Oil Risk Returns; Liability Fractures the Tape

2. Complete macro-event table

Rank	Event and time (EDT)	Actual versus consensus / prior	Surprise and immediate reaction	Why it mattered
1	U.S.–Iran/Hormuz escalation — overnight/session	U.S. struck two IRGC launchers; Jordan intercepted eight missiles; UAE intercepted one drone; no verified mined vessel	Brent +2.71%; WTI +2.83%; 10Y near 4.75%; energy outperformed	Reopened the probability distribution around commercial Hormuz safety and global inflation. (AP)
2	California wildfire-liability compromise — session; final vote pending at 4:00	No numerical consensus; survivor/prevention package retained subrogation and omitted durable utility-liability relief	EIX −23.07%; PCG −20.06%; SRE −3.10%	Repriced catastrophic left-tail exposure, fund solvency and financing/rating risk for California utilities. (bill record)
3	China property-financing rules — first session after Aug. 28 7:00 p.m. CST release	Mortgage maturity ceiling 40 from 30 years; delivery-first disbursement and closed project accounts	Hang Seng Mainland Properties nearly −6%; CR Land −9.40%; broad HSI −0.07%	Long-run trust protection creates a near-term developer cash-cycle shock and consolidation pressure. (PBOC, NFRA)
4	China official PMIs — Aug. 30 9:30 p.m.	Manufacturing 49.8 vs Reuters 49.6, prior 49.2; non-manufacturing 49.0 vs 49.4–49.5, prior 49.0; composite 49.5 vs 49.3	AUD nearly flat; HSI −0.07%; CSI 300 recovered from −0.8% midday to +0.35%	Export/high-tech stabilization did not repair weak services, construction or employment. (China government/NBS)
5	Germany flash CPI/HICP — 8:00	CPI 2.9% y/y, 0.2% m/m vs 3.0%/0.3%, prior 2.8%/0.8%; HICP 2.9%/0.2% vs 3.1%/0.3%, prior 2.8%/0.9%	Brief Bund bounce, but contract remained ~40 ticks lower; EUR/DAX response modest; DAX −1.17%	Softer-than-feared core/services did not remove the oil-driven inflation/ECB problem. (Destatis)
6	India Q1 FY27 GDP — 6:30	Real GDP 7.8% vs Reuters 7.1% / Bloomberg 7.3%; prior Q4 revised to 8.6% from 7.8%; nominal GDP 10.3%	Cash equities/INR had closed; 10Y held near 6.95%, +4 bp; no clean discrete reaction	Investment/manufacturing strength beat, but revisions changed “steady” growth into deceleration. (PIB/MoSPI)
7	Dallas Fed manufacturing — 10:30	General activity 11.6 vs 0.7–1.6 public estimates, prior 1.3; production 16.1 from 10.1; new orders 22.0 from 6.4; input prices 44.1 from 41.3	No isolatable reaction in an oil/Warsh-dominated session	Strong regional demand plus rising input costs supported resilient-growth/sticky-inflation risk. (Dallas Fed)

3. Detailed macro events and opinion clusters

3.1 U.S.–Iran/Hormuz: direct barrels intact, commercial safety unresolved

Facts. CENTCOM said U.S. forces conducted a limited strike on two IRGC launchers on Larak Island after observing personnel preparing rockets carrying sea mines. Jordan’s armed forces said it destroyed eight missiles; the UAE said it intercepted an Iran-origin drone and denied that Al Minhad Air Base was hit. An IRGC claim that a supertanker struck mines was denied by CENTCOM and uncorroborated by UKMTO. Trump’s AI-generated Kharg Island image was not confirmation of a physical attack, and Iran’s state oil company said operations continued. (Jordan/Petra, UAE/WAM, CBS chronology)

Cluster A — Tactical shipping premium, not lost production (high evidence). Lin Ye/Rystad, Giovanni Staunovo/UBS, ANZ, Tony Sycamore/IG, Gelber & Associates and Ole Hansen/Saxo agreed Larak is a military/logistics node rather than an oil-export asset. Continuing Gulf flows capped the rally, but the first direct exchange in a month increased the probability of ship hesitation and repeat strikes. The days-to-weeks implication is support for crude, producers and freight optionality; a verified Kharg hit, falling exports or another disabled tanker would turn a premium into physical loss. (Reuters analyst roundup, Reuters settlement)

Cluster B — Physically open, commercially unsafe (medium-high). Alex Plitsas/Atlantic Council, Ian Ralby/Auxilium Worldwide and the Soufan Center argued tactical strikes cannot destroy every rebuildable missile/drone site or make risk-averse owners, crews and insurers return. Periodic U.S. strikes under consideration are evidence that mine clearance is not self-sustaining. Implication: weeks or months of elevated freight/insurance and oil-inflation tails. Disconfirmers are sustained escorted transits, declining premiums and stable 15–16 mb/d Gulf exports. (RFE/RL, Axios)

Cluster C — Bounded policing signal (medium-low). Jason Brodsky/United Against Nuclear Iran read the limited target set and defensive CENTCOM wording as policing rather than a decisive new war phase. Interceptions prevented publicly verified U.S.-base damage and Pezeshkian called for dialogue. Trump’s threat to “hit them hard” and the proposed recurring-strike plan are the disconfirmers. (RFE/RL, Axios)

Consensus / disagreement / resolver. Consensus: Larak did not remove current production, but raised commercial-shipping risk. Sharpest disagreement: transient enforcement premium versus a repeat-strike regime. Verified transits/export discharge, war-risk premiums, the next U.S. response and any mediator-backed stand-down resolve it. No public CDX/GCC sovereign-spread close was verified; credit reaction is intentionally not invented.

3.2 California wildfire legislation: the market priced omitted liability relief

Facts. The official record at the U.S. close still showed SB 492 in Assembly floor process. The compromise added faster survivor payments, prevention/financing tools and governance provisions but retained insurer subrogation and did not provide the durable liability cap or clearly replenished state backstop investors had anticipated. EIX, PCG and SRE fell 23.07%, 20.06% and 3.10%; broad utilities lost about 1.2% and non-California peers were roughly flat. (official history, California Insurance Commissioner, S&P Global)

Cluster A — Structural-liability bear case (high). Mizuho, BMO, Wells Fargo, Barclays and Jefferies converged on a durable risk discount: Mizuho cut EIX to Neutral and its target to $70 from $86 because the package did not ensure Wildfire Fund solvency or replenishment. The 6–24 month implication is higher equity/debt costs and lower valuation multiples, with EIX most exposed and PCG most sensitive to ratings/sponsor support. New capital, a hard cap or accepted loss-sharing would disconfirm it. (S&P Global, Bloomberg Law)

Cluster B — Risk premium, not current EPS impairment (medium). The countercase is that regulated near-term earnings and approved rate bases did not change, while a 20%–23% one-day decline may overprice a still-pending bill and leave 2027 reform optionality. Sempra’s relative resilience supports exposure differentiation rather than sector-wide operating deterioration. Ratings actions, financing terms, EIX’s Eaton-loss range and PG&E’s scheduled Sept. 2 update are the tests.

