Scan: Will AfD win the next election to the German Bundestag?

1. Planner Fable 5.1 · Fermi decomposition + tool plan

Tools chosen: kalshi_direct kalshi_related polymarket_related claude_news claude_news article_search gdelt_news wikipedia code_execution

Sub-questions (Fermi decomposition)

What is the current Kalshi YES price and volume for AfD winning the next Bundestag election, and how has it trended?
What are AfD's and CDU/CSU's current national polling averages (INSA, Forsa, YouGov, Politbarometer), and is AfD leading or tied for first place?
How likely is a snap Bundestag election before 2029 given the stability of the Merz-led CDU/CSU-SPD coalition (approval, budget fights, coalition conflicts)?
What is the status of any Bundesverfassungsschutz classification or potential party-ban (Parteiverbot) proceedings against AfD that could prevent it from competing or winning?
How does the market define 'win' (plurality of seats/second votes), and historically how often has the party leading polls ~3-4 years out actually won the German federal election?
What does Polymarket price for AfD winning the next German election, and does it diverge from Kalshi?
Are there structural factors (Brandmauer, firewall erosion, CDU cooperation, economic conditions, migration salience) that could shift AfD's vote share upward or downward before the election?

Planner reasoning

The market asks whether AfD will win (presumably a plurality of seats/votes) at the next Bundestag election, likely in 2029 unless the Merz coalition collapses and triggers a snap vote. Key drivers are current polling of AfD vs CDU/CSU, the stability of the governing coalition, the risk of a party ban proceeding, and how predictive polls 3-4 years out are for German elections. Kalshi's own price is the primary anchor; Polymarket offers an independent crowd.

2. Research Parallel tool calls

tool	status	time	items	summary
kalshi_direct	OK	0.2s	1	## This Market's Kalshi Data AfD (KXGERELECTION-29MAR25-AFD) - Current price (probability): 87.00% - 7-day price change: +5.00% - 30-day price change: +39.00% - Average daily volume: 526 contracts - Price range: 48.00% - 87.00% - Data points: 22 days
kalshi_related	OK	3.3s	1	1 related markets / summaries. series KXGERELECTION: 0 markets (skipped 7 no-signal) \| keyword 'German election': ok \| keyword 'AfD': no matches \| keyword 'Bundestag': no matches
polymarket_related	OK	3.3s	0	Scanned 100 active Polymarket markets, kept 0 matches. keyword 'AfD': 0 markets \| keyword 'German election': 0 markets \| keyword 'Bundestag': 0 markets \| keyword 'Germany chancellor': 0 markets
claude_news	OK	34.9s	18	Based on current polling data (as of September 2026): - AfD leads national polls at ~27-28.8%, clearly ahead of CDU/CSU. The September 2026 ARD-DeutschlandTREND shows the AfD remains the strongest force at 27 percent (-1), while the CDU/CSU holds steady at 21 percent . A poll average from Sept
claude_news	OK	31.7s	26	## Findings: AfD Ban Efforts & Brandmauer Status (as of mid-September 2026) Verfassungsschutz classification — currently blocked by courts - The Federal Office for the Protection of the Constitution (BfV) classified the AfD as a confirmed right-wing extremist organization ("gesichert rechtsext
article_search	OK	3.8s	16	Fetched 16 unique articles across 4 queries (lookback=60d). 'AfD polls lead CDU Germany': 10 articles \| 'Merz coalition crisis snap election Germany': 10 articles \| 'AfD ban Bundesverfassungsgericht': 10 articles \| 'next Bundestag election 2029 AfD': 10 articles
gdelt_news	OK	115.3s	30	GDELT: 30 articles across 3 queries (lookback=30d). 'AfD Umfrage Union': 10 hits \| 'AfD Verbot Bundesverfassungsgericht': 10 hits \| 'Koalition Merz Neuwahl': 10 hits
wikipedia	OK	3.3s	4	Fetched 4 Wikipedia entries (0 missing pages).
code_execution	OK	56.0s	0	## Findings Historical base rate (1998–2025, 8 elections): - The party leading national polls ~3–4 years before election day went on to actually win the plurality in 4/8 cases (50%): hits in 2002, 2009, 2013, 2017; misses in 1998, 2005, 2021, 2025. - Interpretation: over a 3.5-year horizon,

