Buy and hold isn't dead.

Oh, I know you've read a lot of obituaries for this investment strategy. And it's certainly true that quite a few buy-and-hold investors have been badly scorched by the end of the decade-long bull market that hit in January 2000 for the

Dow Industrials

and in March 2000 for the

Nasdaq Composite

.

But I'm just not ready to go into mourning yet for this venerable investing strategy (for

TSC's

recent analysis of buy-and-hold investing, see this

special report). I think much of the damage suffered by buy-and-hold portfolios over the past year has in fact been self-inflicted by investors who forgot, while the bull market was running, that buy-and-hold must include rules for selling. And I'm not convinced that the economy has yet changed so much that investors can't still identify a set of companies that deserve long-term loyalty.

In this column, I'll explain why I think most investors' buy-and-hold strategies would benefit from some attention to the discipline of selling, however. And I'll take a look at what this bear market has taught me about competitive advantage for the long haul.

The buy-and-hold obituaries I've seen recently usually list two causes of death. First, that buy-and-hold investing leaves investors helplessly exposed to market downturns that can take away, in a worst-case scenario, more than half of an investor's wealth. Second, that the New Economy moves so fast now that it's difficult, if not impossible, to find companies with a long-term competitive advantage that would make them worth holding for years.

Let's take these arguments one at a time.

Even Buy-and-Hold Investors Sell Sometime

I'm not going to pretend that even the 981% gain recorded by the buy-and-hold Nasdaq investor during the 10 years that ended March 10, 2000, makes anyone indifferent to the 66% plunge from the March high to the April 4, 2001, low. Nobody wants to take this kind of punishment, and nobody in their right mind would follow an investment strategy that didn't give them at least a chance to avoid the damage from a downturn like this. If buy-and-hold strategies said never sell, I'd say this was a fatal flaw.

But only someone interested in debunking buy-and-hold at any cost would actually maintain that a buy-and-hold investor

never

sells. The buy-and-hold investor buys

Cisco Systems

(CSCO) - Get Cisco Systems, Inc. Report

in 1990 and dies 40 years later, leaving the heirs to pry the shares out of cold, clenched fingers? Come on. Buy-and-hold investing has always assumed that investors would sell -- just not very frequently and not in reaction to minor market volatility. Buy-and-hold should properly be called "buy and hold but occasionally sell."

Buy-and-hold-but-occasionally-sell investing should, I'd argue, include two rules for selling. Most buy-and-hold investors would agree with the first of my two rules; it is, after all, a cliche mouthed by those who follow the strategy. A buy-and-hold investor sells a stock when the reasons behind the original buy are no longer convincing.

As I wrote in my

March 16 column, this involves a simple review of why you bought the stock in the first place. If, for example, you bought shares because the company dominated its market

then

, how dominant is the company

now

? Is the lead of a technology leader still as strong or has its edge over rivals significantly decayed? Has the market for the company's products zigged when you expected it to zag, so that customers increasingly prefer new products from competitors instead of the aging bestsellers this company is still pushing? (For more detail on how to conduct this kind of review, see my

earlier column.) This type of sell discipline would have led long-term investors to sell stocks such as

Lucent

(LU)

,

AT&T

(T) - Get AT&T Inc. Report

TheStreet Recommends

and

WorldCom

(WCOM)

before the worst of the recent plunge. The fundamental business reasons for buying those three stocks changed so radically that it was clearly time to step aside from the shares for a while.

But I don't think this kind of fundamental review offers the average buy-and-hold investor enough protection by itself. A buy-and-hold strategy has to include a second selling rule, one based on valuation.

I know this is likely to strike some buy-and-hold investors as heresy, since it smacks of market timing. So be it. I can't think of any other way for the buy-and-hold investor to avoid a significant part of the punishment delivered by one of the stock market's infrequent but extended corrections or bear markets. When a market collapses because stocks have become too expensive, the kind of fundamental review I proposed above won't take an investor to the sidelines quickly enough to prevent major losses. In a valuation-driven correction, fundamentals are likely to be a lagging indicator of declining stock prices, rather than a predictor.

I don't recommend that any buy-and-hold investor try to time the market as a whole. Getting reliable buy and sell signals for the entire market requires mastering one of a number of complex systems, each one of which requires constant study and tweaking. That's an effort that takes buy-and-hold investors away from their focus on the quality of individual companies.

