More Than You Know

by Michael J Mauboussin

  No Progress in Human Nature

  In a lecture delivered in the mid-1940s, Ben Graham noted that while there had been many advances in the art of security analysis up to that date, in one “important respect” there was practically no progress at all: human nature. There is little in today’s dismal market performance, swings in sentiment, and anxieties that is new. The small-town barber’s fear and greed are symbolic of those investors who preceded him as well as those who are sure to follow.

  The central role of sentiment comes through loud and clear in the first edition of Graham and Dodd’s Security Analysis, written over seventy years ago. Here’s what they said about the prevailing investor psychology in the late 1920s:The “new-era” doctrine—that “good” stocks (or “blue chips”) were sound investments regardless of how high the price paid for them—was at bottom only a means of rationalizing under the title of “investment” the well-nigh universal capitulation to the gambling fever. We suggest that this psychological phenomenon is closely related to the dominant importance assumed in recent years by intangible factors of value, viz., good-will, management, expected earning power, etc. Such value factors, while undoubtedly real, are not susceptible to mathematical calculation; hence the standards by which they are measured are to a great extent arbitrary and can suffer from the widest variations in accordance with the prevalent psychology.8

  Investors must bear in mind, too, that all sentiment extremes eventually pass, as Graham and Dodd remind us:But if past experience is any guide, the current critical attitude of the investor is not likely to persist; and in the next period of prosperity and plethora of funds for security purchases, the public will once again exhibit its ingrained tendency to forgive, and particularly to forget, the sins committed against it in the past.

  Maintain Perspective

  The stock market, like the bond market, is a discounting machine. This means investors should expect a mid- to high-single-digit nominal return over time under normal conditions. Sentiment swings, extreme optimism or pessimism, can distort these expected returns. (When investors expect returns to be highest is when they’re likely to be the lowest, and vice versa.)

  In difficult markets such as we had in the early 2000s, investors are well served to try to maintain perspective and avoid groupthink.9 In particular, reflecting on history and carefully considering multiple scenarios can be helpful to provide necessary calibration. Buffett, with an emphasis on how easy it is to get swept up in emotion and a dismissal of overly quantitative approaches, comments:[A]n investor will succeed by coupling good business judgment with an ability to insulate his thoughts and behavior from the supercontagious emotions that swirl about in the marketplace.

  38

  Stairway to Shareholder Heaven

  Exploring Self-Affinity in Return on Investment

  Nature uses only the longest threads to weave her pattern, so each small piece of the fabric reveals the organization of the entire tapestry.

  —Richard P. Feynman

  I Could Do That

  Life magazine created a stir in the late 1940s when it openly questioned whether Jackson Pollock (1912-1956) was “the greatest living painter in the United States.” Pollock wasn’t a standard paint-and-palette guy—he created his abstract art by dripping paint onto huge canvases. While some of his pieces sold for millions, one skeptic suggested his art is like “a mop of tangled hair I have an irresistible urge to comb out.”1 Some critics deridingly implied that they could recreate Pollock’s work by randomly splashing paint on a surface.2 Exhibit 38.1 shows a Pollock painting from the late 1940s.

  Still, Pollock’s work has draw. In an effort to understand the aesthetic appeal of Pollock’s paintings, physicist Richard Taylor turned to the world of mathematics. He found that Pollock’s paintings, while seemingly haphazard, exhibit pleasing fractal patterns. A fractal is “a geometric shape that can be separated into parts, each of which is a reduced-scale version of the whole.”3 In spite of the skeptical sneers, Taylor showed that fractal patterns are by no means an inevitable consequence of dripping paint.

  Fractals are ubiquitous in nature—trees, clouds, and coastlines are but a few examples—and as a result are visually familiar to humans.4 One critical feature of a fractal pattern is its fractal dimension—or degree of complexity (a line has a fractal dimension of 1.0, while a filled space has a dimension of 2.0). Taylor and his collaborators found that humans have a preference for fractals with dimensions between 1.3 and 1.5, whether those fractals are natural or human-made. Many of Pollock’s paintings fall within, or near, this range. As a consequence, scientists can quickly distinguish between a Pollock and non-Pollock.5

  EXHIBIT 38.1 Jackson Pollock, Number 8, 1949. Source: Collection Neuberger Museum of Art, Purchase College, State University of New York, gift of Roy R. Neuberger. Photo by Jim Frank.

  Because fractals are so common in nature, scientists often associate them with self-organized systems. Since economics deals largely with these types of systems, we might expect to see fractals in economic systems as well. And indeed, we do.

