Before the US market opened on Wednesday, Nomura cross-asset strategist Charlie McElligott released a research note stating that US stocks have achieved a "critical stabilization" from the "left-tail risk repricing/growth panic" of the previous two weeks, with market sentiment shifting from fear to relief.
The shift was triggered by two catalysts that landed back-to-back: the explosive revenue figures disclosed in Anthropic's S-1 filing, and SK hynix's earnings that massively beat expectations. McElligott wrote in the note that this represents a "broad 'relief' across the collective psychology of risk markets and the AI trade as a macro phenomenon, after several major days."
At the same time, a large bullish options block of roughly $1.6 billion in Magnificent Seven (Mag7) calls emerged, Nasdaq implied volatility compressed sharply, and the S&P 500 stood just 1% from its all-time high. However, roughly 100 hedge fund managers departed during this round of turbulence, which McElligott characterized as a structural "purge" of the multi-strategy model, while also warning that the next round of stop-loss exits may only be a matter of time.
The note was written before Wednesday's market open, when Alphabet and Microsoft had not yet reported earnings, and the Fed's rate decision had not yet landed.
Anthropic S-1: AI Commercialization "Proof of Concept" Lands
McElligott believes that the release of Anthropic's prospectus "did a great deal of work to rebuild local confidence in the AI narrative"—after September's annualized recurring revenue (ARR) data had raised questions about AI profitability and financing models, and had at one point led IPO investors to mark down its valuation to $1.5 trillion.
According to S-1 data cited by Nomura: Anthropic's revenue jumped from $386 million in 2024 and $4.6 billion in 2025 to $16.3 billion in the first half of 2026, of which the second quarter of 2026 alone reached $11.6 billion, up about 15 times year-over-year; consumption grew 13 times year-over-year, and subscriptions grew 18 times year-over-year.
McElligott specifically highlighted two data points he considers most materially supportive of the AI "proof of concept":
First, Anthropic turned operating profit positive in the second quarter of 2026.
Second, revenue efficiency per unit of compute improved dramatically. He directly cited the data: "A year ago, in the second quarter of 2025, for every $1 spent on compute/infrastructure, Anthropic could generate $0.57 of revenue… while according to the S-1, in the second quarter of 2026, for every $1 spent on compute/infrastructure, it can already generate $2.08 of revenue."
In other words, Anthropic's revenue per unit of compute more than tripled in a year. McElligott also noted that the next key milestone is the October ARR update to be released next week, setting his baseline expectation at about $78 billion, and pointing out that "if that number points toward an expectation of roughly $100 billion by year-end, the subsequent virtuous economic spillover effects will follow."
SK hynix "Sold Out Through 2029": AI Memory Demand Remains Unending
The second catalyst came from SK hynix's results. McElligott said the earnings show that the AI memory supply chain's "demand remains robust, with no sign of slowing—'sold out through 2029.'"
According to data cited by Nomura, SK hynix's revenue grew 328% year-over-year, operating profit grew 632% year-over-year, net profit grew 216% year-over-year, and operating margin reached 80%. The company also announced substantial buyback and dividend plans, and its shares jumped 11% that day.
McElligott also pointed out that the market's ability to absorb large-scale financing deals is likewise an important support for this round of confidence restoration. He mentioned the smooth launch of the first tranche of Meta's equity offering (he expects the final supply scale to be "as much as $98 billion"), as well as Broadcom's $63 billion chip agreement with Anthropic and SpaceX's $40 billion equipment financing lease with Apollo, and concluded that "the most real, hard cash" is "enthusiastically taking down large-scale debt deals."
"Whales" Return: $1.6 Billion in Mag7 Bullish Options Blocks Emerge
As sentiment turned, upside chase capital also entered the market. McElligott (citing Nomura's Mike Lamon data) flagged in the note the "large-scale upside volatility buying" that appeared on Monday and Tuesday:
Apple (AAPL): bought 70,000 January-expiry $380 call options, premium about $5, Delta about $570 million, Vega about $3.5 million;
Google (GOOGL): bought 50,000 January-expiry $360 call options, premium about $8, Delta about $430 million, Vega about $2.5 million;
Microsoft (MSFT): a Nomura client bought 42,000 January-expiry $585 call options, premium about $11, Delta about $630 million, Vega about $3.6 million.
