In this previous article, it was demonstrated that an almost risk-free investment is to buy 2-year Treasury bonds when the FRR2-10 is close to 1.0 and to sell when the Federal Funds Rate (FRR) is at its lowest after recessions. A reader critiqued the article and stated: "[Y]ou don’t know in advance when the Fed is done easing. This backtest is a little like saying you will make a profit by selling when yields reach their lows for the cycle - an ex post truism.”
In response to that critique, we show that it is possible to design a mathematical model using a smoothed FRR2-10 to signal investment in IEF and a moving average of the FFR for sell signals.
- The FRR2-10 is calculated using as input smoothed 2-year and 10-year U.S. Treasury yield rates. When the FRR2-10 is less or equal to 1.02 the model buys (IEF), otherwise (SPY).
- The model sells IEF when the 10-day moving average of FFR divided by the 50-day moving average of FFR is greater than 1.05, otherwise SPY, provided a position has been held for a minimum of 90 days.
Performance Backtest from Jan-2000
A backtest from Jan-2000 to Sept-2019 was performed on the online simulation platform Portfolio 123 with transaction costs and slippage taken into account and compared with the returns of a buy-and-hold investment in SPY.
Performance over almost 20 years is shown in Figure-1. For the model the Total Return and Annualized Return would have been 1,100% and 13.5%, respectively, with a Maximum Drawdown of -22%.
Over the same period, the benchmark, SPY, had a Total Return and Annualized Return of 192% and 5.6%, respectively, with a Maximum Drawdown of -55%.
Since Jan-2000, there were only six realized trades, and the current holding since 4/1/2019 is IEF, all of them winners.
From the analysis, it appears that fairly satisfactory returns can be had by investing in 10-yr Treasury bond funds as indicated by the FRR2-10, and switching back to equity according to signals from the Federal Funds Rate.
The model can be followed live at imarketsignals.com.
Calendar Year Performance
The model would have had only positive calendar year returns, and significantly under-performed the benchmark in 2003 and 2006. (The indicated under-performance in the years 2004 to 2005, and 2010 to 2017, is a result in differences how dividends are reinvested in the benchmark and the model.)
|IEF||4/1/2019||Holding||5.20% (as of 9/13/2019)|
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.