Without trying to brush all stock analysts with the same brush, I really have seen some very illogical analysis from some of this group in the past year.

After a year of trying to keep up with valuations of technology companies, many analysts threw in the towel and started using a real estate valuation device as this year began. If a similar company had a certain market value, then the new company with similar characteristics was assigned a similar value.

If Akamai had a market value of $20 billion, why should Digital Island, which does some similar things but is a little behind in revenues, not be valued at $10 billion. The fact that revenues are less than $100 million and that cash is being exhausted to service current levels of business should not be a concern. With such high market valuations, more cash should be readily forthcoming.

This was the time when the "new paradigm" required new economic models.

As an economist, I cringed at this thinking. After all, an investor is buying a future earnings stream when it purchases shares in a company. If the revenue growth path is strong, costs can be contained, and competition can be restrained by patents or falling costs per unit sold as companies grow (a natural monopoly); then high valuations for companies currently burning cash might be justified.

Yet, few analysts cited those factors when they projected high future stock values for technology companies. Furthermore, they failed to question where additional cash would be acquired if stock values began to fall. Indeed, some acted as if future earnings had nothing to do with stock values.

The merger of a management service company with a web enabling service company was viewed as transforming the assets into high powered revenue growth. This is why some analysts, in March, recommended Marchfirst when it was selling at $52 per share. Even after the first wave of value revisions hit the technology sector in the spring, several analysts liked the stock at $38.

No one doubted that the exhaustion of cash could be managed in the combined company. It had a market presence that was enviable even as it developed a cash drain that overwhelmed the balance sheet. Now, this stock is selling under $2 because investors are not sure whether a cash infusion can be arranged to keep the company floating until it controls costs. Why was this current real worry not even considered when the stock was recommended in July?

Instead of three to five year projections that were based upon a realistic assessment of market shares, growth of the markets, and management of costs, analysts cited importance in a "space" or "first mover".

I will grant that a first mover can be beneficial if there is a natural monopoly (costs per dollar earned continuing to fall as the firm grows up to a very large size). However, very few technology companies fit that model. How many people own a Dumont, the first mover in the production of commercial TV.

When technology is changing so rapidly, one wonders how long a natural monopoly can persist.

I must admit I was fooled by the branding of web sites. I thought that customers would be reluctant to learn new web addresses because of information overload. Therefore, a site establishing a presence, such as Amazon, would have economic benefits because of the knowledge of its name. What I failed to realize is the speed by which users dumped sites that did not perform and sought out new sites that might do better.

As a result, the value of branding was only temporary in most cases (and much less than the advertising costs needed to establish those brands).

I could continue with the failure of many analysts to recognize the large percentage of all internet advertising that was used for site branding, the impact that the failure of branding had upon internet marketing analysis, the degree to which e-commerce was enhanced by equity financed subsidies last Christmas (causing no growth in e-commerce during the first half of this year when the equity based subsidies vanished), and a host of other problems that were ignored by many analysts.

In case you think that analysts are humbler but wiser, I cite the analyst who realized that Micron Technology would only earn about $2 next year instead of the $3.70 that previously was expected. Despite this earnings downgrade, the investment rating was not changed and neither was the expected price target.

Now, if the earnings shortfall was the result of charges needed to streamline costs, then a temporary drop in earnings could still justify the same value of future earnings. Indeed, when all the natural gas producers were forced to write off impaired assets two years ago, they were positioned for spectacular price gains if natural gas prices rose. I can cite several natural gas companies whose stock prices have more than tripled this year because their balance sheet now understates their economic power.

That was not the case with this Micron Technology report, which was issued less than two weeks ago. Again, not all analysts have performed so poorly. But for some, the old refrain "when will they ever learn" is very appropriate.