When stock markets are not doing well, ETFs tied to long-term US treasuries (loans to the US government) are one of the next best options available in the Tines Capital algorithms. Unfortunately, the use of long-term US treasuries over the last year has been a drag on the Tines Capital algorithms. Naturally this raises the question of whether the algorithms should be modified to eliminate US treasuries or reduce their use. In short, I don’t believe they should be removed. There seems to be a lack of evidence for that change.
Rising Interest Rates
The Federal Reserve is taking up an even larger portion of the financial news than normal. This is because with high inflation they have few options other than raising interest rates and in particular the fed funds rate. Common financial wisdom states that when interest rates go up, bond prices go down. This is mostly true, but there is one very important caveat – all else being held constant! It is reasonable to think that only buying bonds during falling interest rates would yield better results. Unfortunately, the data doesn’t seem to support such a simplified trading technique.
To demonstrate the concept I have used Vanguards Long-Term Treasury mutual fund VUSTX and compared it to the federal funds rate. This fund is not currently used by any Tines Capital strategy since Tines Capital primarily uses ETFs. However, VUSTX has a long history going back to the 1980s making it ideal for historical comparisons.
- VUSTX (Rising Rates) tracks the hypothetical performance of buying VUSTX as soon as rates hike and selling VUSTX as soon rates drop, while reinvesting all dividends.
- VUSTX (Falling Rates) tracks the hypothetical performance of buy VUSTX as soon as rates fall and selling VUSTX as soon as rates hike, while reinvesting all dividends.
- VUSTX (Buy and Hold) tracks the hypothetical performance of buying VUSTX and holding it, while reinvesting all dividends.
- Fed Funds Rate tracks the historical federal funds rate as reported by tradingview.com
According to this simple test, changes in the federal funds rate alone do not appear to have significant predictive power for the future performance of government treasuries. This could change in the future, but it seems unlikely. Without clear evidence suggesting otherwise, modifying any investment portfolio due to changes in fed policy alone seems like a poor recipe. Though US Treasuries have been a major drag in the Tines Capital algorithms they will remain a part of the algorithms until clear evidence suggests their removal.