Taking a look at the semiconductor sector
Fellow Senior Quantitative Analyst here at Schaeffer’s, Chris Prybal, noted that the VanEck Semiconductor ETF (SMH), the most liquid semiconductor ETF, fell below its 1,000-day moving average for the first time ever. That didn’t even happen during the coronavirus crash in 2020. It’s been argued that semiconductors are the new transportation stocks. In other words, the sector is a good measure of the health of the overall economy and can signal bullish or bearish times ahead.
This week I’m looking at the semiconductor sector to see how weak it is compared to the broader market and if it has historically indicated where things are going.
SMH Relative Strength as an Indicator
SMH began trading near the end of 2011. Going back to then, I calculated the six-month relative strength of the semiconductor ETF versus the S&P 500 Index (SPX). This tells us to what extent the SMH is outperforming or underperforming the broader market. Currently, the six-month relative strength is in the bottom 10th percentile going back to 2012. In short, semiconductors have been underperforming bigtime.
The table below shows the one-month performance of the S&P 500 based on the relative strength of the semiconductor sector. When the SMH relative strength was in the bottom 10th percentile, the S&P 500 has done poorly over the next month. The index has averaged a loss of over 1% and positive barely 40% of the time. When the semiconductor sector has outperformed by a large amount (90th percentile), the S&P 500 averaged a return of 2.26% over the next month with 80% of the returns positive. SMH performance seems to be a legit indicator for the broader market over the next month.
The table below shows the three month returns of the S&P 500 based on the SMH relative strength. Again, the big underperformance from SMH has foreshadowed poor returns for the S&P 500. When the SMH relative strength was in the bottom 10th percentile, the S&P 500 has averaged a gain of 1.56% over the next three months with 62% of the returns positive. When SMH outperformed by a lot over the prior six months, the S&P 500 averaged a gain of 4% over the next three months with 84% of the returns positive.
Looking six months into the future, the SMH performance versus the S&P 500 means less, especially the underperformance like we’re seeing now. The table below shows when SMH underperformed against the broader market, the returns going out six months hasn’t been much different than when it has moderately underperformed or outperformed. Big outperformance from the semiconductors still tended to lead to bigger returns going forward but that’s not our current situation.
SMH After the Signals
Finally, I did the same analysis as above but looked at the return for SMH after the signals instead of the S&P 500. When SMH has underperformed the S&P 500 by a lot, like now, it tended to underperform over the next month as well averaging a loss over the next month with less than half of the returns positive. When the SMH six-month relative strength was positive by a lot, I tended to do well over the next month.
Like the S&P 500 analysis above, SMH six-month relative strength has been a decent indicator for returns over the next three months. Going out to six months, however, it means very little.
In conclusion, the semiconductor index has been a dependable indicator for stocks going forward. When SMH, the semiconductor ETF, underperforms by a lot over a six-month period, it has paid off to bet on downside for stocks over the next few months. When SMH has beaten the S&P 500 by a lot over a six-month period, it has made sense to bet on further gains.