← Back to scans

When will a supervolcano next erupt? — Before Jan 1, 2050

KXERUPTSUPER-0-50JAN01 · Climate and Weather · 2026-08-02
4%
Agent
23%
Market Price
-19.0%
Edge
medium
Confidence
Volume: 33,434
Spread: 5.9c
Days to resolution: 8552
Markets in event: 1
Final Rationale
Under the strict scientific definition (VEI-8, >1,000 km³), the 2026-2050 Poisson probability is ~0.02-0.16%, and every candidate system is at background: Yellowstone shows no significant deformation since Jan 2026, and INGV plus the 2025 Science paper attribute Campi Flegrei's 161.5 cm uplift and M4.4/M4.7 tremors to non-magmatic caldera inflation with no magma migration. Taking the critique's call for explicit decomposition: P(strict VEI-8 rules) ≈ 0.75 × ~0.001 ≈ negligible, plus P(loose/ambiguous rules crediting a VEI-7 or any eruption at a named 'supervolcano' caldera) ≈ 0.25 × ~0.10-0.15 (global VEI-7 recurrence plus a small Campi Flegrei 2030-2034 regime-change catalyst) ≈ 0.03 — totaling roughly 3-4%. The 28.9% Kalshi anchor is discounted heavily but not ignored: a 24-year-dated contract on ~69 contracts/day makes NO-side selling capital-inefficient, so the 71% NO price mostly reflects carry/time-value and headline attention rather than information; even informed traders monitoring INGV feeds are pricing unrest, not eruption. I therefore land between the two forecasts, closer to the lower one, at 4% Yes.
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 $t{=}0$.
$n = 31$ follow-ups
Re-scan Context
This market has been scanned before. Previous predictions:
DatePredictedMarket PriceConfidence
2026-07-15 3% 24% 80%
2026-07-07 4% 30% 80%
2026-06-30 2% 31% 80%
1. Planner Opus · Fermi decomposition + tool plan
Tools chosen: kalshi_direct kalshi_related polymarket_related wikipedia claude_news gdelt_news code_execution
Sub-questions (Fermi decomposition)
  1. What is the current Kalshi YES price and volume for KXERUPTSUPER-0-50JAN01, and how has it moved over the last 90 days?
  2. How does Kalshi define a 'supervolcano eruption' for resolution — VEI 8 magnitude, a specific list of volcanoes (Yellowstone, Toba, Campi Flegrei, Taupo), or an authoritative source like the Smithsonian GVP?
  3. What is the empirical recurrence interval for VEI 8 eruptions in the geologic record (eruptions per 100,000 years), and what Poisson probability does that imply over the 24-year window 2026-2050?
  4. Are there any currently elevated/unrest signals at candidate supervolcanoes (Campi Flegrei bradyseism, Yellowstone, Taupo, Long Valley, Aira) that would materially raise short-horizon odds?
  5. Do scientists assign any published annual probability to a VEI 8 eruption, and what is the largest eruption (max VEI) recorded in the past 200 years for calibration?
  6. Is there an equivalent Polymarket or other Kalshi market on supervolcano/large eruption that gives an independent price?
Planner reasoning
This is a very-long-horizon base-rate question: VEI-8 'supervolcano' eruptions occur roughly once per 15,000-100,000 years globally, so a 25-year window implies a probability well under 1%. The key research tasks are (a) anchoring on the Kalshi price (which for such far-dated novelty markets is often mispriced high due to lottery-ticket bias and no-carry), (b) pinning down the resolution definition (VEI 8? which volcanoes count?), and (c) computing the Poisson base rate from published eruption recurrence intervals.
