A Random Walk Down Wall Street: The Time-Tested Strategy For Successful Investing

I have enjoyed reviewing Prof Malkiel’s latest edition of his classic, ‘A Random Walk Down Wall Street’. This is a book that I highly recommend. Everyone will learn something, from the academic, to the expert finance professional and finally to the layman. Especially the latter will learn the most. But even the academic who may be very well versed in the finance literature has a lot to pick up from this book—from the historical review of different episodes in the history of the markets, to a great review of the relevant academic literature, to the practical advice in the final part of the book. And lastly, the finance professionals will benefit from all the above, and will be able to put to the test their investment ideas. What the book has to say about investing will definitely challenge buy-side practioners, given that Malkiel advocates indexing, which finance professionals tend to oppose.

The author makes clear in the beginning of the book that it is written for the layman. This means that it does not require a prior knowledge of finance or investments, and that it provides all the definitions needed. Indeed, the explanations are done masterfully in a simple language providing all the necessary intuition in order to accept or reject a result. However, the Author takes shortcuts, but does not cut corners. Matters are explained as simply as possible, but not simplistically. The arguments are given mostly through examples able to provide the intuition. The use of jargon is kept to a minimum helping the flow which is uninterrupted. In a nutshell the book is very well written and easy to read even for the uninitiated.

Having said this, it does not mean that I found the book without flaws. Some of my reservations were issues of presentation that I will discuss below. But a substantive aspect of the book that raised a question mark in my mind was about smart beta strategies. Since these are relatively new, and did not feature in earlier editions of the book, I will look at them in some detail in a latter part of this review.

Let me start from the presentation issues. At times, for example while I was reading the part where the Author gives a very negative account of the usefulness of technical analysis, I got the impression that he was not providing any serious evidence against technical analysis but rather that he was dogmatically insisting on his view. It was only much later, towards the end of that chapter, that he went through several technical analysis systems and the tests done for them in order to inform the reader that these tests provided no evidence that the systems were profitable. I felt the same way in other chapters too: apparently dogmatic statements are presented first, and the buttressing evidence provided later. For those who prefer to be shown the hard evidence first, and the argument in favour or against afterward, the book does not get a high score.

In order to expand on this, I found myself at times disagreeing with the author. I suspect this will be the case for any unbiased reader who has not already made up his mind about the issues discussed in this book, and who first wants to see the evidence. The author makes it clear, right from the very beginning, that he has made up his mind and he does not try to keep an equal distance between the theories he examines. It sounds like Malkiel talks to an old friend to whom he has already repeated his views in the past. And perhaps this is just because he truly has told this story many times in the past—this is the 11th edition since the first one in 1973. In some cases, however, the clinching evidence does not come at all or comes much later in the discussion of a topic, and in addition this evidence is ‘circumstantial’ at best. The former, the lack of timing in providing supporting evidence of his view, left me unhappy. The latter, the lack of hard evidence, also left me unhappy, but to a lesser degree, because I was keeping in mind that the audience he speaks to is not the academic reader.

To conclude my general remarks, I feel that a more robust, balanced and rigorous exposition of the evidence would help interested and inquisitive readers and hence the book itself. I understand that perhaps the author targets a different type of audience and that probably considerations regarding the size of the book would come into play if one incorporated these comments. However, one does not have to go all the way towards creating an academic study. There is a huge spectrum between the approach followed in this book and the standards employed in academic journals. I feel there is a way to move the needle towards more rigour without losing audience or increasing the volume size prohibitively. At the very least, an easy to implement suggestion would be to provide references to academic papers when results from them are used in the book.

Now let me turn to discussing the organization of the book and the content of each part. The book is organized in four parts. The first part starts with some definitions, such as what is the random walk theory. In the first part the Author also explains and constrasts on the one hand the ‘castle-in-the-air theory’, i.e. the theory that says any price is correct as long as there is another trader tomorrow wanting to pay more for what was bought today; and on the other hand the ‘firm-foundation’ theory, i.e. the theory that justifies prices with firm fundamentals like earnings and book value. The rest of part one is very fascinating and is devoted to discussing historical events in the financial markets that correspond to bubbles. The author starts with the tulip-bulb craze of the sixteenth century and walks us through the latest financial crisis of 2008 covering a total number of eight different instances of abnormal rise and bust of different asset prices. These historical references are one of the best parts of the book. The author offers not only his extensive academic knowledge about these historical events but also his personal thoughts and examples drawing from his active involvement with the markets since the sixties. These personal accounts add a lot of flavour to the book.

