The momentous challenge
A successful momentum strategy requires close attention to trading costs and timing and the impact it has on returns
- Momentum has posted strong performance in recent years, raising concerns about whether it can continue to add return after implementation costs
- Practitioners say implementation costs are lower than academic studies suggest
- Overall, momentum remains a proven diversifier relative to value and other factors
It is no exaggeration to say that the momentum factor has chalked up heady performance numbers in the past few years.
So much so, in fact, that some in the investment industry have questioned whether the momentum premium remains a viable source of return for institutional portfolios.
Critiques generally focus on the cost of capturing the performance of equities that are delivering above-market returns – a universe that, by nature, is constantly in flux, requiring investors to sell some stocks to make room for those that top the momentum charts.
But momentum is about more than just chasing returns – the factor adds significant diversification benefits to a portfolio, even in the absence of positive performance of its own. Several recent research papers and in-depth interviews with leading investment managers suggest that the best way to capture momentum returns is to hit the brakes.
That is not to say the momentum factor is being thrown overboard. Instead, managers are refining their approaches to harvesting momentum, using more precise evaluations of total trading costs to reduce the cost of implementation, considering the impact of momentum in relation to other factors in a multi-factor portfolio. They are paying greater heed to long-term indicators of the potential return that can be expected from momentum at different points in the economic, credit and business cycles.
These changes in thinking are taking place as part of a broader shift in the conceptual and analytical framework used to evaluate factors and their respective roles in an institutional portfolio (see ‘The last free lunch’ in this report). The latest thinking is that factors are best used in multi-factor portfolios with weightings that change little over time. Some are looking towards new approaches to harvest factor premia from multiple asset classes, adding currencies and fixed-income markets to the current predominance of equities.
Concern about the future of the momentum factor prompted Research Affiliates to title a recent paper ‘Can Momentum Investing be Saved?’ Noting that the title may seem “unduly provocative”, Research Affiliates head Rob Arnott and his co-authors outline how momentum – while one of the most compelling factors on paper – delivers “surprisingly weak” results based on a study of US mutual funds pursuing momentum strategies.
According to Research Affiliates, no US-benchmarked mutual fund with ‘momentum’ in its name has cumulatively outperformed its benchmark since inception, net of fees and expenses. Worse, because the standard momentum factor gave up so much ground in the last momentum crash of 2008–09, suffering a 54% decline, “it remains underwater in the United States, not only compared to its 2007 peak, but even relative to its 1999 performance peak”. Despite noting, in fairness, that all factors may experience long periods of weak performance, Research Affiliates concludes that “this means 18 years with no alpha, before subtracting trading costs and fees”.
Still, says Research Affiliates, momentum “is a popular and seductive strategy” that is associated with superior performance on paper over long periods in all the main markets except Japan. But with momentum funds failing to beat the market, on average, the strategy “needs saving”.
The solution, Research Affiliates contends, is three-fold. The first step is recognising that the price impact of trading a momentum portfolio makes standard momentum – the trailing 12-month return, excluding the most recent month – a “very expensive strategy…. unless very careful (and clever) attention is paid to implementation”. Better to distinguish between fresh and stale momentum, Research Affiliates contends, and trim stocks that have exhibited strong momentum for two or more years – such companies are typically expensive, and the strength of performance tends to fade – and buy stocks that have recently experienced wonderful news.
Finally, a strong sell discipline is essential: the optimal way to participate in fresh momentum is not to shorten the average holding period of the portfolio, which would lead to “prohibitively high” transaction costs, but to avoid holding on to stale stocks that have been on strong momentum trajectory for two or more years.
Others suggest that the actual costs managers incur trading a momentum strategy are much lower than the generic costs assumed in academic studies, rendering the strategy cost-efficient on its own. And they point out that considerations other than trading costs are important in evaluating the contribution of the momentum factor to an institutional investor’s portfolio.
Based on a study of about $1trn (€800bn) of its own trades in momentum and other strategies over a 20-year period, AQR Capital Management found that actual transaction costs “are an order of magnitude lower than most other studies assumed”, says AQR principal Ronen Israel. “A momentum strategy is a good long-term source of return net of transaction costs,” he says. “It’s not just a theoretical factor return.”
The key is to trade smart. As part of its study, summarised in the research paper ‘Implementing Momentum: What Have We Learned?’, AQR devised a transaction-cost model that estimates the market impact of trading on a security’s price, which is the most significant component of transaction cost, for a given set of characteristics and a given amount of trading as a percentage of daily trading volume and so forth, Israel says.
“It’s true that turnover is higher than value or other factors,” Israel says, “but trading costs are low enough that you can capture the returns of momentum.”
In practice, AQR also reduces implementation costs by evaluating the cost-benefit tradeoff of potential changes to momentum portfolios using a proprietary optimiser that provides a systematic method to slow down the turnover.
While some practitioners take a purist approach and assume that all trades generated by a theoretical momentum portfolio must be executed at a high dollar cost, “in the real world, both of those things don’t have to be true”, Israel says, adding: “We handle this endogenously by explicitly weighing the benefit of changing the portfolio relative to the costs of changing the portfolio, to come up with an optimal point for the most exposure a portfolio wants to momentum but not at the highest cost.” This generates “savings from both the market impact cost per dollar traded being lower as well as the amount of dollars traded being lower, without giving up much of the gross return of the strategy”.
Perhaps most important, Israel says, is that momentum is typically used in a multi-factor context. “You might like a stock because of its momentum, but you may dislike it because of its value,” he says. “That will naturally slow down the turnover.”
Eliminating momentum could also disrupt a portfolio’s diversification. In a simple analysis in the paper ‘Fact, Fiction and Momentum Investing’, AQR determined the optimal weight of momentum in a four-factor portfolio would be about 38%. Even assuming the extreme case of zero return for momentum, the optimal portfolio still places a positive weight on the factor, indicating that the diversification benefit is so great that holding momentum for a zero return would be additive (see figure). “While we believe in momentum as a stand-alone factor,” Israel says, “we’ve always advocated combining it into a broader portfolio, particularly with value.”
Investment managers are launching strategies that reflect the new thinking about momentum. While momentum is a valid factor, returns can be volatile, says Mustafa Sagun, chief investment officer at Principal Global Investors. “You climb the stairs, but you take the elevator down,” he says.
Principal recently launched a “sustainable momentum” exchange-traded fund (ETF), that seeks to smooth volatility and reduce trading costs by monitoring the valuation of momentum stocks – which tend to decline sharply after becoming expensive relative to the rest of the market – and by slowing portfolio rebalancing to every six months instead of monthly.
Principal will also employ three and five-year momentum-trend indicators designed to detect important turning points in momentum stocks; the long-term indicators would get more weight as the momentum factor became more expensive. “Momentum works, but it’s like picking up pennies in front of a bulldozer,” Sagun says. “You want to know when the bulldozer’s coming.”