Cluster C — Survivor/insurance stability must constrain reform (medium-high). Commissioner Ricardo Lara, survivor groups, lawmakers and insurers argued eliminating subrogation would transfer costs to policyholders, shrink coverage and weaken accountability. Utilities respond that unlimited claims can exhaust the fund and raise customer financing costs. The disagreement is distributive, not merely technical; the 2027 governor/legislature must produce new capital or a politically credible loss-sharing formula.

Consensus / disagreement / resolver. Consensus: the framework remains too fragile for a utility rerating. Sharpest disagreement: whether the selloff overstates unchanged near-term EPS or correctly discounts an uncapped tail. Final bill text/votes, rating-agency actions, company financing updates and 2027 reform terms resolve it. Evidence is high on the tape/omissions, medium on eventual reform.

3.3 China property financing: buyer protection versus the lost presale float

Facts. The PBOC/NFRA package was published Aug. 28 between 7:00 and 7:15 p.m. CST, after local trading; Aug. 31 was the first market test. Maximum mortgage maturity rose to 40 years, monthly mortgage service remained capped at 50% of income, and total debt service rose to 60% from 55%. Mortgage disbursement moves from topping-out to formal completion; lead banks must control closed project accounts; development-loan absolute terms extend to five years for presale and seven for completed homes. (PBOC, NFRA mortgage rule, NFRA development rule)

Cluster A — Trust repair can revive demand (medium). Zhang Dawei/Centaline, Zhiwei Zhang/Pinpoint, Yan Yuejin/E-House, Zhao Ke/China Merchants Securities and Zhu Jin/China Securities emphasized safer delivery, transparency and roughly 15% lower monthly payments at a 3% mortgage rate. The 6–24 month bull case is repaired household trust. Weak employment/income, falling-price expectations and higher lifetime interest are the disconfirmers. (Reuters, Securities Times)

Cluster B — Cash-flow shock accelerates consolidation (high near term). Zhang Xiaoxi/Gavekal, Jeff Zhang/Morningstar, Everbright Securities and Li Hao/CICC focused on presales’ dominant funding role, tighter project-account control and the transfer of completion risk back to developers. Large liquid/SOE developers should gain share while weaker private firms cut starts or exit over 6–36 months. Grandfathering, longer project loans and capital-market channels could soften the shock. (Reuters, Reuters market close)

Cluster C — Cleaner bank collateral, longer-duration assets (medium). CITIC Securities, China Securities and CICC expect lead-bank control, delivery-first disbursement and payment restructuring to slow mortgage runoff and improve asset quality. Banks also inherit longer-duration low-yield assets and more construction-period concentration. Same-day bank gains were contaminated by earnings/dividends, so the thesis needs mortgage originations and property NPLs.

Consensus / disagreement / resolver. Consensus: this is structural risk transfer, not conventional stimulus. Sharpest disagreement: whether safer delivery and lower monthly payments offset the lost presale float. Sales/prices, starts/completions, mortgage originations, white-list approvals, developer cash flow/defaults and bank NPLs resolve it.

3.4 China PMI: export/high-tech stabilization, domestic demand still contracting

Facts. Manufacturing improved to 49.8, with production 50.4, new orders 50.6, exports 50.1 and high-tech manufacturing 52.9. Non-manufacturing remained 49.0: services 49.3, construction 46.9, new orders 44.1 and employment 45.4. Composite rose to 49.5. Reactions were muted and contaminated by property rules and geopolitics. (official release, AP)

Cluster A — Narrow export/AI stabilization (medium-high). Capital Economics’ Nguyen Hoang Nam, EIU’s Xu Tianchen and Conference Board’s Max Zenglein highlighted stronger exports, high-tech output and larger-enterprise activity. The implication is tactical support for exporters/AI supply chains and lower near-term hard-landing risk. Trade-policy shocks or new-order relapse would disconfirm it. (AP, Reuters)

Cluster B — No broad recovery (high). ING’s Lynn Song, Pinpoint’s Zhiwei Zhang and CFLP’s Zhang Liqun read the below-50 composite, weak services/construction, 44-month-low non-manufacturing orders and contracting employment as continued household/property weakness. The implication is persistent easing pressure and weak pricing power outside strategic manufacturing. (Reuters, ING)

Cluster C — Weather/fiscal normalization (medium). Capital Economics, China Logistics Information Center’s Wen Tao and NBS’s Huo Lihui expected fading typhoon disruption and faster already-budgeted spending to lift September without proving autonomous private demand. Services, construction orders and employment must cross 50 alongside stable property sales to validate this view. (Xinhua, Investing.com)

Consensus / disagreement / resolver. Consensus: sequential stabilization was real but narrow. Sharpest disagreement: durable export/high-tech turn versus weather/seasonal noise. September PMIs, exports, credit, retail sales and property activity resolve it.

3.5 Germany CPI: energy shock, limited second-round evidence

Facts. Destatis released the flash estimate at 2:00 p.m. CEST / 8:00 a.m. EDT. Energy inflation accelerated to 10.5% from 8.3%, but core stayed 2.4% and services slowed to 2.8%. The lower-than-consensus headline/HICP briefly supported Bunds, yet the contract remained roughly 40 ticks lower as oil/gas dominated; DAX closed 1.17% lower. (Destatis)

Cluster A — Energy, not domestic overheating (high). Elmar Voelker/LBBW, Dirk Schumacher/KfW and Thomas Gitzel/VP Bank focused on stable core and easing services. The print softens the worst second-round narrative; absent persistence, the implication is less aggressive tightening beyond September. (Reuters economist roundup)

Cluster B — Headline relief may be temporary (high). Voelker, Carsten Brzeski/ING, Jörg Krämer and Ralph Solveen/Commerzbank expected renewed energy pass-through and possible above-3% inflation, leaving September tightening live even after the miss. Bund duration remains exposed to oil/gas pass-through over weeks. (Reuters, German roundup)

Cluster C — Peak then slow decline (medium). Union Investment’s Michael Herzum and KfW’s Schumacher saw the energy-heavy rise peaking before a gradual decline and policy pause. Renewed oil/gas escalation or a services/wage reacceleration would disconfirm it. (Reuters economist roundup)

Consensus / disagreement / resolver. Consensus: energy drove the acceleration and broad spillover was not yet visible. Sharpest disagreement: one September hike then pause versus a longer sequence. Final German detail, euro-area flash/final CPI, negotiated wages, services and gas/oil prices resolve it.

