On March 25, Bill Simpson wrote an analysis of Super Micro Computer (SMCI). On March 28, the company raised $64 million with an initial public offering that was priced below a forecast range. The 8 million share offering was sold for $8 per share, compared with a $9.50 to $11.50 forecast range. Today, April 16, the stock is trading at $9.66.
The text of Mr. Simpson's original writeup follows:
• • •
Super Micro Computer plans on offering 9.2 million shares (assuming over-allotments) at a range of $9.50 - $11.50. Merrill Lynch is lead managing the deal, Needham and UBS co-managing. Post-offering SMCI will have 28.6 million shares outstanding for a market cap of $300 million on a $10 1/2 pricing.
*Note* - This SMCI deal is another of those tech ipos structured more like a 1999 ipo than should be in this day and age. This means there are excessive outstanding, already in the money and exercisable, options here. SMCI has a whopping 15 million shares of common stock issuable upon the exercise of stock options outstanding at a weighted average exercise price of $2.12.
You can bet over the next few years these options will be exercised, converted into shares and sold. Factoring in this dilution, SMCI would have 43.6 million shares outstanding for a market cap of $458 million on a $10 1/2 pricing. This is a trend we've been seeing way too much of lately and one I do not care to see. While these options will not effect early trading of SMCI, they do create a pretty significant headwind for a stock over the first few years of trading.
What we will see here with SMCI is a continued stream of insiders exercising and selling options beginning at the 180 day market and most likely continuing quarterly for a few years. SMCI as a company will have to outperform operationally to simply 'stand still' on stock price as these options will be growing market cap, even with stock at same price. Please keep this in mind for those looking to hold SMCI longer term - you will be diluted heavily by insiders. Many of the options shares are part of SMCI's 1998 stock option plan. As most options need to be exercised within 10 years, there is a pretty good chance that many of the 15 million option shares will be exercised and sold by the end of 2008.
Approximately 1/3 of ipo proceeds will be used to repay all debt, 2/3 is for general corporate purposes. Chairman of the Board, President and Chief Executive Officer Charles Liang will own 32% of SMCI post-ipo.
From the prospectus:
We design, develop, manufacture and sell application optimized, high performance server solutions based on an innovative, modular and open-standard x86 architecture. Application optimized servers are configured to meet specific customer needs in contrast to typical servers which are offered in limited standardized configurations.
Customizable server company: SMCI customizes its servers to meet specific customer requirement by adjusting memory, processing power, and/or input/output capabilities. SMCI's servers are based on x86 which is the open standard utilized by both Intel (INTC) and Advanced Micro Devices (AMD) in their microprocessors. INTC and AMD are SMCI's only suppliers of microprocessors.
Note that SMCI does not participate in the traditional UNIX server market. The open architecture server market in which SMCI participates is growing much faster than the traditional UNIX server market. Going forward, the UNIX server market is expected to grow 6%-7% annually the next 4 years, while open architecture servers are anticipated to grow 40%+ over the same period of time.
SMCI's open architecture approach allows for easy integration of its servers as well as allowing for the ability to 'stack' its servers to create scalable server systems. The selling point for open system servers is the ease of scaling when compared to the traditional UNIX server market. SMCI also offers server components to OEM's. Majority of revenues are from individual servers and server components, not complete server systems. As SMCI sells through distributors, this is not surprising. Companies purchasing from distributors can/do stack SMCI's servers and components to create their own specialized server system.
SMCI commenced operations in 1993 and has been profitable every year since. Pretty impressive considering the various tech cycles in those years. SMCI sells primarily through distributors, who accounted for approximately 70% of revenues in the past two years. In 2006, 400 companies in 70 countries purchased SMCI servers and server components.
SMCI believes its advantages include:
1) Customizable and flexible;
2) Rapid time to market;
3) Power efficiency and thermal management;
4) High density - Density is amount of space required by a server. By being 'high density,' SMCI's servers require less floor space.
As of 12/31/06, SMCI did not offer a 'high performance blade server' solution. However SMCI expects in the first half of 2007 to launch a high performance blade server solution, called Superblade. Blade servers are specifically designed for high density by sharing power, cooling, networking and other resources within a single server-rack enclosure, compared to standard servers which each require their own independent resources. By eliminating these repetitive components and locating them in one place, a greater number of blade servers can be used in a smaller physical area as compared to standard servers. This would appear to be SMCI's answer to servers offered by companies such as Rackable (RACK).
No single customer accounted for more than 10% of revenues each of 2005 and 2006.
SMCI utilizes Ablecom Technology for contract design and manufacturing coordination support. SMCI’s purchases from Ablecom have accounted for roughly 1/3 of cost of sales expense the past 5 years. Ablecom is a private company owned by the brother of SMCI President/CEO Charles Liang.
Legal - In 9/05 Rackable Systems filed a patent infringement suit against SMCI. In 2/07 the Court threw out much of the lawsuit, ruling very favorably for SMCI. The remaining claims in the suit are set to go on trial in 8/07. Also SMCI was fined by the U.S. Government for sales to Iran during the 2001-2003 period. In 2006, SMCI settled the charges and paid what appears to be approximately $500,000 in fines to the U.S. government.
$2 per share in cash, no debt post-ipo.
