If you took shop class in high school, you learned the principle of "having the right tool for the job." That principle also applies to money management. Some savings/investing vehicles are suited for certain tasks but not others.

For emergency savings — money you might need at any time — bank money-market accounts (MMAs) are the best choice. Such accounts typically don't pay much interest (even the best rates currently offered by online banks are less than 1%), but the goal of an emergency fund isn't earnings. Rather, an emergency fund account is simply a temporary holding place that keeps your money safe and accessible.

But what about savings that you're setting aside for non-emergency purposes, such as a major purchase that's a couple of years (or more) away? Historically, a short-term bond fund has been a better "tool" for this type of accumulation fund — but having at least a two-year time horizon is important. Here's why. As interest rates fluctuate, short-term bond funds may suffer temporary losses. However, while such funds occasionally lose money over a span of several months, it is rare for them to lose money over a two-year period.

The table below shows the annualized performance over recent two-year periods for various short-term bond funds. The table includes the average, best, and worst performance for each based on 109 two-year holding periods.

(click table to enlarge)

How did we come up with 109 if the table only goes back to 2003? By using "rolling" periods for tracking performance. We began by looking at the results from buying on January 1, 2003, and holding for 24 months. Then we "rolled" to the next month to see what happened if the fund had been purchased on February 1 of that year and held for 24 months. Next, we moved to March 1 and did the same thing. Continuing in this way, month-by-month from 2003-2013, we computed the results for 109 24-month holding periods (in contrast to just 10 such periods if only "calendar years" were considered).

Using the rolling-periods approach gives a more accurate picture of the level of returns — and degree of volatility — that can be expected from an investment. For instance, looking at the calendar-year performance columns would seem to suggest that the worst performance for the Vanguard Short-Term Investment Grade Bond Fund occurred from 1/1/2007 through 12/31/2008 when the annualized return was 0.4%. But because investors don't buy bond funds only at the beginning of the year, that number is slightly misleading. As shown in the "Worst" column, the true worst-case two-year performance for that fund was actually an annualized loss of -0.09% (from 12/1/2006 through 11/30/2008).

Now that you understand the methodology behind the table, let's draw some conclusions. Looking at all the returns in the table, it's clear that these short-term bond funds are likely to beat the less-than-1% yields currently being provided by MMAs. Only the periods impacted by the worst of the financial crisis in 2008 failed to clear this hurdle.

However, while the table shows that short-term bond funds tend to outperform money-market accounts, we want to stress again that you can lose money in a short-term bond fund if you sell into a decline rather than holding for at least two years. In rare cases, you might even suffer a loss over a 24-month period (see the "Worst" column). Occasionally, there may be periods when you would be slightly better off with a more conservative savings account. Typically, however, short-term bond funds will outpace money market accounts over a two-year period.

You may notice that we used the Vanguard S-T Investment Grade Bond fund here rather than the Vanguard S-T Bond Index fund recommended in Upgrading. We did so because the investment-grade fund's returns have been slightly better, although it comes with slightly higher risk. In this specific case, the tradeoff seems worthwhile, though you can certainly use the S-T Bond Index fund instead if you prefer an even lower-risk option.

We included the Fidelity S-T Bond fund for illustrative purposes, because we often have readers ask if they can substitute a seemingly comparable fund at their broker of choice. As you can see, the Fidelity fund has a lower relative risk and average return, but somewhat surprisingly suffered the largest worst loss. That's not to say using a substitute fund at your broker is always a bad idea. You may decide the convenience outweighs the risk of diminished performance. That's a reasonable tradeoff—just don't assume that all short-term bond funds will behave similarly.

Keep in mind that a historical pattern can only show what the tendencies are — there's no guarantee those tendencies will hold up during any specific period. But if you're willing to endure a little more volatility in search of better returns for your accumulation savings and are committed to the recommended holding period of at least two years, short-term bond funds are certainly worth a look.