There is a new theory making the rounds on Wall Street.
It holds that historically low volatility is the product of smarter investors who are armed with more information.
It is a cool theory. One that’s backed by plenty of learned people. It is certainly topical, and convenient.
Unfortunately, this theory is also probably untrue. Even worse, it could be dangerous.
Investors should be aware. They should be taking precautions.
You know the premise. Investors see things more clearly through the lens of history. They have learned the hard lessons of previous investors.
However, it’s different this time. No, really.
That’s the basic idea behind a story in The Wall Street Journal this week about how quantitative investors spread a calming effect over stock averages.
The idea is that these investors, armed with powerful computers running bulletproof algorithms, never get emotional about stocks.
They don’t overreact to news. They calmly buy every dip.
“If there are fewer things that lead to surprises because information is going from completely unknown to completely known, then you should have less volatility,” says Leigh Drogen. He’s a former quant trader and chief executive officer of crowd-sourced earnings-estimate provider Estimize.
Like I said, it is a cool theory.
The rise in quantitative investing follows fantastic advances in data science. More powerful computers coupled with bespoke (customized, or tailor-made) software is facilitating better modeling.
David Lun, a hedge fund manager at Taaffeite Capital Management, honed his craft by studying cell interactions in a petri dish of algae.
The MIT computational biologist spent a decade developing an artificially intelligent algorithm to predict cellular behavior. Making the jump to the financial world was an easy transition. Lun doesn’t see a big difference between the two worlds.
“If you have a causal relationship, you can make predictions and be somewhat confident in them,” he told Bloomberg.
The promise of algorithmic investing is attractive. I’ve seen the promise firsthand, as I have been working with data scientists for almost two decades.
In fact, when I was managing editor at MSN Money, I helped invent the first AI-based stock-rating system widely available for public use.
Understanding, managing and gleaning insights from algorithmic systems is now the biggest part of my approach. And it is part of the service I provide to my members every day.
So, I can say with authority that it’s true: Algorithms are not emotional. Unlike humans, they have no problem selling if the data changes for the worse on a once-favored position.
Ironically, that is precisely why investors should be skeptical about the current Wall Street characterizations.
When algorithms flip, they are not beholden to human frailties. They don’t wait to sell the next rally for fear of loss. And there is no sense of loyalty. When they want out, they sell.
And, like Game of Thrones, a chain reaction of abrupt decisions could get ugly, quickly.
Markets have been prone to booms and busts since inception.
I’m reminded of an insight from Jesse Livermore, a century ago. Widely considered to be one of the best speculators who ever lived, he had a simple guiding principle: Nothing is new.
He believed that everything you see in the market today has been seen before. The market doesn’t change because human nature doesn’t change.
Livermore understood markets are guided by fear and greed. A hundred or so years later, that is still true.
Even the quantitative algorithms developed by mathematicians model fear and greed. Lun’s cellular interaction models likewise are tribal by nature.
It is very tempting to accept everything as part of the process. So-called experts say we must be seeing historically low volatility levels because investors are smarter. They have more information, better tools. Or so the theory goes.
I think this is a mistake. Investors have not changed. It’s our opinion of their prowess that changed.
It is not different this time. So, I’m going to stay alert for abrupt changes in volatility ahead. You should do the same.
Best wishes,
Jon Markman