  Just as we have to analyze a Pollock painting or a coastline to appreciate the underlying fractal pattern, we must take a fresh look at economic systems as well. Order is often hidden.6

  Stairway to Shareholder Heaven

  Self-affinity, or the resemblance of the parts to the whole, is another crucial feature of a fractal. Think of a cauliflower. The whole cauliflower, a large bump, and a small bump all visually resemble one another. Stock price changes are also fractal: after some adjustments, the data look the same whether you look at month-to-month, week-to-week, or day-to-day changes.7

  Analysis shows that distributions of the spread between returns on investment and cost of capital show self-affinity across five levels: country, industry, company, firm, and division. The best way to assess this point is through visual inspection (see exhibit 38.2). Across all levels, we tend to see the same pattern of value creation, value neutrality, and value destruction. To be sure, some of the distributions skew toward value creation and others toward value destruction, but both sides of the spectrum are consistently represented.

  While I show only one industry (diversified chemicals), we can look at any industry and see a similar array of value performance. Ditto for an individual company. So there is nothing unique about the country, industry, company, division, or business line we selected (besides availability of the data).

  Making the Art Less Abstract

  The usefulness of this observation may appear, on the surface, as abstract as a Pollock painting. But I believe these distributions present at least five concrete implications for investors:1. Consider why returns are less than the cost of capital. Generating poor returns is clearly not desirable, but it is important to consider why the returns are low. For example, a company that is early in its life cycle may have depressed returns because it is investing heavily, but its economic future may be bright. Current weak returns may belie a strong outlook.In contrast, a mature company may have poor returns because competitive forces have wrung out all the attractive opportunities, and the industry may be plagued with excess capacity.

  Companies also invest in new businesses where they have little chance of gaining a competitive advantage. So some insight into the nature of poor returns is very useful.

  2. Look for changes in returns (both positive and negative) not anticipated by the market. Empirical evidence shows that changes in returns are strongly correlated to stock price changes. Companies with the greatest return improvement, on average, tend to significantly outperform the companies with the largest return degradation. These data suggest that the market does not fully anticipate the full degree of return changes.8 Investors should carefully gauge market expectations and try to determine whether or not those expectations are likely to change. Many investment processes fail to properly measure and consider market expectations.9

  EXHIBIT 38.2 Self- Similarity of ROI on Multiple Levels

  3.
Judge the likely longevity of excess returns. Reversion to the mean is a powerful force with company-level returns. High-return businesses face competition that drives down their returns, and capital tends to flee low-return businesses, allowing returns to drift up. Discerning how long it is likely to take for excess returns to be competed away is essential.10 The stock market tends to equilibrate shareholder returns via valuation (allowing for risk differences). High-return businesses receive high price-to-book ratios, and low-return businesses garner low ratios. For this reason, good companies are not always good stocks.

  4. Strategy matters. From a company’s perspective, strategy is about pursuing a set of activities that allow it to generate returns above the cost of capital. Successful strategies typically put a company in a unique position, with either a differentiated offering or a low production cost. Strategy is about trade-offs—deciding what to do and what not to do.One noteworthy finding of this work is that even the worst industries include value-creating companies, and the best industries, value-destroying companies. This evidence strongly suggests that competitive strategy matters. A thorough strategy assessment should be integral to a long-term investor’s process.11

  How does management allocate its time? Since exceptional, talented managers are so rare, investors must determine how a company allocates its managerial talent. Often, companies assign their best managers to turn around or fix ailing businesses, instead of letting them drive value at the strongest divisions. For this reason, investors should try to understand the value breakdown of various businesses (which may be in stark contrast to sales or operating income contributions) and judge whether or not the company is intelligently allocating managerial resources.

  Order and Disorder

  Better data and computational tools are allowing researchers to see order in systems previously perceived to be disorderly or random. I suspect that the self-affinity evident in return spreads is symptomatic of the self-organizing properties of global business. While this general observation is intellectually exciting, it also has practical, investment-related relevance. And you don’t have to be a Jackson Pollock fan to see it.

  Conclusion

  The Future of Consilience in Investing

  Since 1993, I have taught a course called Security Analysis at Columbia Business School. As you would expect, the course covers basic investing issues like valuation, financial-statement analysis, and competitive-strategy frameworks. However, in the first class of the semester every year I sound a warning for the students: this class will raise a lot more questions than it will answer.

  For example, we don’t really know how markets aggregate information and what that means for stock price efficiency. Our concept of risk remains incomplete, although we do know that the standard measure of risk is wrong. Most competitive strategy frameworks don’t tell you what strategy is likely to succeed or fail under varying circumstances. And for sure, we still have lots to learn about how our brains work.

  This long list of unanswered questions makes investing both exciting and frustrating. Exciting because we can expect to gain knowledge and improve our understanding in the years to come. Frustrating because we still understand so little, and the market consistently confounds even its smartest participants.

  I firmly believe that consilience among disciplines will play a crucial role in advancing our investing knowledge. Financial economists often greet the investment-related work of physicists, psychologists, and sociologists with skepticism. No doubt, the lack of economic training can put these other scientists at a disadvantage. But ultimately, the insights that researchers gather from cross-disciplinary collaboration will provide the deepest insights—maybe even answers—into the workings of companies and markets.

