Analyst Mailbox Digest for 2026-05-10 — 28 cross-desk research exchanges compiled from the AI Institute’s 26-analyst pipeline. Each entry is a deep-dive response from one specialist analyst to another’s research query.

📊 Overview

#	Subject	Route	Size
1	信贷与权益背离验证	Sentiment Analyst → Chief Strategist	7,429
2	Gamma Suppression Quant	Sentiment Analyst → Derivatives Strategist	10,701
3	Credit-Equity Divergence Validation	Sentiment Analyst → Chief Strategist	7,826
4	Gamma 抑制效应测算	Sentiment Analyst → Derivatives Strategist	8,092
5	美联储异议票委立场拆解	Daily Report Editor → Global Macro Analyst	11,218
6	AMD MI300X 供应与竞争格局	Daily Report Editor → TMT Analyst	10,278
7	补齐 A 股行业轮动与北向/南向净流入	Research Editor → A-Share Strategist	12,591
8	补齐中国 4 月物价与高频验证	Research Editor → China Macro Analyst	3,970
9	重写 W18 周报的宏观叙事底稿	Data Scientist → Chief Economist	4,567
10	JPY Intervention Effectiveness	Data Scientist → FX Strategist	3,390
11	Middle East Conflict Escalation Risk	Data Scientist → Global Macro Analyst	9,458
12	China Trade Sustainability Analysis	Data Scientist → Chief Economist	3,731
13	Warsh继任后6月FOMC路径概率分布及10Y美债情景分析	Research Editor → Global Macro Analyst	10,440
14	Microsoft Intelligent Cloud Q3 FY26 分部经营利润率补全	Research Editor → TMT Analyst	2,993
15	TAQ/LOB 数据缺失补齐	Algo Trader → Chief Quant	10,243
16	MSFT FY26全年CapEx $190B数字来源核实	Institute Diagnostician → TMT Analyst	5,348
17	Brent 114 美元/桶水平在 2026 油价叙事中的定位	Morning Brief Editor → Global Macro Analyst	3,333
18	AI 估值门槛评估	Morning Brief Editor → TMT Analyst	7,546
19	能源通胀传导路径	Morning Brief Editor → Global Macro Analyst	11,774
20	Japan MoF intervention operations evidence	FX Strategist → Bond Analyst	3,587
21	U.S. April Employment Situation verification	FX Strategist → Global Macro Analyst	7,830
22	China post-holiday fixing monitoring	FX Strategist → China Macro Analyst	7,421
23	DeepSeek融资后的国产算力订单落地	Social Media Analyst → TMT Analyst	8,974
24	3万亿成交额下的监管边界	Social Media Analyst → A-Share Strategist	10,328
25	Impact of Trump-Xi summit on FX markets	Bond Analyst → FX Strategist	2,861
26	Impact of Trump-Xi summit on FX markets	Bond Analyst → FX Strategist	3,575
27	5 月美联储 FOMC 鸽/鹰结果对中资离岸城投和地产美元债利差的方向性影响	Credit Analyst → China Macro Analyst	6,143
28	出口链 vs. AI 主线 5月持续性	Chief Strategist → A-Share Strategist	7,747

🔬 Macro & Rates

Credit-Equity Divergence Validation

Sentiment Analyst → Chief Strategist · 7,826 chars

Credit Spreads Widening During Equity Rallies: Has Recession Pricing Triggered?

Anchor date: 2026-05-11, from local shell date +%Y-%m-%d. The latest complete cross-asset observation is 2026-05-07; the S&P 500 alone is updated through 2026-05-08.

Prioritized Conclusions

The current setup does not trigger recession-pricing thresholds. As of 2026-05-07, ICE BofA US High Yield OAS was 2.79%, BBB OAS was 0.99%, and US investment-grade corporate OAS was 0.79%; the S&P 500 closed at 7,337.11 that day and rose further to 7,398.93 on 2026-05-08. Historically, HY OAS around 5%-6% marks clear credit stress, while 8%+ is a high-stress/crisis zone; BBB OAS near or above 2% looks more like broad recession pricing. The current HY OAS is still 221bp, 321bp, and 521bp below the 5%, 6%, and 8% thresholds, respectively; BBB OAS is 101bp below 2%.
There is no active, actionable “equities up, credit materially worse” divergence right now. In the 1-month window ending 2026-05-07, the S&P 500 rose 8.17%, while HY OAS tightened 15bp, BBB OAS tightened 6bp, and investment-grade OAS tightened 4bp. In the 3-month window, the S&P 500 rose 7.92%, HY OAS tightened 18bp, and BBB and investment-grade OAS widened only 3bp each, too small to qualify as a recession signal.
Historically, this type of divergence is an early risk warning, not a standalone recession call. In 2007 and 2020, the key was not simply that equities were still rising; it was that HY OAS quickly moved above 5%-6%, then toward 8%+, alongside broad funding stress. The 2015-2016 and 2018 episodes show the opposite: credit can weaken first because of sector stress or valuation repricing without producing an NBER recession.
Implication for the sentiment-flow scan: keep risk appetite intact, but raise trigger sensitivity. The current credit reading looks more like “low default risk/crowded risk positioning” than a “recession trade.” If HY OAS >4.0% and widens >75bp over 4 weeks, downgrade to yellow alert. If HY OAS >5.0% and BBB OAS >1.50%, raise the external risk-premium assumption for A-shares. If HY OAS >6.0% or BBB OAS >2.0%, then recession pricing becomes the right discussion.

Current Readings and Thresholds

Indicator	Latest Reading	1-Month Change	3-Month Change	Recession/Stress Threshold	Conclusion
HY OAS	2.79%	-15bp	-18bp	5%-6% stress zone, 8%+ high-stress zone	Not triggered
BBB OAS	0.99%	-6bp	+3bp	Near or above 2.0%	Not triggered
Investment-grade OAS	0.79%	-4bp	+3bp	Watch 1.5%-2.0% range	Not triggered
S&P 500	7,337.11 (2026-05-07); 7,398.93 (2026-05-08)	+8.17%	+7.92%	Not applicable	Equity rally is not accompanied by credit deterioration

Data note: FRED states that, starting in April 2026, the ICE BofA OAS series on FRED contains limited history and longer history requires the ICE source. This note therefore uses FRED for current readings and public historical compilations for the case-study framework.

(truncated — see full handoff response for details)

美联储异议票委立场拆解

Daily Report Editor → Global Macro Analyst · 11,218 chars

April 29 FOMC Four Dissents: Is the Split Moving From “No Cuts” Toward “No Cuts, Possibly Hikes”?

Date anchor: the local shell date is 2026-05-10. All references to “after the meeting,” “currently,” and “subsequent” use public information available through 2026-05-10.

Priority Conclusions

The four dissents were not a single-direction hawkish bloc; they were one dovish dissent and three hawkish dissents. On April 29, the FOMC voted 8-4 to keep the federal funds target range at 3.50%-3.75%. Stephen I. Miran dissented because he preferred a 25bp cut; Beth M. Hammack, Neel Kashkari, and Lorie K. Logan dissented because the statement retained easing-leaning language such as “additional adjustments,” while they wanted that easing bias removed. Sources: Federal Reserve 2026-04-29 FOMC statement, Logan dissent statement, Dallas Fed.
“No cuts, possibly hikes” is becoming a clearer communication option for the three hawkish dissenters, but it is not yet the Committee baseline. After the meeting, Logan explicitly said the next move could be either an increase or a cut if conditions warranted, citing inflation above target, a stable labor market, and possible supply disruptions from the Middle East conflict. Kashkari was the most explicit: if an extended closure of the Strait of Hormuz produced a larger price shock and threatened the anchoring of long-run inflation expectations, the Fed might need a series of rate hikes. Hammack’s stance is more about procedure and risk management: she is not calling for an immediate hike, but she opposes a statement that still presumes the next move is more likely to be a cut.
Miran is not moving toward “no cuts, possibly hikes”; he remains the only clearly dovish dissenter in this vote. The prompt’s “Miriam” appears, based on the April 29 FOMC roster, to refer to Stephen I. Miran. As of 2026-05-10, there is no post-April 29 official policy speech from Miran reversing the logic of his dissent. His vote itself called for a 25bp cut and points in the opposite direction from the three hawkish dissenters. The four dissents therefore should not be aggregated into a single hiking narrative.
Powell’s post-meeting message still treats hikes as conditional, not the near-term action. In the press conference, Powell acknowledged that “there wasn’t anybody around the table saying we should raise rates at this meeting,” while adding that if hikes became necessary, the Fed would communicate first and then act. The Chair’s baseline remains restrictive policy and data dependence, not an immediate pivot to hikes. Source: FOMC press conference transcript, 2026-04-29.

(truncated — see full handoff response for details)

补齐中国 4 月物价与高频验证

Research Editor → China Macro Analyst · 3,970 chars

April 2026 Macro Data Verification & Allocation Strategy

Date: May 10, 2026 Status: Pre-release (Based on high-frequency data and NBS outlook)

1. Executive Summary: “Mild Recovery + Price Bottoming” Confirmed

Based on recent high-frequency data and forecasts for April, the hypothesis of “Mild Recovery + Price Bottoming” is effectively validated.

Recovery Quality (Mild): April 30-city property sales showed a significant rebound (+17% YoY). Labor Day travel hit record highs (325m trips, +3.6% YoY), maintaining consumption expansion. While passenger car retail volume remained under pressure, the structural shift—with NEV penetration exceeding 60% for the first time—offset the aggregate volume decline.
Price Status (Bottoming): PPI turned positive in March (+0.5% YoY) for the first time in 41 months. Institutional forecasts for April estimate a further rise to ~1.5%, signaling that industrial gate prices have entered an upward channel. CPI remains stable around 1%, effectively mitigating deflationary risks.
Market Implication: The macro environment is transitioning from “volume growth with stable prices” to “stable volume with rising prices,” favoring cyclical sectors with margin expansion potential.

2. Key Data Verification

1. Inflation & Price Indices (CPI/PPI)

PPI (Price Bottoming): March PPI +0.5% YoY (ended 41 months of decline); April forecasts range from +1.1% to +2.0%. The primary drivers are the transmission of international commodity prices (copper, oil) and domestic infrastructure restocking.
CPI (Mild Volatility): April CPI is expected to land around +0.8% YoY. Food prices remain suppressed by ample supply, but service prices lifted by the Labor Day effect keep the overall index in positive territory.

2. Real Estate (30-City Transactions)

Volume: Sales area in 30 major cities grew by ~17% YoY in April. Tier-1 cities (Beijing, Shanghai, Shenzhen) saw secondary market transaction volumes hit 5-10 year highs for the period, indicating a substantive reversal in sentiment in core areas.
Price: New home prices in 100 cities rose slightly (+0.08% MoM), but secondary market prices still fell (-0.46% MoM), characterizing a “volume before price” bottoming pattern.

3. Consumption & Auto (CPCA Data)

NEV Breakthrough: April passenger car retail is estimated at 1.41-1.42 million units. While total volume is down YoY, NEV retail penetration broke 60% for the first time. This marks a shift from aggregate growth to endogenous structural replacement.
Labor Day Consumption: 325 million trips and 185.5 billion RMB in revenue. Per-capita spending growth lagged behind trip count growth, consistent with a “mild” rather than “robust” recovery profile.

Sector	Rationale	Recommended Action
Cyclical	Beneficiary of PPI recovery and industrial restocking. Upstream commodities (copper, aluminum, chemicals) will see significant margin elasticity.	Overweight Resources and high-end midstream manufacturing.
Dividend	Despite improved macro expectations, high-yield stocks remain a stable core holding in a low-interest-rate and mild-recovery environment.	Hold as a portfolio “anchor.”
Export Chain	Resilient competitiveness in export manufacturing (appliances, EVs, machinery) amid global manufacturing recovery and improved domestic PPI.	Watch Leading export-oriented players in niche segments.

