On the WhoTrades Marketplace, we’re featuring a trader that focuses exclusively in semiconductor stocks. The portfolio is straightforward, with just two stocks. $AMD takes up 55% of the portfolio while $MU comprises 45%. It’s returned 75.2% over the past year with just a 16.2% drawdown. Semiconductors have been one of the best-performing sectors over the past 2+ years. Since Feburary 2016, the PHLX Semiconductor Index (“SOX index” as charted below) is up close to 120%. There have been few down months along the way. Is this a sustainable trend or is the semiconductor trade getting long in the tooth? The semiconductor industry is tough to generate high rates of return on capital over time because the technology tends to quickly become commoditized. Historically chip companies have invested a lot in building factories. Then the basic strategy was to have them generate at maximum capacity in order to spread the high fixed cost over a high number of units to generate a profit. Even though demand has been growing heavily over the decades, this tactic has frequently led to massive oversupply issues (especially since competition has been so high with so many firms trying to take advantage of booming demand). So companies either cut back on their R&D (or get out of the industry) and eventually this lead to undersupply and the cycle repeats itself. These dynamics have favored consolidation in the industry – see $QCOM, $NXPI, $AVGO, plus the $AMD rumors. Consequently, you end up with a few large players who generate positive returns through scale, and certain niche players who try to compete through node scaling, or by offering differentiated products. Leaning toward focused – rather than diversified – portfolios is generally better for smaller firms as they can more realistically achieve a higher share of the market. This will tend to generate cash flow (assuming their tech is commercialized successfully) and benefits R&D, which in turn helps perpetuate market power. Semiconductors are like a commodity industry. The best time to enter is generally when a lot of capacity has been shed from the market and you expect there's a catalyst to stoke demand. The time to get out of these companies is generally when there's a lot of expansion planned, which eventually leads to oversupply in the relevant markets they're involved in. Because it means that earnings will eventually start falling. Markets will tend to anticipate this before it actually happens, even when earnings are still strong or near their peak. Whither the Semiconductor Trade? Whether to invest in semiconductor stocks is a decision to be made based on one’s individual circumstances. Suffice to say, the latest move from February 2016 to the present won’t be as good as the next 2+ years, but some of these companies can be thought of from a long-term perspective. For example, $NVDA and $AMD are both involved in the future of certain forward-thinking technologies, such as machine learning, AI, and autonomous driving. $MU’s thesis is more geared toward the future of the NAND and DRAM market (discussed more in this post), which looks stable through 2018, but could become dicier in 2019 as China begins to bring supply onto the market. For those who trade or invest in the semiconductor sector, this trader may be worth following. Moreover, there are thousands of traders on the Marketplace where you can check out strategies of interest, track their progress over time, and generate trade ideas for your own portfolio.