Repost Pineda's great work.
Every serious investor should read this assessment. Here is the post
by Pineda:
<<
From: Fernando J. Pineda
Jun 13 1996 10:40PM EST
To All: Here's my analysis of FY95-96 ZIP unit sales and revenue
This last week I got curious about what I could learn from
published information (basically my file of posts from this thread).
My goal was to assume as little as possible and to use numbers that
were derived from published data or numbers that did not have a big
influence on the result. The following (long) report is a slightly
edited version of my notes. I think if you work through my analysis
you will (might?) be convinced of the following:
1) Iomega probably gets roughly $110/drive and $11/disk.
2) It is very likely that on average 7.3 disks are bought in the
first month of zip ownership and 1.5 disks are bought the next
month. More disks are sold in the following months but it is
hard to estimate these numbers from the limited information.
3) If Iomega sells 5M zips in FY96, then ZIP+disk revenue in FY96
will be $1.1B. The breakdown is
Q1 $162M
Q2 $250M
Q3 $314M
Q4 $360M
I'm not finished with this, but I have to get back to my life. I
figured if I didn't get this out now, I probably wouldn't get it out
at all.Anyway, I'd like to get some intelligent *and civilized*
discussion going.
-- Fernando
_______________________________________________
STEP 1. Estimating the unit drive sales, FY95-FY96
I estimate the cummulative number of zip drives sold per month
from the following 7 pieces of publically available information. I
also assume that the numbers from month to month vary smoothly
(you have to make some assumptions and smoothness seems to be the
simplest one!).
1) Product was introduced in March 1995.
2) PR News release March 15, 1996
³SHIPMENTS OF ZIP DRIVES PASS ONE MILLION³
3) April 19 Conference Call w/analysts:
²We shipped our first millionth zip drive in February,
less than one year after the first shipment. ³
4) April 19 Conference Call w/analysts:
²We are pleased to note that we shipped as many Zip drives
in the first quarter as we did in all of 1995²
5) PR News release June 3, 1996:
³to date approximately 2 Million sold...²
6) K. Edwards states that Iomega & partner¹s current capacity
in FY96 is approximately 5M drives.
Month/yr zips/mo(x1000) cumm. (x1000)
1/95 0 0
2 0 0
3 6 6
4 9 15
5 14 29
6 22 51
7 40 91
8 60 151
9 75 226
10 100 326
11 135 461
12 175 636
1/96 200 836
2 260 1096
3 293 1389
4 340 1729
5 388 2117
6 415 2532
7 443 2975
8 468 3444
9 490 3934
10 513 4447
11 525 4972
12 538 5511
Several points to make. First, the most reliable numbers are
probably from 8/95 through 3/96. These are the most constrained by
the publically available reports. Second, the very early numbers are
just too small to be reliable. I pretty much fiddled them to satisfy
the smoothness criterion and the known fact that Seiko-Epson wasn¹t
producing any zips early on. Finally, it¹s important to stress that
the whole analysis from here on pretty much hinges on the reliability
of these numbers. Once I established these numbers from the above
published information I didn¹t diddle them any more to make anything
come out favorably or unfavorably. I don¹t see much leeway in the pre
Q1 FY96 numbers, you can argue the remaining FY96 numbers if you
believe 8 million drives in FY96...I don¹t).
STEP 2. Estimating the unit disk sales
In an article written by William Terdoslavich entitled ³The media
is the message² in ³Computer Reseller News² (Ken Marcus, Post #1517)
Michael Collins, the director of channel marketing for IOMEGA makes
the following statements, which are supposedly based on some market
surveys:
³[The user] ...buys an average of 5 to six disks with the
drive...Then 64 percent of the people purchase more media in the
first eight weeks of owning the drive...They buy another six disks...²
Of that portion three fifths of those repeat buyers purchase their
disks within four weeks of buying the drive.
In other words we have the following:
On average 5-6 disks purchased at the same time as the drive
On average 0.64*6= 3.84 more disks purchased within 2 months. Of
these (3/5)*3.84=2.3 disks are purchased in the first month and
(2/5)*3.84 = 1.53 disks are purchased in the second month.