3. Evidence Brief Sonnet 5 · 7450 chars

# Current state As of September 2026, AfD leads all national polls by a wide margin (~27-29%) over CDU/CSU (~18-21%), a lead that has widened since January 2026 and been confirmed by a landslide AfD state-election win in Saxony-Anhalt (43.8%) on 2026-09-06. The next *regular* Bundestag election is due by 2029-03-25, but coalition-collapse speculation is elevated; no snap election has been called and Chancellor Merz has publicly ruled out the minority-government/confidence-vote path that could trigger one. # Timeline of key events - 2025-02-23: 21st Bundestag election (snap); CDU/CSU wins with 28.5%, AfD second with 20.8% (confirmed, Wikipedia). - 2025-05: BfV classifies AfD as "confirmed right-wing extremist" (confirmed). - 2026-01: GMS poll shows AfD's largest-ever lead over CDU/CSU at 27% (reported, Brussels Signal). - 2026-02-26: Cologne Administrative Court (VG Köln) blocks BfV from publicly using "confirmed extremist" label pending main proceedings; no ruling timeline set (confirmed). - 2026-05-05: Merz rules out minority government/snap election, citing AfD (confirmed, Bloomberg). - 2026-06: GFF publishes 1,500-page legal opinion favoring a ban procedure; legal professionals' open letter passes 1,000 signatures (reported). - 2026-08: ARD-Deutschlandtrend: SPD at 12%, Union+SPD combined 33%; "Kanzlerdämmerung" narrative emerges (reported). - 2026-09-06: AfD wins Saxony-Anhalt state election in a landslide (43.8% vs CDU 17.2%); confirmed by exit polls and multiple outlets. - 2026-09-08/15: National poll averages put AfD at 27-28.8%, CDU/CSU 18-21% (ARD-DeutschlandTREND, YouGov, INSA aggregate) (confirmed). - 2026-09-20 (upcoming): Mecklenburg-Vorpommern state election; Forsa pre-election poll shows AfD 37% vs CDU 13% (reported/forward-looking). # Event Will AfD win (finish first by seats/vote share) the next Bundestag election, resolving via Kalshi ticker KXGERELECTION-29MAR25-AFD. # Outcomes to forecast - Yes (AfD wins) - No (AfD does not win) # Kalshi market anchor Current YES price: **87%** — up +5pts over 7 days and +39pts over 30 days, off a 22-day range of 48-87%. Average daily volume ~526 contracts. This sharp, sustained upward trend directly tracks the AfD state-election landslide (Sept 6) and record national polling gap. This 87% is the consensus benchmark to beat. # Sub-question answers 1. **Kalshi price/trend** — 87% YES, +5pts (7d), +39pts (30d); trending strongly upward, moderate but consistent volume (~526/day). [kalshi_direct] 2. **AfD/CDU polling** — AfD leads clearly at 27-29% (INSA, YouGov, ARD-Deutschlandtrend, GMS aggregate ~28.8%); CDU/CSU has fallen to a historic low of 18-21%. AfD has held first place continuously since ~2025 and the gap has widened, not narrowed. [claude_news, article_search] 3. **Snap election likelihood** — Elevated speculation ("Kanzlerdämmerung," "Operation Chancellor Removal") amid coalition dysfunction (only 15% approve of government; Merz approval 13%), but Merz explicitly ruled out minority government/confidence-vote route in May 2026. Atlantic Council assesses collapse is "not insignificant" but coalition could persist to 2029; no snap election confirmed as of Sept 2026. [claude_news, Atlantic Council, Wikipedia] 4. **Ban/classification status** — BfV's "confirmed right-wing extremist" classification is currently judicially blocked (VG Köln, Feb 2026), pending unresolved main proceedings with no timeline. No federal Parteiverbot motion filed despite lawyer petitions (>1,000 signatures) and a GFF legal opinion favoring one; AfD remains eligible to compete and win. [claude_news] 5. **Market definition & base rate** — "Win" implies plurality of seats/second votes (consistent with 