← Back to scans

China x Philippines military clash before 2027?

0x2167ab5c0d163f85195ccea4d11734e4ba53913c838c6720fb815794128c68f2 · World · 2026-08-21
26%
Agent
28%
Market Price
-1.5%
Edge
52%
Confidence
Volume: 233,591
Spread: 3.0c
Days to resolution: 131
Markets in event: 1
Final Rationale
The strict bar (gunfire, melee weapons, hull-breaching ramming, PH Navy/AFP specifically) has gone unmet through three-plus years of intense but sub-threshold friction, and most 2025 incidents involved excluded PCG or civilian fisheries vessels. Only ~5 months of window remain, which mechanically cuts the Poisson-implied fair value below the ~27% derived from an 11-month window; the offsetting factor is a clearly accelerating hazard rate (993 CCG ship-days at Scarborough in June, three clashes in one week in late July, PLA drills Aug 1, first PH Navy-involved melee July 20, no durable de-escalation mechanism), plus genuine resolution ambiguity over whether baton/oar blows count as 'melee weapons' — which asymmetrically adds YES risk if a similar incident recurs. The critique is right that Polymarket's -6pt weekly drift against rising real-world tempo is odd, but a modest decline is plausibly the market correctly reading that July's incidents did not cross the threshold. Netting the shorter window (down) against accelerating tempo plus loose-resolution tail risk (up), I stay essentially at the anchor, shaded marginally lower to 26%.
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 $t{=}0$.
$n = 13$ follow-ups
Re-scan Context
This market has been scanned before. Previous predictions:
DatePredictedMarket PriceConfidence
2026-08-14 28% 34% 53%
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 its trend over the past 90 days for this market?
  2. How many incidents since 2023 would have met this market's bar (melee weapons use, gunfire, or ramming causing hull-breaching damage) between Chinese CCG/PLAN and Philippine Navy/AFP vessels?
  3. Do recent incidents involve Philippine Coast Guard vessels (excluded) rather than Philippine Navy/military vessels (included)?
  4. What is the current escalation trajectory in the South China Sea (Scarborough Shoal, Second Thomas Shoal, Thitu) as of late 2025/2026, including any Chinese 'nature reserve' declarations or new Philippine resupply postures?
  5. Have there been any bilateral de-escalation arrangements (e.g., provisional resupply agreement, hotlines) that reduce clash probability?
  6. What is the implied monthly hazard rate needed to match the market price over the remaining window, and does the historical base rate support it?
Planner reasoning
This is a geopolitical tail-risk question with a strict resolution bar (actual use of force / significant ramming damage, CCG counts as Chinese military but PCG excluded). The key is the current market price anchor, the historical base rate of South China Sea incidents that would have met this bar (e.g., June 2024 Second Thomas Shoal machete/ramming incident), and the current escalation trend into 2026. I'll pull the Polymarket price, cross-venue signals, and news on recent incidents.
2. Research Parallel tool calls
ToolStatusTimeItemsSummary
polymarket_direct OK 0.3s 1 ## This Market's Polymarket Data **China x Philippines military clash before 2027?** - Current price (probability): 27.50% - 7-day price change: -6.00% - 30-day price change: +1.00% - Total volume: $233,591 (USD notional) - Price range: 26.00% - 34.00% - Data points: 28 days
polymarket_related OK 4.1s 0 Scanned 100 active Polymarket markets, kept 0 matches. keyword 'China Philippines': 0 markets | keyword 'South China Sea': 0 markets | keyword 'Taiwan invasion': 0 markets | keyword 'China military': 0 markets
kalshi_related OK 4.0s 3 3 related markets / summaries. keyword 'China Philippines': ok | keyword 'South China Sea': ok | keyword 'China military conflict': ok
claude_news OK 24.9s 19 ## Key Findings: China-Philippines South China Sea Incidents (2025-2026) - **Aug 11, 2025 – CCG/PLA Navy collision (Scarborough Shoal):** a China Coast Guard vessel and a People's Liberation Army Navy warship collided while pursuing a Philippine Coast Guard patrol vessel , with the Philippines cla
gdelt_news OK 173.5s 10 GDELT: 10 articles across 3 queries (lookback=90d). 'China Philippines South China Sea collision ramming': error GDELT rate-limited after retries (429) | 'Scarborough Shoal Philippine Navy China Coast Guard incident': error GDELT rate-limited after retries (429) | 'Second Thomas Shoal resupply missi
wikipedia OK 0.1s 4 Fetched 4 Wikipedia entries (2 missing pages).