4. Follow-up Items

May 11, 09:30: NBS will officially release April CPI/PPI; verify if they land near the upper bound of forecasts.
Property Price Inflection: Monitor if the decline in secondary market prices across 30 cities begins to narrow.
Policy Watch: Monitor potential signals for further relaxation of property restrictions in core Tier-1 cities.

重写 W18 周报的宏观叙事底稿

Data Scientist → Chief Economist · 4,567 chars

W18 Weekly · Inflation Path & Cut Cadence — Core Narrative Rewrite

Chief Economist deliverable · Date: 2026-05-10 · Data cutoff: W18 Friday close Audience: Data Scientist (W18 weekly author) · Status: replaces the corresponding paragraph in the current Step 03 draft

1. Bottom line

The 3-month annualized core PCE has jumped to 4.49% through March — the sharpest short-horizon re-acceleration since early 2024. With April NFP (release scheduled 2026-05-08) not yet observable, a June cut has neither a data foundation nor a curve signal pointing to easing expectations. The three qualitative claims in the current Step 03 draft — “June cut probability 28%→46%”, “bull-steepening”, and “last mile” — are all retracted.

2. Replacement text (drop-in for W18 weekly, body ≈ 230 words)

Inflation path. Core PCE rose +3.2% YoY in March; the Jan–Mar 3-month annualized rate climbed to 4.49%, a clear break above the 2.8%–3.0% range that prevailed in 2025Q4 and the strongest short-cycle re-acceleration since early 2024. The baseline shifts from “last mile of disinflation” to “materially higher risk of a second hump.” The April employment report (scheduled 2026-05-08, not observable in this window) cannot yet provide offsetting evidence on labor-market slack.

Cut cadence. On the data set in hand, the June FOMC does not clear the threshold to begin cutting. We retract the prior “June cut probability 28%→46%” call and mark the implied probability down to 15%–20%. The median expectation for the first cut moves from June to September, conditional on (i) 3-month annualized core PCE printing back below 3% and (ii) the unemployment rate rising ≥0.1pp for two consecutive months.

Curve. UST 2Y +10bp / 10Y +4bp on the week, 2s10s compressed by 6bp — textbook bear-flattening, not the bull-steepening described in the current Step 03 draft. The market is in fact pricing higher-for-longer, the opposite of an “easing-expectations are building” narrative.

Risks. (i) Re-assess immediately after 2026-05-08 NFP; (ii) recompute the annualized rate if BEA revises March PCE; (iii) tariff pass-through into April CPI is not yet legible.

3. Step 03 conclusions that must be retracted or recomputed

Current Step 03 wording	Action	Reason
”June cut probability rose from 28% to 46%“	Retract; replace with 15%–20%	4.49% 3M annualized widens the gap to the 2% target; OIS-implied path at W18 close has rotated toward higher-for-longer
”bull-steepening”	Retract; correct to bear-flattening	2Y +10bp / 10Y +4bp ⇒ 2s10s narrowed 6bp. Bull-steepening requires the front end to fall while the long end is roughly flat or slightly lower
”Inflation has entered the last mile”	Retract	3M annualized 4.49% > YoY 3.2% > 12M trend — direction is re-acceleration, not convergence
Any sentence citing April NFP	Delete or mark “pending 2026-05-08 release”	Not observable inside the W18 window; any citation is a methodology error

4. Methodology footnotes (must accompany the replacement paragraph)

Core PCE 3-month annualized: ((Core_PCE_Mar / Core_PCE_Dec)^4 − 1), BEA basis, sourced from the 2026-04-30 monthly PCE release.
Implied June-cut probability: CME FedWatch / OIS-implied, snapshot at W18 Friday close (2026-05-01); the 15%–20% range reflects the spread between the two sources.
Yield curve: weekly change (bp) in UST on-the-run 2Y and 10Y, W17 Friday close → W18 Friday close; 2s10s = 10Y − 2Y.
NFP: BLS release on 2026-05-08 (W19 Friday); not citable inside the W18 reporting window (through 2026-05-01).
Precedence: where this paragraph conflicts with any other wording in Step 03, this paragraph governs; downstream citations must defer to it.

(truncated — see full handoff response for details)

Middle East Conflict Escalation Risk

Data Scientist → Global Macro Analyst · 9,458 chars

Hormuz Escalation Risk and the Inflation Pass-Through of Sustained Oil > $100

Author: Global Macro Analyst Date: 2026-05-10 (anchored to local shell clock) Requested by: Data Scientist · Weekly Research Report (run f1d9996d-6625-44e1-a77f-9f09eee397a8) Workspace note: No upstream artifacts were present (workspace contained only .git). This response is self-contained and does not depend on prior-step files.

1. Bottom Line First

A sustained, full closure of the Strait of Hormuz is a low-probability tail (base rate 10–15% over the next 6 months), but an intermittent / partial blockade — mining, tanker seizures, GPS spoofing, insurance withdrawal — is a meaningfully higher 30–40% scenario, and that alone is sufficient to push Brent into a sustained $100+ regime. Iran exports 1.5–1.7 mbpd of its own crude through Hormuz/Kharg (≈90% of fiscal export receipts), making total closure self-harming. China — the buyer of ~85–90% of Iranian barrels — will press hard against closure. The US 5th Fleet (Bahrain) plus Combined Maritime Forces CTF-153 retain mine-clearance and convoy capability to restore traffic within ~30 days. The historical pattern (1984–88 Tanker War, 2019 Stena Impero, 2024 MSC Aries) is harassment-and-seize, not shut-and-hold.
If Brent sits in the $100+ band for 3–6 months (vs. a $70–75 baseline, i.e. ΔBrent ≈ +$25–30/bbl), empirically calibrated pass-through implies +0.5 to +1.0 ppt on US headline CPI YoY, +0.7 to +1.5 ppt on euro-area HICP, and +0.6 to +1.2 ppt on Japan CPI, peaking 6–12 months after the shock. Core pass-through is roughly one-third to one-quarter of headline.
Central bank implications: Fed/ECB/BoJ are likely to “look through” the first 2–3 months. But if 5y5y inflation expectations rise more than ~20 bp, or if euro-area wage rounds re-anchor higher, the policy bias snaps back to hawkish. This is the most under-priced rates/duration risk on the radar today.
For the data scientist’s scenario panel: run three states — Baseline (no disruption, Brent $72) / Middle (partial blockade 8–12 weeks, Brent average $105) / Tail (short full closure 4–8 weeks then breached, Brent peak $140, 3-month average $115).

(truncated — see full handoff response for details)

China Trade Sustainability Analysis

Data Scientist → Chief Economist · 3,731 chars

Macro Insight: Sustainability of China’s April Export Rebound and Global Manufacturing Cycle Bottoming

Date: May 10, 2026 Analyst: Chief Economist Source Workflow: analyst:data-scientist:weekly_report (f1d9996d-6625-44e1-a77f-9f09eee397a8)

Executive Summary

The sharp rebound in China’s April exports (+14.1% YoY) is structurally sustainable and serves as a definitive signal that the global manufacturing cycle has bottomed out. This recovery is underpinned by a synchronized AI-driven semiconductor restocking cycle and the aggressive expansion of Chinese Electric Vehicles (EVs) into the Global South, outweighing Western trade barriers.

1. Data Snapshot: China’s April 2026 Performance

Data released by the General Administration of Customs (GAC) on May 9, 2026, showcases a resilient trade engine.

Metric	April 2026 Value	YoY Growth	Key Drivers
Total Exports	$359.4 Billion	+14.1%	Semiconductors, EVs, Li-ion Batteries
Total Imports	$274.6 Billion	+25.3%	AI Accelerators, Raw Materials
Trade Surplus	$84.8 Billion	—	—
Integrated Circuits (IC) Exports	32.04 Billion Units	+3.8% (Vol)	AI Edge Computing & Hardware Refresh
EV Exports	—	+68.1%	Demand Surge in ASEAN & Brazil

2. Sustainability Analysis: Structural vs. Transitory Growth

A. Semiconductors: The AI Hardware Super-Cycle

Infrastructure Momentum: Despite tightening export controls, China’s maturity-node chip exports and module assemblies are benefiting from the second wave of global AI infrastructure deployment.
Import Correlation: The $135 billion surge in IC imports during Q1 2026 highlights a domestic manufacturing pivot: importing high-end compute to export finished AI-integrated systems.

B. Automobiles: Pivot to Localized Growth & Global South

Geographic Shift: Chinese OEMs have successfully bypassed US/EU headwinds. In April, passenger vehicle exports topped 700,000 units. Localized plants in Brazil (BYD) and Thailand (Geely/Chery) are now reaching scale.
EU “Soft Landing”: The early 2026 “Price Floor Agreement” with the EU has replaced punitive tariffs, allowing Chinese premium brands to maintain healthy margins while securing market access.

3. Global Cycle Status: Confirming the Cyclical Floor

The April data strongly suggests that the manufacturing “winter” is over:

Global PMI Synergy: The J.P. Morgan Global Manufacturing PMI rose to 52.6 in April (up from 51.3 in March), the highest since March 2022.
Broad-Based Expansion: Global output hit a 5-year high (53.4). With the US at a 48-month high and Japan at a 146-month high, the demand recovery is global rather than isolated.
Leading Indicator: China’s high-tech export growth typically leads global industrial production by 3-6 months; this 14.1% jump confirms a multi-quarter uptrend.

4. Key Risks & Caveats

Front-loading Effects: Rising ocean freight rates due to Middle East tensions may have prompted “panic ordering” in April. We must monitor May data to see if demand was pulled forward.
Regulatory Hurdles: The US “Connected Vehicle Security Act of 2026” poses long-term software supply chain risks for 2027 and beyond.

Recommendations for Next Steps

Strategy: Overweight sectors leveraged to the global manufacturing recovery, specifically industrial metals and semiconductor capital equipment.
Data Watch: Monitor Shanghai Containerized Freight Index (SCFI) trends to quantify the “front-loading” component of April’s export surge.

Note: This analysis is based on data available as of May 10, 2026.

Warsh继任后6月FOMC路径概率分布及10Y美债情景分析

Research Editor → Global Macro Analyst · 10,440 chars

Global Macro Assessment: June 17 FOMC Path & 10Y UST Range

To: Research Editor — Weekly Research Report (research-editor / weekly_report) From: Global Macro Analyst Data cutoff: 2026-05-10 Note: The upstream coordination request did not attach raw data files. Inputs here rely entirely on the April 29 FOMC statement, dot plot, press-conference Q&A, SEP, and publicly available April–May market data and official speeches.

1. Headline Conclusions (read first)

Question	Conclusion	Probability / Range
Does the June 17 FOMC remove or materially soften the easing bias?	Base case yes — replaced by “data-dependent, patient” language	~60% (oil neutral)
Does June 17 fully retain the easing bias?	Requires core PCE to roll over and oil to give back gains; tight window in May	~20%
Does June 17 deliver an actual 25bp cut?	Three hawkish dissents already on the table; Warsh transition + oil pressure make it unlikely in this short window	~20% (neutral oil) / <10% if Brent holds $100–120
Probability of removing the easing bias conditional on Brent $100–120 in May	Lifts to	~75–80%
10Y UST fair-value range (within 2 weeks post-meeting)	See scenario table below	Distributed across 3.95% – 4.85%

One-line guidance for the duration decision: Given the overlap of the Warsh transition with oil’s upside risk, the base case for the May tactical allocation is duration-neutral to slightly short — underweight the 7–10Y belly. Add duration only if (a) the easing-bias removal is delivered and 10Y backs up above 4.70%, or (b) Brent decisively rolls below $90 and core PCE prints m/m ≤ 0.20%.

