Inflation and war cannot stop the trillion-dollar ETF torrent! American households' "ETF faith" supports the US stock market AI computing power theme follows the index main line.

Despite the ongoing inflation and the new round of geopolitical conflicts in the Middle East sweeping through the markets since the end of February, and the occasional market volatility, Exchange-Traded Funds (ETFs) continue to attract increasingly large funds at a record pace, proving that the belief of American households and institutions in buying low and holding long-term remains strong. The growing influence of ETFs in recent years is also the core driver of the super bull market in US stocks. The massive inflow of funds into ETFs is not just a normal "risk appetite rebound", but ETFs have become the core asset allocation pipeline for American households and institutions to navigate through wars, inflation, and market volatility. More importantly, the trend in ETF investment is shifting from traditional broad-based index investments to themes related to AI computing power infrastructure. In the US stock market, ETFs that have been listed for trading have absorbed about $1 trillion in funds so far this year, reaching this milestone even before the end of June. In comparison, it took a whole year in 2024, and 10 months last year to reach the same level of fund inflows. According to Bloomberg Intelligence research report, the current rate of fund inflows is expected to easily surpass the record $1.5 trillion inflow from last year, which itself exceeded the historical peak of $1.1 trillion set in 2024. So far this year, the Vanguard S&P 500 Index ETF fund (ETF code: VOO) has been leading the way in fund inflows, attracting about $110 billion. The latest trend in ETF investments is that the investment frenzy is expanding from traditional broad-based index investments to themes related to AI computing power infrastructure. The newly launched DRAM ETF fund by Roundhill has attracted over $15 billion since April, targeting global storage chip original equipment manufacturers, covering key bottleneck areas in AI data centers such as HBM, NAND, and DRAM. This also shows that ETFs are not just important tools for buying index ETFs like the S&P 500, but are increasingly becoming investment products for rapid bets on the AI computing power industry chain. The most striking aspect of the recent ETF investment trends is that fund flows into stock ETFs centered around the US stock market have accelerated, despite the ongoing geopolitical conflict with Iran. This indicates that investors still firmly consider the US stock market as the primary destination for risk capital. However, since last week, there have been some outflows in the US ETF market, slightly reducing the total inflow to $987 billion since the beginning of the year. Nevertheless, it is worth noting that in the context of 6 months remaining until the end of the year, this category has already attracted such a significant amount of funds. The US ETF industry is not only expected to break an important record of fund inflows, but is also heading towards a record milestone in the number of new funds. According to the Bloomberg Intelligence research report, more than 600 products have started trading so far this year, at the fastest rate ever. In this mix, one of the most eye-catching fund products is the DRAM ETF fund launched by Roundhill, which provides high-risk exposure to global storage chip original equipment manufacturers, trading under the code "DRAM". It has become the fastest-growing ETF fund in terms of fund inflows, attracting over $15 billion since its launch in April. Its success has also led to the creation of multiple imitation ETF investment products. Deutsche Bank's stock analysts' forecast data shows that the demand for High Bandwidth Memory (HBM) will grow at a compound annual growth rate of about 40% by 2030, with standard DRAM storage chip corresponding growth rate of about 21%. Large-scale cloud providers such as Amazon and Microsoft are paying premiums to secure supply and signing multi-year agreements, further squeezing the market space for other buyers. "I don't want to say that this is an investment frenzy, but it is the continued growth momentum of ETFs as the preferred tool, and an extension of their use, which allows investors to grasp market niches that may not have been so easily accessible before," said Christian Magoon, founder and CEO of Amplify ETFs. In light of the above, the DRAM ETF fund launched by Roundhill has attracted over $15 billion since its introduction in April, targeting the global storage chip original equipment manufacturers with a high-risk exposure - covering key bottleneck areas in AI data centers such as HBM, NAND, and DRAM. This highlights that ETFs are not only important tools for buying index ETFs like the S&P 500, but are increasingly becoming investment products for rapid bets on the AI computing power industry chain. At the same time, more than 600 new ETFs have started trading this year, with the emerging ETF issuer Corgi launching 35 ETFs in a day and planning to introduce over 300 products within a year, reflecting the ETF industry's move to capture the hottest investment themes - AI computing power infrastructure - at a faster pace. For investors, this means that "ETFization" is amplifying two main trends: one is that low-fee broad-based ETFs continue to guide long-term funds into core assets of the US market; the other is that ETFs focused on AI computing power, storage chips, semiconductor equipment, power equipment, and data center construction themes are transforming the unprecedented "AI computing power capital expenditure" narrative of tech giants into tradable asset allocation entryways. According to Goldman Sachs on Wall Street, the global bull market around the AI computing power chain is far from over, and the market narrative has shifted from the long-standing "code-driven software light-asset software valuation expansion" of 2008 to the "re-pricing of AI computing power infrastructure around a series of physical assets". In other words, Goldman Sachs believes that the AI bull market is far from over, but is moving from the "AI chip buying frenzy" to the second phase of "large-scale construction of AI factories" - meaning that the next round of excess alpha returns will not only be limited to the strongest players in the AI GPU/AI ASIC field, but will systematically spread across the full-stack AI computing power infrastructure layer of the "AI factory", including high-performance CPUs in data centers, DRAM/NAND/HBM storage, AI PCBs, liquid cooling systems, data center optical interconnect systems, ABF substrates/glass substrates, MLCC, electronic cloth, and extensive wafer foundry. Nvidia CEO Jensen Huang last Wednesday also expressed that the AI infrastructure is expected to revitalize American factories, and artificial intelligence may usher in a new era of growth for American manufacturing and industry. US-listed ETFs have already attracted about $1 trillion before the end of June this year, while it took a whole year in 2024, and 10 months in 2025 to reach the same scale; according to Bloomberg Intelligence, the total fund inflow is expected to easily exceed last year's record of $1.5 trillion, with stock ETFs attracting over $660 billion this year, the VOO single index fund attracting about $110 billion, and becoming the first ETF to surpass $1 trillion in assets at the beginning of June. These all indicate that investors have not exited risk assets due to the Iranian war, ongoing inflation, and short-term volatility, but continue to "buy core assets in the US" through low-cost, liquid, and transparent ETFs.