The most common way investors select a mutual fund is simply by examining its historical performance, usually just the fund's total returns over a specific time period such as the last 3, 5 or 10 years. Some may even get a bit more sophisticated and select a fund based on its "risk-adjusted" performance. That is, the fund's total return divided by some measure of risk to create a ratio (i.e. how much return do you get for each unit of risk). Intuitively, it is more appealing to select funds that have achieved higher returns per unit of risk taken.
We think these are flawed ways to select funds for your portfolio for a variety of reasons.
The main problem with picking mutual funds simply on the basis of their returns is that the markets are cyclical and returns don't persist (in other words, past performance is a poor predictor of future performance). What has risen must fall and what has fallen may rise. Even using risk-adjusted numbers does not avoid this problem. Since risk-adjusted returns still have the total returns in the numerator, the funds that have had the superior total returns still generally come out on top.
One of the mutual fund industry’s most notable risk-adjusted performance measures is the Morningstar RatingTM. Morningstar's well-known "star-system" ranks mutual funds based on their risk-adjusted performance (five stars being best, one star the worst) over various time periods. The Morningstar RatingsTM may be the most important fund rating system in the industry, at least in terms of directing investment dollars. It has often been stated that 80% of the flows in the mutual fund industry go to funds rated four or five stars.
Unfortunately, it’s generally not until late in a given market cycle that this kind of rating system will start to reward funds that were positioned well for that cycle with high scores. In other words it tends to reward funds that were positioned for what just happened in the market -- instead of those positioned for what is about to happen.
To be fair, relative risks do tend to persist, not only between asset classes but also between mutual funds within specific peer groups. A fund that is riskier relative to another over a certain timeframe will tend to be riskier over the next timeframe as well. There are plenty of exceptions, but this is a good general rule.
So, if the relative returns between funds are highly variable, but their relative risks hold fairly steady, it is indeed a reasonable expectation that risk-adjusted performance will likely persist! To Morningstar’s credit, (and something that nobody in the financial media picks up on) using past Morningstar RatingsTM to predict future Morningstar RatingsTM can actually work if given enough opportunities to make that bet.
But simply using Morningstar RatingsTM, or any other risk-adjusted measure, is not a great way to pick a mutual fund on a stand-alone basis, particularly within a specific peer group -- and it’s definitely not how to build a portfolio of mutual funds.
First and most obvious, you need to understand what drove the return and risk characteristics of the fund in question. Have the people managing or supporting the fund changed? Is management honest and capable? Has the investment philosophy or process changed? Is the strategy understandable and will it be effective moving forward? How was the portfolio positioned over time? What has the market done over this time frame? How do they perceive current market conditions? Here at Kobren Insight Management, we talk directly to each fund that we invest in and ask those (and more) questions. It is critical to picking the right funds.
Let’s pretend for a moment that this is just a quantitative exercise in picking a mutual fund and we'll exclude such "qualitative considerations." Picking a fund solely on its risk-adjusted performance likely means you're buying either a low-volatility fund just after the market has performed poorly, or a high-volatility fund just after the market has performed well. Or it could be a fund that concentrated in one economic sector, market cap or investment style that happened to have strong relative returns. In other words, it’s just another way of chasing performance -- and a strong reason why the typical equity mutual fund investor has only gained 3.5% a year over the last 20 years even though the market has gained 13.0% a year according to Dalbar Research.
Also, if you're building your portfolio by picking funds simply on the basis of their risk-adjusted performance, there is a strong possibility that you may not be taking enough risk to reach your financial objectives. To generate an reasonable return, an adequate amount of risk must be taken. Yet, often investors will pick the funds with the lowest risk -- who doesn't like the idea of low risk? -- only to see disappointing results.
Generally speaking, lower risk funds within a specific peer group tend to have above-average cash positions. So even if an investor thought they were building a diversified portfolio by selecting funds in each category with superior historical risk-adjusted performance, they may actually just be loading up on cash-rich funds. The result is that their effective asset allocations (i.e., how much they have in stocks, bonds and cash) will not be at appropriate levels. (Please note that there are other ways that funds within peer groups can have relatively lower risk besides a lot of cash, but cash levels are indeed one of the leading reasons).
At Kobren Insight Management, we certainly look at risk levels when we examine funds -- along several different dimensions. But, the key point is that a fund’s level of risk is only a descriptive element -- not a value judgment. We're emotionally neutral on whether a fund is "low-risk" or "high-risk." We just need to know what that fund’s risk characteristics are, to help define how it will interact with the other funds in the portfolio. In fact, how the funds in your portfolio interact with one another (the most common statistic here is "correlation") is arguably an even more important characteristic in determining overall portfolio risk than the risk levels of the individual funds. In sum, a fund’s level of risk is one basic building block information that we require to construct an over-all portfolio that is appropriate for a specific client’s risk and return requirements.
Now at the risk of confusion, once a balanced, diversified portfolio of mutual funds has been assembled, then risk-adjusted performance of the whole portfolio becomes a good way to measure a money manager’s over-all performance. Risk-adjusted performance tends to be a win-win for both parties in the client-investment manager relationship. Clients find it easier to "stay the course" with better risk-adjusted performance and staying the course tends to increase their probabilities of attaining their portfolio objectives. Investment managers meanwhile, are able to retain their satisfied clients.
Next month we'll talk about some of the ways we actually use quantitative information to help us identify mutual funds, including our proprietary rating system. 