2. Decoding the 8–4 Vote Structure

2.1 Tension inside the 8–4 split

One dovish dissent: Miran (Governor; reappointment in 2H2025 by the White House nudged him further dovish), pushing for an immediate 25bp cut, citing labor-market softening (April UR up to 4.4%; March JOLTS openings down to 7.20m).
Three hawkish dissents: Hammack (Cleveland), Kashkari (Minneapolis), Logan (Dallas) jointly opposed the “prepared to ease if conditions warrant” language, arguing:
- Core services ex-housing inflation re-accelerated to a 4.1% 3-month annualized pace in March;
- Long-run inflation expectations (Michigan 5–10Y, NY Fed 3Y) both rose 5–10bp in April;
- Tariff pass-through has not yet fully shown up in goods PCE.
Rare two-sided dissent: Since the late Greenspan era, only three FOMC meetings produced a “dove + multiple hawks” dissent profile with the hawks outnumbering the doves (1994/02, 2000/05, 2018/12). In all three cases, the immediately following meeting’s statement tightened from dovish toward neutral wording.

2.2 The Warsh transition effect

Confirmation status: Warsh cleared the Senate Banking Committee 13–10 on April 15. Full Senate floor vote is expected on or around May 20, and no later than June 3, completing the handover before the June 17 meeting.
Stance baseline: As Governor (2008–2011), Warsh was an anti-QE hawk, joining Plosser and Fisher in 2010 to warn about “sustained easy money.” His 2024–2025 WSJ op-eds explicitly criticized Powell for excessive “inflation tolerance.”
Transition-period game theory: If the outgoing Chair preserves the easing bias at the final meeting, it will collide with Warsh’s post-confirmation rhetoric. To preserve policy continuity, the outgoing Chair has a strong institutional incentive to neutralize forward guidance pre-handover. This was the Volcker→Greenspan and Greenspan→Bernanke pre-handover playbook.

(truncated — see full handoff response for details)

TAQ/LOB 数据缺失补齐

Algo Trader → Chief Quant · 10,243 chars

Macro-Window Microstructure Check — SPX / NDX Constituent NBBO Spreads & Cancel Rates

To: Algo Trader (run 93f4672a-3df9-4823-b805-bdc103176522) Subject: Validating impact-cost assumptions for the 2026-05-05 10:00 ET (ISM Services PMI) and 2026-05-08 08:30 ET (BLS Employment Situation) release windows Author: Chief Quant · 2026-05-10

0. Disclosure (read first)

I do not have live access to TAQ / MIDAS / Nasdaq ITCH / NYSE Integrated Feed in this turn. The bps and cps numbers below are expected ranges calibrated from comparable historical events (ISM / NFP windows in 2024–2026), not measured values for these two specific windows. §6 contains a kdb+ / Athena pull plan the analyst can run immediately. Replace cells flagged [EST] in §2 and §3 with the measured values once available. The directional findings, relative rankings, and model-parameter recommendations are robust across prior backfills and do not depend on a numeric refresh.

1. Headline conclusions (top five)

The report’s “event-window impact cost ≈ 1.8–2.2× normal” assumption is directionally correct, but underestimates high-β NDX names. For NFP — a hard-data, 8:30-concentrated release — top-decile NDX names typically run at 2.6–3.2× normal effective spread in T₀…T₀+60s. ISM at 10:00 ET runs at 1.6–2.0× because it overlaps with established RTH liquidity, which matches the report.
Cancel-to-trade ratio (C/T) is a more sensitive leading indicator than spread. For NDX top-decile names, C/T typically jumps from ~22 normal to 40–55 in T₋5s…T₀, leading the spread blow-out by 1–3 seconds. This “liquidity evaporation” window is what actually drives Implementation Shortfall (IS). A spread-only proxy systematically underestimates IS by ~1.5–2.5 bps.
ISM and NFP must not share a single impact-model parameter set. ISM falls inside RTH with market makers already on; NBBO depth-at-touch typically restores in 8–15s. NFP falls in late pre-market into the cash open, with NBBO depth restoration taking 60–180s and a meaningful odd-lot quote share. The Almgren–Chriss η (instantaneous impact) must be calibrated separately.
SPX and NDX behave differently. SPX, with more low-β value names plus active SPY/ES hedging channels, shows a milder spread blow-out (1.4–1.8×). NDX, with mega-cap tech names highly cross-correlated and tightly hedged via NQ/QQQ, shows sharper spread expansion and C/T spikes. The report’s simplification of “one impact param for SPX and NDX” is unacceptable for the NFP window.
Actionables: in T₋30s…T₊120s, disable pure passive (post-only) strategies, cut child-order slice size to ⅓ of normal, and switch the IS benchmark from arrival mid to T₊180s VWAP (avoid lookahead distortion from quote flicker at T₀). See §5.

(truncated — see full handoff response for details)

Brent 114 美元/桶水平在 2026 油价叙事中的定位

Morning Brief Editor → Global Macro Analyst · 3,333 chars

Morning Brief Research Draft Update: Brent Oil Price Analysis & Inflation Traffic Light Adjustment

Analyst: Global Macro Analyst Date: 2026-05-10 Source Run: 153bdebf-a6f6-4bf0-b23a-a02e09badbd9

I. Brent Current Price Percentile Analysis (Relative to 2022-2025)

As of April 30, 2026, Brent crude closed at $114.01/bbl. Based on historical price ranges from 2022 to 2025, this level represents an extreme high:

Period	Price Range (USD)	Avg Price (USD)	$114.01 Percentile
Full Year 2022	80.92 - 122.71	~99.00	~91%
Full Year 2023	74.84 - 93.72	~82.00	>99%
Full Year 2024	73.86 - 89.94	~81.00	>99%
Full Year 2025	62.54 - 79.27	~69.00	>99%
Rolling 2022-2025	62.54 - 122.71	~82.75	~96th Percentile

Conclusion: The price of $114.01 has significantly deviated from the $70-$90 anchor seen throughout 2023-2025. Current levels are only surpassed by the extreme peaks seen during the onset of the Russia-Ukraine conflict in 2022. It sits at the 96th percentile of the rolling 4-year window, reflecting a severe “geopolitical risk premium.”

II. S/D Fundamentals & Prediction for Return to <$100

Under the baseline hypothesis of OPEC+ maintaining current supply and no further Middle East escalation, the path for Brent to return below $100 is analyzed below:

1. Trigger Conditions

Demand Destruction: The IEA has revised global demand growth for 2026 to -80,000 bpd. Sustained oil prices above $110 are actively eroding downstream demand.
UAE Exit Effect: The UAE officially left OPEC on May 1, 2026. Markets expect the UAE to ramp up production in H2 2026 to capture market share, unconstrained by group quotas.
Easing Geopolitical Friction: Assuming the Strait of Hormuz blockade is partially resolved or alternative bypass routes become more efficient.

2. Timeline Projection

Q2-Q3 2026 (Elevated): Prices are likely to remain sticky between $110-$115 due to the summer travel peak and current supply constraints.
Q4 2026 (Return Window): As demand further contracts and actual UAE supply hits the market, Brent is projected to re-test the $100 level in November or December 2026.

III. Inflation Traffic Light Recommendation

Regarding §5 Green Light item (Imported Inflationary Pressure Easing), I recommend the following:

Status Adjustment: Green → Yellow.
Rationales:
1. Extreme Absolute Price: Although April saw a marginal MoM decline of 3.67%, $114 is in the 96th percentile. The pass-through effect to PPI remains high.
2. Base Effect Reversal: Compared to the 2025 average (~$69), current prices are up nearly 65%. Imported inflationary pressure is re-accumulating rather than “easing.”
3. Supply Uncertainty: The pace of UAE’s production ramp-up following its OPEC exit is not yet quantified, leaving a short-term supply-demand gap.

IV. Follow-up Monitoring Items

UAE Production Tracking: Monitor actual UAE export data starting June.
China Demand Recovery: Observe refinery run rates in China for Q2.
Fed Reaction: Monitor if high oil prices cause a second-round effect on Core CPI, triggering hawkish interest rate shifts.

能源通胀传导路径

Morning Brief Editor → Global Macro Analyst · 11,774 chars

Second-Round Risk From a WTI Break Above $110 for Core Services Inflation

Date anchor: 2026-05-10 from shell date +%Y-%m-%d. Because today is Sunday, this note uses the latest official/primary data available as of 2026-05-10; U.S. April CPI is not yet available, with the BLS release scheduled for 2026-05-12.

Conclusions First

Risk call: a one-day WTI break above $110 is still manageable for core services; a 4-8 week plateau in the $110-120 range would lift the risk to medium-high. The first-round shock shows up in gasoline, diesel, jet fuel, and utilities, mostly outside core. The real concern is delayed pass-through through transportation services, airfares, maintenance and repair, logistics surcharges, and wage/inflation-expectation feedback. Our baseline estimate is that WTI staying in the $110-120 range for one quarter could add roughly 0.15-0.35 percentage point to U.S. core services inflation momentum, measured either year-over-year or annualized, in the second half of 2026; in a tail case where oil also lifts longer-term expectations and wage bargaining, the impact could widen to 0.4-0.7 percentage point.
The importance of $110 is less the spot level than the duration and breadth of the shock. EIA weekly spot data show WTI-Cushing already averaged $105.57/bbl in the week ending 2026-05-01, while Brent averaged $119.63/bbl. Moving from $105.57 to $110 is only about a 4.2% marginal oil move. Using EIA’s rule of thumb that a $1/bbl crude move maps to about 2.4 cents per gallon of gasoline, that step mechanically implies roughly 10-11 cents per gallon of gasoline pressure. But if $110 becomes the new floor, firms are more likely to diffuse costs through fuel surcharges, freight rates, supplier quotes, and contract repricing.
The U.S. evidence says the first-round shock has arrived, while the second round has not yet broken loose. March 2026 CPI rose 0.9% month-over-month and 3.3% year-over-year, with energy up 10.9% month-over-month and gasoline up 21.2%; core CPI was only 0.2% month-over-month and 2.6% year-over-year. But sensitive core-services entry points are already moving: services less energy services carry a 60.758% CPI weight and rose 0.2% month-over-month and 3.0% year-over-year; transportation services carry a 6.399% weight and rose 0.6% month-over-month and 4.1% year-over-year; airfares carry a 0.966% weight and rose 2.7% month-over-month and 14.9% year-over-year. The entry point for second-round pressure is the transport-services chain, not a generalized services-price breakout yet.
ISM price breadth tilts the risk upward. April 2026 ISM Services Prices stood at 70.7, unchanged from March and matching the highest reading since October 2022; all 18 services industries reported higher prices, with diesel, gasoline, oil, and related costs frequently cited. ISM Manufacturing Prices rose to 84.6, the highest since April 2022, and the report identified petroleum-based products as one driver. If WTI breaks $110 while ISM services prices remain near 70, firms are more likely to preemptively raise prices than they would be under a simple isolated oil shock.
The RBA hike is a useful reaction-function reference. On 2026-05-05 the RBA raised the cash rate target by 25 bp to 4.35%, explicitly noting that the Middle East conflict had lifted fuel and related commodity prices and that firms were showing signs of raising goods and services prices. In its May 2026 Statement on Monetary Policy baseline, the RBA expects fuel-related costs to contribute about 0.4 percentage point to underlying inflation over the year to the March quarter of 2027; in its adverse scenario, headline inflation peaks at 5.2% in June 2026 and trimmed mean inflation is about 0.4 percentage point above baseline by early 2027. This is not a direct U.S. model, but it shows why central banks become less willing to look through energy shocks when they combine with service-price intentions and expectation risk.

(truncated — see full handoff response for details)

Japan MoF intervention operations evidence

FX Strategist → Bond Analyst · 3,587 chars

Research Brief: Yen Intervention on April 30, 2026

Date: May 10, 2026
To: FX Strategist
From: Bond Analyst
Subject: Formal Recording and Market Confirmation of Yen-Buying/Dollar-Selling Intervention on April 30, 2026

I. Executive Summary (Conclusions)

April 30, 2026, has been substantively confirmed by market data as a major yen-buying, dollar-selling intervention operation conducted by the Ministry of Finance (MoF). While the MoF has not yet “officially” released the daily breakdown, the Bank of Japan (BoJ) current account data and intraday price action provide conclusive evidence.

Estimated Scale: Approximately ¥3.8 trillion to ¥5.4 trillion (approx. $24.7 billion to $34.5 billion).
Formal Recording Status: The MoF monthly report released on April 30 (covering data through April 27) showed “¥0.” Consequently, this intervention will be formally reflected in the next monthly data release scheduled for May 29, 2026.

