Forecasting Models ...
Jim,
<< You're correct, I posted this sometime back on SI- Note that my 2004 est is considerably lower than the Q's current guidance (handset sales increases, ASPs increasing rather than decreasing). Also, the 2005 EPS is close to what S&P is now forecasting. >>
You have a good facility for numbers, so your models are always interesting, but I must say, they are not always presented with adequate detail of how you are arriving at the summary detail you post, and as a consequence I find them to be less meaningful than they might otherwise be.
The board might be interested in knowing ...
1. What are you using for CDMA unit shipments in each of the years you are modelling?
2. What are you using for WCDMA unit shipments in each of the years you are modelling?
3. What specific blended ASP are you using for each of the years?
4. What blended QCT net operating margin are you using for each of the years?
Relative to number 1 & 2, I ask that because in a general sense, handset forecasts for 2.5G GSM, 2.5G/3G (depending on who is categorizing it) 1xRTT and GSM EDGE, 3G 1xEV (DO and DV), and 3GSM UMTS WCDMA (including J-UMTS in Japan) are being moved up by virtually all agencies or analysts, credible or incredible, with good and realistic market visibility or without, and with good historical track records or without.
Most certainly, the latest of Qualcomm's composite "GSM is Going Out of Business" slide most recently shown by Bill Davidson on July 31 (I believe that was last showing) which is based on averages of older forecasts by IDC, InStat, Shosteck Group and Yankee Group - who for the most part aren't the most credible of available agencies for handset forecasts to begin with - are not very insightful. Not only do they obscure the reality of GSM growth, but they also obscure the reality of 3GSM UMTS WCDMA growth and CDMA2000 and IS-856 1xEV-DO growth, and overall paint a poor picture of industry growth.
Brian Modoff obviously has extremely poor visibility on GSM and 'other' growth (and actually even historical sales of same), but by contrast, I consider his recently revised CDMA and WCDMA forecasts to be reasonably credible.
Michael Thelander's recently revised WCDMA forecast as graphically depicted in the recently published IBD article is slightly more aggressive - the most aggressive I've yet seen relative to 3GSM UMTS WCDMA handset ramp - but depending on how quickly early UE gives way to fully type approved UE, how quickly handset ASP declines, user acceptance, and the pace and scale of buildouts, it could also be deemed credible, albeit as a high side at least at this moment.
In the next day or so, I hope to post abstracts of some of the more recent handset forecasts I've collected as well as WCDMA forecast dating back to last September constructed by Herschel Shosteck and Jane Zweig. That forecast references back to Modoff and Thelander's more pessimistic S2N Deutsche Bank USA forecast as it stood at that time. I'll also abstract separately the methodology Shosteck and Zweig used in that older forecast because it is quite sound.
Best,
- Eric -
AI
