Morningstar's Contrarian Portfolio--10 Stocks to Buy

Contrarian investment strategies are popular among value-oriented investors, and in our opinion, rightly so. Stocks that are widely ignored or despised are more likely to sell at an attractive discount to fair value than stocks with heavy analyst coverage or a household name.

At Morningstar, we do a lot of behind-the-scenes research looking to uncover attractive investment returns, and contrarian strategies are a natural fit with our preference for uncovering stocks that are selling with an attractive margin of safety. Over the past couple of years we've been researching our Web site traffic, comparing the returns from popular stocks with the returns from unpopular stocks, and the results surprised us. Investing in the stocks that are most popular with Morningstar's audience offers good odds of achieving market-beating returns, but higher returns seem much more likely in less-popular, or "contrarian," stocks. We're not surprised that many stocks don't attract much attention. We follow about 1,500 stocks, a good portion of which are from overseas markets or obscure industries, and frequently add new coverage. It's probably impossible to keep up with such a lengthy list--but therein lies an advantage. We think that learning more about lesser-known stocks, especially those that have a Morningstar Rating of 4 or 5 stars, can really pay off. To help you get started, we've decided to publicize a portfolio of our 10 best "contrarian" 5-star stock ideas for 2005.

Before we cut the ribbon, we'd like to walk through our research and explain how we selected our "Contrarian 10"--and why our research concluded that this strategy has merit. At Morningstar, we can directly measure a stock's popularity--or lack thereof--with our subscribers. Using our Web site data, we count unique page views for each stock we cover, and then create a page rank that measures that stock's popularity with our clients. We define a page view as one client clicking on the Analyst Report for a stock once. This yields a direct, and unique, measure of "popularity." For example, in 2003,  Pfizer  was our most-viewed Analyst Report (with a page rank of 1), while  Footstar  (page rank 6,495) was least popular. By creating an ordinal ranking of the stocks we cover using our page ranks, we can easily identify the most "contrarian" stocks we cover.

Let us be the first to note that this methodology isn't perfect. Our 5-star stocks are the most popular, so they generate more page views and thus higher page rankings, and well-known stocks tend to get more page views regardless of value or market activity. And a stock that has a strong run through part of the year will get a popularity boost, even if it tanks later on. Our page-rank data is also a partial measure of our subscribers' interests, which tend to lean heavily toward undervalued stocks in general. Finally, we can't prove that Morningstar's subscribers are a representative sample of the financial markets. However, we think the data are reliable enough to prove useful, and any strategy that helps investors uncover good businesses they can buy with an attractive margin of safety has merit. What's more, we also think that awareness of the most-popular stocks among Morningstar subscribers has its own usefulness.

To get started, let's review the 10 most popular stocks among Morningstar subscribers in 2003 and 2004:

2003's 10 Most Popular Stocks
Name	Rank	2003 Return	2004 Return
Pfizer	1	17.76%	-22.31%
Wal-Mart Stores	2	5.73%	0.51%
Microsoft	3	6.81%	8.95%
General Electric	4	30.75%	20.68%
Home Depot	5	49.42%	21.47%
Altria Group	6	42.74%	18.38%
Johnson & Johnson	7	-2.11%	25.17%
Cisco Systems	8	84.96%	-20.26%
General Dynamics	9	15.97%	17.40%
First Data	10	16.28%	3.72%
Average Return (equally weighted)	6	26.83%	7.37%
S&P 500	--	28.69%	10.88%
Outperformance	--	1.86%	-3.51%

 

2004's 10 Most Popular Stocks
Name	Rank	2004 Return
Pfizer	1	-22.31%
Microsoft	2	8.95%
Wal-Mart Stores	3	0.51%
General Electric	4	20.68%
Merck	5	-27.76%
Coca-Cola	6	-16.12%
Johnson & Johnson	7	25.17%
Anheuser-Busch Companies	8	-1.95%
Home Depot	9	21.47%
Berkshire Hathaway B	10	4.30%
Average Return (equally weighted)	6	1.29%
S&P 500	-	10.88%
Outperformance	-	-9.59%

While most of these stocks are unsurprisingly popular, the returns aren't consistently attractive. The 2003 stocks delivered positive returns in both years, but lagged the returns an investor in the S&P 500 would have accrued. The 2004 stocks barely broke even, and both Pfizer and  Merck  declined by more than 20%. These returns certainly aren't horrible, but we'd be inclined to try a little harder.

So let's review the returns of Morningstar's 10 least popular stocks for 2003 and 2004. If our "contrarian" theory has merit, these should be higher:

2003's 10 Least Popular Stocks
Name	Rank	2003 Return	2004 Return
Neiman Marcus	1,907	76.60%	34.57%
Gabelli Asset Man.	1,645	32.56%	26.50%
Knight Ridder	1,467	24.48%	-11.83%
Navistar Intl.	1,403	97.00%	-8.17%
Citizens Comm.	1,288	17.73%	34.09%
Getty Images	1,265	64.09%	37.34%
Sealed Air	1,251	45.15%	-1.61%
VISX	1,210	141.65%	11.75%
Barra*	1,201	*	*
Bausch & Lomb	1,199	46.11%	25.30%
Average Return (equally weighted)	1,384	54.54%	14.79%
S&P 500	--	28.69%	10.88%
Outperformance * Was acquired, data unavailable	--	25.85%	3.91%

 

