Brazil's software sector reaps AI efficiency gains, but long-term growth path unclear

Cost savings are visible and real. Revenue multiplication from AI is still theoretical.
Brazilian software companies are reaping immediate efficiency gains from AI, but investors remain uncertain whether the technology will drive sustainable revenue growth.

No coração do setor de software brasileiro, uma divisão silenciosa se aprofunda: as máquinas já estão cortando custos, mas ainda não aprenderam a criar receitas. O Bank of America, ao avaliar TOTVS, LWSA e VTEX no primeiro trimestre de 2026, encontrou empresas mais eficientes e investidores menos convictos — um paradoxo familiar na história da adoção tecnológica, onde os frutos imediatos raramente satisfazem quem sonha com a colheita futura. A questão que paira sobre o setor não é se a IA funciona, mas se ela crescerá além de si mesma.

  • As margens EBITDA dispararam — VTEX subiu 850 pontos-base, LWSA 470 e TOTVS 200 — mas o mercado reagiu com ceticismo crescente, não com celebração.
  • Analistas do BofA cortaram preços-alvo mesmo mantendo recomendações de compra, sinalizando que a confiança no presente não garante fé no futuro.
  • O corte de custos via IA já é real e mensurável; a multiplicação de receitas via IA ainda é promessa — e essa distância está corroendo a paciência dos investidores.
  • Cada empresa traça seu próprio caminho: VTEX reinventa a plataforma inteira, TOTVS aposta na monetização da base instalada, LWSA foca em eficiência de infraestrutura.
  • O segundo semestre de 2026 surge como o horizonte crítico: se novos produtos não gerarem receita até lá, o ciclo de downgrades pode se intensificar.

O Bank of America divulgou na última quinta-feira uma avaliação estratégica do setor de software latino-americano, com foco nos resultados do primeiro trimestre de TOTVS, LWSA e VTEX. O retrato foi ambivalente: ganhos concretos no presente, incertezas crescentes sobre o futuro.

Os números surpreenderam positivamente. As margens operacionais se expandiram de forma expressiva nas três empresas, impulsionadas pela alavancagem operacional e pelas primeiras eficiências tangíveis da inteligência artificial — automação de processos internos e redução de custos de programação. TOTVS avançou 200 pontos-base no EBITDA; LWSA, 470; VTEX, 850. O padrão era o mesmo: a IA colhendo os frutos mais acessíveis.

Mas o humor do mercado havia mudado. O próprio relatório do banco capturou a tensão: a IA é o debate central do setor, com preocupações crescentes sobre a sustentabilidade do crescimento de longo prazo — mesmo sem impactos visíveis ainda nos resultados. O BofA manteve recomendações de compra para as três empresas, mas revisou os preços-alvo para baixo. O da LWSA caiu de R$7 para R$6; o da TOTVS, de R$47 para R$44.

As três companhias navegam a transição por rotas distintas. A VTEX trata a IA como uma reinvenção completa de plataforma e modelo operacional. A TOTVS aposta na monetização de sua base instalada, usando dados proprietários para aprofundar o vínculo com clientes via agentes e automação. A LWSA concentra esforços em eficiência de integração e infraestrutura de dados estruturados.

O verdadeiro nó da história está no intervalo entre o que a IA já entrega e o que ela ainda promete. Os analistas apontam o segundo semestre de 2026 como o momento em que novos produtos poderão começar a gerar receita adicional. Se esse prazo se confirmar, o ceticismo pode se reverter. Se a fase de eficiência se prolongar sem multiplicação de receitas, a paciência dos investidores continuará se desgastando — e o setor permanecerá suspenso entre um presente tangível e um futuro ainda por provar.

Bank of America released a strategic assessment of Latin America's software sector last Thursday, examining the first-quarter results from major Brazilian players like TOTVS, LWSA, and VTEX. The picture it painted was split: immediate gains masking deeper uncertainty.

