The Momentum Factor Is Real. Too Bad It Doesn’t Work.
Decades of academic research support the existence of the momentum factor, but it can’t be put into practice in reality, according to a new study from Dimensional Fund Advisors.
The momentum factor exists, according to reams of academic literature — but fund managers can’t find a way to profit from it in the real world, according to new research available to clients of Dimensional Fund Advisors.
The momentum factor goes back to a 1993 paper that documented that investors could generate excess returns by buying U.S. stocks that performed well over the previous three to 12 months and selling those that had poor returns over the same period.
Dimensional Fund Advisors studied U.S. momentum-style equity funds with at least three years of history and found that the majority of these portfolios weren’t able to deliver above-market returns for investors after fees.
Dimensional — which excluded so-called mulitfactor funds, or funds that invest in momentum as well as other factors such as value — ultimately studied 11 funds since their start dates. (Factor funds, including momentum, are a fairly new category of fund, with most being live for less than three years; those funds were therefore excluded from the study.) The oldest dated back to 2003.
Most of the funds’ managers were able to actually capitalize on the momentum premium, but they incurred trading costs that wiped out the benefits, Dimensional found.
[II Deep Dive: David Booth Brings Academic Research to Life]
Ten of 11 funds in the latest study underperformed the Russell 3000 by 0.15 percent to 2.35 percent annually. The funds had turnover that ranged from 60 percent annually to 211 percent. The fund with the highest turnover underperformed the benchmark by the largest amount.
Dimensional has conducted larger studies of the momentum factor in live funds before and has come to similar conclusions. But this time Dimensional wanted to examine only U.S. funds that explicitly state that they are looking to capitalize on the premium.
Academics have been able to simulate momentum premiums in the lab, but practitioners haven’t been able to translate that into outperformance in most working momentum funds.
“I think the research is just a good example of why investors should look at live performance when they can, and not base investment decisions on simulated performance,” said Marlena Lee, co-head of research at Dimensional, in a phone interview.
Among other things, back tests don’t account for transaction costs. “[That] can drive a wedge between simulations and what you observe in live results,” she said.
As for why it’s not working, Lee explained that there isn’t a lot of consensus on whether or not this premium has a good reason to exist. In the 20 years since it was discovered, there has been a lot of research into the potential explanations, including ones based on investor behavior and risk-based reasons.
“When you have a pattern in the data, but there is not good reasoning behind it, then you have to ask yourself, ‘Why is it that smart investors haven’t traded the premium away?’” said Lee.
It may come down to trading costs. If it’s too expensive for managers to trade on the premium, then the premium will continue to exist. “If there are ‘limits to arbitrage,’ then what you see in computer simulations is not achievable net of costs,” said Lee.
Lee said there is a lot of discussion in the markets about what factors investors should be pursuing — whether value, growth, or size. She suggests that investors should be thinking about long-term factors compared with short-term ones. A high-turnover strategy for Dimensional is one with a turnover of 25 percent.
That means when a stock falls into the value category, it will earn the value premium for four years. But the information contained in the momentum premium signal loses its value quickly, requiring turnover of 100 to 200 percent each year.
Although momentum-style funds aren’t working, Dimensional says information in the momentum premium can be used when buying and selling stocks in general. Dimensional, for example, uses momentum when it’s already trading stocks for its equity portfolios. It’s one more characteristic the firm will use when making trading decisions, Lee said.