- -
- 100%
- +
In the 1980s an alternative view developed, driven by evidence that the rationality assumption is unrealistic. Further, the mistakes of individuals may not disappear in the aggregate. People are irrational and this causes markets to be inefficient. Behavioral finance was the antithesis.
Synthesis hasn't yet arrived, but behavioral finance is now seen as neither an all-encompassing principle nor a fringe movement. It augments, not replaces, traditional economics.
What have we learned from behavioral finance?
First, behavioral finance has added to our understanding of market dynamics. Even in the presence of rational traders and arbitrageurs, irrational “noise” traders will prevent efficiency. And although it is possible to justify the existence of bubbles and crashes within a rational expectations framework (for example, Diba and Grossman, 1988), a behavioral approach gives more reasonable explanations (for example, Abreu and Brunnermeier, 2003, and De Grauwe and Grimaldi, 2004).
Second, we are now aware of a number of biases, systematic misjudgments that investors make. Examples include the following:
• Overconfidence: Overconfidence is an unreasonable belief in one's abilities. This leads traders to assign too narrow a range of possibilities to the outcome of an event, to underestimate the chances of being wrong, to trade too large, and to be too slow to adapt.
• Overoptimism: Overconfidence compresses the range of predictions. Overoptimism biases the range, so traders consistently predict more and better opportunities than really exist.
• Availability heuristic: We base our decisions on the most memorable data even if it is atypical. This is one reason teeny options are overpriced. It is easy to remember the dramatic events that caused them to pay off, but hard to remember the times when nothing happened and they expired worthless.
• Short-term thinking: This thinking shows the irrational preference for short-term gains at the expense of long-term performance.
• Loss aversion: Investors dislike losses more than they like gains. This means they hold losing positions, hoping for a rebound even when their forecast has been proven wrong.
• Conservatism: Conservatism is being too slow to update forecasts to reflect new information.
• Self-attribution bias: This bias results from attributing success to skill and failure to luck. This makes Bayesian updating of knowledge impossible.
• Anchoring: Anchoring occurs when relying too much on an initial piece of information (the “anchor”) when making a forecast. This leads traders to update price forecasts too slowly because the current price is the anchor and seems more “correct” than it should.
And there are at least 50 others.
It is these types of biases that traders have tried to use to find trades with edge. Results have been mixed. There are so many biases that practically anything can be explained by one of them. And sometimes there are biases that are in direct conflict. For example, investors underreact, but they also overreact. Between these two biases you should be able to explain almost any market phenomena. The psychologists and finance theorists working in the field are not stupid. They are aware of these types of difficulties and are working to disentangle the various effects. The field is a relatively new one and it is unfair and unrealistic to expect there to be no unresolved issues. The problem is not really with the field or the serious academic papers. The problem is with pop psychology interpretations and investors doing “bias mining” to justify ideas.
It is common in science for a new idea to be overly hyped, particularly those that are interesting to lay people (traditional finance is not interesting). In the 1970s there were popular books about catastrophe theory, a branch of physics that was meant to explain all abrupt state changes and phase transitions. It didn't. In the 1990s, chaos theory was meant to explain practically everything, including market dynamics. It didn't. Behavioral finance is being overexposed because it is interesting. It provides plenty of counterintuitive stories and also a large amount of schadenfreude. We can either feel superior to others making stupid mistakes or at least feel glad that we aren't the only ones who make these errors.
And people love intuitive explanations. We have a great need to understand things, and behavioral finance gives far neater answers than statistics of classical finance theory. Even though behavioral finance doesn't yet have a coherent theory of markets, the individual stories give some insight. They help to demystify. This is reassuring. It gives us a sense of control over our investments.
A science becoming interesting to the general public doesn't necessarily mean it is flawed. For example, there have been hundreds of popular books on quantum mechanics. However, behavioral finance does have some fairly serious problems to address.
Just as in conventional finance theory, behavioral finance studies individual decision-making despite the fact that people do not make investing decisions independently of the rest of society. Everyone is influenced by outside factors. Most people choose investments based on the recommendations of friends (Katona, 1975). And professionals are also influenced by social forces (Beunza and Stark, 2012). Over the last 30 years the sociology of markets has been an active research field (for example, Katona, 1975, Fligstein and Dauter, 2007, and references therein), but this work hasn't yet been integrated into behavioral finance. Because behavioral finance largely ignores the social aspects of trading and investing, we don't have any idea of how the individual biases aggregate and their net effect on market dynamics. This is necessary because, even though we don't understand how aggregate behavior emerges, it is very clear that markets cater to irrational behavior rather than eradicate it. For example, the services of financial advisors, stock brokers, and other financial intermediaries made up 9% of the US GDP (Philippon, 2012) despite the fact that they are almost all outperformed by much cheaper index funds and ETFs.
