From the Gilder board on AVNX, interesting: Any thoughts?
Thank You
Matt
Financial, that is, not technical.
.
I've created my own company model for Avanex by reverse engineering
Epoch Partners' model (see www.epoch.com) and then entering new
revenue growth assumptions based on RHK and Epoch estimates for
optical component growth in certain segments. I would like to share
some of the numbers I have come up with.
.
WHY HAVE I DONE THIS?
.
You may have noticed that the brokerages (including Epoch) had
estimated Q-over-Q revenue growth for Q1'01 (which ended last Sept) at
25% and pro forma EPS at break-even ($0.00). The numbers actually came
in at 80% Q/Q growth and pro forma EPS of $0.03, which led all the
brokerages to revise their net revenue and EPS estimates substantially
upward over their projected timeframe. For example, Epoch's numbers
(and most other brokerages) were revised as follows:
.
2001 2002 2003
Net Revenues (original) $134mn $259mn NA
Net Revenues (revised) $193mn $373mn $500mn
Change $(%) +$58mn(43%) +$114mn(44%)
.
EPS (original) $0.14 $0.58 NA
EPS (revised) $0.25 $0.74 NA
Change $(%) +$0.09(79%) +$0.16(28%)
.
WHY DID THIS HAPPEN?
.
Avanex management's guidance to analysts has proved conservative, and
analysts may have further low-balled the numbers because of the early
stage of the company's development. Additionally, analysts like to
low-ball their estimates so they can make upward revisions on their
buy recommendations, spin a good growth story and generate greater
buying interest.
.
Indeed, the AVNX share price increased dramatically after Q2 results.
However, the drawback to this strategy has been that the market has
punished companies which appear "overvalued" on a P/E and earnings
growth basis even though the E in question is way too low because of
analyst low-balling. I expect that this process will start to reverse
itself in the near-term and that P/E's will de-compress as the Fed
cuts rates and investors come back to the market. My expectation is
that AVNX's P/E and its E will both be going up sharply this year.
.
By revisiting Epoch's model, I hope to generate more accurate
projections of Avanex's revenue and earnings growth. The model also
can estimate how much analysts are likely to revise their numbers
upwards after Q2 results are announced on January 23rd.
.
Jeff Strambovsky's (sp?) did a good valuation on Avanex by projecting
ten-years out and working back (see the Global Network board). I'm
doing a somewhat similar exercise, but going from the present to the
future and just focusing on revenues and earnings growth (you can pick
your own valuation method).
.
.
If the analyst numbers are bad, why use them at all?
.
The Epoch model provides a starting point and incorporate management's
views as to the evolution of various margin ratios. This can be seen
from a comparison of the company's three-year projection in its
November presentation and the revised Epoch numbers. They are quite
similar:
.
media.corporate-ir.net.
htm
.
epoch.com
.
.
FIRST RESULTS: WHAT COULD THE REVISIONS BE IF NET REVENUES GROW MORE
THAN 25% Q/Q?
.
Q/Q revenue growth was 62% in Q3, 84% in Q4 and 80% in Q1. Epoch and
the other brokerages project viewed this growth as a one-off increase
over their baseline of 25% quarterly growth and revert to that number
for Q2, with quarterly growth numbers declining to 7% by the end of
2002.
.
Below I have done an "expected revision" exercise based on Q2 net
revenue growth being higher than 25%. Here are the numbers.
.
.
>>30% Q/Q growth => Q2 rev's $45 million, EPS $0.05
.
2001 2002 2003 2004 2005
Rev's $201 $392 $527 $691 $905
YoY% 393% 95% 35% 31% 31%
EPS $0.26 $0.78 $1.12 $1.57 $2.28
YoY% NM 201% 44% 40% 45%
.
>>50% Q/Q growth => Q2 revs of $52 million, EPS of $0.05
.
2001 2002 2003 2004 2005
Rev's $226 $452 $608 $797 $1,045
YoY% 456% 100% 35% 31% 31%
EPS $0.28 $0.89 $1.28 $1.80 $2.62
YoY% NM 215% 44% 40% 46%
.
>>70% Q/Q growth => Q2 revs of $59 million, EPS of $0.06
.
2001 2002 2003 2004 2005
Rev's $252 $512 $689 $903 $1,184
YoY% 518% 103% 35% 31% 31%
EPS $0.31 $1.00 $1.45 $2.04 $2.97
YoY% NM 227% 45% 41% 46%
.
Based on the model's margins, Avanex should easily hit its $0.05 EPS
estimate and will beat its estimate by 1 cent if it grows at least
55%. For earnings to miss its estimate, revenues would have to
decline. To beat earnings, revenues would have to increase in excess
of 100%. Keep in mind that these results are based on holding margins
constant at company expectations. If revenues outperform
significantly, margins could be lower and earnings higher.
