Don asks an excellent question:

*The basic principle of your book, as I understand it, is to invest in uncorrelated securities and periodically rebalance them. We can’t know which investments will go up and which will go down. But it really doesn’t matter as long as, over the long run, the investments make uncorrelated “up and down” fluctuations. Rebalancing will force you to sell high and buy low.*

*I believe in this principle with one reservation. If a particular class of investment is already “pegged” at one extreme of its viable range (that is, it is up against some fixed limit), then that investment can only move in one direction. For an example, consider long-term bonds, which are sensitive to interest rates. The Federal Funds Rate is approximately zero, which means it can only go higher. Therefore, for now, long-term bond prices can only go down, not up. In order for bonds to fluctuate “up and down” to generate rebalancing opportunities, interest rates need to be higher than zero, to give some room on the “up” side.*

*It seems to me that long-term bonds can play an important role in a rebalancing strategy when the Federal Funds Rate is in a “normal” range–at least, greater than zero. This gives bond prices room to move in both directions. But for now, there’s only one way they can go. Therefore a wise investor would not invest in bonds until the Funds Rate rises a little.*

*I expect that you will disagree with this, and I would be interested to learn why.*

* -Don*

In my book, I discuss the dangers of oversimplifying (chapter 15—spectrum #1: simple vs. complex). In this question, there are assumptions about interest rates, bond prices, and the overall makeup of the bond market that have been oversimplified and need to be examined more closely.

The Federal Reserve controls the federal funds rate, which is a short-term rate and it has a large influence over short-term Treasury rates. The Fed has much less influence over long-term rates. The Fed has pegged short-term rates to zero, but long-term rates are significantly higher. As I write this, a 30-year Treasury bond yields 3.34%. Although this is near the lowest rate it has been for a generation, it is still not zero.

If long-term rates were at zero, I would agree with Don that they would not make a very good investment. But 3.34% leaves a fair amount of room before reaching the extreme end of its viable range. If you look at the situation in Japan, you will note that long-term interest rates have been lower than 3.34% for about 20 years. Long-term interest rates can go lower and they can stay low for a very long time.

If the Fed raises the federal funds rate, then do long-term rates rise in sync? Many investors make this oversimplifying assumption. An equal rise in interest rates across the spectrum of maturities will mean that the longest-term rates decline the most. But one does not have to look back very far into financial history to figure out that the assumption of an across-the-maturity-spectrum rise of equal magnitude is not typical. When the Fed last increased federal funds rates from mid-2004 through mid-2006, long-term rates stayed relatively constant. The federal funds rate went from 1% to 5.25%. Ten-year Treasury bonds yielded between 4 and 5.1% for that whole period. Short-term interest rates rose steadily for two years, while long-term rates essentially went sideways.

Once you start trying to play the game of waiting for a rise in interest rates, you have to make a series of other decisions: how much of a rise will it take before investing? How much should I invest at any one time? What do I do if interest rates keep rising? What do I do if they fall back? Professional investors who play this “guess where interest rates will be in a month/quarter/year” game have amply demonstrated that they are not able to consistently win. If the professionals cannot figure out how to win this game, then what are the chances that an investor with fewer resources will do better?

Another dangerous simplification is to assume that all bonds are alike. Corporate, municipal and mortgage bonds all behave differently from Treasuries, although they are strongly influenced by the general level of interest rates. Long-term corporate bond ETFs, such as LQD or VCLT, currently yield between 4 and 5%, which is higher than Treasuries and even further from the zero interest rate peg of the Fed funds rate. They carry the additional risk of a corporate default and this is why investors demand a higher interest rate as compensation.

The justification for waiting is to avoid interest rate risk. But there is another way to avoid interest rate risk—by buying individual bonds instead of buying a bond fund. If you hold a bond to maturity, the course of interest rates between the time you purchase the bond and when it matures does not matter. You will receive your entire principal upon maturity. A bond fund works differently, though. The bond fund manager is usually forced to sell bonds at a loss before maturity if interest rates rise, thus locking in a capital loss.

If an investor decides that to have a sizable allocation to Treasury bonds, buying individual bonds is generally a lower cost way to go. Treasuries can be purchased for no commission and can be held without incurring any additional fees. On the other hand, if an investor buys them via a fund, he or she is paying recurring fees for the management of the portfolio. So the fund buyer adds interest rate risk and incurs higher fees. Funds may be more convenient than buying individual bonds, but for many investors, the convenience may not be worth the cost.

The investor who waits for the Fed to raise interest rates still needs to invest the targeted money somewhere. If this money is invested in cash, then the returns are virtually zero, even before taking inflation into consideration. The opportunity cost of sitting on cash compared to a long-term bond is between 3 and 5% per year. If interest rates don’t rise for several years, this opportunity cost begins to add up.

So I would rather not pay this opportunity cost and wait an indeterminate amount of time for the Fed to act. I will also not assume that long-term rates will track short-term rates in lockstep when rates do rise. Instead, I will monitor the fluctuations in price and by rebalancing, I will add to my long-term bond holdings when prices are relatively low and interest rates are higher, and I will subtract from the holdings when prices are relatively high, and interest rates are low. In short, I will successfully time the bond market without making any predictions about the future course of interest rates. The timing won’t be perfect and it will not comprise the entirety of my position. But in the long run it will likely earn more than alternative approaches based on guesses about future interest rates.