The server market is notorious for lumpy revenues. A large order in any particular quarter can help a server company blow away estimates and, conversely, a company pushing out an order into another quarter can mean a pretty ugly looking quarterly earnings miss. Servers are generally not purchased on long term contracts, or even that far in advance. SMCI notes in the prospectus that most sales in a given quarter come from orders placed in that specific quarter. The fact that SMCI has been able to achieve and maintain profitability for 15 years in this highly competitive 'lumpy' sector is a pretty strong selling point here.
Note that SMCI accounts for revenues upon shipment to distributors, not the final sell-through from distributor to end user/customer. This has led to approximately $1-$3 million annually in 'inventory writedowns' for returned and/or excess product inventory.
SMCI's fiscal year ends 6/30 annually. FY '07 will end 6/30/07. 50% of revenues are derived from U.S. customers, 50% from outside the U.S. Revenues from outside the U.S. are growing more briskly and SMCI should generate more non-U.S. revenues than U.S. revenues in FY '07. Specifically SMCI is focusing future growth in China and throughout Asia.
Revenues have ramped impressively since 2002. Revenues in fy '05 (ending 6/30/05) were $211 million, $302 million in FY '06 and on pace for $400-$425 million through first half of FY '07.
Gross margins are slim in this niche. Through the past 18 months, SMCI has had gross margins in the 18%-20% range. This does not leave much margin for error. To be a successful company with these gross margins, managing operating expenses is crucial. In a niche in which SMCI must continue to develop new/better servers annually, R&D is the heftiest operating expense annually. Fortunately sales & marketing expenses are controlled at a low rate thanks to SMCI selling most of its products through distributors. Overall operating expenses equaled 10% of revenues in FY '06 and 11% through first two quarters of FY '07. SMCI has kept operating expenses in that 10%-11% of revenue area the past three fiscal years. I would expect that to continue to be the case going forward.
Net margins for FY '06 were 6%. EPS for the officially post-ipo share-count of 28.6 million was $0.63. When option shares are included, EPS dips to $0.41. On a 10 1/2 pricing, SMCI would be trading 17 X's trailing earnings without options shares included, 26 X's trailing earnings with future option dilution factored in. Big difference.
FY '07 (ending 6/30/07) - Based on first half of FY '07, we should expect full year revenues in the $400 - $425 million range. This would be approximately a 33% revenues increase from FY '06. Gross margins and operating expense ratios look to be in range of previous few fiscal years. Note however, that gross margins ticked down a bit in the 12/31 quarter from that 18%-20% range to 17%. While not dramatic, this will most likely drop FY '07 net margins by 1% for the full fiscal year. Net margins then should be in the 6% range, instead of FY '07's 7%. Earnings per share based on 28.6 shares outstanding should come in $0.80-$0.85. When factoring in option shares, earnings for FY '07 should be $0.55. On a $10 1/2 pricing, SMCI would trade 13 X's FY '07 earnings when no option dilution is factored in, 19 X's FY '07 earnings when full dilution is factored in.
Rackable Systems is the recent ipo most similar to SMCI. This will even be more the case when SMCI starts shipping its blade server product and components in the first half of calendar year 2007. Let's take a brief look at the two. Note, in looking at SMCI's metrics below, we'll 'split the difference' on options and factor in 50% option dilution. As mentioned above, these options will not be in play until beginning approximately 180 days after ipo, although some may hit 90 days post-ipo. These options will be exercised, so when looking at SMCI we must factor in a good chunk of these options. The numbers below reflect SMCI as if 50% of outstanding in the money options were exercised.
RACK - $525 million market cap. Currently trades 1.1 X's FY '07 revenues and 22 X's FY '07 earnings estimates. Note that these estimates have been reduced substantially in the past 90 days. RACK is expected to post a 27% revenue increase in FY '07.
SMCI - $380 million market cap (assuming 1/2 option dilution) on a $10 1/2 pricing. Would trade just less than 1 X's FY '07 revenue estimates and 16 X's FY '07 earnings with an expected 33% revenues increase in FY '07.
Conclusion - Due to the lumpiness of revenues and the low gross margins, this should never be a sector in which a company is valued aggressively on forward PEs. RACK holders discovered this recently as a few good quarters in a notoriously lumpy sector were then forecast as the 'norm' going forward. Factoring in a huge quarter as the norm going forward will nearly always be a mistake in this sector. SMCI has done a terrific job over the years managing its business and growth through 13-14 profitable years in a row. This tells me a couple of things:
1) SMCI has been able to continually innovate over the years. This is a must in this sector as new products/components are constantly being introduced.
2) SMCI has done an excellent job managing operating expenses in a low growth margins sector.
Both these are exactly what you need in a lumpy low gross margin sector. You need constant innovation and strong management. SMCI's years of continued profitability would seem to indicate they've both.
Revenue growth has really ramped here the past three years. I'm skeptical that this swift growth will continue. However if SMCI is simply able to grow in the 20%-25% range in FY '08, this is a very attractive ipo in range. The wild card for FY '08 and beyond is SMCI's Superblade server being introduced in the first half of 2007.
Keep in mind, the coming option dilution is a big negative here as ipo buyers will be massively diluted over the next few years. However even factoring in this dilution, SMCI still looks attractive to me in range. This dilution, coupled with low gross margins would lead me to pass here on any aggressive opening prints as it is doubtful those would be long-term sustainable.
This is not a sector in which you ever want to pay high multiples. In range and $1-$2 above however, SMCI is a recommend.
Disclosure: Note that tradingipos.com does currently have a position on SMCI.