The Delusion of Market Efficiency
Markets have potentially become less efficient in recent decades. There are several reasons why this might be the case.
Trent Ambler, MSF | Portfolio Manager
Tommi Johnsen, PhD | Advisory Board
Key Point: Markets have potentially become less efficient in recent decades. There are several reasons why this might be the case.
Implication: Market inefficiency means more opportunities for outperformance for sufficiently equipped investors.
____________
The question of market efficiency is of great consequence. We introduced the concept in a recent post and suggested that markets have perhaps become less efficient recently. The post can be found here:
https://substack.com/@5thhorizon/p-147219253
There are at least two significant takeaways from this post.
1. Systematic pricing anomalies persist in the U.S. equity market. ‘An efficient market is one where the market price is an unbiased estimate of the true value of the investment.’[1] Under this definition the U.S. market for public equities (the stock market) is most certainly NOT perfectly efficient. There are persistent and identifiable pricing anomalies in the stock market whose existence is not random.
2. The number of pricing anomalies is increasing and are likely underreported. Idiosyncrasies in the market pricing mechanism have become more numerous in recent decades. Recent studies that examine the timing of portfolio formation around data releases suggest further that perhaps certain academic conventions have actually served to understate the number of pricing anomalies that are available to investors.[2]
Why is this happening and what does it mean for equity investors going forward?
Some possible explanations:
1. Market participants do not correctly process available information. This problem has gotten worse as both the set of available information and our ability to access it have grown. Generating an unbiased estimate of the future value of an investment is made harder when more information is available, and in a way that is relatively easy to access. Discerning amongst all the datapoints to arrive at an unbiased estimate of value is harder as the number of datapoints increases.
“A very rarely discussed property of data: it is toxic in large quantities.”[3]
There is likely to be more noise than signal in a large dataset. Give someone limited data infrequently and they will delay their evaluation response until better information is presented. Give them more data more frequently and they will mistake noise for signal and rush to a faulty conclusion. Overconfidence and confirmation bias become a huge problem here because ideas are sticky, and assessments made from noise are not easily abandoned, in fact conviction for an originally faulty assessment increases as more information, however contradictory, is presented. “The more detailed knowledge one gets of empirical reality the more one will see the noise and mistake it for actual information.” [4] Of course, both scenarios, limited data released infrequently, and lots of data released quickly are sensitive to issues around datamining but this is another topic entirely.
When making assessments about the future people often fail to consider the “predictive validity” of any current or past evidence that is presented. “People do not combine strength and weight [of the evidence] in accord with the rules of probability and statistics … People [instead] focus on the strength of the evidence – as they perceive it – and then make some adjustment in response to its weight.”[5]
2. The number of active investors who are less prone to behavioral bias and thereby heuristic driven pricing errors has decreased, meaning the pricing exercise may increasingly be driven by investors who are prone to identifiable pricing anomalies. A significant rationale for this possibility can be found in the rise of indexing. If a larger number of more rational investors have left the active space in favor of indexing leaving their less informed, less equipped counterparts in the active space then prices will increasingly be set by the latter.[6]
3. The reach for yield has potentially compromised the ability to correctly price risk. In the wake of the 2008 global financial crisis central banks across the globe cut interest rates aggressively and perhaps more importantly kept rates low for a long time. Low interest rates drive down returns on safer, lower volatility financial instruments (fixed income securities, money market funds, etc.) making these instruments less attractive to investors who need to realize tangible returns in their investment portfolios. This is a practical outcome of lower rates, but an even more interesting implication of lower rates is an actual psychological breakdown in the capacity of investors to accurately perceive and evaluate risk.
Several possibilities explain why this might be the case, including the idea that investors use reference points for investment returns and may deviate from rational portfolio theory behavior (mean-variance optimization) when reference points are violated.[7]
4. Rational pricing exercises have suffered with the rise of the internet, social media, and the democratization of equity trading by online brokerages. Market efficiency is in no small part dictated by the capacity of a large number of disparate players to express individualized opinions on the value of an asset. The “wisdom of the crowd” requires that individual members of the crowd act independently when expressing their views. When social media platforms challenge this independence by offering up a space for groups to coalesce around biased, irrational views, a previously wise crowd can turn into a dangerous mob.[8] The stories that garner the most attention on social media do not always achieve that status as a consequence of their accuracy, it’s not all biased junk but occasionally it is.
