investors are concerned with are the terminal price of a stock and the intervening dividends.

In the conclusion of his article, Shiller puts forward two explanations as to how the EMH can still hold with the excess volatility argument. It has been proven that stock market volatility in prices is five to thirteen times higher than the volatility which would be explained by the EMH and new information. Some efficient market theorists try to attribute this excess volatility to changes in expected real interest rates. Shiller claims though that the movements needed in expected real interest rates to explain this excess volatility are far larger that the movements in nominal interest rates over a sample period. The other argument in support of the EMH is that perhaps the fears of the investors are greater than the actual changes in price. There is no statistical evidence to support this and the behaviour of investors during the 1930's depression would seem contrary to this argument. Shiller does however admit that because this argument is an academic one based on 'unobservables' it can neither be supported nor discredited.

In support of Shiller's work is the fact that, at the same time as he was writing his article and totally independent of him, two economists, Stephen Le Roy and Richard Porter were conducting a study which had virtually the same conclusions as those of Shiller. They published their work in 'Econometrica' of May 1981 under the title 'The Present-Value Relation: Tests Based on Implied Variance Bounds'. In this study, which was an in-depth statistical study of excess volatility, Le Roy and Porter observed that based on aggregated and disaggregated data, stock prices are more volatile than the efficient capital markets model would suggest. This conclusion differs greatly from all previous work on the subject by such noted economists as Fama. The importance of the conclusions of this article, the author would suggest, is that it revealed very similar results of studies on stock price movements as Shiller and so neither articles' findings can be dismissed as statistical accident. This again lends more weight to Shiller's theory of the importance of popular models.

In John Dalton's book 'How the Stock Market Works' he suggests that in theory the market is indeed efficient. However, it is undeniable, he claims, that there are in practice several inefficiencies which are open to exploitation for profit. He cites two examples where this might be the case: Not all investors are equally well informed and so insider information can be used to one's benefit as long as no one else is in receipt of the information. This is at complete odds to the strong form of the EMH which claims that all information, both publicly and privately known is incorporated into the stock price. Dalton also says that investors react differently to the same information. Risk averse investors might sell as the market becomes bearish, more speculative investors might sell short to gain high profits while long term 'buy and hold' investors might see a market downturn as a chance to indeed buy low and hold. This, as Dalton sees it, is definite evidence that inefficiencies do exist in the market. Arbitrageurs are those who make a living by exploiting these inefficiencies but if we were to use chartism to predict optimal investment behaviour we would see that the optimum investment strategy would be to buy and hold as the market has risen overall since chartism began. However Dalton says that indices such as the Dow Jones cannot be used to indicate market trends as it is concentrated on big companies and does not reflect the more responsive small businesses (Dalton). This is in contrast to Shiller's insistence on using indices as market indicators in his book 'Market Volatility'.

Dalton also addresses the issue of fundamental and technical analysis in his book. He says that fundamentalists examine the environment surrounding companies and markets while the technical analysts feel this is not necessary as price movements can be predicted by historic movement. Thus while fundamentalists seem to be at odds with the theory of efficient markets (as information is the key to their actions), technical analysts do not try to beat the efficient market, they simply examine its past to predict any future trends.

At this point one feels that it would be worthwhile to examine the situation of extreme market volatility to test the theories outlined above. As such one chose to study the 1987 stock market crash in both Dalton's book and a special paper written by Mark Mullins on the subject. Dalton basically outlines the events of October 1987 and is therefore a good place to gain a basic understanding of the events surrounding the crash. He claims that by October 1987 the market had been bullish for so long that in peoples minds a bear period was unavoidable. So, in line with a 'self-fulfilling prophecy' brokers began to sell heavily, so much so that a meeting was called to decide whether or not to close the market for the day. Eventually it was announced that the exchange would remain open but this announcement prompted investors to think that the market was in danger of closing! There was a headlong rush to sell and by the end of the day it was clear that the market had performed badly. It wasn't until later that it was to become apparent how badly. The Dow Jones had fallen 508.32 points, that constituted 22.6% of the value of the entire stock exchange. Over 600 million shares had been traded (a record) and the value of the stock on the New York Stock Exchange had fallen by over $750 million. It is clear from Dalton's account that he considers the cause of the crash to be irrational investor behaviour, just as Shiller claims. The whole sequence of selling was started due to a psychological belief that things could not go on as they had been. According to Dalton there was no economic explanation for the high levels of selling which occurred. This is something which Mark Mullins does not agree with in his paper on the subject.

