← Back to scans

China x Taiwan military clash before 2027?

0x4c80df6f557bb0b9de08c0301ebdce500dd9bdc58400d9269fb8fab1709ed584 · World · 2026-08-17
6%
Agent
7%
Market Price
-1.1%
Edge
68%
Confidence
Volume: 3,214,189
Spread: 0.2c
Days to resolution: 135
Markets in event: 1
Final Rationale
The Polymarket price of 7.1% is a live, liquid quote that already reflects elapsed time without incident, so Forecast 1's additional time-decay haircut partly double-counts with the base-rate argument — the critique is right on that point. Against that, the resolution bar is genuinely high: every documented escalation (Dec 2025 live-fire inside 12nm, Jan 2026 drone incursion, ongoing CCG activity at Kinmen/Pratas) is explicitly non-qualifying, and there has been no confirmed PRC-ROC direct-fire exchange since 1996 (λ≈3.4–6.9%/yr). The critique's strongest surviving points — accidental/miscalculation channels and a Taiwan shoot-down of a PLA drone converting grey-zone activity into a qualifying kinetic event — argue against pricing meaningfully below the market, while Beijing's demonstrated calibration below the kinetic threshold and active US de-escalation pressure argue against pricing above it. I therefore settle just under the anchor at 6%, splitting the two forecasts and reflecting that the remaining window (including the flagged Oct 2026 decision point) is shorter than the full contract period but still tension-laden.
Price vs. prediction
How the market moved after the agent's call. Solid blue line is the actual market price; dashed line is the agent's static forecast at $t{=}0$.
$n = 17$ follow-ups
1. Planner Opus · Fermi decomposition + tool plan
Tools chosen: polymarket_direct polymarket_related kalshi_related claude_news gdelt_news wikipedia code_execution
Sub-questions (Fermi decomposition)
  1. What is the current Polymarket YES price and how has it moved over the last 90 days?
  2. Do Kalshi or other Polymarket markets price China-Taiwan conflict/invasion/blockade before 2027, and do they agree with this market's price?
  3. What is the historical base rate of qualifying kinetic engagements (gunfire, missile/artillery strikes, damaging ship rammings) between PRC and ROC military/CCG forces since 1996/2000?
  4. Have there been any recent near-miss incidents (CCG boarding/ramming of Taiwanese vessels, Kinmen/Matsu confrontations, drone shootdowns) that could plausibly cross the resolution threshold?
  5. What is the current level of PLA activity around Taiwan (large-scale drills like Strait Thunder/Joint Sword, median-line crossings, blockade rehearsals) and are analysts flagging escalation into 2026?
  6. How does the ongoing China-Japan diplomatic crisis over Takaichi's Taiwan remarks and the US posture (arms sales, Trump-Xi dynamics) shift escalation risk in 2026?
  7. Given a ~13.5-month window, what annualized hazard rate does the market price imply, and what does that imply relative to the historical base rate?
Planner reasoning
This is a low-base-rate geopolitical tail-risk question with a strict resolution bar (actual use of force causing casualties/damage, excluding warning shots and routine incursions; CCG counts on China's side but Taiwan's CGA does not). The key work is anchoring on the Polymarket price, checking cross-venue (Kalshi) pricing, establishing the historical base rate of qualifying PRC-ROC kinetic incidents, and scanning current news for escalation drivers (China-Japan spat, PLA drills, Taiwan election/defense budget cycle) over the ~13.5 months remaining.
2. Research Parallel tool calls
ToolStatusTimeItemsSummary
polymarket_direct OK 0.5s 1 ## This Market's Polymarket Data **China x Taiwan military clash before 2027?** - Current price (probability): 7.10% - 7-day price change: +0.20% - 30-day price change: +1.85% - Total volume: $3,214,189 (USD notional) - Price range: 4.50% - 13.00% - Data points: 90 days
polymarket_related OK 4.9s 0 Scanned 100 active Polymarket markets, kept 0 matches. keyword 'Taiwan': 0 markets | keyword 'China invade': 0 markets | keyword 'China Taiwan blockade': 0 markets | keyword 'China Japan': 0 markets
kalshi_related OK 4.8s 3 3 related markets / summaries. keyword 'Taiwan': ok | keyword 'China military': ok | keyword 'China invasion': ok
claude_news OK 33.4s 17 ## Key Findings (Nov 2025 – Aug 2026) **PLA Drills — "Justice Mission 2025" (Dec 29–31, 2025)** - China conducted its largest military drills off Taiwan in 8 months, surrounding it with warships, aircraft and live-fire exercises after a record U.S. arms sale to Taipei . The drills covered "blocka
gdelt_news FAILED 240.0s 0 timeout after 240.0s
wikipedia OK 0.2s 5 Fetched 5 Wikipedia entries (1 missing pages).
