Exploiting Skill: Newsletters

We can find and exploit skill from investment newsletters even better than mutual funds.

I read The Hulbert Financial Digest to learn what Mark Hulbert has discovered. Mark Hulbert has just completed 25 years of reporting the facts. In the July 2005 issue, he tells us how well the original newsletters have done.

He acknowledges the effects of survivorship bias, which has generally been upward. Then, he tells us what has happened. About one in five has beat the market. He makes it clear that skill exists and that past performance is the best way to identify skilled investors. He cautions that past performance does not guarantee future success.

Mark Hulbert wrote:

“Let me hasten to add that, with 25 years of data for just 13 letters, generalizations must be tentative at best. But the fraction of letters beating a buy-and-hold is close to the 20% that I have consistently found over numerous other time periods. So I am tempted to believe that this 1-in-5 proportion is more than just a fluke.”
…
“But I believe we can do better than that when picking a newsletter. As I have discussed at greater length in several recent issues, past performance can be a creditable guide to future performance when it is measured over a long enough period of time – even if, as the SEC requires all advisers to stress, past performance isn’t a guarantee of future returns.”

Mark Hulbert went on to compare newsletters with mutual funds. (Lipper provided him with data on the 257 funds that were suitable.) Hulbert showed that the range of performance of newsletters is much broader than that of mutual funds. Most often, the returns over 25 years for both groups were between 10% and 15% (nominal, annualized).

Only 5% of the mutual funds returned more than 15% (annualized). The Wilshire 5000 index (with dividends reinvested) returned 12.8% (annualized) over the full 25 years. That is, only 5% of the mutual funds beat the index by 2.2% over the entire 25-year period. In contrast, Mark Hulbert has found that 20% of all newsletters do this well.

[Note: the performance of an index fund differs from that of the index that it seeks to replicate. An index fund has expenses, typically small, and a tracking error, which can be positive as well as negative. The full name of the index is the Dow Jones Wilshire 5000 Value Weighted Total Return Index.]

What is especially interesting is that we can exploit these newsletter results much better than those of mutual funds.

The portfolio values of investment newsletters fluctuate just like everything else. Individual newsletters go through dry spells. This includes the very best newsletters. This is similar to the best mutual funds. The key is to maintain price discipline. Once you have identified writers with skill, you can set price levels for purchasing their portfolios. Wait for 2 to 5 years. Buy only on a deep dip, close to the level of the index.

What is better is that you can get all of the necessary information. You can choose your period for identifying skill. For example, if you wanted to use a period of 10 years, you could go back 10 years or, even better, 12 to 15 years. That way, your selection would be based on what happened 2-5 years ago, which would be suitable for making your purchases today.

So what period should you choose for selecting newsletters? Based on the February 2005 issue of The Hulbert Financial Digest, 10 years is about right. It does as well as 15 years and it is much, much better than 5 years. (Nine years is another good choice.)

I caution you to read Bull Street Bull. We are (highly likely to be) in a secular (long lasting) bear market. What has worked well in an extended, roaring bull market may not work well in the future. Before 2000, the best long-term decision was to be 100% in stocks. Low cost, stock index funds were ideal. Being invested in 100% stocks today is dangerous.
Bull Market Bull

[Index funds generally do well in a bull market but not in a bear market. The typical newsletter has performed better than the Wilshire 5000 index from 2000 through the end of 2003 and it is doing so today. There was about a year during which the index outperformed 50% to 60% of the newsletters.]

A skilled newsletter writer who has performed well during the extended bull market will not necessarily do well in this market. Yet, such a writer has shown ability. You should not dismiss that fact lightly. If your decide to use 10 years, your selection period will include both the top of the bubble and the beginning of the bear market.

Have fun.

John Walter Russell
July 14, 2005

NOTE: My only relationship with Mark Hulbert is that I buy The Hulbert Financial Digest. I receive it as email.