Next, behavioral finance has largely limited itself to the study of cognitive errors. There are many other types of nonrational behavioral inputs into decision-making, including emotion, testosterone levels, substance abuse, and the quest for status.
And behavioral finance gives no coherent alternative theory to the EMH. A catalog of biases and heuristics—the mistakes people make—is not a theory. A list of facts does not make a theory. Of course, sometimes observations are necessary before a theory can be formulated. Mendeleev drew the periodic table well before the atomic structure of matter was understood. We knew species existed well before we understood the process of speciation by natural selection. Still, to be scientific, behavioral finance eventually needs to lead to a unifying theory that gives explanations of the current observations and makes testable predictions.
Behavioral finance can still help. Whenever we find something that looks like a good trading idea we need to ask, “Why is this trade available to me?” Sometimes the answer is obvious. Market-makers get a first look in exchange for providing liquidity. Latency arbitrage is available to those who make the necessary investments in technology. ETF arbitrage is available to those with the capital and legal status to become authorized participants. But often a trade with positive edge is available to anyone who is interested. Remembering the joke about the economists, “Why is this money sitting on the ground?” Risk premia can often be identified by looking at historical data, but behavioral finance can help to identify real inefficiencies. For example, post-earnings announcement drift can be explained in terms of investor underreaction. Together with historical data, this gives me enough confidence to believe that the edge is real. The data suggest the trade, but the psychological reason gives a theoretical justification.
High-Level Approaches: Technical Analysis and Fundamental Analysis
Technical analysis is the study of price and volume to predict returns.
Technical Analysis
Aronson (2007) categorized technical analysis as either subjective or objective. It is a useful distinction.
Subjective technical analysis incorporates the trader's discretion and interpretation of the data. For example, “If the price is over the EWMA, I might get long. It depends on a lot of other things.” These methods aren't wrong. They aren't even methods. Subjectivity isn't necessarily a problem in science. A researcher subjectively chooses what to study and then subjectively chooses the methods that make sense. But if subjectivity is applied as part of the trading approach, rather than the research, then there is no way to test what works and what doesn't. Do some traders succeed with subjective methods? Obviously. But until we also know how many fail, we can't tell if the approach works. Further, the decisions different traders who use ostensibly the same method make won't be the same or even based on the same inputs. There is literally no way to test subjective analysis.
Some things that are intrinsically subjective are Japanese candlesticks, Elliot waves, Gann angles, trend lines, and patterns (flags, pennant, head, and shoulders, etc.). These aren't methods. In the most charitable interpretation, they are a framework for (literally) looking at the market. It is possible that using these methods can help the trader implicitly learn to predict the market. But more realistically, subjective technical analysis is almost certainly garbage. I can't prove the ideas don't work. No-one can. They are unfalsifiable because they aren't clearly defined. But plenty of circumstantial evidence exists that this analysis is worthless. None of the large trading firms or banks has desks devoted to this stuff. They have operations based on stat arb, risk arb, market-making, spreading, yield curve trading, and volatility. No reputable, large firm has a Japanese candlestick group.
As an ex-boss of mine once said, “That isn't analysis. That is guessing.”
Any method can be applied subjectively, but only some can be applied objectively. Aronson (2007) defines objective technical analysis as “well-defined repeatable procedures that issue unambiguous signals.” These signals can then be tested against historical data and have their efficacy measured. This is essentially quantitative analysis.
It seems likely that some of these approaches can be used to make money in stocks and futures. But each individual signal will be very weak and to make any consistent money various signals will need to be combined. This is the basis of statistical arbitrage. This is not within the scope of this book.
However, we do need to be aware of a bad classic mistake when doing quantitative analysis of price or return data: data mining.
This is where we sift through data using many methods, parameters, and timescales. This is almost certain to lead to some strategy that has in-sample profitability. When this issue is confined to choosing the parameters of a single, given strategy it is usually called overfitting. If you add enough variables, you can get a polynomial to fit data arbitrarily well. Even if you choose a function or strategy in advance, by “optimizing” the variables you will the get the best in-sample fit. It is unlikely to be the best out of sample. Enrico Fermini shared that the mathematician and economist John von Neumann said, “With four parameters I can fit an elephant, and with five I can make him wiggle his trunk” (Dyson, 2004).
This mistake isn't only made by traders. Academics also fall into the trap. The first published report of this was Ioannidis (2005). Subsequently, Harvey et al. (2016) and Hou et al. (2017) discussed the impact of data mining on the study of financial anomalies.
There are a few ways to avoid this trap:
• The best performer out of a sample of back-tested rules will be positively biased. Even if the underlying premise is correct, the future performance of the rule will be worse than the in-sample results.
