exp system

[exp]

Can Market-Timing Ever Work?
http://poweredby.morningstar.com/PoweredBy/doc/article/1,3020,83565,00.html?CN=WSC789
by David Kathman / 11-29-2002

Dear Analyst,

In a recent article, Pat Dorsey lambasted the idea of timing the market.

Under the umbrella of timing, is he including such well-known approaches as William O'Neil's of Investor's Business Daily? It seems to me that there's a lot of solid research behind his analysis, plus years of successful results.

Mark H.

I assume you're talking about Pat Dorsey's recent column entitled "Three Ways to Lose Your Shirt in the Stock Market", although Pat also wrote a column earlier this year straightforwardly titled "Market-Timing is Bunk". As that second title makes clear, Pat doesn't have a very high opinion of attempts to time the market, and I tend to agree with him.

O'Neil has a system for stock investing, which he promotes in the newsletter Investor's Business Daily (of which he is the founder) and in his book How to Make Money in Stocks. Market timing is only one part of that system, but it's an important enough part to raise red flags. This is not a system that I would recommend for long-term investors.

CANSLIM, or Can't It?
O'Neil's system, known as CANSLIM, is actually a mishmash of different strategies, some elements of which make sense to me. For example, he advocates investing in companies with new products, markets, or management (the "N" in CANSLIM). That's not a bad thing to look for, in combination with other criteria.

Most of the CANSLIM system, however, is geared toward finding stocks with rapid earnings growth and lots of price momentum. That's a recipe for buying very expensive stocks, something we tend to frown on at Morningstar. Such a strategy can work fairly well during a roaring bull market like the one of the late 1990s, but it's extremely risky, to say the least. Lots of people lost their shirts buying go-go momentum stocks during the market bubble.

But CANSLIM is not really a long-term investment strategy anyway; it's more of a short-term trading strategy, which is where the market-timing--and its kissing cousin, technical analysis--come in. O'Neil recommends buying and selling stocks according to a variety of technical indicators, both for individual stocks and for the market as a whole. He places great importance on "cup-with-a-handle" chart patterns, and goes on at length about "pivot points" and "double-bottom price patterns." Sorry, but call me skeptical. While I'm willing to grant that broad forms of technical analysis can be marginally useful over the very short term (measured in days or weeks), I have yet to see any good evidence that the type of intricate chart-reading advocated by O'Neil works consistently.

Apart from the technical mumbo-jumbo, O'Neil advocates selling any stock if it declines 8% or more, in order to limit your losses. In theory, it's a very good idea to have a sell strategy in place. But such a low threshold is likely to lead to lots of short-term buying and selling, especially in a market as volatile as the one we're in now. All that trading increases transaction costs, which tend to drag down returns and increase your tax burden.

The Seductive Danger of Data-Mining
But what about all the research behind the CANSLIM system? Most of the evidence O'Neil presents in his book is anecdotal ("This person gained 279% in one year using our method!"), but he does describe the study out of which his system grew. O'Neil took the stocks with the greatest percentage gains each year since 1953 and searched for common characteristics, which he codified into CANSLIM.

This may sound impressive at first glance, but it's almost a textbook example of data-mining, or looking for patterns in past returns in an attempt to predict the future. The problem is that such attempts virtually never work for the stock market. Most often the patterns are just the result of random variation rather than correlation, and they don't continue going forward. If a genuine anomaly is discovered, such as the January effect, the market soon eliminates the inefficiency after it gets publicized.


As Exhibit A, I present the New USA Growth Fund, a mutual fund created in 1992 explicitly to follow O'Neil's methodology. It was managed by David Ryan, who had used CANSLIM to win an investing contest in the 1980s with three years of triple-digit returns. The fund was a different story: New USA Growth underperformed the S&P 500 in three of its four full years of existence, before it was merged into MFS Emerging Growth MFEGX in 1997. From its inception on April 29, 1992 until May 30, 1997, the fund had a cumulative return of 89.7%, versus 105.9% for the S&P 500 over the same period. And it achieved this underperformance with above-average volatility and a turnover rate that peaked above 500% in 1996--not a combination for the faint-hearted. On top of all this, the fund had a high expense ratio (higher than 2%) and a 5% load.

The recent history of investing is littered with other systems that were supposed to beat the market (usually based on what would have worked in the past) but which did not pan out so well in practice. James O'Shaughnessy's book What Works On Wall Street mined decades of stock-market returns to find the characteristics of the best-performing stocks, but the mutual funds based his research posted uninspiring returns. (Although, to be fair, those funds have done better since O'Shaughnessy sold them to Edward J. Hennessy.) The Motley Fool touted its Foolish Four as a surefire market-beating strategy, but then, after further research, admitted that the strategy's seemingly great past performance was largely the result of data-mining.

The one lesson to take away from all this is that no mechanical strategy is going to help you consistently beat the market. Investing well takes hard work, skill, and patience; trying to predict where the market will go is a fool's errand.

David Kathman does not own shares in any of the stocks mentioned above.


Market-Timing Is Bunk
If you think otherwise, I'd be glad to sell you a bridge.
http://news.morningstar.com/doc/article/0,,13767,00.html
by Pat Dorsey / 03-27-02

I will never understand why some people insist in making investing more complicated than it is. All this talk of resistance levels, market internals (yuck!), and descending flags drives me nuts--but none of it makes me as annoyed as suggestions that market-timing is consistently feasible. (Why am I bringing this up? Because when I suggested that "market-timing is bunk," on Fox News' Bulls & Bears last weekend, I was roundly booed by two other guests on the show. So, my dander is up.)


