
Inside General Atlantic: How a $100B Growth Equity Firm Invests
What this covers
Patrick O'Shaughnessy talks with Martín Escobari, head of global growth equity at General Atlantic, about 25 years of investing across 19 countries. The anchor is Spear Fishing Investing. Escobari's example is the Brazilian trio Lemann, Telles, and Sicupira, who waited five years for Brahma and then closed in a week. He also describes the dot-com years at Submarino, when roughly seven to ten years of work was compressed into one. The firm has backed over 500 companies in 45 years, about half outside the US.
The middle of the conversation covers structure and risk. Escobari criticizes the five-year fundraising cycle for leaving firms short of dry powder when assets go on sale, and he explains why GA uses a hybrid evergreen vehicle. He says the firm's loss ratio is about 4%, against 20-40% common in venture and growth, because scenario planning means a worst case is only growing into the price paid. He also makes a contested case for diversifying away from the US. He cites 26 times earnings for US equities against 9 to 14 times in Brazil, Mexico, and Europe, and US debt at 125% of GDP.
The last stretch turns to AI and to judgment. Escobari expects a moment of regret over over-investment within 18 months to three years. He says code generation is the one area where GA has felt a pounce moment. He resolves two mentors' conflicting advice, build a checklist versus refuse one, with Educated Intuition. He also covers reference checks run through investing families, learning by predicting investment committee votes, and mentoring.
Escobari argues that General Atlantic's durable edge comes from a partnership culture, patient 'spear-fishing' capital allocation, and risk management built around avoiding binary losses, and that the current AI boom is healthier than past bubbles because incumbents fund it.
- Patient waiting for once-in-a-generation opportunities, then moving fast, creates disproportionate value (Brahma, Sumarino, GFC deal examples)
- A partnership-based compensation and fundraising structure keeps teams collaborative and makes Latin America and other emerging markets work
- The AI investment wave is funded mostly by highly profitable 'magnificent' companies, unlike the railroad and dot-com bubbles
This asset isn't compiled yet
You're seeing its claims, ranked. Compile it to build the argument threads, weight them, and check each claim against your library — the full view.
The traditional five-year fundraising cycle creates distortions: firms must deploy at a set speed and return capital on a cadence to raise the next fund, so during a fundraising winter they run out of dry powder just when assets go on sale. GA's hybrid evergreen structure, with a new fund every two to three years plus managed accounts, removes these fundraising cliffs.
“you're out of dry powder exactly at the time that things are on sale.”
The AI wave is healthier than past bubbles because its capex is funded by highly profitable incumbents (the 'magnificent six') reinvesting cash from dominant positions, unlike the railroad and dot-com eras when junk-bond speculators and thin-margin telecom companies levered up with retail money; capex to revenue remains not crazy, so the wave has more legs even though it will be bumpy.
“the funds are coming from really rich companies the magnificent six who are printing money out of their dominant positions are reinvesting”
Escobari argues that every bubble is born from a truly transformative technology: the promise is spectacular, short-term results disappoint, and long-term results exceed expectations, while many fortunes are made and destroyed along the way.
“all bubbles are born out of a truly transformative technology. And in all the previous bubbles, the promise was spectacular. The short term was disappointing and the long term delivered more than expected.”
GA's AI investing is deliberately slow because it is unclear where value will be created: whether large language models will be more powerful than efficient alternatives, or how much value will accrue to models versus applications.
“Our approach which may prove to be too conservative was to take it slow because it's not clear yet where the value is going to be created.”
Outside the US, valuations are much lower: Europe trades at 14 times earnings, Brazil at 9 times, and Mexico at 10 times, and many 40-50% growers trade at roughly 12 times forward earnings in India, so the case for global diversification is the strongest it has been in decades.
“you can buy Europe at 14 times earnings, you can buy Brazil at nine times earnings, you can buy Mexico at 10 times earnings.”
Escobari believes the AI market has not yet reached a bubble peak, and he expects a moment of regret about over-investment, perhaps 18 months to three years away, because the value capture between large language models, efficient alternatives, and applications is still unclear.
“I know we're going to have a moment where we're all going to wake up and say we've invested too much. I don't know if it's three years away or 18 months.”
Code generation is the first AI area where GA felt the pounce moment, because it produced real revenue within the last 12 months; Escobari cites public figures that Anthropic's revenue rose from about $200 million to over $4 billion in that period.
