Stock Star Rating Review for 2002

With 2002 firmly in the rearview mirror, it's time to check in on the Morningstar Rating for stocks. Although we've made a few rotten calls, the star rating is doing fairly well overall, and it looks like the changes we implemented last year are also working out so far.

First, a brief refresher. To hit 5 stars, a stock must trade at a significant discount to our fair value estimate. The narrower the company's economic moat--its competitive advantage--and the greater its risk, the larger this discount needs to be. A high-risk stock with no economic moat (like chipmaker Micron Technology (MU)) would need to trade 60% below our fair value estimate before it gets 5 stars, while a low-risk stock with a wide economic moat (like data processor Automatic Data Processing (ADP)) would need only a 20% discount to fair value. The same differences apply on the other end of the star-rating spectrum: High-risk, no-moat stocks hit 1 star when they trade just 10% over fair value, while higher-quality firms need to appreciate to 30% beyond fair value before they reach 1 star.

The Envelope, Please…
As in our previous checkups, we measured performance using a hypothetical portfolio that purchased $1,000 worth of each 5-star stock and sold any stocks that subsequently hit 1 star, with a $10 cost tacked on to each purchase and each sale. Since inception (Aug. 8, 2001) through the end of 2002, this portfolio would have lost 7.2% versus a loss of 25.6% for the S&P 500. For calendar 2002, the "buy 5, sell 1" portfolio would have lost 7.6% versus a loss of 23.3% for the S&P 500. So, in aggregate, the star rating has added significant value over a passive indexing strategy.

Of course, buying low and selling high isn't the same as buying and holding an index, so we've also calculated a "comparable market return" that buys and sells $1,000 worth of S&P 500 SPDRs (SPY) at the same time any 5-star stock is bought or 1-star stock is sold. (The SPY is an exchange-traded fund that tracks the S&P 500.) Since inception, this hypothetical portfolio would have lost 10.3%, and in calendar 2002 it would have lost 6.8%. As before, it looks like the star rating is doing a great job of identifying when stocks are cheap.

But it seems to be doing only a middling job of picking out which stocks are cheap. This conclusion is bolstered by some "batting averages" that I calculated, which showed that roughly half the stocks that have been rated 5 stars have outperformed a SPDR purchased at the same time, while roughly half have done worse than the market. On a brighter note, about three quarters of the stocks that hit 5 stars during 2002 wound up returning better than the S&P 500's 23% loss--and roughly half ended the year with positive returns. So, the odds of buying a 5-star stock that wound up beating the market in 2002 were pretty good.

So much for the aggregate numbers. Let's look at what we've gotten right and wrong so far, and what that bodes for the future.

What We Got Right…
We've had a fair value estimate of around $22 on J.C. Penney  for quite some time, and the stock stayed solidly above $22--so we advised investors to wait for a better price. This past summer, however, the firm's restructuring hit some snags and the stock slid all the way to the mid-teens, which we thought was an overreaction. Turns out we were right, and the stock soared 60% all the way to almost $25 by early December, at which point it hit 1 star.

We also had successes with some higher-quality companies when the market tanked in July. Automatic Data Processing is up about 25% since it hit 5 stars in mid-July, while drugmaker Eli Lilly (LLY) has risen about 33%, Baby Bell SBC  is up 26%, and Berkshire Hathaway (BRK.B) has appreciated about 17%. Lilly and SBC have risen too far to be worth buying at this point, but Berkshire and ADP still look undervalued at their current prices.

The star rating also has done a solid job of keeping investors away from overvalued stocks. Sepracor  has plunged more than 70% since we warned investors back in August 2001 that the shares were too speculative. Computer Associates  has tanked about 60% since we started panning the shares around the same time, and Motorola  (MOT) shares have been halved since they hit 1 star in October 2001. We've also flagged some companies that weren't worth owning at any price, like UAL (UAL), US Airways , and Palm , all of which have lost most of their equity value since we began covering (and panning) them.

…and Wrong
I detailed some of our worst errors in last July's performance update, but since we're human, we've continued to make some bad calls. For instance, we called tire manufacturer Goodyear (GT) "undervalued enough to be worth a look" at $17 per share in late July, and the shares ended 2002 at about $7. Lesson: No matter how cheap it looks, a company in a terrible industry is still a company in a terrible industry--and there are few industries worse than tires, in which every advance in technology leads to lower replacement demand. As Warren Buffett has said, "When management with a reputation for brilliance tackles a business with a reputation for poor economics, it is the reputation of the business that stays intact."

We also blew a call on software favorite Siebel Systems , calling a late-June drop in the stock's price "a buying opportunity." Unfortunately, we redefined "opportunity cost" with this one, since the shares fell almost 50% by the end of the year. A good lesson in the difficulty of calling the bottom in just about any tech-related business.

Finally, we missed some drug and biotech firms like Bristol-Myers (BMY), Wyeth  (the drug company formerly known as American Home Products), and ICOS , all of which have lost significant value since we said they looked like compelling investments. In the case of ICOS, we underestimated the odds that the Food and Drug Administration would delay a key product approval, and we could have avoided the bad call on Bristol by paying less attention to the firm's financial solidity and more attention to the folks running the firm. Wyeth was less avoidable, but no less painful a loss: The stock got hurt after the release of a surprising study that indicated that one of the firm's leading products could increase the likelihood of breast cancer and cardiovascular problems in women.

You Win Some, You Lose Some
Overall, though, we've done a better job avoiding the real stinkers than we had when I last updated the performance of the star rating. Some of this is due simply to greater caution and conservatism in our cash flow projections, and some of it is due to the recent changes in the star rating. Since we started risk-adjusting the star rating in late September 2002, our hypothetical "buy at 5, sell at 1" portfolio has gained 5.9%, which is not bad relative to the S&P 500's 2.9% gain and the SPDR portfolio's 1.9% rise.

Though I'm loath to attribute much significance to three and a half months' worth of results, it looks like the changes to the star rating have been for the better. We have far fewer 5-star stocks than we did before we made the change--even during the market's swoon in early October, only about 15% of our coverage universe garnered 5 stars--and we certainly have fewer buys than the typical Wall Street firm. Right now, 26% of all stocks rated by "the Street" have "strong buy" ratings, and fully 45% have "moderate buy" ratings. For comparison, just 3.4% of our coverage universe merits a 5-star rating, with 5.6% getting 4 stars.

On balance, the biggest benefit of the star rating is that it forces patience, discipline, and a contrarian stance: To paraphrase Buffett, being greedy when others are fearful and fearful when they're greedy is a strategy that pays off handsomely over time. Although we've occasionally gone astray by being greedy when we should have been fearful (Goodyear, Bristol-Myers, Siebel, and so on) we'll continue to work hard to ensure that the fair value estimates that drive the star rating are as realistic as possible.

Look for the next star-rating performance update around mid-April, when I'll review the first quarter of 2003.