So in the first month we have, on average, 7.3-8.3 disks sold
and in the second month we have on average 1.53 disks sold. So....the
average number of disks of disks sold in the first 2 months is
8.8-9.8 disks/drive. This is consistent with the statement made by
Terdoslavich (in the same article) in which he claims that 1 million
Zips have been sold and 10 million disks have been sold(posted May 6).
These unofficial IOMEGA statements certainly make me believe that a
tie ratio 10 is a safe estimate.
In fact, given that both the sales report (1 million zips) and
the sales study are based on ³early returns² and cover only the first
2 months of drive ownership, I certainly would find it very easy to
believe that continued disk sales over a year or more would push the
tie ratio much higher. I¹ll *assume* that each drive sold generates
the following monthly pattern of disk sales:
1 2 3 4 5 6 7 ...
7.30 1.53 0.50 0.50 0.125 0.125 0.125 ...
The ³...² in the table indicates 0.125 disks per month for the
life of the drive. In the first year this produces 11.2 disk. For all
subsequent years this produces 1.5 disks/year. This is pretty
conservative. Interestingly enough, nothing changes by more than a
percent or two in FY95 and FY96 if I use a pattern like
1 2 3 4 5 6 7 ...
7.30 1.53 0.50 0.50 0.25 0.25 0.25 ...
This results in 11.8 disks the first year and 3/disk per year
thereafter. Anyway, if I use the first (conservative) table, the unit
drive sales estimated above, produce the following monthly unit disk
sales:
Month/yr cumm.zips disks/mo cumm.disks
1/95 0.0 0.0 0.0
2 0.0 0.0 0.0
3 0.006 0.044 0.044
4 0.015 0.075 0.119
5 0.029 0.119 0.238
6 0.051 0.189 0.427
7 0.091 0.339 0.766
8 0.151 0.521 1.287
9 0.226 0.677 1.964
10 0.326 0.907 2.871
11 0.461 1.227 4.098
12 0.636 1.606 5.704
1/96 0.836 1.895 7.599
2 1.096 2.429 10.028
3 1.389 2.823 12.851
4 1.729 3.292 16.143
5 2.117 3.800 19.942
6 2.532 4.156 24.099
7 2.975 4.503 28.602
8 3.444 4.826 33.429
9 3.934 5.115 38.544
10 4.447 5.412 43.956
11 4.972 5.628 49.580
12 5.511 5.831 55.412
Now some observations. First, note that in Feb.96 the cummulative
drive sales equals 1 million while the cummulative disk sales equals
10 million. Thus our pattern of disk sales is consistent with the
statements in the ³Computer Reseller News² article that 10 million
disks had been sold at the point that 1 million zips had been sold.
Ihave to admit this surprised me because I didn't diddle the numbers
to get this! Second, look at the number of disks sold in Q4 FY96 vs
the number of disks sold in Q3 FY96. Finally, note that from our table
we get 7.15M disk sold in Q1 FY96.I will use this number in the next
step.
STEP 3. Estimating the revenue per drive and the revenue per disk
The transcript of the analysts conference call in April gives us
a way of using our unit sales estimates to estimate the revenue per
disk as a function of the revenue per drive. Moreover, it allows us
to calibrate our estimates so that even if we are off on our unit
sales estimates, we will still produce a reasonable near-term
prediction of total zip revenue(probably good for Q2 and maybe even
Q3 ?).
The statements by K. Edwards are that Zip+Jaz acount for $185 in
revenues. Of this, Jaz accounts for more than 10%. Thus Zip sales must
be less than $166. Now I *assume* that if Jaz sales had been more
than 15%, KE would have said ³...Jaz accounts for more than 15%....²
Consequently, I believe that Zip sales were between $166M and $157M.
Let me split the difference and say that Zip sales in Q1 FY96 were
$162M.