code_execution OK 17.4s 0 ## Key Findings **Window conversion (13.5-month window = 1.125 years) using P = 1 − exp(−λt):** - λ = 1%/yr → **1.12%** window probability - λ = 2%/yr → **2.23%** - λ = 4%/yr → **4.40%** - λ = 8%/yr → **8.61%** - λ = 15%/yr → **15.53%** - (Linear approx λ·t overstates by up to ~1.3pp at λ=15%; negl
3. Evidence Brief Sonnet · 8017 chars
# Current state No qualifying "military encounter" (direct kinetic exchange of fire, or damaging/sinking ramming) between PRC and ROC military forces has occurred as of the latest reporting (mid-2026); resolution requires actual force-on-force violence, not drills, warning shots, or airspace/water incursions. Grey-zone pressure (CCG incursions, record drills, drone incursions) has escalated sharply but remains below the kinetic threshold. Polymarket prices this event at ~7%. # Timeline of key events - 1996 (confirmed): Third Taiwan Strait Crisis — PLA missile tests near Taiwan; no direct clash. [Wikipedia] - 2023-04, 2024-05, 2024-10 (confirmed): "Joint Sword" PLA exercises around Taiwan (largest pre-2025 drills); no kinetic clash. [Wikipedia] - 2024-02 (confirmed): Kinmen speedboat incident — CCG pursuit of Taiwanese fishing boat, 2 Chinese fishermen died in collision; CCG began regular incursions into restricted Kinmen waters afterward. [AEI] - 2025-02 (confirmed): CCG incursions begin recurring near Pratas Islands. [AEI] - 2025-12-28/31 (confirmed): "Justice Mission 2025" — largest PLA drills off Taiwan in 8 months, following record US arms sale; live-fire drills entered Taiwan's 12nm territorial waters for the first time, disrupted 857 flights, cut Kinmen air links. [Fox News, CNN, Model Diplomat] - 2026-01 (reported): PLA drone crossed into Taiwan's territorial airspace for the first time. [Model Diplomat] - 2025-11/2026-01 (confirmed/reported): China-Japan diplomatic crisis erupts after PM Takaichi states a Taiwan attack could trigger Japanese military response; Wang Yi calls it a "red line" crossing; Japan protests China's Dec. drills as escalatory; Trump intervenes urging Takaichi to soften tone. [NBC, AP/Yahoo, PBS/AP, Tribune India] - Through mid-2026 (confirmed): No new fatal incidents or fire exchanges reported since Feb 2024 Kinmen collision; CCG presence expanding but still non-kinetic. [Vigil Consilium] - Flagged: October 2026 — analyst Su Tzu-yun (via Model Diplomat) flags as a potential "next decision point" tied to a CCP political milestone. # Event Will there be a qualifying military encounter (direct kinetic force, not warning shots/incursions) between PRC and ROC military forces between Nov 11, 2025 and Dec 31, 2026? # Outcomes to forecast Yes / No # Kalshi market anchor No direct Kalshi ticker for this exact event found; nearest analog is the Polymarket contract itself. **Polymarket YES price: 7.1%**, up +0.2% (7d) and +1.85% (30d); range 4.5%–13% over 90 days; volume $3.2M — a liquid, actively traded market showing gradual upward drift consistent with rising tensions but still low absolute probability. [Polymarket direct] # Sub-question answers 1. **Polymarket 90-day trend?** Started near lows (~4.5%), rose to a 13% peak (likely around the Dec 2025 drills/Takaichi dispute), now settled at 7.1%, up modestly over 30/7 days. [Polymarket direct] 2. **Do other Kalshi/Polymarket markets on invasion/blockade agree?** No matching Polymarket markets found (0 hits for Taiwan/invasion/blockade keywords). Kalshi's Taiwan-adjacent markets (Level 4 travel advisory 0.1%, Trump recognizing Taiwan 14%) don't directly price conflict; no Kalshi invasion/blockade contract found. No independent cross-check available. [kalshi_related, polymarket_related] 3. **Historical base rate of qualifying kinetic engagements since 1996?** No confirmed direct-fire exchange or qualifying damaging ramming between PRC/ROC militaries since 1996; the Feb 2024 Kinmen collision caused deaths but involved CCG vs. Taiwan Coast Guard (CGA is explicitly excluded from this market's definition) and was accidental, not a use-of-force incident. Effectively ~0–1 ambiguous incidents in ~29 years. [Wikipedia, AEI] 4. **Recent near-miss incidents?