Tools chosen: polymarket_direct polymarket_related kalshi_related wikipedia claude_news gdelt_news code_execution

Sub-questions (Fermi decomposition)

Planner reasoning

This is essentially a base-rate question with a small conditional update for current volcanic unrest: VEI 6+ eruptions are rare (roughly 1 per 50-100 years historically, though Hunga Tonga 2022 was recent), so the prior for any single year is low single-digit percent. Research should anchor on the Polymarket price, verify the historical frequency of VEI≥6 events in the Smithsonian GVP record, and check for any current elevated-alert volcanoes capable of a caldera-scale eruption.

# Current state As of the latest available reporting (news through 2026-08-20), no VEI≥6 eruption has occurred in 2026 or 2025; multiple volcanoes are highly active (Etna, Fuego, Lewotobi Laki-laki, Semeru, Campi Flegrei unrest) but all confirmed activity is far below VEI 6 (typically VEI 2-4). Resolution hinges solely on the Smithsonian GVP's official VEI-6 tally for calendar-year 2026, assessed as of March 2027 — a discrete, verifiable dataset, not press coverage of "large" eruptions. # Timeline of key events - 2022-01-15 (confirmed): Hunga Tonga–Hunga Haʻapai eruption, GVP-rated ≥VEI-5 (some sources cite VEI-6 equivalent atmospheric energy); most recent eruption near the VEI-6 threshold globally. - ~2025 Q2-Q4 (confirmed): Polymarket price for this market trades near its 90-day low (~3.6%). - 2026-02 (reported): Iwo Jima/Ioto shows elevated uplift (~20-30 cm/yr) and submarine unrest; no magmatic climax confirmed (volcanolive.com). - 2026 Q1 (reported): Campi Flegrei uplift reaches ~163.5 cm cumulative at Rione Terra; alert remains "yellow" (Italian Civil Protection). - 2026-06-25 to 2026-08-20 (confirmed): Wave of eruptions — Etna (Italy, ash disrupting flights, early July), Semeru & Lewotobi Laki-laki (Indonesia, repeated eruptions June-Aug), Fuego (Guatemala, Aug 4-5, 1,700 evacuated), three Indonesian volcanoes erupting in succession (Aug 20). All are moderate (VEI 2-4), none approach VEI 6. - 2026-07-31 (reported): M4.7 earthquake at Campi Flegrei, described by INGV as intensifying unrest, "not imminent eruption." - 2026 (last 7-30 days, per Polymarket data as queried): Polymarket YES price surges from ~5% to 35.2% (+30 pts in both 7d and 30d windows), coinciding with the above eruption/unrest news cluster, despite no VEI≥6 event being confirmed. # Event Will a natural VEI≥6 volcanic eruption occur globally between market creation and Dec 31, 2026 (per Smithsonian GVP data)? # Outcomes to forecast Yes / No # Kalshi market anchor No direct Kalshi price was returned by the kalshi_direct tool in this research pass (gap). The only Kalshi data available is a related market: "Will a supervolcano erupt before 2050?" (KXERUPTSUPER-0-50JAN01) at 24.3% YES, down from ~33% 30 days ago, thin volume (~49 contracts/day). This is a much longer, broader-eligibility window and not a reliable direct proxy. **Treat Kalshi anchor as unavailable/uncertain — use Polymarket (35.2%) as the best available cross-market reference, with caution given thin liquidity.** # Sub-question answers 1. **Historical frequency of VEI≥6/yr** — Roughly 4-6 events per 100 years using 20th-century/instrumental-era counts (Katmai 1912, Pinatubo 1991, etc.); longer 1815-2025 window (Tambora, Krakatau, Santa Maria, Novarupta, Pinatubo, possibly Hunga Tonga) gives ~6/210yr ≈ λ0.0286/yr [code_execution model]. 2. **VEI≥6 since 1900 / retroactive assignment** — Confirmed VEI-6+: Santa Maria (1902), Novarupta/Katmai (1912), Pinatubo (1991); Hunga Tonga (2022) was GVP-rated VEI-5/5+ (borderline, not confirmed VEI-6) [Wikipedia]. GVP has occasionally revised VEI assignments upward post-analysis, but no systematic retroactive-VEI6 pattern is documented in this research. 