Mutual Fund Research Newsletter
Copyright 2011 Tom Madell, PhD, Publisher
Feb 2011 Published: Jan 30, 2011

Which Funds Should You Select? Surprise, It's Not Morningstar's Winners

by Tom Madell

Summary: While many investors prefer to invest in specific stock funds (and fund categories) that have previously topped the performance charts, there is now considerable evidence that selecting funds from among categories that show sustained relatively poor performance is a sounder strategy. Unlike Morningstar's "star ratings" which primarily measure degree of high past performance and have been found to be flawed predictors of future performance, using Morningstar's "Buy the Unloved" strategy is a better way to add (or eliminate) funds from a portfolio. But even this approach has its pitfalls. Even better perhaps, is to pick out-of favor categories based on long-term underperformance while avoiding categories that have been in long-term cycles of excessively high performance. Investors should heed the fact that no fund category underperforms or outperforms indefinitely. A good guideline for when reversals of fortune can be expected to occur is after 5 years; by 10 years, a reversal would become quite a high probability event.


For over 25 years now, I have been analyzing how different categories of funds, including various stock, bond, foreign, and specialized sectors, perform. My goal has been to see if there might be a way to do a little better than the average investor over the longer term.

Early on, I tended to think that spotting and then avoiding downward trends in an investment category was the best way to manage one's investments with a particular emphasis on loss avoidance. In that regard, I suspect I was on a similar wavelength as the majority of fund investors.

Later, however, I repeatedly began to observe that performance of investment categories tended to run in multi-year cycles. Thus, while avoiding downtrends might be good strategy if implemented relatively early on, longer-term downtrends (in the range of a full 5 years or even more) for the various categories of stock funds usually turned out to be the precursors of decent and even considerably above average returns ahead.

Predictions Everywhere But Are They Trustworthy?

But such observations, representing essentially "educated guesses" about the future, no matter how honed by many years of observing the markets, can rightfully only be greeted by a healthy degree of skepticism.

To truly elicit confidence in the face of a world already far too filled with investment predictions, investors might want to consider going beyond most "ordinary" predictions, which are essentially just judgments stemming from past observation. Rather, to prove their worth, predictions might be expected to address two elements:

These two elements are indeed the very foundation for a predictive science's most established conclusions and therefore should likewise be part of the methodology for any work that claims to make investment predictions that are most worthy of one's attention. And, that is why I have adopted such a two-pronged approach to arrive at my site's stock fund Model Portfolios which represent an encapsulation of my research-based predictions.

Should You Trust Morningstar "Star Ratings?"

While many investors who come across my website may be unfamiliar with it, the great majority undoubtedly have at least some knowledge of the most widely known fund investor site, Morningstar is perhaps best of all to millions (even those who have never visited its highly recommended site) for its "star ratings" of funds, which show up on numerous other websites (and in fund marketing promotions) as well. In fact, it has been shown that investors do place a great degree of confidence in putting money into funds that receive either a 5 or 4 star rating from Morningstar, the highest of their ratings, while avoiding the lowest 1 or 2 star funds.

This is in spite of Morningstar's own frank acknowledgement that "the star rating is a grade on past performance. ...We never claim that they predict the future.” Since good fund past performance is widely recognized as a relatively poor predictor of good future performance, one must wonder why so many investors seem to gravitate toward the 4 and 5 star funds. (To learn more about the pluses and minuses of Morningstar star rankings, you may want to read this article.)

Morningstar is also the originator of the so-called "Style Box," basically a 3 by 3 grid showing a summary of the category of investments a stock fund emphasizes, taking into account the "growth" vs. "value" orientation as well as a "large-cap" vs. "small-cap" focus.

According to its site, "Morningstar is one of the most recognized and trusted names in the investment industry and serves more than 7.4 million individual investors, 245,000 financial advisors, and 4,200 institutional clients around the world."

Morningstar's "Buy the Unloved" Strategy

Given the proven lack of success in validating its star ratings to foresee future outperformance, it is therefore worth noting that Morningstar has postulated and subsequently attempted to verify the effectiveness of what they hope will be a more predictive strategy called "Buy the Unloved." This strategy has certain underpinnings that appear to add force to my own observation and findings that poorly performing investment categories can wind up being better places to put your money than the best performing ones. (It should be noted that my Model Portfolio was developed totally independently of the Morningstar approach and uses an entirely different methodology.)

Briefly, Morningstar's "Buy the Unloved" entails the following, according to Russel Kinnel, Morningstar's director of mutual fund research:

"Invest in mutual funds from the three equity categories that received the greatest redemptions in the prior year. (Bond funds and asset-allocation funds are excluded.) Then do the same thing the next year and the next year. After three years (or four or five--those time periods work just as well), you start rolling over that original group of funds in the next group of unloved."

In addition, the strategy suggests avoiding those categories that are receiving the greatest inflows from investors. (And for those investors so inclined, one might even "short", i.e. make bets against, such categories using Exchange Traded Funds (ETFs). Note, however, "shorting" should be viewed as a technique only for very aggressive investors. I have never recommended this technique on my site. In fact, currently, there appears to be insufficient evidence available to justify Morningstar's shorting strategy.)

Of course, to follow the "unloved" approach you must have access to data on "flows" in and out of fund categories. Such data, however, is hard for even the most web-savvy of investors to come by.