2. Research Parallel tool calls
ToolStatusTimeItemsSummary
kalshi_direct OK 0.8s 1 ## This Market's Kalshi Data **Before Jan 1, 2050** (KXERUPTSUPER-0-50JAN01) - Current price (probability): 28.90% - 7-day price change: +4.90% - 30-day price change: -1.80% - Average daily volume: 69 contracts - Price range: 16.80% - 32.90% - Data points: 76 days
kalshi_related OK 7.4s 2 2 related markets / summaries. series KXERUPTSUPER: 0 markets (skipped 1 no-signal) | keyword 'supervolcano': ok | keyword 'volcano eruption': ok | keyword 'Yellowstone': no matches
polymarket_related OK 7.4s 0 Scanned 100 active Polymarket markets, kept 0 matches. keyword 'supervolcano': 0 markets | keyword 'volcano': 0 markets | keyword 'Yellowstone eruption': 0 markets
wikipedia OK 7.4s 6 Fetched 6 Wikipedia entries (0 missing pages).
claude_news OK 22.7s 16 ## Findings **Kalshi Market Definition & Pricing:** - The Kalshi contract "KXERUPTSUPER-0" opened Jul 18, 2025, and will close at 2050-01-01T04:59:00Z and resolves to "No" if no eruption occurs by Dec 31, 2049, 11:59 pm ET . (octagonai.co) - As of early 2026, the market currently indicates an imp
gdelt_news OK 173.4s 10 GDELT: 10 articles across 3 queries (lookback=90d). 'supervolcano eruption warning': error GDELT rate-limited after retries (429) | 'Campi Flegrei unrest': error GDELT rate-limited after retries (429) | 'Yellowstone caldera activity': 10 hits
code_execution OK 22.2s 0 ## Poisson Probability Analysis: VEI-8 Supervolcano Eruption Before 2050 **Model setup:** P(≥1 event) = 1 − e^(−λt), where λ = 1/recurrence interval, t = 24 years (2026–2050). ### Probability Table (VEI-8, 24-year window) | Recurrence Assumption | λ (24yr) | P(≥1) % | P(≥1) bps | |---|---|---|---
3. Evidence Brief Sonnet · 7594 chars
# Current state The Kalshi market KXERUPTSUPER-0-50JAN01 currently prices YES at 28.90%, up from a 90-day low of 16.80% and near its 90-day high of 32.90% — a substantial premium over the base-rate physics of VEI-8 eruptions (Poisson base rate ≈0.02%-0.16% over 24 years). Resolution requires an actual eruption (implicitly VEI-8 "supervolcano" class) before Jan 1, 2050; no candidate volcano (Yellowstone, Campi Flegrei, Long Valley, Taupō, Aira) shows imminent-eruption signals as of mid/late-2026. # Timeline of key events - 2025-07-18: Kalshi market KXERUPTSUPER opens (confirmed, octagonai.co via claude_news). - 2025-09: Science paper/AI study finds Campi Flegrei uplift driven by non-magmatic caldera inflation, no evidence of magma migration (confirmed, phys.org). - ~Early 2026: Market YES price reported around 11-20% (varying by source/timing) per octagonai.co/antlersarch.com (reported, conflicts with current kalshi_direct 28.90% — later price likely reflects subsequent uplift/quake activity). - 2026-01: Cumulative ground uplift at Campi Flegrei's Rione Terra GNSS station reaches 161.5 cm; series of M4+ quakes through 2023-2025 (confirmed, sciencedirect.com). - 2026-05-01: USGS notice on Yellowstone — background seismicity, 97 quakes in April, largest M2.5, no significant deformation since Jan 2026 (confirmed, USGS). - 2026-05-21: M4.4 earthquake and swarm at Campi Flegrei, depth ~3km (confirmed, watchers.news). - 2026-07-16: Earthquake near Yellowstone reported widely by syndicated local news (confirmed occurrence, but framed sensationally — no eruption/unrest escalation implied by USGS). - 2026-07-31: M4.7 tremor at Campi Flegrei — strongest in Campania in 40 years, injures 21; INGV states no signs of volcanic eruption (confirmed, travelerstoday.com/INGV). - 2026 (undated study): Projection that Campi Flegrei may reach a "critical mechanical threshold" (possible regime change, not necessarily eruption) between 2030-2034 (speculative/reported). # Event Will a supervolcano (VEI-8 class eruption) occur before Jan 1, 2050? # Outcomes to forecast - Yes (a supervolcano erupts before 2050) - No (it does not) # Kalshi market anchor Current YES price: **28.90%** (kalshi_direct, primary anchor). 