NDV: Strategy determines fate, 2026 H1 Hedge Fund Q2 Panorama Overview

Author: NextGen Digital Venture

Original link:

Disclaimer: This article is a reprint. Readers can access more information through the original link. If the author has any objections to the form of reprinting, please contact us, and we will make modifications as requested by the author. This reprint is for information sharing only and does not constitute any investment advice, nor does it represent the views or positions of Wu Shuo.

Over the past two years, there has been a judgment that almost required no proof: as long as you held large-cap tech stocks, you could outperform everything. Buying the "Magnificent Seven" was almost the standard answer for the global asset management industry. But in the first half of 2026, this answer failed for the first time. When a poker table is full of people, leaving the table itself becomes the most scarce type of return this year.

In 2024 and 2025, the U.S. "Magnificent Seven" (Mag 7, referring to Apple, Microsoft, Google, Amazon, NVIDIA, Meta, and Tesla) collectively contributed over half of the gains in the S&P 500 index. "Buy them" became an unspoken consensus in the global asset management industry—not heavily weighting these seven names would even seem unprofessional.

But in the first half of 2026, this consensus failed for the first time.

As of mid-year, the ETF specifically tracking the Magnificent Seven overall underperformed the S&P 500. Among the seven companies, only Google and NVIDIA saw their stock prices hit new highs; the remaining five—Apple, Microsoft, Amazon, Meta, and Tesla—all lagged behind the broader market.

In other words: if you had put all your chips on the "most effective strategy of last year" at the beginning of this year, your return this year would be worse than that of an index fund that passively tracks the market without picking anything. Those who did their homework lost to those who didn't think.

This review aims to break down three things: what exactly happened within the Magnificent Seven; which global hedge funds made or lost money in the first half of the year and the underlying patterns; and, as we stand at mid-year, three variables truly worth watching in the second half.

I. Inside the Magnificent Seven: Who is Spending and Who is Earning?

The divergence within the Magnificent Seven is no accident; it reflects the AI cycle shifting from "faith-driven" to "performance-driven."

In 2024, buying AI concepts alone could make money because the market believed that every dollar of AI capital expenditure (capex, referring to hard investments like land acquisition, data center construction, and chip procurement) would eventually translate into revenue. But in 2026, the market began asking a more fundamental question: the money has been spent—who actually profits from it?

The answer is becoming clearer. The four super spenders—Microsoft, Meta, Amazon, and Google—are expected to collectively spend over $300 billion on AI-related capital expenditures this year. Spread over 365 days, these four companies are spending over $800 million per day on AI. However, the path to converting these investments into revenue is uneven:

Google, leveraging its search advertising and cloud business for AI monetization, continues to deliver results;

NVIDIA, as the "pick-and-shovel seller," directly reaps the rewards of the AI construction boom;

Microsoft, Meta, and Amazon have yet to receive full market recognition for their AI input-output ratio.

To summarize this year's drama in one sentence: those who spent money didn't outperform; those who made money did.

II. Liquidity Siphoning: The Market is Narrowing

What deserves more attention is an ongoing "liquidity siphoning" phenomenon.

AI enthusiasm has not faded, but funds are moving—from broad AI concepts like the Magnificent Seven to more precise beneficiaries of the AI supply chain, especially memory semiconductors. One of the tightest links is High Bandwidth Memory (HBM, one of the scarcest storage components in AI computing infrastructure). Memory manufacturers like SK Hynix, Samsung, and Micron have significantly outperformed the broader market this year.

When funds accelerate into a narrow sub-sector, they simultaneously drain liquidity from other sectors. Thus, the market becomes narrower: in 2024, the Magnificent Seven sucked up all the money; in 2026, two or three segments of the AI supply chain are siphoning money from the Magnificent Seven in turn.

Core Judgment: Crowdedness is converging into an even smaller circle, and convergence itself implies a buildup of fragility. When the "effective breadth" of the market narrows, any strategy relying on a few stocks to generate returns faces a concentration trap—last year, betting on seven names could win; this year, you need to pick the right two out of seven; next year, perhaps only one will remain. This is the essence of a crowded trade: too many people on one side, and when the direction reverses, a stampede is likely.

III. Valuation Safety Margin is Nearly Exhausted

At the same time, the safety cushion left by valuations is almost used up.

The Shiller CAPE ratio (a valuation metric that uses average earnings over the past ten years to smooth out cycles and measure stock price levels) for the U.S. stock market has exceeded 40, the second highest in 140 years, second only to the peak of the internet bubble in 2000.