code_execution OK 21.9s 0 ## Poisson Base-Rate Analysis: China–Philippines Military Clash Before 2027 **Base-rate calculation (historical anchor):** - Historical data: 1 qualifying force-use incident in ~36 months → λ = 0.333 incidents/year - Over the remaining ~11-month window (t = 0.917 yr): **P(≥1 incident) = 26.3%** - F
3. Evidence Brief Sonnet · 7667 chars
# Current state As of early August 2026, no incident has yet met the market's strict bar (gunfire, melee-weapon use, or hull-breaching ramming) between PLA/CCG and Philippine Navy/AFP forces — the closest call is a July 20, 2026 baton/oar melee at Second Thomas Shoal involving actual Philippine Navy personnel (one sailor injured), which is the strongest evidence yet of trending toward a qualifying event. Kalshi's own YES price was not retrieved (kalshi_direct tool absent from research); Polymarket's parallel market for the same event trades at 27.5%. # Timeline of key events - 2025-08-11 (confirmed): CCG vessel collided with a PLAN warship while chasing a PCG ship near Scarborough Shoal; Philippines claims 3 CCG deaths, 1 CCG vessel disabled — intra-Chinese collision, not a China-Philippines clash. (Wikipedia) - 2025-09 (confirmed): China declares "Huangyan Island National Nature Reserve" over Scarborough Shoal, escalating sovereignty assertions. (Wikipedia) - 2025-09-16 (confirmed): CCG fires water cannon on Philippine ships near Scarborough Shoal amid mutual ramming accusations (10+ PH vessels involved). (NBC News) - 2025-10-12 (confirmed): CCG rams/damages Philippine Bureau of Fisheries vessel (not Navy/Coast Guard) near Thitu; no injuries. (military.com) - 2025-12-12/13 (confirmed): CCG/militia cut anchor lines of 20 Filipino fishing boats at Sabina Shoal; 3 civilians injured, 2 boats water-cannon damaged — civilian, not military. (USNI) - 2026-07-20 (confirmed): Second Thomas Shoal melee — CCG personnel exchange blows (batons/oars) with Philippine Navy personnel escorting resupply craft; PH claims a sailor injured. Closest event to qualifying "military encounter" to date. (CNN, Al Jazeera, Newsweek) - 2026-07-23/24 (confirmed): Three clashes in one week — repeated water-cannon barrages at Scarborough Shoal, a 12-CCG-vessel/3-militia-vessel confrontation, "dangerous maneuvers," near-collision within 7m. Prompts joint US-PH-Japan drills. (Bloomberg, Al Jazeera, CFR) - 2026-08-01 (confirmed): PLA conducts joint naval/air drills near Scarborough Shoal after objecting to PH seabed-rights extension bid off Palawan. (SCMP) - Early Aug 2026 (reported): Additional PH-China vessel collision near Spratlys; analysts flag Beijing acting more assertively, believing US distracted by Persian Gulf. (chinaglobalsouth.com) # Event Will there be a military encounter (gunfire, melee weapons, hull-breaching ramming, etc.) between Chinese and Philippine military forces (CCG counts as Chinese military; PCG does NOT count as Philippine military) before Dec 31, 2026? # Outcomes to forecast Yes / No # Kalshi market anchor Kalshi YES price unavailable in this research pull (kalshi_direct tool not returned). Nearest available anchor: **Polymarket YES = 27.5%**, down 6pts over 7 days, up 1pt over 30 days, range 26-34% over 28 days, $233.6K volume — suggesting the market has cooled slightly from a recent high near 34% despite July 2026 escalation. # Sub-question answers 1. **Polymarket YES price/trend** — 27.5% currently; -6% (7d), +1% (30d); traded between 26-34% over past month. (polymarket_direct) 2. **Qualifying incidents since 2023** — Zero incidents unambiguously meet the bar (gunfire, melee weapons, hull-breaching ramming). Closest: July 20, 2026 baton/oar exchange with PH Navy involvement (injury, but no weapons/hull damage); code_execution's base-rate model counted ~1 "qualifying" incident in 36 months, likely a generous read of that event. 3. **PCG vs Navy involvement** — Most 2025 incidents (Sept, Oct, Dec) involved PCG or civilian fisheries vessels — excluded per rules. The July 20, 2026 Second Thomas Shoal incident is the first to explicitly involve Philippine **Navy** personnel — directly relevant to resolution. (claude_news) 4. **Escalation trajectory** — Sharply up: China's Sept 2025 nature-reserve declaration, near-continuous CCG presence (993 ship-days at Scarborough in June 2026 alone, near full-2025 total), three clashes in one week (July 2026), and PLA drills (Aug 2026) all signal intensifying friction. (claude_news, cfr.org, scmp.com) 5. **De-escalation arrangements** — None identified as durable; a "resupply