- We expand on our previous research on the industry momentum premium, analyzing hypothetical industry momentum indexes across different regions.
- Our simulated industry momentum indexes had historical outperformance across all regions; and, except for emerging markets, the outperformance was in-line with or higher than the simulated stock momentum indexes.
- The industry momentum indexes had greater historical variation than the stock momentum indexes in their active industry allocations, and the industry factor was a major contributor to their historical outperformance.
Academic and practitioner research has shown the historical outperformance of momentum strategies at the asset-class, country, sector and stock level. In our “Industry Momentum” paper, published earlier this year, we used empirical analysis to investigate the historical premium in the Global Industry Classification Standard (GICS®)1 industry-level momentum. In this blog post, we extend our analysis to assess the persistence of a premium across different regions and the ability of a rules-based strategy to help investors as they seek to capture this premium.
A Moment for How We Measured Momentum
For each stock in the universe, we calculate one-month-lagged six- and 12-month risk-adjusted returns and standardize using z-scoring. The stock’s momentum score (exposure) is simply the average of these two scores. For the industry momentum score, we follow a similar process and calculate the return of each GICS industry as the market-cap-weighted average of stocks within the industry. The risk-adjusted industry returns for the past six and 12 months are calculated with a one-month lag, z-scored and averaged. All stocks within each GICS industry are assigned the same industry momentum score.
Capturing the Industry Momentum Premium Across Regions
In our previous paper, we analyzed the industry momentum factor using a global universe. We showed that, while industry and stock momentum overlapped, there was independent information in both. Finally, we showed that the two can be combined to provide a deeper level of information. To assess the persistence of industry momentum across different regions, we used the framework of MSCI Momentum Indexes and simulated three indexes:
- Stock Momentum: The index selects stocks with the highest momentum scores.
- Industry Momentum: The index selects the top 15 industries based on industry momentum scores.
- Combined Momentum: The index selects stocks with the highest average of stock and industry momentum scores.2
The exhibit below shows the key metrics for these three hypothetical indexes. We built the indexes using the MSCI ACWI Index as the parent index and rebalanced semiannually. We performed similar simulations for the MSCI World, MSCI USA and MSCI Emerging Markets Indexes. The results for these regions can be viewed using the dropdown menu at the bottom of the exhibit.
All momentum indexes across the four regions outperformed their respective market-cap indexes for the period of analysis, and the industry momentum and dual momentum indexes performed better than the stock momentum indexes in all regions except emerging markets. Other characteristics such as turnover and number of constituents (by construction) were similar across the momentum indexes.
Key Metrics for the Hypothetical Factor Indexes
Data from Dec. 29, 2000, to Aug. 31, 2021. * gross returns annualized in USD; ** monthly averages; *** annualized one-way index turnover over index reviews.
Sources of Return
To better understand the sources of outperformance of each hypothetical index, we performed a return-attribution analysis using the MSCI Global Equity Model for Long-Term Investors (GEMLT). The exhibit below breaks down the return of each index by its sources. With the exception of the emerging-market indexes, the industry momentum indexes had large contributions to their returns from the industry factors. The contribution of the momentum factor was positive across all the indexes and regions, but it was larger for momentum indexes relative to industry momentum indexes, while it was in between both for combined momentum indexes.
Return Attribution for the Hypothetical Factor Indexes
Data from Dec. 29, 2000, to Aug. 31, 2021.
Variations in Industry Allocations
The larger contribution of the industry factor to the performance of the industry momentum index is intuitive and is due to larger active industry bets, as shown in the exhibit below. Each bar in the exhibit shows the average of absolute active industry weights at each month’s end, averaged over time. As we can see, overall, the industry momentum index had larger active industry exposures than the momentum index. The combined momentum exposures sat between the two.
Active Industry Exposure Across Regions
Period: Dec. 29, 2000, to Aug. 31, 2021. The average of absolute active industry weights for each index at each month end is first calculated, and then its average over the entire time period is plotted above.
Moving in the Right Direction?
We found that industry momentum, a strategy that invests in industries with the highest recent performance, outperformed the broad markets across different regions. Although the hypothetical industry momentum indexes were closely related to hypothetical stock momentum indexes, the contributing factors to their outperformance were different. In our simulations, the momentum factor was the main contributor to the performance of the stock momentum strategy, but the industry momentum strategy benefited from its industry allocation, and the industry factor drove its outperformance. Investors may be able to make more informed decisions about their momentum allocation by considering industry momentum as a stand-alone strategy or combined with stock momentum.
1GICS is the global industry classification standard jointly developed by MSCI and S&P Global Market Intelligence.
2The portfolio-construction logic used in all three indexes is to ensure that the number of stocks in the index are similar and in-line with the number of stocks in the standard MSCI Momentum Index of that region.