Tools chosen: kalshi_direct kalshi_related polymarket_related wikipedia claude_news gdelt_news code_execution

Sub-questions (Fermi decomposition)

Planner reasoning

This is a very-long-horizon geophysical tail-risk question; the dominant inputs are (a) the precise resolution definition Kalshi uses for 'supervolcano' (VEI 8? specific volcano list? magnitude threshold?), (b) the geological base rate of such eruptions per year, and (c) the market's current price, which for far-dated tail markets is often distorted by time value and low liquidity. I will anchor on the Kalshi market, pull the rulebook/definition via news and Wikipedia, and compute a Poisson base rate over the ~24-year window.

# Current state Kalshi's KXERUPTSUPER-0-50JAN01 resolves YES only if a supervolcano (scientifically: a VEI-8 eruption, >1,000 km³ ejecta) occurs before Jan 1, 2050. No official Kalshi rules page listing adjudicating source/volcano list was found; resolution presumably defaults to the standard VEI-8 scientific definition. No VEI-8 eruption has occurred in ~25,600 years (last one: Taupō/Oruanui). Current unrest (Campi Flegrei seismicity, Yellowstone ground deformation) is being monitored but no volcano observatory considers a supereruption imminent. # Timeline of key events - **2004-2009**: Yellowstone experienced notable uplift from magma injection (confirmed, historical) — Wikipedia. - **~2005-present**: Campi Flegrei unrest (bradyseism, seismicity) increasing (confirmed) — claude_news/INGV. - **2025-01**: USGS/Nature study maps Yellowstone subsurface magma in scattered pockets (2-30% melt), not an eruptible reservoir (confirmed) — claude_news. - **2025-05**: Stanford/INGV study attributes Campi Flegrei uplift/quakes to geothermal reservoir pressure, not magma (confirmed) — ScienceDaily. - **2025-07**: New Yellowstone north caldera rim uplift begins (confirmed, GPS data) — claude_news. - **2025-11**: Campi Flegrei seismicity total (3-yr AI-reanalyzed catalog) rises to >54,000 quakes; INGV director calls recent quakes "strongest we've ever seen" (confirmed/reported) — WRVO/OPB. - **2026-01 (mid)**: Yellowstone caldera-rim uplift ceases, no resumption since (confirmed via GPS) — claude_news. - **2026-07-16**: M-magnitude earthquake near Yellowstone reported widely by local media (reported, routine seismicity, not linked to eruption precursor) — GDELT/Dailymail. - **2026-07-31/08-01**: M4.7 earthquake strikes Naples/Campi Flegrei area, "strongest in 40 years," injuries and evacuations reported (confirmed) — ANSA, Independent, multiple outlets. - **2026-08-03/05**: Hundreds displaced/evacuated near Campi Flegrei following aftershocks (confirmed) — ANSA. No event in the timeline constitutes or foreshadows an actual VEI-8 eruption; all are unrest/monitoring/precursor-adjacent activity at sub-supervolcanic intensity. # Event Will a supervolcano (VEI-8 eruption) occur anywhere on Earth before Jan 1, 2050? # Outcomes to forecast Yes / No # Kalshi market anchor **Current YES price: 23.30%** (as of latest snapshot). 7-day change: -0.70%; 30-day change: -5.70% (declining). Range over 158 days: 14.00%–32.90%. Average daily volume: ~124 contracts (thin/illiquid). This price is dramatically above any geological base-rate estimate for a true VEI-8 event. # Sub-question answers 1. **Kalshi's exact resolution criterion** — No official Kalshi rules page found; description says simply "if a supervolcano erupts." Third-party aggregators and Wikipedia consistently define "supervolcano" eruption as VEI-8 (>1,000 km³ ejecta), suggesting this is the implicit standard, but no named-volcano list or explicit adjudicator was located (claude_news, Wikipedia). 2. **Geological base rate of VEI-8 eruptions** — Roughly 3 VEI-8 eruptions in the last ~2.1 million years globally (Yellowstone alone), with the most recent VEI-8 anywhere being Taupō ~25,600 years ago (Wikipedia/Supervolcano). Rougier et al. 2018 estimate a mean global recurrence interval of ~17,000 years for VEI-8 events (code_execution). 