7-day change +4.90%, 30-day change -1.80%. 90-day range 16.80%-32.90%, avg daily volume only 69 contracts (thin market). Secondary source (claude_news/octagonai.co) reports an inconsistent ~14% (11-20% range) "early 2026" figure — likely a stale/different snapshot; the live kalshi_direct 28.90% is authoritative and reflects recent uptick, plausibly driven by Campi Flegrei seismic headlines (May/July 2026 tremors). # Sub-question answers 1. **Current price/volume/90-day trend?** — 28.90% now, +4.9% over 7 days, -1.8% over 30 days, range 16.8-32.9%, thin volume (~69 contracts/day) [kalshi_direct]. 2. **Resolution definition?** — No explicit rules text provided; secondary source states market resolves "No" if no eruption occurs by Dec 31, 2049 [octagonai.co via claude_news]. No confirmed VEI-8-specific or named-volcano list criterion found in raw rules; Wikipedia confirms the standard scientific definition of "supervolcano" = VEI-8 eruption [Wikipedia/Supervolcano]. 3. **Recurrence interval / Poisson probability?** — VEI-8 recurrence estimates range ~15,000-100,000 years; implied 24-year (2026-2050) Poisson probability ranges **0.024%-0.16%** [code_execution]. Most recent VEI-8 event: Taupō Oruanui eruption ~25,600 years ago [Wikipedia]. 4. **Elevated unrest signals?** — Campi Flegrei shows most active unrest (M4.4 May 2026, M4.7 July 2026, 161.5cm cumulative uplift by Jan 2026), but INGV/2025 Science paper attribute this to non-magmatic caldera inflation, not imminent eruption [claude_news/phys.org/travelerstoday.com]. Yellowstone at background/NORMAL levels, no significant deformation since Jan 2026 [USGS]. Long Valley and Taupō show no acute unrest reported. 5. **Published annual probability / max VEI in past 200 years?** — USGS estimates Yellowstone supereruption annual probability ~1-in-730,000 [volcanodb.com/USGS]. Largest recorded eruption in past 200 years is far below VEI-8 (historical eruptions like Tambora 1815 reached VEI-7); no VEI-8 eruption recorded in past 200 years [Wikipedia/List of largest volcanic eruptions]. 6. **Equivalent Polymarket/other market?** — None found; Polymarket scan returned zero matching markets for supervolcano/volcano/Yellowstone keywords [polymarket_related]. # Key facts (high-confidence, factual) 1. [Wikipedia] Supervolcano = VEI-8 eruption (>1,000 km³ ejecta); last VEI-8 was Taupō ~25,600 years ago. 2. [USGS/volcanodb.com] Yellowstone annual supereruption probability ≈1/730,000; not "overdue" (avg. interval ~725,000 yrs between major cycles). 3. [claude_news/phys.org] Campi Flegrei uplift/seismicity attributed to non-magmatic caldera inflation; no magma migration evidence (2025 Science paper). 4. [USGS] Yellowstone background seismicity/deformation normal as of May 2026; classified NORMAL/GREEN. 5. [code_execution] Base-rate Poisson probability for VEI-8 by 2050: ~0.02%-0.16%, several orders of magnitude below current Kalshi price. # Cross-market signals - Kalshi related: only the same KXERUPTSUPER-0 contract found under "supervolcano"/"volcano eruption" keywords; no separate calibration market. - Polymarket: no matching markets exist (0/100 scanned). - Sportsbook implied: N/A (not applicable to this event type). # Analyst opinions and speculation - octagonai.co / antlersarch.com frame the market as pricing tail risk/attention premium rather than pure geologic base rates. - code_execution analysis concludes current-style low-cent pricing would be 12x-200x overpriced under any standard VEI-8 recurrence assumption, unless resolution