The significant demand uncertainty in the end markets of semiconductors has resulted in a challenging environment for managing inventories. Semiconductors are likely to underperform OEM clients in the fourth quarter - an area of concern. While Altera (ALTR) has a strong brand name and plenty of opportunities in a dynamic technology sector, I find that it is fairly priced at the present moment. It manufactures PLDs, ASICs, IP cores, development software, etc. and has major catalysts in 40 nm and 28 nm. In the last two areas, I even anticipate it taking away market share from its main competitor Xilinx (XLNX). Over the last twelve months the stock has appreciated by 13.3% versus 24.3% for Xilinx.
From a multiples perspective, both appear cheap. Altera trades at a respective 14.3x and 16.8x past and forward earnings; Xilinx at a respective 14.9x and 16x past and forward earnings. These multiples are at the lowest-end of reasonably similar companies that have a median PE multiple of around 23.8, more or less evenly distributed from 15 to 31. Even still, given expectations for stagnant EPS growth at least over the next three years, I recommend, again, holding out until a more attractive entry point. The dividend yield of 0.86% at Altera and 2.31% at Xilinx also do not offer the kind of defensive play that would justify a limited upside. By contrast, as I underlined here, I anticipate Intel (INTC) being both a strong and safe investment - an exception to many technology companies.
With that said, Altera is very liquid and has a net cash position of $2.7B, representing 22.9% of market value. Xilinx has net cash representing 10.5% of its market value. Analysts currently rate the former more of a "buy" and the latter more of a "hold."
At the third quarter earnings call, Altera's chairman and CEO, John Daane, noted disappointing performance:
We had expected broad-based end market growth for the third quarter, but instead saw customers aggressively reduce inventory due to concerns of the global economy, along with the downturn in some end businesses. In our Q3 update, we revised guidance downward due to non-Asia communications weakness, as well as broad military test and industrial softness. Late quarter softening in communications in Asia resulted in a close slightly below the updated guidance range.
In the third quarter, telecom wireless revenue decreased 13% sequentially, with broad-geographic declines in both markets. Industrial military automotive decreased 7%, with automotive up strongly, but industrial and military down. Computer, storage and networking, as the recipient of a majority of the short-term ASIC replacement business, increased 31%, other declined 11%. We had one customer represent 12% of revenues in the quarter.
To top it off, management has conservatively guided for 7-11% sequential decline. This range is actually better than what some feared given the volatile demand backdrop. Wireless, communications, and computer networks are expected to decline in the next few quarters, but will be partially offset by automotive industrial military. Strengths are likely to include gains in FPGA and 40nm, the latter of which will enable premium prices. But is this enough to make it a value play?
It is also questionable whether or not PLDs truly represent a catalyst. For the last year, they have not performed well and have arguably detracted from focus elsewhere. Perhaps, however, new products represent a catalyst for Altera - this segment was up 43% sequentially and remains critical for value creation. I am particularly optimistic about how new products can rein in top-line volatility, which, in my view, is a chief concern for management. Another positive for Altera is that it has been successful in cutting costs and the opportunity for premium pricing in 40 nm will help optimize increases to scale. Nevertheless, customers are unsure about future demand and are thus scaling back on inventories, in turn, causing supply problems for semiconductors.
Consensus estimates for EPS are that it will decline by 4.8% to $2.37 in 2011, decline by 6.8% in 2012, and then increase by 16.7% in 2013. Of the 28 revisions to EPS estimates, 27 have gone down. I model the top line growing by 7.5% to $2.1B and then by 2.4% and 13.1% more in the following two years. Assuming a 17x multiple and a conservative 2012 EPS of $2.18 yields an intrinsic value that is roughly in line with the current market valuation.