Fidelity Contrafund: How do they do it?

Motley Fool has a recent article by Selena Maranjian about closet indexing.  It does a great job of setting up the problem - indexing is good, but you should pay a low index fund price for it.

The problem, as we have discussed before, is identifying the funds that are closet indexing.  Ms. Maranjian suggests two different ways - looking at top ten holdings or looking at R-squared.  Looking at top ten holdings has obvious limitations when funds often have hundreds of holdings.  And we've discussed the shortcomings of R-squared in this space before.

So what is interesting about this article is that it presents us with a nice little mystery - Fidelity Contrafund.  As Ms. Maranjian points out, "Fidelity Contrafund, for example, has an R-squared of 94.87%, but it has significantly outperformed the S&P 500, on average, over the past five, 10, and 15 years." Taking a look at the top ten holdings as of 9/30/2011, I see recognizable, large companies like Apple, Google, Disney, Berkshire Hathaway, Wells Fargo, and Coca-Cola.

So how does Contrafund manage to outperform?  Ms. Maranjian chalks it up to a few good picks in stocks like priceline.com, Dollar Tree, and Chipotle.  But Contrafund had 453 holdings as of 9/30 - I don't think three good picks tell the whole story.

So I dug through Contrafund's Q1 2011 disclosures to the SEC to get a better idea of how they're doing it.  Morningstar classifies Contrafund as a US Large Cap growth fund, and it's easy to see why when looking at the top ten holdings list.  But it's complete list of holdings tell a different story.  I counted 27 stocks traded on the Hong Kong exchange including not-exactly-household names like "Hengdeli Holdings Ltd." and "Luk Fook Holdings International Ltd."  I also counted 16 Brazilian stocks including "Mills Estruturas e Servicos de Engenharia" and "Diagnosticos da America."  Even it's US holdings are not universally large cap, with small cap holdings like "Gardner Denver, Inc." and "Clean Harbors, Inc."

So my guess of how Contrafund achieves its outperformance while still having a fairly high R-squared is that its large positions are in US Large Cap growth.  However, it "juices" those returns through smaller investments in higher returns categories like small caps and emerging markets.

Now Contrafund definitely isn't a closet indexer.  But it is a style violator.  Why should you care?  Because you carefully select your asset allocation and rebalance annually.  When you select Contrafund as your US Large Cap growth exposure, you're not counting on this hidden overlap with your small cap and emerging markets exposure, so you're not getting the lack of correlation that you should get.  That will lower the effectiveness of your rebalancing going forward.

Don't panic on treasuries, but corporates need fixing

The recent Shifting Bond Maturities and My Latest Mistake post by Mike at Oblivious Investor got me thinking about bonds.  When you're holding bonds, you shouldn't really worry about rising interest rates that much.  Sure, their value would fall, but you would rebalance, buying more bonds, and you would end up holding your older bonds to maturity.  No big loss - the older bonds will approach par value as they near maturity.  The new bonds that you're buying will be yielding more, so you end up doing fine as explained by a Vanguard research paper.  So I agree with Mike - stick with your intermediate term bonds, and don't worry too much about it.

When we get out of treasuries and into corporates, we face a lot more issues.  Among them are that the indexes overweight high levels of debt as Matt Hougan points out on IndexUniverse.  It seems to me that there are some simple tweaks that could be used to improve the indexes.  If we assume that the market is somewhat intelligent and demands a higher yield for a good reason, why not just divide issue size by the current yield-to-maturity?