Smart Modular Technologies (SMOD) is a manufacturer of customized DRAM/NAND memory modules and solid state drives for OEMs such as HP, Dell, and Cisco. SMOD’s flash modules and drives are used in many IT devices, including routers, switches, servers, storage systems, printers, gaming devices, desktops and notebooks. Their core expertise is in the design and assembly of high capacity memory modules. SMOD does not own any semiconductor fabs, and purchases all its semiconductor chips from Hynix, Micron, Samsung and other suppliers. The company has a reputation for quality and quick turnaround times. There is considerable client concentration. HP, Dell and Cisco each contribute more than 10% of revenues, and the top 10 customers account for nearly 70% of revenue. SMOD sales are mainly international (70% of revenue in 2010 are from international sales), and its customers are biased towards the enterprise/industrial/defense sectors.
My initial reaction upon encountering SMOD is the knee-jerk reaction that most value investors get with IT component companies, namely fear. The IT industry is notoriously competitive and has a high rate of failure. In addition, the industry is also heavily cyclical, declining steeply during down cycles as consumers and businesses alike cut capital expenditures. The customer concentration is worrisome, suggesting that customers have substantial leverage over SMOD. SMOD designs specialized circuits for memory modules that has high capacity, is faster, and is more tolerant of heat, but is open to competition from generic memory modules that are available directly from chip manufacturers at a lower cost. With the exception of the CEO, the rest of management has kept their stock ownership of SMOD at a low and prudent level, consistently selling stock obtained through option exercises. For all these reasons, if you believe that another recession is imminent, you can stop reading right here. Without doubt, SMOD will tank in another recession (but probably will not go bankrupt). For the few brave souls willing to consider SMOD despite the above list of dangers, I will present some qualitative arguments on why SMOD’s business has a moat, and will then present some quantitative projections on SMOD’s value under different circumstances.
Firstly, it is important to clear up a confusion about the company. SMOD is not a chip manufacturer like Micron. Chips are commodities, and fabs are immensely expensive to build and run. Accordingly, chip manufacturers like Micron go through huge booms and busts. Success in the chip industry depends crucially on keeping the fab plants humming at maximum capacity. SMOD is a value-added layer on top of the chip commodities, is much less capital intensive than chip fabs, and is therefore much less prone to boom-bust cycles. I believe that this is the chief source of SMOD’s current undervaluation. In effect, investors believe that SMOD’s business will fall off a cliff at any time. As proof of SMOD’s modest capital needs, it has only $55M of long-term debt (at variable rates), while holding $93M in cash on hand. Management projects that all debt will be retired in 2-3 years. Fixed costs are relatively moderate at $3.3M in annual interest expenses (5.4% rate), plus another $2-3M in annual operating leases. In general, SMOD prices its products based on a fixed margin off commodity chip costs, and therefore has increasing profitability when chip prices rises, and vice versa. So, on superficial inspection, SMOD seems to undergo the same boom/bust chip cycle, but in actuality, is far less likely to actually go bankrupt in a recession.
Next, we move on to SMOD’s strategic moat. It is one of the largest customized memory manufacturers in the world, capable of worldwide manufacturing and distribution, and therefore one of the few which can serve similarly global companies such as Cisco and HP. There is competition, but SMOD’s combination of scale, expertise in design, and reputation for quality products is tough to beat. It has manufacturing plants situated in Malaysia and Brazil, and is well situated to serve the Asian and South American markets. In 2010, sales increased 59% largely off increased demand in Brazil. I see SMOD as a cheap and safer way to play off a secular growth in consumption of IT products in the emerging economies. Note that while SMOD concentrates mainly on the enterprise/industrial/defense markets, it has also indirectly benefited from the strong retail consumer demand pushing up chip prices. Thus far, no major chip manufacturer has indicated plans to build more fabs, so supply is likely to remain tight, a plus for SMOD. Furthermore, most tech analysts predict dramatic demand growth in solid state drives, which will also help push up chip prices, in addition to directly contributing to SMOD’s bottom line. SSDs now account for around 15% of SMOD’s sales, but that number will surely increase in the future.
Now, let’s look at a few postulated valuations for SMOD under different revenue conditions. In the following scenarios, I assume a 24% gross margin. Gross margins have gradually increased from 17% in 2006 to 24% in 2010, as a result of reduced salaries and a 26% reduction in head count effected during the financial crisis, which has led to better plant and personnel utilization. The increase in gross margins in the face of a recession is also proof of management’s competence. I also assume that the company pays off the $55M debt with cash on hand for simplicity, to avoid dealing with interest expenses. Fixed costs are the combined expenses of SG&A and R&D, at $85M per year. Tax rate is 40%. First off, we should note that sales for the past 5 years have ranged from $440M in 2009, to $840M in 2007. Net sales for 2010 came in at $700M.
Breakeven scenario. With a 24% gross margin and $85M in fixed costs, SMOD breaks even at a revenue of $355M. Note that this level is below the trough earnings of $440M reached in 2009 in the teeth of the recession, so SMOD is unlikely to incur losses. Without earnings, I will estimate the replication value of the business. SMOD has a current equity value of around $320M. Intangibles and goodwill make up only around $7M of this equity, and the vast majority of the equity value is in the value of the inventory and receivables, which together total $278M. One should also note that both inventory and receivables are quite high quality. SMOD inventory is stable at $100M despite a 60% increase in sales in 2010, indicating much faster inventory turns, and attesting to inventory quality. And most of the receivables are due from global companies which are unlikely to renege on their debt. Hence, for another company to simply replicate the underlying working capital and properties needed to recreate SMOD’s business, I estimate a minimum of $300M has to be sunk in for the working capital and manufacturing plants. This is a per share value of around $4.75. This is quite a conservative estimate. It is in fact not unreasonable to suggest a replication cost for SMOD of $360M, near its current trading price.
Modest retrenchment of sales. Let’s say sales drop 10% from $700M in 2010 to $630M. Gross profit will be $150M, reduced by $85M for fixed costs and $25M for taxes, giving $40M in net income. Given that current market capitalization of SMOD is around $370M, SMOD is selling at around 9 times reduced future earnings.
Continued growth in sales. Let’s say sales growth of IT products in China, Brazil and other emerging economies boost SMOD sales by 10%, to $770M. Gross profit will be $185M, reduced by $85M for fixed costs and $40M for taxes, giving $60M in earnings. In this scenario, SMOD is currently priced at around 6 times increased future earnings.
I expect that the current drop in DRAM pricing will reverse soon in the face of constrained supply and increased demand, and I personally value SMOD at around $500M, about 10 times trailing earnings of $50M, or a per share price target of around $8.00.
In summary, I think SMOD is one of the cheapest tech stock around. If you are interested in the likes of Micron (MU) and Sandisk (SNDK), then I definitely urge you to take a closer look at SMOD, since it is a much safer and cheaper way to play the same semiconductor trends. Still, the stock is exposed to cyclical factors, and I suggest a portfolio weighting of less than 10% to be conservative. Investors should be willing to wait out any dips in the share price, as it is entirely possible for chip pricing to decline still further before recovery.