The Wall Street Journal created a stir with its takedown last week of research firm Morningstar's "star ratings" system for mutual funds.

From the article, titled "The Morningstar Mirage" (behind paywall, but if you're not a WSJ subscriber you may be able to access it here via Twitter):

From pension funds to endowments to financial advisers to individuals, investors rely on Morningstar's star ratings to help divide $16 trillion among America's mutual funds, in much the way shoppers use Amazon's ratings to pick products. A lot of these investors, and the people paid to guide them, take for granted that the number of stars awarded to a mutual fund is a good guide to its future performance.

By and large, it isn't....

Of funds awarded a coveted five-star overall rating, only 12% did well enough over the next five years to earn a top rating for that period; 10% performed so poorly they were branded with a rock-bottom one-star rating.

Although the Journal gave the story front-page treatment, there's really not much new here. The WSJ reported similar information three years ago in a piece titled "Mutual Funds' Five-Star Curse," and Marketwatch posted a story in 2010 titled "Five-star mutual funds don't live up to their past."

In a written response to the current WSJ story, Morningstar CEO Kunal Kapoor acknowledges the limitations of the star ratings but defends their (limited) predictive capabilities:

We recognize and have often acknowledged the limitations of a measure like the star rating that's based on past performance, but we also believe it can usefully tilt the odds in investors' favor, when combined with other research and tools.... Our research finds that the star rating points investors toward cheaper funds that are easier to own and likelier to outperform in the future.

All this is a tempest in a teapot. Research on fund performance (including a study published by Morningstar itself just last year) demonstrates that that longer-term past success bears little or no relationship to current or future performance. However, more-recent success does have some correlation with near-term future performance. Just as some baseball teams "get hot" and enjoy winning streaks, so do some mutual funds — for a while.

This is why SMI's Upgrading methodology looks only at a fund's most-recent performance — more specifically, its most recent 3-, 6-, and 12-month performance. In contrast, Morningstar's star system focuses on 3-, 5-, and 10-year past performance (more details here in PDF), making it less likely to be predictive than if it concentrated on the shorter-term. Upgrading leads us to the best current performers, irrespective of the number of stars funds may have beside their names in the Morningstar database.

Consider our current universe of 15 fund recommendations. Here's a breakdown of how they align with Morningstar's star ratings.

★ ★ ★ ★ ★ Five-star funds: 5
★ ★ ★ ★ Four-star funds: 7
★ ★ ★ Three-star funds: 3

As you can see, two-thirds of our current recommendations rate fewer than five stars. In the past, we've even had plenty of two-star funds among our Upgrading recommendations, and probably will again in the future.

The star-ratings system, though well-intended, leads many people — as the WSJ and Marketwatch stories note — to make investment decisions that yield inferior results. "Investors pour money into top-rated funds even if their performance declines," notes the current Journal piece. "[They] pull money from low-rated funds even if their performance improves."

Morningstar provides a great deal of useful information to investors. Indeed, it would be difficult for SMI to operate without access to Morningstar's wealth of performance data. But to quote from an old song, when it comes to choosing mutual funds "don't let the stars get in your eyes."