• The size of this bias decreases with larger in-sample data sets.
• The larger the number of rules (including parameters), the higher the bias.
• Test the best rule on out-of-sample data. This gives a better idea of its true performance.
• The ideal situation is when there is a large data set and few tested rules.
Even after applying these rules, it is prudent to apply a bias correcting method.
The simplest is Bonferroni's correction. This scales any statistical significance number by dividing by the number of rules tested. So, if your test for significance at the 95% confidence level (5% rejection) shows the best rule is significant, but the rule is the best performer of 100 rules, the adjusted rejection level would be 5%/100 or 0.005%. So, in this case, a t-score of 2 for the best rule doesn't indicate a 95% confidence level. We would need a t-score of 2.916, corresponding to a 99.5% level for the single rule. This test is simple but not powerful. It will be overly conservative and skeptical of good rules. When used for developing trading strategies this is a strength.
A more advanced test is White's reality check (WRC). This is a bootstrapping method that produces the appropriate sampling distribution for testing the significance of the best strategy. The test has been patented and commercial software packages that implement the test can be bought. However, the basic algorithm can be illustrated with a simple example.
We have two strategies, A and B, which produce daily returns of 2% and 1% respectively. Each was developed by looking at 100 historical returns. We can use WRC to determine if the apparent outperformance of strategy A is due to data mining:
• Using sampling with replacement, generate a series of 100 returns from the historical data.
• Apply the strategies (A and B) to this ahistorical data to get the pseudo-strategies A' and B'.
• Subtract the mean return of A from A' and B from B'.
• Calculate the average return of the return-adjusted strategies, A” and B”.
• The larger of the returns of A” and B” is the first data point of our sample distribution.
• Repeat the process N times to generate a complete distribution. This is the sampling distribution of the statistic, maximum average return of the two rules with an expected return of zero.
• The p-value (probability of our best rule being truly the better of the two) is the proportion of the sampling distribution whose values exceed the returns of A, that is, 2%.
A realistic situation would involve comparing many rules. It is probably worth paying for the software.
There is also a totally different and complementary way to avoid overfitting. Forget about the time series of the data and study the underlying phenomenon. A hunter doesn't much care about the biochemistry of a duck, but she will know a lot about their actual behavior. In this regard a trader is a hunter, rather than a scientist. Forget about whether volatility follows a GARCH(1,1) or a T-GARCH(1,2) process; the important observation is that it clusters in the short term and mean reverts in the long term. If the phenomenon is strong enough to trade, it shouldn't be crucial what exact model is used. Some will always be better in a sample, but that is no guarantee that they will work best out of a sample.
As an example, this is the correct way to find a trading strategy.
There is overwhelming evidence that stocks have momentum. Stocks that have outperformed tend to continue outperforming. This has been observed for as long as we have data (see Geczy and Samnov [2016], Lempérière et al. [2014], and Chabot et al. [2009]) and in many countries (for example, Fama and French, 2010). The observation is robust with respect to how momentum is defined and the time scales over which it is measured. In the trading world, the evidence for stock momentum is overwhelming. Starting from this fact, design a simple model to measure momentum (e.g., 6-month return). Then sort stocks by this metric and buy the ones that score well.
The worst thing to do is take a predefined model and see if it works. Has a 30-day, 200-day moving average crossover been predictive of VIX futures? What if we change the first period to 50 days? I don't know or care.
Fundamental Analysis
Fundamental analysis aims to predict returns by looking at financial, economic, and political variables. For example, a fundamental stock analyst might look at earnings, yield, sales, and leverage. A global macro trader might consider GDP, currency levels, trade deficit, and political stability.
Fundamental analysis, particularly global macro, is particularly susceptible to subjectivity. It also tempts otherwise intelligent people to make investment decisions based on what they read in the Wall Street Journal or The Economist. It is exceedingly unlikely that someone can consistently profit from these public analyses, no matter how well the story is sourced or how smart the reader is.
Consider these statements from “experts”:
“Financial storm definitely passed.”
—Bernard Baruch, economic advisor to presidents Woodrow Wilson and Franklin Roosevelt in a cable to Winston Churchill, November 1929Stocks dropped for the next 3 years, with the Dow losing 33% in 1930, 52% in 1931, and 23% in 1932.
“The message of October 1987 should not be taken lightly. The great bull market is over.”
—Robert Prechter, prominent Elliot wave theorist and pundit, in November 1987. The Dow rallied for 11 of the next 12 years, giving a return (excluding dividends) of over 490%.“A bear market is likely… It could go down 30% or 40%.”
—Barton Biggs, chief strategist for Morgan Stanley, October 27, 1997The Dow had its largest one-day gain on October 28 and continued to rally hard for the next 6 months.
