SuperMicro (SMCI) Long – $8.70
Market Cap: $340MM; EV:
$280MM
SMCI is a recent tech IPO
that has gotten no respect and presents a rare opportunity to buy a profitable,
fast-growing tech company at a single digit multiple. The company has been
growing revenue organically 20 – 30%+ over the past several years and has been
profitable every year since it was founded 14 years ago. The stock trades at 9X
our lowball earnings estimate (ex cash) for the current fiscal year ending June
2008. For FY 2009 (ending June 2009), we believe the company can earn north of
$1.05 (excluding interest and including options expense). At $8.75 and with
$1.50 per share in cash, that puts the stock at 7X FY 2009 EPS. Only one
analyst covers the stock and management is extremely conservative.
SMCI is a server / server
component company that competes with more established players like Dell and HP.
Many people compare it to Rackable (RACK), which has had its problems recently.
But SMCI is a completely different business with a much more diversified
customer base.
We typically wouldn’t bother
looking at a company like this. On the surface, the business seems too
competitive and we’d expect any primary research to yield 100 different
opinions about the company’s future. But it seemed cheap enough to warrant a
few phone calls. So we started calling SMCI’s large distributors and were
amazed by the consistency of the positive comments we heard. Every single
distributor we spoke with expected accelerating growth from SuperMicro products
(meaning they expect 2008 growth to be greater than the growth they see in
2007). The median response for expected 2008 growth was 20% vs. 15% in 2007.
Note that significant incremental sales growth beyond this has come and should
continue to come from the company increasing direct sales as a percent of the
total.
The company’s competitive
advantage centers around its innovation and speed to market. They do not
necessarily compete on price but rather offer “application optimized”
solutions—customized servers designed to perform for specific needs (processing
power, input/output speed, memory optimization, etc). We consistently heard
that SuperMicro offers superior service and products for mid-size enterprises
who are not big enough to be meaningful customers for Dell. We consistently
heard that there is surprisingly little competition in this niche. We found
this hard to believe initially, but we really urge you to call a customer / distributor
or two so you can hear for yourself what they have to say.
New products and product
cycles at Intel drive the business for SuperMicro, so gross margins are lumpy
(have ranged from 14.7% to 21.5% in any quarter over the past several years).
There are numerous reasons why we expect margins to increase. The street is
projecting gross margins in the low 18% range for the next couple years (taking
the low end of management’s conservative margin guidance). We think this is too
low. Below is a chart of historical margins, including our conservative guess
for 2008.
FY End June 30,
|
2004
|
2005
|
2006
|
2007
|
2008E
|
Revenue
|
167,065
|
211,763
|
302,541
|
420,393
|
534,949
|
Gross Margin
|
17.5%
|
15.8%
|
19.9%
|
17.9%
|
19.0%
|
Operating Margin
|
4.4%
|
5.4%
|
9.1%
|
7.7%
|
8.9%
|
To explain the variation in
margins in the above, FY 2006 represented the last major Intel processor
upgrade, which is typically on a two year cycle (note the good margins in FY
2004). In FY 2007, which was an “off” year for Intel, the company started
serving the Internet data center market and had two quarters where it had
trouble sourcing components. Those problems are now solved. Gross margins in
the past two quarters were 18.0% and 19.6%, both solid numbers. But both were
actually artificially low because of higher than usual inventory write-offs,
which we do not believe will continue at these levels. You can see that
adjusted for inventory write-offs, gross margins are almost hitting highs,
despite the fact that the new Intel processors won’t be released until late FY
2008 or early FY 2009.
