jbog - there were 9 weeks, not 14 weeks in Q3 to manage inventory so that there would be no effect in Q4. we also know that hospitals and retailers had to work down their own inventory before they bought from wholesalers. script share data corroborates this, as sales in the last 4 weeks were 30% more than sales in the first 5+ weeks, so working back it clearly took time to work down inventory at retail level, and only then could wholesalers begin to unload their stock to an appreciable degree (and there is no reason to think hospital pharmacies performed differently - particularly in this case where we kinda know erosion in retail was more accelerated than hospital (per SNY). so all in all i think it's a stretch to assume inventory reached steady state by end of Q3
while i am not that knowledgeable in this area, i did live through this once before with recothrom, so i know how inventory can negatively affect reported sales. in Q4 2009 zgen sold 9.2M of recothrom. in Q1 2010 they reported 9M of sales - 2% less that the prior quarter, despite end user demand increasing 14%. so a negative 16% effect from inventory on sales. this too was for a product shortly after launch, but not a situation where they purposefully intended to overstuff the channel in case another competitor entered the market (NVS admits to higher inventory stocking in case another generic was approved), so it's not a stretch for me to think there was an even larger inventory effect in this case
supportive evidence again comes from simple arithmetic - 40% of the market discounted 15% is well over 170M. so either NVS is wrong and they don't have 40% of themarket (something hard to imagine since SNY even admitted to losing 33% market by volume and more than that by dollar sales in Q3 - which makes sense since margins in hospitals historically lower where they kept higher market share), or discounting would have to be on the order of 35-40% (again something hard to imagine given initial pricing was at a 14% discount and MNTA has stated there has not been significant pricing pressure - and why would there be in a duopoly situation)
the one piece of evidence supporting your contention is jeff george's answer during Q&A - where he hemmed and hawed to the tune of yes 40%, 56-57M per month, that's about right
i agree with your last statement time will tell. SNY provided a good amount of data in their Q3 call on US lovenox dynamics, so hopefully in a little over a week we'll get a better sense