Market Timing: Should You Attempt It?

» Kirk - Washington Post: "It doesn't work"

The Trouble With 'Timing': It Doesn't Work

By James K. Glassman

Sunday, November 11, 2001; Page H01

In the stock market (as in much of life), the beginning of wisdom is admitting your ignorance. One of the many things you cannot know about stocks is exactly when they will go up or go down. Over the long term, stocks generally rise at a nice pace. History shows they double in value, on average, every seven years or so. But in the short term, stocks are just plain wild. Over periods of days, weeks and months, no one has any idea what they will do. Still, nearly all investors think they're smart enough to divine such short-term movements. This hubris frequently gets them into trouble.

For example, many investors, believing the economy would collapse after the Sept. 11 terror attacks, sold their shares at the first opportunity. The Investment Company Institute reported that investors pulled $29.5 billion out of stock mutual funds in September -- the largest one-month withdrawal in history.

General Electric fell below $30 a share, Oracle dropped to $10, the Dow Jones industrial average skidded to 8236 and the tech-heavy Nasdaq composite index fell to 1423. But within seven weeks, GE had risen above $40, Oracle broke $16, the Dow was above 9600 and the Nasdaq topped 1800. Many investors who sold their stocks in late September in anticipation of hard times ahead got a rude surprise.

But, as surprises in the market go, it was a pretty common one. The strategy that these investors were employing -- trying to make money in stocks by predicting their short-term moves -- is called "market timing." It doesn't work. Period. John Bogle, founder of the Vanguard Group, one of the world's largest mutual fund houses, once wrote: "After nearly fifty years in the business, I do not know of anybody who has done it successfully and consistently. I do not even know anybody who knows anybody who has done it successfully and consistently."

The reason for the failure of market timing can be summed up in two words: "random walk." The phrase, made popular 30 years ago by Burton Malkiel, a Princeton economist, describes the pattern that stock prices take in the short term. It's random: You can't guess it; no one can. Malkiel's notion was that today's stock price is determined by everything that millions of buyers and sellers of stocks know about those stocks today. Tomorrow's price is unknowable today since it is determined by tomorrow's events. So, from today's perspective, tomorrow's price looks random. For example, investors who want to sell because they see an economic downturn ahead don't seem to realize they are not alone. Other investors, also seeing such a downturn, have already priced stocks for that negative event. The stock price is "discounted" (as the jargon goes) to take the expected hard times into account.

This idea is at the heart of what is called "efficient market theory," which, taken to its logical extreme, concludes that every stock is perfectly priced; that is, the market doesn't make mistakes. Clearly, the market does make mistakes. Investors get hyper-excited about something and bid it up into the stratosphere, or they get overly depressed about something and drive it deep into the ground. But, in general, an individual investor should have a healthy respect for the daily price that all the other investors in the world set for a stock.

At any rate, you don't have to believe in perfect markets to believe you can't guess stock prices from one day to the next. It's impossible to overemphasize this basic truth about investors' own ignorance.

If I sound fanatical, it is only because I have just read an article in Barron's headlined "The Truth About Timing." The article cites a new study by Birinyi Associates of Westport, Conn., that shows that investors who are out of the market during the five worst days each year do astronomically better than those who hold their stocks throughout each year.

For example, in 2000, the Standard & Poor's 500-stock index dropped 10 percent (not including dividends), but an investor who invested in the S&P on every day but the five worst days of 2000 made a profit of 9 percent. In 1998, being out of the market during the five worst days would have given you a return of 56 percent instead of 27 percent. Overall, a dollar invested in the S&P stocks in 1966 became $12 for the buy-and-hold investor (again, without dividends) but an incredible $987 for the investor who missed the five worst days each year.

This kind of analysis is not new. It's no secret that the market moves in spurts, both up and down. If you miss the five best days each year, your returns are vastly reduced. Stocks are very, very volatile in the short run. But the real question is what to make of this information. You might conclude, as Barron's does in a headline, that the key is "knowing when to stay out." But this conclusion raises an important question: How can you know when to stay out? You can't.

Barron's does its readers a disservice by implying that you can.

Say you are betting your brother on coin flips. You take tails, your brother takes heads. A total of 100 percent of your winnings will occur on roughly 50 percent of the flips. But does that knowledge help you make a good guess before you toss the coin? Of course not.