3.6 India GDP: a real investment beat with a large revision caveat

Facts. MoSPI released the estimate at 4:00 p.m. IST / 6:30 a.m. EDT. Gross fixed capital formation grew 11.9%, manufacturing 9.2%, and financial/real-estate/IT/professional services 12.1%; nominal GDP grew 10.3%. Cash INR/equities had closed before release, and the 10-year finished near 6.95%, up about 4 bp, so no discrete price response is assigned. (PIB/MoSPI)

Cluster A — Capex-led resilience (high). Sakshi Gupta/HDFC Bank, Devendra Pant/India Ratings, Radhika Rao/DBS and Tanay Dalal/Axis emphasized double-digit investment and broad manufacturing/business-services growth, supporting high-single-digit FY27 estimates and Indian cyclicals over 6–12 months. (Moneycontrol, Reuters)

Cluster B — Beat, but not yet self-sustaining (medium-high). Dhiraj Nim/ANZ, Madhavi Arora/Emkay, Aditi Nayar/ICRA and Rumki Majumdar/Deloitte accepted the investment beat while questioning whether private capex/consumption can sustain it through the oil shock. The revised 8.6% Q4 also means momentum slowed 0.8 point. (Economic Times, Moneycontrol)

Cluster C — Method/nominal/oil caution (medium). Majumdar, Arora and Rao emphasized that oil and nominal-income channels could weaken H2; the new series’ double-deflation and historical revisions can also change sector contributions. Tax receipts, industrial output and company revenue should corroborate the 7.8% real estimate. (Reuters, Moneycontrol)

Consensus / disagreement / resolver. Consensus: investment/manufacturing breadth was better than expected. Sharpest disagreement: durable private-cycle acceleration versus a public-capex/revision artifact. Q2 GDP, monthly capex/credit, rural consumption, tax collections and the next revision resolve it.

3.7 Dallas Fed: strong orders and higher prices, little clean market signal

Facts. The 10:30 a.m. EDT survey showed general business activity at 11.6 from 1.3, production 16.1 from 10.1, new orders 22.0 from 6.4 and raw-material prices 44.1 from 41.3. No revision was reported. Oil, Friday’s Warsh repricing and month-end flows dominated the event window, so assigning a precise yield/equity move to Dallas would be false precision. (Dallas Fed release, summary table)

Cluster A — Regional demand reacceleration (medium-high). Orders and production strengthened together, consistent with a resilient industrial cycle and positive near-term read-through for Texas capital goods/energy-linked manufacturing.

Cluster B — Stagflationary supply pressure (medium-high). Input-price acceleration alongside the oil shock complicates a duration-bull interpretation. The implication is more near-term inflation compensation if national ISM prices confirm.

Cluster C — Volatile regional survey (high caution). The headline is sentiment-based and oil-intensive; one state survey should not override national ISM, payroll-hours or hard production data. Public named same-day analyst commentary was sparse, so the clusters rely chiefly on the primary detail and contemporaneous market coverage rather than invented sell-side views.

Consensus / disagreement / resolver. Consensus: the August Texas report improved sharply. Disagreement: national leading signal versus energy-region volatility. ISM manufacturing, industrial production, payroll hours and September Dallas orders/prices resolve it.

4. Complete earnings/call table

Times are EDT. “Pre-result” means the local cash close preceded the filing; no earnings causality is implied.

Rank	Company / event time	Result versus consensus or prior	Guidance / call hinge	Realized reaction / read-through
1	SAIC (SAIC) — 6:55 SEC / 7:10 wire; call 10:00	Revenue $1.880bn vs $1.76bn; adj. EPS $3.01 vs $2.31; organic growth 5.3%	Revenue/EPS/EBITDA raised; FCF >$600m retained; quarterly book-to-bill only 0.6x	Opened +11.5%, peaked +13.3%, closed $128.22, +1.79%. (release)
2	LexinFintech (LX) — 1:00; call 7:00	Revenue RMB3.187bn, −11.2% y/y; adjusted profit −76.4%; no defensible public consensus	Q3 originations to fall significantly and company expects a net loss; liquidity first	$1.07, −9.32% on 3.3x Friday volume; negative for China online consumer finance. (6-K)
3	So-Young (SY) — 6:00; call 7:30	Revenue $74.4m vs $70.4m; loss/ADS $0.03 vs $0.09; treatment revenue +129.5%	Q3 treatment revenue RMB352m–362m; no consolidated profit guide	$2.93, +22.08% on extreme volume; clinic unit economics improved. (release)
4	Air China (0753.HK/601111.SS) — Sunday 6:59 release; briefing 3:00	H1 loss RMB2.288bn, inside July warning; fuel cost +34.7%; Big Three combined loss RMB8.2bn	Summer demand below expectations; no H2 fuel/fare/FX or profit guide	H-share −5.04%, A-share −1.85%; peers lower. (filing)
5	China Resources Land (1109.HK) — Sunday 6:55 p.m.; afternoon briefing	Revenue RMB67.87bn, −28.5%; core profit RMB10.16bn, +1.6% and above CICC preview; recurring revenue +17%	Development cash flow weak; completed-home rules short-term negative, long-term quality filter	HK$30.06, −9.40%; policy-heavy developer basket fell 6.2%. (filing)
6	Cango (CANG) — 5:00 p.m.; call 9:00 p.m.	Revenue $50.8m vs one-estimate $60.0m; GAAP loss $81.6m; mining cash cost $73,313/BTC	3MW Georgia AI site expected to produce initial Q3 revenue; replay/Q&A unavailable by cut	Regular session +7.17% before result; final after-hours $2.19, −8.37% versus the $2.39 close. Trading ended before the call. (release)
7	China Merchants Bank (3968.HK/600036.SS) — Sunday 9:30 p.m. call	H1 profit RMB76.45bn, +2.0%; NIM 1.83%; wealth fees +18.4%	NIM pressure may have peaked, not rebounded; H2 loan growth near 5%; retail NPL pressure	H-share +3.94%, A-share +1.96%; positive but sector/dividend contaminated. (filing)
8	Yankuang Energy (1171.HK/600188.SS) — Sunday 9:30 p.m. briefing	CAS profit +47.8%, recurring only +4.0%; Q2 recurring profit +108.5% y/y	H2 must recover volumes and reverse H1 unit-cost inflation; interim payout modest	H-share +1.83%, A-share +2.61%; positive coal-sector tape. (HKEX)
9	Horizon Robotics (9660.HK) — 5:26; call 7:30	Revenue RMB2.055bn, +32.9% but below thin public proxies; adjusted loss RMB1.671bn, +25.4%	>RMB5bn FY revenue, 5m chips, around-2028 breakeven reiterated	−0.53% pre-result; Sept. 1 first reaction. (filing)
10	Momenta (06880.HK) — 4:48; evening call, exact clock unavailable	Revenue RMB1.602bn, +75.9%; GM 73.2%; adjusted loss only RMB14.1m	219 nominations, R7 H2 launches, L3 2027; no numerical FY guide	+5.65% pre-result; Sept. 1 first reaction. (filing)
11	Z.ai (2513.HK) — 6:56; evening briefing	Revenue RMB953.9m, +399.7% but ~29.6% below Bloomberg preview; adjusted loss +12.1%	API now 86.5% of sales; usage run-rate claims not backlog; no formal guide	+9.63% pre-result, heavily driven by MSCI inclusion/model-sector rally. (filing)
12	Zhongsheng (0881.HK) — 12:01 a.m.; briefing ~3:00	Revenue RMB63.0bn, −18.5%; GP +20%; attributable profit −89.1%	NEV stores 102 in June/135 August; target 300 and 35% December vehicle mix	+3.77% after lunch-break filing on 2.75x volume. (Quartr)
13	Black Sesame (2533.HK) — 12:19 a.m.	Revenue RMB458.5m, +81.3%; GM 49.5%; adjusted loss +17.2%	A2000/C1200 H2 SOP; R&D 188% of sales; material funding/going-concern uncertainty	HK$9.92, −7.29% after filing; smart-driving cohort also weak. (filing)
14	OSL Group (0863.HK) — 10:31	IFRS revenue HK$55.8bn is gross principal flow; revenue less asset cost HK$130.9m, −33%; continuing loss HK$860.9m	Volumes +241%, but no net take-rate/breakeven guide; Banxa added loss	−1.35% pre-result; Sept. 1 first reaction. (accounting notice)
15	United Laboratories (3933.HK) — 10:58	Revenue RMB6.166bn, −18% / 4.6% below one estimate; profit RMB347.6m, 2.2% above warning; GM −17.8 pp	Product revenue ex-license nearly flat, but all segment profits −68% to −74%; no dividend	−3.40% pre-result, in line with pharma peers; Sept. 1 first reaction. (filing)
16	China Cinda (1359.HK) — 4:37	Attributable profit RMB784m, −65.6%, within warning; continuing ops loss RMB350m	Property is 79.6% of restructured book; core Tier 1 9.04%; no public call	−1.07% pre-result; Sept. 1 first reaction. (official)
17	CITIC FAMC (2799.HK) — 5:30	Attributable profit RMB6.862bn, +11.3%; ECL −67%, but associates/JVs exceeded PBT	Funding cheaper; capital adequacy −81 bp to 15.71%; no dividend	−1.56% pre-result; Sept. 1 first reaction. (filing)
18	CDB Leasing (1606.HK) — 6:13	Net profit RMB2.908bn, +21.1%; operating-lease income +17.7%; finance-lease income −12.7%	Stage 2 +43.5%; finance-lease NPA 1.36% from 1.05%; no H1 call found	−2.01% pre-result; Sept. 1 first reaction. (filing)

