SMH ETF: Balancing Profit-Taking and Long-Term Growth in the AI-Driven Semiconductor Sector

VanEck Semiconductor ETF: Time To Lock In Your Profit (NASDAQ:SMH)

I do much more than just articles at Envision Early Retirement: Members get access to model portfolios, regular updates, a chat room, and more. Learn More " I last wrote on the VanEck Semiconductor ETF (NASDAQ:SMH) in September 2022 (see the screenshot below). That article was titled Time To Step Up From QQQ. It was a while back, so I thought a little refresh of the overall background would be helpful. SMH has historically been traded at a higher P/E compared to the overall market or the tech sector represented by SPY and QQQ, respectively, due to its focus on high-growth stocks (more on this in the second section). However, the fund suffered large price corrections in late 2022 amid the global logistic network disruptions. The correction has brought SMH valuation to a discount relative to SPY and QQQ. Quote: To wit, SMH suffered a total loss of almost 30% YTD. Recent corrections have brought SMH's P/E to the current level of 17.5x, about a 14% discount from SPY's 20.2x. To add another reference point, the NASDAQ 100 index, represented by the Invesco QQQ ETF (QQQ), suffered a total loss of 22%. Combined with earnings changes, the P/E of SMH now stands at an even larger discount from QQQ (about 25.0x) by a discount of ~30%. Therefore, my previous article urged investors (many of which I assumed to have sizable positions in either SPY or QQQ) to consider a more concentrated bet on the tech stocks held in SMH. Indeed, SMH has delivered terrific returns since then. Its price at my last publication was $108. Compared to the price of this writing, the total return from SMH is over 137%, outperforming SPY by more than 100%. A working thesis tends to defeat itself eventually. As prices advance, valuation expands and risks heighten. In the remainder of this article, you will see why this is exactly what has happened to SMH over the past ~2 years or so. The valuation discount 2 years ago has now been replaced by a huge valuation premium with a P/E of around 45x. To make things worse, the premium is disconnected from SMH's profitability metrics in my view given the recent enthusiasm in AI-related stocks. QQQ is widely known for its tech-oriented exposure. Despite the reputation, QQQ actually has substantial exposure to non-tech stocks as well. In contrast, SMH is a truly pure play on technology stocks as stated in the following fund description: VanEck Semiconductor ETF seeks to replicate as closely as possible, before fees and expenses, the price and yield performance of the MVISUS Listed Semiconductor 25 Index (MVSMHTR), which is intended to track the overall performance of companies involved in semiconductor production and equipment. As you can see from the chart below, SMH allocates a significantly higher percentage of its holdings to the Electronic Technology Sector (86.28%) compared to QQQ (33.6%). While QQQ has substantial holdings of non-tech companies such as retail trade (10.15%) and healthcare (6.3%). Given its concentrated bet on tech stocks, especially those involved with semiconductors and chips, a large driver for SMH's outsized return since my last writing is the rise of AI technologies. To provide a more vivid impression of the role of AI-related stocks in this stock, the chart below shows the P/E ratios of its 4 largest holdings: Nvidia (NVDA), Broadcom (AVGO), Advanced Micro Devices (AMD), and Taiwan Semiconductor Manufacturing Company (TSM). As seen, all of them - with the possible exception of TSM - are currently trading at extremely high P/E ratios, both in absolute terms and relative terms. As seen, NVDA trades at a P/E ratio of 65.52 (on a Non-GAAP and TTM basis), AMD at 46.77, and AVGO at 36.13x, all far above the broader market either represented by the S&P 500 or the Nasdaq 100 Index. To make things even more concerning, SMH's current high valuation is disconnected from the profitability of the underlying holdings, as detailed next. The next table compares the valuation multiples and profitability metrics of SMH to those of QQQ. The data used in the table were gleaned from ETF.com and MarketChameleon on July 21. Given the ongoing market volatility, these numbers may have changed a bit when you read the article. But I think the overall picture should remain the same. As aforementioned, SMH's valuation discount relative to QQQ at my last writing is now replaced by an alarming valuation premium. To wit, SMH's current P/E ratio stands at 45.4x as seen. Compared to QQQ's P/E ratio of 33.2 (which is already among the highest levels in at least 10 years), SMH's ratio represents a 37% premium. Yet, SMH's sizable valuation premium is not supported by the profitability metrics of its underlying holdings. The fundamental reason in my view is that the market's enthusiasm for AI technologies is currently far ahead of the actual financials. For example, the average profit margin for SMH's underlying holdings is 12.9% (inferred from the fund's P/E and P/S ratios). This is lower than QQQ's 15.4% margin by 20%. The average ROE for SMH's underlying holdings is 16.7% (inferred from the fund's P/E