USD/BRL reaches 5.45 by October 25, 2026 in the base case, 5.62 in the bull case and 4.90 in the bear case. The mechanism is a speculative long-real position sitting in the 99th percentile of 31 years of Chicago Mercantile Exchange (CME) data, running into a two-round general election that Brazil’s Monetary Policy Committee (Copom) will sit out entirely.

USD/BRL traded at 5.0956/5.0962 on the Banco Central do Brasil (BCB) PTAX fix of September 3, 2026, roughly 4% below where it started the year. The number behind this call is 67,902: the net non-commercial long position in CME Brazilian real futures on August 25, 2026, the seventh-largest reading in 890 weekly observations since November 1995. What follows sets out why that crowd, that calendar and that volatility history point higher, and the four signals that would prove it wrong.

Key Levels:

• USD/BRL: 5.0956 bid / 5.0962 ask — BCB PTAX fix, September 3, 2026
• Base case target: 5.45 by October 25, 2026 — a one-standard-deviation move at 18.3% annualised volatility, the average realised volatility of the three previous Brazilian presidential election windows (Federal Reserve Economic Data, series DEXBZUS)
• Bull case target: 5.62 — a 1.5-standard-deviation move on the same basis, above the highest end-2026 forecast in the BCB Focus survey (5.60)
• Bear case target: 4.90 — the 2026 low of 4.8975 set on May 11 (DEXBZUS), triggered by a decisive first round that removes the runoff
• Major resistance: 5.4224 — the 2026 high, January 2, 2026 (DEXBZUS)
• Major support: 5.0577 — the second-half 2026 low, July 22 (DEXBZUS)
• Invalidation level: a weekly close below 5.0577 at any point before the October 4 first round

Methodology and what this call does not claim

Spot levels come from the BCB PTAX fix and from Federal Reserve Economic Data series DEXBZUS, the New York noon buying rate. The two differ by roughly 0.3% on a given day because they are struck at different times; DEXBZUS is used for all historical ranges and volatility work, PTAX for the current level. Positioning comes from the Commodity Futures Trading Commission’s Commitments of Traders (COT) report for the CME, futures-only, contract code 102741, with the full 890-week history to November 21, 1995 pulled from the CFTC public reporting database. Volatility is close-to-close, annualised at 252 trading days. Policy expectations come from the BCB Focus survey collected August 28, 2026. This is a directional view on a seven-week window, not a forecast of the election result, and it takes no position on which candidate wins.

The crowd is the trade: what CME positioning shows

Non-commercial accounts held 91,963 CME Brazilian real futures long against 24,061 short on August 25, 2026, a net long of 67,902 contracts on open interest of 131,553 across 112 reporting traders. That is not a record — the record is 82,558, set on September 30, 2025 — but it sits at the 99.2nd percentile of the full history, and only 13 of 890 weeks in 31 years have exceeded 60,000 net long. The position was rebuilt fast: net length was 17,617 contracts on January 6, 2026 and 30,848 as recently as July 7, so the crowd has more than doubled its exposure in seven weeks. One nuance argues against reading the latest week as fresh conviction. Open interest jumped 18,145 contracts week-on-week, but 13,516 of that was spreading, while outright longs added 233 and shorts added 1,080. Net length actually fell by 847. Speculators are not still piling in; they are already there.

Election year	Net non-commercial BRL position, first COT of September	Percentile of 890-week history	USD/BRL, Sept 1 to first print after the runoff	Realised volatility over that window
2014	−3,001	26th	+13.0%	18.3%
2018	−16,771	15th	−11.8%	17.2%
2022	+30,860	88th	+0.4%	19.4%
2026	+67,902 (August 25)	99th	Base case +6.9%	9.0% trailing 30 days

Sources: CFTC Commitments of Traders, CME futures-only, contract 102741, full history to November 21, 1995; Federal Reserve Economic Data series DEXBZUS. Windows run from the first business day of September to the first published rate after each runoff (October 26, 2014; October 28, 2018; October 30, 2022). Data collected September 4, 2026.

