Hi Guy. Zip sales estimates: Revision #3
Thanks for the numbers. Looks like we just about have this puppy nailed! I thought I'd reply to you and take the opportunity to update the forum on my latest revision at the same time.
As you may recall in my first post [reply #2965], the most reasonable numbers that came out of the calculation were $110/drive and $11/disk. I subsequently posted the results of a revised calculation [reply #3055]. This post included the following two improvements: 1) I used information about Jaz sales and 2) I modified my estimation method so that it became, in effect, an optimization algorithm for constraint satisfaction. This resulted in revised numbers of $160/drive and $11/disk [reply #3055]. I received some input on this forum suggesting that $110 was bullish and $160/drive was out of sight.
In fact, I now believe my revised numbers weren't so far off. Why do I believe my numbers? First, I'm on revision 3 of my calculation. I'm using yet another piece of sales information and I've found a sales curve that satisfies all the publically known constraints. Also, I've seen some commentary by Chiros [Thanks mgsam] that suggests that a revenue of >$150/drive in the retail channel is very well accepted by analyists. This number is consistent with your [thanks Guy] statement that you wouldn't believe anything less than $155 (retail). Iomega has stated officially that OEM sales make up less than 10% of their zip sales (through Q1 FY96), so if we assume $155 retail and $95 OEM we come up with an average revenue of $149 for drives. I think you previously thought $11.00 - $11.40 was reasonable for disks. According to my results (revision 3) $149/drive implies $10.29/disk. Here are selected entries from the current table (revision 3) of drive vs disk revenue:
Table 1. $/drive vs $/disk
drive($) disk($)
160 8.76
150 9.78
145 10.29
140 10.80
135 11.32
130 11.83
120 12.85
110 13.88
As before, if doesn't make much of a difference which row in this table you use in the calculation, the quarterly revenue numbers stay pretty much the same. These numbers can be used, "as is" to make good estimates of total revenues from FY96 and Q1FY96. This is because the wholesale price of zip products has stayed constant from introduction through Q1FY95 [conference call] and because 90% of drive sales are through the retail channel. Projecting these numbers into the future is somewhat problematic since we can assume that the OEM fraction of the drive sales will increase, thereby pushing down the revenue per drive.
The two improvements I made in revision 3 over revision 2 are: 1) I'm using the information that zip revenue over Q4 FY95 was twice that of Q3 FY95. This came from the Q4 conference call to analysists [Thank you Ken]. The second change I made was that in revisions 1 and 2 I used $162M as my guestimate of zip sales for Q1 FY96. In revision 3 I made this guestimate an adjustable parameter since I found that I could not get satisfactory agreement with the jaz sales numbers unless I bumped up my estimate of zip revenue in Q1 FY96 from $162M to $165M.
So, were do I stand? Right now I believe $145/drive and $10.29/disk. Both numbers are a little lower than Guy Gordons estimates [$158/drive, $11/disk] based on retail mark-ups, but I'm not concerned. I think the point here is not perfect agreement. Instead, the point is that we have built a fairly good quantitative picture of zip sales. I think this picture pieces together the publically known "facts" pretty well.
Here are the quarterly zip numbers I get for FY95 and FY96. This assumes 5M drives produced in FY96. All numbers are in millions.
Table 2. Quarterly unit sales and $ sales for zip products
Qtr drives drives ($) disks disks($) total($) FY ($)
1/95 0.006 0.87 0.04 0.45 1.32
2 0.039 5.66 0.34 3.48 9.13
3 0.210 30.45 1.80 18.53 48.98
4 0.438 63.51 4.07 41.94 105.45 165
1/96 0.672 97.44 6.56 67.56 165
2 1.047 151.81 10.44 107.45 259.26
3 1.44 208.8 14.89 153.25 362.05
4 1.836 266.22 19.66 202.38 468.6 1255
A few comments concerning these numbers: First, don't believe the first two quarters of FY95 too much. These numbers are small and are not constrained very well by the publically available information. On the other hand, you can believe that the numbers are small. Look at how the production of drives leaps in Q3 and Q4 FY95. Do we see the hand of Seiko Epson appearing here? Second, production of drives ramps up at a faster rate following Q1 FY96 than before. This has to happen in order to meet the production figure of 5M zips in FY96.(I assume a linear ramp up of drive production following 1996). (KE stated that the ramp up in Q4FY95 was constrained by availability of parts.) Third, whereas in previous analyses I saw evidence that disk sales exceeded drive sales, I now find no evidence of this through FY96. This remains true no matter how you price the disks and drives (assuming numbers from table 1 above). Disk revenue and drive revenue stays comparable, but drive revenue is always a bit higher.
Well, that's all for now folks. I gotta get some sleep. I'm happy to answer questions, and I'm very open to constructive criticism. On the other hand I take a dim view of people who jump into the middle of discussions without having done their homework. Click on my name in the header of this post if you want to read my first two posts on this topic. These posts may answer some of your questions.
Good night All.
-- Fernando
"Check the facts first. Then form biased opinions" |
