Greenline Partners sees data mining as a “huge risk” in factor-based investing strategies, according to the firm’s co-founder and chief investment officer Maneesh Shanbhag.
Many factors don’t work in practice, he claimed, and even the most popular factors — the factors known as value and momentum — may turn out to be less effective in producing alpha than historical evidence shows, he said in a blog post this month on Alpha Architect. According to Shanbhag, past results may even defy logic.
“Investors can avoid being fooled by backtests by always keeping in mind that most attempts to beat markets will fail because trading is a zero-sum activity,” wrote Shanbhag, who worked for Bridgewater Associates before forming Greenline in 2012.
Backtests done by the firm show that value investing beats growth stocks with high price-to-earnings ratios, and that “past winners” — how Shanbhag refers to momentum stocks — beat “past losers.” But according to Shanbhag, the evidence from these backtests defies common sense.
Cheap value assets, he points out, are essentially the same as “past losers,” while momentum stocks are highly similar to growth assets. By Shanbhag’s logic, value can’t beat growth if “past winners” beat “past losers.”
“Both cannot simultaneously be true,” the Greenline CIO argued, explaining that smoke can’t be both created by fire and produce it. “If our logic is correct,” he wrote, investors should be able to observe that value is highly correlated to stocks that have recently underperformed, while growth should be tightly linked to recent outperformers.
In a phone interview, Shanbhag blamed human bias and the use of historical data in investment marketing materials for the tendency of people to be misled by factors.
“Eyes light up” when investors see evidence of high, market-beating returns, he said by phone. “It’s human nature to look at the good historical results and come up with a good explanation for it,” he added.
When historical data doesn’t line up with Greenline’s fundamental understanding of investing, the firm will dismiss that evidence in favor of logic, according to Shanbhag’s blog.
“In this era of big data and cheap computing power, it is easy for anyone to create a winning investment strategy in a backtest,” he wrote. “But investing is forward-looking and markets are adept at pricing in known information.”
[II Deep Dive: Here’s the Fastest Growing Strategy in Smart Beta]
Validea, which provides investment research, has also warned this month about the pitfalls of factor-based investing.
“Some investors believe that quantitative and factor investing are areas where mistakes are less common,” Jack Forehand, Validea’s president, wrote in a blog on the firm’s website. “After all, if you have a computer running the show that has no emotions or biases, the result should be a process less prone to mistakes.”
But he cautioned against believing the past always repeats itself, as factors may stop working. It takes patience – possibly a decade of waiting – before factor-based investing proves successful.
“Factor investing works over long periods of time,” Forehand wrote. “In the short-term, the periods of underperformance can be both long and brutal.”
