I recently took over an existing stock portfolio for a new client and proceeded to liquidate a small cap, Taiwanese semiconductor company in favor of other tech stocks I preferred. Since the sale the stock has risen about 10% and the client emailed me wondering why I sold and what my outlook on the little company was. My answer was not as company-specific as it could have been (I knew very little about it and instead preferred to avoid small, non-U.S. chip stocks in favor of other stocks I have spent hours researching), but I did admit that I have an aversion to semiconductor stocks in general (although exceptions sometimes do present themselves).
I find the semiconductor space quite difficult to analyze and even harder to make money in as a long term investor. The industry is very cyclical, certain chips are always being replaced by a next generation product (often from a competitor), and with such high fixed costs required to manufacture chips, profit margins often rise and fall like roller coasters, making for a very volatile stock price environment. Even when you can identify solid semiconductor companies with below-average competition, in growing markets, making money on their stocks can prove quite difficult.
For example, consider flash memory manufacturer SanDisk (SNDK), a current favorite of many hedge fund managers. SNDK is a good company and with demand for flash memory soaring in recent years due to increased penetration of consumer electronics products, sales have been going through the roof. Over the last five years, in fact, SanDisk has seen its annual revenue grow more than 70% from $3.2 billion to $5.6 billion. It would be logical to assume that SNDK stock has been a great investment over that period, but you might be surprised to learn that five years ago today the shares closed at $44.50 each. Yesterday's closing price was $44.51 per share. I know this is only one example, but chip stocks can be tough nuts to crack from an investment standpoint.
In SanDisk's case they actually have done a great job at maintaining their strong position in the flash memory market, as opposed to many chip companies who often find themselves supplying Apple with a chip for one iPod only to see them be cut out of the next generation product in favor of a competing chip. The problem that SanDisk faces, as do most in the sector, is falling prices. If you have bought your fair share of memory cards, you know that every year prices drop. You can either buy the same amount of memory a year later for much less money, or you can spend the same and get a much larger card. There is no pricing power in the industry, which is great for consumers but not good for investors.
The problem is that huge demand and the corresponding unit growth that comes with it can often largely be eaten away by price erosion. Consider a market where prices drop 30% year-over-year for the same chip (not uncommon if you ever shop for digital camera memory cards and similar products). In order to keep your revenue in dollar terms steady, you need to grow units 43% per year. If you want to grow revenue, say 15%, over the prior year, you need to ship 64% more units! SanDisk actually has been fortunate that demand for flash memory has been so strong, as other areas within the chip space have not been nearly as robust.
So while I agree with many smart money managers who have been accumulating the stock that SanDisk is a good company that is serving a growth market, and that its stock does appear to be cheap, I do not share the same optimism about its long term prospects as an investment. It is just really hard to sustain stock price appreciation in an industry with these types of market dynamics. While there are certainly plenty of success stories within the semiconductor stock universe, I suspect for every long term stock market winner there are five or ten big losers, and I personally do not care for those kinds of odds.
Full Disclosure: No position in SNDK but positions may change at any time