Why does Financial Machine Learning Usually Fail???

Why does Financial Machine Learning Usually Fail???

Quantitative finance has a high failure rate, especially for financial ML. Those who are successful, however, can earn a large number of assets and deliver consistent, exceptional performance to their investors. Why do so many people fail in this field? There is one critical mistake that leads to all these failures.

The Sisyphus Paradigm:

Discretionary Portfolio Managers (PMs) make investment decisions that are not reliant on any particular theory or rationale. They consume raw news and analyses but mostly rely on their judgment or intuition. They may rationalize those decisions based on some story, but there is always a story for every decision. Because nobody fully understands the logic behind their bets, investment firms ask them to work independently from one another, in silos, to ensure diversification.

If you have ever attended a meeting with discretionary PMs, you may have noticed how long and aimless they can sometimes be. This is because each attendee is usually only concerned with one particular piece of information, and giant argumentative leaps are made without proof. Even though this does not mean that discretionary PMs cannot be successful, a few of them are. The point is that they cannot work as a team as naturally as other groups. Hence, making them work in silos and minimising interaction makes sense.

When such a formula is often applied to quantitative or ML projects, it always leads to disaster. The boardroom’s mentality is to treat quants the same as discretionary PMs. This approach usually doesn't work out because each Quant will end up frantically searching for investment opportunities that fit their criteria. Eventually, they'll, either.

(1) find a False Positive that looks great in an overfit backtest or

(2) settle for good old academic standard factor investing with a low Sharpe ratio.

Neither of these outcomes will be satisfactory to the investment board, so the project will likely be cancelled.

Source: Marcos Lopez De Prado @ Advances in Financial Machine Learning

Many Factor Funds with Low Sharpe Ratio is launched. This is Old School !!

#Quant #Quantace

Follow?Quantace Research

Introducing?Quantace Chetak (Aggressive Momentum Factor).?This basket of ours has generated 33% returns since Mar 17. In the same time period, the NIFTY has given 3% returns. We generated 30 % Alpha :). Do check us out.