Freestyle chess x investment process
How investors have to adapt their process to incorporate new sources of data and insight
IBM’s Deep Blue beat Garry Kasparov, the greatest chess player of all time, nearly 25 years ago. Kasparov became an advocate for using chess computers to practice and in tournament play, leading to an open-book format called ‘freestyle’, ‘cyborg’ or ‘centaur’ chess. Teams of players could freely consult computers during play. It turned out that man and machine, together, could still beat machine.
There isn’t much tournament play in the format but the cyborgs seem to have a 100-150 Elo rating point advantage over computers1, similar to the difference between ranks one and twenty-five in regular chess today. You’ve heard of Magnus Carlsen, but probably not Dmitry Andreikin (I know I hadn’t).
Machines got better faster than humans and it’s entirely possible that the best chess computer of today (DeepMind’s AlphaZero) beats the best cyborg team. Isn’t it surprising that cyborgs ever beat the best computers? You’d think our biases, unreliable memories and emotions would do more harm than good.
Tyler Cowen explored this in a chapter of his 2013 book, Average is Over. A year later, Michael Mauboussin wrote an excellent article exploring potential applications to investing. Both are worth reading; it’s a fascinating game2 and the investing analogy is fun to ponder.
I’m going to explore investing applications through a 2017 discussion between Kasparov and Cowen (my emphasis in bold):
KASPAROV: I think so. Again, it depends on the qualification of the operator.
COWEN: Sure, if it’s the best operator in the world, whoever that may be. Maybe yourself, maybe Anson Williams.
KASPAROV: By the way, I exclude myself from this category because I’m not a very good operator. I’m a very good chess player. A great operator does not have to be necessarily a very strong player.
COWEN: What makes for a great operator?
KASPAROV: Someone who can work out the most effective combination, bringing together human and machine skills. I reached the formulation that a weak human player plus machine plus a better process is superior, not only to a very powerful machine, but most remarkably, to a strong human player plus machine plus an inferior process.
At the end of the day, it’s about interface. Creating an interface that will help us to coach machine towards more useful intelligence will be the right step forward. I’m a great believer that, if we put together a good operator — still a decent chess player, not necessarily a very strong chess player — running two, three machines and finding the best way to translate this knowledge into quality moves against Rybka Cluster, I would probably bet on the human plus machine.
Kasparov mentions three things I want to explore.
One: In freestyle chess, magic is in how man extracts insight from machines. I think this is increasingly the case in investing - not just quant (where it’s core to the process) but also fundamental, bottom-up investing.
Fintwit just had a heated debate on the merits of reading 10-Ks. Paul Enright said it best: that’s table stakes. The days of finding investment insights by reading 10-Ks is over.3
Short-term investors are looking at credit card data, satellite imaging and footfall; you can’t call the quarter without it.
Long-term investors use similar data sources to estimate churn, marketing efficiency and cohort trends; you can’t estimate intrinsic value without it.4
Even the most long-term, fundamental investors must know how and when to incorporate insights from machines into their investment process. The magic happens in the process through which it all comes together.
Two: The best-rated players aren’t the best at freestyle chess. The first tournament was won by a team called ZackS - two guys from New Hampshire with ratings well below the 50th percentile of rated players. Their opponents were at the pinnacle of regular chess - over the 99.9th percentile. The skillset that wins freestyle is obviously different from what wins in regular chess.
Investing has always involved pulling together a mosaic of insights. But there’s work to do in adapting to new sources of data and insight. The S-curve in investment skill seems to flatten out after about 10-15 years of experience; adapting to become “operators” (see below) boosts skill.5 Failing to do so can leave you in the dust.
Three: Kasparov uses the word “operator” for humans that run the process. I couldn’t help but think of Tank. Is this what the fundamental investor of the future looks like?
As investors, we have a lot to learn about incorporating new sources of insight and data in our investment processes. Freestyle chess is an interesting game to draw from since machines generate the moves (insights), but humans still add value to the process. Lessons are that process matters more than ever and we’ll have to become better “operators” in our investment decision-making.
Cowen, Tyler. Average is Over.
Less fun to watch than chess boxing, though.
I suspect other common sources of insight are also being commoditized, e.g. expert interviews.
There’s probably still a behavioral edge to be had, too.
I’ve always admired Bill Miller for his adaptability and first principles thinking. He sets a good example. Who else has Farfetch, DXC, a SPAC and Uber calls in their top 10? I can’t speak to his process, but the outcome looks very different to most funds.