Access to retail brokerages and trading platforms has exploded in the last decade, trading costs at least appear free, equity investing has been embraced and glorified in entirely new ways. The democratization of equity trading carries many benefits for individuals and society at large. Equities are a powerful tool. The capacity to not only preserve, but also to meaningfully grow wealth via equity exposure is, in many ways, unmatched. It is a good thing that more investors have been granted access to such a remarkable vehicle, but not everyone who participates in equities is completely disciplined and rational in their trading. There is a bit of gamification of equity trading at play here. Some segments of the investor base treat equity trading like gambling. The behaviors exhibited by investors within these segments are certainly not rational nor conducive to an efficient market. If active investing is sufficiently represented by those in the gamification segment, then market efficiency suffers.
Diminished market efficiency means that asset prices are increasingly set from a place of bias, there have always been anomalies in the market but the number of these anomalies and the conditions under which they persist have expanded in recent decades. There is meaningful opportunity for active investors who are sufficiently disciplined, and more so, who are sufficiently equipped to capitalize on pricing errors because of this shift. Some examples of how to answer the recent trend and actually benefit include the following:
1. If more data means more pricing errors, then there is an opportunity to generate excess return by building better systems to process and parse the vast set of available data. The systems used by managers to generate buy and sell decisions have never been more important. Investment managers that know how to cut through the noise and deliver actual signal will have a huge edge in the new era.
2. Fewer “smart money” investors in the active space means more opportunity for those that remain. In some ways active investing has never looked better than it does today. Garnering an edge is not easy but it can be done, the rewards for doing so are bigger today than they have been in a long time.
3. If risk is being systematically mispriced, then there should be ways for an investment manager to capitalize on this trend. Momentum and Sentiment factors have delivered impressive results for at least a decade, it is possible that at least part of their performance can be explained by a reach for yield.
4. Markets don’t care about your emotions. Portfolios which are borne from processes that mitigate the impact of human bias thrive as the sources for and number of these biases increase. Quantitative investing is one example of such processes. One of the greatest strengths of quantitative investing is its dispassionate and regimented approach to building portfolios. The best quantitative investors know how to find investing signals and even more how to represent those signals in a way that moderates the impact of human bias.
The idea that the U.S. stock market is perfectly efficient has always been a bit of a delusion. A perfectly efficient market requires that asset prices are set from an unbiased view of the future. There is room for pricing errors in an efficient market but these errors need to be random and offsetting such that the probability of finding mispriced assets converges on 50-50 over time. If, on the other hand, there are consistent biases in the pricing mechanism markets will be something less than perfectly efficient. It is increasingly clear that such biases do in fact exist and evidence suggests that they have become more numerous and relevant in recent decades.
[1] Damodaran, Aswath. Investment Valuation Third Edition, Wiley, 2012. P112
[2] Bowles, Boone and Reed, Adam V. and Rinnggenberg, Matthew C. and Thornock, Jacop, Anomaly Time (November6, 2023). Journal of Finance, Forthcoming
[3] Taleb Nicholas, Nassim. Antifragile, Random House, 2012. P126
[4] This entire paragraph is derived from insights presented by: Taleb Nicholas, Nassim. The Black Swan Second Edition, Random House, 2010. P144. Direct quotes where appropriate. See also: Silver, Nate. The Signal and the Noise, The Penguin Press, 2012. Pp 329 - 369
[5] Dale Griffin, Amos Tversky. The Weighting of Evidence and the Determinants of Confidence, Cognitive Psychology, Volume 24, Issue 3, 1992. Pp 411-435
[6] Asness, Cliff S., The Less-Efficient Market Hypothesis (August 30, 2024). 50th Anniversary Issue of The Journal of Portfolio Management, Forthcoming
[7] Lian, Chen and Ma, Yueran, and Wang, Carmen, Low Interest Rates and Risk Taking: Evidence from Individual Investment Decisions (August 22, 2018). Review of Financial Studies
[8] Asness, Cliff S., The Less-Efficient Market Hypothesis (August 30, 2024). 50th Anniversary Issue of The Journal of Portfolio Management, Forthcoming