← Back to scans

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

KXERUPTSUPER-0-50JAN01 · Climate and Weather · 2026-06-23
3%
Agent
31%
Market Price
-27.7%
Edge
65%
Confidence
Volume: 31,612
Spread: 0.1c
Days to resolution: 8593
Markets in event: 1
Final Rationale
Scientific base rate for a strict VEI 8 eruption in 25 years is ~0.03-0.05%; no monitored true supervolcano shows imminent precursors. The 30.7% Kalshi price is a thin, retail-driven mispricing. I lift modestly above base rate to ~3% to account for definitional ambiguity (whether VEI 7 events like Campi Flegrei could count) and unknown-unknown tails, consistent with both forecasters.
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 $t{=}0$.
$n = 22$ follow-ups
Re-scan Context
This market has been scanned before. Previous predictions:
DatePredictedMarket PriceConfidence
2026-06-15 4% 28% 65%
2026-05-15 6% 24% 70%
2026-05-08 6% 19% 70%
1. Planner Opus · Fermi decomposition + tool plan
Tools chosen: kalshi_direct wikipedia claude_news code_execution kalshi_related polymarket_related
Sub-questions (Fermi decomposition)
  1. What is the historical frequency/base rate of VEI 8 supervolcano eruptions (how many per 100,000 years)?
  2. Given the base rate, what is the probability of at least one such eruption in the ~25-year window through 2050?
  3. Are there any current signs of imminent activity at known supervolcanoes (Yellowstone, Campi Flegrei, Toba, etc.)?
  4. What definition of 'supervolcano eruption' does the market use, and how strict is the VEI threshold?
Planner reasoning
This is a very long-horizon, low-base-rate geological question with no near-term catalyst. The key drivers are the historical eruption frequency of supervolcanoes (VEI 8 events) and any current monitoring signals from known supervolcanoes. The Kalshi market price anchors expectations.
2. Research Parallel tool calls
ToolStatusTimeItemsSummary
kalshi_direct OK 0.2s 1 ## This Market's Kalshi Data **Before Jan 1, 2050** (KXERUPTSUPER-0-50JAN01) - Current price (probability): 30.70% - 7-day price change: +7.70% - 30-day price change: +1.70% - Average daily volume: 106 contracts - Price range: 14.00% - 32.00% - Data points: 86 days
wikipedia OK 4.2s 4 Fetched 4 Wikipedia entries (0 missing pages).
claude_news OK 42.6s 13 Here are the key findings for the forecasting question: --- **Campi Flegrei (Italy) — Highest Near-Term Concern** - The volcano near Naples is shaking the ground in a way that scientists say it hasn't for centuries, posing risks for hundreds of thousands of people living in the 8-mile-wide crate
code_execution OK 37.6s 1 ## Key Findings: Probability of a VEI 8 Supervolcano Eruption Before Jan 1, 2050 - **Assessment window is only 25 years** (2025–2050), an extraordinarily short span relative to the ~50,000–100,000-year average recurrence interval of VEI 8 eruptions. - **At the higher-frequency bound (1 per 50,000
kalshi_related OK 4.1s 1 1 related markets / summaries. series KXERUPTSUPER: 0 markets (skipped 1 no-signal) | keyword 'volcano': ok | keyword 'eruption': no matches | keyword 'Yellowstone': no matches
polymarket_related OK 4.1s 0 Scanned 100 active Polymarket markets, kept 0 matches. keyword 'volcano': 0 markets | keyword 'supervolcano': 0 markets | keyword 'eruption': 0 markets
3. Evidence Brief Sonnet · 4814 chars
# Event Will a supervolcano (VEI 8) erupt before Jan 1, 2050? # Outcomes to forecast - **Yes** — a supervolcano erupts before Jan 1, 2050 - **No** — no supervolcano erupts before Jan 1, 2050 # Kalshi market anchor **KXERUPTSUPER-0-50JAN01 current YES price: 30.70%** - 7-day change: **+7.70%** (significant recent spike) - 30-day change: +1.70% - Price range over 86 days: 14.00%–32.00% - Avg daily volume: 106 contracts (thin market) - ⚠️ **This price is dramatically inconsistent with the scientific base rate (~0.03–0.05%). The market appears to be mispriced, likely reflecting Campi Flegrei news driving retail speculation.** # Sub-question answers 1. **Historical frequency of VEI 8 eruptions** — Approximately 1 per 50,000–100,000 years using commonly cited estimates; empirically, ~47 events over ~36M years ≈ 1 per 766,000 years [LaMEVE database via code_execution]. Most recent VEI 8 was Taupō ~25,600 years ago [Wikipedia]. 