3. **Poisson probability over ~24.2-year window** — Using ~17,000-yr recurrence: P≈0.14%. Using 45,000-100,000-yr estimates: P≈0.02-0.05%. Yellowstone-only (~730,000-yr): P≈0.003%. If the market loosely admits VEI≥7 events (broader "caldera-forming" definition), P rises to ~2.4-4.7% (code_execution). 4. **Current unrest at known systems** — Campi Flegrei shows the most pronounced unrest (rising seismicity, M4.7 quake Aug 2026, evacuations), but 2025 Stanford/INGV research attributes this to hydrothermal/geothermal pressure, not magma ascent — scientists say eruption "not imminent" but risk "not zero" (claude_news). Yellowstone is at NORMAL alert, magma is fragmented/non-eruptible, and 2025-2026 uplift has ceased (claude_news, USGS). Long Valley is cooling/quieting (Caltech study). Taupo shows no notable unrest reported. 5. **Kalshi price/volume behavior** — YES priced 23.3%, down from a 30-day high near 32.9%, trending down over time; thin volume (~66-124 contracts/day) suggests a "lottery ticket" or long-shot-bias/illiquidity effect rather than efficient pricing, since it vastly exceeds any credible geological probability (kalshi_direct). 6. **Other markets pricing similar risk** — Polymarket has no matching markets (0 hits for supervolcano/volcano/Yellowstone/eruption). No other Kalshi eruption-specific series found besides this one (polymarket_related, kalshi_related). # Key facts (high-confidence, factual) 1. [Wikipedia] Supervolcano = VEI-8 eruption, >1,000 km³ ejecta; last such event was Taupō ~25,600 years ago. 2. [Wikipedia/claude_news] Yellowstone: 3 supereruptions in 2.1M years; implies ~1-in-730,000 annual base rate; currently NORMAL alert, magma not eruptible (USGS Jan 2025 Nature study). 3. [claude_news] Campi Flegrei seismicity surged from ~12,000 to >54,000 quakes (3-yr reanalysis); M4.7 quake Aug 2026 was strongest in 40 years, causing evacuations. 4. [ScienceDaily/Stanford-INGV 2025] Campi Flegrei unrest attributed to geothermal reservoir pressure, not magma — reduces (but doesn't eliminate) near-term eruption concern. 5. [code_execution] Poisson-model fair value for a strict VEI-8 definition: ~0.02%-0.14% by 2050; loosened to VEI≥7: ~2.4%-4.7%. # Cross-market signals - Kalshi related: only this single market found under "supervolcano"/"volcano eruption" keywords; no Yellowstone-specific market exists. - Polymarket: no matching markets at all (0/100 scanned). - Sportsbook implied: none available. # Analyst opinions and speculation - Octagon AI aggregator cites fluctuating implied probabilities (14%-23%) with dramatic "civilization-ending" framing, reflecting entertainment/clickbait pricing rather than rigorous science (claude_news). - Scientific consensus (INGV, USGS, Caltech): no credible near-term VEI-8 eruption forecast anywhere; Campi Flegrei is the most closely watched but experts stress eruption is "not imminent," risk "not zero" (claude_news). # Directional lean per outcome - **Yes**: Supported only by Campi Flegrei unrest and long 24-year window allowing some tail risk; no support from base rates or expert consensus. - **No**: Overwhelmingly supported — strict scientific VEI-8 base rate implies <0.2% probability over this window; all monitored systems (Yellowstone, Campi Flegrei, Long Valley, Taupo) show either declining activity or non-magmatic explanations for unrest; last VEI-8 event was 25,600 years ago. # Gaps / unknowns - No confirmed official Kalshi resolution rules (definition, adjudicating source) — key ambiguity if a VEI-7 "supervolcano-labeled" eruption (e.g., a future Campi Flegrei event) occurs; market could interpret loosely. - Unclear how thin liquidity/lottery-bias affects long-run price convergence toward true probability. # Calibration anchors - Kalshi current YES price: **23.30%** (anchor, likely overpriced vs. geological base rate). - Geological Poisson estimate (VEI-8 strict): **~0.02%-0.14%** over window. - Loose VEI≥7 definition Poisson estimate: **~2.4%-4.7%**. - Precedent: last true VEI-8 eruption occurred 25,600 years ago (Taupō); no VEI-8 event in recorded human history.

