AFOS Daily · Daily Synthesis

July 27, 2026

Prediction Markets × Polls × News

Synthesis generated from auditable data. Every claim cites its source.

With 69 days to the first round, the first national poll with fieldwork entirely after the PL convention confirmed the picture at the top and scrambled the one below it: BTG/Nexus puts Ronaldo Caiado in a statistical tie with Lula in the runoff, 45% to 43%, where the Datafolha poll three days earlier gave a seven-point cushion. The market moved in the same direction and raised Caiado's price in both contracts where he appears. Novo formally nominated Romeu Zema and his price fell, the third launch in a row to end that way. Total accumulated volume in the presidential market is around USD 116.13M.

1. Prediction market

With 69 days to the first round, total accumulated volume in the presidential market is around USD 116.13M (Polymarket via the AFOS proxy, capture of Jul 27 at 4:23pm BRT and re-verification at 5:09pm with no change in the presidential book). Lula stayed at 62.50% (USD 7.68M) for a second straight session, with no variation at all. Flávio Bolsonaro rose 0.20pp, to 23.85% (USD 7.68M), and the gap between the two stood at +38.65pp.

This edition records here a correction it owes to anyone who read the July 26 edition. That edition stated that 62.50% was the ceiling of the AFOS series and that +38.85pp came down from a peak of +39.65pp. Sweeping the full contract series, the ceiling of Lula's price is 63.50% and the widest gap on record is +39.80pp, both marked in the early hours of July 26, five hours after the capture the panel published. The record happened after the edition went out, and the correction comes with the same prominence the original claim had.

The mover was the chasing pack, and one name in it. Ronaldo Caiado rose 0.20pp, to 1.95% (USD 5.15M), and Romeu Zema fell 0.10pp, to 0.55% (USD 4.55M), on the very day he was formally nominated. Renan Santos held at 9.55% (USD 8.56M) after two sessions of losses. Across the rest of the group, Jair Bolsonaro stayed at 1.05% (USD 5.25M), Michelle Bolsonaro fell 0.35pp to 0.40% (USD 9.29M), Fernando Haddad fell 0.10pp to 0.25% (USD 6.38M) and Tarcísio de Freitas stayed at 0.15% (USD 13.64M), holding the largest individual volume in the book, a legacy anomaly that measures traded history rather than current conviction.

In the sub-markets the pattern was clearer. In second place in the first round, Flávio fell 1.50pp, to 76.50% (USD 4.32M in the book), and Renan rose 0.80pp, to 12.65%. In third place, Renan gave up 2.50pp, to 62.00%, and Caiado went from 15.50% to a range of 19.00% to 19.50%, in a book worth USD 481 thousand. Flávio's three contracts moved in different directions in the same session, the opposite of the previous day's clean reading, and that does not support a single narrative. In the Senate seats market, the PL stood at 71.00% (USD 292 thousand in the book) and the MDB at 18.25%. The impeachment of an STF justice before 2027 ranged between 3.35% and 3.40% (USD 83 thousand) across the day's captures, a fourth session essentially flat. In the 2026 inflation market, the modal band remains 5.00% to 5.49%, at 39.20%, followed by the 4.50% to 4.99% band at 26.35%, in a book worth USD 77 thousand, a distribution centre that remains above the target ceiling.

A methodological caveat about the capture, because hiding it would be worse than recording it: the AFOS double-reading lock blocked both rounds on July 27. In the first, the second-place and third-place contracts diverged; in the second, only Caiado's third place, between 19.00% and 19.50%. The presidential book returned identical numbers across all four readings, over 17 minutes, which is why the winner prices enter this edition as firm, while Caiado's placement figure enters as a range rather than a settled price.

2. What the pollsters recorded

The BTG/Nexus poll released on July 27 is the first national survey whose fieldwork sits entirely after the convention that formalized Flávio Bolsonaro's candidacy on July 25, which makes it the first real measurement of that event's effect. In the first round: Lula 42% and Flávio 33%, with Ronaldo Caiado at 6%, Renan Santos at 5% and Romeu Zema at 3%. In the head-to-head runoff, Lula 47% x Flávio 43% (Folha de S.Paulo, CNN Brasil, the institute's own release). Fieldwork July 24 to 26, n=2,004, telephone across all regions, 2pp margin, 95% confidence, TSE filing BR-01489/2026, commissioned by Banco BTG Pactual.