.
I didn't find any whisper numbers for Avanex on earningswhispers.com,
but I seem to recall another whispers site which gave an EPS of $0.06
for Q2. That would imply growth of at least 55% (or lower growth and
better margins).
.
Remember, the annual projections above are based just on ONE quarter
being better than estimated and growth reverting to Epoch's estimates
thereafter.
.
My guess for Q2 is that earnings will be $0.06, revenues will by
50%-60% and analysts will increase 2001 earnings by $0.05-$0.10 and
2002 earnings by $0.15-$0.25.
.
.
WHAT IF WE USE DIFFERENT REVENUE PROJECTIONS OVER THE NEXT FIVE YEARS?
.
Avanex's high growth rate over the past three quarters is due to its
low revenue starting base and the ramping up of new products. We
SHOULD expect revenue growth to slow down, but the question is, at
what pace?
.
Epoch shows these net revenue numbers on a calendarized basis:
.
CY2000 CY2001 CY2002 CY2003 CY2004
CY2005
Original $84 $192 $323 $430 $564
$740
% YoY -- 129% 69% 33% 31%
31%
Revised $108 $283 $438 $574 $752
$986
% YoY -- 163% 55% 31% 31%
31%
.
But here are some of the estimates from RHK and Epoch for the growth
of some of the industry segments in which Avanex's competes:
.
Terrestrial DWDM Optical Comp (RHK of 11/2000):
.
CY2000 CY2001 CY2002 CY2003 CY2004
US$Bn $5 $9.5 $13 $17 $24
% YoY -- 90% 37% 31% 41%
.
DWDM (Epoch, late summer?):
.
CY2000 CY2001 CY2002 CY2003 CY2004
US$Bn $5 $8 $11 $15 $21
% YoY -- 60% 38% 36% 41%
.
Metro DWDM (Epoch):
.
CY2000 CY2001 CY2002 CY2003 CY2004
US$Bn $0.5 $1.8 $3.2 $5.3 $7.9
% YoY -- 260% 78% 64% 50%
% Total 10% 23% 29% 35% 37%
.
Metro DWDM (Epoch #'s revised with new RHK estimates):
.
CY2000 CY2001 CY2002 CY2003 CY2004
US$Bn $0.5 $2.1 $3.8 $6.0 $8.9
% YoY -- 328% 77% 57% 50%
.
So metro DWDM will be one of the fastest growing segments of the DWDM
optical components market, and we know that Avanex should make a big
splash in metro because it will provide the best connectivity,
scalability and flexibility. At the same time, Avanex will become a
leading choice for high-channel-count backbone systems. In the words
of Charlie Burger, "Avanex will be EVERYWHERE, and has certainly made
its first inroads in the backbone, I would guess that by the end of
2002 it will be the MAN that will have really propelled Avanex
forward."
.
Another factor to consider is that the in-house component divisions of
the systems integrators are likely to be spun off at some point. I
think the merchant vendors will gain market share over these freed
captive divisions because the system integrators will no longer feel
obliged to buy company product.
.
Given the strong anticipated growth in Avanex's sector, and the
expectation that Avanex will come to dominate some key sectors, I have
calibrated the model with the following calendarized revenue growth
numbers:
.
Growth in Line with Metro DWDM:
.
CY2000 CY2001 CY2002 CY2003 CY2004
CY2004
US$Bn $120 $475 $881 $1,353 $2,006
$2,868
% YoY -- 295% 85% 54% 48%
43%
.
I don't think these numbers are unreasonable, especially considering
the major product launches in CY2001 such as PowerShaper (this
quarter) and PowerMux NxG (by 2H2000 I believe). It happens that the
annual growth number I project trends down to around the 50% level,
which was what Jeff used it his estimates (so it must be right!).
.
Here are the resulting fiscal year estimates for net revenues and EPS:
.
2001 2002 2003 2004 2005
Rev's $270 $687 $1,096 $1,655 $2,409
% YoY 561% 155% 60% 51% 46%
EPS $0.33 $1.32 $2.28 $3.70 $6.02
% YoY NM 305% 72% 63% 62%
.
So this model projects that Avanex will have $1 billion in revenues
and $2.28 EPS in 2003, more than double the analyst revenue and
earnings estimates for that year.
.
I don't think these numbers are wildly optimistic. Revenues of $1
billion would be just about 8% of the terrestrial DWDM optical
component market in 2003 as estimated by RHK. No assumption is made of
an overnight conversion to a "lambdasphere" world. If the Avanex
solution DOES catche on big time and lead the way to such an outcome
in the next 5-10 years, then Avanex will be HUGE.
.
Uncertainty is a given in investing and the economy. This is my best
"stab in the dark" for Avanex.
.
Regards,
CDKelly
(sorry about the scrunched up tables) |
