NextFin

Short Bets Against US Equities Hit Record as AI Risks Mount

Summarized by NextFin AI
  • Bearish positioning against U.S. equities has reached record levels, indicating skepticism about the AI spending cycle's ability to generate profits.
  • The market is increasingly questioning whether the capital invested in AI will translate into free cash flow, leading to a concentration of short bets against a narrow set of AI-linked stocks.
  • Shorting is driven by concerns over valuation pressure and capital intensity, as the market grapples with the timing of monetization in the AI sector.
  • Investors are not doubting AI's importance but are concerned about the timetable for returns, suggesting a structural issue rather than a cyclical one.

NextFin News - Bearish positioning against U.S. equities has reached a fresh extreme just as the AI trade that lifted the market for two years is being judged less like a one-way growth story and more like a test of earnings conversion. The question underneath the latest wave of short bets is simple: can the spending cycle behind artificial intelligence keep outrunning the profits it is supposed to create? The market is increasingly treating that answer as uncertain.

That tension sits behind a report that hedge-fund short bets against U.S. stocks have climbed to record territory. The point is not that investors have suddenly decided the AI theme is finished. It is that the market is no longer willing to assume every dollar spent on chips, cloud capacity, power, and data-center buildout will translate quickly into free cash flow. That shift matters because the U.S. benchmark has become increasingly dependent on a narrow set of AI-linked leaders, and that concentration gives skepticism more leverage than it would have in a broader market.

In practice, shorts are rising for the same reason crowded long trades eventually run into trouble: the market has moved farther than the evidence. When a handful of megacap stocks dominate index performance, a disagreement about future earnings becomes a disagreement about the whole benchmark. The bearish bet is not only a directional view. It is also a hedge against multiple compression, breadth deterioration, and the possibility that the market has capitalized the AI payback too early.

The basic mechanism is easy to miss if the story is reduced to a simple bull-versus-bear narrative. AI is creating real demand for compute, networking, power, and cooling. That is the first-order effect. The second-order effect is that the capital needed to serve that demand arrives today while the revenue validation often arrives later, and sometimes much later. If investors begin to suspect that the lag is longer than the stock market assumed, the valuation of the leaders can come under pressure even if the technology itself continues to improve.

That is why the latest positioning looks more structural than purely tactical. The shorting surge itself is cyclical: it can reverse quickly if earnings surprise on the upside or if a softer macro backdrop pulls discount rates lower. But the underlying question about the timing of monetization is structural. The market is grappling with whether a technology that is clearly important to the economy is also important enough, right now, to justify the cost of the current buildout.

Put differently, the market is not shorting AI because it doubts that AI will matter. It is shorting because it doubts the timetable.

Why Shorts Are Rising Now

The recent rise in bearish positioning makes sense only if crowded leadership, valuation pressure, and capital intensity are viewed together. The market has spent much of the past two years paying up for companies tied to the AI supply chain, but the newest phase of the cycle asks a harder question: what happens when the investment phase becomes too visible to ignore?

That question is not theoretical. The Federal Reserve’s July 2026 Monetary Policy Report said inflation remains elevated relative to the central bank’s 2% objective, in part because supply shocks have pushed up prices in some sectors, including energy. Sticky inflation is not an AI story by itself, but it matters because a higher-for-longer rate backdrop makes investors less forgiving of long-duration cash flows. If the discount rate stays elevated, the market has to demand faster payback from expensive growth stories.

The short side also tends to build when a trade becomes too concentrated for comfort. That is the political economy of markets as much as their math. A narrow index can keep rising even while breadth weakens, because passive flows and momentum allocations keep rewarding the same winners. But once investors decide that the crowding itself is the risk, hedging demand can snowball. Shorts then serve two functions at once: they blunt market exposure and express skepticism that the same leaders can keep carrying the benchmark indefinitely.

The market has seen versions of this before. The internet buildout created real utility but asked investors to fund infrastructure before profits emerged. Telecom laid fiber before demand fully matured. Cloud migration required years of upfront spending before recurring revenue became obvious. AI does not need to be a bubble to create the same problem. A real technological shift can still produce a bad equity entry point if the market pays too much for the first stage of the cycle.

That is the short-term versus long-term split. In the short run, record bearish positioning is a cyclical statement about crowded trades and a potentially overextended benchmark. In the long run, the issue is structural: whether AI investment turns into a durable profit machine or remains a capital-intensive buildout whose returns arrive more slowly than the market wants.

“The big story in 2026 in labor will be AI,” said Joseph Briggs, who co-leads the global economics team in Goldman Sachs Research. “If we see some job losses pulled forward, that sets stage for potential underperformance relative to our forecast, and that may lead the Federal Reserve to cut rates.”

That comment captures the market’s broader dilemma. AI is already affecting parts of the labor market, but the transition is uneven. Goldman Sachs Research says the technology is already being felt in parts of the tech sector, and its base case assumes it could take about 10 years for firms to adopt AI on a wide scale, with 6% to 7% of workers displaced during the transition. The stock market is being asked to price that future now, even though the payoff curve is still unfolding.

The stronger the concentration, the more dangerous that gap becomes. If the biggest AI beneficiaries keep reinvesting heavily while the revenue curve remains slower than expected, the market can punish multiples even without a recession, even without a collapse in demand, and even while the technology keeps improving. That is the second-order risk. The first-order story is innovation. The second-order story is valuation.

