The Psychology of Successful Trading

Behavioural Strategies for Profitability

Tim Short

• Publisher: Routledge
• Year: 2017
• Print ISBN: 9781138096288
• eBook ISBN: 9781351601016
• Topic: Business
• Subtopic: Business General
• Index: Business

This book is the first to demonstrate the practical implications of an important, yet under-considered area of psychology in helping traders and investors understand the biases and attribution errors that drive unpredictable behaviour on the trading floor. Readers will improve their chances of trading successfully by learning where cognitive biases lead to errors in stock analysis and how these biases can be used to predict behavior in market participants.

Focusing on the three major types of bias—Belief-Formation, Quasi-Economic, and Social—the book provides a rigorous discussion of the literature before explaining how each of these biases plays out in financial markets. The author brings together the fields of philosophical psychology and behavioral finance to introduce "theory of mind," providing readers with tools to predict biases in others as well as using these predictions to form optimal trading strategies for themselves. Readers will also learn to understand their own behaviors, counteracting biases such as overconfidence and conformity—and the "curse" of their own knowledge—to strengthen trade performance. Pairing his skill and experience with an extensive research bibliography, Short positions the foundational sources of cognitive biases alongside concrete examples, experimental designs, and trader's anecdotes, helping readers to apply theoretical guidelines to real-life scenarios.

Shrewd professionals and MBA students will benefit from The Psychology of Successful Trading's intuitive structure and practical focus.

1

Introduction

The central idea of this book is that you can improve your chances of trading successfully by understanding where cognitive biases cause you to make mistakes in stock analysis and, more importantly, in predicting the behaviour of others.

Predicting the behaviour of others successfully is the only way to outperform in financial markets. This is what has been called “second-level thinking” (Marks 2011, Ch. 1). First level thinking is selecting stocks that you think will perform well based on characteristics of the stock-issuing company. To do this, you will engage in an array of standard analysis such as considering P/E ratios, yields, the macro-economic environment, geopolitics and a myriad of other factors including the likely future development of all of the aforementioned factors. This, of course, is what the whole market is doing all the time. This is why it is difficult to make money doing it; others have already got there before you. The key point is that you have to find stocks to buy or sell which are undervalued or overvalued by the market, which is primarily a psychological question and not a financial or economic one. As Marks (2011, p. 1) puts it, practically at the start of his important book, “[p]sychology plays a major role in markets.” With even more emphasis, he also writes that the “discipline that is most important is not accounting or economics, but psychology” (Marks 2011, p. 27). This is underlined by Hirshleifer (2001, p. 1533) who writes that in the newer, less purely rational approach to investor psychology, “security expected returns are determined by both risk and mis-valuation” with mis-valuation being primarily psychological in origin. In this book, I will give you the psychology you need.

This second level psychological thinking will be the major focus of this book. It is called second level thinking because it involves thinking about the first level thinking, or more precisely establishing what the first level thinking of others will be and whether it is correct. I will outline the first level thinking I use so it can be understood sufficiently to underpin the second level thinking that I aim to discuss, but the discussion here of first level thinking will be instrumental. The main aim will be finding examples to illustrate how the psychological points will play out in the market rather than exhaustively setting out the financial questions alone. In sum, the first level thinking will be described only to illuminate the second level thinking. The motivation for this approach is described well by Nofsinger (2016, p. 8) who notes that those who learn about biases “may find opportunities to benefit from the biased decisions of other investors.”¹

I am not saying that you can avoid doing the first level thinking. You can get it done for you or learn how to do it yourself by reading the financial and investment press, or by reading some of the many books which do focus on the first level – but the second level thinking is where the action is. Understanding of the markets is rare enough; understanding of psychology is rarer still. Both are needed for success, which is yet rarer still. This is why you need to gain expertise in aspects of psychology, as well as being able to understand investment cases. If you only have time to do one of these tasks, the psychological one is more important because as said, you can outsource the first level thinking. It is much harder to buy in the second level thinking, because few people are doing it well. In this book, I will not be doing the second level thinking for you. I will be showing you how to do it for yourself.

