Inflation is on the rise and bond investors fear further inflation increases. Active duration management is required to protect fixed income portfolios against the adverse impact of higher inflation. In this note we extend the backtest of the duration model to study its performance in times of rising inflation.
The duration model has predicted bond returns successfully during such periods. Robeco’s quant duration strategy is hence well-suited to navigate bond portfolios through periods with high or rising inflation.
Robeco has applied its quantitative duration model in practice since the nineties. The backtest of the model started in the eighties. Both the live period and the backtest period have been periods with generally declining inflation and falling bond yields. As inflation has started to rise and expectations are for a further rise, we wanted to study the model during times of high and rising inflation. To do so we had to extend the backtest.
US interest rates have been trending lower for nearly 30 years. Commonly cited reasons have been the disinflationary influence of cheap labor and goods from globalization and reduced inflation risk as the Fed gained credibility as an inflation fighter. As a consequence bond returns have been great during this period, providing both coupons and capital gains. Yet going forward it seems that the aforementioned reasons might change direction to cause higher inflation. Emerging markets are now an inflationary source of demand for consumer goods and commodities; and the Fed has provided unprecedented amounts of liquidity to stimulate growth and employment.
For these reasons we decided to investigate the US bond market prior to the September 1981 peak in interest rates. During the seventies and early eighties inflation reached very high levels. This resulted in negative bond returns relative to cash. Hence we had a bear market rather than a bull market. More specifically we find government bonds to perform poorly relatively to cash both in periods of high and of rising inflation.
To extend the backtest we have created a reasonable proxy for the model for the US that goes back to 1951. We find that the duration model performs well during the bear market prior to 1981. And we also show that the model generally does well both in high inflation periods and in periods of rising inflation. The duration model has proven its value in practice in a period of generally declining inflation, but our research shows that the quant duration strategy can also protect fixed income portfolios in times of high or rising inflation.