deal"/agreement over Ayungin (Second Thomas Shoal) is contested — Philippine officials reject Chinese claims about permission requirements, and a domestic probe (Pangilinan) into the "Ayungin deal" suggests friction over even the existing arrangement, not stabilization. (gdelt_news) 6. **Implied hazard rate vs base rate** — Poisson model: historical base rate (~1 incident/36mo) implies ~26-28% fair price for a 11-month window; market's ~27.5% is roughly base-rate-consistent, not pricing aggressive acceleration. However, the code_execution window (11 months) appears to overstate remaining time — actual window from ~Aug 2026 to Dec 2026 is closer to 4-5 months, which would mechanically lower fair-value probability at a constant hazard rate, unless λ has risen due to 2026 escalation (three clashes/week trend argues yes). # Key facts (high-confidence, factual) 1. [Wikipedia] China declared Huangyan (Scarborough) a "national nature reserve" Sept 2025. 2. [CNN/Al Jazeera] July 20, 2026: first PH Navy-involved physical melee (batons/oars), sailor injured — no weapons/hull damage confirmed. 3. [CFR/CNN] CCG presence at Scarborough hit 993 ship-days in June 2026, near full-year-2025 levels. 4. [Bloomberg/Al Jazeera] Three separate clashes occurred within one week in late July 2026. 5. [SCMP] PLA conducted joint drills near Scarborough Aug 1, 2026. # Cross-market signals - Kalshi related: No directly comparable Kalshi market found; adjacent China-risk markets (EUV, GDP, FTA) show no spillover signal. - Polymarket: Same-event market at 27.5%, recently declining from ~34% peak despite escalation — possibly reflecting relief that July incidents didn't cross the weapons/hull threshold. - Sportsbook implied: N/A. # Analyst opinions and speculation - chinaglobalsouth.com: dispute entering a "more dangerous phase," Beijing possibly emboldened by US distraction in Persian Gulf, even risking Mutual Defense Treaty invocation. - claude_news synthesis: July 2026 incident represents "closest approach yet" to qualifying clash but doesn't meet the bar. # Directional lean per outcome - **Yes**: Escalating frequency/intensity of incidents, first direct PH Navy involvement (July 2026), CCG presence at record highs, PLA drills, deteriorating diplomatic posture, analyst warnings of "dangerous phase." - **No**: No incident to date has met the strict definition (no gunfire, no melee weapons proper, no hull-breaching ramming); China/PH incidents to date largely involve PCG/civilian vessels (excluded) or water cannon/minor damage (excluded); short remaining window (~4-5 months) limits opportunity even if hazard rate is rising. # Gaps / unknowns - No Kalshi direct price retrieved — cannot state Kalshi's own consensus; Polymarket used as proxy. - Exact market creation date unknown, affecting true elapsed/remaining window calculations. - Whether "baton/oar" blows would ever qualify as "melee weapons" per resolution source is ambiguous and unresolved by rules text. - No confirmation of any events after early Aug 2026 in this research pull, leaving ~5 months of the window (through Dec 2026) unassessed. # Calibration anchors - Polymarket YES (proxy anchor): 27.5%, range 26-34% over 28 days. - Poisson base-rate model: ~26-28% "fair" price for an 11-month window at historical incident rate (1/36mo); shorter actual remaining window (~4-5 months) would push fair value lower unless hazard rate is judged to have risen (supported by 2026 escalation trend).
4. Forecasters Opus + GPT · in parallel
claude-opus-5
0.48
Yes 26%
No 74%
The market's strict bar (gunfire, melee-weapon use, or hull-breaching ramming between CCG/PLA and AFP/PH Navy) has not been met through early August 2026: 2025's incidents involved PCG or civilian fisheries vessels (excluded) and water cannon/minor ramming (excluded), and the closest call — the July 20, 2026 baton/oar melee at Second Thomas Shoal with an injured PH Navy sailor — apparently did not trigger resolution, since the parallel Polymarket contract still trades at 27.5% (down from ~34%). Escalation is clearly intensifying (993 CCG ship-days at Scarborough in June 2026, three clashes in one week in late July, PLA drills Aug 1), which argues the hazard rate has risen above the historical ~1-per-36-months base rate. But only ~5 months of window remain, which mechanically cuts the probability relative to the 11-month Poisson calculation that produced the ~27% fair value. Netting a higher hazard rate against a shorter window, I land essentially at the 27.5% anchor, shaded marginally down to 26%.
