Accounting Article

Efficient Market Hypothesis - Raising Eyebrows

by Yasir Khan | Published on 12/13/2004

---

What looks like a small project turns into a big project. What looks like a big project turns into a small project. So I don’t think there’s been a lot of planning involved. At least in my case, there hasn’t been.

These words by Eugene Fama are for his work on Efficient Market Hypothesis. Even after enormous work in the fields of computers, finance, economics, corporate sector and so much so Nobel Prize 2003 nomination for economics, only EMH became his identity. His Ph.D dissertation in 1964 played havoc with entire investment advisory markets and it continues even today after 40 years. A theory named Random Walks became a cause of random talks in capital markets all over the world. Making investment in index funds and spending all research time on the beach enjoying holiday is an efficient way introduced by Fama.

The EMH is notionally based on the theory of random walks, which gives an idea of no prediction role of analysts and chartists in stock markets. The security price changes are not accurately predictable under any mechanical or technical method introduced by analysts well before and after EMH. The empirical testing of EMH helps to clarify this fact.

As per EMH, the stock markets are categorized into three types of weak, semi-strong and strong, and all such categories are based on information availability and its reflection in security prices. The existence of weak form market has just become a pretense in the modern world but its existence in the 19^th century is not so far behind. The very existence of weak-form market laid the foundations of preliminary Dow Theory.

The Dow Theory has its origin back in 1884 when it was presented in anonymous editorials of The Wall Street Journal by its architect Mr. Charles Dow.  His work was later publicized by Mr. S.A. Nelson in 1902 in his book The ABC of Stock Speculation. The Dow Theory was presented on the premise of weak form capital market behaviour prevailing in the investment world at that time. The theory is based on the view that stock prices exhibit very significant factors, which affect supply and demand forces in the market. Such factors are considered to be the volume of trade, fluctuation in exchange rates, commodity prices, bank rates and so on. It is worth noting that this theory is applicable not only to share prices but also to foreign exchange rates. Dow believed that the prices exhibit a set pattern on the grounds of History Repeats Itself, and they are also representative of market psyche. The same approach became the base of technical analysis in weak form markets at that time when the corporate sectors were not interested in exposing their assets, cash and debt position to investors. In the absence of such information, the prediction of future exchange prices was confined to the study of past price patterns. In this way, the Dow theory efficiently tackled the problem of investment management at the infancy stage of capital markets.

The time passed away and the capital markets moved from weak to semi-strong form on account of enhanced disclosure requirements of both the corporate sector regulatory authorities and stock market regulations. At last, the semi-strong markets began to prevail all over the world, making weak-from market a bygone event. But the Dow Theory didn’t breathe its last and remained popular even after the emergence of new scenario. The time came when Fama’s theory made a breakthrough followed by immense criticism all over the world.

The markets presented by Fama are nothing but subsets of a great balloon and all this depends on the information transformation into share prices. This phenomenon is called operational efficiency of the market. Each advanced level market encompasses the junior but the reverse is not applicable.

Outperforming the market is the subject matter but the market efficiency makes it impossible for an investor, thus leading to passive fund management instead of active fund management. In weak market, the technical or statistical analysis is all about the future price prediction. In semi-strong, the fundamental analysis gives a way to predict future prices, but the availability of all public information to all market players makes it impossible to beat the market. The use of insider information is legally prohibited in all stock markets and the user of insider information is also answerable to specific exchange regulatory or corporate regulatory authority. The strong-form markets, therefore, do not exist.

One implication of the efficient market hypothesis is that abnormal return, through outperforming the market, is not possible. The abnormal return over the normal return is commonly calculated by deducting the expected return form the actual return. A mathematical formula, as per CAPM theory, is used while determining the return expected from a security or a portfolio of securities. Such a model is called Market Model, which is widely used to assess the efficiency of market.

R_t = a_t + b_t R_mt + e_t (expected result)

Excess Return = Actual Return – Expected Return

Where e_t = Actual – (a_t – b_tR_mt)

Where the estimation techniques do have predictive value, the existence of excess return may be assumed to be the non-existence of random walks. But the marginal existence of excess return does not provide the strong evidence to reject Fama’s view. One important thing to remember is that the excess return, as computed above, excludes the transaction costs. The existence of transaction costs may well nullify the presence of excess return.

The empirical testing of EMH is mainly based on the implications of chartists’ techniques to predict future from the past. Where the correlation coefficient of successive price changes approaches to zero or is too much insignificant, the EMH holds good and chartists’ techniques are declared to be akin to astrology by Fama.

As per serial correlation technique, the Random Walk hypothesis suggests that the following relation should govern the prices

P_t = m + P _t-1 + e_t

Where E (e_t) = 0 and Cov (e_t,e_t-k) = 0

This implies that returns are independent.

Such kind of statistical tests applied by Fama and other researchers indicated the presence of some correlation in asset return at both the short and long horizons. However such serial correlation found was very small with respect to trading cost of implementing the strategy. These tests were also rejected due to some statistical problems embedded in them.

