← Back to scans

Will anyone win the 2026 Colombian presidential election in the first round? — First-round winner

KXCOLOMBIAPRES1R-26MAY31 · Elections · 2026-05-01
8%
Agent
9%
Market Price
-1.2%
Edge
78%
Confidence
Volume: 10,076
Spread: 0.8c
Days to resolution: 394
Markets in event: 1
Final Rationale
My independent estimate of 6.3% was driven by Cepeda polling at 44.3% (6pp short of 50%) with a fragmented field where the top three candidates split ~86% of support, and the historical base rate of only 1-in-8 post-1991 Colombian elections producing a first-round winner. I PARTIALLY ACCEPT the critique on voto-en-blanco mechanics—blank/null votes mechanically lower the effective 50%-of-valid-votes threshold, providing a real (if narrow) path for a 44-46% raw-intent leader to cross. I REJECT the late-polarization concern as deserving a large adjustment since with the election weeks away, late 5-8pp swings concentrated on a single candidate are uncommon when two competitors poll above 19%. Adjusting upward by ~2pp for the threshold mechanic brings me to ~8%, which now closely matches the shallow-orderbook market at 8.9%.
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 $t{=}0$.
$n = 58$ follow-ups
1. Planner Opus · Fermi decomposition + tool plan
Tools chosen: kalshi_data article_search wikipedia_lookup web_search
Sub-questions (Fermi decomposition)
  1. Is the 2026 Colombian presidential election field fragmented enough that no single candidate is polling above 50%? (w=0.40)
  2. Has any first-round winner ever achieved an outright majority in a Colombian presidential election since the 1991 constitution was adopted? (w=0.30)
  3. Is there a dominant frontrunner candidate in 2026 with polling consistently above 50% of valid votes? (w=0.30)
Planner reasoning
The key mechanism here is Colombia's two-round presidential election system under the 1991 constitution, which requires a candidate to receive more than 50% of valid votes to win outright in the first round. The base rate question is critical: in every Colombian presidential election since 1991 (1994, 1998, 2002, 2006, 2010, 2014, 2018, 2022), a first-round winner has been extremely rare — the country's fragmented political landscape almost always produces a runoff. The notable exception was 2002 when Álvaro Uribe won in the first round with ~53%. That is roughly 1 out of 8 elections post-1991, giving a base rate of ~12.5%. For 2026, the relevant questions are: (1) How fragmented is the current field? (2) Is there a Uribe-level dominant candidate with sustained >50% polling? The temporal analysis matters: the election will likely occur in May/June 2027 (Colombia holds elections in May, but this market closes May 31, 2027, suggesting a May 2026 first round — actually with 394 days until resolution from May 2026, the first round is likely around May/June 2026). Given Colombia's political dynamics — with Petro's presidency generating both strong support and opposition, and typically multiple strong candidates — first-round victories remain historically exceptional. Only a scenario where one candidate consolidates both left-wing and centrist support, or the opposition fragments extremely while one candidate dominates, would yield YES. The prior strongly favors a runoff.
2. Research Parallel tool calls · legacy v2
ToolStatusTimeItemsSummary
kalshi_data OK 1.8s -
article_search OK 0.6s -
wikipedia_lookup OK 2.2s -
web_search OK 57.8s -
3. Evidence Synthesis Sonnet · 10 items · legacy v2
#StrengthCredibilityDirectionSourceClaimPriced In
1 STRONG 88 ↓ DOWN web_search The latest Invamer poll (April 15-24, 2026, n=3,800) shows Iván Cepeda at 44.3%, Abelardo de la Espriella at 21.5%, and Paloma Valencia at 19.8% — no candidate is above 50%. No
2 STRONG 82 ↓ DOWN web_search The race is effectively a three-way contest with center candidates (López at 3.6%, Fajardo at 2.5%) collapsing, but votes are still fragmented across at least three major candidates totaling ~86% of polled support. No
3 MODERATE 75 NEUTRAL article_search Violence is escalating in Colombia ahead of the May presidential vote, with 26 attacks by rebel groups in late April 2026, making security a top voter concern — this could shift votes but not consolidate them toward a 50%+ threshold. No
4 STRONG 95 ↓ DOWN wikipedia_lookup In the 2022 Colombian presidential election, no candidate reached 50% in the first round, requiring a runoff between Petro and Hernández. Yes
5 STRONG 90 NEUTRAL article_search The 2026 Wikipedia article confirms the two-round system is in place: a candidate needs more than 50% in round one, otherwise a runoff on June 21, 2026 follows. Yes
