AFOS Analytics
Real-time political pricing
Prediction markets × polls × news
no smoothed averages
Week of the gap that opened without new voting research. Between June 22 and 26, no new national voting intention polls were released (the latest remains the Vox Brasil of June 27, which showed a 2º turno in Empate técnico, 45.3 × 42.8). Even so, the presidential market opened: Lula from 51.50% to 57.50% (↑6.00pp in the week, new cycle peak) and Flávio from 25.05% to 22.25% (↓2.80pp), taking the Lula × Flávio gap from +26.45pp to +35.25pp (↑8.80pp), the widest of the entire period. The tradeable fact is the source of the movement: without new voting data, the record gap was priced on news flow (Flávio with 43% of absences in the Senate, Michelle × Flávio crisis over PL inheritance, Lula ahead in coalitions 25 × 14) and not on polling data. Below the top, the signal is the normalization of Renan: the 1º turno 3rd place contract fell from 62% to 49.5% (↓12.50pp) while the winner contract declined to 11.60% (↓2.65pp), still diverging ~7.60pp from the ~4% of national polls. The presidential contract accumulates ~USD 107M, so the record gap was opened with real money. The next window brings the first voting test since Vox: the AtlasIntel national (June 30 and July 1, n=5.000 each), with a questionnaire that measures the impact of the Michelle × Flávio crisis.
1.Executive Summary
The week of June 22–26 brought no new national voting intention poll: the latest remains the Vox Brasil of June 27, which showed a 2º turno in Empate técnico (45.3 × 42.8). Even without a vote print, the market pushed the gap to a new record: Lula at 57.50% (vol USD 7.0M) and Flávio at 22.25% (vol USD 7.1M), gap +35.25pp (↑8.80pp in the week), the widest of the cycle. The movement was priced on news flow, not on polls: 43% of Flávio's absences in the Senate, Michelle × Flávio crisis over PL estate, and Lula ahead in coalitions (25 × 14). One step down, Renan's peak from Edition №5 normalized: the third-place contract retreated from 62% to 49.5% (↓12.50pp) and the winner contract fell to 11.60% (↓2.65pp). Snapshot closed on June 28, 14:30 BRT (markets operate 7 days; Sunday's data is the most recent over the covered window).
2.Why AFOS does not smooth
The aggregation industry reduces 'the market versus polls' to a single, smoothed reading. This week, that average is precisely the error: mixing the market's record gap with the 2º turno coin flip from Vox cancels out the fact that defines the window—that the market opened the gap to a record without any new vote print to sustain it.
Why it matters: the last polling survey of the window, Vox from June 27, is the tightest 2º turno of the cycle (Lula 45.3 × Flávio 42.8, gap +2.5pp). In the same interval, with no new national print, the market took Lula's probability of victory to a record gap (+35.25pp, ↑8.80pp in the week). An aggregated reading would nullify the most useful information: real money opened the advantage to a record backed by news flow, not by polling, and in a direction opposite to the last voting survey. AFOS Tradeoff reports both side by side and treats the divergence as an observed variable, not as error to smooth. Method note: market gap is probability of victory, not vote margin, so the comparison is one of direction and conviction, not pp to pp.
Unlike AFOS Daily Edition №5, where the record gap resisted an entire national polling battery, this week the signal is that the gap widened even further without any new polling test, driven by news flow. This raises the stakes for the next release: the AtlasIntel national poll (June 30 and July 1, n=5,000 each) is the first vote portrait to confront the record gap and measure the impact of the Michelle × Flávio crisis. The AFOS Tradeoff registers the directional divergence between market and polling and maintains ballot results as the final arbiter.
The same framework we apply to Brazil—measuring the distance between the prediction market and polls and checking the signal against the actual ballot results—has already been tested against nine elections across four continents: South Korea (2025, Asia's first case), Peru and Colombia (2026), Chile, Germany and Canada (2025), and United Kingdom, Mexico and United States (2024). In some cases the signal indicated high divergence; in others, near-zero convergence (Colombia, Germany); and in the United States in 2024 two markets disagreed with each other (the electoral college was correct, the popular vote was wrong), recorded with radical honesty. What validates the method is the actual result, not the direction of the signal, which is why we don't speak of 'validated divergences', but rather of cases checked against what the ballots delivered. It is auditable and open track record, not claims of predictive superiority. The cases are in the AFOS global hub.
