Sobre
I lead Patria’s Buyout portfolio, overseeing more than US$10B in revenues and 65,000…
Artigos de Márcio
Atividades
8 mil seguidores
-
Márcio Tabatchnik Trigueiro publicou istoWhat You Can’t Download One of the sharpest lessons about where AI hits my industry came not from work, but from strength training. I had a shoulder injury. A physician established what was wrong, what I couldn’t do, and when I’d be cleared to progress. Within those constraints, I had AI redesign my routine in three versions: full gym, kettlebells and bodyweight only. Minutes of work for what a good coach does in hours. AI never drew the lines, nor decided when I could cross them. Give it the wrong diagnosis and it returns a confident plan that wrecks the joint. I’ve worked in private equity for 25 years. That little story is my whole industry, and most people won’t say out loud what it means. The production layer in our world is junior work. Models, memos, data rooms. AI now does that layer the way it built my training plan. Faster, cheaper, endless, on demand. Everyone reaches the same conclusion. Fewer juniors, lower cost, seniors move up to judgment. That part is already priced in. Two things aren’t. First, the judgment formation gap, which I flagged earlier this year as one of six structural curiosities about AI. Juniors did grunt work because that is how judgment gets built. You learn to smell a bad deal by building two hundred models for deals that died. Automate that and you lose the training ground for the judgment you still need at the top. You cannot buy judgment. You grow it. Second, the moat moves. When everyone can analyze the same data instantly, analysis stops being an edge, and the only advantage left is what you have to leave the desk to get. Citrini Research, behind this year’s bearish AI call, proved it the hard way. While the market read the same satellite feeds and called the Strait of Hormuz closed, Citrini sent an analyst to Oman. The data was missing half the traffic; the real story was a tolling system, not a blockade. Any data a machine can reach, it commoditizes. The premium goes to what only a human in the field can see. Notice the pattern. Another curiosity was time-to-power, where AI’s limit was physics, not intelligence: turbines and the grid. Same shape here. The limit is never the model. It’s everything the model can’t reach. This is what AIQ, Artificial Intelligence Quotient, measures. Not how smart the model is, but how much you get out of it. The model is the same for everyone now. The operator is the variable. A high AIQ is knowing what to ask, knowing what is true, catching the confident wrong answer, and owning the call you can’t take back. AI gave me a coach for free. It still couldn’t tell me what was wrong, or clear me to push. The work is free now. What decides the outcome is not. If you are early in your career, that is your map. Your edge is no longer the analysis AI does for free. It is the three things it can’t: get what isn’t online, catch what it gets confidently wrong, and own relationships a model never will. None of it is downloadable. #AIinPrivateEquity #LeadershipInAI
-
Márcio Tabatchnik Trigueiro compartilhou issoThe most exciting AI infrastructure story is in space while the investable one is in Ceará. Last week SpaceX filed the largest IPO in history partly on the promise of putting data centers in space. Start with the bull case, because it is genuinely strong. In space, a satellite sits in sunlight 24 hours a day, at solar intensity roughly six times what reaches the ground. Cooling is free — point a radiator at the void, near absolute zero, and heat disappears. Lasers move faster through vacuum than through fiber. SpaceX's own S-1 describes the technology as "novel and untested," and warns the company may spend for years before these efforts become profitable. The rocket science solution is real. It is also, by its own backers' timeline, years away to commercial scale. What is driving all of it is the constraint I flagged three months ago among six structural curiosities: time-to-power. Increasingly, the binding constraint on AI is not compute, not capital, not talent. It is electricity arriving at a specific physical location. And the constraint is most acute exactly where the capital concentrated. Northern Virginia has effectively halted new data center permits in several counties. Dominion Energy publicly admitted it cannot meet demand. US interconnection queues average five years. But there is an answer that requires no rocket science at all. Brazil's electricity mix is 83% renewable, dominated by hydro with substantial wind and solar. The country has abundant land, water for cooling, and roughly 12 GW of available transmission capacity. At our annual investor meeting in New York earlier this month, Rodrigo Abreu, CEO of Pátria's platform Omnia, presented exactly this. The Pecém project: a $9.25 billion data center in Ceará, built with Casa dos Ventos, powered by dedicated wind, with a global anchor tenant committed. It will be the first data center in Brazil authorized to provide data