Interest Rate Forecasting and Your Bond Portfolio
by Cathy Pareto, MBA, CFP®
January 2010
The basic premise of active investment strategies is that the investor is willing to make bets on the future rather than settle with what a “passive strategy” might offer. With regard to bond portfolios, betting on the direction of interest rates is an example of this. But, do bond investment decisions based on one’s anticipation of interest rate changes really make an impact? Not really.
Betting on interest rate changes in your bond portfolio can be a futile process. And yet, investors are fascinated with the idea that forecasting will lead to better returns. In fact, the Federal Reserve has no direct control over the interest rates that bonds pay. The bond auction, not the Fed, sets the rates for Treasuries and investment banks set the yield for corporate issues when they are underwritten. Rather than worrying about the direction of interest rates, wouldn’t investors would be better served to evaluate the risk and return dimensions of the bond they are buying, specifically as it pertains to maturity?
Now, let’s agree on one thing before we proceed. The whole purpose of having bonds in your portfolio is to further mitigate the risk of an already diversified mix of stocks. Unless you’re a bond trader, I don’t know of anyone who got rich on bonds. Bonds are what I’d call your “insurance policy” in a balanced portfolio—funds that you can access in a market downturn without having to sacrifice equities that may be under water. The key is to select the right balance between risk and return.
Balancing Risk and Return
The greater the risk associated with an investment, the higher the compensation should be —that’s the conventional wisdom. Of the investable assets available, arguably, equities are the riskiest while bonds are perceived to be the less risky assets. But, the truth is bonds carry risks that investors tend to overlook, including: default risk, credit risk, interest rate risk, and reinvestment risk. The longer the maturity of a bond, the more risk it carries. Remember that yields move inversely to bond prices. So, as interest rates go up, your bond value goes down and when rates go down your bond value increases.
Knowing this relationship exists, shouldn’t we use it to make interest rate plays in our bond portfolio? The answer is, “not really”. I would argue that no matter what direction interest rates are headed, an investor should never own a bond with a maturity of greater than (roughly) five years. It’s true, long maturity bonds pay investors additional yields than shorter term bonds. However, that yield comes at a much greater risk than you might think.
The best way to determine whether that additional risk is worth it is to evaluate the “return per unit of risk” of the investment as it compares to investment alternatives. This ratio attempts to show the relationship between the investment’s historical return and the risk associated with the generation of that return. A return per unit of risk of 1.00 means that for every percentage point of volatility there has been one percentage point of return. The higher the ratio, the better the risk adjusted performance.
Let’s take the following period as a case study. In its effort to combat inflation, the Fed increased the fed funds rate seventeen consecutive times from June 2004 to June 2006. While the Fed does not directly impact bond interest rates, when the federal funds rate goes up, bond coupons tend to rise as well in anticipation of higher inflation. So, let’s conduct a little study to compare the results of: one month Tbills, one year Treasuries, five year Notes, Long Term Bonds and the S&P 500 over the same period the Fed was adjusting rates upward as well as a longer term period from 1975 to 2008.
We find that the best risk adjusted bond investment for this period was the five year Treasury notes. Although the long term bonds generated a higher return of 5.77%, the additional risk taken for that yield might make that a less attractive investment. The five year notes generated the highest return per unit of risk for the period with 2.50.
Now let’s see if the same holds true for our broader time period:
Source: Dimensional Fund Advisors
This time period captures several years of inflation, increasing rates and decreasing rates. Here we find that the annualized returns for the five year and long term bonds are not that far off, while the level of risk (standard deviation) is substantially higher for the long bonds. In fact, the level of risk for the long bonds is only ~6% lower that the risk of the S&P 500. It’s clear to see that the highest returns per unit of risk were delivered by the Tbills, one year Treasury and five year Notes.
Instruments with maturities longer than five years have not offered returns commensurate with the greater volatility. Some investors may have a preference for bonds with longer maturities to match assets with liabilities. Investors seeking to minimize volatility should consider shrinking away from long-term bonds.
We can conclude from this data that it really does not pay to extend bond maturities beyond five years. And if we hold this to be true, then attempting to forecast interest rates as part of your bond strategy may not be very effective. The Fed has a deeply important role in our global economy, there’s no doubt about that. But it should have no role in determining your most appropriate bond portfolio. In summary, the role of fixed income in a balanced portfolio is to dampen the volatility of equities. An allocation to instruments with short-term maturity and high credit quality may help reduce risk, no matter what the direction of interest rates might be.