5. Detailed company sections and opinion clusters

5.1 SAIC: the operating beat was real; the bookings veto was real too

Facts. Revenue beat by 6.8% and adjusted EPS by about 30%; FY27 revenue rose to $7.2bn–$7.3bn, adjusted EPS to $10.65–$10.75 and EBITDA to $750m–$755m. Management expects second-half on-contract growth near 5%, but a 350 bp RITS roll-off and planned investment pull second-half margins to the high-9% range. Backlog was $22.1bn; book-to-bill was 0.6x quarterly and 0.8x trailing, although a delayed $740m award arrived just after quarter-end. (release, call)

Beat-and-raise cluster (high). Zacks, Benzinga, Newsquawk and the Q&A framing from Jonathan Siegmann/Stifel and Sheila Kahyaoglu/Jefferies emphasized 9% on-contract growth, faster outlays and guidance above pre-print consensus. Implication: upward FY27 estimates over 0–3 months. Risk: one point of growth was nonrepeat materials and the comparator had a favorable settlement. Orders/valuation cluster (high). Jefferies retained Hold/$130; Goldman’s Noah Poponak and BNP’s Matt Akers pressed on leading indicators. The opening surge’s near-total fade made this the marginal price-setting cluster. A sustained >1x book-to-bill disconfirms it. Orbit option cluster (medium). Kahyaoglu, Seth Seifman/JPM and Gautam Khanna/TD Cowen tested the $150m savings and fixed-price upside. Two-thirds is reinvested and the 11% margin target is around FY30, making this an 18–36 month option with execution/underbidding risk. (Zacks, Newsquawk, call)

Consensus / disagreement / resolver. Q2 deserved estimate upgrades; durability did not. The dispute is a durable government-outlay recovery versus a front-loaded quarter. Bookings, submit volume, second-half OCG/margin and quantified Orbit savings resolve it.

5.2 LexinFintech: funding stress became an explicit loss cycle

Facts. Revenue fell 11.2%, adjusted profit 76.4%, and 90+ day delinquency rose to 3.6%. Cash plus restricted cash fell to RMB2.48bn from roughly RMB3.20bn in Q1. Management said a late-June peer-risk event tightened institutional funding; Q3 originations will fall significantly and the company expects a loss as revenue contracts, provisions rise and restructuring costs land. (6-K, call)

Funding-to-loss loop (medium-high). Patrick Donovan/Fly on Wall Street and management’s own guide focused on lower funding, lower volume, precautionary provisioning and delayed capital returns reinforcing one another over 1–3 quarters. Partner whitelisting and current-asset coverage are disconfirmers. De-risking bridge (low-medium). Management and secondary EarningsIQ/Finsee analysis emphasized shorter loans, FPD30 below 1%, AI cost reduction and e-commerce revenue +60.8%. Rising 90+ delinquency and fast cash decline challenge a short trough. Cheap without an inflection (medium-low). Astrada Advisors’ dated pre-print Neutral view remains relevant: valuation is low, but funding, credit and margins need evidence. Same-day public human commentary was extremely sparse; no post-call target change was found. (Donovan, EarningsIQ, Astrada)

Consensus / disagreement / resolver. Liquidity conservation dominates diversification. The sharp dispute is short-lived funding shock versus a multi-quarter adverse loop. Q3 loss, partner capacity, unrestricted cash, collections, provisions and the FY26 dividend decision resolve it.

5.3 So-Young: clinic unit economics inflected before group profitability

Facts. Treatment revenue rose 129.5%; 47 of 65 centers were company-defined profitable and 51 cash-positive, excluding corporate allocations. Same-store sales rose 52%, visits 145%, clinic gross margin reached 28.1%, and the group loss narrowed to RMB22.7m. Consolidated gross margin nevertheless fell to 44.1% from 51.3% as owned clinics displaced higher-margin platform revenue. (release, call)

Unit-economics inflection (medium-high). Citi’s Nelson Cheung and Haitong’s Daisy Chen pressed the margin/breakeven path; public same-day summaries highlighted center maturation and traffic. The 6–12 month implication is operating leverage for scaled clinic networks. Corporate overhead, declining cash and 18 still-unprofitable centers are risks. Lower-margin migration (high). Clinics are succeeding while information/reservation revenue fell 35%; absolute gross profit grew but group margin compressed. Implication: higher revenue with a structurally different multiple and cash profile. Data/supply-chain flywheel (medium-low). CITIC’s Jinpeng He and GF’s James Zhang tested collagen products and AI clinics. More than 66,000 Miracle Collagen units and Q4 AI-clinic plans are evidence; monetization and retention remain unproved. (call, TipRanks recap)

Consensus / disagreement / resolver. Clinic momentum and utilization improved; consolidated breakeven did not arrive. Q3 clinic revenue/margin, corporate cash burn, mature-center cohorts and repeat purchase resolve “operating inflection” versus “growth bought through expansion.”