criteria loosely include VEI-7 events (e.g., Campi Flegrei-scale), in which case pricing becomes closer to fair. - One 2026 projection speculates Campi Flegrei could reach a "critical mechanical threshold" 2030-2034 — explicitly not necessarily eruption, but a regime change (fracturing/fumarole increase) [travelerstoday.com] — speculative, unconfirmed. # Directional lean per outcome - **Yes**: Supported by elevated Campi Flegrei seismicity/uplift trend and rising Kalshi price (+4.9% in 7 days), reflecting recent headline-driven attention; opposed by strong geologic consensus (no VEI-8 in 25,600 years, USGS Yellowstone probability ~1/730,000, INGV explicitly denies eruption signs at Campi Flegrei) and definitional mismatch (Campi Flegrei's largest known eruption is VEI-7, not VEI-8, so it may not even qualify as a "supervolcano"). - **No**: Overwhelmingly favored by base-rate physics, current monitoring (all major sites NORMAL/background), and historical record (zero VEI-8 events in the past 200+ years globally). # Gaps / unknowns - Exact Kalshi resolution rules/source (e.g., Smithsonian GVP, VEI-8 requirement, or named volcano list) not confirmed from primary rules text — rules field was empty; only secondary paraphrase available. - Discrepancy between claude_news (~14%, early 2026) and kalshi_direct (28.90%, current) unresolved — likely reflects price movement/staleness, not data error. - No independent Polymarket price for triangulation. # Calibration anchors - Kalshi current YES price: **28.90%** (anchor). - Base-rate Poisson probability under strict VEI-8 definition: ~0.02%-0.16% over 2026-2050. - USGS Yellowstone annual probability: ~1/730,000 (~0.0033% over 24 years). - No VEI-8 eruption has occurred in recorded/historical times (past 200 years); last VEI-8 was ~25,600 years ago (Taupō).
4. Forecasters Opus + GPT · in parallel
claude-opus-5
medium
Yes 7%
No 93%
Under the standard scientific definition (VEI-8, >1,000 km³), the Poisson base rate for an eruption in 2026-2050 is ~0.02%-0.16%, and USGS puts Yellowstone at ~1/730,000 per year; no VEI-8 has occurred in ~25,600 years. All candidate systems are at background/NORMAL: Yellowstone shows no significant deformation since Jan 2026, and INGV/the 2025 Science paper attribute Campi Flegrei's 161.5 cm uplift and M4.4/M4.7 tremors to non-magmatic caldera inflation with no magma migration. The Kalshi anchor of 28.9% is on an extremely thin (~69 contracts/day), 24-year-dated contract where capital carrying costs make NO-side selling unattractive, so the price carries a large structural/attention premium rather than genuine information. I shade meaningfully below the anchor but leave headroom above pure base rate for ambiguous resolution criteria (rules text unconfirmed; a VEI-7-scale or Campi Flegrei event might qualify) and the speculative 2030-2034 'critical mechanical threshold' projection.
gpt-5.5
medium
Yes 2%
No 98%
Kalshi’s live YES anchor is 28.90%, but the brief indicates this is a thin market and several orders of magnitude above the strict VEI-8 Poisson base-rate estimate of roughly 0.02%-0.16% before 2050. Current unrest evidence points only weakly toward Yes: Campi Flegrei has notable uplift and M4+ quakes, but studies and INGV statements report non-magmatic inflation and no signs of eruption, while Yellowstone remains at background/NORMAL levels. The recurrence and historical evidence point strongly toward No because no VEI-8 has occurred in recorded history, the last was ~25,600 years ago, and USGS estimates Yellowstone’s annual supereruption probability around 1-in-730,000. I leave a small premium above the strict base rate for resolution-rule ambiguity and unforeseen global volcanic unrest, but far below the Kalshi price.