Shull (2011, p. 24) has a useful poker analogy, which can serve to illustrate this distinction between levels. She notes that many people think of poker as a probability game. Now clearly, having the right cards is a very significant part of winning a poker game, but it is not everything. Shull notes that uncertainty arises whenever wagering begins. Her key observation is that winners “in poker rely on the human perception games of the betting.” What she means here is, of course, that the key difference between winning players and losing ones is less to do with the cards they are dealt and more to do with their ability to predict the behaviour of others. This is what is known as a Theory of Mind² task. Theory of Mind is the psychological term for the way we predict and explain the behaviour of others. Analysing stock-market assets is a useful discipline; it is like counting cards to optimise the underlying numerical probabilities. However, outperforming at the Theory of Mind task, in either poker or the markets, is what makes winners.

Shull (2011, p. 54) has also noticed the importance of first and second level thinking, and the fact that they have already been observed in the literature. She notes that Keynes had already gone to the third level. Third level thinking is thinking about the second level thinking; or predicting how others will predict further others will think. He points out, in his famous beauty contest example (2016, Ch. 12), that the task is really to attempt to anticipate what average opinion expects average opinion to be. This is again a Theory of Mind task. It is an example particularly conducive to the simulation account thereof, which I defend. There are two major accounts of Theory of Mind in the literature. One is called Simulation Theory and is based on the idea that we predict others by putting ourselves in their place. The other account is called Theory Theory and is based on the idea that we have a theory of others which we use to predict their behaviour. I argue in Short (2015) that Simulation Theory is the correct account. Simulation Theory has no difficulty explaining Keynes’s multiple levels. Theory Theory, by contrast, would need to postulate that there are rules about what people think about what people think. On a simulation account, the level almost drops away. There is no difference between simulating what someone will think and simulating what someone will think someone else will think. In any case, both authors are correct to emphasise the importance of Theory of Mind to market performance.

This book is partly aimed at retail investors who are investing their own money and need to investigate how psychology and market forces interact in ways which are not necessarily helpful. I will not be explaining any standard financial terminology, but it will mostly be familiar to the fairly experienced individual or at least, capable of investigation in public sources. Professional traders can also gain a lot from specialised psychology, which they will not have come across before unless they have a university level background in psychology and, more specifically, with a focus on bias psychology.

All of the psychology I cover will be explained in depth, with my remarks aimed at an educated individual who has no previous knowledge of academic psychology. I will be citing all of the relevant psychological and financial literature I use thoroughly, so you will be able to pursue the ideas further, if you wish. The point here, though, is that I have read 600 relevant academic journal articles and books – and synthesised here what it means for understanding market participants – so you do not have to. Again, it is worth underlining that understanding psychology is crucial to understanding markets and, in particular, difficult times in markets which can cause enormous damage to investors. Barberis (2013, p. 25), discussing the 2007–2008 financial crisis which continues to have malign effects today in 2016, suggests that “it is very possible that psychological factors were also central to the crisis” as well as institutional failures, which are themselves not immune to psychological causation.

I will not of course attempt to explain the whole of psychology, or even the whole of the psychology of cognitive biases. That would not be possible in a single book. It is also not necessary for the project here. Large parts of psychology, such as, for example, the psychophysics of perception, are not relevant. We will be tightly focussed on cognitive biases because, as I have previously argued (Short 2015; Short and Riggs 2016), they are the dominant causes of Theory of Mind errors. These, then, are the elements of psychology which are the dominant cause of error in predicting the behaviour of others and thus the key to better second level thinking. To the extent we can assume that the bulk of market participants have not read this book, then it will suffice to provide an edge or market advantage.³ If it becomes widely read, we may need to go to third level thinking, or thinking about the second level thinking, but we can cross that bridge when we come to it. I will restrict myself to those cognitive biases which are important in predicting the behaviour of financial market participants. Large parts of the subject deal with abnormal psychology. While it will occasionally be useful to examine some abnormal conditions for what we can learn from them about neuro-typical subjects, in this book we are interested only in how we can expect the majority of people to behave.

All of us are constantly subject to a wide array of biases in our thinking; for example Confirmation Bias (§5.1) where we tend only to seek information which accords with what we already think. The idea throughout will be to alert you to these biases so that you can make trading decisions which are: a) more optimal initially because you have reduced the impact of bias on your own thinking, and b) more optimal again because you have incorporated the possibility of biases into your analysis of the thinking of other market participants. We are especially likely to resort to heuristics and biases in the context of markets. As Dale, Johnson, and Tang (2005, p. 261) note in the context of a discussion of the South Sea Bubble, “individuals increasingly rely on heuristics, non-rational strategies, and biases when faced by a complex information environment.” There is no more complex information environment than a financial market, and so the explanation of Dale, Johnson, and Tang (2005) that investors failed to recognise the South Sea Bubble when it was occurring is plausible. In this case, obviously, the biases of investors did them a great deal of harm. So we can agree with Barber and Odean (2002, p. 456) that “cognitive biases […] for the most part, do not improve investors’ welfare.”

