Brazil's Central Bank (BC) kept the Selic rate unchanged at 15% for the fourth consecutive meeting at the last Copom meeting of 2025, in the second week of December. But this interest rate level — the highest in 20 years — is not expected to last for long, according to projections by Legacy Capital.
The asset manager, a leading player in the fixed income market with over R$16 billion under management, predicts that interest rate cuts will begin as early as the next meeting in January 2026.
“Inflation is plummeting. If the central bank doesn't cut it very quickly now, it will start tightening monetary policy even without doing anything,” says Gustavo Pessoa, head of fixed income and co-founder of Legacy.
The firm projects that, with inflation converging towards the target, there is room for the Selic rate to fall "easily" by 300 basis points (three percentage points). "After that, it will depend more on who wins the election," it predicts.
Pessoa believes that a right-wing candidate's victory could favor more aggressive rate cuts. But little of that is priced in. The co-founder of Legacy estimates that, with Tarcísio de Freitas as governor of São Paulo, Brazil's neutral real interest rate could fall from 8% to 3%.
"Tarcísio is not a candidate, and nobody believes that [Senator] Flávio Bolsonaro will be a candidate . The only credible candidate to beat [President] Lula, in our view, is Tarcísio."
With little certainty about the future of the elections, Pessoa believes that the best asset to hold in Brazil are government bonds indexed to inflation, the NTN-Bs. "With Lula, implied inflation will probably rise, and then you're more protected with NTN-Bs. And, if it's the case of Tarcísio, I think the real interest rate will close too low."
According to Pessoa, the slower pace of issuances and the monetary easing cycle should also favor the appreciation of these bonds next year. "Generally, in a scenario of falling interest rates, NTN-Bs are the best asset to hold," he says. "We believe it's the best asset in Brazil."
Below are the main excerpts from the interview with NeoFeed .
How do you see the scenario for 2026: is there room for the closure of NTN-Bs?
In the case of NTN-Bs (Brazilian Treasury Notes - Series Bs), two factors contributed: the National Treasury started issuing more and experienced lower demand for these assets, which then began competing with infrastructure bonds. This led to a significant widening of interest rates. And, with inflation under control, NTN-Bs are trading below the Selic rate, around 11%. By the middle of next year, the market is already pricing in inflation below 3%. This is a very bad carry.
Is this scenario likely to change?
We believe that private issuances will decrease, as will those of the Treasury, and that the Central Bank will begin cutting interest rates starting in January, which improves the carry of NTN-Bs relative to the CDI. We believe it is the best asset in Brazil.
Even with the carry still low, do you believe it's the best asset to own, given the expected repricing next year?
Generally, in a scenario of falling interest rates, NTN-Bs (Brazilian Treasury Notes - Series B) are the best asset to hold. They are already trading below the real interest rate, and the market is pricing in inflation of 2.90% for the middle of next year, so there is little room for further declines.
According to the Focus Bulletin, inflation expectations for the end of 2026 are above 4%. Should this number be revised downwards?
It's already being revised. There's an unusual amount of resistance, especially because 2026 is an election year. So, they're hesitant to say that inflation will converge 100% to the target because, if the result isn't favorable, the inflation premium could rise significantly. If there's a favorable election, inflation will quickly move closer to the target. This Focus figure, detached from the target, is the election premium.
What should the Central Bank's interest rate cutting cycle look like?
Inflation is plummeting, from 4.70% [in October] to 2.90% [by the middle of next year]. Therefore, the Central Bank will have a unique opportunity to begin cutting interest rates. If it doesn't cut very quickly now, it will start tightening monetary policy even without doing anything, due to the disinflationary effect.
"The Central Bank will have a unique opportunity to begin cutting interest rates. If it doesn't cut very quickly now, it will start tightening monetary policy."
Is there room for inflation expectations to fall even further?
From now on, I think expectations will fall more slowly because people will continue to place electoral risk on the projections. It will be difficult to anchor expectations.
How do you assess the electoral risk?
With the debt so high, the fiscal dynamics going forward are crucial in determining what will happen. The market won't like it if Lula is elected and will price in higher inflation and a more out-of-control fiscal situation. If someone from the right, more fiscally responsible, wins, the market will believe in fiscal adjustment and that inflation will remain low. The neutral real interest rate might be 8% with Lula. With Tarcísio, it could be 3%. When you factor that into the debt trajectory, everything changes.
How much of this positive scenario is already priced in?