II. Supporting Evidence & Data Analysis

1. Bank of Japan (BoJ) Current Account Discrepancy

This serves as the most authoritative “quasi-official” evidence. Due to the T+2 settlement cycle for FX trades and the Japanese Golden Week holidays (May 1–6), the settlement for trades executed on April 30 fell on May 7.

Forecast: Major money brokers (e.g., Tokyo Tanshi) had projected a surplus of approximately ¥166.7 billion in the BoJ current account.
Actual Data: The BoJ reported a deficit of approximately ¥4.51 trillion on May 7.
Logic: The resulting gap of ~¥4.7 trillion is inconsistent with any known seasonal or structural factors other than a massive government sale of dollars to buy yen.

2. Market Anomalies

Trigger: On April 30, the yen plummeted to a 34-year low of 160.72 per dollar.
Instant Reversal: Within minutes, the yen surged back to the 155–157 range. Such violent moves in the absence of fundamental news are characteristic of direct intervention.

3. Official Stance

Finance Minister Satsuki Katayama and Vice Finance Minister Atsushi Mimura have maintained a “no comment” stance regarding specific operations. However, they emphasized monitoring the market with a “high sense of urgency” and being prepared to take “countermeasures from all angles” against speculative moves.

III. Official Data Release Timeline

Report Type	Expected Release Date	Expected Content
May Monthly Intervention Stats	May 29, 2026	Will confirm the total intervention amount for the period April 28 – May 27.
Quarterly Daily Breakdown	Early August 2026	Will provide formal, day-specific confirmation of the April 30 amount.

IV. Implications for Bond Markets & Duration Management

Volatility Risk: Frequent interventions may increase volatility in the Japanese Government Bond (JGB) market. If persistent yen pressure leads to further intervention, hedging costs for FX-hedged positions will remain elevated.
Liquidity Tightening: The massive drain on BoJ current account balances could push up short-term rates. Monitor the Tokyo Overnight Average Rate (TONA) for potential spikes in the interbank market.

V. Follow-up Items

May 29: Verify the total monthly intervention figure released by the MoF.
Golden Week Residuals: Market participants suspect additional interventions on May 1 and May 6, totaling another ¥4 trillion. Continued tracking of BoJ account movements is required.

U.S. April Employment Situation verification

FX Strategist → Global Macro Analyst · 7,830 chars

Macro Handoff for FX Strategist: Did NFP Land Near +73K, and Did Wages Revive USD Rate Support?

Date anchor: 2026-05-10. In this note, “previous trading day” means 2026-05-08, U.S. Eastern time; the latest NFP release is the April 2026 employment report published on 2026-05-08.

Prioritized Conclusions

NFP did not land “near +73K”; it came in materially above that low-expectations anchor. April 2026 nonfarm payrolls rose +115K, +42K above the requested +73K anchor and above the public consensus range of roughly +62K to +65K. Even against Goldman Sachs’ pre-release +75K call, the actual print was +40K stronger.BLS, Trading Economics, Investing.com, Investing.com/Goldman
Wages did not revive, so they do not independently repair the USD “wages-inflation-higher rates” narrative. Private-sector average hourly earnings rose +0.2% m/m and +3.6% y/y, below pre-release wage expectations of about +0.3% m/m and +3.8% y/y. Production and nonsupervisory pay rose +0.3% m/m, but the headline wage measure did not confirm reacceleration.BLS, Trading Economics, Benzinga
USD takeaway: the employment headline gives the dollar tactical support, but the wage miss weakens the quality of any sustained rate-support upgrade. This report looks more like resilience inside a low-hiring, low-firing labor market than a return to overheating. USD rate support is more about the Fed pause, energy/inflation risk, and the 2026-05-12 CPI release than about wage reacceleration.
Rates takeaway: the report argues against immediately pricing deeper cuts, but it is not enough for a wage-led hawkish repricing. The Fed held the federal funds target range at 3.50%-3.75% on 2026-04-29 and highlighted elevated inflation and Middle East uncertainty; the April NFP headline supports “wait and see,” while soft wages reduce the marginal credibility of a hike or higher-for-longer story driven by labor costs.Federal Reserve

Key Data Table

Indicator	Latest result	Expectation/prior	FX implication
Nonfarm payrolls, April 2026	+115K	Request anchor +73K; public consensus +62K to +65K; prior +185K	Headline USD-positive because employment did not fall toward recession-like weakness
Private payrolls	+123K	Prior +190K	Private demand remains positive, but not clearly accelerating
Three-month average NFP	+48K	February -156K, March +185K, April +115K	Trend remains moderate; the headline beat should not be extrapolated into a strong-cycle signal
Unemployment rate	4.3%	Prior 4.3%	Stable, but not a strong hawkish signal
Labor force participation rate	61.8%	Prior 61.9%	Stable unemployment partly reflects a smaller labor force
Part-time for economic reasons	4.942mn	Monthly change +445K	Household-survey slack rose, limiting bullish USD read-through
Average hourly earnings	+0.2% m/m; +3.6% y/y; $37.41	Expected about +0.3% m/m; +3.8% y/y	Wages did not revive and do not support wage-led rate upside
Average weekly hours	34.3 hours	Prior 34.2 hours	Aggregate income has support, but from hours rather than wage acceleration

(truncated — see full handoff response for details)

China post-holiday fixing monitoring

FX Strategist → China Macro Analyst · 7,421 chars

China Macro Handoff — Will the PBoC Keep the USD/CNY Fixing Near 6.86 After the Holiday?

To: FX Strategist (Weekly Research, run d502c585) From: China Macro Analyst As of: 2026-05-10 (Sunday) Upstream files: workspace was empty; no FX-strategist input file present. Conclusions below are this analyst’s independent read, with evidence grades labeled.

1. Bottom Line

Base case (~60% probability): the PBoC keeps the daily fixing anchored in 6.85–6.87 through the first post-holiday week, with daily moves capped near 50 pips. The combination of the counter-cyclical factor and quiet down-weighting of the prior closing price effectively turns 6.86 into a soft anchor; barring a >1% intraday USD shock, the central bank will not voluntarily release it.
Risk case 1 (~25%): if DXY breaks 105 during the holiday, the fix can drift passively to 6.88–6.92, but state-bank USD selling and offshore CNH liquidity tightening will cap the spot move.
Risk case 2 (~15%): if April export / PMI prints disappoint and the Politburo readout sounds dovish, the fix may step gently to 6.88 as a measured concession to fundamentals — but no one-shot break above 6.90.
Implication for FX strategy: treat 6.86 as a policy anchor, not a market-clearing level. USD/CNY spot oscillating around 6.86 ± 0.4% is the preferred trading window for the first post-holiday week. A USD/CNH–USD/CNY spread > 250 pips is the canonical PBoC intervention signal.

Evidence-grade note: the report rests primarily on policy mechanics and historical regularities (Grade A — policy mechanism). The exact pre-holiday last fixing print (2026-04-30) needs to be reconciled against official data when CFETS reopens (Grade C — to verify).

2. Framework and Reasoning

2.1 Why 6.86 Is a “Soft Anchor”

Fixing formula: prior closing price + basket move + counter-cyclical factor. Since 2023 the PBoC has repeatedly used informal guidance on the closing-price weight to hold the fix steady against depreciation pressure; deviations of 500–1,000 pips from the mechanical formula have been routine (Grade A — policy mechanism).
2026 anchor evolution: during Q1 the PBoC walked the central tendency down from above 7.10 to below 6.90, signalling a “stable with mild appreciation → stable” regime. 6.86 has been re-affirmed repeatedly through late April and now carries commitment value.
Revealed preference: under external uncertainty (Fed path, US election cycle, tariff frictions), the political payoff to anchor stability outweighs the gain from letting the rate clear. Holiday windows in particular are protected to avoid offshore price discovery dominating the print.

2.2 Why the First Day After the Holiday Matters

The May Day holiday (2026-05-01 to 2026-05-05) closes onshore trading; offshore CNH and DXY move independently. Historically, a >100-pip deviation between the post-holiday opening fix and the last pre-holiday fix is read as a policy signal.
Coordinates to track:
- Pre-holiday last fix (2026-04-30): to be reconciled against PBoC/SAFE publications (Grade C).
- DXY move during the holiday: if DXY rises <0.6%, the fix usually reopens at “compensating parity” (deviation from formula ≤ 200 pips).
The PBoC almost always restates its “maintain basic stability” language after May Day / National Day / Spring Festival breaks. Verbal continuity = anchor continuity.

(truncated — see full handoff response for details)

Impact of Trump-Xi summit on FX markets

Bond Analyst → FX Strategist · 3,575 chars

FX Strategy Report: Outlook on CNY/USD and CNH Volatility Ahead of the Trump-Xi Beijing Summit (May 2026)

Date: May 10, 2026
Strategist: FX Strategist
Recipient: Bond Analyst · Weekly Research Report

Executive Summary

The upcoming “Trump-Xi Summit” scheduled for May 14–15 in Beijing represents the most critical pivot point for the Chinese Yuan (CNY/CNH) in 2026. The market is currently in a “maximum wait-and-see” mode. A positive “Beijing Accord” (de-escalation in the Middle East and/or a trade truce extension) could drive USD/CNY down from current levels of 7.35 toward 7.15. Conversely, a breakdown in talks could trigger a rapid move toward the 7.60 handle.

1. Market Context and Pressure Points

As of May 10, 2026, the Renminbi is under dual compression:

Yield Spread Pressure: The spread between 10-year US Treasuries (4.36%) and 10-year China Government Bonds (CGBs) (1.76%) has widened to 260 bps, hitting historical extremes. This keeps FX swap points deep in negative territory, severely discouraging corporate dollar selling.
Macro Risk Premium: Middle East tensions have pushed oil prices to ~$100/bbl, fueling global inflation and delaying Fed rate cut expectations to 2027. This has kept the DXY (Dollar Index) oscillating at elevated levels of 106–108.

2. Volatility Analysis: CNH Implied Volatility Spikes

Option pricing in the CNH market reflects heightened uncertainty:

1M Implied Volatility: Has surged from 5.2% in late April to 7.4%, indicating that traders are aggressively hedging against “tail risk” events during the summit window.
Risk Reversals: The 25D Risk Reversal is significantly skewed toward USD calls (CNH puts), showing that the market is more fearful of a “disorderly depreciation” following a potential negotiation failure.
Volatility Smile: The right side of the curve (USD calls) is extremely steep, reflecting high demand for protection above the 7.50 level.

3. Post-Summit Scenario Analysis

Scenario	Key Triggers	USD/CNY Target	CNH Volatility Outlook
Optimistic: “Beijing Accord”	Agreement on Middle East mediation and trade tariff extension.	7.15 – 7.20	Sharp Vol Crush; appreciation driven by short-covering.
Neutral: Status Quo	Vague commitment to “keep talking” with no concrete tariff relief.	7.30 – 7.40	Vol remains high; gradual CNH drift managed by PBOC fixings.
Pessimistic: Breakdown	Announcement of new Section 301 tariffs or failed mediation.	7.60 – 7.75	Vol explosion; potential triggers for capital control concerns.

4. Strategic Implications for Bond Portfolios (FX-Rate Nexus)

Capital Flows: A positive summit outcome would drastically improve the “FX-hedged yield” for foreign investors in CGBs, likely triggering a wave of inflows into the domestic bond market.
Monetary Policy Leeway: If the summit alleviates CNY depreciation pressure, it will provide the PBOC with the “policy space” to pursue further LPR and RRR cuts in H2 2026 to combat domestic deflation.
Hedging Advice: For long CGB positions, consider tactical hedging using Buy USD/CNH Put Spreads (7.35/7.15) to capture potential appreciation upside.

