The yield curve tends to flatten late in the business cycle as it did in the U.S. in mid-August. This movement occurs in tandem with the central bank raising rates beyond expectations in order to curb inflation. But the yield curve will steepen when a recession hits, as the short end will reflect easing monetary policy.

This simple dynamic can serve as a basis for implementing a basic mean-reversion model to capture the portion of expected returns driven by a change in prices. We add this mean reversion to our analysis by moving into the portion of the curve where the expected return derived from mean reversion is highest. (Note, we used a fairly simple mean-reversion model. Investors would likely want to improve upon these results with their own more sophisticated models.)

Applications For Public Plans

Our analysis may be particularly relevant to public plans.
The average public plan (not only in the U.S. but across developed markets) allocates about 22% of its assets to fixed income. Within this category, about 60% of assets are allocated to the Bloomberg Barclays US Aggregate Bond Index. The balance consists of global investment grade, high yield, U.S. long Treasuries, emerging market debt and securitized mortgages (see Figure 1).

To evaluate the optimal contract used to obtain the target duration exposure, we replace the 10% U.S. long Treasury exposure within the fixed income allocation with a Libor exposure and add a swap overlay leveraged to obtain the same duration as a representative U.S. long Treasury index. Given that we aim to determine which position in the swap curve provides the maximum benefit to the portfolio, we calculate the expected Sharpe ratio to serve as a ranking criterion for different contracts across time. Each month, we select the swap tenor with the highest expected Sharpe ratio for optimal positioning along the curve. Likewise, we choose to compare the strategy with the same level of duration achieved by the benchmark Treasury portfolio so we can assess the effectiveness of each contract at reaching the same target. A separate and interesting question is the amount of duration to take to offset equity risk—a question we do not answer in this exercise.

As noted, in many states of the markets and economies, we found that the medium-tenor part of the curve was the optimal point of exposure for the dynamic model used in our research. But this can break down under some circumstances.

For instance, if we experience a structural change in the slope of the curve, either because of a repricing of inflation risk or because of an anticipated new inflation regime, the model will mean revert to the incorrect target. To derive the best benefit from this approach, the manager should use a satisfactory model of yield curve mean reversion. Different views of the target values of the yield curve will inform different optimal positioning strategies.

By using this dynamic swap switching strategy in our research, we were able to improve the total Sharpe ratio fairly materially compared with using long Treasury physical bonds (0.915 versus 0.621). Moreover, the 36-month rolling marginal Sharpe ratio over the benchmark, illustrated in Figure 2, is positive for 99% of the sample before 2012 and drops to 63% from 2012 onward. This strategy was also more effective than a naïve allocation to long Treasuries (physicals) in protecting against an equity drawdown. As illustrated in Figure 3, the maximum drawdown experienced in both the dot-com crash and the financial crisis is substantially lower in the portfolio with the swap overlay.