5.4 Air China: fuel and weak peak-season demand overwhelm capacity growth

Facts. Revenue rose 10.5%, but jet-fuel cost rose RMB8.44bn/34.7% and attributable loss widened to RMB2.29bn. The Big Three lost RMB8.2bn combined. Management said international summer demand exceeded domestic, but total peak-season demand was below expectations and issued no H2 fuel, fare, FX or profit guide. (filing, Reuters)

Fuel-demand trap (high). HSBC’s airline team projected a RMB16.8bn full-year combined Big Three loss versus the market’s prior RMB1.3bn profit expectation; Flight Master and aviation-market commentary focused on fares/capacity that cannot absorb fuel. Implication: estimate cuts and weak airline cash/equity over 3–9 months; crude retreat or materially higher yields disconfirm. International/capital-repair option (medium). Li Dan/Guojin’s Buy framing and management highlighted international recovery, Cathay contribution, lower finance cost and recapitalization. That is a 2027 option, not an H2 profit bridge. Thin named public post-briefing research limits confidence. (Reuters/HSBC, Investing.com)

Consensus / disagreement / resolver. H1 was pre-signaled; the summer/fuel outlook worsened. H2 fuel per tonne, passenger yield/load factor, international capacity, monthly traffic, FX and any capital raise resolve the recovery horizon.

5.5 China Resources Land: a recurring-income winner inside a development reset

Facts. Development-property revenue fell 39.1% and its gross margin was about 10%; recurring revenue rose 17%, recurring core profit 10.4%, group core profit 1.6% and consolidated gross margin 1.4 points. Gearing was 41% and average funding cost 2.63%; the interim dividend stayed RMB0.20. (filing, Quartr)

Recurring flywheel (medium-high). CLSA, UOB’s Liu Jieqi/Damon Shen and management emphasized malls, offices and fee income supporting core profit/dividend despite completions. Implication: relative SOE quality over 12–24 months. Presale cash shock (high near term). Nomura, Everbright and management said delivery-first rules pressure cash flow. The 9.4% fall, versus a 6.2% developer-index loss, priced both sector and company development risk. Demand is binding (medium-high). Standard Chartered’s Shuang Ding argued weak employment/income and falling-price expectations make cheap funding insufficient. (CLSA recap, UOB, Reuters)

Consensus / disagreement / resolver. Recurring assets make CR Land a survivor; they do not immunize development economics. Sales, completed-home conversion, land spend, cash collections, development margin and rental growth resolve whether the selloff overstates relative quality.

5.6 Cango: mining retrenchment, accounting damage and unproved AI optionality

Facts. Mining revenue fell to $47.4m from $98.4m in Q1 as operating hash rate declined to 27.58 EH/s and output to 656 BTC. Cash cost improved 4.7% to $73,313/BTC; adjusted EBITDA remained negative $10.7m. A $42.9m machine impairment, $8.5m disposal loss and $4.1m crypto fair-value loss contributed to the $81.6m GAAP loss. Cango ended with 1,056 BTC and $10.1m cash and said its converted 3MW Georgia site should begin AI-infrastructure revenue in Q3. (release)

Efficiency-first reset (medium). Management argues retiring S19 machines, leasing capacity and hedging improve unit economics; lower cash cost and sharply better adjusted EBITDA support the claim. Lower output/revenue and still-negative EBITDA limit it. AI optionality (low-medium). CEO Paul Yu and dated independent context from Boaz Sobrado/Forbes frame distributed inference sites as a higher-value reuse of power/infrastructure. No material revenue, customer economics or utilization was disclosed; 3MW is pilot scale. Balance-sheet/accounting veto (high). Impairments, a thin cash balance and an unreconciled BTC treasury bridge dominate near-term risk. No public institutional post-print analyst note was available.

Consensus / disagreement / resolver. Mining efficiency improved, but the smaller operation is still loss-making and AI is pre-proof. Q3 AI revenue/customers, BTC sales/treasury reconciliation, cash, self-mining EH/s and adjusted EBITDA resolve it. At 9:34 p.m. EDT, Cango's IR site/API still ended with the June 1 Q1 event: no Q2 player, recording or genuine transcript was available, and phone access/replay required code 8654407. A Benzinga page labeled Q2 recycled Q1 material and was excluded. Q&A is therefore unverified—an access gap, not evidence that the call was canceled. After-hours trading ended at 8:00 p.m., before the 9:00 p.m. call, so no live post-call price reaction exists.

5.7 China Merchants Bank: fee moat versus the new retail-credit fault line

Facts. H1 net operating income rose 4.8% and profit 2.0%. Broad wealth income rose 18.4%, retail AUM 8.0% and Q2 NIM fell only 1 bp sequentially to 1.82%. Retail NPL rose to 1.16%, card NPL 1.90% and consumption-loan NPL 1.39%; corporate-property NPL improved to the official 4.27%. CET1 was 14.07% and the 35% interim payout framework stayed. (filing, call)

Wealth moat reaccelerates (medium-high). Guosen’s Tian Weiwei/Wang Jian maintained Outperform; HSBC’s Gary Lam and CICC’s Zhang Shuaishuai tested sustainability. Funds/trusts and non-deposit AUM support fee diversification over 6–12 months, but management said the H1 pace cannot persist indefinitely. NIM trough, not rebound (high). Goldman’s Yang Shuo and Guolian’s Wang Xianshuang pressed funding competition; management said the hardest compression may have passed while refusing a rebound call. Quality over scale (medium-high). Morgan Stanley’s Xu Ran and Guotai Junan’s Ma Tingting focused on slower loan growth and retail risk. Lower growth protects capital; card/consumer NPL formation is the disconfirmer. (Guosen, call)

Consensus / disagreement / resolver. CMB remains a fee-rich, better-quality bank; the dispute is whether wealth/NIM stabilization outweighs household deleveraging. H2 NIM, retail NPL formation, fee/AUM growth, loan growth and payout settle it.