The second part of the book is devoted to discussing the merit of using the technical and fundamental analysis. The author’s view is clearly against using any of these methods and he does not hide it. In the beginning this may sound unjustified to the reader, but in the end he discusses several tests of technical analysis rules that do not seem to offer trading profits. The author is careful to emphasize that what he wants to convey is not that these trading rules do not generate profits. They can and do at times generate big profits, but, the Author argues, they cannot consistently generate profits especially after accounting for transactions costs and taxes.

Most readers may think that attacking technical analysis is a ‘soft target’. More surprisingly—at least coming from an academic—are the criticisms he levels at fundamental analysis. The author discusses where the difficulty arises when the analysts try to predict future stock performance and this lies in accurately and consistently predicting earnings growth. He gives a number of different reasons why this is a difficult task for the analysts as well as the potential conflicts they may face in their work. After the extensive discussion of the difficulty of predicting earnings growth, the author goes on into examining whether the analysts are any better at predicting stocks that will outperform the market as a whole. He reports several studies in which the researchers have found no ability that the analysts can—persistently and consistently—forecast outperforming stocks. The coverage of the fundamental analysis issues is very thorough and satisfying. Here is where the author also introduces the efficient market hypothesis and he blends it in very nicely with the previously discussed concepts.

In order to help his readers understand the point he tries to make in this section, Malkiel repeats again and again that the failure of these methods is in failing to consistently outperform the market. Outperformance can occur randomly but one cannot depend on it. The same is true about the performance of mutual fund managers that will be visited in the next part. Consistently good performance comes very rarely and it occurs as frequently as pure random chance can explain. This is what he calls a random walk down Wall Street.

The fourth part presents modern portfolio theory—connecting risk and return, the CAPM and its shortcomings and the Fama–French three-factor model. Malkiel then presents the main findings from the study of behavioural finance demonstrating how susceptible to mistakes humans are. The value of this part is to educate readers about the common behavioural biases that are embodied in trading in order to control them.

The next section examines the merits of ‘smart’ beta strategies and whether they are truly smart. The main premise here is that all smart-beta strategies are trading themes placing different weights on known risk factors (value, growth, size, momentum, market etc.), hence achieving higher or lower returns than the market overall. Most interestingly, Malkiel looks at the return time series of known smart-beta strategies and tries to explain exactly where the seeming outperformance of these strategies comes from. In his reading, the excess retrurns are due to the increased risks associated with them by overweighting intentionally or unintentionally certain risk factors or industries or placing too much weight on too few names (lack of diversification). He concludes that any smart or not-so-smart strategy’s performance is just proportional to the risk it carries. So, once again, there is no free lunch or alpha. Since this is one of the new topics touched upon in the new edition, it is perhaps worthwhile looking at this in a bit more detail.

Let’s start from what pre-Fama financial theory used to tell us about where returns ‘come from’. If a CAPM-like model held true, then the excess return, , on security i would be given by(1)

with the intercepts, , statistically indistinguishable from zero. (In Equation (1(1) ) denotes the riskless rate, and the return on the market portfolio.) One can derive Equation (1(1) ) starting from some assumptions about preferences, returns, etc. This is the route that leads to the CAPM theory. Or, as simple-minded empiricists, we can pragmatically say that we regress the excess return of security i, , on the excess market return, , because the excess market return seems a reasonable explanatory variable for the return of any security. In both cases the ‘market beta’ is of course the same, and it is given by(2)

Whichever way we look at the problem, let’s now assume that empirically we observe that (some of) the intercepts are statistically significantly different from zero. This means that the return on a particular security is not fully explained by the security’s exposure to the market factor.

A hard-core empiricist may want to stop here: why look for new factors, she may say? Let’s just focus on, and build portfolios from, the securities that, for whatever reason, have high and positive intercepts.

In reality, one is never sure that the intercepts are stable and different from zero, and, in addition, one would like to understand why the extra returns are earned. This is where the search for understandable and financially justifiable factors comes into play.

