Semiconductor ETF Trends & Picks; Mid-Week Market Recap

In partnership with

Your IRA, made to order

Choose where and when you want to retire, and a Betterment IRA can help make your money hustle all the way there.

Learn more

Mid-Week Market Recap

Stocks have had some success in carrying forward last week’s post-Fed rally into this week so far, but it’s really been a mixed bag.

Tech and cyclicals are still leading the way this week, which is a reflection of the bullish mood with which investors are viewing the Fed’s rate cut. Growth is leading value, but high beta & low volatility are showing little separation. Small-caps are also trailing the S&P 500 by nearly 2% on the week, which is a divergence from what we saw last week. Small companies are generally expected to see a greater benefit from rate cuts, but that’s not being reflected in stock performance right now.

Treasury yields on the long end of the curve continue to inch higher as the rally feeds into optimism that the soft landing can be achieved. I’m not sure that optimism is necessarily warranted longer-term, but the stimulus package announced by China could ultimately inflate commodity prices, restart inflation and eventually lift interest rates higher. With China in the equation, there are suddenly a lot more moving parts.

Semiconductor ETFs: Trends & Picks

Since the start of 2023 and the advent of the AI boom, semiconductor stocks have been one of the best ways to play it. Well, one stock in particular, NVIDIA. Right now, chip stocks are more than fully valued, but it’s all about potential. It looked like they were set to consolidate over the past couple months with the Fed heading into its rate cutting cycle, but investors seem confident that the soft landing will become a reality. In that event, this group may have some more life in it.

Semiconductor ETFs have drawn more than $10 billion in new money this year, making them some of 2024’s most successful performers and money draws.

Therefore, you might be surprised to see that ETF performance has actually been very mixed. You won’t be surprised though to hear that performance has been directly tied to its exposure to NVIDIA. A lot of these funds have weighting and selection strategies that de-emphasizes the mega-cap names and that actually has most of them actually underperforming the S&P 500 year-to-date.

For as much as investors have enjoyed the magnificent 7 stocks over the past couple years, it’s created issues for those looking for a little more diversification.

Let’s take a look at a few ETFs getting attention.

VanEck Semiconductor ETF (SMH)

SMH has been the unquestioned top performer in this group and it’s not even close. The reason for this is that it tracks a simple cap-weighted index that overweights the biggest companies. That ends up giving NVIDIA a 20% weighting in this fund. Add in allocations to Taiwan Semiconductor and Broadcom, you’ve got 40% of the fund accounted for.

The biggest problem with SMH is its concentration. Outside of just the top-heaviness of the portfolio, it only holds 25 stocks. That really requires it to hit the market’s sweet spot in order for it to excel. When large-caps are rolling, this fund probably will too. If not…..

Outlook: Depends on what you think of NVIDIA. The stock trades at about 44 times forward earnings, so it’s already priced for perfection, but with so many companies investing billions of dollars in AI development, demand will remain robust for the foreseeable future.

Large-cap concentration hasn’t been an issue lately, but it could be longer-term. Funds with 40% of the portfolio is just three stocks generally won’t rate highly in my opinion.

iShares Semiconductor ETF (SOXX)

SOXX builds its portfolio very similarly to SMH. It concentrates on 30 stocks instead of SMH’s 25, but it also cap-weights the resulting components. The big difference is that SOXX caps the individual weights of its top 5 holdings at 8% and remaining weights at 4%. Therefore, the bell curve on individual holdings in SOXX is flatter and more spread out.

That’s impacted performance though to the tune of a staggering 20% gap between it and SMH. In the end, I probably prefer this ETF to SOXX because I appreciate that it’s a little more balanced out. From a standpoint of overall risk, they’re pretty similar, but the relatively greater better diversity should help cap some of the more extreme highs and lows.

Outlook: Probably better built for the long-term. It’s also worth noting that this fund tilts a little heavier towards some of the “old school” names, such as Intel and Texas Instruments. Again, if you think mega-caps or NVIDIA, specifically, are due for a pullback, this is probably a better choice.

XTrackers Semiconductor Select Equity ETF (CHPS)

CHPS debuted about a year with two potential differentiators. With an expense ratio of just 0.15%, it instantly became the cheapest fund in this sector. It also layers on an ESG screen to eliminate some of the biggest offenders. In the end, the correlation between CHPS and the other major semiconductor ETFs is still very high and overlaps are pretty similar.

This is another example of how an ETF comes into the space late, tries to compete on cost, but fails to gain any traction. After more than a year, it’s garnered just $8 million in assets and could be at risk of shuttering at some point in the future. From a tradeability standpoint, the low expense ratio is currently being offset by higher trading costs due to its small size, so the low cost advantage hasn’t turned out to be an advantage.

Outlook: The perceived differentiators haven’t really been differentiators. The portfolios are similar enough to SMH and SOXX that they’ll likely perform very similarly. On total cost basis, CHPS holds no real advantage and I don’t think investors really care about ESG as much as the market thought they might.

VanEck Fabless Semiconductor ETF (SMHX)

VanEck already has SMH, so it wouldn’t seem like they need another semiconductor ETF, but this one is kind of interesting. The idea of “fabless” is that it targets companies that design and develop semiconductor chips, but outsource the actual manufacturing of them. These companies are likely to have much lower operating costs and could be a better pure play exposure to the semiconductor space.

The fund is only a month old, so it’s far too early to come to any judgments about it. The strategy, however, is logical (perhaps even better in chip sector targeting) and it’s competitive on cost. SMHX may get drowned out by SMH, but I’m not sure it should.

Outlook: I can see this easily becoming a $500 million, but it might take some work to get there. The term “fabless” won’t be intuitive to most investors and they’ll need to be educated on why it could be advantageous compared to traditional semiconductor ETFs.

If it can get a little scale and bring its trading costs down, I like how this one is positioned.