The table isolates what is genuinely unusual about 2026. Brazil has held three presidential elections inside the FRED daily series, and in two of them speculators entered September net short the real. In 2022, the only prior cycle where the crowd was long going in, it was long by 30,860 contracts at the 88th percentile — less than half today’s position. Realised volatility in all three windows ran between 17.2% and 19.4% annualised. USD/BRL has realised 9.0% over the past 30 days and 10.5% over 90 days, roughly half the rate that has actually prevailed in Brazilian election windows. The base case does not require a crash. It requires volatility to revert to its own election-window norm while the largest speculative long-real position in a generation is exposed to it.

“This seems the first day that politics has really started to hit the real this year. We would not chase the real lower, however. 13.4% implied yields through the one-month non-deliverable forwards and Brazil’s position as a net energy exporter should keep the currency reasonably in demand. Positioning is probably quite crowded long the real now, but we suspect it would require a broadly stronger dollar, rather than local news, to send USD/BRL through 5.22.”

— Chris Turner, Global Head of Markets and Regional Head of Research for UK and CEE, ING (ING Think, August 12, 2026)

Why the mechanism runs through a 49-day Copom blackout

Brazil’s Superior Electoral Court (TSE) has set the first round for October 4, 2026 and the presidential runoff for October 25. Copom’s published 2026 calendar has meetings on September 15–16 and then November 3–4. There is no October meeting. That leaves 49 days between the September decision and the next one, a stretch containing both rounds of the vote and every poll released in between.

That matters because the carry doing the work here is extraordinary and entirely policy-dependent. The Selic target stands at 14.00% against an effective federal funds rate of 3.63% on September 2, 2026, a nominal differential of 1,037 basis points, and Brazilian headline inflation ran at 4.44% in the 12 months to July, leaving an ex-post real policy rate above 9%. Focus respondents expect the Selic to end 2026 at 13.75% and inflation at 5.01%. That combination is precisely what drew 91,963 gross long contracts into CME futures. It is also why the position is fragile: a carry trade this crowded pays a known coupon while holding an unhedged short position in event volatility, and the institution best placed to lean against disorder has no scheduled voice for seven weeks.

The polling gives that blackout something to work on. A Quaest survey for TV Globo published September 2, 2026, with fieldwork from August 30 to September 1, put President Luiz Inácio Lula da Silva on 42% against Senator Flávio Bolsonaro on 41% in a runoff — inside the two-point margin of error, and narrower than the three-point lead the same pollster measured on August 14. A one-point spread seven weeks out is not a forecast; it is a guarantee that every subsequent poll is tradable. The steelman for the other side is straightforward: Brazil holds US$371.9bn of international reserves as of September 2, 2026, runs a real rate no developed market can match, and in 2018 the real strengthened 11.8% across the same window when the market decided it liked the likely winner.

What the model misses

Three limits deserve naming. First, positioning is a condition, not a catalyst. Speculative length was already at the 99th percentile through May 2026, when net longs hit 71,651 on May 26, and the real strengthened anyway to its 4.8975 low a fortnight earlier. Second, the 2018 analogue is a genuine two-sided risk: a first-round result markets read as decisive and reform-friendly would compress event premium immediately, and the bear case to 4.90 is the honest expression of that. Third, consensus is split and reasonable people sit on both sides. The most recent verifiable Reuters poll of 31 foreign-exchange strategists, conducted November 3–5, 2025, put the median 12-month USD/BRL forecast — a horizon ending October 30, 2026 — at 5.48, slightly above this call. ING’s published August 6, 2026 forecast table runs the other way, at 5.15 for one and three months, 5.00 at six months and 4.75 at 12. The Focus median for end-2026 sits at 5.20, between them.

“Latin America looks quite attractive relative to tech-heavy Asia. There’s uncertainty over whether the pace of the AI spending boom is sustainable, so if that investment impulse eases, you might see a pullback in some Asia names.”

