Quant funds are expected to account for US$1 trillion in assets this year. But when algorithms are managing this money, who manages the algorithms?
Computer-powered, data-driven strategies have continued their spectacular march towards dominating the hedge fund industry.
According to research by HFR, the amount of money being managed by quant hedge funds rose to more than $940 billion by the end of October 2017, and is on course to pass the $1 trillion mark at some point this year.
This is an enormous amount of money being managed by machines.
As the Financial Times noted, quant funds are the fastest-growing segment of the hedge fund industry, and money is flowing towards systematic strategies at an increasing pace.
For example, one of the world’s largest quant funds, Two Sigma, grew by more than 700 per cent (from $6 billion to more than $50 billion) between 2011 and 2017.
Black box worry
Not everyone is heralding the meteoric rise of quantitative investing. But it’s not the role and impact that data will play in hedge funds – and across investment industries – that’s in question.
Data will unquestionably play a critical role in the future of investment strategies, and anyone who says otherwise will find themselves on the wrong side of history.
What is worrisome about this approach is the “black box” nature of many quant funds.
They are run by complex algorithms that are really only understood by the engineers and data scientists that created them.
As the FT put it, some fund managers “fear that the money gushing into both simple and complex algorithmic trading strategies is making markets both more complex and fragile”.
We’ve already had a glimpse of what can happen when machines are left to make trading decisions alone, and how an unexpected event can cause a negative chain reaction.
In addition to several “flash crashes” over the past few years, the Quant Quake’s 10-year anniversary last summer served as a reminder of the inherent risks of placing too much trust in computers.
There’s no way to avoid risks completely, and there will always be unexpected events that rock Wall Street.
But humans can mitigate this risk by taking a thoughtful and measured approach to these types of events. That’s the benefit of Man Plus Machine.
Right now, when algorithms are managing the money, it’s important to ask: who’s managing the algorithm?
For funds that trade automatically, an unexpected event means one that a quant failed — or didn’t have the foresight — to code into his or her algorithm. And when that unforeseen event occurs, bad things happen.
Artificial intelligence may have the answers to these challenges one day, but we’re definitely not there yet.
In the meantime, maintaining human involvement while developing and executing trading strategies will be key.
The continued push to give data a central role in trading decisions will add complexities to human involvement, especially as firms fight for the limited talent that truly understands this technology.
So the only way to move forward is to make that technology easier to use for everyone.