What do you think about the end of Bill Miller's streak at Legg Mason?
Statistical analysis of Bill Miller's track record leads us to conclude that his outstanding results over the years were most likely due to skill rather than luck. As we have noted in the past, active management success is based on superior forecasting skill (whether of fundamental values or the future behavior of other investors, or both), which in turn is based on some combination of superior information or a superior forecasting model. We have also noted that superior sources of information can dry up or become widely known (and thus less effective), while superior models can either be copied or have their assumptions invalidated by ongoing changes in the economy or investor psychology. The best active managers have a unique ability to keep these wolves at bay. Yet Miller's case proves that even for the best managers, this power is not infinite. Of course, there is another possible explanation for Miller's performance - his views may still be right, but unfortunately not proven within the calendar year performance measurement period which anchors the psychology of so many investors. In this regard, we're reminded of a famous quote from Julian Robertson (founder of Tiger Managemenet) in 2000, at the height of the internet bubble: "The current technology, internet and telecom craze, fueled by the performance desires of investors, money managers and even financial buyers, is unwittingly creating a Ponzi pyramid destined for collapse. The tragedy is, however, that the only way to generate short-term performance in the current environment is to buy these stocks. That makes the process self-perpetuating until the pyramid eventually collapses under its own excess." It may well be that Miller's forecasts are right, but that they won't be proven so until more time has passed. The final thought we have about Bill Miller is that his skill as an investment manager became obvious only over time and with hindsight. It remains extraordinarily difficult for an investor to accurately forecast (i.e., identify) the future Bill Millers and Warren Buffetts.
How do you treat inflation in your model portfolio calculations?
We do not explicitly take future inflation into account in our model portfolio analyses. There are two reasons for this. First, we believe that investors are primarily interested in maintaining targeted levels of future real consumption. To put it differently, they want to be able to buy the same grocery basked regardless of the prices of the items. Hence, we assume that investors are primarily interested (consciously or not) in real (inflation adjusted) rather than nominal returns. The second reason we use real returns is that we assume we have a fallible crystal ball when it comes to projecting future inflation. Were we to attempt to forecast future inflation, and were our forecast to be wrong (which seems the most likely outcome, by far), if we used nominal returns in our calculations our conclusions could be invalidated. By eliminating an inflation forecast, we also eliminate one source of estimation error that could adversely affect the conclusions of our model portfolio analyses. We acknowledge that, based on the frequency with which we're asked this question, that thinking in real rather than nominal terms is a rather unnatural act for many investors. For example, we often have to explain that our assumptions about annual savings or portfolio withdrawals are expressed in real terms and must therefore be converted into nominal terms in the "real world." For example, a real annual savings or withdrawal of $10,000 in year one rises to $10,300 in year two after a year of inflation at 3%, and $10,815 if inflation rises to 5% in the third year. We recognize the need to help investors make these calculations more easily, and hope to make changes to our site in the future along these lines. However, we remain very strongly committed to the logic of expressing all our calculations in real terms, as that is what ultimately maintains an investor's purchasing power (and avoids disappointment) over time the long term.