  Here are some quick thoughts about where a multidisciplinary approach might help our investing knowledge:• Decision making and neuroscience. Throughout this book, I refer to Daniel Kahneman and Amos Tversky’s prospect theory, which describes how people systematically make decisions that deviate from the theoretical ideal. Prospect theory catalyzed the field of behavioral finance, dedicated to the study of cognitive errors and decision-making biases in business and investing settings.Though it represents a huge leap forward, prospect theory still fails to reveal why people make the decisions they do. Advances in neuroscience now allow researchers to peer into the brains of subjects, providing the first tantalizing glimpses of what’s actually going on as people decide. Economist Colin Camerer likens the plunge into neuroscience to the first family on the block to have a television in the 1950s: the picture may be fuzzy and you may need to tweak the rabbit ears, but the new images and insights are exhilarating. The pictures will only get better with time.

  • Statistical properties of markets—from description to prediction? When describing markets, financial economists generally assume a definable tradeoff between risk and reward. Unfortunately, the empirical record defies a simple risk-reward relationship. As Benoit Mandelbrot has argued, failure to explain is caused by failure to describe.Starting in earnest with Mandelbrot’s work in finance in the early 1960s, statistical studies have shown that stock price changes are not distributed along a bell-shaped curve but rather follow a power law.1 Practitioners acknowledged this fact long ago and have modified their models—even if through intuition—to accommodate this reality. Yet even if we can properly describe and categorize the market’s statistical features, the challenge to figure out cause and effect remains.

  • Agent-based models. Most economic models gloss over individual differences and simply assume average individuals. An agent-based model confers limited but varied abilities on individual agents, and lets them loose in silico. These models show that individual differences are important in market outcomes and that feedback mechanisms are rampant. For example, people often make decisions based on what other people decide. These simple models may dramatically improve on our intuition about why markets behave the way they do and eventually may lead to useful predictions.

  • Network theory and information flows. Stanley Milgram is best remembered for his 1960s idea of six degrees of separation—you can connect any two people in the world through five intermediaries. The problem is that Milgram’s research was shoddy at best. For decades, the six-degrees notion was popular but not proven. In the late 1990s, a new generation of scientists addressed the problem using much more sophisticated analytical tools, including computers. They not only rigorously showed the validity of the six-degrees concept, they described the key features of network structure.2

  Our improved understanding of networks has clearly been a multidisciplinary effort, with liberal exchange between the hard, biological, and social sciences. Network research intersects a number of areas, including epidemiology, psychology, sociology, diffusion theory, and competitive strategy. Network theory is likely to add substantially to our understanding of how product and capital markets develop and change.

  • Growth and size distribution. The distribution of firm sizes in industrialized countries is highly skewed: there are very few large firms and many small ones. Scientists have observed this pattern for nearly a century. But no one has been able to explain the mechanism that leads to this distribution.The distribution of animal size and metabolic rate is also skewed, and in a very similar way to firm size. Notably, scientists have successfully explained the physical conditions between size and metabolic rate.3 Extension of some of these biological and physical principles to the social sciences holds substantial promise.

  • Flight simulator for the mind? I have always been very impressed by the flight simulators that pilots use for training. These sophisticated machines simulate myriad conditions, providing pilots with important experience and feedback in a realistic but safe setting.Is it possible to build a simulator that serves the same purpose for investors? One of the major challenges with investing—especially long-term-oriented investing—is feedback. Studies show that clear and consistent feedback helps professionals in probabilistic field
s. While weather forecasters and handicappers get accurate and timely feedback, long-term investors don’t. Maybe one day we’ll create a simulator that provides investors the training they need to make better decisions. Of course, the result will be markets that are even harder to beat.

  Trillions of dollars are exchanged in global markets every day. Yet despite the high stakes and considerable resources researchers have committed to understanding markets, there is much we do not grasp. This book celebrates the idea that the answers to many of these questions will emerge only by thinking across disciplines.

  NOTES

  1. Be the House

  1 J. Edward Russo and Paul J. H. Schoemaker, Winning Decisions: Getting It Right the First Time (New York: Doubleday, 2002), 3-10.

  2 Alfred Rappaport and Michael J. Mauboussin, Expectations Investing (Boston, Mass.: Harvard Business School Press, 2001), 106-8. In this discussion, we assume investors running diversified portfolios are risk-neutral. For techniques to capture risk aversion, see Ron S. Dembo and Andrew Freeman, Seeing Tomorrow: Rewriting the Rules of Risk (New York: John Wiley & Sons, 1998).

  3 Michael Steinhardt, No Bull: My Life In and Out of Markets (New York: John Wiley & Sons, 2001), 129.

  4 Steven Crist, “Crist on Value,” in Andrew Beyer et al., Bet with the Best: All New Strategies From America’s Leading Handicappers (New York: Daily Racing Form Press, 2001), 64. Crist’s chapter is one of the best descriptions of intelligent investing I have ever read. I also highly recommend Steven Crist, Betting on Myself: Adventures of a Horseplayer and Publisher (New York: Daily Racing Form Press, 2003).