June 10, 1998
Buy-and-Hold Beats Rapid Trading
Individual investors can beat the market -- but only if they avoid frequent trading. That's the conclusion of an important new study by Brad M. Barber and Terrance Odean, professors at the graduate school of management at the University of California at Davis. Unfortunately, the study also found that the average household turns over 80% of its common stock portfolio each year, in line with the 77% average annual turnover of mutual funds. Such excessive trading helps explain why both individual investors and mutual funds generally underperform the market. The lesson is clear: Investors who think of themselves as actual owners of a business are far more likely to generate enviable long-term returns than are the active traders who try to time the market by rapidly moving in and out of stocks.
Barber and Odean studied the trading activity of 78,000 households that used one large discount broker between February 1991 and December 1996. They found that the average household scored a 17.7% annualized gross return during this period. However, they then factored in reasonable transaction costs, such as average round-trip commissions running at 5% of the principal invested and a spread between the bid and ask prices amounting to 1% of an investment. After such costs, the average net annualized return fell to just 15.3%. By contrast, a value-weighted index of stocks on the New York Stock Exchange, Nasdaq, and the American Exchange (a proxy for "the market") delivered a gross annualized return of 17.1% during this period. For context, the Vanguard fund designed to mimic the performance of the S&P 500 delivered a 16.9% annualized return over this period.
These figures do not account for taxes, which could claim up to a third of the gains for those generating mostly short-term profits in taxable accounts. Since the Vanguard index only trades to mirror McGraw-Hill's changes in the S&P index, its turnover typically amounts to less than 20% a year, minimizing short-term capital gains and thus maximizing long-term return.
Furthermore, the study found that individual investors have a preference for smaller, higher-risk (high beta) stocks as well as a slight bias toward value stocks (low price-to-book ratios). Beta is dubious proxy for risk. Still, if one assumes it's useful, the risk-adjusted net returns of the average investor underperformed the market index by 4 percentage points annually. So for the study period, an investor could have done better with an index fund even while enduring far lower risks to capture that performance.
While 51.8% of investors produced average gross returns that beat the market, only 46% outperformed after expenses. In typical bell curve fashion, 25% of the households turned in net outperformance averaging 7 percentage points a year while 25% turned in net underperformance of 9 percentage points a year. The crucial difference in performance was not so much stock selection as turnover. And transaction costs weren't the only problem. As Odean notes in a related paper, investors sell profitable investments twice as often as unprofitable ones. That is, they tend to hold on to their losers and sell their winners. This proves a faulty strategy because the stocks they sell subsequently do better than the ones they buy.
The penalty for frequent trading is striking. The study breaks down the households into quintiles based on how frequently they turn over their portfolios. The low turnover group averaged just 1.44% turnover per year, meaning they rarely traded out of stocks. The high turnover group sported a 283% average annual turnover rate (115% minimum); investors in this group basically swapped out their entire portfolios three times a year. Gross returns in the highest two turnover groups, as adjusted for their generally higher risk stock selections, underperformed the market by 3 to 4 percentage points annually. The average portfolio in the highest turnover group delivered just a 10% annualized net return while the group with the lowest turnover scored a market-beating 17.5% net return on average. Adjusted for risk, the net return of the high turnover group lagged the market by 10.9 percentage points per year.
Factor in taxes and the frequent trading would look even worse given the preferential treatment of long-term capital gains. This is a major concern because the study found that investors trade more frequently in their taxable accounts (96% annual turnover) than they do in their tax-deferred accounts (72% turnover), exactly the opposite of one might expect if investors were acting rationally to maximize net after-tax returns.
The study is particularly valuable because it examines a huge group of actual investors, folks with typical holdings for people relatively early in their investment lives. For example, the average household in the study owned four stocks worth $41,344, while the median household owned 2.5 stocks worth $14,579. That mix makes sense given the demographics of typical discount brokers during the study period. It also seems likely to reflect a fairly large block of the folks reading this column.
There are some important caveats. First, the transaction costs used in the study to figure net returns have both been reduced substantially in the past 18 months. In January of '97, Nasdaq introduced new trading rules that have helped slice the spread between the bid and the ask prices on Nasdaq stocks, in some cases by a third or more. Also, commission charges for online trading have plunged to as low as $8 a trade from costs of $30 or more just a few years ago. And the most active traders have flocked to online brokers, cutting the cost of frequent trading.
On the other hand, the Barber-Odean study covered a period when small-cap stocks favored by individual investors outperformed large-caps, a phenomenon that no doubt accounted for part of the gross outperformance of individual investors relative the market index. Just the opposite has been true of late, as large-caps have rallied to new highs. Also, the widening of the tax-privileged status for long-term capital gains adds a further reason why a low portfolio turnover should produce superior net after-tax returns today.
A recent Wall Street Journal article drew out two interesting corollaries to this study. First, stock turnover in general has soared in the past few years. In 1997, 69% of NYSE shares changed hands versus just 46% in 1990. In part due to the rise of day-traders who use the Small Order Execution System (SOES), turnover of Nasdaq shares was 199% last year, double the rate seen in 1990. Morningstar also reports that turnover by mutual funds has jumped 16% since 1993.
Second, small-cap growth and value funds, curiously enough, have managed to do better by trading more frequently. That apparent anomaly may suggest that while individual investors may not have an information edge regarding these stocks (though they apparently believe they do), mutual funds actually might. Morningstar's Don Phillips told the Journal that faster trading in and out of these stocks may allow funds to take advantage of good news while avoiding some of the pain when bad news crushes a stock. Of course, there's no escaping the tax effects of such high turnover.
The bottom line is that the Barber-Odean study makes a very strong case for taking a buy-and-hold approach. As the Fool has said repeatedly, people should invest in businesses rather than speculate on stock movements. If you're not mentally prepared to do that, you should really just stick your money in an index fund.