Now, if each drive produces revenue of ³A² dollars and each Disk
produces revenue of ³B² dollars then the total revenue ³R² in a given
quarter is
R=A*Z+B*D
where Z is the number of Zip drives sold and D is the number of
disks sold. The values for R, Z and D that we have already estimated
are R=$162M, Z=0.753M, and D=7.15M. Accordingly we can solve our
little equation and obtain the following estimates for A vs.B.
Revenue/Drive (A) Revenue/Disk (B)
$215 $0
$190 $2.65
$175 $4.23
$150 $6.87
$130 $8.97
$120 $10.02
$110 $11.08
$105 $11.60
$100 $12.13
$90 $13.18
$75 $14.75
$50 $17.40
$25 $20.03
$0 $22.66
Looks to me that the plausible range for revenue seems to be
between $130 - $100 for a drive and $8.97 - $12.13 for a disk.
Guy Gordon (Post #2816) estimated Iomega¹s revenue per disk by
taking into account the markup added by the retailer and the
distributor. He concluded that Iomega was probably getting
$11.40 to $11 per disk. This falls in the middle of the above
table, so I think it is a pretty reasonable estimate.
STEP 4. Estimating Zip Revenues (FY95 & FY96)
Let me use revenue of $11.08 per disk and $110 per drive.This
corresponds to a retail mark-up of 81% ($200 retail) for the drive
and between 35%-80% ($15-$20 retail) for the disks. Let me also
assume that IOMEGA produces and sells 5M drives in FY96. Finally,
let me assume that these numbers don¹t change very much over the two
years of FY95-FY96. Under these assumptions I get the following
monthly (quarterly) revenues. (In units of $M):
Month/yr disk zip zip+disk quarterly
1/95 0 0 revenues
2 0 0
3 0.49 0.66 1.14 Q1=1.14
4 0.83 0.99 1.82
5 1.32 1.54 2.86
6 2.10 2.42 4.52 Q2=9.20
7 3.75 4.40 8.15
8 5.77 6.60 12.37
9 7.51 8.25 15.76 Q3=36.3
10 10.05 11.00 21.05
11 13.59 14.85 28.44
12 17.79 19.25 37.04 Q4=86.5
1/96 21.00 22.00 43.00
2 26.91 28.60 55.52
3 31.28 32.27 63.54 Q1=162
4 36.47 37.40 73.87
5 42.10 42.72 84.81
6 46.06 45.65 91.70 Q2=250
7 49.90 48.77 98.66
8 53.48 51.52 104.99
9 56.67 53.90 110.57 Q3=314
10 59.97 56.47 116.43
11 62.32 57.75 120.07
12 64.61 59.22 123.82 Q4=360
STEP 5. Pause and return to the wife and kids....
So here is what I learned in this excercise and where I am going.
1. I think I have a pretty good idea of revenue per disk and
revenue per drive. I believe $11/disk and $110/drive.
2. In June of 1996 IOMEGA reaches critical mass! In this month
disk revenues are greater than drive revenues for the first time.
Actually this is not such a big deal, since from the beginning it is
clear that disk revenues are comparable to drive revenues. It becomes
a big deal in later years if the pattern of disk purchases for drives
that are older than one year is much higher than my very conservative
estimate of 1.5 drives per year.Unfortunately, I can't extrapolate
that far because revenue per disk and revenue per drive will most
certainly not be lower in a couple years.
3. The total revenue for zip products in FY96 will be $1.1B
assuming 5Million disks sold in FY96. This is the same number a lot
of analysts are getting, but now I understand where this number comes
from!
4. OK, so now that I understand how to do this, I can be wild
and extend my spreadsheet to FY97 and/or look at 8 million disks...
5. When Q2 figures come out and if Iomega gives us a little more
information, I can continue this analysis with more current data.
6. I¹d like to see two things. First, some discussion of these
numbers (can someone do their own spreadsheet and check that I don't
have any programming errors?) and perhaps some refinements or further
constraints. Second, I¹d like to see someone who knows what they are
doing, take these numbers and try to estimate gross margins and eps
from the publically available information. I assume of course that
you would share your approach.
Any takers?
-- Fernando
>>
KM |
