And today, the cheapest smartphone on the market can blow that computer out of the water.
The point is that computing power has grown a trillion times since then. Its impact on our lives is profound. So, then why are we still investing like it was in the 1950s?
They have already moved on in the US – 60-70% of trading activity is rules-based. Mutual funds managed by human experts are losing flows to rules-based instruments like ETFs.
You can’t blame them.
SPIVA, S&P’s research arm, shows that globally, human mutual fund managers are unable to beat their benchmark.
“The evidence from more than 50 years of research is conclusive: for a large majority of fund managers, the selection of stocks is more like rolling dice than like playing poker. At least two out of every three mutual funds underperform the overall market in any given year.” – Thinking Fast and Slow.
Also, who do you think the most successful investor of all time is? That’s right it’s a quant fund. Renaissance Technologies Medallion Fund has generated a 60%+ gross CAGR since the 1980s.
John Bogle, the founder of Vanguard (which today has $7 trillion AUM), predicted in the 1970s that over the long term, humans would find it impossible to beat the index.
At last count 1 in 3 hedge fund managers in the US use quantitative strategies. In China, quant AUM has doubled to nearly $100bn in 2020.
The problem lies in our genetic code. All of us, whether experts or not are subject to the same biases.
It is pretty well established that herd mentality, recency bias, confirmation bias and loss aversion, extrapolation bias, and many more affect us all.
Just to be clear, technology-based investing or quant investing is not a style of investing. It is a tool that is applied to various niches – momentum, arbitrage, macro, value, growth, etc.
With technical-based trading, the name of the game is analysing patterns in price volume charts, so technology is a natural fit.
But when it comes to long term fundamental based investing, it still remains a domain of humans. The reason is that it’s a more complicated process for sure, and companies need to be looked at contextually.
However, this contextual outlook is not only the boon but the bane of human investors as well. As we discussed above, it’s where all the biases that plague us come into play.
Everyone human manager will talk about discipline and claim to have their rules and systems in place. But, humans can’t help but break their own rules.
There is always a reason for the exception. Sure enough, over time the framework gets corroded, while trying to chase the latest fad.
In today’s context, this means machine learning. Tomorrow it could be more sophisticated techniques. The actual technology that powers these machines is changing dramatically – factor investing or static excel filters were great in the past but the next evolution is self-learning algorithms.
Earlier computers were used, just codifying human foibles. Today with ML, we have gone one step ahead and can now learn at a meta-level.
For example – if a human expert says 20x PE is the magic number. Everything above it is overvalued and vice versa. The machine will say why is “20” THE number? It’s a moving target contextual to a myriad of other things in the market, impossible for a human to parse.
But, with the help of machine learning, we might not get a perfect answer but one that is on average better than the rest of the market.
The key is to do better than average most of the time. Not perfect at all times. Too often, the barometers for machines become an unreasonable one.
Even here human bias ends up costing us – because we would rather pick an even more unreliable human over an imperfect (but better) machine.
(The author is Co-founder of Upside AI)
(Disclaimer: Recommendations, suggestions, views, and opinions given by the experts are their own. These do not represent the views of Economic Times)