Thanks for mentioning that JAVA had two upgrades, which prompted me to take a look over at C, which had a SELL on JAVA going into earnings:
Sun Microsystems Inc (JAVA): Upgrading to Buy; All Lines on the P&L Trending Favorably
6 November 2007 - 12 pages
Sun Microsystems Inc (JAVA) Change in
Upgrading to Buy; All Lines on the P&L Trending Favorably opinion Y
Rating change Y
* Upgrading to Buy - We are upgrading shares of JAVA from Sell to Buy w/$7.25 tgt due to improving product line, restructuring, and share repurchase. We are also lowering risk rating from Speculative to High to reflect better earnings predictability, return to sustainable profitability,
and ample working capital.
* Attractive Risk/Reward - We expect solid 2FQ08 (December)
results on strong gross margin, headcount cuts, and share
repurchases. While revenue momentum should improve more
gradually, we see worst-case downside to $5 and best-case
upside to $8 within one year, a very favorable risk/reward.
* Product Positioning Is Improving - With the introduction of
its second-generation differentiated Niagara chip
architecture, the addition of Intel to its x86 lineup, and
the introduction of Rock in CY08, Sun looks to have a
competitive server line from top to bottom for the first time
in seven years.
* Aggressive Restructuring - Operating expenses should
decline by nearly $400M between FY07 and FY09 via real estate
and IT consolidation, headcount reductions, etc. Consensus
estimates do not give the company full credit for this.
* Share Repurchase - Sun repurchased an impressive $1.25B in
stock during 1FQ08, and has another $1.55B remaining in its
current authorization. With $1-2B in free cash flow during
FY08 and FY09, there is ample room for additional repurchase
once the current authorization is exhausted.
Buy/High Risk
1H
from Sell/Speculative
Price (05 Nov 07)
US$5.71
Target price
US$7.25
from US$5.25
Expected share price return
27.0%
Expected dividend yield
0.0%
Expected total return
27.0%
U p g r a d i n g t o B u y
We are upgrading shares of JAVA from Sell to Buy with a revised target
of $7.25. In the near term, gross margin upside, operating expense
reductions, and share repurchase should drive solid earnings despite
the absence of revenue upside.
For this reason, clients should begin building positions now and simply
average down if product transitions prompt a 5-10% pullback in the
share price. Our thesis is as follows:
* Within one year, Sun looks to have a competitive, low- to high-end
server product portfolio for the first time in seven years.
* Despite fairly-specific guidance regarding planned operating-expense
reductions, sell-side consensus estimates do not give the company full
credit for a $400M reduction in annual operating expenses between
FY07 and FY09.
* Sun is reducing its share count more quickly than we had expected.
Moreover, with $1-2B in annual free cash flow during FY08 and FY09,
the company appears to have plenty of flexibility to authorize
additional repurchase activity once its current $3B authorization is
fully used.
While product transitions and a change in channel revenue-recognition
policy (from sell-in to sell-through) could limit revenue upside for the
next several quarters, we expect revenue momentum to improve
throughout FY08 as the company's server line-up improves.
Sun has been steadily diversifying its server product portfolio for a
number of years:
* In 2002, Sun acquired Afara Websystems, a developer of multi-core,
multi-threaded microprocessors based on Sun's UltraSPARC
architecture. The first UltraSPARC T1 'Niagara' systems based on this
innovative architecture were introduced in 1CQ06 and achieved $100M
in quarterly revenue in just one quarter. During 3CQ07, Sun introduced
'Niagara II,' the second generation of this architecture, which offers
almost twice the price/performance of its predecessor and
significantly-improved floating point performance. The 'Niagara'
architecture offers significantly-better price/performance and
performance/watt than competing x86 chips for certain applications,
including web, network infrastructure, and online transaction
processing. With the introduction of Niagara II, we expect production
deployments of Niagara to accelerate.
* In 2004, Sun added x86 servers to its proprietary line-up in recognition
of the fact that the x86 architecture offers superior price/performance
for most applications and the flexibility to run a number of different
operating systems. Initially, Sun limited its x86 portfolio to AMD
microprocessors, which in turn limited the company's available market
to 15-20% of the total X86 market. In 1CQ07, Sun announced its
intention to offer a full line of both AMD and Intel-based x86 servers,
which should significantly expand the company's available market. We
expect to see a full line of Intel-based servers in Sun's portfolio by the
end of FY08. We note that Sun is the only Unix server vendor providing
support for its proprietary Unix, Solaris, on high-volume, low-cost x86
architecture.
