Some thoughts on quantamental investing.

Quantamental is the combination of quantitative and fundamental trading.

I’m thinking back about an article that made some good points about how algorithms, big data and cloud computing are changing all industries including finance and investment management. But it also noted that there is still a long way to go.

For faster trading where there is a lot of data, it is clear that computers dominate.

For slower trading with a lot more uncertainty, there are a lot more humans.

The human brain has an amazing ability to analyze and synthesize disparate data sources. Unfortunately this is offset by our computational limitations and many cognitive biases.

The plan naturally follows that we should develop deep learning approaches that replicate what the human is doing, while avoiding our limitations and biases.

Hmmm...

In the same way that the Wright brothers designed and built an airplane rather than a bird (albeit they studied birds), it may be that we don't need to replicate a human trader to be able to trade.

Stepping back, trading is trying to efficiently aggregate information into the price of various securities so as to be close to a theoretical fair value. This is valuable because accurately priced assets lead to better resource allocation and greater long term economic growth.

It is not obvious that replicating human trading is the best way to do this.

Perhaps instead we should be moving up one level to think about what we are trying to achieve in the financial markets (fair pricing of securities) and how best to achieve it.

Anyway, at Noviscient we have some thoughts on this. We call it Structured Dynamic Learning. I'll talk about it next week.

Struggling to raise capital? It’s time to do it the modern way.