explanation, page-65

I know you're trying to be realistic/conservative TDA, but I'll keep reinforcing that it is more realistic to use sales projections rather than actual data when the market is growing rapidly.

You're using actual annual US Arixtra sales of US$203M for Oct08-Sep09, which is now on average 12 mths out of date, and more by the time of an FDA decision happens.

My regression model suggests that US sales will have increased to approx. US$290M pa by Jun09. Using your other assumptions, that comes out to be earnings per share of 20.1c ....

20.1c x 15 PE = $3.01
20.1c x 20 PE = $4.02
20.1c x 25 PE = $5.02

I can't vouch for your other assumptions, but I think the basis behind US sales is too conservative. The only thing that might change is if a Lovenox generic has a negative impact on Arixtra volume (although its hard to imagine a doctor making the decision between enoxaparin and fondaparinux would consider the prices of related generics.)