Two weeks ago, an SA contributor, Kingsley Park Capital, pointed out that shares of Xyratex (XRTX) have a 70% upside. I suggest KPC is being conservative. I think its opinion is priceless, but I would like to offer my financial model to support this thesis and show that the moment the market realizes what's going on, XRTX should be worth $20-$30 a share a year or so from now. Let's get started.
I have been following XRTX for the last couple of years as I cover the HDD industry very closely. XRTX sells HDD capital equipment, and following them helped me get a better feel for the CAPEX environment for the HDD duopoly. I witnessed XRTX's revenues plunge from about $455M a quarter back in 2010, to under $200M a quarter in Q1-13, a decrease of approximately 65% in revenues and more than 85% in net profits. This was mainly because NetApp (NTAP), once responsible for more than 45% of XRTX's Enterprise Data Storage Solution business, or $150M a quarter, started to source their disk arrays in-house. This trend was true with Dell (DELL) as well. I've summed up, NetApp's, Dell's and IBM's (IBM) contributions to XRTX's revenues for the past 8 quarters in the Enterprise Data Storage Solution segment.
XRTX's Large Customers In The Past 8 Quarters in EDSD Segment
By now you get the point. XRTX looked like the worst investment idea back then. The stock played along. Since the problems started, it plunged about 65% at the lowest point in late 2012. As someone who listens regularly to XRTX management's earning calls, it seemed like they were perfectly fine with this "free fall" the company experienced. The CEO and CFO both took home $1.5M in 2011. In their 2012 annual report they buried this number even deeper, but we can safely assume it was the same. The company was still profitable, but the direction it was heading in was clear. In January of 2013, Baker Street Capital Management, then with about 22% ownership, sent a letter to the board confirming this lack of initiative from the XRTX leadership and demanded action.
Baker Street bought into the company fast. Since October of 2012 they have bought almost 7 million shares in 3 months. Baker made it very clear that they would like to see some bang for their buck. Since they started buying the stock it has almost doubled in value. After a few months, Baker cleaned up the company's leadership and got 2 of their own choices as directors. The company is still in the process of finding a new CEO. One of the most interesting findings was that, the former CEO said that it only needed $40M-$50M of R&D expense to maintain its core business. At that time, the company was running at a rate of almost $100M in annual R&D expenses. That entire R&D expense was going towards the new "Growth Segment" for the HPC segment, or High Performance Computing, with the ClusterStor as the main product in this segment.
The Near Future
This all leads us to the present day. I built a financial model based on the following assumptions, most of which are sourced from XRTX earning calls:
1. HDD capital spending is about to pick up modestly in 2014, and grow more in 2015. This is based on the conservatism of the HDD companies, which saw a depressed TAM in recent quarters. They haven't invested in new equipment in a while, and they have multiple new products launching nowadays, like the 5mm HDD that is meant for tablets.
3. R&D expenses will start winding down to the rate of $80M a year, by the end of 2014. The former CEO did say they only needed $40M-$50M a year to maintain their core businesses, but he wasn't referring to the R&D needed to maintain the HPC segment.
4. Gross margins will go up, as XRTX will be able to get 45% of the gross profits from the HPC business.
5. NetApp and Dell will keep lowering their orders from XRTX, at the same rate as they did recently, until they finally stop being a customer of the company, and in-source all of their needs.
I broke down the HPC revenue out from the Enterprise Storage Solutions segment, as XRTX is not obliged to do so until it becomes 10% of total revenue. We do know it was between $10M and $15M in the last quarter, according to their CFO.
We know the assumed growth ratio, and the fact that XRTX management guides to about $60M in ClusterStor revenues in 2013. So here are the results of my model.
Segment Revenue for XRTX 2013-2014
Enterprise Data Storage Solutions
Operating Expenses For XRTX 2013-2014
GAAP EPS for XRTX 2013-2014
All numbers are in millions of USD except per share data.
Valuation and conclusion
Staying on the conservative side, I get an EPS for 2014 of $2.50, while the 3 analysts that Yahoo! Finance poll for an estimate, hold a mean estimate of $0.25 for the entire year of 2014. I believe none of them have taken into consideration the HPC segment, or the pressure Baker Street is putting to bring R&D costs down. With $90M in cash, XRTX had more than $3 a share in cash. That means the enterprise value of the share is about $7 (as of September 25th, 2013). This gives the company an EV/P of 2.8 for 2014 earnings. At the rate that XRTX can, and I think will, grow their bottom line, it is very conservative to give the business an EV/P of 8.5. But we do want to be conservative. Assuming XRTX will continue to grow their cash at a rate of $5M a quarter, we will get to $115M in cash by the end of 2014, or $4.11 a share. The EV/P of 8.5 and $4.11 in cash gets us to a share price of $25.36. Of course, the situation here has to be monitored throughout the year to make sure the assumptions get validated, or see whether any change is needed. But hey, this is what investing is all about, and this is why I'm here. Looking forward to your comments and ideas.