In his book, The Success Equation, Mauboussin also writes about the importance of differentiating between experience and expertise. He includes a quote from Professor Gregory Northcraft, a psychologist at the University of Illinois, in the book: "There are a lot of areas where people who have experience think they're experts, but the difference is that experts have predictive models, and people who have experience have models that aren't necessarily predictive (page 23)." He talks about the need for deliberate practice (the hard work and thousands of hours needed to master a skill), and how investors should seek to differentiate between experience and expertise to understand the value of predictions about the future.

One of the clear signs of expertise, for example, is one's ability to make accurate predictions, or use a model that effectively ties cause to effect. In investing, for example, the Valuentum team has built and updated 20,000 discounted cash-flow models over nearly 10 years, testing cause and effect. Other firms may have done the same. One might conclude that this expertise alone may give a leg up on the experience of some investors that may just be passive onlookers of market price movements (even those watching for decades) or just crunching the numbers, making little connection between what actually drives stock returns.

The Success Equation goes into many examples of how "statistical significance" is hurting many professional fields, perhaps one of my favorites a comical reference to a paper published in the peer-reviewed journal The Proceedings of the Royal Society B. You might have heard of this one. The work suggested that women who ate breakfast cereal were more likely to give birth to boys than girls. It's difficult to believe this "entertainment" gets published in peer reviewed journals, but such work may be no more significant than some of the most widely disseminated beliefs in finance, which brings us to the main point of this article.

Whether it's Nobel laureate William Sharpe's arithmetic (which can easily be dismissed by assuming one large investor underperforms leading to hundreds of smaller investors outperforming) or how Vanguard's founder Jack Bogle uses dividend yield as an incremental measure of total return (as opposed to capital appreciation that would have been achieved had the dividend not been paid) or the very basic construct of factor investing, itself, (which falls flat as most factors are based on ambiguous, realized and backward-looking data)--Mauboussin's take on the paradox of skill in investing is just as disappointing.

Let's define the paradox of skill first. The paradox of skill is the view that as more and more participants of a sport or business become more skilled, the variance between the outcomes narrows. It may be fair to say that the opposite may also be true. As more and more participants of a sport or business become less skilled, the variance between the outcomes might narrow (for example, if a group of Little Leaguers doesn't have the strength to hit homeruns, the league home run tally will converge to 0.). Mauboussin uses the examples of batting averages and earned run averages in baseball as well as marathon runners to illustrate his view. However, the author indicates that there are other endeavors such as professional basketball, where, as skill has improved, the variance of outcomes has actually increased. 

---
The statistical tendencies speak for themselves. Holding more than 20 stocks in a stock portfolio, for example, pretty much eliminates unsystematic (firm-specific) risk and should push the returns of heavily diversified portfolios to the market average, reducing the standard deviation of excess returns. Over-diversification is the enemy of outperformance (outsized standard deviation of excess returns), and the means by which to achieve mediocrity. Here's a quick take from Morningstar that hits the nail on the head on the risks of over-diversification:

The author, while highlighting in his own words in The Success Equation (see quote at top of this article) how stock portfolios have conformed over time due to a reduction of active share brought about by myriad influences in how active managers are "playing the game," doesn't emphasize enough what we believe is the correct conclusion for the observation of declining standard deviations of excess returns relative to holdings-based benchmarks.

Investors are conforming to a similar playbook due to conflicting incentives, and then they are being measured by holdings-based benchmarks that pretty much match their exposures. Oftentimes, active managers are even penalized if they deviate from their benchmarks (by allocators looking for targeted exposure), and sometimes and ironically, they are punished when they beat their benchmarks by too large of a margin -- because some may believe those managers are taking on too much risk. It's near-impossible for the standard deviation of excess returns to expand in such an environment.

Maubousin notes that famous researchers Amos Tversky and Daniel Kahneman wrote in their 1971 paper, "Believe in the Law of Small Numbers," that humans have "'strong intuitions' that are 'wrong in fundamental respects,' a shortcoming 'shared by naïve subjects and by trained scientists.' The flaw, simply stated, is that (humans) have a tendency to believe that a small sample of a population is representative of the whole population (page 216, 217)." In this case, humans believe that there is a paradox of skill in investing based on a small sample of the population, the 15% of the stock market that makes up active fund managers. With that said, here is what could actually be driving the reduced standard deviations among excess returns, from Value Trap:

The peculiar connection between 1) the propagation of spurious (quant) factors, 2) quant's influence on fundamental investment frameworks (e.g. some fundamental value investors think in terms of quant value factors)--an effect that reduces the dispersion of skill and returns as professional approaches become more congruent and undifferentiated (think conformity)--and 3) severe active fund underperformance [in part, because of holdings-based benchmarking] during the past 15 years is the primary reason why (we) do not believe professional active investors are becoming more skilled, and that's why they are underperforming [their benchmarks] after fees, as outlined in the popular "paradox of skill" thesis. [Professional active investors may actually be becoming less skilled as quant and passive investing proliferates.] (Our) view [instead] is that, as indexing and factor investing proliferates, alpha may actually (be) shifting from professional active investors to other parts of the stock market [as Sharpe's arithmetic may imply on a grander scale]. For example, other investors--defined as hedge funds, pension funds, life insurance companies, and individuals by the Investment Company Institute--hold roughly 70% of the stock market, as of 2019. Households own about 30-40% of the stock market, by some estimates. [We posit simply that wider dispersion of returns can be found in the other 85% of the market not impacted by the various factors outlined in this article.]

There's a lot of informational value in reading The Success Equation (and everyone should pick up a copy), but please be careful to come to your own conclusion. From where we stand, there is not a paradox of skill in investing (or baseball, for that matter). The games have simply changed based on new incentives. Some wise person may have written this before: Be careful not in what you read, but rather in the conclusions you draw from your reading. We wish Mauboussin could re-write The Success Equation considering some of the thoughts in this article. Maybe he will!