“anthropic revenues and coaching went from 200 million to over 4 billion in 12 months”
Operating in the dot-com era compressed roughly seven to ten years of normal work into one year: Submarino opened in six countries with warehouses, customers, and teams within a year, when a normal timeline would have been three to four years, and this kind of compression comes from the need to be first mover in fleeting opportunities.
“You can do seven years of work in one year.”
General Atlantic's partnership culture is what makes its global model work: the firm has invested in over 500 companies in 45 years, about half outside the US, and its culture expels behavior inconsistent with being a good partner, which is why teams in Latin America, China, India, and Southeast Asia can compete for capital within a global fund.
“It's so hard to do unless your culture is about partnership. The culture demands good partnership.”
GA's compensation is pooled: partners receive a percentage of total firm performance rather than individual performance, which Escobari initially disliked but says increases collaboration; to prevent free riding, anyone not pulling their weight is off the boat, making the system a meritocracy of team wins and losses.
“we have a communist system of compensation which is you all get a percent of the total performance not your individual performance.”
Investment professionals learn best through apprenticeship: Escobari predicted each investment committee member's questions and votes by reading memos, reaching 80-90% accuracy after a year, which forced him to adopt the perspective of proven investors and learn vicariously.
“By the end of a year of doing this, I had I was up to 80 90%.”
Winning in a lottery-like portfolio depends on the capital allocator at the top: even when an opportunity is available to many, only a spear fisherman, who can seize it, captures the value, so some people can do this and others cannot.
“there was a spear fisherman at the top to capture an opportunity that was available to many, but they seized it.”
A good reference check should be run through a family that has invested with you, since they will tell the truth about a founder's trustworthiness because their money is at stake; for hiring, reference calls should describe the specific challenges of the role honestly rather than asking generic questions about whether the person is a good guy.
“Because he has money with you. He tells you the truth.”
Having 32 case studies in a sector does not guarantee a copyable edge, because competitors can copy what worked but cannot see the failed experiments that GA learned to avoid. Scale and experience help only when firms are deliberate about learning from them and building capabilities in areas that matter.
“you don't know the things that didn't work that we tried that we're going to prevent you from trying”
The best investment evaluation combines a checklist with gut feeling, which Escobari calls 'educated intuition': after completing a checklist of huge TAMs, sound business models, strong management with go-forward capabilities, inorganic growth opportunities, and strategic value to acquirers, the investor closes their eyes and checks their gut, because a super-interviewer who did this consistently outperformed.
“the framework I use for what's the perfect martin or g ide deal is the combination of a checklist with my gut which I call the educated intuition.”
The best investors act like spear fishermen: they wait patiently for the big fish, ignoring small opportunities, and strike quickly when a once-in-a-generation opportunity appears. Brahma's owners (Lemann, Telles, Sicupira) exemplify this by waiting five years to buy the beer company, closing the deal in one week when the Swiss owners got scared before an election.
“You let little fish go by because you you're not there to hunt little fish. You're waiting for the big fish.”
Steve Denning advised Escobari to build a checklist of characteristics of winning GA deals, while Dave Hodgson said to refuse the temptation of a checklist; Escobari resolved the paradox with an 'educated intuition' that combines a checklist with gut feel.
“the framework I use... is the combination of a checklist with my gut which I call the educated intuition”
US total debt is 125% of GDP, the highest in the OECD and above post-World War II levels, and under current plans it will reach about 145% of GDP within five years, higher than Greece and Italy, which Escobari cites as a reason to limit US exposure.
“total debt to GDP is 125% of GDP. That is the highest of the OECD.”
The US equity premium is at a historic extreme: US public equities trade at 26 times earnings for about 4% forecast growth, at the 97th percentile of the last 25 years, while the dollar is about two standard deviations from neutral, so Escobari argues investors should hold far less than 90-95% of assets in the US.
“US public equities are trading at 26 times earnings for a 4% forecasted growth which is at the 97th percentile of the last 25 years.”
Avoiding binary risk means in scenario planning a company's worst case is that it grows into the valuation the firm paid, which limits how early the firm enters new industries; the firm leaves some money on the table but also leaves much more risk on the table, and the combination of reasonable returns with low risk is a superior product.
