Tools chosen: polymarket_direct polymarket_related kalshi_related claude_news claude_news gdelt_news wikipedia code_execution

Sub-questions (Fermi decomposition)

Planner reasoning

This is a base-rate question with a critical resolution-mechanics wrinkle: GVP historically averages ~3-6 VEI 4+ eruptions per year (so 'exactly 0' should be near-impossible), but VEI assignments lag substantially, so the GVP 'eruptions by year' table for 2026 as read on 2027-03-31 may still show 0. The main research tasks are pinning the true base rate, measuring GVP's assignment lag (what did the 2025 row show in early 2026?), checking for any 2026 eruptions already large enough, and anchoring on the Polymarket price plus sister markets in the same event.

# Current state As of late August 2026 (per research cutoff), no VEI 4+ eruption has been publicly confirmed or classified by GVP/USGS for 2026. Several volcanoes (Sheveluch, Klyuchevskoy, Lewotobi Laki-laki, Kanlaon) have erupted repeatedly but plume heights/intensity reported so far correspond to VEI 3 or lower. The resolution source (GVP "Eruptions by year" table) is checked as of March 31, 2027 — GVP historically takes many months to over a year to finalize VEI assignments for a calendar year (2025 review only "completed" per an Aug 7, 2026 database update), so a currently-blank 2026 row would not yet reflect true finality. # Timeline of key events - 2024-04: Ruang (Indonesia) VEI 4 eruption — confirmed, prior-year precedent (Wikipedia). - 2025 (year-round): GVP database review of 2025 eruptions ongoing; not "completed" until 2026-08-07 per changelog — confirmed (GVP changelog). - 2026-03-16: Sheveluch (Kamchatka) explosive eruption, ash to ~10.4 km — reported (watchers.news). - 2026-05 (by): Kanlaon (Philippines) had 7 eruptions in 2026, all low intensity (Alert Level 2) — reported (volcanodb.com). - 2026-07-05: Sheveluch high-level eruption, ash to 12.2 km, Aviation Color Code raised to Red — reported (watchers.news). - 2026-08 (through): Klyuchevskoy VAAC advisories, ash plumes 6.1–9.1 km — reported (volcanodiscovery.com). - 2026-08 (through): Lewotobi Laki-laki continues erupting but at modest scale (1–2.5 km ash columns), far below its July 2025 18 km column — reported (techtimes.com). - 2026-08-07: GVP database update (v5.4.0) completes review of 2025 eruptions — confirmed (GVP changelog), illustrating ~7+ month lag even for the prior year. # Event Will there be exactly 0 confirmed VEI 4+ volcanic eruptions worldwide in 2026 (per Smithsonian GVP data, checked March 31, 2027)? # Outcomes to forecast Yes (exactly 0 VEI 4+ in 2026) / No (1 or more VEI 4+ in 2026) # Kalshi market anchor No Kalshi-direct price available for this specific ticker in the research; only a related, non-equivalent Kalshi market found: "Will a supervolcano erupt before 2050?" (KXERUPTSUPER-0-50JAN01) trading at 31% (7d +1%, 30d +8%, low volume ~52 contracts/day). This is NOT a direct comparable (supervolcano ≠ VEI 4+ annual count) — treat as weak signal only. # Sub-question answers 1. **GVP historical VEI4+ annual counts / zero-years** — Research did not retrieve exact year-by-year table values; qualitative sources state VEI4+ frequency "has remained impressively constant for more than a century" and academic estimates put VEI4 at "almost annual" (λ≈1/yr per single-event estimate, though GVP's own average is higher, ~3-6/yr per code_execution model). No explicit zero-year count found in research. 2. **GVP finalization lag / 2025 row status** — GVP completed review of 2025 eruptions only as of 2026-08-07 (GVP changelog), confirming multi-month-to-year-plus lag; current-year rows are provisional and likely undercount VEI4+ until later review. 3. **2026 plausible VEI4+ candidates** — Sheveluch (ash to 12.2 km, July 2026, Red alert) and Klyuchevskoy (ash to 9.1 km, Aug 2026) are the closest candidates but reported plume heights (6-12 km) are generally associated with VEI 3, below the ~15+ km / 0.1 km³ tephra threshold typically signaling VEI 4. No confirmed VEI4+ reported as of late Aug 2026 (claude_news). 