May 10, 2013, 3:43 p.m. EDT
Investors can’t beat the machines
Commentary: Computer-dominated trading takes over
By Mark Hulbert, MarketWatch
It has always been difficult for investors to consistently beat index funds. It has been nearly impossible lately.
And there’s a double whammy: The small number of advisers who outperform the market rarely can keep doing so.
One big culprit, experts say: the rise of sophisticated computer trading programs.
Consider the 51 advisers out of more than 200 on the Hulbert Financial Digest’s list who beat the market in the decade-long period that ended April 30, 2012, as measured by the Wilshire 5000 Total Market index, including reinvested dividends.
Of that group, just 11 — or 22% — have outperformed the overall market since then. On average over the last year, they have lagged the Wilshire index by 6.2 percentage points.
That’s no better than the percentage that applies to all advisers, regardless of past performance. In other words, going with a recent market beater doesn’t increase your odds of future success.
“Before the era of computer-dominated trading, it was slightly easier to identify winning advisers in advance, because you could more easily understand and evaluate what they were doing,” says Lawrence G. Tint, chairman of Quantal, a risk-management firm for institutional investors, and former U.S. CEO of Barclays Global Investors.
One major reason why machines are winning is our inability to process lots of financial data, which is getting more complex and voluminous every year.
Terrance Odean, a finance professor at the University of California, Berkeley, has extensively studied the behavior and performance of individual traders. He points out that there used to be another human being on the other side of the trade when an individual bought or sold a stock. “Now it’s a supercomputer you’re competing with,” says Odean.
“Individuals are no longer playing against Grandmasters; they’re playing against Deep Blue,” he says, referring to the famous battle in the 1990s between chess’s Grandmasters and International Business Machines’ (NYSE:IBM) supercomputer Deep Blue. Individual investors “will almost certainly lose.”
Another reason traders are losing out to machines is their general inability to assess complex data. They look at the same set of facts on different occasions and reach different conclusions, and they unwittingly let their emotions dominate their intellect.
Daniel Kahneman, professor emeritus of psychology and public affairs at Princeton University and the 2002 Nobel laureate in economics, has widely studied this phenomenon. In his 2011 book “Thinking Fast and Slow,” he reviewed more than 200 academic studies over the past five decades that analyzed head-to-head contests between human beings and mechanical algorithms.
Kahneman reports that man consistently loses out to machine in a wide variety of pursuits, ranging from medicine to economics, business, psychology and even things like predicting the winners of U.S. football games and judging the quality of Bordeaux wine. In each of these domains, he reports, “the accuracy of experts was matched or exceeded by a simple algorithm.”
Betting on the pros
Some traders hold out the hope that they can beat the market by following the lead of an investment adviser. But it is close to impossible to identify these advisers in advance, according to Tint.
“The average reader of The Wall Street Journal simply won’t be able to identify these market-beating advisers,” he says. After all, “repeated studies have shown that even the best institutional investors have been unable to identify them in advance.”
Tint adds that there is an above-average chance that an awful adviser will continue to perform terribly. This creates the mathematical illusion that there also is persistence among high-ranking managers and that we can beat an index fund by following one of those top performers, he argues. But all it really tells us is that it’s a good idea to avoid a terrible adviser.
Regulators Urged to Beef Up Insider Trading Rules
Former SEC commissioner has urged the government to beef up rules that regulate corporate executives’ stock sales.
This persistence at the bottom of the rankings is well-illustrated by the adviser on the Hulbert Financial Digest’s monitored list who, one year ago, was at the very bottom for trailing 10-year performance: Charlie Buck’s Situational Strategies. Sure enough, it has been a bottom performer in the 12 months since then, falling 33% vs. a 17% gain for the overall stock market. The newsletter’s publishers didn’t respond to requests for comment.
There’s another reason why it is so hard for top-performing advisers to beat the index over the long term, says Tint, even when their numbers were powered by genuine ability rather than sheer luck.
Once the adviser turns in impressive performance, lots of new money flocks to his fund, diluting the ability to continue performing well.
That phenomenon appears to be what contributed to the downfall of legendary fund manager Bill Miller of Legg Mason Value Trust. At the end of 2005, Miller had one of the hottest hands in U.S. mutual-fund history, beating the Standard & Poor’s 500-stock index for each of the previous 15 years. His fund attracted lots of new money, and he found it impossible to continue his remarkable record.
From 2006 to 2011, he lagged the market in all but one year, and in 2012 he resigned as manager of that fund.
Miller, in an interview, says it is “mathematically true” that there is a portfolio size “beyond which it is difficult, if not impossible, to beat the market.” Assets under management in Legg Mason Value Trust grew markedly over the years.
