(FORTUNE Magazine) – Fall often has an invigorating effect on biotechnology stocks. Clinical trial results pour out of the scientific and investment conferences that tend to bunch between September and January. The Street watches closely, to sort out the biotech firms whose drugs will succeed in the marketplace and those that won't make it to the market at all. Attention is particularly keen this year, as reforms at the Food and Drug Administration are speeding up the approval process. Already, the biotech sector is up 18% since mid-August, after a horrendous spring and lazy summer.

But don't let that jump mislead you, because something else has changed this year as well: Biotech no longer moves as one beast. Traditional leaders such as Amgen and Biogen have actually been suffering stock price declines, while stocks like Agouron Pharmaceuticals are soaring. Even after AutoImmune and Liposome Co. each recently plunged 75% because of product failures, other biotech names reacted mildly. "This is the first year biotech stocks are trading on their own merits," says UBS Securities analyst Tim Wilson. "And thank God."

For the individual, this means that while biotech looks attractive, figuring out how to distinguish among specific stocks is more important than ever. Clearly, this sector has always demanded a different kind of perspective. For starters, most of these companies don't yet have earnings. So how to assess their value?

To start answering that question, remember that, in effect, a bet on a biotech stock is less an investment in a company than in a drug. Or more precisely, it's a bet on whether--and when--a given chemical compound will finish making its way through all the tortuous phases of clinical trials, win final approval by the FDA, and thus gain permission to be sold as a drug. You also want a company to accomplish all this before a rival product makes it through.

The trial process is key. It goes on for years, and the final FDA review alone has taken an average of 18 months in the past. More recently, though, the FDA has been granting fast-track approval status to AIDS and cancer drugs. And now Congress is codifying this practice into law and expanding it to cover drugs that target any life-threatening disease. Senate and House bills, if signed into law, will also make it easier for companies to add new uses to existing drugs. The upshot is that companies with successful products can start making money faster.

The trial and approval process breaks down into phases, and Wall Street tends to make its bets as a drug crosses a threshold from one phase to another. At each step, the odds of a drug advancing are different, and understanding those odds can give an investor a leg up. Lehman Brothers analyst Tony Butler figures that fully half of all drugs in Phase II, for instance, don't make it any further. (The testing involved in this phase gives onlookers the first sense of a drug's effectiveness.) Of all those candidates that make it to Phase III, about a third will not come out alive. Finally, Phase III survivors have a 10% chance of failing the last hurdle, which is the approval of a new drug application (NDA) by the FDA.

So you should minimize your risk by waiting as long as possible to plunk your money down, right? Yes and no. Getting in earlier means more risk--but with each clinical milestone, shares also get more expensive. Consider one biotech company that the Street now loves: Agouron Pharmaceuticals. Its AIDS drug Viracept is an early beneficiary of the FDA's fast-track program, having gone from Phase I trials to approval in the unusually short time span of just three years. Available since March, Viracept already accounts for 27% of all prescriptions for HIV protease inhibitors, and in September, Agouron posted its first profit, 11 cents a share. It all sounds great--if you bought the stock a year ago. It's up 126% since then. But today, with a fiscal 1998 P/E hovering around 75, it looks, to put it mildly, fully priced.

So it pays to get in a bit early--if the product gets approved. If you happen to invest in one of the 10% of final-stage drugs that don't get past the FDA, you're going to get hammered. The tradeoff is simple: The odds of suffering a fall may be lower as a drug moves up the approval chain, but the size of the fall is likely to be greater if rising expectations aren't met.

How do you minimize the risks that investing in biotech inescapably entails? There are some fundamentals here. First, look for a drug that will address a large therapeutic need for which few or unsatisfactory treatments exist (such as cancer, AIDS, or Alzheimer's disease) so that the potential payoff will be significant. Another good sign: major drug companies that have signed on as partners (presumably a Pfizer or an Eli Lilly will feel pretty good about a biotech firm's science before spending millions for the future rights to a drug). Third, look at the company's cash situation, since biotech research burns up piles of it long before the business can even hope to earn a penny (compare the amount of cash the company has on hand with the amount it throws at research and development annually; anything less than two years' worth should raise an eyebrow). And last, try to assess how skilled management has been both at attracting capital (a canceled stock offering, for instance, would be a troubling sign) and at hitting its own target dates for getting its products closer to market.