6 STRONG 80 ↓ DOWN web_search Historical context from the forecasting question indicates only Álvaro Uribe in 2002 won in the first round (~53%) among all post-1991 Colombian elections — a base rate of roughly 1 in 8 elections (~12.5%). Yes
7 STRONG 88 ↓ DOWN web_search Frontrunner Iván Cepeda leads at 44.3% in the most recent poll — well below the 50%+ threshold needed for a first-round win, and approximately 6 points short of the required majority. No
8 MODERATE 72 ↓ DOWN web_search Cepeda's support appears tied to Petro's approval recovery (rising from 38% to 49% favorable), but Petro himself remains near-even at 49%/49% favorable/unfavorable — suggesting a ceiling on Cepeda's potential to break 50%. No
9 MODERATE 75 ↓ DOWN web_search Paloma Valencia has nearly doubled her support from ~10% to 19.8% as centrist voters consolidate behind her, suggesting the opposition is consolidating but not behind a single candidate capable of overtaking Cepeda with >50%. No
10 STRONG 70 ↓ DOWN kalshi_data The Kalshi prediction market prices a first-round winner at 8.90%, down sharply (-10.1% in 7 days, -5.1% in 30 days), signaling market participants strongly expect a runoff. No
Information Gaps
  • No data on how the Cepeda polling number translates to 'valid votes' only (blank/null votes excluded from the 50% threshold calculation) — actual valid vote share could differ from raw voting intentions
  • No polling trend data showing whether Cepeda was ever above or approaching 50% earlier in the campaign cycle
  • No data on undecided voter share or likelihood to vote, which could shift the effective threshold
  • No information on whether the security crisis (late April attacks) could dramatically consolidate votes around any candidate
  • Absence of second-round polling scenarios, which would help assess if there's any path toward consolidation before May 31
  • No data on ballot blank ('voto en blanco') levels, which reduce the effective total of valid votes cast for candidates
  • Limited detail on whether all major candidates are confirmed to be on the ballot or if any withdrawals/consolidations are likely before election day
Key Uncertainties
  • Whether Cepeda's 44.3% could surge to >50% in the final weeks before May 31 — a ~6 point gap is large but not impossible given momentum
  • Whether further opposition fragmentation (or consolidation) changes the first-round calculus
  • Impact of the late April security crisis on vote preferences — could security fears benefit or hurt incumbent-aligned candidates
  • Blank vote share: if high blank vote counts reduce total valid votes, the threshold for 50%+ might be harder or differently calibrated
  • Whether any late candidate withdrawals or endorsements could rapidly concentrate the anti-Cepeda vote behind a single challenger, paradoxically helping Cepeda consolidate his lead or preventing him from reaching 50%
Market Context
Price: 0.089 · 7d: -0.101 · 30d: -0.051
Volume: LOW-MODERATE · Depth: SHALLOW
Reliability: LOW-MEDIUM — moderate daily volume (~431 contracts), only 22 days of data, price range of 5-19% suggests significant uncertainty; sharp recent decline aligns with new polling data showing Cepeda at 44.3% with two strong competitors
4. Independent Forecaster Opus · No market price · legacy v2 Did NOT see market price
Independent estimate: 6% (81% confidence)
base rate
8%
75% conf
evidence driven
5%
85% conf
contrarian
6%
82% conf
Spread
3pp
Low disagreement
sq1: Is the 2026 Colombian presidential election field fragmented enough that no single candidate is polling above 50%?
96%
Latest Invamer poll shows Cepeda at 44.3%, with two other candidates at 21.5% and 19.8%. No candidate is close to 50%. The field is clearly fragmented with three major candidates capturing ~86% of support. Probability that the field IS fragmented (no one polling >50%) is very high: ~96%.
sq2: Has any first-round winner ever achieved an outright majority in a Colombian presidential election since the 1991 constitution was adopted?
99%
This is a historical fact question. Uribe won outright in 2002 with ~53%. So YES, it has happened. Answer is essentially certain YES at 99%. Note: this sub-question's framing seems oddly correlated with the main question - if it's asking whether it has EVER happened, the answer is yes, but this provides base rate info (1/8 ≈ 12.5%).
sq3: Is there a dominant frontrunner candidate in 2026 with polling consistently above 50% of valid votes?
4%
Cepeda is at 44.3%, ~6 points short of 50%. No evidence of any candidate consistently polling above 50%. Petro's approval is split 49/49, capping Cepeda's ceiling. Probability that there IS a dominant >50% candidate is very low: ~4%.