What The Market Is Really Pricing

The consensus view is still that AI is a long-term earnings driver. That is exactly why the short-interest story matters. A clean consensus bearish call would not be interesting; what is interesting is that investors appear willing to stay exposed to the AI winners while hedging against the possibility that the earnings conversion rate disappoints. That is not a vote against the technology. It is a vote against the market’s confidence in the timetable.

What seems priced today is not a collapse in AI spending. It is a slower and more contested return on that spending. The market is effectively asking whether hyperscalers, chipmakers, and the broader AI supply chain can keep expanding capacity at the same pace without a matching lift in monetization. If the answer is no, the pressure does not stay confined to one sector. It spreads through index weights, factor models, and passive ownership channels that have amplified the same winners for most of the rally.

Goldman Sachs Asset Management’s July 2026 market commentary still said it remained constructive on global equities, particularly in the U.S. and emerging markets, because AI investment continues to support corporate earnings. That is the counter-thesis in its cleanest form: the buildout is still early enough, and earnings are still strong enough, that the market should continue to reward exposure to the AI stack. Under that view, elevated shorts are fuel rather than warning.

That counter-thesis is not easy to dismiss. It argues that the market is treating capex like a cost when it should be treated like an investment, and that the current criticism simply reflects the usual lag between the start of a new industrial cycle and the appearance of visible profits. On that reading, bearish positioning can persist for a long time without being right, because a strong earnings cycle can keep forcing shorts to cover.

But the burden of proof still sits with the bulls. The falsifying signal for the cautious view is concrete: if the largest AI spenders keep raising capital spending while operating income, free cash flow, and AI-linked revenue fail to accelerate over the next two reporting cycles, then the bearish positioning will look less like a crowded trade and more like a correct read of the capital cycle. The market does not need AI to stop working. It only needs the profit curve to be flatter than the price curve.

That distinction matters because it changes how the next phase of the market should be read. If the AI trade remains strong but less explosive, the broad index can still survive while leadership rotates. If the monetization gap keeps widening, the market may move from paying for potential to paying for proof. That is the real risk behind the record shorts.

Who Wins If The Trade De-Risks

In the short term, heavier short interest can make the market more fragile. A crowded bearish trade can unwind violently if earnings stay strong or if rates drift lower, because covering shorts creates its own demand. That means the next few market moves may say more about positioning than fundamentals. But over a medium horizon, the more durable winners are likely to be companies that can show earnings without relying on the same level of capital intensity.

The exposed names are the ones closest to the market’s most expensive narrative: megacap platforms that are still spending heavily on AI infrastructure, semiconductor suppliers whose valuations assume long demand runs, and broad equity benchmarks whose returns are still dominated by a narrow leadership group. If AI keeps growing but at a slower profit rate, the damage may show up less in technology adoption than in the multiples investors were willing to pay for it.

There is a broader rotation risk here. A de-risking of the AI trade would not automatically mean a bear market for U.S. stocks. It could mean breadth improves while headline returns become less dependent on a small cluster of leaders. That would be healthier over time, but it can look like a correction while it is happening because the benchmark has become so concentrated.

The base case is that short bets remain elevated because investors are still uneasy about concentration and timing, but they do not trigger a broad reversal unless the next earnings cycle confirms slower monetization. The upside case for bulls is that capex stays high, revenue keeps accelerating, and the market decides AI is still early. The downside case is sharper: if the largest spenders keep leaning into capital expenditure while margins flatten, the market may conclude that the buildout is real but the equity returns are already behind it.

Short term, the tape will still be ruled by positioning and liquidity. Medium term, earnings will decide whether AI is a compounding machine or a capital-intensive story that needs more time. Long term, the market will keep rewarding the companies that turn intelligence into durable cash flow rather than those that simply spend faster than everyone else.

The market is not shorting AI because it doubts the technology. It is shorting because it doubts the timetable.

Explore more exclusive insights at nextfin.ai.

Insights

What are the core principles behind short selling in stock markets?

What historical events have influenced the current state of short selling in U.S. equities?

How has the rise of AI technology affected investor sentiment towards U.S. stocks?

What recent trends have emerged in bearish positioning against U.S. equities?

What was highlighted in the Federal Reserve's July 2026 Monetary Policy Report regarding inflation?

How might the investment cycle in AI impact long-term stock market performance?

What challenges do investors face when assessing the profitability of AI-related investments?

What are some notable comparisons between the current AI investment landscape and past technological shifts?

How do concentration risks in stock indices affect market dynamics?

What factors might lead to a reversal in short selling activity in the market?

How does the current state of the U.S. equity market reflect investor confidence in AI technology?

What implications does high short interest have on market volatility?

What are the potential long-term impacts of AI investment on labor markets?

What is the significance of the current bearish bets against AI-linked stocks?

What role does capital intensity play in the valuation of AI companies?

What are the potential consequences if the earnings from AI investments do not meet market expectations?

How do investor strategies differ in crowded long positions versus crowded short positions?

What historical case studies illustrate the risks of investing heavily in emerging technologies?

How might future developments in AI technology influence financial markets?

Search, ask or analyze anything...