Instead, I'd suggest playing to that focus by studying the valuation of individual companies. Stocks, even the stocks of great companies, run in cycles between undervaluation (times when they're cheap) and overvaluation (times when they're expensive). A buy-and-hold investor would love to buy near the bottom of that cycle and should try to sell when overvaluation reaches an extreme.

There are many ways to do this. Previously, I've suggested charting a stock's

price-to-earnings ratio against that of the market as a whole, as represented by the

S&P 500

index. When a stock gets near the top of its historic range -- when it becomes as expensive in comparison to the rest of the market as it has at other peaks -- it's time to sell. That system doesn't work for stocks with only short histories, however.

I think you can get roughly the same result from charts. Let me use Cisco as an example. Look at an annual price performance chart for Cisco over the past decade. The years beginning in 1991 and stretching through 1999 were clearly a great period for buy-and-hold investors in this stock and culminated in two years of spectacular returns in 1998 (150%) and 1999 (129%), before the bottom fell out in 2000 with a 28% loss.

Cisco's Annual Price Performance
1991-99 were great years for buy-and-holders

Source: Big Charts

Now switch over to a 10-year price history chart to get another view of the stock's performance. To get a smoother long-term trend, go to the analysis tab on the chart and draw in the 200-day moving average. Study the way the actual price history line for the stock climbs roughly in pace with the 200-day moving average until August 1999, when the black daily price line begins to rapidly diverge from the 200-day moving average. That gap continues to widen until the Nasdaq peaked in March 2000.

A buy-and-hold investor looking at this chart would have received a pretty strong signal that valuations at Cisco had climbed to dangerous levels. Combining that alert with some study of the fundamentals would at some point during 2000 have told even the confirmed buy-and-hold investor to sell Cisco Systems. (Buy-and-hold investors comfortable with technical analysis could improve this signal: For example, as my colleague Jon Markman noted to me, a rule that said sell Cisco when the stock spends three to six days below its 200-day moving average would have resulted in a sell on Cisco at around $60 in September 2000, and at no other time in its history.)

Cisco's Daily Price History vs. 200-Day Moving Average
The gap created when the two lines diverge signaled valuations had risen to dangerous heights

Source: Big Charts

My "pictorial" alert isn't a very sophisticated sell indicator, and I don't think it would be particularly useful if you were trying to trade Cisco. I doubt, for example, that the peak in February 1994 would have been enough to send a clear sell signal, so that a short-term investor could have avoided the slump in the shares from April 1994 to October 1994. But, in conjunction with fundamentals, it is good enough for the long-term buy-and-hold investor who really doesn't want to generate sell signals more than once every five or 10 years. The point isn't to avoid all the dips, just the most punishing collapses.

Competitive Advantage and the New Economy

The bear market of the past year or so has sent a lot of great companies reeling. But it's important not to make sweeping generalizations about the destructive power of the New Economy out of what is really just a normal example of how U.S. capitalism works out the demand-supply imbalances generated by an overabundance of cheap capital. It's certainly far too early to say that the New Economy is peculiarly hostile to a company maintaining a long-term competitive advantage.

For example, I think you can make a good argument that profit margins at

EMC

(EMC)

will be lower in the future than they've been over the past five years because the high profit margins of the past five years have attracted new competitors to the storage market. But that's the way capitalism has always worked. Excess profits attract competitors.

That's very different from saying that in the future EMC won't dominate the storage market because its installed base, its massive and aggressive sales force and its economies of scale are no longer important competitive advantages. I haven't seen anything during the bear market in technology that suggests that some other company has been able to negate these advantages.

Our feeling that competitive advantage doesn't last very long anymore is attributable, I think, to recent mistakes in deciding that some relatively untested companies had managed to achieve a lasting edge over their competitors. Two years ago, for instance, many of us thought

Amazon

(AMZN) - Get Amazon.com, Inc. Report

had established such an edge in retailing that it would simply roll up as much of the sector as it wanted.

I'd argue that Amazon has failed to do that not because some character of the New Economy attacked and eroded that competitive advantage, but because that competitive advantage was a lot more potential than actual. And in the testing, Amazon has, so far, failed to demonstrate that it has achieved any competitive advantage over well-run traditional retailers.

Wal-Mart's

(WMT) - Get Walmart Inc. Report

competitive advantage seems stronger than ever.

It's simply a lot harder to build a sustainable competitive advantage than it seemed to be a year ago -- and a lot harder than it seemed then for investors to accurately identify it in an immature company.

At the time of publication, Jim Jubak owned or controlled shares in EMC.