2004's 10 Least Popular Stocks
Name	Rank	2004 Return
Equity Lifestyle Properties	6,393	20.88%
Pemstar	6,377	-44.99%
WebMethods	6,324	-20.77%
TIM Hellas Telecom. SA ADR	6,278	45.09%
Distribution y Servicio D&S ADR	5,816	-11.12%
Molex	5,732	-13.66%
Marsh Supermarkets A	5,667	8.57
Kindred Healthcare	5,630	15.24%
United National Group A	5,494	5.38%
Com. Brasileira de Distribuicao	5,445	2.76%
Average Return (equally weighted)	5,916	0.74%
S&P 500	-	10.88%
Outperformance	-	-10.14%

And generally speaking, they are. The returns on our least popular 2003 stocks were more than double the returns of the most popular, and they outperformed the S&P 500 in 2003 and 2004 by a healthy margin. That's more like it. However, the 2004 returns lagged both the S&P and the most popular stocks, although not by much, so we have more research to do. One aside is important at this point--note the average page rank of these stocks. For the 2003 stocks it was 1,384, and for the 2004 stocks 5,916. In other words, on average during 2004 there were more than 5,900 stocks that were more popular among Morningstar's subscribers, so we're definitely in contrarian territory. (Morningstar only covers about 1,500 stocks, so the other rankings reflect page views for stocks we don't cover). Now we need a further reality check--in the form of the returns of the 10 best-performing stocks we covered during 2003 and 2004. For our contrarian theory to hold up, the average page rank of these stocks ought to be low:

2003's 10 Best Performers
Name	Rank	2003 Return
Capital One Financial	63	106.74
Netflix	206	396.73%
DR Horton	207	152.07%
Guidant	244	96.11%
Countrywide Financial	252	97.41%
Millennium Pharmaceuticals	295	133.42%
Centex	298	117.54%
Jack Henry	331	72.51%
National Semiconductor	624	162.56%
Ann Taylor Stores	1,076	90.99%
Average Return (equally weighted)	360	142.61%
S&P 500	-	28.69%
Outperformance	-	113.92%

 

2004's 10 Best Performers
Name	Rank	2004 Return
Cheniere Energy	1,473	444.44%
Elan ADR	186	295.50%
Chicago Mercantile E	52	218.54%
Autodesk	328	209.62%
Apple Computer	65	201.36%
American Eagle Outfitters	478	188.04%
Metris Companies	1,814	187.16
AK Steel Holding	943	183.73
TXU	332	177.67
BanColombia SA ADR	4,426	176.40%
Average Return (equally weighted)	1,010	228.25%
S&P 500	-	10.88%
Outperformance	-	217.37%

And sure enough, in both cases our best-performing stocks boasted an average page rank in the bottom third of our research universe. (Morningstar covered 500 stocks in 2003, 1,500 during 2004, and even more today). Even  Chicago Mercantile Exchange , the most popular of these stocks, was only the 52nd most-popular stock among our subscribers during 2004. So it seems safe to conclude that our contrarian strategy has merit. To put our research into action, early last year Morningstar formed a 10-stock contrarian portfolio. In January 2004, we chose 10 stocks with a 5-star rating, subject to the constraint that the average page rank of the portfolio had to be in the bottom third of our research universe. The results confirmed our previous research:

2004 Morningstar Contrarian Portfolio
Name	Rank	2004 Return
CarMax	163	0.39%
Chicago Mercantile Exchange	175	218.54%
Accenture	215	2.58%
AutoZone	226	7.16%
Biogen IDEC	247	81.50%
Expeditors Intl of WA	334	49.04%
Washington Post (WPO)	385	25.21%
Triad Guaranty (TGIC)	513	20.12%
TransCanada (TRP)	515	20.44%
CH Robinson Worldwide (CHRW)	979	48.14%
Average Return (equally weighted)	375	47.31%
S&P 500	-	10.88%
Outperformance	-	36.43%

Our "Contrarian 10" delivered a much-healthier return than the 10 most popular stocks we highlighted earlier, and trounced the S&P 500. These are definitely the kind of returns that catch our eye. And true to contrarian form, the average page rank of these 10 was 375--easily placing them amongst the least popular of 2003.

Introducing Morningstar's 2005 Contrarian Portfolio
We think that the promise of our Contrarian strategy warrants releasing it for investors to consider. To that end, we'll evaluate our 2004 and 2005 contrarian portfolios over time, and if the concept continues to generate attractive returns we may publish these annually. While we won't release page rank data, we will provide an occasional update on our contrarian portfolios. Here are our 10 contrarian stock ideas for 2005, along with their page ranks and 2004 returns:

2005 Morningstar Contrarian Portfolio
Name	Star Rating	Rank	2004 Return
Micrel		2,865	-29.18%
Fed. Investors (FII)		151	5.00%
Lear (LEA)		477	0.85%
Group 1 Auto. (GPI)		893	-12.96%
Millennium Phar.		600	-34.91%
Intersil		474	-32.28%
DeVry (DV)		420	-30.92%
Allied Waste		293	-33.14%
eSpeed		271	-47.322
Fifth Third Bank (FITB)		162	-17.89
Average Return (equally weighted)	--	661	-23.28%
S&P 500	--	--	10.88%
Outperformance	--	--	-34.16%

The higher average page rank of this year's contrarian portfolio reflects a relative lack of 5-star stocks--itself a comment on the market's valuation. And all of these stocks performed poorly during 2004, helping them trade at the margin of safety required to earn our 5-star rating. Based on our research, we think our contrarian portfolio has a better-than-average chance of earning attractive returns. It's also a good way to focus your research on some undervalued names you may be less familiar with.

But the real proof is in the investment returns. We'll check back in with these portfolios for 2004 and 2005 later in the year.