Big potential gains may appear lucrative, but don't forget that all-important word "risk." By Andrew Feinberg, Contributing Columnist March 31, 2007 Most great investors have a remarkable ability to assess risk and reward. They may not be as fanatical as Warren Buffett in shunning risk (Buffett says, "Rule number one: Never lose money.Rule number two: Never forget rule number one"), but they do treat risk with profound respect. Investors who trail the averages usually focus much more on the reward side of the equation -- a bias that can be disastrous. Here are a couple of cases that underscore my point. Biotech disasters December was a bloody month for biotech stocks. One victim was Northfield Laboratories, a single-product company that has developed PolyHeme, a blood substitute. On December 19, the company announced that PolyHeme had failed in a late-stage (phase III) trial. The next day, the stock fell 58%. On December 26, Telik announced that its cancer drug, Telcyta, one of two products it was developing, failed in three late-stage trials. That day the stock plunged 71%. I would argue that anyone who had conducted a thorough risk-reward analysis would never have owned either of these stocks in the first place and that it was almost criminally illogical to do so. Advertisement Here's how I view these cases. Given that only about 60% of biotech drugs in phase III trials get approval from the Food and Drug Administration, there was a significant chance that Northfield's product would fail. How much would the stock decline if the drug flopped? Perhaps 70%. And the upside? If PolyHeme worked, the stock might have gone up 70% in a day. Well, isn't this a good deal, given that approval is the more likely outcome? No. A probability calculation would show that your expected gain from buying Northfield is a mere 14%. That's a very low payoff if you have a 40% chance of a catastrophic loss. The risk-reward analysis for Telik was murkier because it had two products in development. But, clearly, the market was pinning most of its hopes on the drug that failed -- and any investor who followed the company would know that. So here, again, a prudent investor should have concluded that an expected gain of 14% wasn't enough to compensate for the enormous risk. Conclusion: The risk-reward equation for a small biotech company with one or two products in development is almost always lopsided against investors. You must avoid such companies like Ebola unless you have a PhD in molecular biology or the company has an important asset other than its patents (such as a lot of cash or a manufacturing plant that could be sold). Most good investors won't buy a stock unless the reward-risk ratio is five to one or, at a minimum, three to one. Sounds great, but how can you possibly calculate the ratio? It's not that difficult. Let me use a recent investment as an example. Advertisement I bought USG, the nation's leading maker of wallboard, last September at $51.75, about the time that Warren Buffett, the company's largest shareholder, was increasing his position. At the time, the company had a liquidation value of $55. Did that mean the stock couldn't go down? Six-to-one in my favor Hardly. By investing in USG, I was betting that the housing slump would not turn catastrophic. If it did, the stock could sink to $40. But if housing recovered, USG would likely return to $122, the price it reached in April 2006. So in buying shares of USG, I had 70 points of upside potential and 12 points of downside risk -- a nice six-to-one ratio. Plus, I had the tail wind of Buffett's endorsement. In mid February, USG traded at $53. Low risk and high potential reward -- that's what the game is all about. Editor's note: USG closed April 5 at $46.99. Columnist Andrew Feinberg writes about the choices, challenges and frustrations facing individual investors.