4. **Polymarket ladder pricing** — Only this exact-0 market's data was retrieved: 47.5% current price, down from a 73.5% high, 30d trend -16.5%, volume ~$499K. No adjacent bucket (1, 2, 3, 4+) prices were found (polymarket_related returned 0 matches), so ladder coherence cannot be verified. 5. **Kalshi comparable** — Only a loosely related "supervolcano by 2050" market exists (31%); no direct VEI4+ 2026 count market found on Kalshi, so no cross-venue disagreement can be assessed. 6. **Analogous 2025 market performance** — No data found on how a 2025-vintage equivalent market priced or resolved. # Key facts (high-confidence, factual) 1. [GVP/Wikipedia] Ruang, Indonesia had a confirmed VEI 4 eruption in April 2024. 2. [GVP changelog] GVP completed its review of 2025 eruptions only on 2026-08-07 — evidencing lag of 7+ months into the following year. 3. [academic/arxiv] VEI4-equivalent (M4) events occur at "almost annual" base rate; VEI5 ~every 5 years; VEI6 ~every 30 years. 4. [claude_news] As of late Aug 2026, no VEI4+ eruption has been publicly confirmed by GVP/USGS for 2026. 5. [Polymarket] "Exactly 0" contract has fallen from a high of 73.5% to 47.5%, a 16.5-point 30-day decline, suggesting market is pricing in rising likelihood of a VEI4+ confirmation (or reduced confidence in "0" holding) as the year progresses without a fully lagged final row. # Cross-market signals - Kalshi related: Only a "supervolcano by 2050" market (31%, rising) — not directly comparable. - Polymarket: This exact market itself at 47.5%, sharply down from earlier highs (73.5%), high volume (~$499K), suggesting active repricing amid ongoing eruptive activity (Sheveluch, Klyuchevskoy) without confirmed VEI4 classification yet. - Sportsbook implied: None found. # Analyst opinions and speculation - code_execution Poisson/mixture modeling: pure Poisson (λ=3–6, GVP-consistent) gives P(zero VEI4+)≈0.2%–5% — very low. - Blended model incorporating GVP reporting-lag risk (p_lag 5–35%) raises "fair" YES probability to roughly 5%–38%, with a "central" estimate of ~12–21% (λ≈4, p_lag≈10–20%) — meaning the lag/finalization mechanic, not true eruption absence, is the dominant driver of any material YES probability. # Directional lean per outcome - **Yes (0 VEI4+)**: Supported by GVP's slow finalization process (2025 review only completed Aug 2026, implying 2026 row will likely still show fewer/no VEI4+ events by the March 2027 resolution date even if a qualifying eruption occurred) and by the fact no VEI4+ has been confirmed as of Aug 2026. Opposed by strong historical base rate (VEI4 "almost annual") and two currently active high-plume candidates (Sheveluch, Klyuchevskoy) that could still be upgraded upon full review. - **No (≥1 VEI4+)**: Supported by base-rate statistics (λ≈3-6/yr historically) and multiple ongoing large eruptions with red alerts in 2026. Opposed by the resolution mechanic's dependence on GVP finalization timing — a real VEI4+ event might not be classified/confirmed in the table by March 31, 2027, causing the market to resolve YES on a technicality even if a qualifying eruption occurred. # Gaps / unknowns - No confirmed exact GVP year-by-year VEI4+ historical counts (2000-2024) retrieved to compute empirical zero-year frequency. - No data on adjacent Polymarket buckets (1, 2, 3, 4+) to check ladder coherence with the 47.5% "0" price. - No analogous 2025-market resolution/pricing history found. - Uncertain whether Sheveluch/Klyuchevskoy 2026 activity will later be upgraded to VEI4 upon full GVP review (key swing factor). - GDELT news tool failed (SSL errors), limiting most-recent-news coverage beyond Aug 2026. # Calibration anchors - Polymarket "exactly 0" price: 47.5% (down from 73.5% high; strong downward 30d trend), on $499K volume — no Kalshi-direct price exists for this specific ticker. - Poisson/mixture modeling suggests fair value plausibly in 5%-38% range depending on lag assumptions, with 12-21% as a central estimate — notably below the current 47.5% Polymarket price, suggesting the market may be overpricing YES relative to base rates (or pricing in higher lag-risk than the model's central case). - Historical precedent: VEI4+ eruptions confirmed in most recent years (e.g., 2024 Ruang), consistent with "almost annual" frequency claims.

