China quant-fund AUM estimated above ¥2.6T as quant strategies beat discretionary managers by ~20 points

Why it matters (spine). Capital is voting with its feet: money is flowing hard toward systematic, machine-driven stock-picking in the world's second-largest equity market, and the reported outperformance gap over human discretionary managers is roughly 20 points in a single year. That is a money-weighted verdict on quant/AI strategies at national scale — but only as strong as the estimates carrying it, which here are brokerage and data-provider figures, not a regulator's ledger.

What would prove it wrong (falsifier). An AMAC or CSRC primary shows materially lower quant AUM or outperformance than the trade-press figures — for example, a quant-specific regulator dataset landing well below ¥2.6T, or an official performance series that closes the ~20-point discretionary gap.

Live-status check + triangulation. Quoted from the Bloomberg article (Bloomberg News, 2 July 2026): China quant funds' AUM "more than doubled in less than a year to more than 2.6 trillion yuan." The cited Bloomberg URL returns HTTP 403 (paywall); the full article was verified in-session via Yahoo Finance's syndication of it (confirming the >¥2.6T figure, the Citic Securities 44.7% / 20.3ppt performance claim, Industrial Securities' 6,296 new quant products = 46% of new hedge funds, and Ubiquant's ¥2.6B raise in under two hours). Corroborators fetched in-session: Eastmoney (2026-01-20) reports 2025 quant index-enhanced products averaged 45.08% return with 90% positive excess, consistent with Citic's 44.7%; Xinhua Finance (2025-12-17, PaiPaiWang data) shows 833 quant long-only products averaging +17.25% excess return through Nov 2025, 91.48% positive; Xinhua Economic Information Daily (2026-04-18) puts quant private-fund AUM conservatively above ¥1.8T at end-Q1 2026 (+28% QoQ, 61 firms above ¥10B) — materially below ¥2.6T but a narrower, private-only scope and a quarter earlier. The true primary here is AMAC's monthly registry (linked as source_url), which shows private securities investment funds at ¥7.46T (end-March 2026) and ¥7.92T (end-May 2026) — matching Bloomberg's "8 trillion yuan hedge fund industry" — but AMAC publishes NO quant-specific AUM. This is why the card sits at signal, not verified: the ¥2.6T headline is unattributed within the article, the underlying Citic and Industrial Securities research reports are not publicly posted, and no regulator dataset breaks out quant AUM. Independent private (¥2.5T) plus public (¥0.7T) quant estimates bracket Bloomberg's number, so the falsifier is partially engaged but not tripped. Do not promote to verified unless AMAC or CSRC publishes a quant-specific dataset supporting ≥¥2.6T.

Correction carried forward (discipline). Four corrections applied against the original draft: (1) the draft said "AI strategies beat discretionary managers by >20 points" — the primary (Bloomberg citing Citic Securities Co.) says long-only stock QUANT strategies gained 44.7% in 2025, beating discretionary rivals by 20.3 percentage points; quant is not synonymous with AI. (2) The draft presented ¥2.6T AUM as an established figure — the number is unattributed within the Bloomberg article; independent Chinese estimates range from ¥1.8T (quant private funds only, end-Q1 2026, PaiPaiWang-derived, called "conservative") to ~¥3T (including ~¥0.7T public quant mutual funds), with no regulator dataset. (3) The draft's framing of a "national-scale, money-weighted verdict that AI beats human stock-pickers" is overreach — it rests on one calendar year (2025) in a small-cap-led bull market, measured by brokerage estimates, with Bloomberg itself noting crowding concerns as 71 managers now exceed ¥10B each. (4) The draft's "doubled" baseline cannot be independently anchored — no public dataset establishes the mid-2025 starting AUM Bloomberg implies (¥1.3T).