Shiller argues that “mass psychology may well be the dominant cause of movements in the price of the aggregate stock market” (Shiller, Fischer, and Friedman 1984). The way this works, as is succinctly explained in the appended commentary by Fischer, is that “smart-money investors [look] ahead to try to predict both dividends and the value of shares the blockheads will be holding in the future” (Shiller, Fischer, and Friedman 1984, p. 502). What this means, simply enough, is that “[c]hanges in expectations of the holdings of blockheads, as well as changes in expected dividends, will change the price” now (Shiller, Fischer, and B. M. Friedman 1984, p. 502). Forecasting future changes in dividends, if it can be done, will be an excellent way to predict future stock price movements. More significantly for our purposes, changes in the value of future holdings by blockheads – which here simply means any change not explicable rationally on the basis of rational reasons to expect future dividend changes – are more important. These changes, I will propose, are driven by cognitive biases on the part of the blockheads. In final comments by Fischer and Friedman, it is complained that while Shiller explains that investing in ‘fads’ or fashions explain excess stock volatility, he does “not explain how fads are formed and why they subsequently disappear” (Shiller, Fischer, and Friedman 1984, p. 510). I will be adding that missing piece by suggesting that cognitive biases can also be explanatory of that point.⁴ For example, Conformity Bias (§7.4), which is just the tendency to copy others, can lead to herding behaviour (Nofsinger 2016, pp. 104–105).

A further underlining of the importance of the idea may be derived from Soros’ (1994) subtitle, ‘Reading The Mind Of The Market.’ Certainly this is the key task, and informally within psychology, ‘mind-reading’ is often used as a synonym for Theory of Mind. Soros (1994) even mentions biases, but he is mostly concerned with a simple prevailing bias – by which he just means the weight of market participants at a given time on the bullish/bearish spectrum. So while the task identified is the correct one, there is a lot more work to do on elucidating how biases play out in markets. That will be my task in this book.

You may be thinking that bias psychology does not apply to you. Everyone else may be making biased decisions all the time, but you do not. Unfortunately you are wrong about this (I obviously do not claim any immunity for myself either). Pronin, Gilovich, and Ross (2004, p. 781) find that we tend “to see others as more susceptible to a host of cognitive and motivational biases” than ourselves. They ascribe this to what we might describe as introspective asymmetry. The underlying claim involved in introspection is that I know what is going on in my mind directly and unmediatedly while I have no such access to what is going on in your mind. Introspection is the method by which I know what is going on in my mind; you can think of it as a contraction of ‘internal’ and ‘perception’ if you like. At least the second part is true, because otherwise there would be no need for Theory of Mind abilities. The first part appears to be true to everyone except some philosophers and psychologists, though it has found some defenders even within those disciplines (Rey 2013).

It has been shown that well-documented psychological biases play a major role in mis-pricing. An analysis by Daniel, Hirshleifer, and Subrahmanyam (1998) simulated the effects of the tendency to ascribe success to one’s own abilities, and failure to bad luck or unknowable external factors. I see this tendency as a species of Self-Presentation Bias; it is known in the literature as Self-Attribution Bias. Daniel, Hirshleifer, and Subrahmanyam (1998, p. 1866) state that their “key contribution” is to show that this bias can “induce several of the anomalous price patterns documented in the empirical literature.” This is good evidence in relation to a single bias; in this book I will aim to cover all of the most important ones.

One sort of objection here suggests that what I am proposing is a kind of Error Theory. In the view of the objector, I am proposing an account whereby many people are wrong much of the time. This is seen as implausible on the grounds that widespread error does not seem to be the sort of occurrence which could remain widespread and uncorrected. My response to that, as will be discussed more below, is to suggest that cognitive biases are not exactly errors. They may result in sub-optimal decision making but we have them because often they are fast and good enough. One of my aims in the book will be to explain some of the key cognitive biases so you can see when your thinking and that of others may have been influenced by them. That is an essential precursor to the subsequent decision as to whether the output of the cognitive bias is actually the optimal decision in the current instance.