export services. Construction is underway. Operations begin 2027. Not in orbit. Not in 2031. In Ceará, in 2027. There is a second reason the ground endures, beyond timeline. Orbital compute, at least at first, will be controlled by a handful of operators. Not every company, sovereign, or jurisdiction will route its most sensitive workloads through infrastructure it does not control. Renewable-rich, neutral geographies become not a stopgap, but a structural home for compute that lives on the ground. I have been mapping AI as a capital allocation problem in layers. The technology can be right and the capital structure still wrong. Right, and the budget still wrong. Right, and the prices still wrong. And right — while in the markets you assumed it would land, the power simply never arrives. Capital used to follow data centers. Now data centers follow electrons. While the most exciting AI infrastructure story might be in space, the investable one is already pouring concrete in the global south. #AIinPrivateEquity #LeadershipInAI
-
Márcio Tabatchnik Trigueiro publicou istoLast week I asked whether the prices feeding enterprise AI budgets are real prices at all. Microsoft just answered. Effective June 1, GitHub Copilot moves to usage-based pricing. Microsoft was unusually candid: it is now common for a handful of requests to incur costs exceeding the monthly plan price. The signals are consistent. Anthropic recently disclosed in a legal filing that its cumulative inference and training spend exceeds revenue by more than 2x. Last week the WSJ reported that OpenAI's CFO is in open conflict with Sam Altman over compute commitments. The vendor side has been the focus of most commentary. The buyer side concerns me more. Companies are quietly making three decisions that all assume today's AI prices are durable. They are reducing headcount in roles agents can perform. They are redesigning workflows around agent-first architectures. They are embedding inference into product economics, where the cost is now load-bearing in unit margins. Each is rational under current pricing. Each is hard to reverse. Each carries an unspoken assumption — that the input price does not move materially. When that breaks, a 5x increase is not margin compression. It is a strategic event forcing workflows, headcount, and customer pricing to reopen at once. There is also a second-order risk most have not priced in. The reset assumes the vendor you built around is still your vendor in 2028. That is not obvious. Markets in this shape consolidate. The survivors set the new prices — and not to be your friend. Vendor risk and price risk compound. A paradox worth flagging. Mobile telecoms in emerging markets followed a similar pattern — late entrants who arrived after prices cleared built around better economics than early movers. The companies most exposed to the AI reset are those who moved fastest. Those positioned to benefit may be those who build later. Which industries and geographies that favors is a question I will keep watching. Three months ago I wrote that the technology can be right and the capital structure can still be wrong. Last week I wrote that even when the technology is powerful and useful, left unchecked, it can make any budget wrong. Today I add a third layer. Even when the technology is right and the budget under control, the prices we are budgeting against may not be real. Companies pricing operating plans against today's API rates are budgeting against a market that may not exist in 24 months. At the board level, the right questions are not what the latest agent can do. They are: what does this cost per outcome? What price assumptions are we baking into the operating plan? Have we stress-tested for a 3x or 5x reset? If we built operating leverage on these prices, do we still have a business when they clear? These are the questions I would want any boardroom approving an AI program today to care about. Most are not being asked yet. #AIinPrivateEquity #AIinPrivateMarkets #LeadershipInAI
-
Márcio Tabatchnik Trigueiro publicou istoAI got cheaper by a factor of 280. Enterprise AI bills tripled. Both numbers are real. According to Deloitte's 2026 Tech Trends, inference costs have dropped roughly 280-fold over the past two years. Over the same window, enterprise generative AI spending climbed from around $11.5 billion in 2024 to $37 billion in 2025. The unit economics improved by orders of magnitude. Total spending more than tripled. This is not a market failure. It is Jevons Paradox — the 19th-century observation that when the cost of a useful resource falls, total consumption tends to rise faster than the price falls. Coal got cheaper, England burned vastly more coal. Tokens got cheaper, enterprises consume vastly more tokens. The mechanism is structural. Single-call chatbots got replaced by agentic workflows that hit a model 10 to 20 times to finish one task. Retrieval systems stuff thousands of pages of context into every query, just in case the model needs them. Always-on monitoring agents consume compute around the clock. Inference now represents