Given the high regard accorded to Morningstar's research, we find it highly encouraging that they can now report successful forecasting results using their Buy the Unloved method. By using their definition of "unloved" stock fund categories to buy, and then holding 3-5 yrs, their research shows you will outperform the averages. More specifically, examining 17 full years of results, Morningstar's Kinnel reports:

"From the beginning of 1994 to the end of 2010, the unloved earned 308% cumulatively or 9% annualized. That’s far better than the loved, which earned 157% cumulatively or 6.1% annualized. The MSCI World Index returned 4.6% annualized, and the S&P 500 returned 8% annualized."

One fascinating aspect of these results is that they are quite comparable to the research reported on my site which also shows that selecting stock fund categories based on those that are "out-of-favor" is an outperforming strategy. However, my results, compiled between 2000 and 2010, while also showing about a 3% outperformance, achieve that outperformance in comparison to the S&P 500 itself, not just between the most favored vs unfavored categories; the comparable S&P 500 comparison for Morningstar data is a mere 1%.

The latter Morningstar article also confirms something we recently pointed out to our readers: The method doesn't work as well when category returns are all bunched together, such as in the last year or two. Thus, there are periods especially during bear markets, when nearly all stock fund categories show poor performance. The practical effect is that no matter which category you previously invested in an attempt to get the best returns, there is little outperformance possible. (See our Oct 2010 Newsletter and below for more on this topic.)

I believe that one conceptual problem with the Morningstar strategy could be that its "unloved" categories are selected based on only one year's data. But fund investors often likely continue to sell funds for far longer than merely a year. If so, this would result in a continued drag on the category's subsequent performance considerably beyond a single year.

Even More "Unloved"

We favor a method that identifies long-term underperformance as a measure of how out-of-favor a category has become, not merely a category that is relatively disliked by investors for just a single year. To the contrary, an out-of-favor category, and therefore in our view, a particularly desirable new investment to make, has performed quite poorly over the last 5 to 10 yrs. Additionally, we usually recommend that it has shown signs of a sustained comeback over the last year. (Note: Our precise method of predicting which investment categories have the best forward-looking potential is proprietary; we report the categories our selection methods identify on our site but do not intend to publish the exact selection rules.)

Highly "Loved" Funds, and Those That Are Slipping

In our methodology, a "loved" category can simply be viewed as one that has a high annualized return over the last 5 or 10 years. (Strong past long-term returns are almost always a sure ticket to category popularity and would suggest the category has likely become overvalued as well.) An in-favor category, just as with an out-of-favor category, often remains so far beyond a single year. If so, when would we finally deem it as an undesirable investment? We consider not only high absolute returns, but also weigh a considerably substantial tapering off in performance as a sign that the category appears to have lost its edge.

All told, I would suggest that my method of identifying popularly favored vs out of favor categories appears to be better matched to the longer-term nature of how categories remain in and out-of favor. Additionally, the act of selecting underperformers is easier for the average investor to implement than delving into fund flows.

The Recent Dry Spell

While my method has proven to have had considerable success since its inception at the start of 2000, a perusal of our 11 year track record will reveal that index-beating Model Portfolio results have not been without some "dry spells." As noted above, outperformance is highly dependent on the availability of outperforming categories. Thus, beginning in the 4th quarter of 2007, the majority of all major categories of stock funds began doing as poorly or worse than the S&P 500 as the market as a whole tanked big time. That being the case, it became extremely unlikely for even previously underperforming categories, or any for that matter, to outperform the Index.

Until early 2009, investors have generally chosen to minimize their risks by investing in the "safest" of stock investments. They also turned to bond funds. As a result of the latter fact, investors who did not put all their "eggs" into just a stock basket were able to make up for a considerable amount of the underperformance that was seen in stocks. Bond funds, while "loved" by investors recently are likely to wind up disappointing investors over the next five years as the 2000-2010 decade's favorable cycle is gradually reversed by expectations of higher interest rates and more inflation.

We anticipate that underperforming stock fund categories, once they escape the influence of the prior decade's period of highly depressed overall stock prices, will once again show the market-beating outperformances seen in the above-referenced table for most of the past 11 years.

Similar Current Recommendations from Both funds-newsletter and Morningstar

Given that both our and Morningstar's recommendations are based on the similar premise that investors should pick fund categories that tend to be trailing the pack rather than leading it, one might expect a fair degree of agreement as to investment recommendations looking forward. And this is indeed the case for recommendations issued this January.

Specifically, both our and the Morningstar method seem to agree that Large Growth is the best category now (see my Jan. Newsletter), and that Large Caps are a better place to be than smaller stocks or emerging markets. The latter category, counter to current public opinion, seems to be the one with the poorest potential for outperformance over the next 3-5 yrs.

It should be noted that both Morningstar's 2011 predictions and my model portfolios are designed for investing from a moderately longer-term perspective. Importantly also, neither method implies that stock results will be good/great; it just suggests that relative to the indices, the named categories will do somewhat better. It should be remembered too that the majority of investors do worse than the S&P 500, so just nearly equaling it, or even better, beating it by a small amount is a big accomplishment.

Unfortunately, it appears that most investors think mainly in terms of where the overall market might be headed based on shorter-term considerations when deciding when and where to invest. In simple terms, we are all "wired" to want to avoid losing money. Therefore, when making our decisions, we tend to primarily weigh what we see happening right now and what we fear may happen over perhaps the next 6 months or year. We think this hard-to-break-away-from human tendency is what makes it difficult to follow what this very counterintuitive research suggests.


Click here to read "The Housing Barometer" by Steve Shefler.