Data Science can’t crack the Stock Market

Data science is eating jobs, so why can’t algos (e.g. RenTech) “consistently” beat returns the same way as programs can beat high ELO rated players at chess?

The markets are more poker than chess, like life itself – where luck plays a defining role. Algos can break down factor-based investing and even use retail sentiment indicators (such as ICICI Direct’s RIBI) to make bets.

A data scientist will break LUCK as “stacked probabilities” in strategies backed by back-testing. This lies at the heart of algorithmic trading.

Let’s take the Put Call Ratio or PCR. PCR over 1 is seen as a bearish indicator but also as a contrarian indicator. Can an algorithm bake in “contrarianism”?

What happens to valuation defying momentum returns in the grip of panic like the crash in FAANG stocks?

What about Trump’s ever changing stance on trade wars?

Business is war. War by nature is asymmetric and unpredictable. In 1942, Germany’s surrender was thought to be around the corner due to its failure to secure Russian oil fields. Germany ground on till 1945, partly due to innovations in equipment and in using synthetic fuel (Fischer-Tropsch process).

Companies similarly are an ever-changing, non-ergodic process, with new inventions, new information and predicting their moves can be like predicting the weather on mars.

So you can take the emotion out of the markets but can you ever take the market out of emotions?

Nevertheless, those who sell shovels for this new gold rush to extract Alpha out of Data will get incredibly rich in the process.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s