At the heart of sector rotation strategies lay trend-following relative strength signals, which select the best-performing sectors over a specified period of time. The concept of trend-following as an investment strategy can be traced back to practitioners from more than a century ago. Since the 1990s, there has been increased attention from academics and practitioners, resulting in a wealth of empirical work being published on the subject. The effectiveness of the strategy in consistently generating positive excess returns over relevant market benchmarks has been well documented for many asset classes, including commodities, equities, and foreign exchange strategies. More recently, these types of factor strategies have been proposed in fixed income and empirical work focusing on particular asset types has appeared, such as the cross-sectional ranking of 10-year government bonds.
Broadly, trend-following strategies generally take on two forms. “Absolute momentum” is based on the price trend of a single asset, typically in the form of holding long the asset when price appreciation is signaled and holding short otherwise. “Relative momentum” or “relative strength” is based on a cross-sectional analysis of multiple asset price series, in which each asset is scored and ranked in terms of expected performance. One simple example of the former is to use 12-month price returns as the trading signal; on each rebalance day, if the asset shows a positive 12-month trailing return, the investor would buy the asset, otherwise, they would short sell the asset (or hold Libor).
The same signals used in the single asset strategies can be used to rank a universe of assets in relation to one another. Assets with the highest historical returns, for example, are preferred over those with the lowest returns. One example of a relative strength fixed income strategy is to select from 10-year government bonds from the G10 countries, and hold a long position in the three 10-year government bonds with the highest 12-month returns and a short position in the three 10-year government bonds with the lowest 12-month returns.
In this paper, we examine a stylized sector rotation framework using a relative strength signal, which selects from industry sectors of the S&P 500 ® Bond Index. We specify a simple prototype model that produces a score for each of the 10 industry sectors of the index based on historical price changes. The score is used to rank the sectors in relative terms of expected performance over the following month. The portfolio takes long positions in the top three ranking sectors. We compare the performance of this hypothetical portfolio with the headline S&P 500 Bond Index, which consists of long positions in all 10 industry sectors.