* Sun should complete the rejuvenation of its server portfolio in the
second half of calendar 2008 with the introduction of 'Rock,' a version
of the multi-core, multi-threaded Niagara processor with support for
large symmetric-multiprocessing systems. At this point, Sun will have a
mid-range and high-end proprietary line able to compete more
effectively with IBM on price/performance and virtualization capability.
Cost reductions
Sun reduced pro forma operating expenses by approximately $400M
between FY06 and FY07 via headcount reductions, real estate
consolidation, and other restructuring actions. We expect the company
to reduce operating expenses by an additional $400M between FY07 and
FY09 through IT upgrades and consolidation, real estate consolidation,
and further headcount reductions. The Street is not giving Sun full credit
for this level of expense reduction, in our view.
In the near term, the company should benefit from a 1,500 person
headcount reduction announced in August 2007. The 3CQ results
suggest that virtually none of this reduction occurred during the
September quarter, but we still expect it to be largely completed within
two quarters. This relatively-small reduction should alone yield $150M in
annual operating expense savings.
Given Sun's recent strong track record of delivering on promised
operating-expense reductions, our confidence in the company's ability to
reduce opex by $400M between FY07 and FY09 has grown. This still
leaves the company one percentage point below its stated goal of a 10%
operating margin in FY09, but a nickel above consensus EPS in FY09.
Share repurchase
Sun completed an impressive $1.25B of its $1.8B outstanding share
repurchase authorization during 3CQ07. Because the repurchase
occurred throughout the quarter, diluted share count only dropped 95M
shares despite the fact that Sun repurchased a total of 245M shares.
Another significant drop in diluted share count will likely occur during
the December quarter as Sun enjoys the full impact of 3CQ's large
buyback.
Looking forward, it is reasonable to expect Sun to continue buying back
shares aggressively for two reasons: 1) The company seems very
confident in its ability to deliver 10% operating margin during FY09,
which would suggest that it believes JAVA shares are undervalued; 2)
Sun ended 3CQ07 with a cash and marketable debt-securities balance of
$5.2B and should generate another $1-2B in free cash flow annually
during both FY08 and FY09. In fact, we would not be surprised to see a
second-consecutive quarter of $1B+ in repurchase activity during
4CQ07.
With its current cash balance, Sun could theoretically buy back 25% of
the company at the current share price.
R a i s i n g E s t i m a t e s a n d T a r g e t P r i c e
We are significantly raising EPS estimates for FY08, FY09, and FY10 to
reflect higher gross-margin assumptions, lower operating-expense
assumptions, and lower share count assumptions.
Of these three changes, our increase in gross margin assumptions is
sure to be the most controversial. We concede that a significant portion
of recent improvements in hardware gross margin have been driven by
mix and favorable component pricing, but we expect mix to remain
favorable as customers consolidate infrastructure onto larger systems. In
addition, barring a sharp slowdown in demand, it seems unlikely that
Sun's hardware margins will deteriorate 300-400 basis points during the
coming year in the face of significant improvements in the company's
relative competitive position.
* We are raising FY08E EPS from $0.17 to $0.27 to reflect higher gross-
margin assumptions, lower operating-expense assumptions, and lower
share count assumptions. We are now $0.05 above the pre 3CQ07-
results consensus.
* We are raising FY09E EPS from $0.24 to $0.38 to reflect higher gross
margin assumptions, lower operating expense assumptions and lower
share count assumptions. We are now $0.05 above the previous
consensus.
* We are raising FY10E EPS from $0.30 to $0.52 to reflect higher gross
margin assumptions, lower operating expense assumptions and lower
share count assumptions. We are now $0.10 above the previous
consensus.
Valuation: Risk/Reward Looks Attractive
Without significant changes to our P/E or DCF assumptions, our revised
estimates suggest a new 12-month price target for JAVA shares of
$7.25, 31% upside to Monday's after-market price of $5.55. The target
price implies a multiple of 15X. Given estimated 34-117% EPS growth
between FY08 and FY10, we view a 15X target multiple as conservative.
Given our comfort with near-term earnings estimates, the P/E would
have to decline to 13X in order to pull the shares down to $5, which we
view as unlikely.
Our DCF valuation assumes 3% growth in free cash flow between 2011
and 2018 and 2% perpetual growth thereafter. We also view these
assumptions as conservative.