Against the institute's own July 13 round, Lula goes from 40% to 42% and Flávio from 34% to 33%, which widens the first-round difference from 6pp to 9pp (JOTA, Money Times). Both moves fit inside the margin, so the widening is nominal and the correct reading is stability with a favourable tilt, not advance. The PL convention, measured for the first time, produced no launch effect: Flávio came out one point lower. His rejection stood at 50%, the highest in the survey, against Lula's 48% (Poder360).

The poll's finding, however, is not about the top two. Caiado ties Lula statistically at 45% to 43%, the narrowest of the four runoff scenarios tested, ahead of Zema (46% x 42%), Flávio (47% x 43%) and Renan Santos (47% x 39%) (Metrópoles, CNN Brasil). On approval, the institute shows 47% against 49% disapproval, versus 47% x 47% on July 13, and government ratings at 36% excellent or good against 43% poor or terrible (Estado de Minas). The Datafolha poll of July 24 showed the reverse on approval, 49% x 48%: both readings carry a 2pp margin and both are statistical ties, which describes a narrow band around 48% rather than a swing.

Also on July 27, a cut of the Datafolha round of July 22 and 23 circulated, with n=2,004, in which 52% believe Trump's tariff push is meant to favour Flávio Bolsonaro's candidacy, against 37% who disagree and 11% who do not know, running at 73% among Lula voters and 29% among his own (Metrópoles). It is an opinion cut on an external topic, not a vote-intention measurement.

📅 Polling calendar — next 7 days

Polls registered with the TSE scheduled for publication between Jul 28 and Aug 3. Inclusion in the table does not mean publication is confirmed, since institutes can delay or cancel a release. Filter applied: sample ≥ 1,000. Each filing is linked to the TSE public search.

DateInstituteSampleScopeTSE filingConf.
Jul 28AtlasIntel 🔥5,000nationalBR-08602/20260.9
Jul 29Alfa Inteligência2,700undeclaredBR-00972/20260.7
Jul 30PoderData2,400nationalBR-07845/20260.7
Jul 31Alfa Inteligência2,700undeclaredBR-04488/20260.7
Jul 31Vox Brasil2,100nationalBR-01084/20260.7

Source: public TSE registry via the AFOS API. 🔥 marks samples ≥ 3,000. The window holds 24 filings with a sample ≥ 1,000; the table shows the 3 with national scope plus the 2 from Alfa Inteligência, which filed a 2,700 sample without declaring the universe, and the other 19, all state-level, were left out for readability and not for lack of a filing. The AtlasIntel poll of July 28 has fieldwork from July 22 to 27 and is the largest sample in the window. Status "filed ≠ published": confirming an actual release requires verification from two primary sources before any figure is cited.

3. What the press covered

The calendar event of July 27 was the national convention at which Novo formally nominated Romeu Zema for the presidency, in Brasília, by unanimous approval and still without a running mate (Estado de Minas, Gazeta do Povo, Brasil 247). The deadline to register tickets runs to August 5. In his speech, Zema criticised the PT and the STF and said he knows the "real Brazil".

The day's second thread came from the financial market and it is reporting, not a finding by any authority. Folha de S.Paulo reported that fund managers, economists and executives speaking on condition of anonymity are hoping Flávio Bolsonaro withdraws in favour of the Caiado and Kassab ticket, citing among the reasons his lack of transparency in dealing with his closeness to banker Daniel Vorcaro, of Banco Master, from whom he asked for money (story summarised by Brasil 247 and by Jornal de Brasília). Senator Rogério Marinho, who coordinates the pre-campaign, replied publicly. Anonymous sources describe a stated preference, not a verifiable fact, and that is how it enters here.

The third thread is the fourth beat of the same sovereign escalation running through the week. After Itamaraty denied visas, after the US denial of any intent to undermine the election and after the ambassador in Buenos Aires was recalled, on July 27 Xi Jinping told Lula that China supports Brazil in rejecting external interference (G1. On the same day, justice Cármen Lúcia defended democracy and said no one should interfere in the country, without naming Trump.