In his article 'Meltdown Monday or Meltdown Money: Consequences of the Causes of a Stock Market Crash' Mullins claims that the Meltdown Monday explanation is incorrect. This is the name given to the theory that technology and 'herd-like' behaviour of investors was the cause of the crash. He says that on the contrary, investors were reacting rationally to changing economic circumstances. He also cites government intervention in the financial markets as a cause of the crash. Mullins points out that leading up to 1987 there had been a five year bull market due to the economic recovery in the US and falling interest rates. From August 25th 1987, it can be seen that prices began to fall at a slow rate until they accelerated into a sheer drop on October 19th. The important point to note is that is that Mullins sees this crash as a rational reaction. This is in contrast to the 'Meltdown Monday' approach. The author, however, suggests that it would be inaccurate to view the crash as a bubble bursting. Price movements from 1982 to 1987 have been shown to be random in the US so there was no bubble to burst. If a bubble were to burst surely it would have occurred on August 25th when prices were at their highest. Thus, this argument has no grounding in statistical evidence (Mullins). However, an important point to note is that Mullins accepts that perhaps the crash could be explained by investors believing that a bubble existed and fearing that it was about to burst! Indeed, 38% of normal investors felt that the market was overvalued in the week October 12th to 16th, 1987. Thus we can see that although Mullins is a supporter of the EMH, Shiller's popular model argument can be applied here. Mullins defends the EMH by claiming that the assumptions about investor rationality, preferences and ability to act are too restrictive. If one relaxes these assumptions one can then examine the fundamentals within the context of the EMH.

Evaluation and Conclusion

In addition to the thoughts the author has expressed already while examining one's readings on the subject, one feels it is worthwhile to summarise the arguments put forward and this author's opinions of them. The author found Shiller's work to be by far the most convincing. Shiller established several inarguable points supporting his belief that the EMH cannot be the complete story. Excess volatility is an accepted phenomenon and cannot be adequately explained by fluctuations in real interest rates. In this essay the author has already shown that it is more likely that capital gains or losses can be attributed to the actions of individual investors, rather than movements in ex-post values of which the ordinary investor would have very little knowledge. The readings detailed above concerning the 1987 stock market crash serve to support this theory. The independent work of Le Roy and Porter shows that Shiller's work holds weight, in that two independent studies had almost identical results. One has found that there do exist unquestionable market inefficiencies and the simple fact that there are so many arbitrageurs in the market who profit by these inefficiencies proves, to this author, that the market cannot be fully explained by the existing theory of efficient markets. Even the work by Mullins (who was the only supporter of the EMH) can be interpreted as supportive of some of Shiller's findings.

However it must be recognised that the author, like Shiller, does not totally reject the EMH. There are substantial tomes of work by such noted economists as Fama which show statistical evidence to support the theory. The author would postulate that the theory is only an extreme of certain limited markets. While it does hold to a certain extent in all markets, it is not the full explanation. The excess volatility which has been examined has been adequately explained by the theory of popular models, seems to be totally logical in its assumptions and is well supported by several texts as the author has demonstrated. For this reason one's answer to the question posed by this essay 'Stock Market Volatility - Psychological Phenomenon?' would be that volatility is a function of both economic variables and psychological factors. The assumption though that the economic variables should be concentrated on to the detriment of research on psychological factors, is something the author has tried to oppose with this essay.

Bibliography

Shiller, R. J. (1990) Market Volatility - pages 1-4, 71-76, 197- 214, MIT Press, London

Schiller,R.J. (1981) Do Stock Prices Move Too Much To Be Justified by Subsequent Changes in Dividends? The American Economic Review, Volume 71, June, No.3 pages 421-435

Dalton, J. M. (1988) How The Stock Market Works 2nd edition pages 175-200, Institute of Finance, New York

Mullins, M (1988) Meltdown Monday or Meltdown Money:Consequences of the Causes of a Stock Market Crash - Special Paper Number 11 from the LSE Financial Markets Group Special Paper Series

Le Roy, S. F. and Porter, R. D. (1981) Tests Based On Implied Variance Bounds Econometrica Vol. 49, May, No..3 pages 555-574.