In talking to thousands of investors each year, there are nearly as many ways to choose mutual funds as there are investors choosing mutual funds. These range from "choose last year’s top-performers" to "my neighbor is the fund manager," and everything in between. At MutualFundAlliance.com, we have spent years studying various fund selection methodologies before designing and putting our own methodology through a series of rigorous statistical tests. Below, we describe a few of the techniques we employ in creating and testing the MutualFundAlliance.com investment methodology and how you can use these techniques to test your own fund-picking prowess.
The first and most obvious step is to create your own investment methodology. In the MutualFundAlliance.com Investor Library, we discuss the MutualFundAlliance.com investment methodology and how we have developed and refined it over time. In addition to our methodology, there are many other ways to choose mutual funds for your portfolio. Broadly speaking, there are two schools of thought relative to investment methodologies:
Fundamental analysis focuses on the financial performance of the underlying investments within the mutual fund’s investment portfolio. Investors who employ fundamental analysis focus data such as the price/earnings ratio, price/book ratio, and the earnings growth rate. There are two ways of analyzing fundamental data for mutual funds: 1) analyzing these data for the major holdings of the funds or 2) analyzing the average fundamental data for all of the funds as found in such sources as Morningstar. Either way, these data is changing on a daily basis as the fund manager buys and sells investments. As we discuss elsewhere in the MutualFundAlliance.com Investor Library, we have created statistical models that analyze fundamental data for the underlying fund portfolio and estimate the daily holdings of each mutual fund.
Technical analysis focuses on the price pattern of investments. The underlying assumption is that the pattern of prices that investors pay for a stock or mutual fund gives insight into the future direction of the investment. When you hear portfolio managers on CNBC talking about a “trading range”, “retracement”, or “trading within a band”, they are referring to technical analysis.
Please see the MutualFundAlliance.com Investor Library for the latest analysis on both fundamental and technical analysis and references to books on the topic.
Once you come up with an investment idea that you think will outperform your current investment methodology, you need to put it through a series of historical tests known as backtesting. Backtesting is applying your new investment methodology to historical data, to determine what your strategy would have accomplished historically, over a significant period of time.
Backtesting is not only a good step because it puts your methodology to the test, but because it also forces you to quantify the details of your theory. Backtesting your investment methodology based on fundamental analysis is very difficult since historical fundamental, such as Price/earnings ratios, is not readily available to individual investors. At MutualFundAlliance.com, we have been collecting this data for over 12 years and have it available in our databases and use these data to continuously test and refine our investment models.
On the other hand, historical price data for stocks and mutual funds is readily available for free on the internet. Therefore, it is very easy to test moving average methodologies using such software MarketBrowser.
Once we conducted an adequate series of historical tests on our methodology, we put our methodology to work on a forward-looking basis, by using it with a phantom portfolio. This is as equally important a test as backtesting since there are many more real-world variables that rear their heads going forward than you can overlook historically. We may choose to run our phantom portfolio for months, or even years. In developing the MutualFundAlliance.com Model Portfolios, we ran our phantom portfolios for years. Furthermore, whenever we update the estimating parameters of our Model Portfolios, we once again test the portfolios using phantom portfolios.
Finally, once we have put our methodology through a series of rigorous statistical tests and are satisfied with the results, we are ready to roll it out and begin using it. And in spite of all the testing, the methodology’s ultimate effectiveness can only be judged by actual returns and how much money has been made through its implementation.