For most of October the AI selloff was a semiconductor problem. Friday it became a software problem, and that distinction matters because software is where the highest multiples live.

The numbers are blunt. AI-focused software shed about $260 billion in value on October 16 after a Goldman Sachs note raised alarms about debt financing. The S&P 500 fell 1.09% to 7,659.44, the Nasdaq Composite plummeted 2.15% to 26,140.03, and the Dow slipped 0.34% to 52,002.41. The Nasdaq booked its worst week since June, falling 4.50%, as investors grew more skeptical that heavily debt-funded AI expansion can stay profitable while borrowing costs remain elevated.

The Catalyst: Credit Before Equity

This rotation started in bond markets, not stock screens. US dollar high-yield AI-related debt spreads widened by about 25 basis points between October 13 and 14, reaching around 360 basis points over Treasurys. By contrast, non-AI-related debt spreads were only marginally wider. Goldman’s Friday note made that credit divergence impossible to ignore for equity desks.

Credit spread widening is one pressure on software multiples; falling AI inference costs are another. SaaS names carrying elevated multiples face a second de-rating vector as model price cuts compress the unit economics underlying their growth forecasts.

CoreWeave (CRWV) is the stress case that crystallized the concern. CoreWeave reported roughly $640 million of quarterly net interest expense in Q2 2026, versus about $128 million of operating income on an adjusted basis. Its unsecured notes due 2032 traded below 80 cents on the dollar this week, implying yields above 15%, according to TRACE data cited by Bloomberg. Bloomberg has also reported the company had about $35 billion of debt at the end of June, plus roughly $16 billion of debt-like lease liabilities. CoreWeave dropped about 12% Friday and finished down roughly 25% for the week, with its bond yields still above 15%. The 10-year Treasury yield was near 5.19% Wednesday after the September CPI release, before easing to about 5.09% by Thursday and Friday. That level is the anchor. Every AI company refinancing debt is pricing against it.

Palantir: Correction Territory, Premium Multiple

Palantir Technologies closed at $186.03 on October 16, down 7.22%, one week after its record close of $209.05 on October 9. Shares now sit 10.99% below that record, past the 10% decline that marks a correction. The move erased roughly $34 billion in market value in a single session.

The business itself has not deteriorated. Palantir’s Q2 revenue surged 93% to $1.94 billion, driven by 149% growth in US commercial revenue to $764 million. Management raised its 2026 revenue forecast to $8.150 billion to $8.158 billion, implying roughly 82% growth year over year. But the multiple is the problem. Palantir trades at a premium valuation versus the broader software complex, which leaves less room for error when rates and credit spreads move against long-duration cash flows. When a 5.09% risk-free rate competes with that kind of forward promise, the math compresses fast.

The iShares Expanded Tech-Software Sector ETF (IGV) fell about 4.8% on October 16. In a widely circulated software basket, nearly every name finished lower, with combined market-value losses estimated around $260 billion, including roughly $77 billion to $80 billion at Microsoft based on its day-over-day market-cap change.

The sector-wide damage on October 16 reversed a trade that had looked durable just weeks earlier. Capital had been rotating into IGV and away from semiconductors as recently as mid-September, when chips fell 5.9% in a single session while software gained.

Technical Framework

PLTR closed below the $188 level that technically confirmed correction status. The 200-day moving average, which traders had watched as structural support, sits materially below current price. A consensus of 32 analysts carries a Buy rating with an average 12-month target of $206.43, implying roughly 11% upside from the October 16 close. Several firms remain constructive with targets clustered in the low-to-mid $200s. Those targets assume the credit concern resolves; they do not price in further spread widening.

Scenario Modeling

Bull Case: AI debt spreads reverse the October 13 to 14 widening, CRWV stabilizes above $60, and PLTR reclaims $200 heading into November 2 earnings. Q3 US commercial revenue growth above 140% year-over-year would reassert the fundamental floor and challenge the spread-driven de-rating.

Base Case: The 10-year holds near 5.10%, software multiples remain under pressure, and PLTR consolidates in the $175 to $195 range through earnings. IGV finds a floor after its year-to-date gains are fully erased, with rotation into energy and financials continuing.

Bear Case: CRWV bond yields stay above 14%, additional bank notes widen AI high-yield spreads past 400 basis points, and PLTR breaks below the $170 area. Another hot inflation release would accelerate that scenario by pushing the ten-year back toward 5.30%.

Active Trader Strategy Framework

Three levels matter most for PLTR right now: $186 (Friday’s close and the correction confirmation close), $175 (the next meaningful volume cluster from August), and $209 (the October 9 record that would need reclaiming to signal recovery). Any position sizing in high-multiple AI software should account for the fact that Goldman credit strategist Amanda Lynam has flagged some potential for near-term stabilization in parts of US high-yield AI-related debt, while also warning that dispersion across issuers is likely to remain high. That dispersion is the trading environment: single-name moves of 7% to 12% in a session will persist until credit stabilizes.

Volatility in IGV and PLTR options has moved sharply. Traders monitoring the credit-equity linkage should watch CRWV’s 2032 bond price daily. A sustained move back above 85 cents on the dollar would be an early signal that the de-rating pressure is easing.

Preparation defines the outcome here. The macro crosscurrents, a 5.09% ten-year, AI debt spreads near recent highs, and software multiples still pricing extraordinary growth, do not resolve in a single session. Position discipline and predefined levels, not conviction about direction, are what protect capital through this recalibration.