5. Devil's Advocate Sonnet · red-team critique
1. Both forecasters anchor heavily on the strict VEI-8 base rate but diverge significantly (7% vs 2.5%) on how much weight to give the resolution-ambiguity scenario (VEI-7/Campi Flegrei qualifying); neither explicitly assigns a probability to "rules are loose enough to include VEI-7" versus "strict VEI-8 only," which is the single biggest driver of any deviation from the ~0.02-0.16% base rate and deserves explicit decomposition rather than an intuitive "shade" adjustment. 2. Both forecasts land far below the Kalshi anchor (28.9%) with little justification for why the market is so persistently mispriced over a sustained period (90-day range 16.8-32.9%, recent +4.9% 7-day move) rather than treating this as a possible signal of information the forecasters lack — e.g., a genuine loosening of the resolution criteria that the brief flags as unconfirmed but plausible, which could rationally justify pricing near 20-30%. 3. Neither forecaster incorporates the 2030-2034 "critical mechanical threshold" projection with any explicit weight — it's mentioned in both rationales as a caveat but not decomposed into a scenario probability, even though it is one of the only forward-looking (post-2026) catalysts explicitly named in the brief and could materially raise near-term Campi Flegrei-linked resolution risk if rules include lower-VEI candidate eruptions. 4. Both forecasts assume the "thin market / structural premium" explanation for the Kalshi price gap without considering the alternative that thin markets with directional momentum (+4.9% in 7 days, near 90-day high) can also reflect real-time absorption of unrest headlines by informed traders monitoring INGV/USGS feeds — this alternative should lower confidence in dismissing the price as pure noise. 5. Neither model quantifies a scenario where a non-VEI-8 but headline-dominant eruption (e.g., a large Campi Flegrei event) causes the market to resolve YES on a technicality or through ambiguous adjudication, which is a tail risk distinct from both the "strict VEI-8" and "normal geologic base rate" framings that dominate both rationales. 6. Forecast 2's 2.5% is nearly 20x the upper base-rate estimate (0.16%) while Forecast 1's 7% is over 40x that upper bound — this wide spread between two forecasters using the same evidence highlights unresolved disagreement about how much "resolution-ambiguity premium" is appropriate, and neither justifies their specific multiplier with a transparent framework (e.g., P(loose rules) × P(any eruption under loose rules by 2050)).
6. Aggregator Opus · final calibrated forecast
4% (medium confidence)
Yes: 4% No: 96%
Under the strict scientific definition (VEI-8, >1,000 km³), the 2026-2050 Poisson probability is ~0.02-0.16%, and every candidate system is at background: Yellowstone shows no significant deformation since Jan 2026, and INGV plus the 2025 Science paper attribute Campi Flegrei's 161.5 cm uplift and M4.4/M4.7 tremors to non-magmatic caldera inflation with no magma migration. Taking the critique's call for explicit decomposition: P(strict VEI-8 rules) ≈ 0.75 × ~0.001 ≈ negligible, plus P(loose/ambiguous rules crediting a VEI-7 or any eruption at a named 'supervolcano' caldera) ≈ 0.25 × ~0.10-0.15 (global VEI-7 recurrence plus a small Campi Flegrei 2030-2034 regime-change catalyst) ≈ 0.03 — totaling roughly 3-4%. The 28.9% Kalshi anchor is discounted heavily but not ignored: a 24-year-dated contract on ~69 contracts/day makes NO-side selling capital-inefficient, so the 71% NO price mostly reflects carry/time-value and headline attention rather than information; even informed traders monitoring INGV feeds are pricing unrest, not eruption. I therefore land between the two forecasts, closer to the lower one, at 4% Yes.
Pipeline Timing
Total pipeline time: 273.5s
Per-tool research timings shown in the Research section above.