Pure skill, dumb luck

Welcome to Money Buff! Feel free to check out my other posts, and follow me on Twitter for more regular thoughts! For more about the hows and whys of investing, subscribe here:

Chess players make hundreds of decisions in the course of one game. It’s complex and difficult to perfect. Like other games, participants learn the right lessons from experience. There are stable variables and players see the immediate consequences of their actions through accurate feedback.

The game offers decision-makers a “kind” learning environment. This is a term coined by psychologist Robin Hogarth, meaning that lessons learned from experience are reliable. Unfortunately, investing is rarely this way and Hogarth would contrast it as a “wicked” learning environment, where information is clouded in such a way that we struggle to identify causality.

Take Bill Miller, famous for beating the S&P 500 for 15 straight years. Money Magazine called his fund the “scariest portfolio ever”. Thanks to his bold bets on questionable companies, Miller generated a cumulative return 2.13 times that of the S&P 500 between 1991 and 2005.

But nothing good lasts forever, and the fund lost its way during the Great Recession. Success in prior crashes gave him deadly confidence, and the playbook of past decades was long gone. Miller told the Journal he failed to recognize, “the severity of this liquidity crisis”. Fifteen years of outperformance was erased within two. His mistakes overshadowed decades of hard work and success.

If you asked investors in the 90s about “the streak”, many would say Bill Miller was a market wizard. In hindsight, it’s more likely he experienced a stretch of outstanding luck. Granted, he recognizes this:

Nonetheless, his story points out a hard truth about investing — it's very difficult to distinguish between luck and skill. Asset returns are so variable that if there is skill, it’s very difficult to tease out. Michael Mauboussin outlines this phenomenon using a luck-skill continuum in his book The Success Equation:

On the left side are slots and roulette. There is an inherent house edge, so despite any strategy, large wins are driven by luck. You can’t intentionally lose playing these games, further indicating that it is pure luck.

On the other side is chess, where player ratings range from 100 - 3000. If someone with a score of 2500 played someone with a score of 500, the 2500 player would win every time. There is little variance. Also, one can lose chess on purpose, showing a massive skill component.

In sports, skill is emphasized by the number of possessions a team has. Said another way: the number of opportunities to use your skill. Basketball and tennis are sports where possession is constantly changing sides, thus having a higher skill factor. This is why we often see superstars of these sports rise to the top.

You’ll notice that investing is towards the luck end of the equation, but not entirely. No one can consistently forecast market movements but that doesn’t mean people are spinning roulette wheels on Wall Street. Investing has its own luck-skill continuum and it is almost entirely a function of variance over time.

Bill Miller’s fund ran with a very high standard deviation. And in probability distributions, you get to the tail by increasing the variance. Yes he finished with a higher geometric return than the S&P 500, but his record becomes less impressive on a risk-adjusted return basis.

Today we hear about the crypto bros who turned pocket change into millions overnight. The losers may have had great methods, but they were unlucky.

In the short run, one can do everything right and fail, or do everything wrong and succeed. Success stories often downplay the role of luck, while stories of failure tend to overplay the role of luck.

Trust the process

A consistently good process will diminish variance, leading to success in the long run. It’s far less likely that such a process will frequently produce poor results. Warren Buffett has had down years, but he’s true to his process and it remains rock solid.

You’re probably wondering, “how do I know if a process is good?”. Causality is messy in today’s world and it’s too easy to form the wrong conclusions.

“Since Y happened after X, Y must have caused X.”

You wore new cleats and then missed a penalty kick. Must have been the cleats.

We rarely consider that the outcome could have been in spite of X. But our brains like stories of cause and effect. When faced with uncertainty around random or unlikely events, we are inclined to place a tidy narrative on these events to make them seem foreseeable - often changing our strategy as a result of one outcome.

Casinos play into this by giving slot machines an illusion of predictability, showing ‘near misses’ to keep players coming back for more.

In The Man Who Solved The Market, Gregory Zuckerman writes:

Some feedback is misleading and sometimes we simply misinterpret things. Your rationale for a market movement could be the exact opposite of reality even if it arrives at the same result. Reliable feedback can help recognize uncertainty and understand what influences outcomes.

As we saw with Miller, markets are fluid and domain expertise that outperformed then may not today. Extrapolating from small samples can be detrimental to long-term outcomes, as short-term results will not reveal skill. But with large enough samples, investors get tighter feedback loops for which to critique their process.

On the surface, generating alpha is the clearest evidence that an investment manager knew what he or she was doing. But there are a lot of very lucky investors viewed as geniuses. And it makes sense that investing is largely rooted in luck. The number of data points, tools, and analysts per stock is constantly growing.

The paradox of skill states that as the skill of participants increases, luck plays a larger role in the outcomes. You may think Tesla is a buy while someone else could have an equally good case for why it’s a sell. You’re both predicting the unpredictable, so someone has to get lucky in the short run.

Separating luck from skill gives a better understanding of the odds at the table you’re playing on. Professionals identify where an outcome exists on the luck-skill continuum and adjust future actions accordingly. They understand the role of both, while amateurs attribute success to skill and failure to luck.

When it comes to investing, “Adam Smith” (the pen name of George J.W. Goodman), once wrote, “If you don’t know who you are, this is an expensive place to find out.”