Mistakes were made. Now, let's just try not to repeat them.

On the way to


5000 and beyond, a lot of people made a lot of errors as they tried to figure out what stocks were worth.

Sell-side analysts, investors and, yes, even reporters goofed up with optimistic assumptions, dubious yardsticks and shaky comparisons. In 20/20 hindsight, the methods people employed to price tech stocks were rationalizations at best and ridiculous at worst. But kicking the nasty habit will be hard.

In other words, investors who forget the past are condemned to repeat it. But those who remember may repeat it anyway.

So where did we all go wrong? Let's take a look down memory lane to document some common valuation mistakes, and try to figure out whether we're still walking down the same path.


Comparable valuations

: In its simplest form, it goes like this: If Company X is worth $Y, then Company P must be worth $Q. If the market is raising one tech firm at a certain multiple of revenue or earnings, another tech firm should be valued similarly.

The problem here is that the benchmark company may turn out to be wildly overvalued itself. Look, for example, at

a March 2000, article on




(CNET) - Get Report



. The author of that article suggested that ZDNet, then with a market cap of $2.3 billion, didn't deserve to be trading at such a discount to CNet, then worth $4.7 billion. But a year later, CNet, having since absorbed ZDnet, is worth $1.5 billion, less than ZDNet alone one year earlier. Makes one wonder about the value of that original comparison.


Extended time horizons

: Valuing a company based on the present value of future cash flow is a respectable, time-tested,

Graham and Dodd tool for figuring out the value of a company. You make some projections about how much money a company will generate over five or 10 years, then work your way back to the present and figure out how much a stock should be trading for, given the expected future cash flow.

But with the commercialized Internet so young and so volatile, looking out a decade would be enough to scare off

Jeane Dixon, were she still alive. When


analyst Kelly Flynn initiated coverage on



at the close of 1999, part of the valuation was based on annual revenue estimates going out to 2010. Given that HotJobs.com was only about 2 1/2 years old at the time, forecasting that far out seems awfully brave. In fact, the first year of the forecast was way off, with the $41.8 million estimate turning out to be less than half of HotJobs.com's $96.5 million in actual sales. Despite the outperformance, Hotjobs.com stock is down about 90% from back then. Go figure.


Miscounted discount rate

Going out 10 years for discounted cash-flow analysis isn't necessarily wrong, says Jeffrey Hooke, author of the book

Security Analysis on Wall Street.

More of a problem, he says, is the number that people plug into discounted cash-flow models to reflect the cost of capital and the risk of the investment. A company with a track record, say,

Procter & Gamble

(PG) - Get Report

deserves a discount rate of 12% or 13%, Hooke says; for venture capitalists, the rule of thumb is a 100% discount rate.

Over the past few years, he says, analysts and others talking about tech stocks have consistently low-balled the discount rate. And that low-balling makes a big difference, he says, using the example of a hypothetical company he assumes will be worth $30 a share in five years. With a discount rate of 25% -- standard in Net stock sell-side research literature -- the stock is worth $10 today, Hooke calculates. But at the 100% VC rate, the present value is $1. Split the difference and use a 50% discount rate, and you get a $4 present value. Given the risks in the tech market over the past few years, the discount rate should be closer to the VC rate than people admit, he says.


Overstated market potential

: Shrinking market size is one of those variables that contribute to the above-mentioned risk. From 30,000 feet up, it's easy to say something like the offline market for Product A is $B billion, but if this online business snares just C% of that market, it'll have revenue of $D million in five years.

Sounds good, but it understates various execution risks: marketing effectiveness, market size, pricing pressure and competition. When

Janney Montgomery Scott

, for example, initiated coverage on



two years ago, analyst Tomas Isakowitz wrote, "

Forrester Research

estimates that Internet Advertising will grow at a 50% compound annual growth rate (CAGR) over the next five years. ... We estimate DoubleClick's segment of the market to comprise 40% of total Internet Advertising, and to grow, therefore from $480 million in 1998 to $5 billion in 2003."

Unfortunately, once Internet advertising stops growing at a 50% CAGR -- right now, apparently -- the rest of the pieces don't fall into place. DoubleClick, which was trading at more than $40 at the time, is down below $14.


Overvalued market share

: In the tech run-up, losses weren't a danger sign: they were the cost of staking a claim in a business, building a brand, say, or winning

first-mover advantage. For example, when

Gruntal & Co.

initiated coverage on



in August 1999, analyst Anthony Vendetti called drugstore.com's position as the first online pharmacy with a wide selection a "crucial competitive advantage." Perhaps it was, but not as big an advantage as people thought. The online druggist, then trading at about $50, is now below $1.50.


Focusing on the top line, not the bottom


Price-to-earnings ratios have been a traditional tool for valuing stocks. But if you have no earnings, you can't have a P/E ratio. That led people to focus on revenue instead. That's a reasonable choice for a start-up -- but it can gloss over the cost of that revenue, and how long it might take for it to turn into profits.

In an October 1999 research report on




First Union Securities

analyst Charles Wittman put a price target of $85 on the stock, reflecting a 63x multiple of expected 2000 revenue. That made sense in light of similar ratios from




Commerce One


. (Return to the first point on this list for the hazards of comparable valuations.) VerticalNet's stock is about a tenth of where it was then, or about one-and-quarter times projected 2001 revenue.


Home-grown yardsticks

: In addition to income-statement measurements, people also tried to judge companies on operational ratios, such as, among Internet stocks, price per page view and price per user. They didn't have earnings, but something had to fill the explanation vacuum for billion-dollar stocks. In October 1999, writing about


-- now part of



-- and the e-consulting industry, analyst Susan Lacerra of


could find only one factor that correlated to stock value: buzz. "Stocks that are well known and well publicized are also being afforded the highest market capitalization," she wrote, adding, "We think stocks with stories that are easy to tell and understand have an advantage in this environment."

That environment didn't last, of course. USWeb/CKS stock, worth $39 then, is worth about a dollar in postmerger MarchFIRST stock.

So, we've learned from all our mistakes, right?

Maybe not. Just like a party-hearty college student waking up with a hangover and a promise of "Whoa! I'll never do


again!," we're still skating on the edge. Just look at a Jan. 11 report from

Robertson Stephens


FuelCell Energy

(FCEL) - Get Report

. The fuel cell company has development-stage technology, losses for the foreseeable future, serious competition and a need to raise additional funding. But, hey, it's got a big potential market, it's reasonably valued in comparison to similar companies. And if you perform a discounted earnings-per-share analysis -- relying, admittedly, on a bunch of variable assumptions, including a discount rate of 20% -- the stock looked like a buying opportunity at around $64.31. (It closed Thursday at $45.62.)

Of course, doubters risk missing out on the chance to buy into the next


(MSFT) - Get Report

. But, as indicated by the tech cave-in, they also can avoid plenty of headaches.

TheStreet.com has a revenue-sharing relationship with Amazon.com under which it receives a portion of the revenue from Amazon purchases by customers directed there from TSC.