Beware of geeks bearing formulas

We live in a time when decades’ worth of financial data is stored and easily accessible. With computers, we’re now able to store decades of financial and market data, retrieve it and analyze it all within a few minutes. Numerous free websites can calculate betas and derive alphas, not to even mention all of the option “Greeks.” Speaking of Greeks, Buffett once channeled Laocoon’s warning from Virgil’s Aeneid (Wikipedia), when explaining how sophisticated financial institutions were wiped out : “…they had all these types from Wall Street, you know, and they had advanced degrees, and they look very alert, and they came with these — they came with these things that said gamma and alpha and sigma and all that. And all I can say is beware of geeks, you know, bearing formulas.” (CNBC)

The point is that measurement of the past can be comprehensive and precise, but cannot predict the future.

The best analog to this issue is baseball statistics. Baseball statistics can be found as early as the 1800s, similar to a lot of market data. Player and game data was recorded by hand and stored in paper format, until the latter half of the 20th-century (also like a lot of financial data). The advent of computers allowed decades worth of data to be aggregated and analyzed. New statistics were developed, beyond the standard existing statistics. For instance, statisticians could now look at hitters’ “batting average on balls in play” (BABIP), rather than simply relying on the traditional “batting average” (BA). Teams can now measure every player’s performance to the fifth decimal place, but the breadth of measurement and degree of precision does not always translate to winning. Below are at least three reasons why:

Assuming all the data, analysis, and statistical output is perfect, someone still has to decide what data is important and what data is not. What statistics measure meaningful factors and which ones are noise? This is a challenging task for both baseball teams and investors.
Knowing what data is important is not enough though, since the future is not known. It may or may not be like the past. A young player may become better or worse, or he might stay the same. Who can accurately predict this consistently? Very few, in any.
Even if the above challenges could be met, in a competitive market (whether for baseball prospects or financial assets), prices move up and down to arbitrate supply and demand. If a player contract or stock is bought at too high a price, team or portfolio performance can suffer.

There is A LOT of value to performance measurement and analysis (that’s another post). but it cannot predict the future. As nearly all investment literature discloses: Past performance does not guarantee future results.