roughly 85% of enterprise AI budgets, up from a negligible share three years ago. What this looks like inside a company is almost never dramatic. A team ships an agentic feature. In testing, the AI handles a customer ticket in three or four steps. In production, the world is messier — the customer asks a follow-up, a tool fails, the AI retries, then retries again. Three steps quietly become thirty. Nobody notices, because no one is watching token consumption the way they watch cloud spend. Then the monthly bill arrives. Finance panics. Engineering kills the feature. Everyone agrees agents are too expensive. They aren't. The company just built a system with no speedometer and no brakes, and was surprised when it crashed. Most boards are flying the same plane. They approved the AI program. They have no instrumentation to know whether it is working, what it costs per outcome, or when it is about to overrun. Approval is cheap. Execution is where the real questions begin — and most companies haven't gotten there yet. A few months ago I wrote that the technology can be right and the capital structure can still be wrong. The same logic applies one layer up. The technology can be right and the budget can still be wrong. Most companies will not feel this as a strategy problem. They will feel it as a cash flow surprise. And there is a more uncomfortable question I will come back to soon. We are debating whether enterprise AI budgets are too high. We should also be debating whether the prices feeding those budgets are real prices at all. #AIinPrivateEquity #AIinPrivateMarkets #LeadershipInAI
-
Márcio Tabatchnik Trigueiro compartilhou issoSomething unusual just happened in AI — and it’s worth looking at it from a few angles. Anthropic is signaling that it has built a very powerful model — Mythos — but instead of releasing it broadly, it is restricting access under Operation Glass Wing, a controlled environment with limited users, monitoring, and clear boundaries. At the same time, references are circulating suggesting this is not just another incremental step, but potentially another ChatGPT-type moment. What stands out is simple: we are hearing about restriction before utility. That opens up a few interpretations — and they can coexist. One is operational. A more capable model is also harder to run: cost, reliability, edge cases. Restricting access buys time to make it economically and operationally viable before a broader rollout. Another is strategic. Scarcity drives attention. Positioning something as “too powerful” is an effective way to shape perception before the product is widely available. But there is a third angle that is more interesting. Anthropic has been shipping consistently, and arguably faster than most. One plausible explanation is not just execution, but underlying capability — models improving in a way that compounds. This is where Mythos becomes relevant. If part of that progress is coming from early forms of recursive learning — systems improving through feedback loops on their own outputs — then the pace of development changes. You move from discrete releases to something closer to continuous improvement. And continuous improvement at that level does not feel continuous from the outside. It feels like jumps. We saw that once with ChatGPT. It didn’t arrive as a smooth curve — it felt like a step-change. Maybe AI is starting to move even faster than we are used to — and certainly faster than most organizations can absorb. If that is even partially true, the next few months may not be linear. They may be discontinuous — in ways that feel both phenomenal and uncomfortable at the same time. #AIinPrivateEquity #AIinPrivateMarkets #LeadershipInAI https://lnkd.in/eEcvaKFBEp 752: Why Anthropic’s New Mythos Model is Terrifying and How it May Change How Business Gets DoneEp 752: Why Anthropic’s New Mythos Model is Terrifying and How it May Change How Business Gets Done
-
Márcio Tabatchnik Trigueiro compartilhou issoEm momentos onde vemos uma certa desolação com perspectivas futuras, vale ressaltar aquilo que teima em funcionar no Brasil.Márcio Tabatchnik Trigueiro compartilhou issoMarco histórico para o ITA Endowment! Atingimos a meta da nossa Campanha de Fundação - alcançamos R$ 25 milhões em compromissos de doações, um passo fundamental na construção de um fundo patrimonial sólido e perene para o Instituto Tecnológico de Aeronáutica - ITA. Essa conquista é fruto da confiança, do engajamento e da visão de longo prazo de uma comunidade que acredita no poder transformador da educação de excelência. Nosso profundo agradecimento a todos os doadores que tornaram esse avanço possível. Seguimos avançando: a Campanha de Fundação busca por 100 doadores que estabeleçam o compromisso de contribuir com R$300.000 ou mais até 2029 com fluxo de desembolso flexível. Ainda temos 35 vagas para Doadores Fundadores, que serão reconhecidos em painel de reconhecimento a ser instalado de forma permanente no ITA. Se você acredita no fortalecimento do ITA e no impacto que ele gera para o Brasil, junte-se a essa iniciativa. 🤝 Contamos também com o seu apoio para compartilhar esse marco e ampliar ainda mais essa rede de transformação. #ITAEndowment #ITA #AEITA #ITAEx #FundosPatrimonais #Endowment
