AFOS Daily · Daily Synthesis

August 15, 2026

Prediction Markets × Polls × News

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

The polls narrowed and the price pushed apart, over the same pair of days. Quaest showed both distances between the top two falling against the house's own previous round, while the market, holding USD 123.55M cumulative in the presidential book, widened the distance because the leader broke through a seven-day level. And the candidate registration deadline closes on Aug 15.

1. Prediction market

The confirmed reading of Aug 15, at 13:33 BRT, brings the first substantial movement of the week in the presidential market. Lula stands at 64.50% (USD 8.32M), up 1.00pp, breaking the 63.50% level he had held since Aug 9 and returning to where he last stood on Aug 8. The total cumulative volume in the presidential book is USD 123.55M.

Flávio Bolsonaro stands at 28.15% (USD 8.25M), up 0.30pp and a second straight day of gains. Even so he ended up further behind: the distance between the two widened from 35.65pp to 36.35pp, because the leader rose more.

⛔ Neither value is an extreme. Among the 174 points recorded since May 18, 17 had the leader at an equal or higher level than 64.50%, with a peak of 66.50% on Aug 1; among the 88 days with a recorded distance, 32 had a gap equal to or greater than 36.35pp, with a peak of 41.80pp. The fact of the day is the break, not a record.

For Renan Santos, for the trailing pack and for the second and third place contracts, the Senate one and the STF impeachment one there is no new confirmed reading on Aug 15. The values of Aug 14 stand: Renan at 7.15% (USD 9.59M), Caiado at 1.05% (USD 5.70M) and the impeachment contract at 3.90% (USD 83 thousand), the thinnest among those tracked.

2. What the pollsters registered

Genial/Quaest published on Aug 14 at 21:00, six hours after that day's capture. In the first round, Lula 38% and Flávio Bolsonaro 31%, with Caiado and Renan Santos at 4% and Zema at 2%. In the runoff, 43% against 40%. Fieldwork Aug 10 to Aug 13, n=2,004, registration BR-06773/2026.

The comparison that counts is the house against itself, and it is symmetrical. On Aug 5 Quaest measured 39% x 30% in the first round and 44% x 39% in the runoff. On Aug 14, 38% x 31% and 43% x 40%. In both rounds the leader gave up 1 point and the runner-up gained 1 point, and each distance narrowed by 2pp. The 38% is the floor of the nine national polls since Aug 5, whose range runs from 38% to 44%.

The same round brought two new pieces. Rejection came out separated, at 54% for Flávio Bolsonaro, 52% for Lula and 35% for Caiado, the last with high non-recognition declared by the house. And government approval came in at 46% against 48%.

No new record entered the TSE database on Aug 15. Over 15 days it holds 11 national polls against 55 state ones.

📅 Polling calendar — next 7 days

Polls registered with TSE scheduled for publication between Aug 16 and Aug 22. Inclusion in the table does not mean publication is confirmed, because pollsters may delay or cancel release. Filters applied: national scope and sample ≥ 1,000. Each registration is linked to the TSE public search.

DatePollsterSampleScopeTSE registrationConf.
Aug 17Nexus2,000nationalBR-03317/20260.7
Aug 19Real Time Big Data1,600nationalBR-05423/20260.8

Source: public TSE registry via the AFOS API. 🔥 marks samples ≥ 3,000, and none in this window reaches that size. The table lists the 2 NATIONAL polls of the window; 16 state records from the same period were left out by scope. The status is "registered ≠ published".

3. What the press covered

The candidate registration deadline closes on Aug 15, and from Aug 16 the absence of a registration stops being a pending matter and becomes a fact.

On the judicial thread, Justice Flávio Dino suspended on Aug 14 the effects of the TRF-1 conviction against former senator Romero Jucá, removing his ineligibility. ⚠️ Two qualifiers matter and rarely appear in the headline: it is a single-justice decision, and it depends on ratification by the First Panel to become settled. Dino took the case to the STF on the grounds that the TRF-1 lacked jurisdiction, and the STJ granted an injunction to the same effect. Jucá is a pre-candidate for federal deputy in Roraima, not for the presidency.

Justice André Mendonça pledged to handle the Master and INSS cases impartially, with no measured effect on the polls.

4. Divergences of the day

Poll × market: the two instruments moved over the same pair of days and in opposite directions. Quaest narrowed both distances between the top two, from 9pp to 7pp and from 5pp to 3pp, while the price widened them, from 35.65pp to 36.35pp. One measures declared voting intention, the other measures probability of winning, and the panel does not reconcile them.

House × house: PoderData of Aug 13 tied the rejection of the top two at 48%; Quaest of Aug 14 separated them at 54% and 52%. On the approval net the difference between the two is 5 points, 43% x 50% against 46% x 48%. Two houses, two days, and disagreement on the sign.

Polls × price in the pack: Caiado has 4% in the Quaest first round and rejection at 35%, the lowest on the board, against 1.05% of winning price in the last confirmed reading. ⚠️ Low rejection with high non-recognition is not acceptance: it means most of the electorate has not yet formed an opinion.

In summary

  1. The price broke upward and the poll narrowed, in opposite directions, and that divergence is the finding of the day, not an error in either instrument.
  2. In Quaest the movement is symmetrical, 1 point on each side in both rounds, which is more precise than saying the lead fell.
  3. The candidate registration deadline closes on Aug 15, and the decision on Romero Jucá is a single-justice ruling that still depends on ratification by the First Panel.

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 Aug 15, 13:33 BRT), TSE registry (official electoral polls).

Sources cited in this text: Polymarket, TSE (registro público), O Globo, G1, Folha de S.Paulo, Tribuna do Norte, GZH, Genial/Quaest

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 — August 15, 2026 | AFOS Analytics