In most situations it is just mean to make fun of people's mistakes. We all make mistakes. But the people I have quoted have proclaimed themselves experts in a field where real expertise is very, very rare.
And evidence of this is more than anecdotal.
The poor prediction skill of experts is a general phenomenon. Gray (2014) summarizes the results of many studies that show that simple, systematic models outperform experts in fields as diverse as military tactics, felon recidivism, and disease diagnosis. Expertise is needed to build the models, but experts should not make case-by-case decisions.
Koijen et al. (2015) show that surveys of economic experts (working for corporations, think tanks, chambers of commerce, and NGOs) have a negative correlation to future stock returns. They were also contraindicative for the returns of currencies and bonds. This effect applies across 13 equity markets, 19 currencies, and 10 fixed income markets. A simple “fade the experts” strategy would have given a Sharpe ratio of 0.78 from 1989 to 2012.
Financial advisors are equally bad. Jenkinson et al. (2015) look at the performance of advisors in picking mutual funds. They conclude with, “We find no evidence that these recommendations add value, suggesting that the search for winners, encouraged and guided by investment consultants, is fruitless.” And fund managers themselves can't consistently beat the averages. Due to costs, most managers underperform and there is no correlation between performances from one year to the next. So, managers can't pick stocks and it is pointless to try to pick good managers.
It is also likely that much of the alpha generated by fundamental analysis is smart beta, compensation for exposure to a certain risk factor. There is absolutely nothing wrong with this. Trading profits are profits, no matter whether they are due to smart beta or alpha. But before we ascribe a trader's results to skill, we should know what is causing the profits. Beta should cost a lot less than alpha.
Conclusion
It is difficult to make money in financial markets. The EMH isn't completely true, but it is closer to being correct than to being wrong. If a trader can't accept this, she will see edges in noise and consequently overtrade. Behavioral finance, technical analysis, and fundamental analysis can all be used as high-level organizing principles for finding profitable trades, but each of these needs to be believed only tentatively, and the most robust approach is to look for phenomena that are independently clear. For example, momentum can be discovered through technical analysis but also understood as a behavioral anomaly. The observable phenomenon must come before any particular method.
Summary
• Exceptions to the EMH exist but they are rare.
• Exceptions will either be inefficiencies, temporary phenomena that last only until enough people notice them, or poorly priced risk premia.
• Risk premia will persist and can form the core of a trader's operations but the profits due to inefficiencies will decay quickly and need to be aggressively exploited as soon as they are found.
• A promising trading strategy is one whose basis is independent of the specific methods used to measure it. Start with observation, then move to quantification and justification.
CHAPTER 3
Forecasting Volatility
All successful trading involves making a forecast. Some traders (for example, trend followers) say they don't forecast, they react. I don't know why they say this, but in any case, they are wrong. The moment a trader enters an order, she has implicitly made a forecast. Why would you get long if you didn't think the market was going up? No matter how it was arrived at, the forecast is, “the market is going up.” Except for a pure arb (which are practically extinct), to get positive expectation we need to make a forecast that is both correct and more correct that the consensus.
In this chapter we will concentrate on making correct forecasts of volatility. But, first, here are some principles that are applicable to any financial forecasting:
• Pick a good problem. Some things are impossible to forecast. No one can predict the price of AAPL in 25 years. Some things are hard to predict. Forecasting the S&P 500 index in two days is a hard problem. Some things are trivial to predict. The FED funds rate in the next day is almost certainly going to be unchanged. Aim to find problems that are solvable but are hard enough that you will be able to profit from the predictions. Volatility is a perfect candidate for this.
• Actively look for comparable historical situations. What happens when the government shuts down? What is the link between recessions and the stock market? This is a good general principle, but it is also vital if you are looking for catalysts that could lead to volatility explosions. Good periods to be short volatility can often be deduced from financial data alone, but long trades generally need a catalyst (that isn't priced in) to be successful. Don't trust your intuition or what you think is true. These will be biased by your experiences, environment, and political persuasion. If you don't have data, you don't have knowledge.
“When my information changes, I alter my conclusions. What do you do, sir?”
—J. M. Keynes• Aim to balance being conservative and reactive. All good investors have a Bayesian model in their head and update their forecasts as new information arrives but you also shouldn't update too aggressively.
• Actively look for counterarguments. Every person has biases. If you are convinced that every article you read is a harbinger of chaos, be open to the possibility that you are wrong. And the same holds if you are a habitual volatility seller.
“It is impossible to lay down binding rules, because two cases will never be exactly the same.”
—Field Marshall Helmuth von MoltkeRemember that there are no certainties when predicting the future.
Model-Driven Forecasting and Situational Forecasting