<cite index="21-3">Since 2022, bullish leveraged ETF exposure has roughly quadrupled. Mapping that exposure to the underlying stocks shows the share tied to AI-related companies climbing from 26% to 58%. In mid-2026, that AI-related exposure was clustered in a small group of stocks. Just 10 companies accounted for two-thirds of it, including Nvidia Corp., SK Hynix and Micron Technology Inc.</cite> <cite index="21-3">The concentration is even starker among single-stock leveraged ETFs, an increasingly popular way to amplify bets on individual shares. Four major memory-chip makers alone account for nearly half of the category's exposure.</cite>
<cite index="21-1">There are elements of the growth of levered ETFs that remind observers of the meme frenzy. Even for investors who want nothing to do with levered ETFs, it is important to understand that it can still impact them.</cite> <cite index="21-2">While total global assets in the funds hover around $250 billion, measured by daily turnover — the amount of fund shares that change hands each day — they are hugely over-represented, making them one of the busiest parts of the fund universe. Their share of daily trading activity is far greater than the share of assets they command.</cite>
<cite index="24-3">If similar selling spreads across the broader AI ecosystem, passive fund flows could magnify the downturn well beyond what company fundamentals alone would justify.</cite> For architects allocating capital to AI infrastructure, concentration risk in a handful of semiconductor suppliers means volatility in leveraged derivatives can spill into wider market repricing of hardware capex and inference costs.