and P/B ratios). This is lower than QQQ's 22.7% ROE by about 30%. Besides the above valuation risks, SMH is also exposed to the same set of macroeconomic risks as other large-cap index ETFs such as QQQ and SPY. Therefore, here I will focus more on the risks that are more specific to the fund itself. SMH also charges a much higher fee than most passively indexed funds. As you can see from the next chart below, its expense ratio of 0.35% is almost double that of QQQ (0.20%) and quadruple that of SPY (0.09%). In addition, SMH, with an AUM that is orders of magnitude smaller than QQQ and SPY, also tends to suffer more trade friction, another issue you may need to consider depending on how actively you trade. On the positive side, despite its fewer holdings (26 currently), SMH has a turnover rate of about 18%, actually lower than QQQ's 22% turnover rates as seen. A lower turnover rate could lessen tax headwinds. Although over the broader scheme of passively index funds, an 18% turnover is quite high (e.g., SPY's turnover is only ~2%). Another key upside risk involves the ongoing geopolitics. SMH suffered a sizable price correction off its peak levels, largely due to the market's fear about export control and presidential candidate Donald Trump's recent remarks regarding Taiwan. Things can change unexpectedly (and quickly, as reflected in the dramatic events in the past few days) over the election and the potential candidates have vastly different platforms on trade and tariffs. Such geopolitical developments can send favorable news for SMH and trigger a large price rebound considering SMH's concentrated exposure to advanced chips, a focal point of the ongoing U.S.-China tension. All told, my verdict is that the overall reward/risk profile is unfavorable. To recap, the dominant forces here are A) SMH's oversized valuation premium and B) the disconnection of its valuation premium and profitability metrics. In my last writing, I recommended investors consider a more concentrated bet on the tech stocks by replacing their QQQ/SPY holding with SMH. Given the updated considerations in this article, I now urge SMH investors to cash out and better diversify their tech exposure with other market segments.

SMH: Let Your AI Winners Run Further

Looking for a helping hand in the market? Members of Ultimate Growth Investing get exclusive ideas and guidance to navigate any climate. Learn More " VanEck Semiconductor ETF (NASDAQ:SMH) suffered downside volatility last week as investors reacted to Donald Trump's commentary on Taiwan Semiconductor Manufacturing Company (TSM). The Republican presidential nominee highlighted that "Taiwan should compensate the US for defense," potentially worsening geopolitical risks for the industry. It also coincided with reports indicating that the Biden administration is considering imposing the "most severe trade restrictions" against semiconductor companies. The move is designed to tighten China's access to the most advanced AI chips to slow its AI advances further. As a result, geopolitical concerns likely dominated investor sentiment on SMH last week, coinciding with underlying market rotation. I last updated investors in a bullish SMH article in April 2024. Given the ETF's exposure to the AI growth wave, I reminded investors that the pullback in semiconductor stocks in April shouldn't be missed. My conviction has played out accordingly, as the ETF has significantly outperformed the S&P 500 (SP500) since then. However, profit-taking on VanEck Semiconductor ETF amid potentially heightened geopolitical risks and investor reallocation shouldn't have surprised SMH investors. Accordingly, SMH notched a 1Y total return of over 57%, significantly outperforming its tech sector (XLK) peers and the S&P 500. Given the possibility of a higher-for-longer Fed, Wall Street is mixed about a sustained rotation to small caps. In addition, worries about weaker profitability and a less financially sound balance sheet could also hamper a more robust rotation to small caps. As a result, large caps and secular growth tech stocks are still expected to remain in vogue, and they don't have to rally in a "straight line." With that in mind, SMH is well-positioned to capitalize on the long-term growth drivers that benefit the semiconductor industry. As seen above, SMH's top ten holdings account for over 70% of its total exposure. Investors must pay close attention to these companies, as critical developments could shape the ETF's future potential upside. Nvidia (NVDA), TSMC, and Broadcom (AVGO) are SMH's top three holdings, accounting for more than 40% of the fund's exposure. These companies are key beneficiaries of the AI infrastructure gold rush. It's no longer a secret that AI infrastructure companies comprise critical semiconductor players vital to the AI value chain. Nvidia is the foremost AI chips designer with a stranglehold in merchant AI chips. Mizuho Securities's estimates show that "Nvidia controls between 70% and 95% of the market for AI chips used for training and deploying models like OpenAI's GPT." As a result, having an almost 20% exposure in the leader is likely justified. As my recent Broadcom article explained, AVGO is the crucial player in the