The Barron's article quotes Bob Brinker, a market-timing advocate, as saying that the Birinyi study confirms his own approach. Brinker told his readers to switch their portfolios in January 2000 from 100 percent stocks to just 40 percent stocks. "I think the probabilities are we've entered a secular bear market," he said.

That was a good call, no doubt. But Brinker missed the huge run-up since Sept. 21, and, more important, his long-term record is nothing spectacular. According to the Hulbert Financial Digest, for the 10 years ending Dec. 31, 2000, Brinker's long-term growth portfolio rose an annual average of 13.7 percent while the S&P rose at 15.1 percent. I won't deny that Brinker is good compared with other timers, but why go to the trouble of jumping in and out of the market if you can't beat the averages?

[EC: Brinker is wrong about the Birinyi study confirming his report. That study showed how missing the worst 5 days of the year in a market would give you higher returns... Brinker has NEVER said to get in or out for only a day... and the 5 days could be spread over a full year. Brinker is about long term market timing but that is another issue...]

The thrust of the Barron's piece is that the trick to making money in markets is to get out before downturns. Are some timers particularly brilliant as bad-news bears? The Hulbert newsletter, the authoritative record-keeper, recently looked at the 10 best-performing newsletters (nearly all of which employ timing of one sort or another) during October 1987, a month in which the market lost 2 1/2 times as much as it did in September 2001. Mark Hulbert, the editor, then asked: How have these market timers done since? The answer: Rotten.

If you had followed these newsletters, moving in and out of stocks and mutual funds as they suggested, you would have achieved an average annual return of just 4.5 percent between November 1987 and September 2001. "You could have made more than that just by investing in 90-day Treasury bills," wrote Hulbert. The market as a whole returned 12.9 percent -- nearly triple the rate of the timers.

What is the alternative to timing? It is finding good stocks and mutual funds at reasonable prices and buying them with the intention of holding them for a long time -- meaning five years or more. Sell only if something has happened to the underlying company (its management, products, competitive position), not to its stock price. And never sell because you think you're the one market-timing genius that John Bogle didn't happen to come across in his 50 years' experience.

While market timing is bunk, there are better and worse times to be buying individual stocks. Last month, for example, Merrill Lynch's David Anders reported that gaming stocks "appear to be on sale." In a careful analysis, he showed that the cash that casino companies are likely to be earning over the next few years -- if things return to normal -- justifies considerably higher stock prices. The companies he specifically mentioned -- including Harrah's Entertainment, Mandalay Resort Group, MGM Mirage and Park Place Entertainment -- had been driven down sharply by the attacks of Sept. 11. They have bounced back since the Anders report but remain below his target prices.

I cite gaming stocks not because they're guaranteed winners but because they are the kind of potential bargains you should spend your time and energy pursuing. Forget about timing the market; concentrate on buying good companies.

James K. Glassman invites comments at jglassman@aei.org, but he cannot answer all queries. Of the stocks mentioned in this column, he owns General Electric.

© 2001 The Washington Post Company

-- posted by Kirk

» SteveT - Re: Washington Post: "It doesn't work"

-- posted by SteveT

» mdorsey - Can it be done? Look at the results.

December 31, 2001
Vanguard 500 Index (VFINX) 
YTD total return  -12.02% .
Since 12/31/99 return  -20.00%

Vanguard Prime MM Inst  (VMRXX) 
YTD total return  4.37% . performance vs 500 Index +16.39%
Since 12/31/99 return  9.79%…..…………………………+29.79%

Vanguard GNMA (VFIIX)
YTD total return 7.94%.  performance vs 500 Index +19.96% 
Since 12/31/99 return  20.13%…………………………+40.13%

Vanguard Total Bond Mkt Index Inst (VBTIX)
YTD total return 8.45%. performance vs 500 Index +20.47%
Since 12/31/99 return 20.95%…………………………+40.95%
Vanguard Short-Term Bond Index (VBISX)
YTD total return 8.88%. performance vs 500 Index +20.90%
Since 12/31/99 return 20.09%…………………………+40.09%

Wether in MM or bonds, timers had a great relative performance. Will it be three in a row? My expectation is that some time this year you will be able to reenter stocks at levels at least 10%-20% lower than 12/31/01 prices and earn MM or bond returns till that time. OCICBW