The Epistemology of Price Prediction

“The blind transfer of the striving for quantitative measurements to a field in which the specific conditions are not present which give it its basic importance in the natural sciences, is the result of an entirely unfounded prejudice. It is probably responsible for the worst aberrations and absurdities produced by scientism in the social sciences.” - F.A. Hayek, The Counter Revolution of Science

I’ve spent lots of time of late thinking about where investing fits into the philosophical schools I studied in graduate school at Oxford and later taught to undergraduates, and especially where to place Dan’s unique approach. I thought I’d share some of those reflections with you.

Investing, ultimately, is the attempt to predict future price movements and use those predictions to either try to beat the index or to produce returns in a less volatile fashion than the index (Verdad focuses on the former).

Investors differ widely on what methodology they believe will be successful in leading to future price knowledge. On the one end of the spectrum, pure quants see investing as a pattern recognition problem and, on the other end of the spectrum, fundamentalists believe deep knowledge of industry, company, and management will lead to superior forecasts. At which temple do we worship? Which prophets-of-profit are closest to the truth?

This need not be a question of faith. We believe this is, ultimately, a question for epistemology, the branch of philosophy that interrogates the difference between justified beliefs and opinions, and the methods we humans can use to make such a distinction.

Perhaps we should start with a simple question: Have we learned anything at all definitive about the field of investing since academic inquiry began in the 1930s?  So much of academic finance has proven faulty — from the Dividend Discount Model to the Capital Asset Pricing Model and many other stale theories pushed upon business school students.

And so, at the very least, we’ve learned that investing is not a science like physics. Those who look on the quants and their “mathematical hocus pocus”[1] with suspicion are right if they restrict themselves to arguing that the certainty of quants is crude and subject to being overruled by the evolution of a dynamic market. The best quantitative models often have an R-Squared of about 0.3; that is, they can explain 30% of reality. Physics this is not.

So we believe that to the extent that quantitative finance is part of the high-modernist quest for a rational design of social order equal to the scientific understanding of the natural sciences, it is a failure. The “fatal conceit” common to these ambitious plans was recently expressed well by the political anthropologist James Scott in Seeing Like a State: “Any large social process or event will inevitably be far more complex than the schemata we devise, prospectively or retrospectively, to map it.”  In academic finance, there is still a spectre of the high-modernist philosophy haunting the profession.

But the purpose of investment knowledge is not to place a rocket on the moon. In most cases, it is to launch a portfolio of securities or assets that outperforms the market. A marginal advantage over a purely random guess may secure valuable advantage.

And we have learned a few things about markets. Markets often display persistent trends and suffer from temporary exuberances along fashionable lines of thought. Harvard professor Andrei Shleifer noted “people do not deviate from rationality randomly, but rather most deviate in the same way…investor sentiment reflects the common judgement errors made by a substantial number of investors, rather than uncorrelated random mistakes.” There should be opportunity for arbitrage where execution costs are low (such as long-only strategies), and the empirical approach should offer some hints at where opportunities might lie.

We believe this should be evident from the consistent outperformance of value investing since Benjamin Graham and David Dodd formalized the approach in the 1920s. Academics such as Fama and French document the nuanced but persistent outperformance of value; Piotroski shows that including analysis of historical financial statements improves returns; and others have conducted rigorous empirical research into the benefits of considering select factors, such as momentum.

Our own research suggests that investing in leveraged small value equities that are paying down debt and have a low risk of bankruptcy may achieve outperformance of public markets comparable to what LBO firms achieved in the 1980s and 1990s.

On the other hand, the pure fundamentalist believes his predictions are defensible on account of more in depth knowledge of a given industry or company. We believe the “better knowledge” line of argument is dubious in that it is impossible to distinguish which facts are relevant and unknown. History and the literature on expertise within the realm of complex social phenomenon with disaggregated causal data suggests that base rate assumptions are more accurate than the predictions of the most recognized experts in such fields. The track record of fundamental active management seems to confirm this critique in the investment profession as well.

However, to transition from patterns-of-past to plausible predictions-of-future, we need some theoretical hypothesis of causality. We justly would expect that value investing may continue to work not merely because a regression analysis shows that it has for 100 years in different environments. We believe that value investing is an epistemologically defensible hypothesis because it marries a plausible understanding of persistent elements of human nature with a record of empirical evidence in history.

Where the two approaches combine, we may begin to make epistemologically defensible hypotheses about future price movements. Using empiricism to identify persistent historical trends and a philosophical eye to probe theory and identify cracks in the armor of fashionable theories should be in our opinion the surest methodology — and epistemology — for discovering the elusive goal of future price knowledge.

Dan and I are both humanities scholars turned investors; investors who make extensive use of the quantitative end of the spectrum. We find it intellectually indispensable alongside rules-based principles as a disciplining mechanism against the hubris of overconfidence in our flawed and often anecdotal human reasoning. At the same time, we believe that the future is fundamentally less predictable than our psychological biases guide us to believe. The more confident an individual is that he can predict a complex outcome over extended horizons, the more we are suspect. The more we see this hubris in the aggregate, the more we smell potential opportunity worth investigation. [1] A term I borrow from an 1880s German philosopher who criticized the esoteric and inaccessible language used by Spinoza, the “Great Rationalist” of 17th century philosophy, for its heavy reliance on mathematical formulations and Greek symbolism: “[T]he hocus-pocus of mathematical form with which Spinoza clad his philosophy…in mail and mask, to strike terror at once into the heart of any assailant who should dare to glance at that invincible maiden and Pallas Athene: how much of personal timidity and vulnerability this masquerade of a sick recluse betrays!”