2025 election framing). Historical base rate: party leading polls ~3-4 years out won the eventual plurality in only 4/8 elections (1998-2025, 50%). Monte Carlo modeling gives ~50-60% depending on horizon/lead size, but these models assume small (0-2pt) leads — AfD's current 7-9pt lead is far larger, pushing empirical probability well above these baseline estimates. [code_execution] 6. **Polymarket** — No matching markets found (0/100 scanned); no cross-market comparison available. [polymarket_related] 7. **Structural factors** — Brandmauer is eroding at the voter level (51% of Germans, 55% of CDU/CSU voters no longer support categorical exclusion of AfD cooperation) though no formal CDU policy shift yet. Record state-election wins (Saxony-Anhalt 43.8%) and coalition unpopularity are pushing AfD support upward; deep unpopularity of Merz government amplifies this. [claude_news] # Key facts (high-confidence, factual) 1. [kalshi_direct] Current YES price 87%, up sharply over 30 days. 2. [claude_news/article_search] AfD polls first nationally at 27-29%, ~7-9pts ahead of CDU/CSU (historic low 18-21%). 3. [Wikipedia] Next regular election due on/before 2029-03-25; AfD is currently the largest opposition party (151 MPs) after finishing second in 2025 with 20.8%. 4. [claude_news] BfV's extremist classification judicially suspended pending litigation; no ban procedure formally initiated. 5. [article_search] AfD won Saxony-Anhalt state election Sept 2026 with 43.8%, more than doubling 2021 result; CDU collapsed to 17.2%. # Cross-market signals - Kalshi related: Chancellor market prices Friedrich Merz at only 2% to be next Chancellor, consistent with high AfD win probability (though AfD winning plurality ≠ AfD forming government, given the firewall). - Polymarket: No active market found; no comparison possible. - Sportsbook implied: None available. # Analyst opinions and speculation - Atlantic Council: coalition could survive to 2029 but "not insignificant" collapse risk given negative economic outlook. - Söder (CSU): warns a second coalition collapse would be a "penalty kick with no goalkeeper" for AfD. - Historical/statistical modeling (code_execution) suggests base rates for "leader wins" are closer to 50-60% at long horizons — creating tension with Kalshi's 87%, though AfD's current lead is unusually large and reinforced by real election results (Saxony-Anhalt), not just polls. # Directional lean per outcome - **Yes**: Strong, widening, multi-poll-confirmed lead; validated by actual state election landslide; coalition deeply unpopular; firewall eroding at voter level; no ban imminent. - **No**: Long time horizon to 2029 election allows substantial mean reversion (historical base rate only 50% for polling leaders 3-4 years out); firewall/ban risk, though currently stalled, could resurface; CDU/SPD could stabilize if economy improves; "win" typically means plurality not government, but electoral coalitions/dynamics could shift by 2029. # Gaps / unknowns - No Polymarket comparison available. - Unclear whether VG Köln will resolve main case before election, and how a ban ruling might affect eligibility/support. - No clarity on snap election timing/probability itself (distinct sub-market not addressed here). - Historical base-rate model doesn't fully account for AfD's unusually large current lead vs. modeled 0-2pt scenarios. # Calibration anchors - Kalshi current YES price: **87%** (up from 48% a month ago) — primary anchor. - Historical base rate: parties leading polls ~3-4 years out won 4/8 times (50%) since 1998. - Monte Carlo estimates (small-lead scenarios): 50-64% depending on horizon/gap — likely underestimates given AfD's much larger current lead and real-election confirmation.