Net, net, we believe risk/reward is very attractive at current levels, with
worst-case near-term downside to $5 (-10%) or 13X, and best-case 12-
month upside to $8 (+44%) or 16X.
R e v i e w o f 1 F Q 0 8 R e s u l t s
Revenue of $3.22B (+1% yoy, -16% qoq) was in line with our estimate and about
$50M (one percentage point) below consensus. Both product revenue of $1.98B
(+1% yoy) and service revenue of $1.24B (+1% yoy) were in line with our
forecasts. Excluding $113M in restructuring charges, non-GAAP EPS of $0.05
exceeded both our $0.01 and consensus $0.03.
Server revenue of $1.31B (-1% yoy) was $5M above our forecast. ASPs
of $17.6K grew 1% yoy versus our 3% forecast. Total server shipments
only declined 2% yoy versus our 4% forecast, thanks to good upside in
Niagara volume outweighing slight shortfalls in UltraSPARC and x64
volumes.
UltraSPARC revenue of $970M (-9% yoy) was $80M (seven percentage
points) below our forecast. UltraSPARC ASPs rose 18%, below our 25%
forecast. Despite $45M upside in Niagara revenue of $170M, Sun's total
Unix-server (UltraSPARC plus Niagara) sales declined 2% versus IBM
pSeries' 6% rise. By contrast, x64-server revenue of $165M grew 10%
yoy, resuming growth after a flat 2CQ and comfortably ahead of IBM
xSeries' +6%.
Storage revenue of $505M were $5M below our forecast; however, 3%
yoy growth was better than both overall revenue and IBM storage's flat
yoy performance.
The 48.0% (+530bp yoy, +60bp qoq) product gross margin was 360bp
above our estimate, even better than 2CQ's 200bp upside. Server
margins again benefited from favorable component pricing; DRAM
prices only rose 2% qoq in 3Q after having dropped by more than 50%
qoq in 2Q.
Professional Services revenue of $260M exceeded our forecast by $10M,
compensating for $10M downside in Maintenance revenue of $769M.
However, the 49.2% (+430bp yoy, +210bp qoq) gross margin was 380bp
better than our forecast and likely the highest margin in at least ten
years, indicating superb utilization.
Thanks to the sharp gross-margin upside in both business segments, the
corporate gross margin of 48.5% (+500bp yoy, +130bp qoq) exceeded
our forecast by 370bp and was the highest since 2CQ00's 52.1%. In
turn, the 5.5% (+680bp yoy) operating margin was 580bp better than our
forecast. Opex of $1.39B was $65M lower than our forecast despite the
in-line revenue with all the savings coming from R&D, indicating
excellent cost control.
The sharecount was about 65M below our forecast; during the quarter
the company repurchased 245M shares, but due to the timing of the
repurchase, it should not fully impact sharecount until 4CQ. Cash flow
from operations of $574M was the highest since 1CQ01's $767M
(excluding the Microsoft settlement-benefiting 2CQ04), and free cash
flow of $447M more than doubled the $157M of a year ago.
K e y p o s i t i v e s
* UltraSPARC ASPs of $28.4K rose 18% yoy, marking the eighth-
consecutive quarter of double-digit growth. Although the lack of
UltraSPARC-volume growth remains an issue, Sun-as with other server
vendors-continues to benefit from the industry-wide trend toward
virtualization. (Please see IT Hardware: Virtualization is a Good Thing
for the Server Industry, our contribution to the CIR software team's 23
September 2007 initiation report on VMware,
citigroupgeo.com, for more details.)
* Niagara server volume grew 70% versus our 25% forecast. Despite
Niagara II-based servers shipping in early 4CQ, Niagara sales during
3CQ surprised us by not showing any signs of disruption.
* Opex management remains very solid. R&D spending $65M below our
forecast indicates that the company is finally making progress in
holding the line here; as a percentage of revenue, R&D declined by
90bp yoy and 230bp from two years ago. Surprisingly, the company
was able to manage opex without needing to execute during 3CQ the
1,500 headcount cuts it announced in early August, indicating the
potential for additional savings once the cuts occur.
* Contrary to IBM's comments regarding 3CQ, management specifically
cited financial services as an area of good demand despite (or, given
the transaction volume-oriented nature of many of Sun's products,
perhaps because) of the recent turmoil in global capital markets.