5.8 Yankuang Energy: genuine Q2 recovery under one-off H1 headlines

Facts. The Xintai sale added RMB2.84bn; recurring H1 profit rose only 4%, but Q2 recurring profit rose 108.5% y/y. BofA calculated Q2 coal unit profit more than doubled q/q and chemical unit profit rose 54%; H1 self-produced coal cost nevertheless rose 7.2%, and capex was only 29% of plan. (SSE filing, BofA public summary)

Recurring inflection (high current / medium duration). BofA raised its H target to HK$15 and retained Buy; Guolian Minsheng and Sxcoal highlighted coal/chemical margin recovery. Payout/cost cap (medium-high). BofA said the RMB0.20 dividend could disappoint; H2 must raise volume, execute capex and reverse unit-cost inflation. Sector beta (medium). China Coal outperformed and peers rallied, so not all of Yankuang’s move was briefing alpha. The public briefing transcript/minutes were unavailable; no Q&A is reconstructed. (BofA summary, Sxcoal)

Consensus / disagreement / resolver. Q2 economics improved, while headline H1 growth is not sustainable. Q3 volume, realized prices, unit costs, chemical spreads, capex/FCF and final dividend resolve durability.

5.9 Horizon Robotics: share and licensing gains still consume extraordinary R&D

Facts. License/services revenue rose 52.7% at 90.4% margin and Journey shipments rose 12.1% to 2.22m; bundled controllers pulled product margin down eight points. R&D was RMB2.76bn/134% of revenue and adjusted net loss widened 25.4%. The RMB3.78bn IFRS profit was driven by fair-value and deconsolidation gains, not operating breakeven. (filing, call)

Platform flywheel (medium-high). Management, Xinhua Finance and thin public CMBI context emphasize share gains and high-margin licensing. Loss/cash veto (high). Morgan Stanley’s Tim Hsiao challenged R&D returns; UBS’s dated preannouncement view cut estimates despite retaining Buy. H2 must deliver at least RMB2.95bn revenue and 2.78m chips to meet guidance math. R&D moat or trap (medium). Journey 7/HSD design wins can build a multi-year moat; delays, OEM insourcing and controller price competition would lock in losses. No same-day post-call target revision was public. (call, CMBI preview)

Consensus / disagreement / resolver. Share/licensing momentum is real; headline IFRS profit is not. H2 shipments/revenue, product margin, adjusted loss/cash, HSD program conversion and 2028 breakeven progress resolve the capital-intensity debate.

5.10 Momenta: near adjusted break-even, with NRE and L4 caveats

Facts. Installations rose 83.7% to 321,000, mass-produced models reached 105 and nominations 219, including 114 pre-SOP. Technical development/NRE remained 62% of revenue. The RMB16.5bn IFRS loss was mostly preferred-share remeasurement that converted at IPO; adjusted loss narrowed 96.6%, but operating cash outflow was RMB381m. (filing, call)

Mass-production flywheel (high current / medium duration). Phate Zhang/CnEVPost and Q&A framing from CICC’s Liping Zhao and Morgan Stanley’s Junming Zhang focused on scale outrunning R&D and narrowing cash burn. Backlog versus ASP pressure (medium-high). BofA’s Joey Yang pressed price-war effects; management conceded future volume discounts while defending R7 adoption. L4/global optionality (low-medium). Deutsche Bank’s Bin Wang and CITIC’s Yiqi Lian tested regulation and Robotaxi economics. Permits/partners are real; paid unit economics and safety evidence are not. (CnEVPost, call)

Consensus / disagreement / resolver. Core operating leverage is strong; the dispute is sustainable FCF versus a near-break-even high-water mark before R7/L4 spending. H2 installations, pre-SOP conversion, license ASP, cash flow and paid autonomous deployments resolve it.

5.11 Z.ai: an API usage explosion with subscale economics

Facts. API/open-platform revenue rose 27-fold to RMB825m/86.5% of sales and reached positive 24.6% segment gross margin. Blended gross margin fell to 26.4%, R&D was RMB2.13bn/223% of revenue and adjusted loss widened 12.1%. Management’s $1.6bn monthly and $2.0bn weekly annualized rates are spot usage velocity—not contracted ARR/backlog—and revenue missed the Bloomberg preview materially. (filing)

MaaS inflection (medium-high). CMBI’s Saiyi He/Ye Tao and Bernstein’s dated Buy views, management and same-day Chinese media see migration from private deployments to API as validation of model usage and pricing. Growth without economic scale (high). The Bloomberg miss, halved blended margin and R&D/loss expansion say usage is far from self-funding; Goldman/HSBC’s dated neutral context supports caution. Strategic funding/index optionality (medium-low). Placement cash and MSCI inclusion improve capital access, but the pre-result 9.6% rally was mechanical/sector-contaminated. No named same-day sell-side note or verifiable call roster was available. (CMBI public extract, Bernstein summary, KraneShares)

Consensus / disagreement / resolver. API demand is real; profitable monetization is not. Q3/H2 API take-rate and gross margin, compute cost, R&D intensity, cash burn, enterprise retention and first post-result price/broker revisions resolve it.

5.12 Zhongsheng: gross-profit repair before cash/profit repair

Facts. Group gross profit rose 20% and margin 260 bp; the new-car loss narrowed 74% but remained RMB631m and aggregate new-car margin was −0.4%. After-sales margin rose to 48.4%. Attributable profit and operating cash flow fell about 89%, while free cash flow was negative RMB735m. NEV stores increased from 102 in June to 135 in August, with 300 targeted by year-end. (Quartr)

Margin inflection (medium-high). Same-day China Wealth/HSTong coverage and dated Citi/UBS/China Merchants/CITIC positive baselines emphasize inventory discipline and after-sales resilience. Cash-quality skepticism (high). BofA’s Underperform and Morgan Stanley’s Equal-weight priors plus primary cash data show accounting gross-profit repair has not reached owners. NEV transition option (medium). Rapid store build and 35% December mix can restore growth, but requires working capital and may dilute margins. No full public Q&A or same-day named broker note was available. (ETNet broker panel)

Consensus / disagreement / resolver. Dealer economics improved at gross-profit level; net profit/cash did not. H2 new-car margin, inventory days, OCF/FCF, NEV store productivity and OEM rebates settle whether +3.77% was justified.

5.13 Black Sesame: product traction meets a funding veto

Facts. Revenue rose 81.3%, shipments exceeded 6m and gross margin doubled to 49.5%. Adjusted loss widened 17.2%; R&D was RMB860m/188% of revenue and operating cash burn nearly doubled to RMB1.02bn. The filing identifies material going-concern uncertainty dependent on timely borrowing/private placements. (filing, RTTNews)

Auto-chip ramp (medium-high). Guotai Haitong’s Zhao Muyang, CMSI and Huatai/BofA dated baselines emphasize A2000/C1200 H2 SOP and domestic ADAS penetration. R&D moat versus burn (high risk). Bernstein’s Qingyuan Lin and the primary cash disclosures frame the central issue: growth and margin do not cover development expense. Robotics/Eeasy optionality (low-medium). New markets broaden TAM but also increase execution/capital demands. No public call/transcript or same-day human post-print note was found. (Guotai Haitong, CMBI, Bernstein analyst record)

Consensus / disagreement / resolver. Product traction is credible; financing self-sufficiency is not. H2 SOP revenue, customer concentration, gross margin, R&D/cash burn, borrowing/private placement completion and first post-result reaction resolve it.