-
Márcio Tabatchnik Trigueiro compartilhou issoEveryone loves music. Now you can program it with AI. I love Spotify. I also used to waste an embarrassing amount of time doing this: scroll → sample 12 seconds → skip → repeat Choosing one song at a time. How boring... Now I can do a playlist in 60s. Open your LLM of choice, and use a prompt like this for example: "Create a 2-hour playlist with classics of electronic music from top DJ's like Tiesto, Armin van Buuren, David Guetta and Above & Beyond, then list all songs formatted for Spotlistr". The key is the formatting for Spotlistr. In Spotlistr.com, login with your Spotify account, copy → paste into Spotlistr.com (Textbox tool), follow the instructions and then, magically, the playlist shows up in your Spotify account directly! Once there, you can continue editing of course; for example, for electronic music, you can sort it upwards in bpm to have a crescendo. Bottom line: Spotify gave you access to all the music there is; as usual, the challenge now is curation. Don’t let an algorithm guess what you want today. Generate the list. Automate the import. Then put the headphones on and go enjoy the music you curated. #AIinPrivateEquity #AIinPrivateMarkets #LeadershipInAI
-
Márcio Tabatchnik Trigueiro compartilhou issoJack Dorsey is a super respected tech leader, and this is what is happening at his Company today: a 40% reduction in workforce, in a single day. Why? This is the key paragraph: "we're not making this decision because we're in trouble. our business is strong. gross profit continues to grow, we continue to serve more and more customers, and profitability is improving. but something has changed. we're already seeing that the intelligence tools we’re creating and using, paired with smaller and flatter teams, are enabling a new way of working which fundamentally changes what it means to build and run a company. and that's accelerating rapidly." Block is a sophisticated Company led by a visionary leader. Initiatives like that take time to trickle down to large corporates, and then SMEs. But, how long exactly? https://lnkd.in/etNZreEF
-
Márcio Tabatchnik Trigueiro compartilhou issoCitadel Answered a Substack. That Tells You Everything About AI and Markets Right Now Three days ago, before the market opened Monday, I published a post on the Citrini Research scenario — a fictional 2028 post-mortem on AI-driven collapse. I flagged it as a thought experiment and derived three stress tests worth running regardless of whether it materializes. Then the week happened. The market reacted as if it were a forecast: IBM -13% — worst day since 2000, $31B in market cap gone overnight American Express -7.2%. Mastercard -5.8%. Visa -4.5%. The iShares Software ETF -4.8% in a session. Then Citadel Securities published a formal macro note to rebut it. Context matters here. Citadel Securities is not a bank or a fund. It is the infrastructure of US markets. It executes ~20% of all US equity volume and ~35% of all retail trades, every day. Largest designated market maker on the NYSE. Active across 50+ markets globally. 10 consecutive years as Risk Awards Flow Market Maker of the Year. $3.4B in net trading revenue in Q1 2025 alone. That firm wrote a formal rebuttal to a Substack. 1. They named the debate I've been writing about: capabilities vs. deployment Citadel's core argument: markets are conflating the recursive potential of AI with the expectation of recursive economic deployment. Capabilities can compound. Deployment follows an S-curve — constrained by infrastructure, regulation, liability, and trust. Software engineer job postings are up 11% YoY. Compute costs rise with demand. The data doesn't support non-linear adoption at work — yet. Both things can be true: capabilities accelerating AND deployment slower than the scenario assumes. That gap is where the real risk and opportunity live. 2. The signal nobody is discussing Two independents wrote a thought experiment. One of the most powerful trading firms on earth formally rebutted it with Fed data and S-curve charts. Narrative velocity now outpaces institutional credibility. A well-constructed scenario with the right distribution moves billions before anyone can fact-check it. This is the information environment we operate in now. 3. A market that moves on narrative is fragile IBM didn't miss earnings. A thought experiment went viral the same day Anthropic posted about COBOL modernization. $31B vanished. The market is migrating from irrationally bullish to irrationally bearish. Every announcement is now assumed negative. That asymmetry is worth watching — in both directions. The canary is still alive. But it has everyone's full attention. #AIQ #AIinPrivateEquity #AIinPrivateMarkets #LeadershipInAIThe 2026 Global Intelligence Crisis - Citadel SecuritiesThe 2026 Global Intelligence Crisis - Citadel Securities
Márcio Tabatchnik Trigueiro comentou uma publicação
2 sem
Sucesso para vc Lorena!