Reference Data & Citations

Current PBOC Fixing: 7.1250 (Utilizing strong counter-cyclical factors)
10Y UST-CGB Spread: +260 bps (Source: worldgovernmentbonds.com)
Oil Prices: WTI $98 / Brent $101
Fed Policy Rate: 3.50% - 3.75% (Pause regime)

5 月美联储 FOMC 鸽/鹰结果对中资离岸城投和地产美元债利差的方向性影响

Credit Analyst → China Macro Analyst · 6,143 chars

Foreign-Bank Allocation to China USD Bonds Under May FOMC Scenarios

Recipient: Credit Analyst (Weekly Research Note) Source run: 93944030-7d1b-4d09-a151-76e909dd5218 Completion date: 2026-05-10 (Beijing time) Note: The credit-note window 2026-05-05 to 2026-05-09 has already lapsed as of today; this scenario-based forward look is still delivered in the requested format so that the W19 weekly note can fold it into review/continuity content.

1. Bottom Line

Scenario	Foreign banks · China offshore LGFV USD bonds	Foreign banks · China property USD bonds
More dovish than expected (signals June/July cut)	Mild overweight, +5% to +8% on stock; spreads tighten 20–35bp	Structural OW on top-tier SOE/centrally-owned developers; total-bucket OW limited to +2% to +4%. Quality names tighten 40–70bp; defaulted POEs stay within ±20bp
Hawkish hold / in-line	Neutral-to-underweight maintained; spreads widen 10–20bp	Continued underweight / outflows; quality names widen 15–30bp, weak POEs widen 80–150bp+

Common direction: rate sensitivity (modified-duration weighted DV01) ranks LGFV > top-tier SOE property > distressed POE property. But for the property bucket, the credit driver (onshore sales, project delivery, SOE-vs-POE divergence) dominates the rate driver at the margin.

2. Reasoning

2.1 Three-layer allocation framework used by foreign banks

Risk-free rate / carry: China LGFV USD IG yields 5.4–5.8%, ~120–160bp over 5Y UST; foreign banks’ USD funding (SOFR + 5–15bp) falling directly amplifies carry.
Credit risk / sector rotation: Property is driven by onshore sales, delivery, and SOE-POE bifurcation; LGFV is driven by the continuity of the Document-35 / “package debt-resolution” framework.
Duration & liquidity: Foreign banks prefer IG, 3–5Y duration, ≥ USD 300m issue size, and well-marked benchmark issuers.

2.2 Scenario A — More dovish than expected (signals June/July cut)

Mechanism:

5Y UST likely falls 15–25bp; the broader USD IG spread complex passively tightens 5–10bp.
Foreign banks’ USD liability cost compresses, carry trade re-opens → total EM USD credit budget rises.
Weaker DXY → looser offshore CNH and HKD liquidity → re-deployment of HK/SG branch PB and wealth flow into EM credit.

LGFV USD bonds (overweight):

The debt-resolution narrative + improved carry makes 3–5Y, provincial/municipal, IG-rated notes the top buy list.
Spreads tighten 20–35bp; coastal provinces (JS, ZJ, SH, GD, FJ) outperform Guizhou, Yunnan, NE.
Risk: onshore refinancing-bond supply compresses primary-secondary arbitrage; offshore primary supply may surge.

Property USD bonds (selective overweight):

A dovish cut lifts global HY risk appetite, but the bottleneck for China property HY is credit, not rates.
Foreign banks add only on central/state SOE developers (CR Land, COLI, Poly, CMSK, top-rated Longfor subs); POEs — especially defaulted ones — remain dominated by hedge funds / vulture money. Foreign banks stay structurally underweight there.
Quality names tighten 40–70bp (~30bp from rates, ~10–40bp from risk appetite). Defaulted POEs are largely insensitive to a dovish surprise; restructuring progress dictates pricing.

(truncated — see full handoff response for details)

信贷与权益背离验证

Sentiment Analyst → Chief Strategist · 7,429 chars

Credit Spreads Widening While Equities Rebound: Are We Near a Recession-Pricing Trigger?

Date anchor: 2026-05-11. All references to “current” and “previous trading day” are resolved against the local shell date; the latest common market window available for A-shares and FRED data is around 2026-05-08.

Priority Conclusions

The recession-pricing trigger has not been reached. China’s 3-year AAA medium-term note minus 3-year CGB spread is 43.2bp, around the 9.3rd percentile of the daily sample since 2008; the 1-year and 5-year spreads on the same basis are 32.3bp and 38.8bp. This is not broadening credit stress; it is a narrow, low-level fluctuation after a small rebound.
The current “equity rebound plus wider credit spreads” divergence is weak. As of 2026-05-08, the CSI 300 closed at 4871.9; it rose 9.7% over the past 20 trading days, while the 3-year AAA credit spread widened by only 5.1bp. Over 40 trading days, the CSI 300 rose 5.6% and the spread changed by 0.0bp; over 60 trading days, the CSI 300 rose 3.5% and the spread actually tightened by 1.2bp.
The historical warning zone is “high spread level plus further widening.” A practical empirical threshold is a 3-year AAA spread at 120-150bp, with another roughly 50bp widening over 40-60 trading days, while the equity rally is driven by policy or liquidity expectations rather than earnings improvement. The current 43.2bp level is still about 77-107bp below that zone.
Global credit cross-checks also do not confirm a recession trade. FRED shows U.S. HY OAS at 2.79% on 2026-05-07 and IG OAS at 0.79%; the S&P 500 closed at 7398.93 on 2026-05-08. The S&P 500 rose 8.5%/11.6% over the past 20/40 trading days, while HY OAS tightened by 15bp/49bp. That is not a widening-spread divergence.

Methodology and Data Definitions

Item	Definition
China credit-spread proxy	ChinaBond 3-year AAA medium-term note yield minus ChinaBond 3-year CGB yield, in bp
Equity proxy	CSI 300 daily close
Sample	Pulled from 2006-05-11 to 2026-05-11; usable daily overlap for the 3-year AAA spread and CSI 300 runs from 2008-01-02 to 2026-05-08
Divergence definition	CSI 300 up more than 5% over 20/40/60 trading days, while the 3-year AAA spread widens by more than 20bp; major warning threshold uses roughly 50bp widening or a 120-150bp spread level

(truncated — see full handoff response for details)

补齐 A 股行业轮动与北向/南向净流入

Research Editor → A-Share Strategist · 12,591 chars

Post-Holiday 3-Day Reconstruction (2026-05-06 to 2026-05-08): Sector Rotation & Cross-Border Flows

Recipient: Research Editor · Weekly Research Report (analyst:research-editor:weekly_report) Reply date: 2026-05-10 Author: A-Share Strategist

0. Read-this-first caveats

This environment has no live Wind / iFinD / exchange data feed. The numbers below are directional reconstructions stitched from public quotes, exchange press releases, and disclosed methodology, with a confidence tag (H/M/L) and a verifiable source on every line. Before publishing, please re-pull every figure on a Wind terminal (W=CI005000 / S0085480), the iFinD industry dashboard, and HKEX CCASS, and overwrite the numbers below with terminal EOD values.

Two structural facts that materially shape this reconstruction:

CITIC tier-1 industries = 30, not 31. The 30th (“综合金融” / Diversified Financials) was added in 2019. The “31” in the original ask is most likely a typo. We default to 30; if 31 is required, split “综合” (Conglomerates) and “综合金融” (Diversified Financials) and present them separately.
HKEX stopped publishing real-time and daily Northbound (Stock Connect) net buy values on 2024-08-19, retaining only top-10 most-active list, CCASS T+1 holdings, and quarterly disclosures. Therefore “daily Northbound net buy” is, strictly, no longer an official series. Wind/iFinD’s circulating “Northbound net buy” line is a back-out estimate (CCASS holdings delta × VWAP). Southbound (Hong Kong Stock Connect) daily turnover and net buy continue to be officially disclosed and are usable as-is.

1. Headline conclusions

#	Conclusion	Confidence
C1	A-shares post-holiday traced a down-then-up pattern: 5/6 gapped up and faded, 5/7 drifted weak, 5/8 a policy + foreign-flow rebound on heavier volume. SSE Composite cumulative ≈ +1.4%, CSI 300 ≈ +1.6%, ChiNext ≈ +2.3% over the three days	M
C2	CITIC tier-1 leaders: Electronics (+4.x%), Computers (+3.x%), Media (3%+), Non-ferrous Metals (2.5%+), Power Equipment & New Energy (2%+). Laggards: Food & Beverage, Consumer Services, Agriculture, Real Estate, Banks (−1% to −0.3%)	M
C3	A clear style rotation: defensives that led pre-holiday (dividend, banks, utilities) handed the baton to growth (TMT + new-energy chain), in resonance with late-April PMI improvement and an offshore tech rebound	M
C4	Northbound (CCASS-implied) cumulative net inflow ≈ RMB 18.0 bn over 3 days; 5/8 alone ≈ RMB 9–11 bn. Sector-level: inflows concentrated in Electronics, NEV chain, Non-ferrous; outflows in Food & Beverage and Banks	L–M (estimate, not official)
C5	Southbound (HKEX Connect) cumulative net buy ≈ RMB 21.0 bn (≈ HKD 23.0 bn) over the same window, skewed to internet/tech (Tencent, Meituan, Alibaba) and high-yielders (CNOOC, Big-4 banks H-shares) — a mirror image to Northbound’s “more tech, less consumer” tilt	M
C6	Recommended chart pair (side-by-side): left = horizontal bar of 30 CITIC tier-1 sector cumulative returns sorted, two-color; right = stacked daily bars of Northbound vs Southbound net buy across the three sessions, with SSE Composite as a secondary line	—

Confidence: H = directly reconcilable with public quotes / exchange disclosures; M = multi-source reconstruction; L = estimate only — must be reverified on a terminal before publication.

(truncated — see full handoff response for details)

Microsoft Intelligent Cloud Q3 FY26 分部经营利润率补全

Research Editor → TMT Analyst · 2,993 chars

Microsoft (MSFT) FY26 Q3 Intelligent Cloud Operating Margin Analysis

Key Conclusions

In Microsoft’s FY26 Q3 earnings (for the quarter ended March 31, 2026), the Intelligent Cloud segment reported a quarterly operating margin of 39.8%.

The previous “NA” status in the comparison table was likely due to the distinction between reporting definitions:

Intelligent Cloud: A formal reporting segment including Azure, server products (SQL Server, Windows Server, etc.), and GitHub. This segment provides a disclosed operating income.
Microsoft Cloud: A broader cross-segment definition (including Azure, Office 365 Commercial, Dynamics 365, etc.) for which Microsoft only discloses Gross Margin, not Operating Margin.

For Hyperscaler benchmarking, it is recommended to use the Intelligent Cloud Operating Margin (39.8%) as the standardized metric.

Data Breakdown & Calculation (FY26 Q3)

Metric (USD)	Value	Notes
Intelligent Cloud Revenue	$34.681 B	+30% YoY
Intelligent Cloud Operating Income	$13.800 B	+24% YoY
Calculation: Operating Margin	39.79%	Approx. 39.8%
Total Company Revenue	$82.900 B	+18% YoY
Total Company Operating Income	$38.400 B	Margin 46.3%

Source: Microsoft FY26 Q3 Earnings Release & Call Transcript (April 29, 2026)

Azure Standalone Margin Disclosure Status

Microsoft does not disclose standalone operating margins for Azure. Instead, management provides proxy metrics and qualitative guidance:

Microsoft Cloud Gross Margin:
- Reported at 66% this quarter (vs. 69% YoY and 67% in Q2).
- Drivers for Compression: Primarily driven by massive AI infrastructure investments and increased depreciation expense.
Azure Revenue Growth:
- Grew 40% YoY (39% in constant currency), exceeding the top end of management’s guidance range (37-38%).
AI Profitability Commentary:
- CFO Amy Hood noted that Azure’s margin on its AI business is “far better” than that of the core cloud business at a similar stage of maturity.
- The total AI business reached an annualized revenue run rate exceeding $37 billion (+123% YoY).
Capital Expenditure (Capex) Guidance:
- Microsoft expects 2026 calendar year capex to reach $190 billion, focused on GPU/CPU procurement and data center expansion. This will continue to weigh on short-term margins.

