The continued surge for chip stocks may be a reminder for investors to check how the exposures of their portfolios have changed, even for those who bought exchange-traded funds specifically to increase their semiconductor holdings. Semiconductor stocks have been on a tear for more than a year, led by Nvidia , which has emerged as a leading stock in the excitement around artificial intelligence. That has helped the S & P 500 enter a new bull market and made the chipmakers an increasingly influential part of the major market indexes. “The % weight of the Semiconductors within the S & P 500 is now at nearly 9%. That’s up from 2% a decade ago, and easily the largest gain for an industry group. NVDA has jumped from .05% in 2014 to 3.6% today,” Strategas ETF strategist Todd Sohn said in a note to clients. That growth means Nvidia and chip stocks in general are now a much bigger holding for investors who hold passive funds that track the S & P 500 or other broad indexes that are weighted by market cap. For those who want to add more exposure to semiconductors, the concentration of funds gets still more complicated. The five biggest non-levered sector ETFs for chip stocks all track different indexes, and the weights of Nvidia range between roughly 3% and nearly 22%. The differences in exposure can be caused by how the indexes are created and how often they rebalance, as well as fund-specific rules around concentration. Those choices appear to be driving big performance differences in January. For example, the best-performing fund in the category so far this year is the VanEck Semiconductor ETF (SMH) . The ETF tracks the MVIS US Listed Semiconductor 25 Index , which has a modified market cap weight approach. The index rebalances twice a year and will cap the biggest stocks at 20% at that time, but Nvidia’s rally has pushed its weight to 21.8% in that index for now. On the other hand, the S & P Semiconductors Select Industry Index followed by the SPDR S & P Semiconductor ETF (XSD) takes a modified equal weight approach, putting the weight of Nvidia at 3.2%. No one approach to index construction is guaranteed to be the winning formula in any particular period. For example, if other chip stocks start to catch up to Nvidia, then the SMH will likely underperform its rivals instead of pulling away. Just as the weighting differences can affect performance during big rallies, they can also matter during pullbacks. Sohn said trading volume in a major leveraged semiconductor fund is showing signs of exhaustion. “Notably, leveraged long Semi ETF volume ( SOXL ) is accelerating and the group’s rolling % change vs. the index is in extreme territory. Momentum isn’t something to stand in front of, but perhaps these are stats to keep on the back-burner,” Sohn said.
