Jay: I don't sense a mean bone in your body...
Besides, responses like yours are why I bothered to post
my analysis. So, where are we...
Yah, in the light of day I'm thinking $160 is way too high, the problem is that if I reduce this, then I'm forced into a higher disk price. Right now my suspicion is that my fit to the constraints is not good enough. I could try to get a better solution, but it is tedious to do this by hand. (Damn, I knew I should have written some C-code!). I'm hoping to enlist some others in the forum to diddle the numbers in an attempt to find the unit sales curve that best fits the constraints. To this end I've put my spreadsheet on the web.
olympus.ece.jhu.edu
Right now it is only
usuable by folks who have Macintosh computers with Excel. Maybe on Monday I may be able to convert it to Excel for Windows. Unless someone does it for me (hint-hint).
Another thing to keep in mind, the numbers that come out of the analysis are biased towards the past. They can't possibly take into account the relative mix between OEM and retail that will occur in the future.The analysis has two parts: 1) backing out historical sales figures and unit revenue figures 2) using these to predict the future. I'm sure my method will work to get the latter, but the former requires further assumptions about OEM's partners, etc. Just like you said.
Finally, concerning the 5M capacity figure, I should have pointed out in my revised post, the observation you have made. Which is the following: A 5M capacity in FY96, coupled with an assumption that the monthly capacity changes relatively smoothly, implies that the capacity is ramping up! This is very very good! :) You can see the ramp up if you plot the monthly sales figures. I think it is pretty much inevitable that the bears will be surprised by the FY97 performance!
Anyway, thanks for the comments. I'm still thinking. I want others to participate in this!
-- Fernando |