3.Weighted scenarios for the week
Three paths for the screenshot window June 29-July 3, with the AtlasIntel national poll (June 30 and July 1, n=5,000 each) as the first vote test since the Vox poll of June 27:
A AtlasIntel (June 30 and July 1, n=5.000) confirms Lula's advantage in line with recent national polls (1º turno ~+10pp; 2º turno in the +4 to +7pp range), and the market maintains the gap in the +32 to +36pp range. Renan remains ~4% in polls, preserving the divergence of ~7.60pp in the winner contract (11.60%). Market and polling converge on the leader's direction; the market continues pricing an advantage larger than the 1º turno. Net-neutral for BRL.
A AtlasIntel, que mede o impacto da crise Michelle × Flávio, mostra reconcentração da direita acima do precificado (Flávio acima de 33% no 1º turno) ou confirma um 2º turno apertado na faixa da Vox (45.3 × 42.8), e o gap de mercado comprime para +25 a +30pp. Implícito: o mercado abriu o gap a um recorde sobre fluxo de notícia e começa a convergir para o sinal das pesquisas quando o print de voto chega.
Institutional event of high impact displaces the outlook from electoral dynamics: acute development of the Master/Vorcaro case at the STF, with Mendonça's request to the PGR regarding Dark Horse film financing, or a domestic print far outside consensus. The impeachment contract at 2.65% signals that the market does not price in imminent institutional rupture; repricing here would require a new trigger.
4.Indicator Grid
| Contract | Current | Δ week | Vol USD acc. | Implied reading |
|---|---|---|---|---|
| Lula · winner | 57.50% | ↑6.00pp semana | 7.0M | New cycle peak; record gap opened without new polling survey |
| Flávio · winner | 22.25% | ↓2.80pp semana | 7.1M | Declined during the week amid adverse flow (43% absences in Senate, family crisis) |
| Lula × Flávio Gap | +35.25pp | ↑8.80pp semana | - | Longest throughout the entire period; driven by news, not by voting polls |
| Renan · 3rd place 1º turno | 49.50% | ↓12.50pp semana | 93k | Strong normalization after the 62% peak; still the favorite, ahead of Caiado (16%) and Zema (13%) |
| Renan · winner | 11.60% | ↓2.65pp semana | 7.5M | Highest volume among active contracts; diverges ~7.60pp from 1º turno polling (~4%) |
| Flávio · 2nd place 1º turno | 75.50% | ↑7.00pp semana | 84k | Isolated leader in 2nd place; consolidated even as Lula surged in the winner |
| STF impeach <2027 | 2.65% | ↓0.95pp semana | 82k | Back to floor after June 26 peak; thin volume, fragile signal |
| PL plurality Senate | 76.50% | ↓1.00pp semana | 244k | Broad favoritism maintained; no close contender (MDB 11.15%) |
| Inflation 2026 modal band 5.0-5.49% | 38.65% | ↑8.70pp semana | - | Modal band rose to near 39%; consensus 4.50-5.49% totals 69.60%, high tail remains priced in |
5.Liquidity and market structure
The five largest volumes in the entire book remain in probability contracts ≤2.95%: Tarcísio (USD 13.10M), Carlos Massa (10.17M), Eduardo Bolsonaro (9.91M), Michelle (8.39M) and Eduardo Leite (7.58M), all legacy positions never dismantled. It is legacy volume, not current traction; none are active contenders in that range. Read 'low price + high volume' as concentrated conviction already priced in and resolved, not as live interest. The relevant data for the week sits just below the top-5: the contenders with material probability, Renan (7.5M), Flávio (7.1M) and Lula (7.0M), remain clustered around USD 7M, and Renan still leads accumulated volume among active contenders, sustaining 11.60% on the winner (diverges ~7.60pp from 1º turno polls) while his third-place contract normalized from 62% to 49.5% in the week.
The 5 largest accumulated volumes account for ~USD 49M (~46%) of the presidential market (total ~USD 107M). The cross-reading of price × volume for the week is at the top: the opening of the gap to a record (+35.25pp) was made on real flow (Lula USD 7.0M, Flávio USD 7.1M), not on shallow book, and without any new voting research in the window. Low volume spike (USD <500k) in individual contract should be treated as noise until confirmation of recurring flow.