We set out to find biotech companies that look likely to get a good drug approved but aren't quite there yet. Our picks fall into two groups: Companies that have been down this road before, having already brought at least one drug to market, and those that still have to prove they can. The first group is theoretically less risky, because the managements have proven that they can navigate the FDA process. Within the groups, the individual selections are ordered by how close their next drug is to the end of the road.

NEXT UP?

AIDS drugs are currently administered in a cocktail combination that almost always includes BioChem Pharma's 3TC. The drug, which has mild side effects, boosts the effectiveness of the other inhibitor drugs in the cocktail. Resistance to AIDS therapies is starting to crop up among some patients, but those who readjust their cocktail will probably switch one of the other inhibitors, not 3TC. As a result, sales of 3TC are expected to reach $770 million this year and approach $1 billion next year; BioChem splits these revenues with its partner Glaxo Wellcome.

Now BioChem is preparing to file an NDA for a hepatitis B drug, which, if accepted, could eventually become another blockbuster. (It's already filed for approval in China, where hepatitis is widespread.) In fact, BioChem's stock could be positioned similarly to Amgen's back in 1990. At that time, Amgen's Epogen (1996 sales: $1 billion) was already on the market, and Phase III trials for its new drug, Neupogen, were complete. Amgen's stock continued to climb steadily, and Neupogen became another billion-dollar drug.

As it happens, there's another lesson built into Amgen's story: Its stock is now stumbling because Epogen and Neupogen are proving tough acts to follow. "Amgen has two beautiful products that have had extraordinary earnings power. But what's next?" asks UBS's Wilson. His No. 1 pick is BioChem.

Centocor is the perfect Halloween stock: It's recently back from the dead. Just two years ago, it was digging a grave for itself at around $10 a share, even though its cardiovascular drug ReoPro had been approved as an anticlotting agent for high-risk angioplasty. Then an FDA panel acted favorably on extending the drug's uses, and Centocor's stock shot to $30 in two months. Today it's at $52.

While the company lost 19 cents a share last year, Lehman Brothers' Butler expects it to earn 50 cents a share this year and $1.18 in 1998. That would put its 1998 P/E at 44--a bit pricy, but the shares still may have room to grow. Why? First, final approval to use ReoPro for lower-risk angioplasties and unstable angina could double sales of the drug to $500 million by 1999.

Even more enticing, though, is Centocor's recent completion of Phase III trials for another drug. This one treats Crohn's disease, an extremely painful condition that can result in inflammation that sears holes through the bowel, allowing abdominal fluids to leak onto surrounding organs. The Phase III data show that in 62% of patients, Centocor's drug closes half or more of these holes. That may not sound so terrific, but when the alternative can be as extreme as partial removal of the stomach, it's easy to see why people think the drug has a good chance of being approved by next year.

Neurex already has one drug approved but probably won't be profitable unless its second one, now in Phase III, makes it too. The drug that's already on the market is an injectable that reduces high blood pressure quickly when oral medications cannot be given. Its use is limited to emergency-room and surgical settings.

The second product, a novel way to treat severe pain, has enormous potential; Warner-Lambert and medical-device giant Medtronic have corporate alliances with Neurex on this one. Today the most commonly prescribed drug for severe pain is still morphine, which dates from the early 19th century. Neurex's drug is a man-made version of a toxin found in sea snails that blocks a special pain-related neurotransmitter in humans. The drug is unique in that it seems to relieve pain caused from both tissue and nerve damage. (The latter is virtually untreatable today.)

The drug's biggest drawback is that in order to avoid serious side effects, in most cases it must be administered directly into the spine. Still, such treatment might allow patients in the late stages of cancer or AIDS, for instance, to live their last months relatively pain-free (using an implantable pump; hence the Medtronic alliance). And there's another potential use: to protect brain cells following serious head injuries. In this instance, the drug would be injected normally into the bloodstream. An NDA for the first use could be filed as early as mid-1998.

Meanwhile, Neurex has $75 million in cash, which should last it about 3 years at its current burn rate. Analyst Scott Sacane of Nationsbanc Montgomery Securities predicts the $17 stock will hit $40 in a year.