The fund now run by Miller, called Legg Mason Opportunity Trust, has less than 10% of the assets under management that his prior fund did at its peak, and was one of the best-performing mutual fund last year, with a return of 40%.
Miller says he thinks the primary cause of his hot hand turning cool was simply “bad decision making,” and that, in addition to good decision making on his part, luck played a big role in his fund being ranked so highly last year.
What about Warren Buffett, chairman of Berkshire Hathaway, who has beaten the market by a large margin over the past four decades? Isn’t he an exception?
It is certainly possible that he has been more skillful than his competitors. But with a portfolio that is now so huge, Buffett will have a more difficult time in the future picking stocks that will perform better than an index fund, Miller says. Buffett himself has said that he expects Berkshire’s future returns to be only slightly better than the S&P 500’s.
Tint says he believes much of the value that Berkshire has added in recent years has derived not from Buffett’s stock-picking skills but from the huge amount of cash at his disposal and his negotiating skills, which combine to give him an enormous advantage in managing his company and extracting very favorable terms from those who need that cash and cannot easily get it elsewhere.
The trading trap
The implication of all this for individual investors is straightforward: Don’t trade. Short-term trading has become so dominated by Wall Street’s computers that individuals—and professional managers—almost certainly will lose out to them over time. The obvious alternative, experts say, is to buy and hold diversified index funds with very low expenses.
The portfolio that is most widely diversified, of course, reflects all publicly traded stocks, both in the U.S. and abroad. One exchange-traded fund that provides such total-market exposure is the iShares MSCI ACWI Index Fund (NASDAQ:ACWI) , which is benchmarked to MSCI’s All-Country World Index. The ETF has an expense ratio of 0.34%, or $34 per $1,000 traded.
For U.S. equities, one low-cost way to get total-market exposure is through the Vanguard Total Stock Market ETF (NAR:VTI) , which has an annual expense ratio of just 0.05%, or $5 per $10,000 invested. Almost as diversified is a fund that mimics the S&P 500. A low-cost vehicle for exposure to that index is the iShares Core S&P ETF (NAR:IVV) , with an expense ratio of 0.07%, or $7 per $10,000 invested.
If you’re interested in developed countries’ stocks, a low-cost choice is the iShares MSCI EAFE ETF (NAR:EFA) , which is benchmarked to MSCI’s Europe Australasia and Far-East index. Its expense ratio is 0.34%, or $34 per $10,000 invested.
The advice to trade as little as possible and be diversified also applies to fixed-income investing, since bond investors are at a disadvantage against machines in this arena as well. For fixed-income exposure in the U.S., a low-cost option is the Vanguard Total Bond Market ETF (NAR:BND) , with an expense ratio of 0.1%, or $1 per $10,000 invested.
For exposure to international bonds, be on the lookout for a fund that Vanguard says it is close to launching: The Vanguard Total International Bond Fund, which the firm has indicated will have an expense ratio of 0.2%—which, if so, would be among the lowest in the category.
Low-cost index funds are also an obvious choice for getting diversified exposure to other asset classes. iShares offers an ETF that is benchmarked to the Goldman Sachs Commodity Index: the iShares S&P GSCI Commodity-Indexed Trust, with an expense ratio of 0.75%, or $75 per $10,000 invested. To invest just in gold, one popular ETF is the iShares Gold Trust (NAR:GLD) , with an expense ratio of 0.25%, or $25 per $10,000 invested.
A new partnership
Is there still a role for man in a world where he so consistently loses to machines?
Yes, according to Brad Barber, a finance professor at the University of California, Davis, who has extensively studied performance and behavior of individual traders. There are some things that computers either can’t do or can’t do well, such as determining whether one of the myriad patterns that emerge from data crunching makes sense.
“If you don’t understand the reason for a pattern, you’re vulnerable to following a mindless algorithm that is quite likely to perform poorly,” Barber says.
Computers are also ill-suited to thinking outside the box and devising new hypotheses and models of what might be able to beat the market in the future, Tint points out. He envisions a man-and-machine partnership in which we use computers to rigorously test our hypotheses and trade on those that survive statistical muster.
Of course, most people don’t have the computer hardware and extensive databases required to take advantage of what computers have to offer.
And even professionals who do have access to such resources need to first recognize the limitations of their decision-making abilities. Until they know what they’re not well suited to do, they are likely to perform poorly — even if they have powerful computers at their disposal. That’s because they are likely to exaggerate what they bring to the table and play down the role that computers can play.
Odean says some of the poorest performances in his studies were turned in by traders who were the most confident of their abilities. This led them to trade even more often and incur even more risk.
This often leads them to do precisely the wrong thing. A series of academic studies over the past decade compared stocks that traders buy with those that they sell, both in the U.S. as well as in some foreign countries.
The average stock traders sell goes on to outperform the average stock they buy, he says.