“that product of you know reasonable returns with low risk is a great product”
Returns in venture and growth equity are highly skewed: about 10% of the best deals produce around 50% of returns and 5x-plus outcomes, while losses are small; the best outcomes are better than the upside case in the memos because unforeseen good developments occur, so the team actively looks for lottery tickets.
“10% of our best deals and we get lucky and they produce 50% return. So we lose money very little and then 10% we get 5x plus”
Persuasion equals desire minus fear: according to John Kim of General Catalyst, fundraising success depends on how strongly investors want the opportunity relative to how much they fear losing, and most investors swing from FOMO to fear, so fundraising is hardest when markets are cheap and fearful.
“persuasion equals desire minus fear”
General Atlantic's loss ratio on capital is about 4%, far below the 20-40% loss ratios common in venture and growth equity, and this is achieved because the firm avoids binary bets and models scenarios so that a worst case means a company only grows into the valuation paid.
“Guess what our loss ratio is? 4%. On capital or on capital”
Having no Latin America fund helped General Atlantic do well in the region, because a dedicated regional fund would pressure managers to deploy money at market tops and sell at bottoms, whereas the global structure lets them buy at the bottom and sell near the top.
“we don't have a Latin America fund. because if we had a Latin America fund we're going to put money in Latin America. We're going to buy at the top but sell at the bottom.”
GA's investments span 18 'power alleys' across five sectors, and its loss ratio on capital is about 4%, against 20 to 40% loss ratios common in venture and growth equity.
“Guess what our loss ratio is? 4%. On capital. for this kind of like when in venture and growth equity loss ratios of 20 to 40% are common”
Chinese entrepreneurs in their 30s and 40s are children of the Cultural Revolution, whose families lost everything, so they are scarred and driven to recover what was stolen, producing a work ethic comparable to post-World War II refugees who built businesses in the US.
“they are scarred and they have something to prove because they think something was stolen and they will get it back.”
The most driven entrepreneurs are often driven by foundational trauma; understanding a founder's trauma helps investors empathize with them and assess whether they can manage and channel the intensity productively, which Escobari calls a wonderful engine of transformation and a form of healing.
“I find a lot of the most driven people are driven because of foundational traumas.”
GA trained an AI 'IC robot' on 45 years of deal data as a sixth investment committee member who votes on every deal; backtesting shows it is much better than humans at the past, but it has only about three years of concurrent data, so it will need four or five more years to be proven against live decisions.
“We've been training this sixth member of the IC based on 45 years of data. So she votes on all our deals”
Escobari and his mentor Don Soul ran a paired-company study of Brazilian firms in the 1990s: 10 highly successful companies were compared with 10 much less successful peers that had similar size and value in 1990, and by 1999 the winners were at least 5x more valuable, which showed what the winners did differently.
“look at 10 very successful companies in Brazil that had made tremendous wealth creation in the crazy '90s”
Investing outside the US involves three differences: more volatility and surprises, so agility matters; low-trust cultures where people lie more often, so rigorous referencing is essential; and enormous low-hanging fruit because many services work poorly, so a great service captures value for a long time.
“we're also low trust cultures... So the value of referencing is much more important.”
During downturns, the best fundraising move is to go to parties that still have excess capital, and GA cultivates entrepreneurial families in each geography for their insights rather than just their money.
“there's always someone in the world that has excess capital even in a time of fear and you go there”
Chuck Feney, General Atlantic's founder, built a duty-free business after noticing commerce was free at Pacific naval bases during the Korean War, then concluded that the purpose of wealth is to improve the human condition now rather than later, and planned to give away his fortune, which shaped GA's mission to back innovators and direct proceeds to good causes.
“The purpose of wealth is to improve the human condition now, not tomorrow, now.”
Investment careers are an apprenticeship business: firms should pair junior investors with different mentors, have them make recommendations from an early age, and push them to answer the 'so what' and state their level of conviction rather than just completing tasks.
“It's an apprenticeship business. So, pairing them up with different people with different skills, different styles is super important.”
A dominant fixed-income exchange platform in Brazil, with about 80% margins, was bought at six times its earnings during the 2008 financial crisis, when competitors dropped out. Escobari argues that a dominant platform is always worth more than six times earnings, so panic is an opportunity to double down rather than shut down.