A variable has emerged on the denominator side. After Kevin Warsh took office as Fed Chair, market expectations for rate cuts significantly narrowed. By May 30, the 30-year U.S. Treasury yield had risen to 5.18%. This means: lending money to the U.S. government, taking no risk, yields over 5% annually.

When risk-free assets start offering real returns above 5%, the earnings growth threshold required to sustain valuations above 40 times is systematically raised.

Core Judgment: The Magnificent Seven haven't become worse; they now need to "beat expectations" every quarter just to hold their current prices—a company surviving solely on "quarterly beats" is itself a form of fragility.

IV. Hedge Funds in Q2: Strategy Determines Fate

Zooming out from the seven companies to the entire hedge fund industry, we also see a test.

A review of Q2 performance for 23 well-known hedge funds reveals a clear pattern: your strategy type explains your quarterly results better than your team size or historical track record.

Commodity strategies shot ahead. Pierre Andurand's commodity fund led the pack with +30.8%—a $1 million investment at the start of the year turned into approximately $1.31 million by mid-year. He benefited from extreme crude oil volatility in the first half: a key strait in the Middle East went from blockade to reopening, and WTI crude experienced a full cycle from $60 to $95 and back down.

Here's a counterintuitive point: Andurand's excess returns did not come from guessing the direction correctly. Oil prices surged and then plunged; someone who holds a single direction to the death just rides a roller coaster and ends up where they started. But a trader capable of repeatedly entering and exiting amid volatility can capture multiple slices of profit from the back-and-forth. When macro narratives dominate pricing and facts persistently diverge from expectations, commodity traders willing to bet on non-consensus directions have a natural advantage. This also explains why commodity strategies significantly outperformed in both 2022 (Russia-Ukraine conflict) and 2026 (Middle East crisis), two geopolitically driven years.

Quant and Asia long/short advanced steadily. Quantedge (+12.3%) and Point72 (+9.2%) share a common trait: they do not rely on single directional judgment but instead capture spreads across multiple markets and cycles through diversification and systematic approaches. FengHe Asia (+11.2%), a representative Asia equity long/short (simultaneously going long on favorites and shorting on underperformers), also shined—indicating that beyond the Magnificent Seven-dominated U.S. stock narrative, Asian markets still offer independent sources of excess returns.

Multi-strategy platforms diverged. Millennium (+6.0%), Citadel Tactical (+5.3%), and Winton (+6.3%) delivered above-average results, showcasing the risk diversification capabilities of multi-strategy platforms (where hundreds of independent trading teams under one roof hedge losses from any single direction). However, ExodusPoint (−1.3%), Walleye (−0.9%), and Balyasny (−1.6%), also labeled as "multi-strategy," recorded losses. The difference may lie in the latter having larger directional equity exposures; when tech stock momentum reversed, the "multi-strategy" umbrella didn't always hold. Labels do not guarantee stability; underlying strategy composition and risk discipline are the true dividing lines.

Tech long-only funds were broadly hurt. Tiger Global (−6.2%) suffered the largest loss of the group, while Coatue (−2.1%) and Viking Global Equities (−0.7%) were also negative. Their common trait is a long-term core holding of tech and growth stocks—which accumulated handsome returns during the Magnificent Seven's surge, creating path dependence. When the Magnificent Seven turned from a "source of excess returns" into a "drag," the most faithful believers were hurt the most. Tiger deserves special mention: this flagship fund, which once managed nearly $100 billion in 2021, suffered massive losses in 2022 and has yet to escape the gravitational pull of tech stocks.

The sharpest contrast comes from the same firm: Marshall Wace's Eureka (quant-driven) returned +1.1% in Q2, while Tops (fundamental-driven) managed only +0.1%. Same team, same infrastructure, but due solely to different strategy types, returns differed by nearly tenfold.

V. When Consensus Crumbles, Who Makes Money?

Stringing the information together, the profile of winners in the first half of 2026 has emerged, with three common traits:

First, they are not in the most crowded tracks. Whether it's Andurand's commodities, Quantedge's systematic multi-asset, or FengHe's Asia long/short, excess returns do not come from the Magnificent Seven. When everyone is crowded at the same table, choosing to go to another table is itself alpha.

Second, they embrace volatility rather than fight it. Commodity strategies lead not because they guessed the direction correctly, but because they could repeatedly trade amid extreme volatility.

Third, risk control is not just a slogan but part of the strategy. The dividing line between positive and negative in Q2 often wasn't who judged more accurately, but who lost less when wrong. Citadel and Millennium didn't have spectacular returns, but they also didn't lose money—among 23 funds, that alone means risk control worked when it mattered.

Conversely, the loser profile is equally clear: heavy tech stock holdings, reliance on directional exposure to a few names, and a lack of error-correction mechanisms in crowded trades.

