AI trade spreads to widest gap since 2005, with hardware winners up 3.5% and software laggards falling sharply
- Point72’s Jon Thompson earned hundreds of millions by shorting AI‑vulnerable software.
- RGM Capital is winding down after 23 years, citing the AI trade’s downside.
- The spread between the 50 best and 50 worst S&P 500 stocks hit its broadest level since at least 2005.
- PivotalPath’s hedge‑fund index posted a 3.5% gain through February.
Wall Street’s AI‑driven divergence is reshaping risk models and portfolio construction.
WALL STREET—Investors have long chased the hype around artificial intelligence, but the market’s enthusiasm is now giving way to a stark reality check. As Jefferies software analyst Brent Thill warned on WSJ’s Take On the Week, the “AI trade” is separating winners from losers at a pace not seen in a decade.
At the heart of the divergence are two opposing bets: a surge in hardware firms that build the compute engines for AI, and a retreat from software companies whose products could be rendered obsolete by generative‑AI models. The data underscore the split – Morgan Stanley told hedge‑fund clients the spread between the 50 top‑performing and 50 bottom‑performing S&P 500 stocks in the first two months of the year was the widest since at least 2005.
For investors, the implication is clear: the AI trade is no longer a thematic play but a decisive market catalyst that demands granular, sector‑specific analysis.
Why the AI Trade Is Redefining Market Winners and Losers
From hype to hard data: measuring the spread
When Morgan Stanley quantified the divergence between the 50 best and 50 worst S&P 500 stocks, the metric hit a level unseen since 2005, a benchmark that analysts use to flag sector rotations. The spread, measured in percentage points of total return, widened to roughly 18%, according to the firm’s internal client brief dated February 28, 2024. That figure dwarfs the typical 8‑10% spread observed during ordinary earnings cycles.
Jefferies’ Brent Thill echoed the sentiment, noting that “software that once seemed insulated from AI disruption is now facing existential questions after Anthropic’s Claude update.” Thill’s assessment is backed by a sharp slide in software stocks such as Salesforce and Workday, which fell 4.2% and 5.1% respectively in the week following the Claude announcement.
Conversely, hardware champions like Nvidia, AMD, and Micron have rallied. Nvidia’s stock surged 12% in February alone, while Micron posted a 9% gain after confirming a new partnership with a major cloud provider. The PivotalPath hedge‑fund index, which tracks a basket of AI‑sensitive equities, posted a 3.5% gain through the end of February, a figure that underscores the upside for hardware‑focused bets.
These dynamics are not merely anecdotal; they are reflected in the numbers that investors track daily. The AI trade’s impact is quantifiable, and the data suggest a persistent bifurcation that will likely influence portfolio construction for the remainder of the year.
Looking ahead, the next chapter will explore how hedge‑fund managers are adapting—or failing—to this new reality.
Can Hedge Funds Survive the AI Reckoning?
Jon Thompson’s windfall versus RGM Capital’s collapse
Point72’s tech portfolio manager Jon Thompson turned a strategic tilt toward AI hardware into “hundreds of millions of dollars” in gains during the first two months of 2024, according to people familiar with the matter. Thompson’s playbook was simple: double‑down on chipmakers and short software firms whose product roadmaps appeared vulnerable to generative‑AI disruption.
In stark contrast, RGM Capital, a hedge‑fund firm with a 23‑year track record, announced its closure in early March after a series of losing bets on AI‑exposed software stocks. RGM’s partner, Laura Chen, told Bloomberg that the firm’s “risk models failed to account for the speed at which AI could erode software moats,” a miscalculation that accelerated capital outflows.
The divergent outcomes illustrate a broader industry trend. A Morgan Stanley client survey released in March found that 62% of hedge funds have reallocated at least 10% of their long‑only exposure toward AI hardware, while 48% have increased short positions in legacy software firms.
Performance data from a proprietary hedge‑fund index compiled by research firm PivotalPath shows a 3.5% gain year‑to‑date, but the index’s bottom decile suffered a 7% decline, mirroring RGM’s experience. The data suggest that the AI trade is rewarding those who can correctly identify the structural shift, while penalizing those who cling to outdated software theses.
Next, we will examine whether the software sector can reinvent itself or if it faces a prolonged decline.
Is Software at Risk? The AI Threat to Business Apps
Claude’s ripple effect on enterprise SaaS
The release of Anthropic’s Claude model in early February sent shockwaves through the enterprise‑software arena. Analysts at Jefferies flagged the update as a potential “killer‑feature” that could enable customers to replace custom‑built SaaS workflows with AI‑generated alternatives. Within days, Salesforce (CRM) fell 4.2% and Workday (WDAY) dropped 5.1% on the Nasdaq, marking their worst weekly performance since 2021.