A variant of this objection is based on evolution. Since, the objection runs, we are evolved creatures, our methods of reasoning will not be as imperfect as is suggested by the widespread presence of biases. We would have evolved them away since they are sub-optimal. There are a number of responses to this objection, of which I will only briefly canvass a few.

The first response is to note that although it is true we are evolved, we are still sub-optimal. ‘Sub-optimal’ means merely that and not maximally so: evolution produced many innovative and intricate solutions to the problem of how reproductive fitness can be enhanced (Pinker 2015, p. 167). However, we do not have bones made of titanium alloy or some super material, despite the fact that this would result in superior engineering properties. This is because evolution does not have any ability to plan. It can become trapped in ‘local minima,’ meaning that even if it would be better to have titanium bones, if there is no path from here to there in which most intermediate stages are improvements over their predecessor, there is no way for evolution to get there. Our bones are also not made from string. Evolution has done a pretty decent job under the circumstances. But we should not expect perfection either here or psychologically.

The second sort of response is decisive, I think. It suggests that we have these biases not because they are imperfections but because they are, on average, an improvement. It is simply impracticable and inefficient to expend vast cognitive resources on all questions that arise, even if sometimes such expenditure would result in a more optimal solution. Pinker (2015, p. 138) notes the telling example of a hiker wanting to return before sunset who spends 20 minutes planning how to make the route back 10 minutes shorter. Quickly finding a route that is imperfect but good enough is a much better practical solution. Similarly, many cognitive biases are ‘good enough’ for general purposes. My aim in this book is not to eliminate them, were that possible, but to enable people to notice their operation and decide when to let them make the call and when further work is called for.

It is also worth bearing in mind evidence as noted by Kramer (2008, p. 128) to the effect that “risk tolerance depends on age, income and wealth, gender, and marital status [and also] ethnicity, birth order, education, and personality traits such as self-esteem [and even] levels of hormones and neuro-chemicals.” Now, risk tolerance feeds into all financial decisions. Since the factors mentioned are features of psychology or brain chemistry and not based on characteristics of the financial question under consideration, not all of the decisions made can be correct. There is an optimal level of risk tolerance under various market conditions and it does not depend on any of the factors listed. This means that anyone strongly influenced by any of these factors is likely to be some distance away from the correct level of risk tolerance. Sometimes this may be appropriate. Someone with a spouse who is a high earner and relaxed about losses can probably take more risk than otherwise. But there is no way that a rush of testosterone can improve your trading.

I will be suggesting that the influence of cognitive biases is one way that these various factors make themselves felt in our trading and other behaviour. I will also be suggesting, as I have in Short (2015), that failure to take account of the emotional state of others and then, as a result, failing to simulate what they will do accurately is a major cause of making bad predictions of what others will do. As Korniotis and Kumar (2011, p. 1513) point out, “aggregate forces generated by investors’ systematic behavioural biases have the ability to influence stock prices and trading volume,” so there are wider implications beyond the individual. Understanding how these biases play out more widely is essential to delivering good market performance. Note also the emphasis on systematic biases. It is the systematic nature of the biases – meaning that they can be expected to occur every time similar circumstances reoccur – that makes understanding biases in ourselves and others a regular source of value. These effects are not small. Korniotis and Kumar (2011, p. 1550) conclude that individual biases harm the economies of US states: “people’s sub-optimal investment decisions aggregate up and the adverse effects of their biases can be detected even in the aggregate, state-level macro-economic data.”

One astute commentator has asked why it is that we do not correct our cognitive biases in economic questions when failing to do so costs us money. Besharov (2004, pp. 13–14) gives three reasons: “individuals have limited knowledge of the system of interacting biases; […] there may be a set of biases that result in the efficient level of action” and “even an individual with full information about the nature of the biases may rationally choose to correct them only partially when correction is costly.” All three, I think, are correct. The first reason asserts that it will be difficult for people to correct for biases if they do not know what they are. This seems unarguably true, and one task I conduct in this book is to describe the key biases and situate them in the trading context, so that first reason is addressed. The second reason is interesting: it points to the fact that since we are influenced by so many biases, more than one of them may be operating at a given time. In reality, I think this is qu...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Brief Contents
• Table of Contents
• Acknowledgements
• 1. Introduction
• 2. Beating the Market
• 3. Theory of Mind
• 4. Belief-Formation Biases I
• 5. Belief-Formation Biases II
• 6. Quasi-Economic Biases
• 7. Social Biases I
• 8. Social Biases II
• Bibliography
• Index