1. Both forecasters anchor heavily on the Polymarket 47.5% price despite the brief's own modeling explicitly flagging that this price appears to overstate fair value (central estimate 12-21%, range 5-38%) — neither forecaster explains why they discount the model-implied ceiling of ~38% only modestly (F1 lands at 43%, above even the model's upper bound) rather than treating 47.5% as a potentially stale/illiquid signal to be substantially discounted. 2. Neither forecast grapples with the critical distinction between "0 confirmed VEI4+ eruptions occurred" versus "0 VEI4+ eruptions appear in the GVP table by March 31, 2027" — the brief notes these are different questions and that the lag mechanic could cause YES to resolve even with a real VEI4 eruption having occurred; both treat this as one blended factor rather than decomposing P(no real VEI4+ occurs) × P(lag delays confirmation | one occurs), which would produce a more defensible joint estimate. 3. Both forecasts effectively ignore the "Gaps/unknowns" caveat that no empirical GVP zero-year frequency was retrieved — they lean on the qualitative "almost annual" (λ≈1) academic estimate but also cite the code-execution model's λ≈3-6/yr (which yields P(zero)≈0.2-5%), a huge discrepancy the brief flags, yet neither forecaster reconciles which λ estimate is more credible or discusses how that choice swings their base-rate component dramatically. 4. Neither forecaster explicitly quantifies how much probability mass should attach to Sheveluch (12.2 km ash, Red alert) or Klyuchevskoy (9.1 km) actually being upgraded to VEI4+ on later review — this is called out in the brief as "the key swing factor" but both treat it as a vague qualitative offset rather than estimating, e.g., a 15-30% chance either gets upgraded. 5. Both converge on similar reasoning and directionally similar numbers (43% vs 32%), but this convergence masks the fact that the Polymarket signal itself is called "weak" and non-Kalshi-direct in the brief — treating it as the "best consensus anchor" without more skepticism risks anchoring bias, especially given the 16.5-point 30-day decline could reflect thin/volatile order flow on $499K volume rather than genuine information updating. 6. Neither forecast considers the tail scenario where GVP's finalization lag extends beyond typical patterns (e.g., if 2026 review isn't even substantially underway by March 2027, mirroring or exceeding the 7+ month lag seen for 2025), which could push YES probability higher than either forecaster's estimate if resolution effectively defaults to "whatever is confirmed in a still largely provisional table."

42% (48% confidence)

Yes: 42% No: 58%

The only direct price signal is Polymarket's $499K-volume 'exactly 0' contract at 47.5% (down 16.5 pts in 30 days), which is a weak but non-trivial anchor. Decomposing as the critique urges: with a realistic GVP VEI4+ rate near 1-2/yr (the λ=3-6 model appears inflated relative to actual GVP year-by-year records like Ruang 2024, Shiveluch 2023, Hunga Tonga 2022), and conditioning on no confirmation through eight months of 2026, P(no real VEI4+) is perhaps 35-45%, while the chance Sheveluch's 12.2 km July event or Klyuchevskoy gets upgraded on review is maybe 20-30%. Adding the documented finalization-lag mechanic (2025 review only completed Aug 2026), a genuine late-2026 VEI4 might not appear in the table by the March 31, 2027 check, which pushes YES up rather than down — so I do not discount all the way to the 12-21% modeling central case. Netting the base-rate pull toward 'No' against the conditioning and lag pulls toward 'Yes', I land just below the market at 42%, close to Forecast 1 but with more explicit weight on the eventual-upgrade risk that favors 'No'.

Total pipeline time: 176.7s