Deconstructing and Rebuilding IBM’s Q4 Server Sales
February 9, 2009 Timothy Prickett Morgan
Well, it seems that I am not the only one who plays around with the limited amount of information that IBM gives to Wall Street analysts. As I said last week, when I tore apart IBM’s numbers from the fourth quarter and then tried to put them together as revenue streams, not percent changes, this game is more fun when lots of people play.
Maybe we can use Newton’s Method of Approximation across a large group of people making estimates as a means of getting to the truth. (The wisdom of crowds, after all, can guess the number of jelly beans in a jar with surprising accuracy despite a wide variation in individual guesses, or plot a course through a maze in the smallest possible steps, among other things.) To do that, though, more people have to play the game.
Each quarter, as I do this story, I invariably get a few emails from private equity firms or other financial industry players who want to bend my ear about my numbers and what they may or may not mean. But I also got one email from a reader last week who wanted to play the IBM Server Sales Guessing Game. Here’s his take:
I, too, track System and Technology product revenues–largely using historical relative (percent) gain and loss data to create enough equations to back out the actuals–along with the occasional random tidbit offered up by someone at Big Blue.
In any event, I think my numbers differ from yours, but still thought I might share mine, having enjoyed and benefited greatly from your remarks.
For 4Q08, my guestimates are:
In Q407, my figure for p plus i was $2.064 billion. Thus, I believe in 4Q08 p plus i was down over 19 percent. Also the game of propping up p revenues by “consolidating” i revenues is effectively over.