Show full reasoning trace
base rate: Only 1 of 8 post-1991 Colombian presidential elections (Uribe 2002) produced a first-round winner, giving ~12.5% base rate. Colombian political fragmentation typically produces runoffs.
evidence updates: Latest April 2026 Invamer poll shows leader Cepeda at 44.3%, ~6 points below threshold. This pushes probability DOWN from 12.5% base rate by ~5% because the leader is meaningfully short and time is limited (election in May/June 2026). Three-way fragmentation with viable second/third candidates further reduces consolidation potential. Petro's polarized approval (49/49) caps Cepeda's ceiling.
combination method: The first-round win requires the leader to actually exceed 50% of valid votes. Given Cepeda's 44.3% polling and structural fragmentation, and adjusting historical base rate downward given specific evidence, I estimate ~8%.
final: Strong evidence points to a runoff. Probability of first-round winner: ~8%.
ensemble: {'ensemble_count': 3, 'perspectives': {'base_rate': {'p_yes': 0.08, 'confidence': 0.75}, 'evidence_driven': {'p_yes': 0.05, 'confidence': 0.85}, 'contrarian': {'p_yes': 0.06, 'confidence': 0.82}}, 'spread': 0.03, 'individual_p_yes': {'base_rate': 0.08, 'evidence_driven': 0.05, 'contrarian': 0.06}}
5. Devil's Advocate Sonnet · legacy v2 Market price revealed
Missing Info Overconfidence
Challenges
  1. Resolution date confusion: The question close time is 2027-05-31, but the Colombian first-round election is in May 2026 (imminent given today is 2026-05-01). The forecaster correctly focuses on the near-term election but should note that resolution is essentially within weeks, not 394 days, making current polling extremely informative and reducing uncertainty about late shifts.
  2. Sub-question 2 is misinterpreted/irrelevant: The forecaster correctly notes sq2 is a historical fact (Uribe 2002 = yes) but this doesn't directly inform the main question. It appears the ensemble combination didn't actually use sq2 mechanically, which is fine, but the 0.063 combined value isn't clearly derived from sq1*sq3 or any principled combination — it's essentially the ensemble mean. Worth making the combination logic explicit.
  3. Slight overconfidence on the downside given information gaps: The forecaster lists meaningful gaps (valid-vote vs raw-intent conversion, voto en blanco rates which MECHANICALLY LOWER the 50% threshold of valid votes, security shock potential, late consolidation). High blank-vote rates in Colombia historically can push a 44-46% raw share over the 50% valid-vote threshold. This factor is acknowledged as a gap but not incorporated into the probability, which may be slightly too low.
  4. The 6-point gap between Cepeda (44.3%) and 50% is not enormous; late polarization or a security-driven rally could close it. Historical Colombian polling has shown 5-8 point shifts in final weeks. The forecast at 6.3% may be slightly underweighting tail scenarios.
Suggested adjustment: +2pp
Modest upward adjustment to account for (a) the voto-en-blanco mechanic that lowers the effective threshold on valid votes — a real path by which 44-46% raw intent could clear 50% of valid votes, and (b) tail scenarios from late polarization or security shocks. The forecaster identifies these as gaps but doesn't price them in. This is a reasoning correction, not market-chasing (the adjustment happens to move toward market but is independently justified).
Market comparison: Forecaster: 6.3%, Market: 8.9%. The 2.6pp divergence is small and within reasonable disagreement. Market reliability is LOW-MEDIUM with shallow orderbook. The forecaster's reasoning is largely sound; the market's slightly higher price likely reflects the same valid-vote threshold mechanics and tail risks the forecaster acknowledges as gaps but doesn't fully incorporate.
6. Aggregator Opus · final calibrated forecast
8% (78% confidence)
My independent estimate of 6.3% was driven by Cepeda polling at 44.3% (6pp short of 50%) with a fragmented field where the top three candidates split ~86% of support, and the historical base rate of only 1-in-8 post-1991 Colombian elections producing a first-round winner. I PARTIALLY ACCEPT the critique on voto-en-blanco mechanics—blank/null votes mechanically lower the effective 50%-of-valid-votes threshold, providing a real (if narrow) path for a 44-46% raw-intent leader to cross. I REJECT the late-polarization concern as deserving a large adjustment since with the election weeks away, late 5-8pp swings concentrated on a single candidate are uncommon when two competitors poll above 19%. Adjusting upward by ~2pp for the threshold mechanic brings me to ~8%, which now closely matches the shallow-orderbook market at 8.9%.