SGI Challenged by Analysts

Today Silicon Graphics International (SGI) stock plunged over 20% following yesterday's release of December 2011 (fiscal Q2 2012) numbers. A group of analysts on the investor conference call challenged statements by SGI management and tried to pin down details on why profits were lower on higher revenues.

SGI manufactures what are still sometimes called supercomputers: high capability computer systems for datacenters, scientific research, and high performance business uses. Amazon.com is the largest single client by revenue. The U.S. government, in aggregate, is an over 10% customer.

Revenues for fiscal Q2 were $195.2 million, up 9% sequentially from $178.9 million and up 10% from $177.5 million in the year-earlier quarter. Bookings were even stronger in the quarter, partly due to the newest marvel, the ICE X HPC platform, which booked $90 million.

Costs, however, were up enough to reduce GAAP EPS to negative $0.07 per share. Non-GAAP EPS was positive $0.04, but that represents non-GAAP net income of just $2.3 million on $195 million of revenue, or about 1%.

What analysts wanted to understand was exactly why SGI has gone back to the old SGI way of making great products and generating revenues while losing investors' money. Also, if it can be fixed, exactly how that fix would take place. If you have followed SGI you know the current version was created when Rackable bought the old, bankrupt SGI. Rackable itself had been consistently losing money, but had a lot of cash and bought SGI for essentially nothing.

Even before the analyst question segment began the new temporary CEO, Ron Verdoorn, gave a number of reasons for the earnings shortfall. The three main ones were fixed costs in Europe that are too high, forex exchange rates issues, and the transition to new product lines. Success in selling the rackable line of server systems, which have lower margins than other lines, also led to lower overall margins.

The Europe problem would appear to be fixable by either cutting costs in Europe or selling more product while keeping costs level. SGI has lost money in Europe for over 2 years, hoping the sell more option would work out. Now restructuring is planned, but as usual that will create short term costs before generating long term savings.

I was surprised that management did not try to divert analysts in another direction. Research and development expense was up almost $3 million from the December quarter of 2010. I consider that a form of reinvestment. If ICE X alone generated $90 million in orders, that would certainly rationalize the $5 million increase in sales and marketing expense.

An even bigger bite out of revenue came from cost of revenue, which would be cost of goods sold for the hardware, but includes cost of providing services, too. That was up almost $18 million from the year-earlier quarter. Maybe SGI is pricing its products and services too low, but there could be a time lag effect as new systems are developed, manufactured, and sold before revenue is recognized.

Inventories were listed as worth $117.6 million, which is an astonishing increase of $36.6 million from $81.0 million year-earlier. I doubt those inventories consist of unsold or unsellable supercomputers. They should represent orders in hand and a reasonable estimate of pipeline demand.

If that is true, maybe management's statement that things will look a lot better in the second half of calendar 2012 is believable. Of course such estimates always assume no Japanese earthquakes, Thai floods, or Euro collapse issues, which were bad assumptions in 2011. Despite all of the problems of 2011, SGI made a big impression on the computer world with its new products, generating record revenue.

Analysts are right to challenge management's statements and assumptions. Most investors would like to see a lot more of that on investor conference calls. It is tough on management, but I feel I knew a lot more about SGI's real situation by the time the call was over.

I remain optimistic about SGI, but consider it one of my risky investments, with a limited place in my portfolio until it can demonstrate consistent higher profit margins.

Disclaimer: I am long SGI. I will not initiate any changes in my position for the next 2 weeks.

Keep diversified!