SOMETHING TO PROVE

What about companies winding through the approval process for the first time? They're riskier, but their stories are intriguing. IDEC Pharmaceuticals looks to be on the verge of becoming one of the next biotech powerhouses--and that point of view is reflected in the 66% rise in its stock since mid-August.

Still, it's reasonable to think there might be more to come. The company has filed for final-stage approval on a drug that treats cancer of the lymphatic system, which the FDA should make a decision on shortly. It also has a second drug that treats rheumatoid arthritis and a third that treats solid tumors, both in Phase II trials. Among IDEC's corporate partners: Genentech and SmithKline Beecham.

In the case of the first cancer drug, the company uses monoclonal antibodies to target B cells in the lymphatic and circulatory systems. The advantage, explains CEO William Rastetter: "Monoclonal antibodies can do what chemotherapy can't do--selectively destroy cells." Some patients with this type of cancer, called non-Hodgkin's B cell lymphoma, die directly from chemotherapy.

An advisory panel last July recommended that the FDA approve IDEC's drug, but that approval already seems reflected in the stock's current price. What may not be fully appreciated is that the drug's arrival on the market may open the door to off-label uses for other types of nonsolid tumor cancers, since the side effects are benign compared with alternatives. Says one biotech hedge fund manager: "I think docs will give patients a poke with the needle and see what happens."

Next is Cephalon, a comeback play. Last May its shares dropped from $20 to $13 in a single day when an FDA advisory panel concluded that the company did not have enough evidence to prove the effectiveness of its drug for Lou Gehrig's disease. Nevertheless, Montgomery's Sacane figures that there is still a 25% chance the FDA will approve the drug by mid-November. How come? Cephalon's problem is that only one of its Phase III trials was deemed statistically significant by the advisory panel. The other one wasn't, but it was trending in the right direction. The FDA normally requires statistically significant data from two pivotal trials. But given that the panel questioned the drug's effectiveness, not its safety, and that there is only one existing treatment for Lou Gehrig's disease, there's a chance that the FDA may bend in this instance. "Cephalon is a tough one," says Stuart Weisbrod, a money manager at hedge fund Oracle Partners, "but my gut is that it goes through."

At $13 there's not much downside in the stock. And even if the drug fails, the company still has $140 million in cash (or about $6 per share), which it is burning at a rate of $65 million a year. And Wall Street is giving Cephalon barely any credit at all for the narcolepsy drug that was filed as an NDA last December.

If Cephalon is a turnaround, COR Therapeutics is like trying to ride a rocket that's already taken off. The stock has charged from about $7 in April to $23 in mid-October. Still, COR is just now starting to pop up on the radar screen of most institutional investors, as it nears the $500 million market-cap level. The company (whose big-name partner is Schering-Plough) is about to submit an amended NDA for an anti-coagulant drug called Integrilin, which is similar to Centocor's ReoPro but targets a slightly different niche. Since Integrilin is less potent, it can be given in instances that are less severe--say, when a doctor wants to send an angina patient home after treatment. Integrilin should also cost significantly less than ReoPro. If the drug is approved--a decision is expected within six months--bulls say the stock could double again from present levels. Of course, if the product flops with the FDA, this rocket will come crashing back to earth. So unless you've got a high tolerance for pain, buy a little, not a lot.

Our last stock may take some patience. Texas Biotechnology is just starting to get attention among professional investors after a recent road show. An NDA has been filed for an anticoagulant to replace the drug heparin, which can create serious complications for a small percentage of patients. SmithKline Beecham signed a $32 million deal to distribute this drug, pending approval.

As in the case of Neurex, however, it's Texas Biotech's second compound that's really exciting. Although still in Phase II trials, this drug for congestive heart failure addresses a billion-dollar market. Lehman's Butler says the stock can easily double. While his firm is one of the company's underwriters, it was Butler himself who helped get Lehman involved after he came across a Ph.D. thesis by a cardiologist at Sweden's Karolinska Institute. The thesis compared Texas Biotech's compound with several others from major drug companies and concluded that it was at least as effective, if not more so. "It was independent information that I just found," says Butler. He figures the compound will attract a major drug company as a partner, and an actual product might be on the market by 2001.

INSIDE: What's next for Southeast Asia? page 304...One small-cap manager's quest for value, page 306...A modest new derivative, page 308...Best PC deals, page 311