Swedroe: Is ‘Momentum’ Faltering? Part I

Factors, like momentum, don’t always work the way investors expect them to.

This is the first article in a two-part series about the momentum factor and whether or not its premium has permanently disappeared. The second article in this series will appear later this week.

For the decade from 2004-2013, the momentum premium—as measured by the Fama-French momentum factor—experienced a negative compound return of -1.2 percent per year. (This number was calculated using the monthly momentum premium figures from the Fama-French data series. Note that returns on factors are generally expressed as annual averages, not annualized returns, and during this decade the annual average premium was 1.8 percentage points.)

This decade-long period of poor performance has led many investors to ask whether the momentum premium has now permanently disappeared, because it’s become so well known and so many are now trying to exploit it.

While we cannot know the answer to the question of whether momentum has indeed disappeared for good (only time will provide the answer), we can look at the historical evidence to see if there have been other decade-long episodes of negative premiums. That might provide insight into whether the most recent period was truly unusual.

Before looking at the data, it’s important to note that one of the biggest problems investors face is adhering to a strategy when that strategy experiences long periods of underperformance. Discipline is a big problem, because there is a nonzero chance that the returns to any asset class or factor—be it beta, size, value or momentum—will be negative, no matter how long the horizon.

The following is a good example. For the 40-year period 1969-2008, long-term (20-year) Treasury bonds returned 8.9 percent per year, outperforming the CRSP Total Market Index, which returned 8.8 percent per year.

Another powerful example is that from January 1990 through June 2014, Japanese large-cap stocks lost 0.4 percent per year, producing a cumulative loss of -5.7 percent. That’s almost 25 years without any positive return.

Value stocks, international stocks and emerging market stocks have all experienced long periods of relatively poor performance, periods that tested the faith, and thus the discipline, of investors.

10-Years Of Negative Returns To Momentum

Using the Fama-French momentum factor and CRSP data, we can look at annualized returns attributable to the momentum premium over 10-year periods. Examining only calendar years, for the period 1927-2014, we find 14 different 10-year periods when momentum produced negative total returns. Even considering nonoverlapping periods, we find two such occurrences.

Yet over the full period, the momentum premium was an annual average of 9.62 percent a year, producing a compound return of 6.87 percent. In other words, the last decade wasn’t unusual at all, and it’s hard to draw any conclusion from that period.

212 Years Of Price Momentum

The authors of the 2013 study “212 Years of Price Momentum” examined the long-term evidence to determine if the most recent decade’s performance for momentum was unusual. Their study looked at the evidence all the way back to 1801.

To create the data set, the authors merged three known 19^th- and early-20^th-century data sources into one testable data set from 1800 to 1927. Those data sources are from the International Center of Finance at Yale, the Inter-University Consortium for Political and Social Research, and Global Financial Data.

“Between 1800 and 1927, the merged dataset contains an average of 272 securities per month, making it robust for security-level studies,” the study said.

When the authors extended the data back to 1801, they found that there were seven other decade-long periods of negative returns. The following is a summary of their findings:

• In the pre-1927 data, the momentum effect remains statistically significant and is about half that of the post-1927 period.
• From 1801 to 1926, the equally weighted top third of stocks sorted on price momentum outperformed the bottom third by 0.28 percent per month (t-stat 2.7) compared to 0.58 percent per month (t-stat 3.6) for the period 1927-2012.
• Linking the two periods together generates a 212-year history of momentum returns, averaging 0.4 percent per month (t-stat 5.7).

The authors concluded that the most recent decade-long underperformance of momentum is not unusual.

They wrote: “Momentum profits are highly variable over time. Nevertheless, over the long run, the trend-following strategy would have generated significant market outperformance, in a different century than the one in which it was discovered and tested. Our study adds to the evidence that momentum effect is not a product of data-mining but is highly variable overtime.”

Among their other conclusions, the authors determined that the momentum premium shifts with “regime” changes. In the first year of a new regime, momentum makes a negative contribution to returns. However, as the regime persists, momentum makes a positive contribution to returns. In other words, momentum is subject to crashes and reversals, such as we experienced in March 2009.

In 2009, the Fama-French momentum factor produced a return of -52.6 percent. As the regime persisted, the momentum factor once again turned positive, returning 4.8 percent per year from 2010-2013. The compound return during this period was 4.7 percent.

In addition, after producing a return of -17.8 percent in 2003 (the regime switched following the bear market of 2000-2002), momentum provided an annual average return of 9.9 percent from 2004 through 2008. The compound return during this period was 9.3 percent.

It seems hard to argue that momentum has disappeared when its returns have been negative in only two of the last seven calendar years, and in one of them (2012), the loss was just 1.1 percent. It’s also worth noting that these crashes in momentum only occur in a long-short portfolio. Long-only momentum portfolios don’t experience such crashes.

For example:

• The two worst months for the momentum strategy are July and August in 1932, when the strategy lost 206 percent. However, past winners gained 30 percent, while past losers gained 236 percent.
• For the three months from March through May 2009, the strategy lost almost 150 percent. However, past winners gained 6.5 percent, while past losers gained 156 percent.

The takeaway here is that the momentum factor’s performance over the last decade hasn’t been unusual. Rather, the historical evidence shows that the momentum factor has been consistent and ubiquitous. Later this week, we’ll continue our exploration of the momentum factor by discussing its out-of-sample record and the benefits of exposure to it for purposes of portfolio diversification.

Larry Swedroe is the director of research for the BAM Alliance, a community of more than 130 independent registered investment advisors throughout the country.