** Feb 2024 Kinmen incident (excluded, CGA not counted); ongoing CCG incursions near Kinmen/Pratas since 2024-2025; Dec 2025 PLA drills breached 12nm territorial waters (first time) and Jan 2026 drone airspace incursion (first time) — both explicitly excluded from resolution per market rules (non-violent incursions don't count). No qualifying gunfire/missile/ramming incidents reported. [Model Diplomat, AEI, CNN] 5. **Current PLA activity level, analysts flagging 2026 escalation?** Record-setting: Dec 2025 "Justice Mission 2025" drills were largest in 8 months with unprecedented territorial-water live-fire and blockade rehearsal messaging (Eastern Theater Command). Analysts (The Diplomat, CNN, Vigil Consilium) describe steady normalization/testing of the status quo, with Su Tzu-yun flagging Oct 2026 as a possible escalation decision point tied to CCP milestones. [Fox, CNN, Diplomat, Model Diplomat] 6. **China-Japan crisis and US posture impact?** Takaichi's remarks triggered a serious diplomatic rupture, with China calling it a "red line" and linking drills partly to Japan tensions; record US arms sale to Taiwan coincided with Dec 2025 drills. However, Trump has actively pushed de-escalation (urging Takaichi to soften tone), suggesting a moderating force. Net effect: elevated rhetorical/military tension but with major-power efforts to contain kinetic escalation. [NBC, Fox, Tribune India] 7. **Implied hazard rate given 13.5-month window?** Market prices (5-12%) imply annualized hazard rates of ~4.6%-11.4%/yr; historical base rate (1-2 incidents/29yrs) implies ~3.45-6.9%/yr. Current 7.1% Polymarket price implies λ≈6.3%/yr, near the upper end of historical base rate — arguably fair-to-slightly-elevated given documented escalation trend. [code_execution] # Key facts (high-confidence, factual) 1. [Polymarket] Current YES price 7.1%, volume $3.2M, 90-day range 4.5-13%. 2. [Fox/CNN] Dec 2025 drills were largest in 8 months; first-ever live-fire within Taiwan's 12nm waters. 3. [Model Diplomat] Jan 2026: first PLA drone incursion into Taiwan airspace. 4. [AEI/Vigil Consilium] CCG incursions near Kinmen (since Feb 2024) and Pratas (since Feb 2025) ongoing through 2026, no kinetic clash. 5. [NBC/AP] China-Japan diplomatic crisis over Takaichi's Taiwan remarks, unresolved as of research window. 6. Market definition explicitly excludes CGA-involved incidents, warning shots, and non-damaging incursions — raising the bar substantially above observed grey-zone activity. # Cross-market signals - Kalshi: no direct match; adjacent Taiwan markets don't price conflict directly. - Polymarket: only this contract found; no companion invasion/blockade markets to cross-check. - Sportsbook: N/A. # Analyst opinions and speculation - CNN/Chieh Chung: PLA drills aim to deter foreign intervention, not launch war. [CNN] - The Diplomat: warns "routine" drills are normalizing pressure, "testing a core element of the status quo." [Diplomat] - Su Tzu-yun (via Model Diplomat): flags Oct 2026 as a possible escalation decision point. - Vigil Consilium: CCG grey-zone activity is the primary risk vector, deliberately calibrated to avoid triggering kinetic response. # Directional lean per outcome - **Yes**: Supported by record drills, territorial-water breaches, drone incursions, unresolved Japan crisis, and Oct 2026 flagged as elevated-risk date. Opposed by: no historical precedent of qualifying kinetic clash in ~29 years; Trump actively de-escalating; Beijing's demonstrated preference for grey-zone over kinetic escalation. - **No**: Favored by base rates, current trajectory (escalation confined to grey-zone/non-kinetic acts explicitly excluded by resolution criteria), and market pricing under 10%. # Gaps / unknowns - No Kalshi-specific price for this exact contract confirmed (only Polymarket anchor available). - Unclear whether Oct 2026 "decision point" will produce anything beyond further drills. - GDELT news tool failed (timeout), limiting real-time incident-level cross-check. # Calibration anchors - Polymarket current YES price: 7.1% (anchor). - Historical base rate implies λ≈3.4-6.9%/yr → window prob ≈3.8-7.5%. - No confirmed qualifying kinetic PRC-ROC military clash since 1996.