One can try to identify factors by assuming that, even if the CAPM model does not tell the full story, securities are nonetheless rationally priced, i.e. returns reflect ‘fair’ compensation for bearing exposure to some source of risk (to some ‘factors’). Since these factors could not be accommodated by the standard CAPM view of the world, they are often referred to as ‘anomalies’. (Note that saying ‘anomaly’ does not necessarily imply ‘irrationality’.)

If we accept this working hypothesis, we may want to try to discover what these non-market factors are. The existence of factors other than the market may be due to institutional constraints, cognitive biases, etc. In this framework, the identification of factors can be formally described as the identification of n significant regressors, , , such that Equation (1(1) ) becomes(3)

with the new intercepts now either zero (there are no further ‘anomalies’) or such that(4)

Suppose that we have indeed identified n additional factors (‘understandable’ regressors with explanatory power). Each positive regression slope means that for the ith security exposure to that factor produces returns in addition to the return from the market term (we have assumed that all the are now zero). Conversely, for a negative regression slope .

Suppose now that an established market index exists, which has been built on arbitrary (in the sense of being non-factor based) criteria—e.g. capitalization weights and equal weights. The question arises as to whether the identification of the non-market factors in (3(3) ) allows the construction of more desirable portfolios (in a sense to be clarified) than the index-based portfolio.

Clearly, if, on the basis of the identification in (3(3) ), one loaded one’s portfolio on the securities with the highest expected return, one may have a poorly diversified portfolio—a portfolio with a higher-than-market-beta expected return, but with a lot more risk as well. However, a priori, it is not self-evident that it should be impossible to create a diversified portfolio which has more attractive risk and reward charcteristics than the market portfolio, or a high- or low-market-beta portfolio.

Whether it is possible to do so or not, is an empirical question, which the Author does not address—and that many academics answer in the affirmative. The observation that some of the anomalies uncovered are associated with low volatility stocks or bonds, or with more liquid securities, suggests that it may indeed be possible to do so, and that it is not necessarily true that all of the smart-beta excess return comes from taking on ‘disguised risk’.

The important point to stress is that the construction of smart-beta indices rests on two pillars: one is the identifaction of (justifiable) factors in (3(3) ); the other is the construction of efficient portfolios, once the existence of these factors has been identified. Surely, if (3(3) ) rather than (1(1) ) holds true, the optimal portfolio will be different from the ‘traditional’ one.

Of course, if one starts from the position that all risks only get rewarded in proportion to their market beta, then extra return must come from a riskier portfolio. But it is exactly this starting point that many academics and practitioners challenge. One could argue that the identified factors are simply the result of over-zealous data mining, but this would be a different argument—and one that the Author does not make. So, the part of the book on smart beta seemed to this reviewer not to do full justice to the topic.

Finally, the last, fourth part of the book, is perhaps the most relevant to the investor. Here Malkiel puts to work all the theoretical and empirical results of finance demonstrated in this book and the author presents a series of very practical step-by-step suggestions on how an investor should think about investing his/her wealth across the life-cycle. Malkiel covers topics like what are the main asset classes one should consider providing return characteristics and correlations among them, what investment vehicles are optimal (mutual funds, ETFs, etc and fee structures), how to project returns for stocks and bonds, what are the optimal percentage allocations across asset classes and across the life cycle, and useful trading/investment habits one should have in order to maximize return and reduce risk (dollar averaging, persistent saving, sticking with the strategy across the business cycle, etc).

The main conclusion of the book is that the investor will be much better off investing in indices (since two-thirds of the actively managed funds underperform the index every year) through no-load mutual funds or inexpensive ETFs and actively incorporating tax-considerations, than by following any other investment strategy. Even with the weaknesses I have identified above, the book does force you to rethink the soundness of you professional or personal investment strategy and really puts active managers in a difficult spot. For more details I urge you to read carefully Malkiel’s book. I very strongly recommend it; it is an enjoyable and very useful read and I confess that I am planning to go over it again one more time soon, especially the last part with the practical investment advice.

Notes on contributors

Antonios Sangvinatsos is currently senior vice president at Pimco. Prior to that he worked at Moodys and held academic jobs at NYU and USC. He holds a PhD and an MPhil in Finance from NYU, an MSc in Management Science from Stanford and a BSc in Mathematics from the University of Athens. His research interests include Credit, Fixed Income and Equities.

Antonios Sangvinatsos
Senior Vice President, Pimco
© 2016, Antonios Sangvinatsos