Further, management indicates that financial-services exposure for
Sun is 20% or less of total revenue, significantly less than we had
believed.
K e y n e g a t i v e s
* UltraSPARC volumes declined by double digits for the eighth-
consecutive quarter, despite volume availability of the new Fujitsu
SPARC64-based models during 3CQ. The strong ASP growth has
mitigated this to some degree, but UltraSPARC revenue still declined
9% yoy in the quarter.
S u n M i c r o s y s t e m s I n c
C o m p a n y d e s c r i p t i o n
Founded in 1982, Sun Microsystems develops, manufacturers and
markets hardware, software, storage & services for enterprise computing
networks. Its portfolio includes: desktops and workstations for graphics,
software development, design & other business and technical apps;
entry-, mid-range, and enterprise-class servers; storage software and
disk/tape systems; SPARC family of RISC processors; Solaris UNIX OS;
middleware software; and svcs (deployment, consulting, sys integration
& maintenance). In FY07 (ending June), total revs was $13.8B (45.2%
overall GM), with 63% from products (45.1% GM) & 37% from svcs
(45.2% GM). Geographical rev breakdown was 41% US & 59% int'l (6%
Canada/LatAm, 36% EMEA, 17% APAC). Although Sun continues to
generate significant revs from the telecom, gov't, and fin svcs verticals,
it continues to diversify into others like mfg, education, retail, life
sciences, media & entertainment, transportation, energy/utilities &
healthcare.
I n v e s t m e n t s t r a t e g y
We rate shares of JAVA Buy/High (1H) with a 12-month target price of
$7.25. Sun's continued efforts to develop and market an end-to-end
enterprise computing platform places it in direct competition with
significantly larger competitors such as Microsoft, Intel, IBM and
Hewlett-Packard. However, the company has made meaningful recent
strides in improving both COGS and opex, resulting in significantly-better
margins even without substantially-better revenues. At the same time,
Sun has made good progress with new products (such as x64, Fujitsu
APL, and T1) and future offerings (such as Rock) that together offer the
real prospect of a future revival in topline growth.
V a l u a t i o n
We have derived our JAVA target price based on price/earnings (P/E)
and discounted-cash flow (DCF) analyses. Sun's inconsistent recent
history of growth and profitability make it challenging to determine the
appropriate target P/E for the shares, and we do not believe the shares
deserve to trade at a significant premium to peers or the market. IBM
shares have historically traded at 16-20X F12 EPS, excluding unusual
periods such as the Internet bubble and the subsequent downturn. HPQ
shares have historically traded in a core range of 11-20X F12 EPS. The
S&P500 is currently trading at 15X F12 EPS. As a result, we have chosen
to use a target multiple of 15X for JAVA shares. The application of this
multiple to F12 operating EPS of $0.42 for the four quarters beginning
2FQ09 and the subsequent addition of $1.43 in net cash per share yields
a 12-month fair value of $7.73. DCF analysis, assuming 3% revenue
growth between FY11E and FY18E, 2% terminal growth in free cash flow,
a beta of 1.22 (formerly 1.11), and an 8.8% (formerly 8.5%) WACC
produces a 12-month fair value of $6.84. Given these inputs, we
establish a 12-month target price of $7.25.
R i s k s
We rate JAVA shares High risk due primarily to the relative lack of scale
and inconsistent recent history of profitability compared to larger
competitors such as IBM, HP, Dell, Intel, and Microsoft. However, given
Sun's ample cash-to-debt ratio and substantial improvements in efficient
working capital management, we do not view liquidity as a meaningful
near-term issue. While Sun currently enjoys a strong position in Unix
servers, we are concerned that Wintel and Lintel servers with superior
price/performance are likely to take share from low-end Sun servers for
"edge-of-the-network" and even some application tier workloads in
coming years. While Sun now sells a line of low-end Lintel, SPARC, and
Solaris-on-X86 servers in order to boost its relative value proposition at
the low end, we believe these servers are likely to dilute margins over
time and, unless incremental to gross profit dollars on a net basis, de-
lever the company on relatively-fixed proprietary processor R&D
spending. Sun derives 45% of total revenue from the telecom and
financial services verticals and is vulnerable to any IT spending
slowdowns within these areas. Should these factors have a greater
impact on the company than we are anticipating, our target price might
not prove to be conservative enough. Similarly, should factors turn more
positive than anticipated, the shares might continue to trade above our
target price and could materially outperform it. |