AI networking space, leading the advances as Ethernet scales up in backend infrastructure. It's also leading the drive in custom AI chips as big tech and hyperscalers look beyond Nvidia's merchant AI chips to differentiate their offerings further. TSMC needs no further introduction. I enunciated in my TSM article that the foundry leader controls the market for advanced process nodes manufacturing. As a result, Nvidia and Broadcom depend primarily on the foundry leader to produce the chips they need to exert their market leadership. Notwithstanding the intensified geopolitical headwinds dominating the media narrative currently, Taiwan Semiconductor Manufacturing Company's leadership should remain intact over the next five years. Given the necessity and urgency of making available AI chips for the semiconductor supply chain, TSMC's capacity utilization is expected to be stretched through 2026. Consequently, it should extend its advantage over its foundry peers. Maintaining solid yields in more advanced nodes in semiconductor manufacturing is critical to TSMC's growth story. The foundry has the tech and anticipated yield advantage to fend off its peers keen on untethering its leadership. While the narrative on AI chips has dominated investor interest over the past year, SMH also offers diversification into other areas in the semiconductor industry. With access to Intel (INTC) and AMD (AMD), investors can gain exposure to more traditional data center CPU growth drivers. They can also capitalize on the upgrade cycle from the AI PCs refresh cycle through 2027/28. Hence, investors can expect a more robust diversification into other growth vectors in the semiconductor value chain. The trio of ASML (ASML), Applied Materials (AMAT), and Lam Research (LRCX) underpin the wafer fab equipment opportunities. These WFE suppliers provide the necessary equipment for logic and memory makers like TSMC and Micron (MU). Hence, they are critical players expected to benefit from advanced and mature process nodes in the value chain. Therefore, investors can gain broad exposure to these key WFE players through SMH. Given the semiconductor upturn we have been partaking in, investors are likely curious whether we could be near the peak of its business cycle. BofA's (BAC) analysis suggests we are still in the earlier stages of the AI-driven upcycle. Wall Street is also optimistic about the industry's short-term (forward 1Y) and 5Y earnings growth forecasts. As a result, it aligns with BofA's analysis, suggesting we are likely still in the earlier stages of a multi-year semiconductor industry growth inflection. Moreover, the industry's valuation is far from expensive when adjusted for growth. Based on the semiconductor industry's adjusted forward P/E of 30.3x, the adjusted forward PEG ratio is 0.672. Therefore, bulls could even argue that the industry seems undervalued, bolstering the buying thesis. SMH's P/E of 31x aligns closely with the industry's average. Notwithstanding my optimism, investors must not understate the risks that SMH might be vulnerable to. More intense geopolitical headwinds linked to China could hurt the earnings growth prospects of the leading semiconductor companies. Nvidia, AMD, and Broadcom "have derived about 20% of their sales from China over the past year." In addition, concerns suggest that the AI gold rush could be overstated. Therefore, execution risks must be considered as we are still in the earlier stages of the AI investment cycle. If the benefits from AI aren't as significant as anticipated, it could lead to a moment of reckoning for the AI infrastructure companies, as earnings potentially get downgraded. Consequently, the specter of considerable valuation de-rating must also be considered, possibly hurting SMH, given the industry's relatively aggressive growth projections. SMH's price action is incredibly robust, underscoring its "A+" momentum grade. It's a fundamentally strong ETF with more than 90% of its holdings assessed to have at least a narrow economic moat (according to Morningstar). Consequently, steep pullbacks have been robustly defended over the past year, with significant bottoms assessed in October 2023 and April 2024. The recent pullback from its July 2024 high has resulted in a decline of more than 12%, indicating a correction. However, given the industry's solid fundamentals and reasonable valuations, I have not assessed red flags on SMH's bullish optimism. Therefore, investors should capitalize on the recent correction to add exposure to a high-quality ETF and participate in the anticipated multi-year growth in the semiconductor industry. Rating: Maintain Buy. Important note: Investors are reminded to do their due diligence and not rely on the information provided as financial advice. Consider this article as supplementing your required research. Please always apply independent thinking. Note that the rating is not intended to time a specific entry/exit at the point of writing unless otherwise specified. Have constructive commentary to improve our thesis? Spotted a critical gap in our view? Saw something important that we didn't? Agree or disagree? Comment below with the aim of helping everyone in the community to learn better!