Editor's Note: On Wednesday, biotech columnist Gabe Hoffman discussed how biotech stocks trade around "reality points," the dates when actual news is released. Today he shares how this insight can help time purchases of specific stocks. Click here for Part 2.

Knowing the Difference

Unfortunately, too many people think that you should buy any biotechnology stock after a company issues a press release reporting results of a clinical trial that the company considers positive. How to tell the difference?

You've got to do your homework beforehand to have an idea of when the data were fully expected, and when they are unexpected, and what constitutes "great data." That way, when the data are released, you can almost immediately determine if they are truly robust and meaningful.

What I mean by meaningful data is not results from some preclinical model or animal tests, but strong human data from a tightly controlled, well-designed, reasonably large-scale clinical trial.

For example,

this notice from

Immune Response

(IMNR)

is not meaningful data. Neither is

this notice from

Advanced Tissue Sciences

(ATIS) - Get Report

. If you read either press release and figured that it was good because the stock was up sharply that day, you got burned badly. Actually, our fund shorted both stocks on the days of the respective press releases and covered a few days later.

It's also important to have done enough research on the market for the product to be sure that the commercial opportunity for the product is substantial.

This assumes, by the way, that you feel comfortable analyzing the quality of clinical data and market research on your own. Part of what this column will try to do is to help you understand how to invest in biotechnology stocks. But it can't all be explained in even a year of columns.

So if you can evaluate this data -- and many people can -- that's great. I run into a great many people who can't do this for themselves but invest in individual biotech stocks anyway. In my mind, that's gambling, not investing. So if you don't feel experienced enough to do this kind of spadework on your own, my advice would be to participate in the biotech sector through a good mutual fund.

Other Uses of This Insight

The theory of perception vs. reality is also applicable to small companies that develop and manufacture devices or diagnostics. If a company's prospects depend on a single product, the price movements can be nearly as severe as with the biotechs. The difference here is that the reality points are primarily quarterly sales because FDA approval is usually not a risk in the devices or diagnostics business.

The theory of perception vs. reality can apply even to specific products within large pharmaceutical companies -- note the hype in 1998 surrounding

Pfizer's

(PFE) - Get Report

Viagra and

Eli Lilly's

(LLY) - Get Report

Evista, both of which later disappointed. The impact on the share price of a major drug company is less drastic, though definitely meaningful.

The key for investors in biotechnology, small device or diagnostics companies is to know what your stock's reality points are and around what time they are expected to occur. If you can't stand an instantaneous decline of 50% to 70% or more on the stock price, haven't really done your homework on a company's lead product, or just don't feel comfortable playing amateur scientist, then you usually can be safe selling before the reality point. Always be aware of where the stock is in this cycle, and don't think a biotech stock is attractive because it is going up -- because it might be going up entirely on perception, not reality.

Gabe Hoffman is the biotech and pharmaceuticals analyst for New York City-based Welch Capital Partners, LLC, an institutional money manager with $250 million in assets. Prior to that, he was at Paramount Capital, a biotech merchant banking firm and hedge fund manager. Under no circumstances does the information in this column represent a recommendation to buy or sell stocks. Hoffman appreciates your feedback at

ghoffman@thestreet.com .