07 Aug Sector Rotation vs. the Momentum Factor
Sector Rotation is one of the most popular investment strategies. In this strategy, portfolio capital is re-allocated to different sectors based on various indicators. These indicators can include business cycle indicators, market indicators, trends, momentum and more. How re-allocation occurs depends largely on the investor or manager, but generally involves investing in groups of companies that either are doing well or are expected to do well based on the chosen indicators.
One popular sector rotation system is based on relative strength, and works by investing in sectors that have been performing well in the recent past, usually not more than the previous 12 months. Inexpensive sector targeted funds and readily available returns data make this version of sector rotation particularly easy to implement and maintain.
There are questions, though, regarding this strategy and whether or not it is any different from the momentum factor strategy. Taking advantage of the momentum factor involves selecting individual stocks that have been performing well in the medium term. But while the sector rotation strategy is based on the assumption that some types of companies or industries will perform better under certain market conditions than others, the momentum factor is based strictly on performance.
To determine if these two strategies are the same, one company, Newfound Research did a small study comparing them. They started with a sector strategy that defined 49 different quite narrow sectors, buying the best performing ones and shorting the worst and compared that to a momentum factor strategy using data from 1927-2017. While they found that their sector portfolio outperformed their momentum portfolio, it did so due to one outlier period between 1943 and 1944. Starting their analysis after that period, they discovered that while the momentum portfolio slightly outperformed the sector portfolio on an absolute basis, risk adjusted returns were about the same.
Because working with 49 sectors probably isn’t realistic for most investors, they also extended their analysis to 30 and 17 sector groups, focusing on date from 1944-2017. These groups underperformed compared to the momentum portfolio, with returns of 6.03% for the 30 sector group and only 3.86% for the 17 sector group compared to a return of 6.95% for the momentum system. Furthermore, while the lower return of the 30 sector system was accompanied by lower risk, that wasn’t the case for the 17 sector system. The graph below compares the 30 and 17 sector group performance to that of the momentum strategy.
30 and 17 Sector Rotation, Long/Short vs UMD Factor
Additional analysis helps to answer the question posed at the beginning of this post: are sector rotation and momentum really the same strategy? What they found was that while the sector rotation strategies all show statistically significant alpha against size and value factors, that significance disappears when momentum is introduced. And when momentum is regressed against size, value and the sector rotation system, significant alpha remains. What this means is that while this sector rotation system does have positive alpha, that alpha is really just a proxy for momentum.
A finally analysis was done on a 10 sector system over the full 90 year period to reflect the most common ETF based strategies. Performance continued to deteriorate compared to the momentum portfolio, with an annualized return of only 3.72%. Nor was volatility significantly reduced from the momentum portfolio. Annualized volatility was 14.02% compared to 17.96% for momentum. This means that the sharp reduction in returns is essentially uncompensated.
One bright spot for the sector rotation system was a possible reduction in downside risk. During the market drops in 2008 – 2009, the 10 sector system had a max drawdown of -34.62%, compared to -57.53% for the momentum factor portfolio. This implies that the sector rotation portfolio could be more stable during market crashes. However, since this data is based on one short period, it is possible that its stability compared to the momentum portfolio was just a fluke. It does, though, bring up a potential problem with a pure momentum strategy. In 2008, momentum was heavily exposed to financials, energy and materials, leading to momentum based portfolios that lacked diversification. Those sectors ended up being the worst performing sectors that year, dragging down the performance of that strategy. One possible solution is a multi-factor strategy, such as the Global Beta All Cap Multi-Factor Strategy, which uses valuation to help protect against overexposure.
Based on these results, it seems likely that sector rotation is just a diluted momentum system with similar risks. Though sector rotation seems to work better the more narrow the sectors, what’s really happening is that the system becomes closer and closer to momentum as the number of sectors increases. Once the sector division gets to the point where every company is in its own sector, what’s left is a true momentum factor system. When a relative strength sector rotation system is implemented on a more realistic basis, though, it provides limited access to the momentum factor though low returns while maintaining most of the risk.