— Katie Exum, Co-head of Sovereign Research, Gramercy (Reuters, August 17, 2026)

What would invalidate this call

The base case to 5.45 breaks if any one of these four signals fires:

• A weekly close below 5.0577 before October 4. That is the second-half 2026 low. Breaking it before the first round shows the market absorbing election headlines without demanding compensation — the opposite of the premise.
• Net non-commercial BRL length falls below 45,000 contracts in a weekly COT report before the first round. A reduction that size means the position has been unwound in an orderly way in advance, removing the risk the call is built on.
• Copom delivers a hawkish surprise on September 16 — a hike, or guidance removing the Focus-implied 25 basis-point cut by year-end. A wider carry cushion going into the blackout compensates holders for exactly the volatility this call expects.
• A single candidate polls above 50% of valid votes in two consecutive major surveys before October 4. That puts a first-round win in play, cancels the runoff and collapses the event window the target is priced against.

What to watch next

The Copom decision and statement on September 16 is the last scheduled policy signal before the vote, and the guidance language matters more than the rate. The weekly Focus readout, published each Monday, will show whether the end-2026 median migrates away from 5.20. CFTC Commitments of Traders reports land each Friday at 15:30 Eastern for the preceding Tuesday, so the September 1 and September 8 readings reveal whether the crowd is trimming or adding. Technically, 5.2388 — the August 14 high — is the first level that has to give.

Frequently asked questions

Why is a crowded long-real position bearish for the real?

Because the marginal buyer is already invested. With 91,963 gross long contracts against 24,061 short on August 25, 2026, the pool of accounts left to add is small while the pool that could be forced to reduce is large. A crowded carry position earns a steady coupon and carries an implicit short position in volatility, so an event that widens the distribution of outcomes hurts holders faster than it helps them.

Is 67,902 contracts a record long position?

No. The record across the full CFTC history back to November 21, 1995 is 82,558 contracts, set on September 30, 2025, and net length reached 71,651 as recently as May 26, 2026. The current reading is the seventh-largest of 890 weekly observations, at the 99.2nd percentile. It is extreme, but calling it a record would misstate the data.

Does Copom meet between the two election rounds?

No. The published 2026 calendar has eight meetings, the last before the vote on September 15–16 and the next on November 3–4. There is no October meeting, leaving a 49-day gap covering both the October 4 first round and the October 25 runoff. Scheduled communication resumes only after the presidency is decided.

How was the 5.45 target derived?

USD/BRL realised 18.3%, 17.2% and 19.4% annualised volatility across the September-to-runoff windows of the 2014, 2018 and 2022 elections. Applying the 18.3% average to the 51 days from September 4 to October 25 gives a one-standard-deviation move of 6.8%, taking 5.0956 to roughly 5.44. The target is that one-sigma move, not a guess about a specific poll.

What is the strongest argument against this call?

That the carry is large enough to absorb the volatility. ING’s published forecast table of August 6, 2026 sees USD/BRL at 5.15 in three months and 4.75 in 12, on the view that implied yields through the forwards and Brazil’s terms of trade keep the real in demand. In 2018 the real strengthened 11.8% over the identical window. Brazilian election risk has resolved to the currency’s benefit before.

Compare this with the carry-compression case in USD/MXN, the USD/ZAR call built on South Africa’s 3% inflation target and the USD/TRY path running through the CBRT’s restarted repo auction. Brazil’s macro backdrop also sits behind our analysis of arabica’s path to 250 cents on a record Brazilian crop. Primary data referenced throughout is available from Federal Reserve Economic Data series DEXBZUS, the BCB Focus market readout and the Copom meeting calendar.

This article is informational analysis only and is not financial, investment, or trading advice. Foreign-exchange, commodity, and equity markets are highly volatile and can lose substantial value rapidly. Leveraged products carry total-loss risk and may exceed the initial margin posted. Past performance and historical correlations do not guarantee future results. Do your own research and consult a regulated financial adviser before making any investment decision.