4. Forecasters Fable 5.1 + GPT-5.6 Sol · in parallel

claude-fable-5-1

46% conf

Yes	80%
No	20%

Kalshi prices YES at 87% after AfD's Saxony-Anhalt landslide (43.8%) and a widening 7-9pt national polling lead (27-29% vs CDU/CSU 18-21%), with the extremist classification judicially blocked and no ban motion filed. However, the regular election is up to 2.5 years away, Merz has ruled out the snap-election path, and the historical base rate for poll leaders 3-4 years out winning is only ~50% (though AfD's lead is far larger than the 0-2pt scenarios that drive that rate). The Kalshi price has also moved +39pts in 30 days, suggesting some momentum overreaction to a single state result. I shade modestly below the anchor to ~80% to reflect mean-reversion, ban/firewall tail risk, and the long horizon, while keeping AfD a strong favorite.

gpt-5.6-sol

68% conf

Yes	79%
No	21%

The 87% Kalshi YES price is the primary anchor, supported by AfD's sustained 27-29% polling, its 7-9-point national lead, and the 43.8% Saxony-Anhalt victory. Snap-election risk and severe government unpopularity favor AfD, while the blocked extremist classification and absence of a formal ban motion mean it remains eligible. I discount the market because the regular election may be roughly 2.5 years away, historically only about half of polling leaders this far out ultimately win, and the recent price rose extremely rapidly from 48% to 87%. AfD's unusually large lead and validation in an actual election nevertheless justify a probability well above the historical baseline.

5. Devil's Advocate Sonnet 5 · red-team critique

Both forecasters converge tightly (0.79-0.80) and discount the 87% anchor by nearly identical amounts (~7-8pts) using the same base-rate argument, but neither explains why that specific discount magnitude is correct rather than, say, 5pts or 15pts—the reasoning is qualitatively identical without differentiation, suggesting anchoring bias on each other's likely output rather than independent analysis.

Both cite the "50% base rate for leaders 3-4 years out" but the brief explicitly flags this dataset uses "small (0-2pt) leads" while AfD's lead is 7-9pts confirmed by an actual landslide election (Saxony-Anhalt 43.8%, more than double 2021's result) — neither forecaster attempts to quantify how much this reference-class mismatch should shift the base rate upward, they just gesture at it and then apply a discount anyway.

Neither forecaster meaningfully engages with the "win" definition distinct from "govern" — the brief notes Merz's Chancellor market is at only 2%, implying markets treat AfD's plurality win as highly likely while governance remains blocked by the firewall; this is a *supporting* signal for the Yes case that both mention only in passing without incorporating its weight.

Both treat the ban/classification risk as a modest generic tail risk, but the brief notes a 1,500-page GFF legal opinion and 1,000+ signature petition are pushing toward an actual Parteiverbot motion — neither forecaster assigns explicit probability to a scenario where a ban proceeding materially disrupts AfD's ability to compete/win before 2029, despite this being a distinct, evidenced pathway to "No" beyond simple polling reversion.

Neither forecaster addresses the possibility of a snap election happening *sooner* (before the scheduled 2029-03-25 date) despite Merz's ruling-out being non-binding and coalition collapse risk being called "not insignificant" by Atlantic Council — an earlier election could either lock in AfD's current strong lead (favoring Yes) or occur before further momentum consolidates, and this timing uncertainty is unaddressed in both rationales.

Both forecasts sit almost exactly at the same value despite differing confidence levels (0.46 vs 0.68), and neither explains what asymmetric evidence justifies GPT's substantially higher confidence in essentially the same point estimate — this inconsistency in stated confidence vs. probability convergence is unexamined by either party.

6. Aggregator Fable 5.1 · final calibrated forecast

83% (57% confidence)

Yes: 83% No: 17%

Both forecasters discounted the 87% Kalshi anchor to ~0.80 on a 50% base rate for poll leaders 3-4 years out, but the critique correctly notes that reference class is built on 0-2pt leads whereas AfD holds a 7-9pt lead validated by an actual 43.8% state landslide and a cross-market signal (Merz at 2% for next Chancellor) consistent with a near-certain AfD plurality. Snap-election risk, though Merz has ruled it out, is asymmetric toward Yes since an earlier vote would lock in the current lead. The genuine No pathways are 2.5 years of mean reversion (e.g., a CDU leadership change restoring competitiveness) and a Parteiverbot, but a federal ban procedure realistically cannot conclude before 2029 and the extremist classification is judicially suspended. I land at 0.83 — slightly below the market to reflect the rapid 48→87 run-up and long horizon, but above the two forecasts because their discount was under-justified relative to the size of the lead.

Pipeline Timing

Total pipeline time: 213.7s

Per-tool research timings shown in the Research section above.