What a difference seven months can make! In April 2022, Australia’s cash rate was 0.1%, unchanged for 18 months. Starting in May, seven consecutive monthly rate hikes by The Reserve Bank of Australia (RBA) to combat inflation, brought the cash rate to its current level of 2.85% by November. In the United States, The Federal Reserve, on a similar mission to curb spiralling prices, raised the discount rate from 0.25% in April to 4.00% by November. There truly has been a regime change for interest rates.
Rising rates will likely lower existing bond prices as investors favour purchasing newly issued bonds that pay higher interest rates rather than older bonds paying a lower fixed rate. This negative pricing pressure is particularly observed in older, long-term bonds which lock-in investor funds for up to thirty years at lower interest rates, compared to newly issued short term bonds.
Adjusting SCTR and INDU algorithms
Unhedged’s Sector Rotation and Industrial Activity algorithms were also impacted by changes in interest rates given their calculated exposure to long term bond exchange traded funds (ETFs) during market downturns. Contrary to our namesake, we hedged potential losses from holding long term debt securities in both algorithms by also including assets that benefited from Central Banks raising interest rates. Concurrently, our quant team explored whether there was a better way to handle market downturns coupled with rising interest rates.
Our solution involves looking at the shape of the US Treasury Yield Curve, a widely accepted leading indicator of the economy. When yields on thirty year Treasury bonds are higher than yields on one year Treasury bonds, the yield curve has a steep slope. Steep yield curves are viewed as net-positive for the economy as investors can borrow at lower short term rates, lend at higher long term rates and facilitate expansionary growth. By correctly forecasting continued steep yield curves, we increase our confidence in holding long term bonds.
Conversely, periods where the yield curve is inverted – where the thirty year Treasury yield is lower than the one year yield, imply that investors anticipate long term economic uncertainty, and would prefer to invest their funds in short term bond market instruments. By correctly forecasting continued inversion in yield curves, we increase our confidence in exiting long term bonds and holding rising rates instruments, instead.
Our statistical forecasts rely on well researched econometric models often applied to financial time series, that use lagged values of yield curves and lagged deviations of the yield curve from its average yields to predict week-ahead and month-ahead yield curve shapes.
Suppose our one-month ahead yield curve forecast indicates an upward sloping (steep) yield curve for the upcoming month. Recall that a steep or upward sloping yield curve occurs where long term interest rates are higher than short term interest rates. These conditions facilitate expansionary economic growth as investors anticipate holding longer term debt instruments increases their returns. Our algorithms would respond to this forecast and the anticipated increase in demand of long term debt instruments by exiting any positions in rising rates instruments and purchasing long term bond ETFs.
Conversely, if our one-month ahead forecasts indicate a downward sloping (inverted) yield curve, our algorithms will pivot to selling long term bond ETFs and buying rising rates instruments. Remember, an inverted yield curve tells us that long term interest rates are lower than short term interest rates, suggesting continued investor uncertainty about long term economic prospects. In such a scenario, we expect an increase in demand for rising rates instruments that benefit from short term interest rate increases, while expecting a decrease in demand for long term bonds or fixed interest securities which tie up investor income for longer durations.
Alongside our month-ahead forecasts, run at the start of each month, we also run one-week ahead forecasts of the yield curve, at the start of each week. This is to ensure our weekly and monthly forecast directions are consistent and to warn us in case of upcoming changes in the yield curve shape. If our month ahead and week ahead forecasts both predict upward or downward sloping yield curves, then there’s no problem. If they don’t match, this could signal a change in interest rate regimes, which we would expect our weekly signal to pick up first.
A final caveat here: While we are using historical values of the yield curve to forecast trends in its shape, we are not timing the bond market or when a yield curve will “switch” from a high interest rate to a low interest rate regime. Our forecasts rely on lagged yield curve values and hence we may be slightly delayed in picking up a switch from a rising rates to a flat or falling rates regime. Nonetheless we use the forecasts of yield curve shape to determine whether to continue holding long term bonds or rising rates instruments.
While the quantified impact of these changes on our portfolios will be measured over the next few months of trading, in ten-year backtests (that incorporated several periods of rising and falling interest rates) we observed improved risk-adjusted returns relative to our benchmarks and outperformance relative to our Sector Rotation and Industrial Activity without yield curve forecasts incorporated.
Summary
Interest rate environments are changing. Our Algorithms are no different. Developed in a period of historically low interest rates, we’ve now updated them to account for rising and falling interest rate environments using yield curve driven forecasts to timely allocate between long- and short-term bond ETFs.
All investments carry risks and you may lose your money. Past performance is not indicative of future performance. The information in this report is general information only and does not take into account your personal circumstances, financial situation or needs.