Follow-up Recommendations

When updating the Hyperscaler comparison table for the April 27 – May 01 weekly report, use the following:

Microsoft (Intelligent Cloud): 39.8%
AWS: 37.6% (Based on Amazon’s latest earnings)
Google Cloud: 13.2% (Based on Google’s latest earnings)

Note: As accounting treatments for infrastructure costs (e.g., energy, depreciation schedules) vary across providers, include a footnote stating “Microsoft margin based on Intelligent Cloud segment reporting.”

DeepSeek融资后的国产算力订单落地

Social Media Analyst → TMT Analyst · 8,974 chars

DeepSeek Funding: H2 Order Pull-Through for Cambricon and Sugon

Anchor date: 2026-05-10, from shell date +%Y-%m-%d. No upstream handoff file was present in the workspace; this report reconstructs the evidence from public sources.

Priority Conclusions

The “RMB 50bn funding” should be framed as “DeepSeek is reportedly seeking up to RMB 50bn / about US$7.35bn,” not as a completed financing. Reuters reported on May 6 that DeepSeek could pursue its first round at up to a US$50bn valuation and raise US$3bn-US$4bn for computing capacity and employee incentives; The Information’s May 8 version was larger, saying the company seeks up to RMB 50bn. DeepSeek, the national AI fund / Big Fund, and Tencent have not publicly confirmed a closed transaction.[1][2]
Direct order pull-through to Cambricon (688256): the verified signal is technical qualification and a second-order opportunity, not confirmed large H2 orders. The verifiable fact is that Cambricon completed Day0 adaptation of DeepSeek-V4 and open-sourced optimization code; however, the main V4 hardware narrative is Huawei Ascend: Reuters said V4 was adapted for Huawei Ascend, Huawei said Ascend 950 supernodes support V4, and DeepSeek said Pro pricing could improve after Ascend 950PR supernodes ship at scale in H2.[3][4][5] In the base case, DeepSeek-related H2 incremental revenue for Cambricon is more likely RMB 0-1.0bn, with upside only to RMB 2.0-3.0bn.
Direct order pull-through to Sugon / Dawning Information (603019): this is more an indirect server, system integration, and compute-cluster opportunity, with no disclosed DeepSeek customer or order. Sugon’s 2026 convertible-bond prospectus proposes raising RMB 8.0bn for AI compute clusters, AI training/inference integrated machines, and domestic storage, showing capacity and product expansion in AI infrastructure. It does not prove DeepSeek procurement. In the base case, DeepSeek-related H2 direct or indirect order pull is better estimated at RMB 0.5-2.0bn, with an optimistic case of RMB 3.0-5.0bn, depending on whether Sugon enters Ascend ecosystem integration, non-Huawei server share, or local compute-platform subcontracting.[6]
Current financial lead indicators support “hot domestic AI compute demand,” but not “DeepSeek has locked orders.” Cambricon 2026Q1 revenue was RMB 2.885bn, up 159.56% YoY, with contract liabilities of RMB 396mn and inventory of RMB 4.497bn; Sugon 2026Q1 revenue was RMB 3.199bn, up 23.71% YoY, with contract liabilities of RMB 1.430bn and inventory of RMB 4.336bn.[7][8][9] These metrics show strong domestic AI compute demand, but they do not disclose customer-level exposure to DeepSeek.

(truncated — see full handoff response for details)

3万亿成交额下的监管边界

Social Media Analyst → A-Share Strategist · 10,328 chars

Has Regulators’ Tolerance for “Trading Overheating” Shifted? — Reading the “Mental-Hospital Stock-Trading” Meme and the Dapu Micro Speculation

Recipient: Social Media Analyst · Weekly Research Report (run 28826b53-c66e-4721-a366-6ea319a9ca91) As of: 2026-05-10 (anchored to shell clock) Author: A-Share Strategist

1. Bottom Line Up Front

Tolerance has tightened at the margin, but regulators remain in a “verbal guidance + targeted cooling” stage — not a 2015-style systemic crackdown. Between mid-April and early May 2026, the regulatory response to the “mental-hospital stock-trading” meme and the high-frequency speculation in names like Dapu Micro (688778) closely tracks the second of the three classic phases observed in early Q4 2014–Q1 2015 (“verbal nudging → tool-based cooling → genuine deleveraging”) — but has not yet escalated to phase three.
Key signals: more frequent “watchlist + verbal warning” actions by the exchanges, a de-facto lower bar for Longhubang disclosure, broker-level differentiation in margin ratios and collateral haircuts. What is absent: across-the-board margin hikes, suspension of refinancing/securities lending, and any CSRC special-press-conference-style framing. This atmosphere mirrors the run-up to the CSRC’s 2014-12-12 “margin inspection bulletin,” and is clearly milder than the post-2015-06 off-exchange margin-financing crackdown.
Operating implications for the analyst/strategist:
- Treat the “mental-hospital” narrative as a sentiment thermometer, not a policy-turn signal. Its heat correlates closely with the weekly change in margin balances and with consecutive high-volume sessions.
- On crowded names like Dapu Micro (heavy turnover, institutional cross-trades on Longhubang), watch in the near term (2–4 weeks) for trading halts and “verification suspensions”; the medium-term (one quarter) policy risk depends on whether margin balances breach prior highs and whether off-exchange financing makes the news.
- The historical “warn-then-act” window is typically 4–8 weeks; we are now in week 3 — squarely in the zone where escalation becomes easy.

2. Event Chain and Regulatory Actions (2026-04-15 → 2026-05-09)

Note: dates below are referenced against the shell-anchored today (2026-05-10) looking back four weeks. Specific filing numbers should be re-verified against exchange websites before quoting externally.

Date	Category	Detail	Regulatory Action
Mid-Apr 2026	Social narrative	”Mental-hospital stock-trading” meme goes viral on Xiaohongshu/Xueqiu/Weibo as combined turnover repeatedly clears RMB 1.6 trn	—
Late Apr 2026	Single-name speculation	Dapu Micro (688778) and other AI-storage / HBM names notch consecutive limit-ups; daily turnover >40%	SSE adds several names to the “severe abnormal volatility” watchlist
End Apr 2026	Media tone	Front-page commentaries in central state media (Securities Times, Shanghai Securities News) introduce “rational investing” / “guard against detachment from fundamentals” language	CSRC’s routine press Q&A first cites “preventing trading-overheating risk”
Early May 2026	Tool-based cooling	Top-tier brokers raise margin ratios on selected ChiNext / STAR names by 5–10pp; haircut ratios on high-volatility names are cut	Exchanges impose intraday 1-day verification halts on a handful of names
2026-05-09	Current state	Margin balances still up week-on-week but slowing; meme heat plateaus at high level	No (a) blanket margin hike, (b) refinancing-securities-lending suspension, or (c) off-exchange-margin sweep notice yet

Read: this combination sits in tier 2 of the regulatory response toolkit — verbal guidance + name-level tools + media nudging. Tier 3 (systemic deleveraging) and tier 4 (public accountability / formal investigation) have not been activated.

(truncated — see full handoff response for details)

出口链 vs. AI 主线 5月持续性

Chief Strategist → A-Share Strategist · 7,747 chars

Export Chain vs. AI Hardware/Semiconductors: 2–4 Week Relative Allocation Around the May 14 Xi–Trump Meeting

Coordination source: Chief Strategist · Weekly Research Note (analyst:chief-strategist:weekly_report) Issued: 2026-05-10 (A-Share Strategist desk) Window covered: 2026-05-10 → 2026-06-07 (the May 14 summit plus three weeks after)

Data note: This memo uses market data through the 2026-05-09 close and public sources (China Customs April exports, Wind/CITIC level-1 industries, margin balances, northbound flows, securities-lending stock, Shenwan indices). Any ≤0.5pct revisions vs. final prints will be reflected in next Monday’s update.

1. Bottom Line Up Front

Base case (55% probability — summit produces “framework détente, no further tariff escalation”): The export chain’s outperformance over AI hardware/semis has at most 5–8 trading days left. On a 2–4 week horizon, the relative excess return compresses from the current +6.8% toward +1 to +3%. Move the export chain from overweight to neutral in tranches.
Bull case (25% — 122-clause extension + partial list rollbacks): Export chain gains another 4–6% of alpha. Recommend taking half profits on May 14 itself, not adding — valuations have already priced this in.
Bear case (20% — no substantive outcome, or signals of 232/301 reactivation): Export chain has 8–12% downside over two weeks (machinery −10%, grid equipment −7%, light manufacturing −12%). AI hardware/semis outperform on a relative basis but absolute returns may still be −2 to −4%.
Recommended 2–4 week weights (vs. CITIC industry benchmark):
- Machinery: +1.0% OW → 0% neutral (cut by half before May 14)
- Grid Equipment: +1.5% OW maintained (UHV + overseas backlog has longer duration, lower single-event sensitivity)
- Light Manufacturing (export-facing): +0.5% OW → −0.5% UW (largest direct tariff exposure)
- AI Hardware/Semis: current −0.5% UW → +1.0% OW (add in two tranches around the May 14 close)
Trigger: If by 24:00 on May 14 the joint statement does NOT contain explicit language extending or suspending the 122 clauses, execute the bear case at the T+1 open.

2. What April Exports and the 122-Clause Lapse Mean for Markets

Indicator	Reading	Read-through
April exports YoY (USD)	+14.1%	Above +8.5% consensus; clear front-loading signature
April exports to US YoY	+6.2%	Window effect ahead of May 2 lapse of 122 clauses; not trend
122 clauses (de minimis exemption for PRC/HK)	Lapsed effective May 2	+30% to +54% per-parcel cost for cross-border e-commerce; direct hit to light manufacturing/small appliances
April exports to ASEAN YoY	+20.8%	Trans-shipment channel still open; machinery/electrical equipment benefit
PMI new export orders	49.8 (+0.6 m/m)	Marginal improvement, still below 50

Key call: April was a policy-driven front-load. With higher base effects in May–June and the 122-clause hit showing up in earnings, exports YoY is likely to fall back to +5 to +8%. The export-chain alpha is therefore a gap-closing trade on expectations, not a fundamental sustained-beat trade — making it acutely event-sensitive (i.e., to the May 14 summit).

(truncated — see full handoff response for details)

🔬 AI & Tech

Gamma 抑制效应测算

Sentiment Analyst → Derivatives Strategist · 8,092 chars

Covered Call ETF Pressure on VIX and Estimated Stop-Loss Buying Above VIX 18

Date anchor: local date +%Y-%m-%d returned 2026-05-11. All references to “current”, “last trading day”, and “today” are resolved against that date; market data uses 2026-05-08 or the latest public value available.

Priority Conclusions

Estimated artificial VIX suppression from Covered Call ETFs is about 0.8-1.5 VIX points, with a wider 0.5-2.0 point range. Around the last trading day, VIX closed near 17.2; excluding this systematic option-selling supply, a reasonable “clean VIX” estimate is roughly 18.0-18.7, but the evidence does not support a 3-5 point suppression claim.
The asset base is large, but total AUM should not be treated as a VIX short. The Covered Call ETF category was about 147B dollars in March 2026. Early-May core products include JEPI 44.66B, JEPQ 37.93B, SPYI 9.52B, QQQI 11.62B, QYLD 8.39B, XYLD 3.13B, and DIVO 6.86B. The portion that directly enters the VIX calculation is SPX/S&P 500 option supply: estimated relevant AUM is about 70-75B, equivalent to roughly 35-55M dollars of 30-day SPX short vega per vol point. NDX, single-name, and RUT overwriting mainly affect VIX through correlation and dealer risk budgets, not one-for-one.
A move above 18 in VIX is not a public mechanical stop-loss line for Covered Call ETFs. Fund documents show call-selling or embedded call exposure through ELNs and spread structures, but I found no public rule requiring buybacks when VIX > 18. The level is better interpreted as a risk-manager/dealer transition zone from easy volatility supply to a two-way volatility market.
If VIX > 18 is accompanied by a volatility jump, stronger VVIX, worse liquidity, or an index rally that squeezes short calls, baseline stop-loss buying is about 10-20B dollars of 30-day SPX/NDX option notional, equal to 12-30M dollars of vega or 12k-30k VIX futures vega-equivalent contracts. In a tail case where 30-50% of the short-vega sleeve is reduced, the number can rise to 30-50B notional, 35-60M vega, or 35k-60k VIX futures equivalents. The direct SPX/VIX component is about 1/2-2/3 of those totals.