1. Both forecasters converge on heavily discounting the Kalshi anchor, but neither seriously entertains that Kalshi's ambiguous resolution wording could actually resolve using a real adjudication source (e.g., Smithsonian GVP) that might apply VEI-8 strictly with no ambiguity at all — in which case even 7% is too high; they should have shown a wider split between "strict-definition" and "loose-definition" scenarios rather than a single blended point estimate. 2. Forecast 1's jump to 7% is poorly justified quantitatively — the brief's "loose VEI≥7" Poisson estimate tops out at ~4.7%, yet Forecast 1 lands above that ceiling while also citing the same evidence Forecast 2 uses to land at 2.5%; this inconsistency between using the same brief but landing on opposite sides of the stated loose-definition range is not explained. 3. Neither forecaster explicitly assigns weight to unknown/unmonitored supervolcano systems (e.g., Toba, Taupo reawakening, or an entirely undetected caldera) — the brief notes "unknown systems" as a gap, but both treat the four named systems (Yellowstone, Campi Flegrei, Long Valley, Taupo) as an exhaustive set. 4. The declining Kalshi price trend (32.9%→23.3%, -5.7% over 30 days) could reflect the market itself converging toward the geological base rate over time, which both forecasters cite as evidence for heavy discounting, but neither considers this trend as a signal that could continue further, potentially supporting an even lower probability than either estimate. 5. Both forecasters imply high confidence (0.62, 0.68) despite explicitly built-in large uncertainty about resolution criteria — a genuinely unresolved structural ambiguity (no rules page found at all) should arguably lower confidence more than stated, since the outcome could hinge entirely on an undocumented adjudication decision rather than any geological forecast. 6. Neither model addresses the scenario where Kalshi resolves "No" by default/void due to the market lacking a clear supervolcano occurring and ambiguous rules causing the contract to expire worthless or be voided — a structural/administrative resolution risk distinct from the geological question, which could affect probability mass allocation.

2% (65% confidence)

Yes: 2% No: 98%

Strict VEI-8 (>1,000 km³) supereruptions have a global recurrence of ~17,000-100,000 years, giving only ~0.02-0.14% over the 24-year window; no VEI-8 has occurred in ~25,600 years and no observatory forecasts one. All monitored systems argue against: Yellowstone's magma is fragmented/non-eruptible (USGS Jan 2025 Nature) with uplift ceased, Long Valley is cooling, and Campi Flegrei's dramatic unrest (M4.7 Aug 2026, >54k quakes) is attributed by Stanford/INGV to hydrothermal/geothermal pressure rather than magma ascent. The 23.3% Kalshi anchor is almost entirely non-informative: a 24-year-dated contract carries a large cost-of-carry discount on NO plus classic long-shot bias in a ~124-contract/day market, and it has been trending down toward fundamentals. I sit slightly above the strict geological base rate (and closer to Forecast 2 than Forecast 1) purely to cover undocumented resolution criteria that might loosely admit a VEI-7 caldera-forming event or a popularly-labeled 'supervolcano', plus tail risk from unmonitored systems — but Forecast 1's 7% exceeds even the brief's loose-definition ceiling of ~4.7% and is unjustified.

Total pipeline time: 234.1s