Folks, this is one of the all-time great myths of investing. There is no strategy that consistently tells you when to be in the market and when to be out of it. None. Anyone who tells you otherwise has something to sell you--usually a market-timing service.

No Way, No Day
Don't believe me? Listen to someone who's been around the block a lot longer than I have--Vanguard's Jack Bogle: "After nearly 50 years in this business I do not know of anybody who has done it successfully and consistently. I don't even know anybody who knows anybody who has done it successfully and consistently."

Now, the timers out there are saying to themselves, "Yeah, but Bogle is Mr. Index Funds--he has a vested interest in a buy-and-hold strategy!" Well, not necessarily--after all, a lot of market-timers use modified index funds from shops like Rydex and ProFunds to do their little dances in and out of the market. Bogle is, however, a big fan of controlling what you can (costs and taxes) and not worrying too much about what you can't control (how the market will do from month to month or year to year).

Moreover, when's the last time you heard someone advocating market timing who wasn't trying to sell you a proprietary system or service to get you out at the top and in at the bottom? (Wow, it's quiet in here… )

Another question: If the systems being touted are so darn good, why are the creators wasting their time trying to get a few hundred bucks in subscriptions here and there from easy marks, rather than leveraging themselves to the hilt based on their own signals? Logically, if you discovered a surefire way to time the market, the wealth-maximizing strategy would be to A) keep it quiet so that whatever inefficiency you'd discovered doesn't get arbitraged away, and B) leverage yourself as much as possible on the way up and the way down so that you could retire in style to Bora-Bora.

The Results Are In
If you still think the stock market can be timed--and judging by the number of dodgy services that pop up after a Google search on "market-timing," there must be plenty of people who think it's possible--consider some of the following studies:

An interesting piece in the February 2001 issue of Financial Analysts' Journal studied the difference between buy-and-hold and market-timing strategies from 1926 through 1999 using a very elegant method. Without getting into the gory details, what the authors did was to essentially map all of the possible market-timing variations between 1926 and 1999, with different switching frequencies. They assumed that for any given month, an investor could either be in T-bills or in stocks, and then calculated the returns that would have resulted from all of the possible combinations of those switches. (For the curious, there are 2-to-the-12th-power--or 4,096--possible combinations between two assets over 12 months.) Then they compared the results of a buy-and-hold strategy with all of the possible market-timing strategies to see what percentage of the timing combinations produced a return greater than simply buying and holding.

The answer? About one third of the possible monthly market-timing combinations beat the buy-and-hold strategy. When the authors looked at quarterly switching over five-year periods, the results got even worse for the timers: Only one fourth of the timing strategies beat buy-and-hold strategies. Annual results were even more grim: One fifth of annual timing strategies beat buy-and-hold strategies.

Now, some of you are probably thinking, "Heck, I have a 33% chance of beating the market if I try to time it? I'll take those odds!" But before you run out and subscribe to some timing service, consider the following.

First, the results in the paper I cited above overstate the benefits of timing, because they looked at each year as a discrete period--which means they ignore the benefits of compounding. (As long as you assume that the market will generally rise over long periods of time, that is.) To some extent, this is also demonstrated in the fact that buy-and-hold strategies outperform timing strategies as the length of time between switches--month, quarter, or year--increases.

Second, consider that stock-market returns are highly skewed--that is, the bulk of the returns (positive and negative) from any given year come from relatively few days in that year. This means that the risk of not being in the market is also quite high for anyone looking to build wealth over a long period of time. Barron's ran an interesting piece last fall (citing data from Birinyi Associates) that highlighted this issue quite well. According to the article, $1,000 invested in the S&P 500 in 1966 through late October 2001 would have grown to $11,171. If an investor had been lucky enough to have missed the five worst days of each year, that $1,000 would have grown to $987,120, but if they'd missed the best five days of each year, the $1,000 would have shrunk to $150. No, I'm not making this up.

I've read some articles by timers who claim that this study vindicates what they're attempting to do--after all, isn't the possibility of turning $1,000 into almost a million worth the risk of losing 85% of your initial investment? Maybe it is if you're playing the lottery, but not if you're planning on sending your kids to college or retiring someday.

Put another way, which of the following two options would you take: a high-probability expectation of 7% to 8% annual returns for some period of years, or a low-probability bet between achieving 30% annual returns and cumulatively losing most of your money? Doesn't seem that hard to me.

The Bottom Line
We invest today so that we can achieve some financial goal in the future. With intelligent diversification, reasonable return expectations, and prudent saving and spending habits, anyone can realize a fairly high probability of reaching most reasonable financial goals. Maybe those goals won't be as lavish as you'd like, but at least you'll have a strong degree of confidence in where your finances will be in 10 or 20 years. By trying to time the market, you are essentially gambling with your financial future at very poor odds.

Etc.
Hubris alert: General Electric GE CEO Jeffrey Immelt appeared on CNBC last week after PIMCO's Bill Gross made some scathing comments about GE's disclosure practices. When asked whether GE will be able to continue growing at double-digit rates, he responded: "Think of the lunacy of that question."

You know, I don't think it's a loony question at all. I think it's a damn good one that any potential investor in GE ought to be asking.


(My take: excellent arguments against market timing as a successful investing strategy. Yet, I believe that it is possible to successfully market time investments/trades but with two caveats:
(1) relatively few will greatly succeed at it so it is not practical for most investors/traders
(2) it requires a system/approach that is based on some original insight/principle which is not popularized/revealed to avoid having it arbitraged away)