Not all bond funds are created equal because not all bond fund managers are created equal. One of the primary differences between bond fund managers is their willingness to tap into lower-rated issues. In previous years, such aggressiveness tended to boost performance. More recently, however, this appetite for less-than-AAA issues has been a recipe for disaster.

A Cautionary Tale

Consider the wreck of Regions MK Select Intermediate Bond Fund (RIBIX). Among its peer group, it was a top 10 performer in 2004, 2005 and 2006. Then in 2007, this fund lost 50.1% and was dead last among all 288 U.S. intermediate bond funds. To put this meltdown into perspective, consider that the average intermediate bond fund had a return of 5.2% in 2007.

How does a bond fund lose more than 50% in one year? The answer is straightforward: risky ingredients (specifically, heavy exposure to subprime mortgages). Ironically, exposure to lower-rated bonds provided a nice tail wind for some bond funds in 2004, 2005 and 2006. But alas, a tail wind (or excess return in a bond fund) may be a sign of an impending storm. The modestly higher returns of RIBIX from 2004 to 2006 were more than wiped away in 2007. In fact, by March 31, 2008, RIBIX had lost nearly 72% over the prior 12 months. So, risking a dollar to win a dime didn't pay off.

Is the performance of RIBIX an anomaly? Yes, but only in its extremeness. The 15 months from Jan. 1, 2007, to March 31, 2008 revealed which bond fund managers were playing with matches and which ones were not.

Consider the following: According to data from Morningstar Principia, there were 288 intermediate U.S. bond funds with a full five-year performance history by year-end 2007. This count includes only distinct funds; only one share class was included for funds with multiple share classes. In 2003, 2004 and 2006, all 288 funds had a positive return. In 2005, three funds had a negative return, although the worst return was only a loss of only 1.1%. Then, in 2007, seven of these 288 intermediate bond funds had a negative return. Sign that a storm was coming? Indeed. From April 1, 2007, to March 31, 2008, nearly 10% of this group of bond portfolios (27 funds) had a negative return. And the storm may be far from over.

Let's examine the contents of the bond funds that sailed around the subprime storm as well as those funds that drowned in it. The top performing quartile of bond funds during the 12-month period ending March 31, 2008, minimized its exposure to bonds with a BBB rating or lower (see "Subprime Exposure"). The average return of these 72 funds was 8.57%, and not one of them had a negative return during this 12-month period. Interestingly, the 72 funds in the bottom quartile outperformed the top quartile funds during 2003 to 2006 for precisely the same reason they dramatically underperformed in calendar year 2007 and in the 12 months ending March 31, 2008—exposure to riskier bonds (see "Boom and Bust").

The recent losses in a significant number of bond funds are a vivid reminder that investors need to understand the source of performance in a portfolio—whether bond or stock—in order to anticipate the potential risks. In the case of bond funds, a portfolio with greater exposure to lower-rated bonds presents greater risk. This greater risk is often masked, for a season or two, by higher returns. It will eventually manifest itself, and when it does, it may exact a higher price than the investor anticipated. Making matters worse, poor performance may initiate large-scale redemptions from an underperforming fund, which exacerbates the fund's meltdown. This scenario has certainly played out for several bond funds recently.

Recency Bias

Going forward, bond investors will likely pay more attention to the quality of the holdings in the bond funds they select, at least for a while. There is something odd at work here, and it relates to the human tendency to downplay risks not recently observed and dramatically react to events most recently experienced.

For instance, if we drive by a terrible car accident, we tend to reduce our speed for a while. But within a short period of time, we're once again attempting to hit warp speed on I-80. It's a form of recency bias—a well-documented perceptual tendency in which we tend to overweight recent events and downplay events not recently observed.

This behavior suggests that the memory of investors is rather short, and that within several years subprime notes (or whatever the financial steroid is at that time) will once again flow like IPOs during the 1990s, producing another educational experience in which investors relearn the fundamental correlation between risk and return. However, those who choose to remember the bond fund experience of 2007 will have learned—once and for all—that bond funds do carry differential credit risks based on the credit quality of their holdings and that it pays to know what is under the hood of your bond fund.