Morgan Stanley has warned that US stocks could face their sharpest correction since the Iran war as rising artificial intelligence financing costs, uncertainty surrounding the US midterm elections, and tighter financial conditions threaten to undermine the market rally.
The Wall Street bank expects the S&P 500 to fall as low as 7,100 in the near term before recovering to end 2026 at 8,000. A decline to 7,100 would represent an approximately 8% drop from the index’s recent levels, while a subsequent recovery to 8,000 would imply a 12% gain from that projected trough.
The forecast reflects a growing divide between headline stock-market performance and the underlying health of the broader equity market. Although major US indexes have remained close to record highs, gains have become increasingly concentrated in technology stocks, particularly companies benefiting from investment in artificial intelligence.
Register for the next Tekedia Mini-MBA.
Register for Tekedia AI in Business Masterclass.
Join Tekedia Capital Syndicate and co-invest in great global startups.
Mike Wilson, Morgan Stanley’s chief US equity strategist, said the bank believes equities have been undergoing a correction since June, even as the headline indexes have concealed weakness across individual stocks.
According to an analysis by The Kobeissi Letter, approximately 84% of S&P 500 constituents were trading at least 10% below their respective all-time highs, suggesting that much of the market has already experienced significant declines.
“Bottom line: We think the correction that began in June is now in its final stages, with much of the damage already done beneath the surface,” Wilson wrote.
A market supported by a relatively narrow group of large technology companies can appear resilient even as financing pressures, weaker market breadth and rising borrowing costs undermine other sectors. Morgan Stanley expects several developments over the coming weeks to test whether the remaining leaders can continue supporting the broader index.
The first major pressure point is the artificial intelligence investment cycle, where investors are increasingly questioning both the cost of infrastructure expansion and the returns that companies can generate from it.
Concerns intensified after reported third-quarter revenue figures from Anthropic indicated a sharp slowdown in growth compared with the previous quarter. The figures have contributed to a wider debate over whether AI companies can sustain the spending required to expand computing capacity while generating enough revenue to justify their investments.
The market is caught between two competing risks. A slowdown in AI capital expenditure could weaken demand for semiconductor manufacturers, data-center operators and other infrastructure suppliers. Continued spending, meanwhile, could increase borrowing requirements and deepen concerns about whether the industry is generating sufficient returns on its expanding asset base.
Financing conditions are adding to the pressure. Credit spreads on data-center-related bonds have widened in recent months, reflecting higher borrowing costs, greater investor caution about AI-related debt and a growing supply of new securities as companies seek to finance infrastructure development.
When investors demand higher compensation for lending to AI infrastructure companies, financing becomes more expensive. That can reduce the profitability of projects, increase the burden of servicing existing debt, and make it harder for companies to raise fresh capital on favorable terms.
CoreWeave has become a focal point for these concerns after its debt moved into distressed territory in October, raising questions about the financial resilience of companies whose growth depends on substantial capital investment and continued access to credit.
The potential consequences extend beyond individual borrowers. If investors begin demanding higher yields across AI-related debt, companies with large financing needs could face more expensive refinancing, while lenders and investors exposed to the sector could become more cautious about extending credit elsewhere.
Morgan Stanley expects these concerns to converge over the next three to four weeks as markets absorb a sharp increase in debt issuance. Strategists said the semiconductor sector could find support around mid-November, after falling another 10% to 15%, as financing concerns begin to reach a low point.
It is all happening when semiconductor stocks have become central to the AI investment thesis. Their valuations depend not only on demand for computing power but also on customers’ willingness and ability to continue financing new infrastructure.
If funding conditions deteriorate further, the pressure could spread from highly leveraged data-center operators to chipmakers and other companies whose revenue expectations are linked to continued AI spending. Conversely, stabilizing credit markets and evidence of sustained demand could help restore confidence in the sector.
Impacts of Midterm Elections and Higher Yields
Morgan Stanley’s second concern is the approaching US midterm elections, which historically have coincided with periods of subdued stock-market performance.
Goldman Sachs has found that the S&P 500 delivered a median return of zero between August 1 and election day in November across midterm years since 1974.
However, the historical pattern has generally improved after the vote. Morgan Stanley’s analysis of market data from 1950 to 2022 found that the index gained an average of 8% in the three months following midterm elections and 15% over the subsequent six months.
The bank argues that the removal of political uncertainty can support equities regardless of the election outcome. Investors may become more willing to commit capital once the immediate uncertainty surrounding the composition of Congress and the policy outlook has passed.
“Markets also loathe uncertainty and so just having it behind us can provide a positive catalyst in and of itself no matter what the outcome,” Morgan Stanley strategists said.
That historical tendency could provide support if the market experiences a sharp decline before the election. But past performance does not guarantee a similar outcome, particularly if credit-market stress, energy-price shocks or monetary tightening overwhelm the usual post-election recovery.
The third pressure point is the tightening of financial conditions. Wilson highlighted the Federal Reserve’s return to rate increases in September, slower growth in the central bank’s balance sheet and the economic consequences of higher energy prices.
Rising Treasury yields can tighten financial conditions even without additional increases in the federal funds rate. Higher yields raise financing costs for companies and households, increase the discount rate investors apply to future corporate earnings and make relatively safe government bonds more competitive with equities.
Higher energy prices compound the problem by increasing production and transportation costs, reducing household purchasing power and potentially keeping inflation elevated. The combination could weigh on economic growth while making it harder for the Fed to ease policy.
Wilson said the market was effectively tightening financial conditions on the Fed’s behalf. He argued that, together with the increase in energy prices, this could slow growth and bring inflation back towards the central bank’s target. However, the same process could generate financial stress severe enough to force policymakers to reverse course before inflation is fully controlled.
Such a reversal could initially support asset prices by easing borrowing conditions and improving liquidity expectations. But it would also signal that economic or financial strains had become sufficiently serious to warrant a policy response.
Investors were awaiting the Fed’s latest interest-rate decision on Wednesday, with markets pricing in a 90% probability that policymakers would leave rates unchanged, according to the CME FedWatch tool. Further rate increases were nevertheless seen as possible through the end of 2026 and into early 2027.
However, it is not clear if the Fed’s decision and accompanying guidance will bolster expectations of prolonged monetary restraint or offer investors a reason to anticipate relief.
Morgan Stanley’s outlook indicates that the market’s next direction will depend on more than the performance of its largest technology companies. AI investment must withstand closer scrutiny of its financing and returns, political uncertainty must pass without a lasting deterioration in sentiment, and bond yields must avoid tightening financial conditions enough to damage growth.
The bank still expects the S&P 500 to finish 2026 at 8,000, indicating that it sees the potential correction as a setback rather than necessarily the end of the broader rally. But its projected decline to 7,100 highlights the vulnerability of a market whose headline strength has increasingly depended on a narrow group of stocks while borrowing costs, energy prices, and AI-related debt risks build beneath the surface.