6.Calendar of price-relevant prints
| Date | Sample | Why it matters | |
|---|---|---|---|
| Mon Jun 29 | Nexus (national) | n=2.000 | The week opens; national reinforcement, cross-check of Lula's advantage |
| Tue Jun 30 | AtlasIntel (national) | n=5.000 | Tier 1, first voting print since Vox; central test of record gap +35.25pp |
| Wed Jul 01 | AtlasIntel (national) | n=5.000 | According to AtlasIntel screenshot; survey measures the impact of the Michelle × Flávio crisis |
| Wed Jul 01 | Real Time Big Data | n=1.600 | State-level registration; scope to be confirmed |
| Thu Jul 02 | Instituto Gazeta | n=3.000 | Large sample, state-level scope; regional context reinforcement |
Source: TSE registry via AFOS API. The June 29–July 3 window brings the dual national AtlasIntel poll (n=5.000 each, Tier 1), the first vote test since the Vox of June 27, with a questionnaire measuring the Michelle × Flávio crisis. Registered ≠ published: inclusion does not guarantee disclosure or figures. Filter applied: sample ≥ 1,000.
7.Watch list, week triggers
- AtlasIntel dupla (June 30 and July 1, n=5,000). The first vote test since Vox. A 1º turno below +10pp, or a 2º turno in the Vox range (45.3 × 42.8) confirmed, strains the record gap of +35.25pp opened without polling.
- Impact of the Michelle × Flávio crisis. The AtlasIntel questionnaire measures the conflict over the PL succession and the Wagner case; it is the first voting data to quantify what the market has already priced in via news.
- Convergence or divergence of the gap. Watch whether the market maintains the gap above +32pp or begins to converge toward the 1º turno polling (~+10pp) when the vote print arrives.
- Renan's normalization. The third-place contract fell from 62% to 49.5%; watch whether it stabilizes or if Caiado (16%) and Zema (13%) continue closing in, and if any snapshot records Renan above 4% in the 1º turno.
- Case Master / Dark Horse in the STF. Mendonça's request to the PGR regarding Vorcaro's financing of the Dark Horse film; any action that moves the impeachment contract away from 2.65%.
8.Methodology
AFOS Tradeoff aggregates three signals without averaging them into a composite: Polymarket (denominated in USD, latency ~30min), polls registered with the TSE (declared intent, variable frequency) and 400+ press sources (event flow). When the three diverge, the divergence is the signal, not consensus. USD volume is reported alongside implied probability to separate conviction from artificial spike. Liquidity (book depth) is not cited inline because low liquidity on Polymarket does not mean wrong price (active arbitrage in minutes), and exposing the technical number generates misread in lay audience.
Weekly Δ series derive from snapshots persisted daily on Neon (June 21–28), on a basis consistent with the dashboard pipeline. Source code, raw data and daily editorial synthesis (PT/EN/ES) under Apache 2.0 license: afos-analytics.com · github.com/AFOS-Analytics.
9.Additional reading · macro coverage
Relevant articles and sections from the week in reference outlets. AFOS Tradeoff is the primary source (Polymarket + TSE + surveys); the references below are supplementary reading for macro context alignment. Attention: some content operates behind a paywall.
- Valor Econômico · Flávio Bolsonaro was absent in 43% of nominal votes in the Senate in 2026 (June 28)(paywall), valor.globo.com/politica/noticia/2026/06/28/flavio-bolsonaro-se-ausentou-em-43percent-das-votacoes-nominais-do-senado-em-2026.ghtml
- O Globo · In conflict with Flávio, Michelle uses PL structure in dispute over Bolsonaro's legacy (June 28)(paywall), oglobo.globo.com/politica/noticia/2026/06/28/em-conflito-com-flavio-michelle-usa-estrutura-do-pl-na-disputa-por-espolio-de-bolsonaro.ghtml
- Reuters · Americas/Brazil section, political flow and country risk, www.reuters.com/world/americas/
- Estadão · Renan bets on crowdfunding, MBL base and Faria Lima to run in 2026 without party machinery (June 28), www.estadao.com.br/politica/renan-aposta-em-vaquinha-base-do-mbl-e-faria-lima-para-disputar-2026-sem-maquina-partidaria/
- Bloomberg Línea · Brazil/emerging markets and BRL coverage(paywall), www.bloomberglinea.com.br/
- InfoMoney · politics and market (open access), www.infomoney.com.br/politica/
Macro context references. The primary signal of this edition (record gap Lula × Flávio open without new voter polling, and Renan's normalization in 3rd place) is direct observation of Polymarket pricing crossed with TSE records, not derived from the articles above.