On the legislative front, nothing moved on the electoral agenda: the Senate remains in recess, the government's priority bills are still unvoted and Davi Alcolumbre is still holding back promulgation of the pauta-bomba pending a conversation with Lula. The day's movement was sectoral, with power-sector groups sending Alcolumbre a letter against "jabutis" in a Senate bill. Across the cycle, the government has accumulated 23 defeats in Congress, a figure that contextualises the bottled-up agenda without this edition attributing it to any event of the day.

4. Divergences of the day

Institute × institute: the day's sharpest divergence is not between market and polling, it is between two top-tier houses. BTG/Nexus puts Caiado two points from Lula in the runoff (45% to 43%) and measures him at 6% in the first round; the Datafolha poll of July 24 gave a seven-point cushion (47% to 40%) and 4% in the first round. Fieldwork two days apart, opposite conclusions about which challenger is the most competitive, and no third measurement to break the tie. AFOS records the divergence and does not arbitrate which one is right.

Polling × price, agreeing this time: in the same session in which the poll put Caiado in a statistical tie, the market raised his price in both contracts where he appears, 0.20pp in the winner contract and a rise from 15.50% to a range of 19.00% to 19.50% in third place. Convergence between the two sources is more informative than price moving alone, and in his case it had not happened for two weeks. It is worth recording with the same prominence a divergence would get.

Launch × market, third occurrence: on July 25 the PL formalized Flávio and his price fell 0.10pp; on July 26 the PSD formalized Caiado and it fell 0.15pp; on July 27 Novo formalized Zema and it fell 0.10pp. Three consecutive occurrences are no longer an isolated coincidence, but they still come with no cause asserted: these are small books, single-instant captures, and nothing guarantees the next convention repeats it. It stands as an observation to track until August 5.

Institutes × Renan Santos: the new poll measures Renan at 5% in the first round and breaks the convergence at 3% that Datafolha, Gerp and Indexa had formed last week. The panel is back to three levels among recent readings: 3% at three institutes, 5% at Nexus and 9% at Real Time on July 21. The distortion against his market price still exists and now has a less clear numerator, because the polling itself has started to disagree with itself.

In summary

  1. The first poll with fieldwork after the PL convention confirmed the top, with Lula at 42% and Flávio at 33%, and recorded no launch effect for the candidate formalized two days earlier.
  2. The finding is below: Caiado ties Lula statistically in the runoff, inverting the Datafolha reading of three days earlier, and the market moved in the same direction in the same session.
  3. Three candidates were formally nominated in three days and all three ended the session lower in the book. The week still brings AtlasIntel on July 28, at n=5,000, to say which of the two readings on Caiado holds up.

Fontes consultadas

matérias com link direto para a notícia (veículos âncora):

matérias secundárias (URL Google News redirect — clique resolve à matéria):

Fontes técnicas: Polymarket (live quotes via the AFOS proxy, fetched Jul 27, 5:09pm BRT), TSE registry (official electoral polls).

Sources cited in this text: Polymarket, TSE (registro público), Nexus/BTG, Folha de S.Paulo, G1, CNN Brasil, JOTA, Gazeta do Povo, Estado de Minas, Estadão, Metrópoles, Money Times, Poder360, CartaCapital, VEJA, Brasil 247, Jornal de Brasília, O Antagonista, Pleno.News, NSC Total, Direito News

Method: this synthesis is generated automatically from auditable data on the AFOS Analytics platform, under code-versioned rules in git. All claims can be verified on the platform or in the linked sources. Understand the automated governance.

Integration: for live data and detailed candidate analyses, access the full dashboard. To understand the method in depth, read The Method.

Glossary: Brazilian political terms used in the syntheses (TSE, STF, BolsoMaster, lideranças envelhecidas, etc.), definitions in 3 languages. See the full glossary

Get it by email

Every day, AFOS Daily cross-references prediction markets, polls registered with the Brazilian electoral court, the press and the day’s coverage.

No spam. Unsubscribe anytime.

After signing up you choose whether to receive it in English, Portuguese or Spanish.

AFOS Daily - July 27, 2026 | AFOS Analytics