-
Márcio Tabatchnik Trigueiro gostou dissoMárcio Tabatchnik Trigueiro gostou dissoIt is with enormous joy that I share that Alucinação, the documentary I co-directed with Renato Terra and Leonardo Caetano, has been selected for the Rio Film Festival, the largest film festival in Latin America. The documentary takes as its starting point Belchior’s landmark 1976 album of the same name, but I have never seen it simply as a film about music. Nor is it merely a documentary about history, politics, or sociology. Alucinação exists precisely at the intersection of all these things. Above all, it is a documentary about the Brazil that might have been. We return to 1976 to encounter a country at the end of a dream. A Brazil that had lived through a decade of dictatorship, social transformation, modernization, repression, and extraordinary cultural ferment — and that was beginning to look at itself with the disillusionment of a country confronting the distance between the future it had once imagined and the nation it had actually become. Fifty years later, Belchior’s questions are still with us. Perhaps because Alucinação is not only about that Brazil. It is also about ours and about the distance between what we were, what we are, and what we once imagined we could become. It was a privilege to make this documentary alongside Renato, Leonardo, an extraordinary team, and Inquietude, the production company that helped turn this idea into a film. It is an even greater joy now to present it at the Rio Film Festival. ___________________ É com enorme alegria que compartilho que Alucinação, documentário que codirigi com Renato Terra e Leonardo Caetano, foi selecionado para o Festival do Rio, o maior festival de cinema da América Latina. O documentário parte do álbum homônimo lançado por Belchior em 1976, mas nunca o enxerguei simplesmente como um filme sobre música. Tampouco é apenas um documentário sobre história, política ou sociologia. Alucinação está justamente na interseção de tudo isso. Sobretudo, é um documentário sobre o Brasil que poderia ter sido. Voltamos a 1976 para encontrar um país no fim de um sonho. Um Brasil que atravessara uma década de ditadura, transformações sociais, modernização, repressão e extraordinária efervescência cultural — e que começava a olhar para si mesmo com a desilusão de quem percebia a distância entre o futuro que havia imaginado e o país que efetivamente se tornara. Cinquenta anos depois, as perguntas de Belchior continuam conosco. Talvez porque Alucinação não fale apenas daquele Brasil. Fale também do nosso, e da distância entre aquilo que fomos, aquilo que somos e aquilo que um dia imaginamos que poderíamos ser. Foi um privilégio construir esse documentário ao lado de Renato, Leonardo, de uma equipe extraordinária e da Inquietude, produtora responsável por transformar essa ideia em filme. E, agora, uma alegria ainda maior poder apresentá-lo no Festival do Rio.