2. **Probability in 25-year window** — Poisson model yields **0.025%–0.050%** depending on assumed recurrence rate (0.0033% using empirical LaMEVE rate). Consensus: ~0.036%. Would need a ~500–1,000-year window to reach even 1% probability [code_execution]. 3. **Current signs of imminent activity** — Campi Flegrei shows the most unrest: ~140 cm ground uplift since 2005, seismicity 47–70 events/week (Feb–Mar 2026), M4.6 quake Jun 2025 (strongest since monitoring began), elevated SO₂/CO₂ [VolcanoDB, NPR]. However: (a) Campi Flegrei's largest known eruptions are VEI 7, not VEI 8 [Wikipedia]; (b) Italian Civil Protection remains at **yellow alert only**; (c) magma reservoir would need 2–3 more decades to reach size of the 1538 eruption [Nature Communications, Jan 2026]. Yellowstone: background levels, no imminent eruption [USGS, 2026]. Toba: low near-term likelihood [Brighter Side]. 4. **Market definition of 'supervolcano eruption'** — No explicit rules stated. Standard definition per Wikipedia and VEI scale: VEI 8, ejecta > 1,000 km³. Campi Flegrei does NOT qualify even if it erupts (max historical VEI 7). Market rules are silent on whether VEI 7 would count. # Key facts (high-confidence, factual) 1. [Wikipedia] Most recent VEI 8 eruption: Taupō Volcano, ~25,600 years ago 2. [Wikipedia] Campi Flegrei's largest eruptions are VEI 7 — it is **not** a true supervolcano by strict definition 3. [USGS, 2026] Yellowstone last magmatic eruption ~70,000 years ago; no imminent eruption detected; magma is not currently eruptible 4. [code_execution] Poisson P(VEI 8 in 25 yrs) ≈ 0.025%–0.050% at standard recurrence rates 5. [Nature Communications, Jan 2026] Campi Flegrei reservoir needs 2–3 more decades to grow to pre-1538 eruption size # Cross-market signals - **Kalshi related:** Only this market found; no corroborating volcano/eruption markets - **Polymarket:** No matching markets found - **Sportsbook:** None applicable # Analyst opinions and speculation - [VolcanoDB/arXiv 2026] Campi Flegrei may reach "critical mechanical threshold" 2030–2034, but authors explicitly state this ≠ eruption certainty - [Cowboy State Daily, Sep 2025] Yellowstone's next eruption likely a lava flow, not catastrophic explosion - [Brighter Side] Aira Caldera (Japan) and Taupō Zone rated most likely to erupt among supervolcanoes - The +7.7% weekly price spike on Kalshi almost certainly reflects Campi Flegrei media coverage — but Campi Flegrei cannot produce a VEI 8 eruption # Directional lean per outcome - **Yes (supervolcano erupts by 2050):** - Supporting: Campi Flegrei unrest is real and accelerating; multiple supervolcanoes show ongoing geological activity - Opposing: Base rate ≈ 0.03–0.05%; no scientific body forecasts VEI 8 eruption; Campi Flegrei doesn't even qualify; Yellowstone at background; Toba low near-term; 25-year window is geologically negligible - **No:** - Supporting: Overwhelming — base rate, absence of precursors at true VEI 8 systems, scientific consensus, monitoring agencies all at low/normal alert - Opposing: Unknown unknowns; definition ambiguity could theoretically expand resolution criteria # Gaps / unknowns - Market resolution rules are silent: would a VEI 7 count? If so, Campi Flegrei becomes more relevant (~2–3% base rate for VEI 7 in 25 years) - Thin market (106 contracts/day) makes price highly susceptible to noise/media cycles - Definition ambiguity is the single largest source of forecast uncertainty # Calibration anchors - **Kalshi current YES price: 30.70%** — treat as a strong **fade signal** given the science - **Scientific base rate: ~0.03–0.05%** for strict VEI 8 in 25 years - **If VEI 7 counts:** ~2–3% probability (still far below 30%) - **Reasonable forecast:** 1–3% Yes, accounting for definition ambiguity and tail unknowns — the Kalshi price appears to overstate probability by **10–30×**
4. Forecasters Opus + GPT-5.5 in parallel
claude-opus-4-8
high
Yes 3%
No 97%
Scientific base rate for a true VEI 8 eruption in a 25-year window is ~0.03–0.05%, and no monitored supervolcano shows imminent VEI 8 precursors (Yellowstone at background, Campi Flegrei only capable of VEI 7). The Kalshi YES price of 30.7% appears badly mispriced, driven by retail speculation on Campi Flegrei media. I lean modestly above pure base rate (3%) to account for definition ambiguity (whether VEI 7 might count) and tail unknowns.
