How to lie with statistics: Investment performance charts (Part 1 of 2)
Darrell Huff wrote a short and very informative book, "How to Lie with Statistics," which was first published in 1954 and was amusingly illustrated by Irving Geis. This book is still in print and remains very popular (Amazon book rank #2,040 in June 2007). It plainly and humorously discusses how statistics can be distorted and misused to serve the self-interest of the presenter.
Historical ETF and mutual fund investment performance charts are a case in point. While the numbers they present might be historically accurate, their presentation in advertising, on line, and in printed materials can amount to lies from several perspectives. ETF and mutual fund performance charts are designed to lure gullible individual investors with an implied promise that superior past performance will continue. The financial research literature tells us clearly that on average this is a promise that cannot be kept. In other words, historical fund performance charts are a veiled lie. They may report factual information, but their purpose is to deceive.
These lies or deceptions include:
1) SELECTING ONLY "WINNERS" TO PROMOTE.
When selling to you, securities industry sales people and the fund companies that advertise performance select only those historical investment fund performance charts that show superior historical performance. The industry sells its ETF and mutual fund winners, and it ignores or hides its losers. Charts for their loser funds are available, but ETF and mutual fund sales representatives are not eager to present them. You have to dig them out yourself on the web. Or, these inferior or average performance charts will be mailed to you AFTER you have bought what you thought was a "superior" fund, but, gosh, things just did not remain superior.Except for very, very poor historical performance, which tends to be an indicator of excessive costs, the financial research literature tells us that historical mutual fund performance is meaningless. The industry knows that many investors naively project past fund performance into the future. Yet the scientific finance literature simply does not support such assumptions. If investing were this easy, then those who buy ETFs and mutual funds based on past performance would be consistent winners in the future and would grow relatively richer and richer. The opposite turns out to be true. [See articles in these categories on The Skilled Investor website: Luck versus Skill (4 articles) and Selecting Investment Funds (16 articles) ]
For your amusement when you are being sold to by a securities industry sales person, ask to see an asset-weighted chart that combines the entire historical performance of all the funds for a mutual fund family. Good luck in getting to see that one! I could list a dozen reasons why you will be told that such a fund family chart does not exist. However, the real reason is that this aggregate historical performance chart would likely show that the entire fund family trails a very broad market index by almost as much as the fund family charges in fees.
I use the word "almost," because professionally managed mutual funds have shown a slightly positive ability to pick individual securities. Unfortunately, this slightly positive gross returns advantage is more than wiped out by mutual fund management fees and transactions costs, which are several times greater than this small gross returns gain. Then, of course, there are the mutual fund sales loads and 12b-1 marketing fees and the percent-of-assets management fees that you pay to your broker or investment advisor. In return, your broker and investment advisor will do you a dis-service by only pushing selected funds with "superior" performance charts and higher costs. These sales and asset management fees just drag your returns down even more year after year after year. (See this article on The Skilled Investor website: "Pay less to get more" and this category of 15 articles: Controlling Investment Costs )
2) EASY INDEX BENCHMARKING.
Historical investment performance charts will compare a particular fund's performance against some market index benchmark. A market index is a market index, isn't it? The question that individual investors should ask is whether the market index benchmark really is appropriate. All index benchmarks are not the same, and there can be very significant differences between market index benchmarks -- even when indexes seem to match the particular investment style of the ETF or mutual fund in question.
When you look at a performance chart, do you investigate whether the fund company picked a challenging index or an easy hurdle that they could more easily stumble over? For more about the variations between index benchmarks, see Craig L. Israelsen's article, "Variance Among Indexes: Don't judge an index by its title" in the May/June 2007 issue of the Journal of Indexes (Pages 26 to 29) Dr. Israelsen analyzes the various indexes published by the six major U.S. index providers (Standard & Poors, Frank Russell, MSCI, Morningstar, Lipper, and Dow Jones). He finds very wide performance variations even with market indexes that supposedly represent the same "style" of investing.
Dr. Israelsen concluded his article by commenting: "It is important to recognize that significant performance differentials among prominent indexes can lead to misleading conclusions about mutual fund performance. Funds with mediocre performance histories can be made to look better by being compared to a prominent benchmark with a weaker performance history. At the very least, the industry needs to recognize the existence of potentially sizable performance differentials among various U.S. equity indexes, and therefore view performance comparisons between Mutual Fund A and Index B for what they are: marketing materials."