In addition to these statistical techniques, the Runs Tests are also very famous. In simple words, the statistical techniques are often used to test any presumption called hypothesis. Different tests are used to test the hypothesis. Many statistical tests assume a set parameter or normality of data to be used for testing any hypothesis. Such tests are parametric tests. Whereas to overcome the demerit of any preliminary assumption about the data and hypothesis to be tested, few statistical tests have been presented by statisticians that entirely ignore any preliminary supposition about the data to be used and hypothesis to be tested. These tests are called Non-parametric tests in statistics literature. The Runs Tests are one of these non-parametric tests.

A Run is a combination of similar elements or signs (say +VE or –VE). Where share prices rise or fall consecutively, the positive (+) and negative (-) signs together make up a run. The presence of too much runs in a sequence support the randomness of data. The total number of signs every moment the share price falls or rises also helps to clarify the randomness or non-randomness of data. Consider the following patterns.

    - - -              + + + + +          - - - - -            ++        - -

     R1                     R2                  R1                R2        R1

In the above example, Run (R1) has symbolized negative signs and Run (R2) has specified positive signs. Run (R1) contains 3 negative signs which may be interpreted as consecutive falls in share price. Whereas, Run (R2) specifies rise in share prices 5 times in a line. In this way runs tests help to identify the presence or absence of randomness of data. The complete procedure of applying the runs tests is not of our interest here. This procedure incorporates the use of test statistics and use of critical values table for accepting or rejecting a null hypothesis.

Fama himself and many other analysts performed these runs tests and identified a very weak set pattern in the share price changes over time. The results of such tests, also interpreted by Fama in his dissertation, strutted the Random Walk Theory.

The main criticism on the statistical techniques is that the serial correlation coefficient and runs tests are not very sophisticated to capture the complicated set patterns embedded in price changes. Moreover, the opponents, on the premise that any temporary downward change in price of a security immediately cancels the previous upward change and the pattern concealed in the entire data remains undiscovered, rejected the runs tests.

One more technique, which is widely used to foretell the future share prices in the stock exchange and for making decisions to actually outperform the market, is filter test technique. The filter test technique defines the rules about the correct timings of purchase and sale of shares in stock exchange. The filter tests technique was presented when the weak-form markets were in existence. While using filter rules, an investor buys or sells the stock when the share price exceeds or falls below filter limit of 5 %. The filter test technique was presented by Alexander and was criticized on the grounds of arbitrary filter limit of 5%. The results of filter tests, as interpreted by Fama in his theory, are also not much impressive to support its validity to beat the market consistently.

The semi-strong markets prevailing all over the world support the efficient theory. The share prices in the capital market exhibit the company’s attributes like P/E ratio, proposed dividends and other financial and non financial information disclosed by the enterprise. Whereas, no use of insider information serves as a hurdle towards absolute efficiency of market.

The efficient market hypothesis presented by Eugene Fama has been appreciated all over the world. The empirical testing of EMH has no absolute support for this theory to be authenticated. But even after much debate, EMH is considered to be the best way to describe the zigzag dances of stock prices in any market. Most of the empirical testing results showed short memory (where the results support technical analysis in short run) while several times, the result of such empirical testing also exhibited long memory position (where the technical analysis supports technical analysis in the long run). Even after such evidences, EMH made its way, as the results were not much impressive to lead to active investment strategies instead of passive investment strategies. The importance of EMH in making decisions about share prices, therefore, can hardly be questioned.

In Pakistan, no study in this respect has ever been conducted by professionally qualified accountant or capital market brokers. Practically, the share prices are greatly affected by random factors like political instability, economic recessions and many other factors. All such factors affect the share prices in the random manner. Theoretically, such factors should not affect the share prices in the stock market unless the financial impacts of such favourable and unfavourable factors do have implications. Moreover, these implications are only relevant in the efficient markets where the players in the market make rational decisions. The irrational behavior of investors prevails and such behavior results in the inappropriate decisions. Such decisions affect the supply and demand forces in the market and the share prices, in consequence, expose random walks. In this way, the random walks is an idea that really works.

The efficient market is defined as “A market having a large number of rational profit maximisers, actively competing with each trying to predict future market values of individual securities, and where important current information is almost freely available to all participants”. The definition covers all practical aspects normally a capital market possesses. But one major aspect of rationality is not always in existence. And this most important decision making factor may cause the investors to sell and buy overpriced and underpriced securities, while in reality it may not be the way to outperform on account of securities being fairly priced due to market efficiency. Moreover, in recent years, much work has been conducted in the field of Behavioural Finance. This emerging field is concentrating on studying emotional and cognitive factors, which affect decisions and lead to irrationality as Albert Einstein says “Only two things are infinite, the universe and human stupidity, and I am not sure about the former”.

On the other extreme, we cannot totally reject the view of active investment management. Fama’s work, no doubt, has opened new horizons in the field of investment management but the existence of Investment advisory organizations all over the world has also proved that a knowledgeable investor performs better than a passive investor. The reality is neither black nor white but it is something like gray.

The author, Yasir Khan works in Mahboob Sheikh and Co., Multan. He can be contacted at yasirkhan80@hotmail.com

Article courtesy of Yasir Khan

---