-
Márcio Tabatchnik Trigueiro gostou dissoMárcio Tabatchnik Trigueiro gostou dissoWhat if something you posted at 16 cost you a job at 26? Think about how much your opinions have changed over the years. Now imagine something as simple as sharing one online leaving a digital footprint that could resurface years later, when you're applying for college, looking for a job, or simply no longer the person who posted it. With employers and universities looking at social media, students have more reasons to think twice before sharing their opinions. But at what point does being careful turn into being afraid to speak your mind? There's also the other side of the argument. If our words can reach millions, shouldn't we be held accountable for them? I was curious about this myself, which led me to explore the fine line between accountability and self-censorship, and whether cancel culture is changing what we believe or simply how willing we are to express it. And with everything we post potentially following us for years, it raises the question... How much should our past opinions define our future? #CancelCulture #DigitalFootprint #FreedomOfSpeech #StudentVoice #SocialMedia
-
Márcio Tabatchnik Trigueiro gostou dissoMárcio Tabatchnik Trigueiro gostou dissoMuito feliz e orgulhoso de passar a fazer parte de Laplace. Aos clientes e amigos, espero vê-los em breve e ajudá-los a fazer seus negócios prosperarem e seus objetivos mais relevantes serem alcançados. Renato Carvalho, Allan Libman, Daniel Assa, Guilherme Rocha e todos da Laplace: muito animado com o que vamos fazer juntos!
-
Márcio Tabatchnik Trigueiro gostou dissoMárcio Tabatchnik Trigueiro gostou dissoGustavo Fonseca, CEO da WAIKEN ILW Brasil, é o mais novo palestrante confirmado na 24ª edição do #CONAREC. 📍 23 e 24 de setembro | Transamerica Expo Center - São Paulo O maior evento de Experiência do Cliente do mundo reunirá, mais uma vez, as maiores lideranças e especialistas para discutir os caminhos do setor. Serão: + de 15 mil congressistas 500 palestrantes (dentre eles + de 120 CEOs) 8 trilhas de conteúdo em 7 arenas simultâneas 80 horas de conteúdo E muito mais! Garanta seu ingresso o e faça parte dessa imersão: https://lnkd.in/dZZx62pJ *O tema está sujeito a alterações sem aviso prévio.
-
Márcio Tabatchnik Trigueiro gostou dissoMárcio Tabatchnik Trigueiro gostou dissoDepois de mais de quatro anos, encerro um ciclo muito importante da minha trajetória no Patria Investments. Foram anos intensos, de muito aprendizado e experiências muito ricas, trabalhando com diferentes negócios, setores e empresas do portfólio, ao lado de profissionais que admiro e com quem tive a oportunidade de construir grandes relações. Sou muito grato pela confiança, pelas oportunidades e por tudo que aprendi ao longo dessa jornada. Um agradecimento especial ao Felipe Caram, Raphael Denadai Sanchez, Márcio Tabatchnik Trigueiro, Roberto Cerdeira, Fabrício Amaral e a todos com quem tive a oportunidade de trabalhar de perto ao longo desses anos. Levo comigo muito aprendizado, boas lembranças e, principalmente, grandes relações que certamente continuarão fazendo parte da minha trajetória. Obrigado a todos que fizeram parte desse ciclo!
-
Márcio Tabatchnik Trigueiro gostou dissoMárcio Tabatchnik Trigueiro gostou dissoAcompanhei desde a sua fundação o crescimento rápido e constante da Mercurio Partners no setor de energia. O empreendedorismo e a competência são a marca registrada do Alexandre Americano, do Eduardo Miranda e dos demais sócios. Estou muito feliz e motivado por me juntar a este time como sócio da holding e da gestora e como CEO da vertical de termelétricas, tendo como objetivo primeiro implantar as termelétricas de Pilar e Biribeira bem como operar a termelétrica de Paulínia. Avante Mercúrio!
-
Márcio Tabatchnik Trigueiro gostou dissoMárcio Tabatchnik Trigueiro gostou dissoOntem 12 de setembro de 2026 meu pai, Ernesto Reibel, teria completado 100 anos. Ele queria muito chegar a essa data. Chegou perto — partiu aos 98 — depois de uma vida longa, plena e de enorme impacto na vida de tantas pessoas. Não conseguiu celebrar os 100 anos como gostaria. Mas construiu algo muito maior do que esse marco: uma história, uma família, amizades, valores e um legado que continua vivo em todos nós. Ontem foi dia de lembrar, com muita saudade e gratidão, do privilégio de ter convivido com um homem tão especial.