3. **Current elevated-alert volcanoes with caldera-scale potential** — Campi Flegrei (yellow alert, ongoing bradyseism/uplift, no imminent-eruption signal per INGV); Ioto/Iwo Jima (rapid uplift, submarine unrest, no magmatic climax confirmed); Axial Seamount (inflating, predicted possible 2026 eruption but effusive/basaltic — incapable of VEI≥6) [claude_news]. None show confirmed VEI-6-scale precursors. 4. **Polymarket price/trend** — 35.2% current, up from a 90-day low of 3.6%; +30.05% (7d) and +30.55% (30d); total volume only $139,376 over 91 days [polymarket_direct] — thin liquidity, single sharp rally. 5. **Comparable markets** — No direct Kalshi VEI≥6-2026 market found; closest is Kalshi's "supervolcano before 2050" (24.3%, declining) [kalshi_related]. No other Polymarket volcano/disaster markets exist currently [polymarket_related]. 6. **Time remaining / partial-year exposure** — Window is full calendar year 2026; as of latest news (Aug 2026) ~8 months have elapsed with no VEI≥6 event, leaving ~4 months of exposure. Poisson model: remaining-period probability at λ≈0.03-0.04/yr for a 4-month window ≈1.0-1.3%, far below full-year estimates [code_execution]. # Key facts (high-confidence, factual) 1. [code_execution/Wikipedia] No VEI≥6 eruption since Pinatubo (1991); Hunga Tonga (2022) rated VEI-5, not confirmed VEI-6. 2. [claude_news/GDELT] All 2025-2026 active volcanoes (Etna, Fuego, Lewotobi, Semeru, Klyuchevskoy) are VEI 2-4, well below threshold, through Aug 2026. 3. [code_execution] Long-run Poisson base rate: full-year P(VEI≥6 in 2026) ≈ 2.8%-4% (best estimate), remaining-year (post-Aug) ≈1-1.3%. 4. [polymarket_direct] Polymarket YES has surged from 3.6% to 35.2% in 30 days on thin volume (~$139k total). # Cross-market signals - Kalshi related (supervolcano-by-2050): 24.3%, declining — much longer horizon, not comparable. - Polymarket (same event): 35.2%, sharply rising despite no confirming VEI≥6 event. - No sportsbook or other prediction markets found. # Analyst opinions and speculation - code_execution model flags likely "fade/short" edge if market price is materially above ~4-6% given historical base rates; a 35% Polymarket price appears strongly disconnected from base-rate math. - VolcanoCafe/claude_news analysts flag Ioto/Iwo Jima as historically "most dangerous" but confirm no imminent magmatic climax. - Axial Seamount widely discussed for 2026 eruption but universally noted as incapable of VEI≥6 explosivity. # Directional lean per outcome - **Yes**: Supported only by heightened public/market attention to unrest (Campi Flegrei quake, multiple simultaneous eruptions) and a speculative Polymarket price surge; not supported by any confirmed VEI≥6-scale precursor or event. - **No**: Strongly supported by base-rate Poisson modeling (1-4% for remaining window), absence of any VEI≥6 eruption through Aug 2026, and consensus that all currently unrest-prone systems (Campi Flegrei, Ioto, Axial) show no imminent caldera-forming signal. # Gaps / unknowns - No live Kalshi_direct price was retrieved for this specific ticker — cannot confirm the "consensus" anchor precisely; Polymarket used as substitute. - Reason for Polymarket's 30-point rally is not explained by any confirmed VEI≥6 event — could reflect illiquid/manipulated pricing, hedging against Campi Flegrei tail risk, or data artifact (avg daily volume = total volume, suggesting possible data quality issue). - No GVP dataset update cited directly in research for 2026 findings. # Calibration anchors - Polymarket YES: 35.2% (thin volume, sharp recent rally) — best available cross-market anchor; Kalshi direct anchor unavailable. - Kalshi related (supervolcano-by-2050): 24.3%, declining. - Base-rate Poisson estimate: ~2.8-4% full-year, ~1-1.3% for remaining 2026 exposure — the most defensible structural anchor, given no VEI≥6 event has occurred and ~2/3 of the year has already passed without one.