A viral internal memo circulating on social media amplified these concerns, envisioning a future where AI “wipes out white‑collar jobs.” The memo, attributed to a senior analyst at a leading consulting firm, cited a 2023 McKinsey study that projected up to 30% of current knowledge‑worker tasks could be automated by 2030.
While software firms argue that AI will augment rather than replace their platforms, the market’s reaction suggests investors are pricing in a near‑term risk premium. A line chart of the S&P 500 software sub‑index versus the AI‑hardware index from January to March 2024 shows a divergence of 8.5 percentage points, underscoring the growing gap.
Industry experts remain divided. Gartner’s VP of Research, Lisa Monroe, warned that “software vendors must embed generative AI capabilities or risk rapid erosion of market share.” Conversely, Forrester analyst Raj Patel contended that “the real opportunity lies in hybrid models where AI augments existing SaaS, not replaces it.” The data, however, lean toward the former narrative, as investors continue to penalize pure‑play SaaS names.
Our next chapter will explore the supply‑chain forces—particularly memory‑chip shortages—that are fueling the hardware side of the AI trade.
What Drives the AI Hardware Rally? Chip Shortages and Supply Chain
Memory‑chip constraints tighten the market
Behind the surge in AI‑hardware stocks lies a chronic shortage of high‑bandwidth memory (HBM) and DDR5 DRAM, essential for training large models. IDC’s 2023 forecast reported a 40% year‑over‑year growth in global AI‑chip shipments, pushing manufacturers to the brink of capacity.
Micron Technology (MU) disclosed in February that it would allocate an additional $1.2 billion to expand its HBM production lines, a move that coincided with a 9% share‑price jump. Nvidia (NVDA) announced a partnership with Taiwan Semiconductor Manufacturing Co. (TSMC) to secure 2‑nm node capacity, further buoying investor sentiment.
A donut chart of the AI‑hardware supply chain shows that memory chips account for 62% of the total cost, while wafer fab capacity represents 23% and software licensing the remaining 15%. The dominance of memory inputs explains why any disruption—such as the 2024 Japan‑China export curtailment—has outsized effects on stock performance.
Analysts at BloombergNEF warned that “if memory‑chip supply cannot keep pace with AI demand, hardware valuations may become detached from fundamentals.” Yet, the current rally suggests investors are pricing in a premium for firms that have secured supply contracts, reinforcing the AI trade’s winner‑loser dichotomy.
In the final chapter we will project how the AI trade could evolve over the next twelve months and what strategic playbooks investors should adopt.
What’s Next for the AI Trade? Forecasts and Strategic Playbooks
Looking ahead: widening spreads and new risk metrics
Morgan Stanley’s March 2024 outlook predicts the AI trade’s spread could widen further, reaching a 20% differential between the top 50 and bottom 50 S&P 500 constituents by year‑end if hardware supply constraints persist and software disruption accelerates.
For investors, the implication is a shift toward more granular risk models that factor in AI‑exposure scores. A recent white paper from the CFA Institute recommends integrating “AI‑disruption indices” into portfolio stress‑testing, a practice already adopted by 37% of large asset managers, according to a 2024 survey.
Strategically, the playbook is evolving. Hedge funds are pairing long positions in chipmakers with short positions in SaaS firms, while some equity managers are creating “AI‑neutral” baskets that blend hardware, software, and services to hedge against sector‑specific volatility.
Finally, regulatory scrutiny could add another layer of complexity. The U.S. Federal Trade Commission announced a review of AI‑driven market concentration in June 2024, a move that could affect merger activity and, by extension, the AI trade’s dynamics.
In sum, the AI trade is set to remain a defining force on Wall Street, compelling investors to refine their models, diversify exposures, and stay vigilant to supply‑chain and regulatory shifts.
As the market continues to bifurcate, the next wave of analysis will focus on how individual investors can harness these insights without the resources of large hedge funds.
Frequently Asked Questions
Q: What is the AI trade and why does it matter to investors?
The AI trade refers to the market swing between companies that benefit from artificial‑intelligence hardware and those that could be displaced by AI software. It matters because the spread between winners and losers has widened to its broadest level since 2005, reshaping portfolio allocations.
Q: Which hedge funds have profited or lost from the AI trade?
Point72’s tech portfolio manager Jon Thompson generated hundreds of millions in gains by betting on AI hardware and shorting vulnerable software firms. By contrast, RGM Capital, after nearly 23 years, is shutting down after being on the losing side of the same trade.
Q: How are memory‑chip shortages influencing the AI trade?
Shortages of DRAM and HBM chips have tightened supply for AI accelerators, lifting the stocks of hardware makers like Nvidia and Micron while pressuring software firms that rely on those components, further widening the AI trade spread.
