Silicon Graphics International (SGI) took a big fall (from $20.14 to $15.05) the day after reporting June quarter earnings on August 8. My view of SGI's future is beginning to dim, although I believe the sell-off was overdone.
I began a long position in Rackable back in June 2006. Rackable looked like a bright, shiny upstart that could take over much of the datacenter computing market with its energy-efficient server rack systems. But eventually much larger companies (HP and Dell, for instance) began competing on price, driving RACK margins into the ground. Strategies were changed several times, and acquisitions were made. In particular a lot of money was invested in acquiring and developing a data storage company, which was going to save the day. Eventually it had to be written off. But all along Rackable had a large cash cushion, so management, or various managements, kept trying new strategies and tactics.
Meanwhile a once-famous computer company, Silicon Graphics, Inc., went bankrupt. Known as an early graphics innovator, SGI was also famous for losing any and all money invested in it. It was killed largely by ATI (now part of AMD) and Nvidia (NVDA) putting capabilities on chips (GPUs) that SGI sold expensive workstations to do. Its history is a great example of not seeing the emergence of a lucrative new tech trend branching off of the death of an old trend, much like when PCs killed minicomputers.
In 2009 Rackable purchased Silicon Graphics, Inc., by then a maker of supercomputers, out of bankruptcy and adopted Silicon Graphics International as its name with stock ticker SGI. It looked like a great deal because the SGI assets seemed worth so much more than the $25 million paid by Rackable, which still had a large cash cushion for the new company. While good things have sometimes happened since then, the overall picture has been of a company unable to operate at consistently profitable margins.
Fast forward to the June quarter of 2013 (SGI's Q4 fiscal 2013). Revenues were $170.5 million, down 17% sequentially from $232.6 million and down 5% from $179.5 million in the year-earlier quarter. GAAP net income was negative $4.5 million, down sequentially from $9 million, but up from negative $18.4 million year-earlier. EPS (earnings per share) were negative $0.13, down sequentially from $0.27, but up from negative $0.58 year-earlier. In other words, not profitable on a GAAP basis.
Non-GAAP net income was better at $6 million, flat sequentially from $6 million, but up from negative $3 million year-earlier. Non-GAAP EPS was $0.17, down 5% sequentially from $0.18. Adjusted EBITDA was $8 million, down sequentially from $9 million.
So SGI is in the twilight world between being unprofitable on a GAAP basis and profitable on a non-GAAP basis. The good news is that SGI is debt-free and ended the quarter with $175 million in cash, up $22 million in the quarter. If I wanted to argue that SGI is a buy, the argument would largely be about cash generation in the quarter.
What has gone wrong lately? The "legacy" cloud infrastructure business. Isn't the Cloud the biggest thing since the Internet, and are not numerous companies big and small making Cloud profits? Yes, but for some reason SGI can't compete. So what is essentially the old Rackable datacenter business is being de-emphasized. SGI will no longer chase low-margin deals in the space.
That sounds okay, but keep in mind that after operating expenses, SGI's other products have not proven to be very profitable either. SGI will continue with various supercomputers used for HPC (high-performance computing) and Big Data computing.
The other strategic switch? Into the data storage business again. Management says the storage pipeline is seeing multiple orders for the new products. Initial shipments are occurring in the present quarter and should make a meaningful contribution to revenue in the second half of fiscal 2014. In other words, far enough in the future to be pie in the sky (clouds?). It might work out, there are certainly many complaints about the data storage solutions currently available, but data storage is extremely competitive. Meaning SGI is likely to again do well with a limited set of loyal customers, then run into margin issues once it has to compete on bids. If I see revenue and good margins in the storage business in 2014, I will change my mind.
Call me cynical, but at this point it would take a solid year of SGI beating guidance to get me to take SGI long-term guidance as something of a guarantee.
September quarter guidance (first quarter fiscal 2014) is for revenue to decline to a range of $160 to $170 million. GAAP EPS is expected between negative $0.14 and negative $0.07, with non-GAAP EPS from positive $0.07 to $0.14.
Long term, full fiscal year 2014 (ending June 2014), SGI hopes to deliver $1 per share non-GAAP EPS, but loaded towards the second half.
If I believed that I would set a price target of $20, based on a growing business with a non-GAAP P/E of 20. Note the 52-week high this year was already above that at $20.97 on August 6.
I last added to my SGI position at $5.27 on May 21, 2012. I last sold SGI on July 22, 2013 at $18.43. I am still long SGI, but with a minimal position. In other words, even before the June results came out I thought $18.43 was a good price given what I knew about SGI.
SGI closed on Tuesday, August 13 at $16.58, getting a $0.99 dead-cat bounce during the day. I see SGI as worth between $16 and $18 per share to me, barring new developments. But with a market capitalization of just $561 million, it does not take much volume to generate exaggerated price swings in either direction.