Ferri: Technical Analysis Hurts Returns

Even Buffett conceded that technical analysis doesn't work.

Trying to predict price trends by studying charts is a popular investment technique. In fact, technical analysis is probably the most common form of securities selection because anyone can do it without much training. Predicting that prices will rise because they have been rising is an easy decision. Avoiding securities that have trended down is another easy decision.

For many investors, technical analysis goes well beyond projecting future prices from the latest trend. Ardent followers attempt to identify trading opportunities by using moving averages, relative strength, patterns such as head and shoulders or top/bottom reversals, lines of support, resistance, channels, flags, pennants, cup and handle patterns, and the list goes on.

The study of price trends is popular among individual investors and professionals, but does it lead to the right buy and sell decisions? That’s a tough question to answer.

Surprisingly little research has been conducted on the actual results from technical analysis. The existing literature essentially ignores the real-world performance and instead focuses on the theoretical effectiveness over select time periods when abnormal profits have been earned.

Until recently, the only study attempting to measure the trading records of investors who used technical analysis dates back to 1980. In “Portfolio Design and Portfolio Performance: The Individual Investor,” researchers Wilbur G. Lewellen, Ronald C. Lease and Gary G. Schlarbaum matched transaction records to investor survey responses from the period 1964-1970. Their findings suggested that technical analysis severely degraded the performance of individual investors’ portfolios.

Researchers Brad Barber and Terrance Odean published several papers beginning in the late 1990s that document the actual performance of thousands of individual investors who traded their own accounts. They found that those who traded more underperformed than those who traded less.

Undoubtedly, some of those investors studied by Barber and Odean relied on technical analysis, and it could be inferred that the investors who traded more than others may have used it. However, the focus of their studies were on overall results rather than how investors made decisions; thus, there is no hard evidence that technical analysis hurt or helped the investors.

Recently, two researchers from the Netherlands published a paper that specifically measures the results of trading by investors who rely on technical analysis. In Technical Analysis and Individual Investors, Arvid Hoffmann of Maastricht University and Hersh Shefrin of Santa Clara University analyzed real-world trading records of Dutch investors for the period 2000-2006. These investors were users of technical analysis.

In 2006, the brokerage firm supplying the data administered a survey among all of its clients. The survey included a question asking participants to select which strategies they use as a basis for his/her investment decisions. Technical analysis was one of the selections they could choose. About 32% of these investors stated they used technical analysis exclusively or in conjunction with fundamental analysis. Hoffmann and Shefrin obtained a sample of 5,500 clients and corresponding accounts for which both transaction and survey data are available.

Overall, the results from statistical tests indicate that individual investors who use technical analysis to make investment decisions are disproportionately prone to speculate on short-term stock-market trends, hold more concentrated portfolios, turn over those securities at a higher rate than people who do not use charts, and earn lower returns.

Hoffmann and Shefrin found the investors using technical analysis had lower performance, on average, approximately 50 basis points (0.5%) per month in raw returns from portfolio selection decisions and 20 basis points (0.2%) from additional transaction costs. Adding these together, the marginal cost in investor returns due to technical trading was 70 basis points per month — approximately 8.4% per year.

I admit to looking at a chart or two over my career. I’ll even admit to believing at one time that the strategy works. Technical analysis is a compelling idea that tends to consume us because it comes across like finding a treasure map that leads to a chest full of gold. Even Warren Buffett admits to spending years trying to master the technique.

Yet, for all the promises that technical analysis brings, studying charts has actually detracted from the bottom line. Buffett reportedly joked about this dilemma with an audience at Vanderbilt University in 2005, “I realized that technical analysis didn’t work when I turned the chart upside down and didn’t get a different answer.”