Performance Guides: The Longer the Track Record, The Sharper the Image

    by Mark Hulbert

    The bear market that began in March 2000 may have been devastating to many investors’ portfolios. But it did have at least one silver lining: It provided the ultimate test of whether performance-monitoring systems can separate advisers with genuine ability from those whose past returns are attributable to nothing more than luck.

    This silver lining exists because, during bull markets, even advisers with no ability can look like they know what they are doing. This is the origin of the sage wisdom on Wall Street not to confuse brains with a bull market.

    For example, when the stock market was performing as well as it was in the 1990s, more than a handful of advisers were producing regular annual returns in the high double digits. Not all of them were geniuses. But how was an investor to know which ones were truly worth following?

    These thoughts prompted me to re-examine the Hulbert Financial Digest’s (HFD) database of investment newsletter returns in light of the bear market that began more than four years ago. The good news: Newsletter track records did a creditable job of identifying those with genuine ability.

    Nevertheless, I found that, in order for past performance to do this creditable job, it had to be measured over the very long term. Shorter-term track records, on the whole, were a disappointing basis on which to choose.

    Setting Up the Test

    I constructed this test of the HFD database in a way that almost appears rigged against it. I imagined a hypothetical investor who, near the stock market’s all-time high on March 31, 2000, was picking which newsletter he would follow.

    There were only two grounds rules:

    • He had to follow the selected newsletter at least until April 30, 2004, and
    • The only information he could use in making his selection would be the HFD’s performance database (which, as of the March 31, 2000, selection date, contained nearly 20 years’ worth of track records).

    Short-Term Track Records

    Consider first how this hypothetical investor would have fared had a newsletter been chosen on the basis of short-term performance—say, the 12 months prior, that is, from March 31, 1999, to March 31, 2000.

    As fate would have it, that method of selection would have been disastrous, since the newsletters that did the best over the 12 months prior to the stock market’s top were among the biggest losers in the subsequent bear market.

    How would this investor have fared over the next four years if this selection method had been used?

    There are several ways of reporting this. One way is to calculate the average return between March 31, 2000, and April 30, 2004, of the five newsletters that were top-ranked over the 12 months prior to the market’s top. Over this four-year-plus time period, these five erstwhile high-fliers produced an average annualized loss of 12.0%, more than double the 5.3% annualized loss of the broad market (as measured by the Wilshire 5000).

    Would five-year records have done a better job than one-year records of identifying advisers who would navigate the bear market?

    Only barely.

    TABLE 1. Newsletters That Beat the Market: 6/30/80 to 4/30/04
    Newsletter Annualized
    The Prudent Speculator (949/497-7657) 18.86% 0.15
    The Value Line Investment Survey (800/634-3583) 15.52% 0.17
    NoLoad Fund*X (800/763-8639) 14.98% 0.17
    Wilshire 5000 Index 12.89% 0.14

    The average return between March 31, 2000, and April 30, 2004, of the five newsletters with the best five-year returns prior to the market’s top was an annualized loss of 7.2%. That’s better than the 12.0% annualized loss realized by those with the best one-year track records, but still worse than the Wilshire 5000’s return.

    Figure 1.
    Track Records

    Long-Term Track Records

    How would our hypothetical investor have done if he had chosen a newsletter based on performance over longer periods?

    Here, the situation begins to improve. He would have done progressively better as the time period lengthened (Figure 1).

    Notice from Figure 1 that the newsletters chosen on the basis of 10-year returns more or less equaled the Wilshire’s return, and that choosing on the basis of 15- and 20-year returns did even better. In fact, the five newsletters with the best returns between mid-1980 and March 31, 2000, produced an average gain of 1.8% from March 31, 2000, to April 30, 2004.

    Of course, averaging the subsequent returns of the top five performers is not the only way to measure the efficacy of a performance-tracking system. But other measures led to similar conclusions. For example, the same pattern emerged when I measured the degree to which a ranking of all newsletters over each time period was correlated with a ranking of those same newsletters over the period between March 31, 2000, and April 30, 2004.


    The bear market storm that began in 2000 brought us some good news. If you focus on performance over the very long term, it is possible to at least begin to identify those few advisers with genuine ability.

    With that cheerful conclusion in mind, Table 1 reports those newsletters monitored by the Hulbert Financial Digest that have beaten the Wilshire 5000 over the nearly 24 years the HFD has been tracking newsletters.

    Mark Hulbert is editor of the Hulbert Financial Digest, a newsletter that ranks the performance of investment advisory newsletters. It is published monthly and is located at 5051B Backlick Rd., Annandale, Va. 22003; 703/750-9060; This column appears quarterly and is copyrighted by HFD and AAII.

→ Mark Hulbert