4. Forecasters Opus + GPT · in parallel
claude-opus-5
0.71
Yes 5%
No 95%
The Polymarket anchor sits at 7.1% for the full Nov-2025-to-Dec-2026 window, but no qualifying kinetic exchange has occurred through mid-2026, so a meaningful chunk of the window has already passed without incident, mechanically shrinking the remaining hazard. The resolution bar is high — CGA-involved incidents, warning shots, and non-damaging incursions are excluded — and all documented escalation (Dec 2025 'Justice Mission' live-fire inside 12nm, Jan 2026 drone incursion, CCG activity near Kinmen/Pratas) is explicitly grey-zone and non-qualifying. Historical base rate implies λ≈3.4-6.9%/yr with zero confirmed PRC-ROC direct-fire exchanges since 1996, and Beijing has consistently calibrated pressure to stay below the kinetic threshold while Washington actively pushes de-escalation. Offsetting risks (unresolved China-Japan Takaichi crisis, Su Tzu-yun's flagged Oct 2026 decision point) keep this above a pure base-rate floor, so I land slightly below the market anchor at ~5%.
gpt-5.5
0.66
Yes 7%
No 93%
Polymarket's liquid anchor is 7.1% YES, with a recent range of 4.5%–13%, so I start near 7%. The trend and recent unprecedented PLA activity point toward Yes because record drills, live-fire inside Taiwan's 12nm waters, drone incursions, and the China-Japan crisis raise the near-term accident/escalation risk. However, the historical base rate and market rules point strongly toward No because there has been no confirmed qualifying PRC-ROC military kinetic clash since 1996, and many recent incidents involve CCG/CGA or non-violent incursions that are explicitly excluded. The remaining risk is an Oct 2026-style escalation or miscalculation, but evidence suggests Beijing is still calibrating grey-zone pressure below the kinetic threshold.
5. Devil's Advocate Sonnet · red-team critique
1. Both forecasters anchor near Polymarket's 7.1% but neither explicitly addresses the time-decay math: the window is Nov 11 2025–Dec 31 2026 (13.5 months), and by "mid-2026" evidence cutoff roughly half the window has elapsed without incident — Forecast 1 mentions this "shrinking hazard" but doesn't quantify it, and Forecast 2 ignores it entirely, treating the full-window anchor as still fully applicable. 2. Neither forecast meaningfully grapples with the accidental/collision risk pathway — given rising CCG/PLA proximity incidents (Kinmen 2024, Pratas 2025, 12nm live-fire, drone incursions), the base rate for an *unintentional* kinetic incident (e.g., a collision or miscalculated live-fire hitting a Taiwanese vessel) may be rising faster than the deliberate-escalation base rate, and this is a distinct hazard channel from "Beijing chooses war." 3. Both treat the Oct 2026 "decision point" (Su Tzu-yun) as a minor tail risk without discussing what CCP political milestone it's tied to or why analysts specifically flagged it — this deserves more explicit weighting or dismissal with reasoning, not a passing mention. 4. Neither forecast addresses potential resolution ambiguity: if a CCG-Taiwan Coast Guard incident escalates (as almost happened in Feb 2024) and ROC Navy gets involved, or if a PLA drone is shot down by Taiwan (converting a grey-zone act into a kinetic ROC-initiated response), this could flip a "non-qualifying" incident into a qualifying one — a scenario not modeled by either forecaster. 5. Both forecasts sit at or below the market price (5% and 7%) without addressing the asymmetry that the Polymarket 90-day range topped at 13% during the Dec 2025 drills/Takaichi crisis — implying the market itself has shown it can reprice significantly higher on realistic escalation scenarios, which slightly undercuts confidence in staying anchored near the low end. 6. Forecast 1's specific claim of landing "slightly below the market anchor at ~5%" is not strongly justified — the time-decay argument is reasonable but partially double-counts with the historical base-rate reasoning already used to justify sub-market pricing, risking over-confidence (0.71 stated confidence) on a fairly thin analytical distinction between 5% and 7%.
6. Aggregator Opus · final calibrated forecast
6% (68% confidence)
Yes: 6% No: 94%
The Polymarket price of 7.1% is a live, liquid quote that already reflects elapsed time without incident, so Forecast 1's additional time-decay haircut partly double-counts with the base-rate argument — the critique is right on that point. Against that, the resolution bar is genuinely high: every documented escalation (Dec 2025 live-fire inside 12nm, Jan 2026 drone incursion, ongoing CCG activity at Kinmen/Pratas) is explicitly non-qualifying, and there has been no confirmed PRC-ROC direct-fire exchange since 1996 (λ≈3.4–6.9%/yr). The critique's strongest surviving points — accidental/miscalculation channels and a Taiwan shoot-down of a PLA drone converting grey-zone activity into a qualifying kinetic event — argue against pricing meaningfully below the market, while Beijing's demonstrated calibration below the kinetic threshold and active US de-escalation pressure argue against pricing above it. I therefore settle just under the anchor at 6%, splitting the two forecasts and reflecting that the remaining window (including the flagged Oct 2026 decision point) is shorter than the full contract period but still tension-laden.
Pipeline Timing
Total pipeline time: 328.2s
Per-tool research timings shown in the Research section above.