Commentary

Winners and losers in the AI shakeup, and the new market landscape

Plus: the latest on the Iran conflict for markets and the economy.

Market remains orderly since the campaign in Iran began Feb. 28

The conflict in Iran continues, but market reaction so far has not reflected a material change in urgency, mindset, or decision-making. The market in recent days has had an orderly tone to it—not one of panic. Jobs data has softened, but energy and inflation remain the bigger watchpoints.

  • Energy prices have risen, with Brent crude about 26% higher since the start of the conflict and up nearly 50% YTD. Volatility is also elevated, but the bigger debate is whether this becomes a lasting supply disruption vs. a risk premium that fades.
  • February payrolls fell by 92,000 and unemployment rose to 4.4%, but wage growth stayed firm at 3.8% YoY. Much of the softness appears tied to temporary factors unrelated to the Iran conflict, and broader indicators suggest the labor market is slowing rather than breaking. While softer hiring may influence Fed policy decisions, rising energy prices linked to events in the Middle East are a more important swing factor for inflation and growth than any single jobs report.

Software rotation underway as durability of business models is being questioned

The software sell-off reflects markets adapting to a faster AI future—not capitulating. Volatility is the cost of that adjustment, and selectivity is the opportunity. Markets continue to navigate a shifting narrative around artificial intelligence (AI), with recent volatility highlighting a growing divide between perceived winners and losers. Over the past six months, investors have rotated aggressively within technology, selling long-duration software stocks while favoring areas tied more directly to AI infrastructure, hardware, and industrial demand. This reallocation reflects structural questions about business models—not a deterioration in economic fundamentals.

What is (and isn’t) driving the sell-off in software stocks?

Uncertainty around terminal value is a sell-off driver

Investors are concerned that rapid advances in generative and agentic AI may introduce uncertainty around traditional software pricing, moats, and long-term profitability, rather than enhance growth. This has led investors to question terminal values for many software companies, even where near-term fundamentals remain solid (Exhibit 1).

Exhibit 1: Software has lagged other tech sectors over the past six months amid questions about AI, even where near-term fundamentals are solid.


Extreme positioning is a sell-off driver

Heavy derisking, crowded positioning, and forced selling acted as accelerants in the sell-off. Software has experienced one of the largest non-recessionary drawdowns in decades, with investors rotating away from software toward AI-linked infrastructure and semiconductors, and sectors such as energy, materials, and consumer staples.

Recession fears, interest rates, and near-term earnings weakness are NOT drivers of the sell-off

The Atlanta Fed is currently estimating real GDP growth of 3.1% for the first quarter of 2026. Based on our business cycle model, the U.S. economy remains in the expansion phase of the business cycle. The Fed has cut interest rates by 1.75%, and while the yield curve is now upward sloping, rates are not very restrictive by historical measures and financial conditions remain accommodative. The S&P 500 is currently estimated to grow earnings by 14.5% in 2026. Thus, this is a valuation and duration reset, not a sell-off caused by economic or earnings concerns.

The sell-off reflects long-term uncertainty, not collapsing revenues or margins. Technology earnings overall remain strong, but software-specific forward expectations are weakening as investors reassess durability under AI competition.

What we’re seeing beneath the surface

Valuation multiples are turning more cautious

Recent earnings announcements show that current and expected revenue and earnings growth remain strong and management teams are broadly constructive. However, investors have turned cautious and are no longer willing to pay the multiples that software stocks once commanded. The S&P 500 Software & Services Index is currently trading at a multiple of 24 times next 12-month earnings versus its post-COVID peak of 33 times in August 2021. This reflects growing consensus that AI introduces longer-term margin and competitive risks.

Markets are actively sorting winners and losers

Importantly, the current environment reflects differentiation rather than broad risk aversion. Investors are not abandoning AI; they are sorting through which companies can adapt and which may struggle. Software companies with proprietary data, mission-critical workflows, and deep customer relationships are better positioned to evolve and potentially thrive. Meanwhile, as the valuations of longer-duration software stocks are repriced for lower terminal values, investors are shifting down the AI stack, favoring infrastructure and semiconductors over application-layer software.

Case for cyclicals remains intact

At the same time, this repricing is occurring against a backdrop of improving cyclical momentum. Manufacturing activity has turned positive, reinforcing the view that the economy remains on stable footing and that parts of the market tied to real activity may benefit. Historically, periods of rising efficiency—driven by general-purpose technologies like AI—have supported both the producers of the technology and the users of it. As AI investment continues to drive demand for chips, data centers, power infrastructure, and industrial capacity, the case for cyclicals and infrastructure remains intact.

Insights

The reset may create opportunity—but selectively

Once positioning stabilizes, valuation dispersion could create attractive entry points in companies with clear data advantages, essential systems, or defensible architectures. The focus remains on selectivity rather than wholesale exits.

What it all means for investors

Fears of structural cracks caused by AI in long-duration business models are leading to a rotation in market leadership, even as economic conditions and corporate earnings remain supportive. Manufacturing momentum turning positive adds a tailwind for select cyclicals, while AI continues to reshape leadership within technology. Ongoing AI innovation means continued reassessment of these long-duration assets. Volatility is likely to persist, but it is also creating opportunity for disciplined, active investors focused on durability, pricing power, and real-economy exposure.