If I offered you a 10% chance to increase your bet 20 fold, would you take it?

The odds are in your favor but, obviously, you can't bet your entire savings; there's a 90% chance you'll lose everything. Perhaps a small bet might be in order.

That's basically where pharmaceutical companies sit with Alzheimer's disease. We really don't know enough about how to treat the disease to assign a chance of approval much larger than 10%, but the potential upside from the large market and lack of effective treatments make the bets worthwhile.

Witness the 90%
There were quite a few Alzheimer's drug failures in 2012.

Medivation (NASDAQ:MDVN) and Pfizer (NYSE: PFE) started the year off registering the third failure for Dimebon. Considering the previous failures, the odds of success were pretty low.

Pfizer and Johnson & Johnson (NYSE:JNJ) teamed up with Elan (UNKNOWN:UNKNOWN) through acquisitions to develop bapineuzumab. Elan and Wyeth, the original developers, proceeded onto phase 3 trials with phase 2 data that wasn't necessarily stellar. Not surprisingly, the phase 3 program crashed and burned.

Eli Lilly's (NYSE:LLY) solanezumab failed two phase 3 trials this year. The results weren't quite as bad as Eli Lilly's attempt a few years ago, where patients taking its Alzheimer's drug candidate, semagacestat, seemed to be worse off than those taking placebo. With solanezumab, at least there was a hint of efficacy, just not enough to pass the clinical trials.

And yet pharma presses on
Despite all the failures drugmakers continue to seek out the elusive Alzheimer's drug that actually works.

Baxter's (NYSE: BAX), Gammagard is currently in Phase III trials set to read out next year.

Eli Lilly is designing a new phase 3 program for solanezumab, trying to tease out the hint of efficacy observed in the first set of phase 3 trials.

Merck (NYSE:MRK) recently announced its plan to move MK-8931 into a phase 2/3 study. The pharma will use General Electric's (NYSE:GE) imaging agent to identify patients who have Alzheimer's disease, which may make it easier to demonstrate efficacy. Of course, the drug still has to work; identifying the right patients to treat with an ineffective drug isn't all that useful.

Good investment?
Let's do some back of the envelope calculations. Figure it costs $2 billion  to develop an Alzheimer's drug, and let's assume $5 billion in average annual sales over 10 years, or $50 billion over the life of the drug. Figure a 50% profit margin for the drug itself. That includes cost to manufacture and sell the drug, but doesn't include fixed administrative costs and funds for research and development of future blockbusters; drugmakers' profit margins are much lower, but we're looking at the incremental addition of a single drug.

Using pretty conservative numbers, we've got a $2 billion cost to develop a drug, versus $25 billion in profits. If you want to be less conservative, it's not hard to see how profits could be double that for an Alzheimer's drug that worked really well, or one that had a longer patent life. And the development cost is considerably lower if we're just considering whether to fund a phase 3 program. The expenses up to that point are sunk costs that shouldn't be included in the financial decision.

Put another way, if there's a 10% chance that the drug will work, it makes financial sense to risk the money now for the potential profit latter. Is there a 10% chance that an Alzheimer's drug will work? That's the multi-billion question now, isn't it?

I'm inclined to say go for it with the obvious caveat that drugmakers should make as much effort as they can to eliminate obvious losers in phase 2 development before costs balloon up. And the long-shot bets must be balanced out with more conservative drug development that has a higher likelihood of success.

Biotechs don't really have that kind of luxury, because they usually don't have that many drugs in their pipeline. Elan selling off half its rights to bapineuzumab might have been the best move in this space.

Fool contributor Brian Orelli has no positions in the stocks mentioned above. The Motley Fool owns shares of General Electric and Johnson & Johnson. Motley Fool newsletter services recommend Elan and Johnson & Johnson. Try any of our Foolish newsletter services free for 30 days. We Fools may not all hold the same opinions, but we all believe that considering a diverse range of insights makes us better investors. The Motley Fool has a disclosure policy.