In the 1990s, Narasimhan Jegadeesh then of the University of Illinois and Sheridan Titman of the University of Texas published several pieces of research, finding something particularly unexpected with the stock market. If you purchased stocks that had risen in price and sold stocks that had fallen in price, you would have historically earned positive returns over time. Not what most would have expected, in fact, this strategy's unexpectedness may be part of the reason it apparently works. Historically by following this approach of buying winners and selling losers for your portfolio from 1991 to 2016 you would earned, on average, around 0.6% a month before trading costs, albeit with various ups and downs along the way. That return would compound to 7.4% a year, and that's actually a lower return for this strategy than in several past decades.
This factor has been referred to as "momentum" the ability of stocks at the extremes of good and bad price performance to essentially keep travelling in the same direction. Better performing stocks, on average, keep moving up in price, weaker performers keep moving down in price. As the authors pointed out, momentum is a relatively short-term effect, working best when you examine stocks that have performed well over the past 6 months or so and then hold them for no longer than a year.
Monthly Returns To Popular Factor Strategies Before Costs: 1991-2016, data: Kenneth French
But Not In January
Interestingly, momentum generally does not generally work in January. This is also when another factor, value, or buying cheap stocks, tends to work particularly well. What generally happens in January is that recently poor performing stocks can rise in price relative to the broader stock market, hence contradicting the 'trend following' aspects of momentum. In fact, whereas value delivers a lot of its return for the year just in the month of January, momentum achieves its return mostly spread more uniformly over the other months of the year. Value and momentum are similar that they are both investment strategies expected to deliver a positive return based on history and using factors, but they seem to work best at different times.
And Not For Too Long
The value of momentum diminishes over time. Though it can work for up to a year, holding a 'stale' momentum portfolio with the same stocks after a year can actually hurt performance. The same analysis that suggests that good performers continue to perform well suggests that the opposite starts to occur after a year, good performers can start to lag the market and poor performers start to rebound, on average. In a sense, this isn't too surprising, if momentum continued to work forever then certain stocks would keep rising unchecked. This need to adjust a momentum portfolio at least once a year does mean that momentum involves more turnover (more trading) than some other strategies, however this is less of a problem than you may think because momentum has a bias towards holding winners and selling loses, and so the stocks with the largest capital gains actually have a tendency to remain in the portfolio rather than be sold. More trading does have a cost, but the profitability of the strategy has historically more than offset those costs when trading is done at large and efficient scale.
Monthly Returns To A Global Momentum Portfolio Before Costs: 1990-2016, data Kenneth French
And Perhaps Less So Recently
Despite the 1% monthly returns found in the early research on momentum, more recently returns have been closer to 0.6% per month, so lower than the roughly 1% monthly return initially found in the 1993 research paper, this could be because momentum is gaining in popularity and as more people follow a strategy so its returns fall. In fact, in 2016 momentum had a negative return in the US market. This is not unprecedented, like all factors, the steady return implied by broad simple monthly average statistics doesn't account for the variation in returns over time. Momentum loses money, before implementation costs, about 1 month out of 3 in recent history, in part due to the January effect outlined above. So it's likely that 2016 was a blip for momentum, just we've seen within many other factors over decades. The very fact that factors don't earn money steadily every month like a bank account is probably what accounts for the fact that their returns have generally persisted even after their discovery. In order to potentially take advantage of momentum, and sleep well at night, you need a holding period measured in years, ideally decades. Even though using factors may be a sensible strategy for longer term investors, it sill involves risk. And of course, risk comes on both sides, momentum has enjoyed some stellar months such as February 2000, when during the height of the tech bubble momentum delivered a 17.8% return in a single month. It's important not to lose sight of the return contribution that momentum has historically bought to portfolios.
Potential Risks Of Momentum
Momentum may be one of the riskier factor strategies in terms of shorter term returns. Momentum losses money in about one third of months based on history, and can lose over 5% in a month about one month in 16. Yet, these bad outcomes don't appear to be randomly distributed over time and can be more extreme around market tops and bottoms, as you might expect with a strategy that can be loosely described as trend following, changes in trends are harmful. During the tech bubble, momentum performed very well on the way up in late 1999, but then extremely badly later in 2000 when the strategy lost more than 10% for three consecutive months.
This is as you would expect - momentum was fueling up on dot com stocks, and then still holding them in the early stages as they cratered, only to then exit those companies earlier than many unfortunate investors. Then again in 2009 as the broader market was just beginning to show positive returns, momentum delivered another 3 months of terrible returns, losing over 24% in the single month of April 2009, which was a generally good month for the broader financial markets. Basically, when the market changes focus, moving away from enthusiasm for tech stocks in 2000 or moving away from financial companies in 2008-9 then momentum will be on the wrong side of the trade, but only temporarily. It does mean that momentum can add to a portfolio's volatility at times when volatility in the markets is already high. However, as Cliff Asness and colleagues have written in a paper that debunks many of momentum's myths, it's important not to focus exclusively on the negative aspects of momentum, because doing so loses sight of its potential to deliver returns to a portfolio fairly consistently on a year-to-year view, and combining momentum with other factors, particularly value, can be a smart move. Often when momentum does particularly poorly, value can do particularly well, and again for a long-term investor it's quite possible, if not always easy to have the discipline to look through shorter term market movements, provided you don't need access to the money you have invested for shorter term expenses.
Implementing Momentum
As Ben Carlson points out in describing the S&P 500 as the "world's largest momentum strategy", you may already own a mild momentum strategy without realizing it through indexing, the S&P 500 and Nasdaq have biases to momentum in terms of how the index weights are determined. A stock that rises in price will ultimately be a greater portion of the index because as its market capitalization increases, so it represents a larger portion of the S&P 500 or Nasdaq index. Essentially, stocks that go up then have larger weightings in most broadly followed indices. Also, stocks that have fallen in price dramatically may drop out of the index entirely. So by owning an index fund, you're not owning a full on momentum focused portfolio, but there will be a momentum effect in your returns. If you pick stocks, then including recent price performance in your screening criteria, and focusing on the better recent performers over the past 6 months as well as not rushing to sell your best performers, can give your portfolio some momentum exposure. Increasingly, Exchange Traded Funds (ETFs) are being offered that support many factors including momentum, as I mentioned combining momentum and value can be a sensible strategy, and many ETF providers are now offering a combination of factor-based strategies that include momentum. As one example, The Goldman Sachs ActiveBeta US Large Cap Equity ETF combines momentum with other factors offering a low expense ratio and relative liquidity.
.