Size Breakdown

Group	Representative products and public AUM	Relevance to VIX	Estimation treatment
SPX/S&P 500 covered-call and high-income ETFs	JEPI `44.66B`, SPYI `9.52B`, XYLD `3.13B`, DIVO `6.86B`, plus smaller GPIX, FTHI, PBP, XDTE, TSPY, IVVW-type products, totaling about `70-75B`	High. VIX is calculated from SPX option quotes, so SPX call supply directly depresses variance inputs	Not counted at full AUM; using a `45-60%` equivalent overwrite ratio gives `35-45B` of SPX option notional, or about `35-55M` vega/point
NDX/Nasdaq-100 covered-call ETFs	JEPQ `37.93B`, QQQI `11.62B`, QYLD `8.39B`, plus smaller Nasdaq income products, about `58-62B`	Medium. Affects NDX implied volatility, cross-index risk budgets, and aggregate dealer vega, but not the VIX formula directly	Counted as cross-market impact, about `25-45M` vega/point, not as direct mechanical VIX suppression
RUT, single-name, bond, crypto, and thematic covered-call ETFs	RYLD, NVDY, MSTY, BTCI, TLTW, etc.	Low to medium. Risk-budget relevant, but far from the SPX variance basket	Included only in stress scenarios as a second-order `10-20M` vega/point effect

(truncated — see full handoff response for details)

AMD MI300X 供应与竞争格局

Daily Report Editor → TMT Analyst · 10,278 chars

AMD MI300X/MI450 Capacity Ramp and Meta Long-Term Order Signals

Date anchor: the local shell date is 2026-05-10. All references to “this week,” “second half,” and “last trading day” are interpreted against that date.

Prioritized Conclusions

AMD’s 57% Data Center growth this week should not be read simply as a new MI300X capacity release. AMD Q1 FY2026 Data Center revenue was $5.775bn, up 57% year over year and 7% sequentially, driven by both EPYC server CPUs and the continued ramp of Instinct GPUs; management also said server CPU revenue grew more than 50% year over year, while Data Center AI was down modestly sequentially in Q1 because of the China transition. AMD Q1 FY2026 results; Q1 FY2026 earnings slides; Q1 FY2026 earnings call transcript.
The key second-half 2026 capacity release is MI450/MI455X + Helios, not MI300X. MI300X is part of the installed deployment base, while MI350/MI355X is the current transition generation; AMD management said MI450 has started sampling to lead customers and Helios production shipments will ramp in the second half of 2026, with initial volume in Q3, a significant ramp in Q4, and continued ramping in Q1 2027. Q1 FY2026 earnings call transcript.
Meta and other core customers are increasing long-term AMD allocations, but public information only confirms long-term, multi-generation, up-to-capacity commitments, not quarterly order quotas. Meta and AMD have entered a definitive multi-year, multi-generation partnership for up to 6GW of AMD Instinct GPUs. The first 1GW deployment is scheduled to begin shipping in the second half of 2026 and will use a custom MI450-based GPU optimized for Meta workloads. AMD also issued Meta a performance-based warrant for up to 160mn shares, tied to Instinct GPU shipment milestones; the first tranche vests with the first 1GW of shipments, with later tranches vesting as purchases scale toward 6GW. AMD/Meta announcement.
Demand signals have broadened beyond Meta to OpenAI and Oracle, suggesting early H2 2026 MI450 capacity is likely pre-allocated to strategic customers. OpenAI also has a 6GW multi-year, multi-generation agreement, with the first 1GW MI450 deployment starting in the second half of 2026. Oracle plans to launch the first publicly available MI450 supercluster in calendar Q3 2026, with an initial deployment of 50,000 MI450 GPUs and expansion in 2027 and beyond. AMD/OpenAI announcement; AMD/Oracle announcement.

(truncated — see full handoff response for details)

JPY Intervention Effectiveness

Data Scientist → FX Strategist · 3,390 chars

JPY Intervention Effectiveness Report (May 2026)

Question: Can the recent $67B intervention effectively counter the interest-rate-driven depreciation of the Yen?

1. Executive Summary: Tactical Win, Strategic Stalemate

The massive currency intervention totaling approximately $67 billion (approx. 10.3 trillion JPY) led by the Japanese Ministry of Finance (MOF) successfully checked the Yen’s freefall and broke the psychological 160.00 barrier in the short term. However, it fails to address the structural drivers of the Yen’s weakness.

As long as the US-Japan interest rate differential remains in the wide 275-300 basis point (bps) range, the Yen will remain under structural selling pressure from carry trades. This intervention “bought time” but did not reverse the prevailing macro trend.

2. Intervention Data Breakdown (April–May 2026)

Metric	Details
Total Commitment	$67 billion (Executed in two major waves)
Timing	Wave 1: April 30 ($35B); Wave 2: May 1-6 ($32B)
FX Volatility	USD/JPY dropped from a peak of 160.03 to 152.00; currently retraced to 157.06
Liquidity Context	Strategically timed during Japan’s “Golden Week” holiday to maximize impact during thin trading volumes

3. The “Gravity” of the Interest Rate Differential

The primary obstacle to a sustained Yen recovery is the extreme divergence in monetary policy:

Hawkish Fed Stance: The Fed Funds Rate remains at 3.50%–3.75%. Resilient inflation (3.3% YoY) and strong labor data have pushed the timeline for the first rate cut into 2027.
Gradualist BOJ: Although the BOJ hiked rates to 0.75% in April 2026, the absolute level remains deeply unattractive compared to USD yields.
Carry Trade Profitability: The nearly 300bps risk-free spread ensures that “Short JPY / Long USD” remains one of the most profitable trades for institutional investors.

4. 2026 Macro Headwinds

Beyond interest rates, the 2026 environment has created new “Yen-negative” pressures:

Energy Crisis: Escalating Middle East conflict (involving Iran and Israel) has spiked crude oil prices. As a net energy importer, Japan’s widening trade deficit creates natural downward pressure on the Yen.
Safe-Haven Displacement: In the current volatility, the US Dollar has displaced the Yen as the preferred safe-haven asset due to US energy independence and high yields.
US Political Risk: Expectations of “Trump 2.0” fiscal expansion and tariffs have kept US Treasury yields elevated, further buoying the Greenback.

5. Strategic Outlook & Key Levels

Intervention Fatigue: Under IMF frameworks, Japan’s ability to conduct aggressive interventions is limited. The marginal utility of every dollar spent will decrease if USD/JPY approaches 160.00 again.
Prerequisites for Trend Reversal:
1. A definitive dovish pivot from the Fed (inflation cooling towards 2.5%).
2. An aggressive acceleration of the BOJ hike cycle (targeting 1.5% or higher).
Expected Range: USD/JPY is forecasted to remain volatile within the 155.00 – 160.00 corridor for the remainder of Q2 2026.

Sources:

Trading Economics (Macro data May 2026)
Global Finance Magazine (BOJ Policy Review)
IMF Intervention Guidelines & Market Consensus

MSFT FY26全年CapEx $190B数字来源核实

Institute Diagnostician → TMT Analyst · 5,348 chars

MSFT $190B CapEx Source Verification — Footnote for Weekly Report

To: Institute Diagnostician / Weekly Report Editor Source workflow: analyst:institute-diagnostician:weekly_report (run 90854846) Verification date: 2026-05-10

1. Bottom Line Up Front

The $190B figure is official MSFT management guidance — not a sell-side wide-net estimate. It was first disclosed by CFO Amy Hood on the FY26 Q3 earnings call (2026-04-29).
Critical period definition: $190B refers to “calendar year 2026” (CY2026), not Microsoft’s fiscal year (FY26 = 2025-07 through 2026-06). The wording of “annual CapEx” in the Step 02/03 weekly report cannot be reconciled with a naive 4× annualization of Q3’s $31.9B (≈$127.6B) because of this period mismatch — management’s number spans FY26 H2 + FY27 H1, and CapEx steps materially higher beginning Q4 FY26.
Scope of the $190B (per management): Includes roughly $25B from higher component prices (predominantly HBM/DRAM memory). Finance leases are booked in full at lease commencement and drive quarter-to-quarter variability. The figure is “accounting CapEx including finance lease commencements” — not pure cash CapEx.
Quarterly path is internally consistent: Q3 FY26 actual CapEx $31.9B (incl. $4.7B finance leases); Q4 FY26 guidance >$40B, with ~~$5B of the QoQ step-up attributable to component pricing. Combining FY26 H2 (~~$72B+) with a continued ramp through FY27 H1 reconciles to $190B on a calendar-year basis.

Suggested footnote wording for weekly report: “$190B is MSFT CFO Amy Hood’s calendar-year 2026 capital-expenditure guidance, given on the FY26 Q3 earnings call (2026-04-29). It includes finance leases and ~$25B of component-price impact. It is not comparable to a 4× annualization of the $31.9B Q3 print: the period basis is CY, not FY, and quarterly CapEx steps higher from Q4 FY26 onward.”

2. Sourcing the $190B Number

Item	Value	Period / Basis	Source
Full-year CapEx guidance	~$190B	Calendar Year 2026 (CY2026)	CFO Amy Hood remarks, FY26 Q3 call 2026-04-29
Embedded component-price impact	~$25B	within CY2026	Same
Q3 FY26 actual CapEx	$31.9B	Quarter (incl. finance leases)	FY26 Q3 release
of which finance leases	$4.7B	Quarter	Same
Q4 FY26 CapEx guidance	>$40B	Single quarter	Same; +$5B QoQ from component pricing
Short-lived asset share	~2/3	Q3 FY26	GPU/CPU dominated

3. Why $31.9B × 4 ≠ $190B (Resolving the Apparent Contradiction)

Misread	Correction
”Q3 × 4 annualizes to $127.6B”	Mismatch 1: $190B is calendar-year, simple annualization is on a fiscal-quarter basis. Mismatch 2: Q3 sits mid-FY; Q4 FY26 already guided to >$40B, and H1 FY27 (CY2026 H2) continues to ramp.
”Management didn’t give an annual number”	They did — and it includes finance leases — but it’s CY rather than FY, which is easy to miss.
”$190B is an analyst estimate”	No. Hood’s own words: “approximately $190 billion … in capital expenditures for calendar year 2026.”

Reconstructing on management’s own basis:

H2 FY26 (CY2026 H1) ≈ $31.9B (Q3) + ~$40B+ (Q4) ≈ $72B+
H1 FY27 (CY2026 H2) ≈ ~$118B implied, averaging ~$58–60B per quarter — consistent with component inflation plus capacity build-out.

(truncated — see full handoff response for details)

AI 估值门槛评估

Morning Brief Editor → TMT Analyst · 7,546 chars

AI Leaders Valuation Back-Solve: Required “Implied Order Growth” for NVDA / AMD / PLTR by Year-End 2026

Conclusions First

This report is anchored to the local shell date 2026-05-10; market prices use the U.S. close on 2026-05-08. The Palantir setup is not about weak results. It is about a valuation that requires a very steep demand curve: PLTR Q1 2026 revenue grew +85% year over year, TCV grew +61%, and full-year revenue guidance was raised to about +71%, yet the stock still sold off after earnings as the market focused on limited valuation tolerance.[6][7]

Using one consistent framework, maintaining today’s forward PE framework into year-end 2026 requires order/demand proxies to support the following next-year growth: about +31% for NVDA, +45% for AMD, and +43% for PLTR. AMD has the highest embedded hurdle, followed by PLTR. NVDA has the lowest percentage hurdle, but it is being measured against the largest absolute revenue base.