-
Márcio Tabatchnik Trigueiro gostou dissoMárcio Tabatchnik Trigueiro gostou dissoRecentemente completei 11 anos no Pátria Investimentos. Foi uma caminhada muito vitoriosa. Agora, no entanto, terei um novo desafio à minha frente. Por isso cumprimento a todos os sócios e colaboradores na pessoa do André Sales e agradeço todo o apoio recebido. Que o Pátria continue a sua trilha de muito sucesso. Um grande abraço a todos e um abraço especial aos colegas da Arke (UTE Marlim Azul).
Experiência e formação acadêmica
Experiência de voluntariado
-
Mentor and Adviser
Endeavor Brasil
- o momento 24 anos 8 meses
Empoderamento econômico
Endeavor is a non-profit institution devoted to foster entrepreneurship. As an adviser, I participate in the selection process validating entrepreneurs and their business plans while as a mentor I provide free consulting to entrepreneurs that face strategic and/or management issues.
Idiomas
-
English
Nível avançado
-
Spanish
Nível avançado
-
Portuguese
Nível nativo ou bilíngue
Organizações
-
Young Presidents Organization (YPO)
-
- o momento
Outros perfis semelhantes
Ver mais publicações
-
Kit Yu
Globally, small and medium-sized enterprises (SMEs) are navigating trade disruptions, inflation and interest rate pressures along with technological shifts. While SMEs may be more vulnerable in the current economic environment, their agility and focus could position them to adapt and thrive. MEI finds that tariffs are disproportionately affecting U.S. SMEs as compared to larger businesses. Based on Mastercard’s aggregated and anonymized data, U.S. spending at SMEs softened in mid-2025 after the announcement of the spring tariffs, while demand at large businesses was more stable. This divergence is particularly evident in import-sensitive sectors such as consumer electronics, toys and apparel; 2025 saw a widening in the spending gap for these sectors relative to 2024, with small businesses exhibiting deeper underperformance compared to large businesses. For example, MEI found that annual spending growth at SMEs was 20 percentage points below that of large companies in the consumer electronics sector in 2025, compared to just 6.6 percentage points in 2024. In these tariff-sensitive goods sectors, SMEs do not benefit from the economies of scale that allow larger firms to shift supply chains or front-load inventory. Conversely, SMEs that benefit from niche demand, such as florists and specialty food stores, have experienced stronger spending than their larger counterparts; consequently, spending at these SMEs continued to outperform larger retailers in 2025. Looking ahead, MEI is optimistic that broader technological access and adoption will buoy small businesses. Digital tools are increasingly helping SMEs streamline operations, reduce costs and compete more effectively. The rise of online-only SMEs illustrates this shift: According to Mastercard’s aggregated and anonymized data, 44% of new card-accepting U.S. businesses in 2024 were online–only, up 20 percentage points since 2019 (classified as businesses receiving only ecommerce transactions at the start of operations). Services-related businesses lead this transition; for example, over 85% of new travel agencies and professional service firms launching as digital-first operations.
-
Michael Esrubilsky
LAVCA released its 2026 report on venture capital activity in Latin America. We were happy to see Kiara Capital mentioned among the active seed investors in Brazil. While the report places us in the Brazil category, our work goes beyond that. We invest across other countries, as long as we can support founders rethinking how finance is built and who are within our trusted networks. Early-stage investing in these markets requires patience, local depth, and global perspective, especially in tighter cycles. Seed today is also very different from what it was a few years ago. Standards are higher, and capital is more intentional. We remain focused on backing the next generation of finance players, many of them using AI to rethink financial systems. Grateful to the founders and partners building alongside us. The work continues! #KiaraCapital #VentureCapital #EmergingMarkets #Finance
-
Steve Monnington
My monthly Dealmakers column is now available to subsribers of Flashes & Flames: The Global Media Business Weekly. Four transactions in UK, China, India and Brazil plus who is building the future while we're buying the past? - a look into the M&A demand and supply conundrum and what companies like Manta Media Capital, Events Venture Group and SCALE London are doing to support founders. https://lnkd.in/ejR-v6dw