gpt-5.5
0.56
Yes 28%
No 72%
Sub-question 1 points toward Yes around 27.5% because the closest available market anchor, Polymarket, prices the event at 27.5% despite a recent decline from the mid-30s. Sub-questions 2 and 3 point toward No because no incident has unambiguously met the strict bar yet, and most prior clashes involved PCG or civilian vessels, though the July 20 PH Navy melee makes the risk more direct. Sub-questions 4 and 5 point toward Yes because the tempo and severity of confrontations have risen sharply, with repeated late-July clashes, record CCG presence, PLA drills, and no durable de-escalation mechanism. Sub-question 6 is mixed: a historical base-rate model is near the 27.5% anchor, but the short remaining window argues lower while the 2026 escalation argues higher, so I stay close to the market at 28%.
5. Devil's Advocate Sonnet · red-team critique
1. Both forecasters treat Polymarket's 27.5% as a hard anchor without questioning why it fell 6pts in the last 7 days right after the most escalatory week (three clashes, PLA drills) — this divergence between falling price and rising real-world hazard rate is flagged in the brief but neither forecaster interrogates whether the market itself is mispriced or lagging, they just defer to it as "roughly correct." 2. Neither forecast meaningfully engages with the ambiguity noted in the brief that "whether 'baton/oar' blows would ever qualify as 'melee weapons' per resolution source is ambiguous and unresolved" — this is a live resolution-criteria risk that could flip a future similar incident to YES, and deserves explicit uncertainty-widening rather than a point estimate treated as settled by Polymarket's continued sub-30% price. 3. Both forecasters use the July 20 non-trigger as evidence the bar is "hard to meet," but this is a single data point from an unresolved market with no confirmed adjudication — they're inferring resolution logic from price movement (a circular reference-class error) rather than from actual published resolution criteria. 4. The escalation trend itself (993 ship-days in June, three clashes in one week in late July, PLA drills Aug 1, analyst warnings of a "more dangerous phase," Beijing "possibly emboldened by US distraction") suggests a hazard rate that may not be constant/Poisson but accelerating — both forecasters acknowledge this but still net out to essentially the historical base-rate-implied figure (26-28%), likely underweighting tail scenarios of rapid escalation in the remaining Aug-Dec window. 5. Neither forecaster addresses the "additional PH-China vessel collision near Spratlys" reported in "Early Aug 2026" as a distinct, more recent data point beyond the July 20 melee — this event is buried in the brief and could represent further escalation not yet priced by either the Poisson model or Polymarket's -6% weekly move. 6. Both converge extremely tightly (26% vs 28%) despite explicitly using the same weak, incomplete anchor (Kalshi price unavailable, Polymarket as proxy, ambiguous resolution criteria) — this convergence looks more like anchoring-on-each-other-via-the-same-proxy than independent analysis, and neither stress-tests how sensitive the estimate is to the window-length assumption (4-5 months vs 11 months) which they both acknowledge "mechanically" lowers fair value but then don't actually apply a lower adjusted figure.
6. Aggregator Opus · final calibrated forecast
26% (52% confidence)
Yes: 26% No: 74%
The strict bar (gunfire, melee weapons, hull-breaching ramming, PH Navy/AFP specifically) has gone unmet through three-plus years of intense but sub-threshold friction, and most 2025 incidents involved excluded PCG or civilian fisheries vessels. Only ~5 months of window remain, which mechanically cuts the Poisson-implied fair value below the ~27% derived from an 11-month window; the offsetting factor is a clearly accelerating hazard rate (993 CCG ship-days at Scarborough in June, three clashes in one week in late July, PLA drills Aug 1, first PH Navy-involved melee July 20, no durable de-escalation mechanism), plus genuine resolution ambiguity over whether baton/oar blows count as 'melee weapons' — which asymmetrically adds YES risk if a similar incident recurs. The critique is right that Polymarket's -6pt weekly drift against rising real-world tempo is odd, but a modest decline is plausibly the market correctly reading that July's incidents did not cross the threshold. Netting the shorter window (down) against accelerating tempo plus loose-resolution tail risk (up), I stay essentially at the anchor, shaded marginally lower to 26%.
Pipeline Timing
Total pipeline time: 272.3s
Per-tool research timings shown in the Research section above.