It's not very expensive. Tarcísio isn't a candidate, and nobody believes Flávio Bolsonaro will be. In our view, the only credible candidate to beat Lula is Tarcísio. Any Bolsonaro running against Lula would result in a heavy defeat. If Flávio were to run as the Bolsonaro candidate, he'd make it to the second round, but would lose badly afterward. I think it's very difficult for him to reverse this rejection, which is even greater than Lula's. It seems like nobody is supporting him, not even his own party.
Could Tarcísio's non-candidacy cause the market to throw in the towel, leading to a price drop?
I believe so. But the market continues to believe that the Bolsonaros will form an alliance with Tarcísio, with Tarcísio as the key figure. They will probably be the most harmed if they lose the election. But there's no way to know if Flávio's candidacy is calculated or just driven by emotion.
"Lula lost the market's goodwill . The market wanted to believe that he would be more fiscally responsible. But what happened was excessive spending, with much higher taxes."
With Lula being elected, what should we expect?
Lula lost the market's goodwill . The market wanted to believe that he would be more fiscally responsible and that he would have more qualified people in the Finance Ministry, with a credible fiscal plan. But what happened was excessive spending, with much higher taxes. The market no longer believes he can be responsible. In a new term, it would be very difficult to believe. It would even be difficult to find a name to lend credibility to his government. If the president doesn't have good intentions, in particular, it's very difficult for Congress to pass anything.
Despite everything, the stock market is doing well, inflation is under control, and unemployment is low.
We're riding this wave from abroad. Inflation is low all over the world. You look at other emerging countries, they're doing very well. The other countries that don't have fiscal disarray like ours have much lower interest rates.
Were we supposed to be much better off?
Brazil should be in a very good position, but we have to do our homework: we need to adjust public finances and be more capital-friendly, and have more tax and sectoral predictability as well.
In Mexico, the interest rate is at 7%.
So, you look at the trajectory, the expected debt history of Brazil versus these other emerging countries: we are much worse off, much worse, and that condemns the country. Almost all countries have cut interest rates this year. This 15% interest rate doesn't exist.
"Almost all countries have cut interest rates this year. This 15% interest rate doesn't exist."
When does it start to fall?
I think Brazil will eventually realize this. We even think it will start to happen in January. There's room for it to drop 300 bps easily. After that, it will depend more on who wins the election.
With the Brazilian real relatively stable and this level of real interest rates, I imagine that carry trade has been a good bet in the market. How much does the drop in interest rates reduce its attractiveness?
Brazil is "alarming" because it has the highest carry in the world. And, if you take carry divided by volatility — which is usually the metric used — Brazil jumps even higher, because the volatility of the real is also very low. Even with the drop in interest rates, it will remain very high compared to other countries.
There's a lot of talk about seasonal effects on the exchange rate: that the end of the year tends to favor a fall in the real and the beginning of the year an increase. Do you believe the real could fall even further at the beginning of 2026?
There has been a very large movement of dividends due to the tax changes for next year. Part of it has been used to buy dollars to send abroad. But, when next year begins, this flow will pass, and the harvest tends to be good, which fuels the dollar flow. This should anchor the real. If the election is divided, I think the real will remain well anchored, especially in this world of a weak dollar.
Given the electoral risk, are NTN-Bs the best bet?
With Lula, implied inflation will likely rise, and then you're more protected with NTN-Bs (Brazilian Treasury Notes - Series B). And, if that's the case with Tarcísio, I think the real interest rate will close too much. The real interest rate could close much more than it opens.
What explains this favorable trend of lower inflation in Brazil and the world?
There are three driving forces: a weak dollar, China flooding the world with consumer goods, and artificial intelligence. This is happening especially in the United States, but the rest of the world is also beginning to benefit: producing more with the same available workforce. AI is being implemented more intensively in the United States. It's clear that the American job market is losing momentum, slowing down, with unemployment rising, even though the economy is doing well.
Could artificial intelligence lead to a structurally higher unemployment rate?
In some cities with self-driving cars, Uber drivers have to do other jobs, so there's a greater demand for labor. And in the United States job market, unemployment is increasing even with fewer immigrants entering the country. Many jobs will be impacted. It may reach a point where we have more artificial intelligence and less manual labor. It's difficult to know what that mix will be for each sector, for each activity.
So, the trend is towards less inflation in the coming years, thinking on a global level?
What we're going to do is increase productivity, increase supply, as if it were an artificial workforce related to this — services and products related to this.