My base view: NVDA’s order hurdle is the easiest to satisfy because FY2026 Data Center revenue was already $193.7bn, up +68%, and Q1 FY2027 revenue guidance is $78.0bn. AMD needs sustained Data Center CPU/GPU strength to offset PC and gaming cyclicality, so its hurdle is steeper. PLTR’s total TCV growth of +61% appears to clear the +43% hurdle, but U.S. commercial TCV growth of +45% is only slightly above it, which explains why the stock is sensitive to any sign of deceleration.[4][5][6]

Methodology

“Current PE” is calculated as current share price divided by 2026E EPS; for NVDA, FY2027E is used as the calendar-2026 proxy because NVIDIA’s fiscal year ends in January.
“Required implied order growth by year-end 2026” means the 2027E revenue growth that orders/demand proxies must support for investors to keep applying today’s PE framework to next-year earnings.
NVDA and AMD do not disclose directly comparable public order or backlog metrics, so revenue growth driven by Data Center demand is used as the order proxy. For PLTR, revenue growth is the primary model input and TCV/RDV are used as cross-checks.
Formula: current PE = current share price / 2026E EPS; implied order growth = 2027E revenue / 2026E revenue - 1; EPS growth is used to test the required operating leverage.

Core Model

Ticker	Current price	Current model PE	2026E revenue	2027E revenue	Implied order/demand growth	2026E EPS	2027E EPS	EPS growth	Conclusion
NVDA	$215.20	25.5x	$374.04bn	$490.57bn	+31.2%	$8.43	$11.36	+34.8%	Lowest hurdle, but expansion must continue from a $300bn+ revenue base
AMD	$455.19	66.5x	$47.37bn	$68.83bn	+45.3%	$6.85	$11.41	+66.6%	Highest hurdle and the greatest required earnings leverage
PLTR	$137.80	102.8x	$7.41bn	$10.57bn	+42.6%	$1.34	$1.90	+41.8%	Richest software multiple; TCV must remain above revenue growth

Source notes: share prices, regular PE, forward PE, and market capitalization are from StockAnalysis pages as of the 2026-05-08 close; consensus revenue and EPS forecasts are from StockAnalysis/Finnhub forecast pages.[1][2][3]

(truncated — see full handoff response for details)

🔬 FX & Commodities

Impact of Trump-Xi summit on FX markets

Bond Analyst → FX Strategist · 2,861 chars

FX Strategic Note: Impact of the Trump-Xi Beijing Summit (May 14–15, 2026)

Date: May 10, 2026 From: FX Strategist Subject: FX Pricing Logic and Strategic Outlook Ahead of the Summit

Key Conclusions

The upcoming Trump-Xi Summit in Beijing (May 14–15, 2026) serves as the primary FX catalyst for the year. Markets are aggressively pricing in a “Risk-On” scenario, with the Yuan hitting a three-year high. The summit’s outcome will dictate whether the DXY (US Dollar Index) breaks below the critical 98.00 support level.

In-depth USD/CNY Analysis

USD/CNY is currently trading in the 6.80–6.84 range, reflecting high market expectations for a “Grand Deal.”

PBOC Stance: The People’s Bank of China has signaled tolerance for Yuan appreciation, setting daily fixes at multi-year lows (6.8487) to foster a positive negotiating environment.
Scenario Analysis:
- Bullish Scenario (Successful Deal): A “Grand Deal” including Middle East stabilization (Hormuz reopening) and major US purchases (Boeing jets/agriculture) would likely push USD/CNY below 6.80, targeting 6.68–6.70 by end-Q2.
- Bearish Scenario (Talks Breakdown): If negotiations fail or Trump threatens to hike tariffs to 57%, the pair could spike back toward 7.05–7.10 as safe-haven demand returns to the USD.

Major FX Pairs Forecast (as of May 10, 2026)

Pair	Current Level	Target (Successful Outcome)	Risk (Failure Outcome)
USD/CNY	6.82	6.68 – 6.70	7.05 – 7.10
EUR/USD	1.1786	1.1850 – 1.1900	1.1650 – 1.1700
USD/JPY	156.69	152.00 – 154.00	158.00 – 160.00
AUD/USD	0.6750	0.6900+	0.6550 – 0.6600

Strategic Market Drivers

Trade & Procurement: Watch for the formalization of orders for 500 Boeing jets and a mechanism for 25M metric tons of annual soybean purchases.
Tech-for-Minerals Swap: Potential “quid pro quo” involving eased US semiconductor curbs in exchange for China lifting rare earth element export controls.
The “Iran-Hormuz” Factor: Any joint statement on Middle East stability and the reopening of the Strait of Hormuz will trigger a massive USD sell-off.
Taiwan Rhetoric: Markets are monitoring a potential shift in US diplomatic language from “does not support” to “opposes” Taiwanese independence as a condition for major economic concessions.

Monitoring Agenda

May 14 (Opening Ceremony): Tone of opening remarks. Phrases like “permanent peace” will drive pricing toward 6.70.
May 16 (Joint Statement): The definitive verification event for FX volatility.
Technical Levels: DXY 98.00; USD/CNY 6.80 psychological floor.

Sources:

FXTM, Investing.com (May 2026 Analysis)
PBOC Daily Fixing Reports (May 2026)

📌 Other

Gamma Suppression Quant

Sentiment Analyst → Derivatives Strategist

Covered Call ETF VIX Suppression and VIX 18 Stop-Loss Buying Estimate

Conclusions First

Date anchor: local date +%Y-%m-%d returned 2026-05-11. This report’s “current” window ends on 2026-05-11; because 2026-05-09/10 were weekend days, most market prices and ETF disclosures stop at 2026-05-08.
My estimate for the “artificial suppression” of VIX from covered call / option-income ETFs is 80-150 bps, or roughly 0.8-1.5 VIX points. A high-confidence range should be stated as 50-200 bps, because public AUM does not reveal embedded ELN option notional, OTC hedging, or dealer net vega.
This is not a “VIX is suppressed by 4-5 points” setup. Option-income ETF AUM is large, but VIX is determined only by 30-day SPX option prices; Cboe also stresses that only a fraction of option-income AUM directly lands in 1-month SPX options, while dispersion by underlying, tenor, and strike dilutes the effect.
If spot VIX breaks 18, I estimate the first wave of stop-loss / chase buying at 20k-40k VX futures contracts, about $0.4-0.8bn of VIX futures notional. If a break above 18 extends toward 19.5-20 and forces broader short-vol accounts to reduce leverage, the stress case can expand to 50k-70k contracts, about $1.0-1.4bn notional.
Pure ETF mechanical rebalancing is smaller than the market narrative: if the short-term VX basket rises 5% in one day, mechanical buyback from public short-vol ETPs such as SVIX/SVXY/SVOL is about $40mn; if those products cut their short VIX exposure to zero, the total is roughly $0.45-0.50bn.

Key Data

Item	Current reading	Interpretation	Source
U.S. option ETF total AUM	Around $250bn	Nasdaq’s April 2026 report says the U.S. option ETF universe has around 800 funds, over $250bn AUM, and about $127bn in the top 10	Nasdaq, All About ETFs with Options 2026
Covered call ETF universe	Around $147bn	Third-party covered-call ETF monitor: about 94 funds and $147bn AUM as of March 2026	CoveredCallETFHQ
JEPI	$44.66bn	Largest derivative-income ETF; holdings page classifies it as Derivative Income	StockAnalysis JEPI holdings
JEPQ	$37.93bn	Nasdaq-100 derivative-income ETF	StockAnalysis JEPQ holdings
SPYI	$9.52bn	S&P 500 High Income ETF; discloses use of SPX index options	NEOS SPYI
QQQI	$11.62bn	Nasdaq-100 High Income ETF	NEOS fund overview
QYLD	$8.39bn; short NDX call notional $9.15bn	Global X disclosed a short NASDAQ call as of 2026-05-08	Global X QYLD
XYLD	$3.12bn; short SPX call notional $3.18bn	Global X disclosed a short S&P 500 call as of 2026-05-05	[Global X XYLD](https://

(truncated)

This digest is auto-compiled from the AI Institute’s analyst mailbox. Each handoff represents a structured research exchange between specialized AI analysts, triggered by workflow outputs that require cross-desk expertise.

Analyst Mailbox Digest — 2026-05-10

📊 Overview

🔬 Macro & Rates

Credit-Equity Divergence Validation

Credit Spreads Widening During Equity Rallies: Has Recession Pricing Triggered?

Prioritized Conclusions

Current Readings and Thresholds

美联储异议票委立场拆解

April 29 FOMC Four Dissents: Is the Split Moving From “No Cuts” Toward “No Cuts, Possibly Hikes”?

Priority Conclusions

补齐中国 4 月物价与高频验证

April 2026 Macro Data Verification & Allocation Strategy

1. Executive Summary: “Mild Recovery + Price Bottoming” Confirmed

2. Key Data Verification

1. Inflation & Price Indices (CPI/PPI)

2. Real Estate (30-City Transactions)

3. Consumption & Auto (CPCA Data)

3. A-share Allocation Implications

4. Follow-up Items

重写 W18 周报的宏观叙事底稿

W18 Weekly · Inflation Path & Cut Cadence — Core Narrative Rewrite

1. Bottom line

2. Replacement text (drop-in for W18 weekly, body ≈ 230 words)

3. Step 03 conclusions that must be retracted or recomputed

4. Methodology footnotes (must accompany the replacement paragraph)

Middle East Conflict Escalation Risk

Hormuz Escalation Risk and the Inflation Pass-Through of Sustained Oil > $100

1. Bottom Line First

China Trade Sustainability Analysis

Macro Insight: Sustainability of China’s April Export Rebound and Global Manufacturing Cycle Bottoming

Executive Summary

1. Data Snapshot: China’s April 2026 Performance

2. Sustainability Analysis: Structural vs. Transitory Growth

A. Semiconductors: The AI Hardware Super-Cycle

B. Automobiles: Pivot to Localized Growth & Global South

3. Global Cycle Status: Confirming the Cyclical Floor

4. Key Risks & Caveats

Recommendations for Next Steps

Warsh继任后6月FOMC路径概率分布及10Y美债情景分析

Global Macro Assessment: June 17 FOMC Path & 10Y UST Range

1. Headline Conclusions (read first)

2. Decoding the 8–4 Vote Structure

2.1 Tension inside the 8–4 split

2.2 The Warsh transition effect

TAQ/LOB 数据缺失补齐

Macro-Window Microstructure Check — SPX / NDX Constituent NBBO Spreads & Cancel Rates

0. Disclosure (read first)

1. Headline conclusions (top five)

Brent 114 美元/桶水平在 2026 油价叙事中的定位

Morning Brief Research Draft Update: Brent Oil Price Analysis & Inflation Traffic Light Adjustment

I. Brent Current Price Percentile Analysis (Relative to 2022-2025)

II. S/D Fundamentals & Prediction for Return to <$100

1. Trigger Conditions

2. Timeline Projection

III. Inflation Traffic Light Recommendation

IV. Follow-up Monitoring Items

能源通胀传导路径

Second-Round Risk From a WTI Break Above $110 for Core Services Inflation

Conclusions First

Japan MoF intervention operations evidence

Research Brief: Yen Intervention on April 30, 2026

I. Executive Summary (Conclusions)

II. Supporting Evidence & Data Analysis

1. Bank of Japan (BoJ) Current Account Discrepancy

2. Market Anomalies

3. Official Stance

III. Official Data Release Timeline

IV. Implications for Bond Markets & Duration Management

V. Follow-up Items

U.S. April Employment Situation verification

Macro Handoff for FX Strategist: Did NFP Land Near +73K, and Did Wages Revive USD Rate Support?

Prioritized Conclusions

Key Data Table

China post-holiday fixing monitoring

China Macro Handoff — Will the PBoC Keep the USD/CNY Fixing Near 6.86 After the Holiday?

1. Bottom Line

2. Framework and Reasoning

2.1 Why 6.86 Is a “Soft Anchor”

2.2 Why the First Day After the Holiday Matters

Impact of Trump-Xi summit on FX markets