5.14 OSL: gross stablecoin flows are not economic revenue

Facts. OSL retrospectively presents principal digital-asset sales gross: HK$55.813bn of “revenue” less HK$55.682bn asset cost leaves HK$130.9m, down 33%. Ex-market-making client volume rose 115%, but continuing loss widened to HK$860.9m; Banxa contributed HK$2.04bn gross revenue and an HK$81.6m loss. February’s placement expanded shares 13.6%. (accounting notice, release)

Regulated rails (medium). Kevin Cui/OSL and CoinShares’ Satish Patel see regulated stablecoins moving into institutional payments; licenses/banking relationships may matter over 1–3 years. Gross-flow normalization (high). The economically relevant spread fell despite headline revenue growth; conventional price/sales comparisons are invalid. License moat versus cost/dilution (high risk). Geographic expansion, MiCAR/AFSL scope and USDGO are strategic assets, but operating costs, Banxa and financing burden remain unproved. No public call, transcript or named sell-side note was available. (CoinShares, RTTNews)

Consensus / disagreement / resolver. Volumes and regulatory footprint grew; unit economics did not. Net take-rate/contribution margin, monthly cash burn, Banxa synergies, non-APAC client mix, USDGO economics and dilution settle it.

5.15 United Laboratories: licensing normalization exposed commodity margins

Facts. Novo/UBT251 license income fell RMB1.33bn and explains nearly all top-line decline; product revenue excluding licensing fell only 0.3%, and finished goods ex-license rose 8.6%. Yet gross margin fell 17.8 points and profits in intermediates, bulk medicine and finished products fell 68%–74%. FX loss was RMB186m; no interim dividend was declared. (filing)

API trough (medium-low). Guoyuan’s Lin Xingqiu, First Shanghai and dated UBS context argue 6-APA/API prices are bottoming and vertical integration creates operating leverage. H1 bulk revenue +5% with profit −68% challenges the timing. Pipeline/formulation value (medium). Guoyuan/First Shanghai and Novo development disclosures treat UBT251, insulin and finished goods as long-duration value. Commodity cyclical (medium-high caution). Latent Tensor/MetroHK and the primary margin/cash data say the pipeline is optionality atop weak current economics and dilution. No same-day named human post-print note was found. (Guoyuan, First Shanghai, Novo)

Consensus / disagreement / resolver. H1 weakness was pre-signaled; UBT251 is the long-duration asset and API spreads drive the near term. Sept. 1 Q&A/reaction, 6-APA spreads, insulin uptake, clinical milestones, FX/capex and dividend restoration resolve it.

5.16 China Cinda: reported attribution masked a loss-making core

Facts. Attributable profit fell 65.6%, but total profit fell only 17.4% because losses allocated to minority interests widened. Continuing operations lost RMB350m; the core AMC segment lost RMB1.01bn, while held-for-sale Cinda Securities contributed RMB1.10bn discontinued profit. Property was 79.6% of restructured assets; property gross margin was −0.2%. Impairments/funding costs fell, but RMB1.98bn distressed fair-value income was unrealized. (official, HKEX)

Policy AMC franchise (medium-low). JPMorgan’s dated Aug. 4 context and Western Securities’ Sun Yin/Chen Jing see national AMCs benefiting from workout supply and state support. Core-quality bear (high). Continuing/core losses, property concentration and unrealized marks say the profit warning understates economic fragility. Capital/restructuring option (medium-low). Securities disposal and lower funding can repair capital, but core Tier 1 of 9.04% and no current call/Q&A leave uncertainty. No same-day post-print sell-side note was public. (JPMorgan extract, Western Securities, Xinhua Finance)

Consensus / disagreement / resolver. Policy relevance and cheaper funding remain; core cash earnings are weak. Disposal proceeds, realized recoveries, property exits, capital ratios, impairments and Sept. 1 reaction resolve it.

5.17 CITIC FAMC: lower credit cost, but associate and mark dependence

Facts. Attributable profit rose 11.3%, normalized income 36.6%, ECL fell 67% and funding cost fell 16 bp. Associates/JVs contributed RMB11.06bn—more than PBT—while tax benefit doubled and fair-value marks drove distressed-asset gains. Capital adequacy fell to 15.71% and no dividend was declared. (official results)

Self-reinforcing turnaround (medium). Fitch’s July BBB+/Stable action, Guotai Haitong/Western dated positive ratings and same-day Zhitong analysis link CITIC integration, cheaper funding and lower credit costs. Quality/capital cap (medium). CICC/Guosen dated Neutral baselines and primary accounting show profits still depend on associates, fair value and tax while capital declines. Property/NPL supply duality (medium-low). More distressed supply is opportunity only if recoveries and exit liquidity hold. No public H1 call or same-day named human sell-side note was found. (Zhitong/StockStar, broker history, ETNet)

Consensus / disagreement / resolver. Operating momentum improved; sustainable cash earning power is unproved. H2 cash recoveries, ECL/marks conversion, property exits, funding/capital, payout and Sept. 1 price/credit reaction resolve it.

5.18 CDB Leasing: funding advantage against leading credit deterioration

Facts. Operating-lease income rose 17.7%, interest expense fell 5.2% and impairments 23.8%, lifting net profit 21.1%. Aircraft PBT rose 27.3%, shipping 70.6% and energy 40.3%. Against that, Stage 2 exposure rose 43.5%, special mention 33.4%, finance-lease NPA ratio to 1.36% and provision coverage fell to 404%. (filing)

Funding/operating-lease flywheel (medium-high). Same-day HSTong/Xinhua analysis and China Securities International’s dated Buy thesis emphasize quasi-sovereign funding and 7% aircraft net yield. Hidden credit migration (high facts / medium loss implication). Leading indicators worsened precisely as provisions fell; large coverage and stronger capital are the offsets. Better franchise, more asset-cycle risk (medium). New rules favor direct equipment leasing, but aircraft commitments, shipping rates and fast green-energy growth replace borrower risk with residual/utilization risk. No public call or same-day named individual analyst note was found. (HSTong, Xinhua, China Securities)

Consensus / disagreement / resolver. Current earnings are strong and buffers ample; the dispute is whether provision relief is premature. H2 Stage 2/NPA/credit cost, aircraft placements/sales, shipping rates, energy delinquencies and final payout resolve it.

6. Cross-event themes and notable contradictions

1. The day’s inflation signal came from barrels, not the scheduled CPI. German headline/HICP undershot consensus and core/services did not broaden, but Hormuz lifted crude nearly 3% and kept Bunds/Treasuries pressured. The contradiction favors near-term headline inflation compensation without proving a domestic wage-price spiral.

