Excerpted from Bogle on Mutual Funds by John C. Bogle, pages 83-85

Since there are variations in investment characteristics even among funds with the same investment objective, closer evaluation is required if you prefer to fine-tune the analysis. For example, if the value fund in which you are interested has a lower ExMark, a lower Beta, and a higher gross yield than the average for its group, fair comparison requires that you select a subset (or peer group) from among funds with investment characteristics that are more similar. Two examples may suffice to make this point, as shown in Table 4-6.

Let's consider this table in two segments. The selected value fund appears to have provided a subpar return (+13.4% versus +13.6%). But the fund had less of its return explained by the market, took substantially less risk, and earned a higher income yield than its objective group. When compared to a peer group with comparable risk characteristics, the fund was actually a superior performer (+13.4% versus +12.7%). The selected growth fund, on the other hand, is clearly a winner when compared to its broad objective group. But, compared to a peer group of growth funds that also have been more diversified, assumed higher risk and earned somewhat lower yields, it is actually a slightly below-average performer.

These portfolio statistics, however scientific they may appear, should not be considered precise evaluators of differences among funds. Henry Clay was right when he warned statistics are no substitute for judgment. But these three factors, taken together, substantially narrow the parameters in which to evaluate a fund's relative performance. Thus, they provide useful information.

To perform the sort of analysis outlined in this section, you will need to gather all of the data I have discussed. Facts on fund asset size, age, manager tenure, and cost are readily available through prospectuses and major financial publications. While information on portfolio characteristics and portfolio statistics is more difficult to obtain, several statistical services provide it. Chapter 8 offers a full discussion of the sources and uses of mutual fund statistics and information.

In the final analysis, I cannot emphasize enough the importance of fairness (and common sense) in assuring that you compare only funds that have similar investment policies and characteristics. Sensible performance comparisons can be made only after establishing that fairness. Even then, as I shall now show, selecting equity funds on the basis of past performance is likely to be a futile effort on your part and virtually certain to be a futile effort for fund investors in the aggregate.