1. Both forecasts converge tightly (3% vs 5%) near the pure Poisson base rate but neither seriously grapples with why Polymarket sits at 35.2% — a 7-10x divergence from their own estimate; simply calling it "thin/illiquid" doesn't fully explain a sustained 30-point rally over 30 days, and neither forecaster attempts to quantify what fraction of that gap might reflect real information (e.g., insider awareness of Campi Flegrei/Ioto data) versus pure noise. 2. Both treat the Poisson base rate as a fixed constant (λ≈0.03/yr) without discounting for the observation that 2025 also passed without a VEI≥6 event, and that VEI-6 classification itself carries retroactive-revision uncertainty (the brief notes "GVP has occasionally revised VEI assignments upward post-analysis") — this could push effective probability of a "surprise" reclassification (e.g., an ambiguous 2026 eruption later upgraded to VEI-6) slightly higher than the raw structural estimate suggests, a scenario neither model addresses. 3. Neither forecast explicitly addresses classification/resolution risk: if a large 2026 eruption (e.g., a major Kamchatka or Indonesian event) is initially reported as VEI-5 but the GVP tally isn't finalized until March 2027, there could be ambiguity in resolution timing or borderline VEI-5.5–6 assignment (as happened with Hunga Tonga) — this tail scenario is unaddressed and could justify slightly fatter tails than 3-5%. 4. Both anchor heavily on the same base-rate math and reach nearly identical conclusions (3% and 5%), which is reasonable given the evidence, but this convergence means neither forecaster is independently stress-testing the other's discounting of the market signal — an adversarial "steelman" of the Yes case (e.g., possible undisclosed seismic/GVP internal data driving the Polymarket rally) is largely absent from both. 5. Neither forecast discusses the asymmetry that Kalshi's direct price was unavailable, so both are essentially "flying blind" on the true consensus anchor for this specific contract; treating Polymarket as the "best available" while heavily discounting it risks anchoring too confidently on the structural model alone without acknowledging the wider uncertainty band this data gap should introduce (e.g., confidence intervals like 2-8% rather than point estimates of 3% or 5%). 6. Both forecasters note Axial Seamount is "incapable of VEI≥6" and dismiss it, but neither considers other under-monitored candidate systems (e.g., Kamchatka's Klyuchevskoy group, historically volatile Indonesian arc volcanoes, or unmonitored submarine calderas) that could produce a surprise VEI≥6 event outside the currently tracked short-list, understating true tail risk from an incomplete global monitoring net.

4% (75% confidence)

Yes: 4% No: 96%

The structural anchor dominates: VEI≥6 eruptions occur ~3-6 times per century (λ≈0.03/yr), none since Pinatubo 1991, and roughly two-thirds of 2026 has already elapsed with all confirmed activity at VEI 2-4 — leaving a remaining-window Poisson exposure of only ~1-1.3%. Unrest at Campi Flegrei (INGV: no imminent eruption), Ioto, and Axial Seamount (basaltic, non-explosive) provides no caldera-scale precursor. The Polymarket 35.2% print rests on ~$139k volume with an unexplained 30-point rally and apparent data-quality artifacts (avg daily volume ≈ total volume), so it warrants heavy discounting rather than partial-credit averaging. Per the critique, I widen slightly above the pure base rate to account for GVP retroactive-reclassification risk, borderline VEI-5/6 assignments (Hunga Tonga precedent), under-monitored submarine/Kamchatkan systems, and the missing Kalshi anchor — landing at 4% Yes rather than the raw ~1-2% remaining-window estimate.

Total pipeline time: 268.7s