2. China has a production recovery without a domestic-demand recovery. Manufacturing orders, exports and high tech improved; services, construction, property funding and household credit remained weak. Horizon, Momenta and Z.ai show strategic-tech adoption, while developers, airlines, online lenders and auto dealers reveal the cost of weak private demand and expensive inputs.

3. Risk transfer repeatedly beats nominal stimulus. China’s mortgage rules lower household delivery risk but move funding risk to developers and lead banks. California’s bill strengthens survivor/insurance protections but leaves utilities bearing an uncapped tail. In both cases, the protected stakeholder is clearer than the provider of new risk capital—hence sharp equity repricing.

4. Headline accounting numbers were unusually unhelpful. Horizon’s IFRS profit and Momenta’s enormous IFRS loss were both liability/deconsolidation accounting; Z.ai’s usage run rates were not ARR; OSL’s HK$55.8bn revenue was gross principal flow; Cinda/FAMC profit quality depended on minorities, associates, marks or tax; Yankuang’s headline included a disposal. Adjusted cash economics, not statutory direction, were the reliable comparison.

5. Funding costs improved while credit quality weakened. CMB’s NIM decline slowed, FAMC funding/ECL fell and CDB Leasing earned more on cheaper liabilities, but retail NPLs, property/workout dependence and Stage 2 migration rose. Lower funding costs are a bridge, not evidence that household/property credit has healed.

6. The market rewarded revenue scale only when a path to unit economics was visible. So-Young rose 22% on traffic, same-store and center cash generation; SAIC’s opening beat faded on weak bookings; Lexin fell on an explicit loss/liquidity guide; Black Sesame sold off despite revenue/margin growth because R&D/cash burn threaten financing. Momenta is the strongest operating-leverage example, but its first post-result price test is still pending.

7. First-session chronology prevents false narratives. CR Land, Air China, CMB and Yankuang had valid Monday reaction/call evidence. Horizon, Momenta, Z.ai, Cinda, FAMC, OSL, United Labs and CDB Leasing filed after their local closes; their Monday moves are positioning, not reactions. That distinction materially changes what can be inferred from the tape.

7. Coverage audit

Calendars and source sets checked

• Economic/macro calendars: Borsa Italiana/FactSet global calendar; Briefing/Kiplinger U.S. calendar; official schedules/releases from the Dallas Fed, Destatis, India MoSPI/PIB, China NBS/PBOC/NFRA/CSRC, U.S. Treasury, and relevant government/legislative/military authorities. AP and Reuters market-close/cross-asset reports were used to cross-check realized reactions.
• Earnings inventories: EarningsWhispers and MarketBeat U.S. release/call lists; company IR/SEC filings; HKEX predefined “Today’s Results” search and issuer title feeds (1,573 records inspected, with 289 Aug. 31-HKT/EDT-window records narrowed by size, move and sector importance); ASX results/calendar checks; First Shanghai/HK financial calendars; public call/transcript services only where their period labeling could be verified.
• Primary company sources: SAIC, Lexin, So-Young and Cango IR/SEC; HKEX filings and company/roadshow archives for all 14 included Asian issuers; official or company-hosted call material where available. Public consensus was reported only when the sample/definition was defensible; otherwise the report says none.
• Opinion research: named same-day sell-side, economist, buy-side/rating or expert commentary was prioritized. When no public same-day human note existed, dated pre-print context and clearly labeled market/editorial expertise were secondary. Analyst questions are evidence of the debate, not inferred ratings. No access to unavailable paywalled notes is implied.

Borderline events and companies excluded

• Chicago PMI: some calendars displayed Aug. 31, but the actual August release was Friday Aug. 28 (47.1) and was covered in that date’s report; excluding prevents double counting.
• New Zealand business confidence; Australian profits, inventories and private credit; Hong Kong retail sales; routine sovereign bill auctions/settlements: checked, but no realized or plausible U.S./global-market impact comparable with the seven included events was found. Their effects were small, regional or overwhelmed by China/property/Hormuz.
• BioLineRx (BLRX): reported/called, but the micro-cap pipeline/cash update produced only a small move and no important sector read-through.
• Pyxis Tankers (PXS): Q2 revenue $12.1m and EPS $0.27 versus a thin $0.20 estimate, but approximately $56m market value and no outsized move/novel tanker read-through. (release)
• Aduro Clean Technologies (ADUR): calendar entry did not produce a new qualifying earnings result/call by cut.
• Other HKEX deadline filings: hundreds of small issuers were screened and excluded where market capitalization, price response, sector importance and new information were all immaterial. Inclusion was not based on a fixed watchlist.

Pending calls, reactions and material gaps

• Cango: at 9:34 p.m. EDT, company IR/API still had no Q2 player, recording or genuine transcript; live/replay phone access required code 8654407. Its Q&A remains unverified, which is an access gap rather than evidence of cancellation. A Benzinga page labeled Q2 recycled Q1 material and was excluded. The final after-hours print was $2.19, −8.37%, but the 8:00 p.m. trading close preceded the 9:00 p.m. call.
• No public full transcript/Q&A found by cut: Cango, Yankuang, Z.ai, China Cinda, CITIC FAMC, OSL, CDB Leasing and Black Sesame. China Resources Land’s exact briefing start time and Momenta’s exact call clock were unavailable. Air China’s SSE event was limited text Q&A, not a conventional analyst transcript.
• Scheduled after this trading date: United Laboratories’ results briefing is Sept. 1 HKT; its Q&A and the first post-result Hong Kong session are unavailable here.
• First price reaction pending Sept. 1 HKT: Horizon, Momenta, Z.ai, China Cinda, CITIC FAMC, OSL, United Laboratories and CDB Leasing. Their Aug. 31 closes are explicitly labeled pre-result. Cango has after-hours U.S. trading but no ordinary overnight cash close comparable with the 4:00 p.m. session.
• Consensus gaps: no robust public line-item consensus for Lexin, Air China H1, Momenta, Cinda, FAMC, OSL, CDB Leasing or Black Sesame; Horizon/CR Land/United Labs/Cango estimates were one-source or thin and are labeled accordingly. Estimate-provider accounting definitions conflicted for CMB and Cango EPS.
• Market-data gaps: no trustworthy public same-day CDX/GCC sovereign spread, California utility CUSIP/CDS basis-point close, China developer-bond spread or tanker war-risk-insurance close was found. No numbers were invented. Several asset windows were contaminated by Hormuz, Friday’s Warsh repricing, month-end flows, China’s simultaneous PMI/property package and broad sector moves.
• Source caveats: company-cited market-share/safety/run-rate metrics are labeled as claims; machine/editorial summaries are never passed off as sell-side opinions; post-close Asian price moves are never called result reactions. The official SB 492 history was rechecked at 9:33 p.m. EDT and still showed “ordered to third reading,” with no Aug. 31 final action posted.