|
|
|
9/05
|
12/05
|
3/06
|
6/06
|
|
9/06
|
12/06
|
3/07
|
6/07
|
|
9/07
|
|
|
F2005
|
Q106
|
Q206
|
Q306
|
Q406
|
F2006
|
Q107
|
Q207
|
Q307
|
Q407
|
F2007
|
Q107
|
Revenue
|
|
$211.8
|
$64.5
|
$72.1
|
$75.9
|
$90.0
|
$302.5
|
$90.2
|
$113.6
|
$105.7
|
$110.9
|
$420.4
|
$117.9
|
COGS
|
|
178.3
|
52.0
|
58.5
|
61.1
|
70.7
|
242.2
|
72.2
|
94.6
|
87.6
|
90.9
|
345.4
|
94.9
|
Gross Profit
|
|
33.5
|
12.5
|
13.7
|
14.8
|
19.3
|
60.3
|
18.0
|
19.0
|
18.1
|
20.0
|
75.0
|
23.0
|
Gross Margin
|
|
15.8%
|
19.4%
|
18.9%
|
19.5%
|
21.5%
|
19.9%
|
19.9%
|
16.7%
|
17.1%
|
18.0%
|
17.8%
|
19.5%
|
Inventory write-down
|
|
1.4
|
0.8
|
0.8
|
1.0
|
0.3
|
2.9
|
0.6
|
0.6
|
1.7
|
2.7
|
5.6
|
2.0
|
Adjusted COGS
|
|
176.9
|
51.2
|
57.7
|
60.1
|
70.4
|
239.4
|
71.6
|
94.0
|
85.9
|
88.2
|
339.8
|
92.9
|
Adjusted Gross Profit
|
|
34.9
|
13.3
|
14.5
|
15.8
|
19.6
|
63.2
|
18.6
|
19.6
|
19.8
|
22.7
|
80.6
|
25.0
|
Adjusted Gross Margin
|
|
16.5%
|
20.6%
|
20.1%
|
20.8%
|
21.8%
|
20.9%
|
20.6%
|
17.3%
|
18.7%
|
20.5%
|
19.2%
|
21.2%
|
The following things will
continue to drive margins higher:
- Scale. This allows the company to optimize
manufacturing. They only recently started assembling in Asia
for Asian/European sales. Prior to this, all assembly was done in California.
- New product innovations like blade servers, which
only started shipping in volume quantities in the Sept quarter. We are
told these new products carry higher margins.
- Continued trend of greater percent of sales from
whole server systems rather than server components (systems were 36.3% in
FY 2007, 34.5% in 2006, and 31.4% in 2005)
- Greater percent of direct sales vs. through
distributors. From FY 2005 – 2007, direct sales were 17%, 27%, and 33% of
total revenue. We expect this trend to continue.
- The company plans (next year, we believe) to
release a software product for server management (typical 90%+ gross
margin).
Factoring in these benefits
and the upcoming Intel product launches, it’s hard to imagine that we should
expect margins in the 18.2% range going forward as the street has modeled. We
very conservatively model 19% gross margins for the current FY 2008 (remember,
Q1 margins were 21.2% excluding the inventory write-down, 20.1% adjusted for
the “excess” write-down above a normalized amount) and come to a GAAP EPS of 77
cents. On a non-GAAP basis (excluding stock option expense), to be comparable
to current estimates, this is around 84 cents (vs. Needham at 77 cents). We don’t see any reason
why margins can’t be 20%, which adds another 9 cents to our estimate. There is
some theoretical limit on gross margins given the competitiveness of the
industry, but the company is far from its scale potential, so there should be significant
room to go. Obviously a successful software product introduction changes
everything.
Opex was a bit higher than we
expected in Q1, much of this due to a one-time R&D charge related to the
introduction of new products. We expect that once the company shows further significant
operating leverage (which could happen as soon as next quarter, with revenue
guidance up 12% sequentially), they will start to get a growth multiple.
The stock is worth at least
$15 (15x CY 2008 earnings of 90 cents or more, plus cash) today and more if our
research is right about accelerating revenue growth (or even just continuing
anywhere near the 20% range). FY 2009 should be a big year for revenue and
margins given Intel’s expected processor rollout. Once SMCI is “discovered,”
it’s not hard to see the stock north of $20 within a year.
Risks:
Two major things bother
everyone. The first is that RACK is really one of few comps people can think
of. We have also done work on RACK in the past. The company serves a much more
competitive Internet data center market and has huge customer concentration
issues. Its top two customers in Q2 were 54% of revenue. During 2006, the top
three were nearly 70%. SMCI currently has no 10% customers. RACK is not
profitable and not growing nearly as much as SMCI, but both companies trade at
around .5x 2008 projected revenue.
The second issue is that 20 –
30% of SMCI’s COGS is paid to a related party contract manufacturer called
Ablecom (which is run by the brother of the CEO and owned by the CEO’s family).
We don’t like this but are comfortable that it’s not such a big deal. The
bankers did significant diligence on this prior to taking the company public.
The company has other contract manufacturers and operates with Ablecom at
similar pricing.
Yesterday (11/5) was the
first day employees can sell since the IPO. There are several that have been
with the company for 10+ years and seem to want to get paid. We think this is
creating a huge opportunity in the stock. Most of these potential inside
sellers are engineers (not necessarily valuation experts) and ought to have
some desire to diversify their holdings after so long. We don’t have much
insight into how much stock is for sale now, but from past conversations with
the CEO/founder, we don’t think much or any of it will come from him.
demonstration of operating leverage in FQ2, software release, other new products gaining traction