Over the years, I've found investors trying to evaluate and choose mutual funds by using many approaches. Some of these are good and others are bad, but by far the most useless is one that focuses primarily on the stocks that a mutual fund invests in. And yet, there's no shortage of people who try and do this. Unfortunately, this style of analysing mutual funds is prevalent not just among investors, but even among some analysts and in parts of the financial media.
Generally, the approach taken by this style of analyses is to find out a fund's latest investment portfolio and then try to read up on the individual stocks and see if they fit the current definition of stocks that are likely to go up in the near future.
There are two problems with this approach: One, it places upon the investor the onus of being a better judge of stocks than the fund manager. The starting premise itself is that the investor (or his advisors) know which stocks are going to move and they should judge the fund's last declared portfolio on that basis and then decide whether the fund is any good or not. Generally, this approach is taken by investors who have primarily been investing in stocks and have decided to dabble a little bit in mutual funds. Based on the research reports I get from brokerages, my belief is that the prime drivers of this approach are stock broking outfits which have also added on the business of selling mutual funds. A lot of the investors who try to follow this approach seem to be interacting with these broking houses.
These outfits have given over the task of researching funds to their stock researchers. A typical piece of logic in such reports goes something like: "The top two holdings of this fund were SBI and ICICI Bank; neither are on our buy list right now, so this fund is not recommended". Stock research is what these people do, so that's what they do in mutual funds also. When all you have is a hammer, then everything looks like a nail.
The other problem with this approach is one that is seen more often in the financial media. I recently read an amusingly misguided article by a financial publication which had picked individual stocks (in isolation) from funds, followed the fund managers' actions on those stocks through a couple of years and then declared those actions to be illogical. The amusing part was that the funds that had been picked out were mostly those that had outperformed the benchmarks as well as their peers over the last few years.
This highlights the basic lack of logic of using fund's holdings as the primary way of evaluating a fund. A mutual fund's portfolio (as should be any portfolio) is more than the sum of its parts. A stock may be held or sold for reasons that have nothing to with itself. Perhaps, another stock in that industry has become more attractive and there's an internal limit on an industry. Perhaps there's an internal limit on total exposure to a certain size of company.
The most important input in analysing a fund is the returns that it has generated for investors in comparison to the benchmarks and its peers. The only role that its holdings play is to analyse the past and potential quality of the returns, and that too more at an aggregate level. These are the kind of questions that an analyst could expect to answer by looking at a fund's portfolio: How concentrated is a fund across industries or capitalisation bands? Are the top few holdings a lot larger than the smaller ones? Does the fund change its portfolio a lot in response to market conditions or only a little? When all other information is in, then you can look for explanations in the portfolio of a fund - using it for any primary analysis is truly pointless.