The three trades totaled roughly $1.6 billion in Delta. In addition, some bullish call spread combinations in Amazon and Google were also rolled lower.
McElligott also pointed to a structural positioning divergence: Nomura's Beta monitoring shows that long-short hedge funds' Beta to the Mag7 basket (BM7T) is at the 96th historical percentile, with Microsoft at the 99th percentile, Tesla at the 97th, Nvidia at the 94th, and Amazon at the 89th; while mutual funds' Beta to the same basket is only at the 10th percentile.
His conclusion: when benchmark fund managers are so heavily underweight the largest constituents of the index, every Mag7 earnings beat becomes performance pressure for them—meaning Alphabet's and Microsoft's results (not yet released when the note was written) will determine the outcome of this long-versus-short battle.
Momentum Strategy's "Lost October"
Behind the repair in AI sentiment, McElligott pointed out that pressure at the cross-asset risk premium level has not dissipated. He wrote that momentum strategies "remain trapped in a mechanical, persistent de-risking/position management process," with original trend signals "either still reversing, or at least not returning to their original direction."
The bond market is the hardest-hit area. McElligott said that although the most acute short-covering phase has passed, bonds "are showing new resilience and are no longer selling off as they did before"—but this is precisely the short-covering rally that previously destroyed many "high-Sharpe" trend trades.
Nomura QIS monitoring shows that more than 30 trend signals were flagged as "reversed," covering US 2-year, 5-year, 10-year, 20-year, and 30-year Treasuries, 10-year government bonds of Spain, Germany, the UK, Japan, Australia, and Canada, as well as the DAX, Nikkei, KOSPI, Russell 2000, and even commodities such as sugar, cocoa, and coffee.
McElligott concluded that cross-asset momentum strategies have returned about -4.59% month-to-date in October, US equity momentum about -7.16%, and rates momentum/trend strategies fell 3.5% to 4.6%; while energy stocks and crude oil/inflation hedge positions posted double-digit gains.
"Everyone Blew Up": Personnel Losses at Multi-Strategy Funds
The cost of this turbulence fell on specific people. According to headhunter statistics, about 100 fund managers lost their jobs amid the impact of this round of AI "growth panic," the blow-up of French government bond (OAT) spread trades, and cross-asset momentum liquidation.
McElligott cited three industry comments from eFinancialCareers to outline the purge: a London macro hedge fund said "everyone blew up"; in the European high-yield market, "young people have never experienced a real bear market"; and "hedge funds hired too many fools."
He wrote: "…against a low-volatility backdrop, a large number of fund managers were doing a large number of the same 'rational' trades, accumulating persistent positive returns, but the rapid expansion of assets under management created旺盛 demand for talent, an almost universal need to hire (possibly diluting talent quality), and the high leverage required by relative value strategies or deployed by option convexity strategies—together creating these rolling stop-out periods… This is just reality, a byproduct of the strategy's own success."
"This Too Shall Pass"—But the Next Purge May Only Be a Matter of Time
McElligott ultimately characterized this purge as the multi-strategy model's "self-validation": precisely because losing positions are mechanically and emotionlessly liquidated, multi-strategy funds can maintain long-term stable returns and continue to win favor from asset allocators. He wrote: "This too shall pass; this is just the nature of the industry, and performance dynamics then take over, flushing excess positioning and leverage out of the system, and potentially bringing some much-needed hiring 'rebalancing.'"
He also flagged several potential risks: the 10-year US Treasury yield back above 5%, Brent crude near $115, a Fed December rate hike now almost fully priced by the market, and Goldman Sachs historical research showing the current hiking cycle is most similar to 1999.
"The next round of stop-loss exits may only be a matter of time," McElligott hinted.
Comments