Russel Kinnel is director of mutual fund research for Morningstar and editor of Morningstar® FundInvestor℠, a monthly newsletter.
As you may recall from my previous update
, the weighted batting average is the best measure of success for our picks. It's not as simple as comparing everything with the S&P 500, but it's a more accurate apples-to-apples measure. Basically, we compare each fund with its peer group and ask whether it outperformed its peer group during the time when it was a pick. Next, we weight the picks based on how long they were picks so that a fund that was a pick for three years counts three times as much as one that was a pick for one year. Then, we tally them all up to see how often we made a good call. I'll provide some index comparisons later in the piece, but I think this is the best measure.
For background on our picks, click here
. Premium members can see all our picks by clicking here
.Third Quarter Successes
Although the true measures of success are long term, it's nice to see that our batting average perked up to 71% in the third quarter. That was above the trend of 67% from our last update that covered the trailing five years ending June 30, 2006.
In the second quarter, we added a slew of target date funds to our picks list, and all but one outperformed their categories in the third quarter. The picks were from T. Rowe and Vanguard, and only Vanguard Target Retirement 2040 lagged its peer group (by 4 basis points) in the quarter. In addition, Fidelity Tax-Free Bond, which we added in June 2006, beat its peer group with a return of 3.46% versus 3.20% for the category average.Long-Term Batting Averages
Since funds are long-term investments, the three- and five-year figures are much better gauges than three months. The longer-term numbers are good, but not quite as good as the third-quarter figures. For the trailing three years ending in the third quarter, the batting average is 69%. For the trailing five years, it's 65%.
Our picks are intended to be funds with attractive risk-reward profiles, so another good way to look at them is to see how many land in the top quartile while they are picks and how many are in the bottom. For the trailing three years, 45% of our picks made the top quartile, and just 8% were in the bottom. For the trailing five years, 41% of our picks made the top quartile, and 9% were in the bottom.Some Index Comparisons
Investors are accustomed to seeing performance compared with the best-known indexes. Just keep in mind that these aren't the best benchmarks to use for something like our Fund Analyst Picks. For example, the S&P 500 is tilted toward large caps, so it's a poor benchmark for, say, our small-value picks. It's worth noting that the past five years have mostly favored smaller stocks over larger ones, which tends to make it easier to beat benchmarks like the S&P 500 and MSCI EAFE.
The Fund Analyst Picks from the nine domestic-stock style box categories (small value, large growth, and so on) produced an annualized five-year return of 10.50%, compared with 6.97% for the S&P 500 and 8.64% for the Wilshire 5000. The Wilshire is a better, but still imperfect, measure than the S&P 500, since it covers the whole U.S. stock market like our domestic stock picks do.
Using the five foreign-stock style box categories, our foreign picks bested the MSCI EAFE: 15.63% annualized to 14.27%.
On the bond front, our intermediate bond fund picks returned 4.78% versus 4.81% for the Lehman Brothers Aggregate Bond Index. That's a great result when you consider that funds have costs and you can't actually buy a cost-free index. If you went with one of the better low-cost Lehman index like Vanguard Total Bond Index
, you would have a 4.31% annualized return, so our picks bested a real-world index. Incidentally, Vanguard Total Bond is itself a pick.Conclusion
I'm pleased to see that our picks delivered superior returns and generally stayed out of the bottom quartile. Not all of our picks have been winners, of course, and we're constantly running studies and conducting manager visits in an effort to improve our process and end result. If you have questions about any of the above, please e-mail