gpt-5.5
medium
Yes 2%
No 98%
Kalshi’s YES anchor is 30.70%, but the brief explicitly notes this is a thin, likely retail-driven market and dramatically inconsistent with the scientific VEI 8 base rate. Historical frequency points toward No because standard recurrence estimates imply only about 0.025%–0.050% probability in a 25-year window, with the last VEI 8 around 25,600 years ago. Current unrest at Campi Flegrei points only weakly toward Yes because it is geologically active, but its known maximum eruptions are VEI 7, Italian authorities remain at yellow alert, and true VEI 8 systems like Yellowstone show no imminent signs. I assign a small premium above the strict base rate for definition ambiguity and unknown tails, but still strongly favor No.
5. Devil's Advocate Sonnet · red-team critique
1. Both forecasters correctly fade the Kalshi price but neither adequately interrogates whether the market's resolution criteria might expand "supervolcano" beyond strict VEI 8. The brief flags this as "the single largest source of forecast uncertainty" — if VEI 7 counts, the brief explicitly estimates ~2–3% probability, yet both forecasters land at 2.5–3% without clearly attributing that elevation to the definitional ambiguity rather than treating it as a minor tail adjustment. 2. Neither forecaster engages with the Aira Caldera (Japan) and Taupō Zone being rated "most likely to erupt among supervolcanoes," nor the Brighter Side/analyst reference to these systems. Given that Taupō is a genuine VEI 8 candidate system and its zone is flagged as active, briefly dismissing all non-Campi systems as "background" without explicit treatment of these named candidates is an underweighting of the brief's own evidence. 3. Both forecasters treat the +7.7% weekly Kalshi price spike purely as a fade signal driven by retail noise, but neither considers the possibility that sophisticated market participants have already incorporated the definitional ambiguity (VEI 7 counting) into the price, which could partially justify a higher YES probability than the strict scientific base rate alone implies. 4. The forecasters' probability estimates (2.5–3%) are numerically indistinguishable from the brief's "if VEI 7 counts" estimate of ~2–3%, yet both frame their reasoning primarily around VEI 8 base rates (~0.03–0.05%). This creates an internal inconsistency: their final numbers implicitly assume VEI 7 might count, but their stated rationales emphasize VEI 8 strictness without explicitly resolving how much weight to give the definitional ambiguity. 5. Neither forecaster quantifies or explicitly models the "unknown unknowns" tail — i.e., a sudden undetected magmatic intrusion at a system not currently under heightened monitoring (e.g., Toba, Long Valley). The brief notes Toba has "low near-term likelihood" but doesn't rule it out, and a 25-year window across all global supervolcanic systems collectively warrants at least a brief systematic treatment rather than a passing mention.
6. Aggregator Opus · final calibrated forecast
3% (65% confidence)
Yes: 3% No: 97%
Scientific base rate for a strict VEI 8 eruption in 25 years is ~0.03-0.05%; no monitored true supervolcano shows imminent precursors. The 30.7% Kalshi price is a thin, retail-driven mispricing. I lift modestly above base rate to ~3% to account for definitional ambiguity (whether VEI 7 events like Campi Flegrei could count) and unknown-unknown tails, consistent with both forecasters.
Pipeline Timing
Total pipeline time: 111.3s
Per-tool research timings shown in the Research section above.