The numbers themselves looked solid. Revenue tracked with the prior year, and profitability surprised to the upside. Operating margins expanded noticeably, driven by two forces working in tandem—operational leverage and the first tangible efficiencies from artificial intelligence deployment. TOTVS saw its EBITDA margin rise 200 basis points year-over-year. LWSA jumped 470 basis points. VTEX, most dramatically, climbed 850 basis points. The pattern across all three was identical: internal cost-cutting and programming automation, the low-hanging fruit of AI implementation.

Yet the market's mood had shifted. Recent waves of recommendation downgrades reflected a creeping skepticism among investors about what happens next. The bank's own report captured the tension plainly: "AI remains the central debate across the software sector, with growing concerns about the long-term growth sustainability of software companies, even though we don't yet see impacts in the results." Cost savings are visible and real. Revenue multiplication from AI is still theoretical.

Bank of America maintained buy ratings on all three companies, betting that margin expansion would anchor profit growth. But the bank also revised its price targets downward, incorporating more conservative long-term views. LWSA's target fell from R$7 to R$6, reflecting expectations of slower revenue growth as AI matures. TOTVS dropped from R$47 to R$44, adjusted for quarterly performance and softer revenue in its R&D, Techfin, and Linx divisions.

The three companies are pursuing distinct paths through the AI transition. VTEX is taking a transformational approach, framing AI as a reinvention cycle across its entire platform and operating model. TOTVS, leveraging its sector-leading position and proprietary data, is positioning AI as a monetization opportunity for its installed customer base—focusing on agents and automation that could deepen customer lock-in. LWSA is narrower in scope, concentrating on integration efficiency and channeling value toward infrastructure and structured data.

The gap between what AI is delivering now and what it might deliver later is the story's real tension. Efficiency gains are immediate and measurable. New revenue streams remain nascent. The bank's analysts note that AI monetization in Latin America's software sector is still in its infancy, though they signal that new products could begin generating additional revenue starting in the second half of 2026. That timeline matters. If the efficiency phase extends much longer without revenue multiplication, investor patience will erode further. If new products gain traction by mid-year, the skepticism could reverse. For now, the sector is caught between the tangible present and an uncertain future.

AI remains the central debate across the software sector, with growing concerns about the long-term growth sustainability of software companies, even though we don't yet see impacts in the results.
— Bank of America report
VTEX is adopting a more transformational approach, framing AI as a reinvention cycle across all facets of the business.
— Bank of America analysis
The Hearth Conversation Another angle on the story
Inventor

Why would investors worry about long-term growth if the companies are already showing margin expansion?

Model

Because margins can expand while the business stagnates. You cut costs, profits rise, but you're not growing the top line. Eventually that well runs dry.

Inventor

So the AI is just a cost-cutting tool right now?

Model

Mostly, yes. TOTVS, LWSA, VTEX—they're all using it to automate internal work, reduce headcount pressure, squeeze more output from existing teams. That's real and valuable. But it doesn't create new customers or new products.

Inventor

And the bank thinks new products are coming?

Model

They're signaling it. They expect AI-driven products to start generating revenue in the second half of 2026. But that's a forecast, not a fact. The market is skeptical because the companies haven't proven they can monetize AI yet.

Inventor

Why did they cut price targets if they still recommend buying?

Model

Because they're being honest about near-term reality while staying bullish on direction. The companies are worth less today than the bank thought three months ago, but they still think they're worth owning. It's a recalibration, not a reversal.

Inventor

Which company looks best positioned?

Model

TOTVS has the most leverage—it has proprietary data, mission-critical systems, and an installed base it can sell new AI tools to. VTEX is betting on deeper transformation. LWSA is the most cautious, focusing on efficiency rather than reinvention.

Inventor

So we're waiting to see if this is a real business shift or just a temporary margin boost?

Model

Exactly. The next six months will tell you whether AI is a tool these companies use or a product they sell.

Contact Us FAQ