Core Judgment: This does not mean tech stocks are done—quite the opposite, Google and NVIDIA prove that companies with genuine AI monetization capabilities are still sought after by the market. The problem is not tech stocks themselves, but using a "buy and hold all seven" approach to express a theme that requires precise stock selection.

VI. Three Dimensions to Watch in the Second Half

First, marginal changes in AI capex. Capex guidance from the four giants was still being revised upward in the first half, but market tolerance is declining. Notably, the AI application companies that provide the ultimate return justification for these capital expenditures are seeing their revenue growth "acceleration" slow: OpenAI's annualized recurring revenue (ARR) grew from about $4 billion at the end of 2024 to nearly $10 billion by the end of 2025—still astonishing in absolute terms, but quarterly sequential growth has dropped from "doubling" to 20–30%. Anthropic's trajectory is similar—revenue is rising, but the slope is flattening.

In other words, the "second derivative" of growth (the direction of change in the growth rate itself) has turned negative. This means a key assumption is loosening: if application-layer revenue growth no longer accelerates, how will the upstream annual capex of over $300 billion be recouped? Currently, the market is ignoring this question. But if any giant sharply cuts AI spending in the Q3 earnings season, or if OpenAI/Anthropic's growth significantly misses expectations, the siphoning effect on memory semiconductors could reverse instantly, with funds flowing back from the supply chain to the broader market. For the currently crowded trade chasing HBM, this is the biggest tail risk (low probability but high impact if it occurs).

Second, the rate path. The policy framework of the Warsh era is still being digested by the market. If inflation remains sticky and rate cuts are repeatedly delayed in the second half, high-valuation growth stocks will face greater pressure. Conversely, if inflation drops quickly, opening a window for rate cuts, tech stocks may experience a valuation recovery—but the beneficiaries might not be all seven names.

Third, catalysts in the commodity market. El Niño has been confirmed to form, with a 63% probability of developing into a "very strong" level. Historically, El Niño's impact on agricultural prices usually peaks 4–5 quarters after its onset—if the peak falls in the winter of 2026, price effects could extend into the second half of 2027. This provides commodity strategies with a different but equally significant opportunity source compared to the first half.

Conclusion: Why I Keep an Eye on This

Perhaps the biggest lesson from the first half of 2026 is: in a market dominated by macro forces and where consensus is constantly upended, "what I allocated to" matters more than "how much I allocated."

I am also someone who manages money. Over the years, I have developed a habit: when observing the market, I first look at where big money is moving, what constraints bind it, and what it fears, rather than listening to what retail traders are shouting. This fund scorecard is not industry gossip to me; it is a mirror—repeatedly saying the same thing: leave that most crowded table. This is also the origin of the words "non-consensus": not for the sake of being contrarian, but because when everyone is on the same side, that position often carries the highest risk and offers the thinnest returns.

The exam paper for the second half has already been handed out. The answer depends on whether you are willing to leave that most crowded table.

Where the Judgment Could Be Wrong (Tail Risks)

Let me say upfront: the three most likely mistakes in this review are:

1. AI might be a genuine paradigm shift.

If it is structural rather than cyclical, then the names that look most crowded and expensive today may not be the top, but the starting point of a new world not yet adapted to. The caution of "don't get too crowded" could cause one to miss the biggest move. This is the most critical counterargument to watch.

2. Commodity strategies' success relies on volatility.

They lead this year precisely because volatility is extreme. Once the macro environment returns to calm, they will be the first to fail—every reason to praise them is, in reverse, their risk.

3. Limitations of a single-quarter league table.

The Q2 ranking is just a snapshot of one quarter, subject to survivorship bias and not representative of the full year. OpenAI and Anthropic revenue figures are public estimates; it is recommended to verify before publication.

Data and Information Sources

Hedge fund Q2 performance data is from publicly available industry tracking reports; Magnificent Seven and S&P 500 performance data is from public market data; Shiller CAPE data is from Robert Shiller/Yale; U.S. Treasury yield data is from U.S. Treasury/FRED; AI capex data is from public financial reports and guidance of Microsoft, Meta, Amazon, and Alphabet; OpenAI and Anthropic revenue data is from public news reports and industry estimates; HBM and memory semiconductor market data is from public market data; El Niño probability and intensity forecasts are from official NOAA releases.

Disclaimer

This article is compiled based on publicly available information, intended for financial explanation and educational purposes only. It does not constitute any investment advice, nor is it directed at any specific security, target, or timing. Content marked as "Core Judgment / Tail Risk" represents the personal views of the author, separate from factual statements, and may be incorrect. Investing involves risk; decisions should be made with caution.