“we were able to buy a market dominant high margin business that's six times”
GA's founding documents set a partnership ethos as foundational, and the firm's reputation for being good partners to founders and clients differentiates it, even though some competitors call it naive, with Escobari embracing the 'dolphins in a sea of sharks' identity.
“We're good partners to each other to our founders and to our clients. The partnership ethos is fundamental.”
Risk, which can be quantified or imagined, differs from pure uncertainty; Escobari says he never invests a lot without visibility, that fog has layers that slow you down but can be pierced to reveal clarity, and one should engage with unpredictability until one sees an opportunity before others.
“there is fog and different layers of fog. And what the fog does is it slows you down, but if you look pierce through, you see clarity.”
Investors who make their money in Brazil are massively underdiversified: when asked what percent of their net worth they hold in Brazil, wealthy families typically answer 90-95%, but when asked to imagine being a citizen of the world, they answer about 3%, so GA uses that gap to persuade them to diversify.
“the number is typically between 90 and 95%. ... and they're like, oh, 3%.”
General Atlantic holds open investment committee meetings for the whole firm every Tuesday, with about 190 people signing up, where there are no presentations; the deal lead answers questions and five IC members plus an AI 'IC robot' ask questions, which builds culture and shared judgment.
“our investment committees are open to everyone. ... every Tuesday 190 people sign up and there's no presenting.”
Mentoring is most effective when it is uncomfortable: Escobari goes straight to a mentee's sensitive points, scathingly showing that an obvious opportunity is being missed, because with limited time (sometimes only two sessions through Endeavor) mentoring must be like crash therapy, compressing years of therapy into an hour.
“It's like crash therapy like three years of therapy in one hour.”
Every four or five years there is a once-in-a-generation opportunity, and investors must be ready and willing to move quickly to capture it; if they do, they can create disproportionate value for their company, investors, and employees.
“every four or five years there's a once in a generation opportunity that you have to be ready and be willing to move quickly to capture”
GA has deliberately deployed AI aggressively across its portfolio: it runs about 500 portfolio projects this year, a third of them AI projects, and it pounces on use cases once it sees real ROI and revenue to the service provider, using economic models of cost-to-serve and long-term profitability.
“This year we'll do 500 projects with the portfolio. A third of them are AI projects.”
In a 1990s Brazilian beer-industry comparison, companies matched for size and value at the start of the decade diverged by at least 5x by decade's end. Brahma's shareholders ended up holding 95% of the equity value of the merged company, about 10x the value creation of its comparison twin, Antarctica.
“by a decade later when these two merged uh the Brahma shareholders kept 95% of the equity value.”
People tend to overinvest in what they understand, which leads to home-country concentration; because the world has become riskier, the one free lunch in finance, diversification, is especially valuable, and diversifying strategically (across whom and where) is hard but worth it.
“people have a natural tendency to overinvest in that in which they understand”
Brahma's investment compounded from an initial $80 million into a value of more than $60 billion (excluding dividends) because the owners repeatedly waited for the right moment across decades, including buying Antarctica in a distressed devaluation within three months and closing the largest deal a decade later.
“basically over this period $80 million initial investment in Brahma became a $60 billion plus excluding dividends”
Roughly every three or four years a unique market distortion appears that demands very fast action, and investors who hesitate during panic miss it. The panic is usually wrong about the end of the world, since a dominant platform remains valuable.
“there's been you know every every three or four years there's one such unique distortion that you have to move very quickly”
GA's competitive edge is its partnership brand plus scale in operations: a 100-person portfolio operations team, a 15,000-executive talent database, and 32 case studies across its 16 power alleys let it help portfolio companies with pricing, sales effectiveness, AI, and hiring, and experience also helps as long as the firm learns deliberately from failures.
“We have 100 people in operations. ... an in-house human talent team that has taps into a database of 15,000 vetted executives”
General Atlantic is structured so employees hold about 8% of the funds it administers and over $5 billion of its own capital, so the firm treats its investing as managing its own family wealth rather than other people's money.
“employees of General Atlantic have about 8% of the funds we administer over $5 billion of our own capital.”
Predicting the future is very hard, while explaining the past is much easier, and Escobari distinguishes between quantifiable risk and pure uncertainty when deciding how much of his success came from embracing uncertainty.
“the thing about predicting the future is it's really hard. Explaining the past is a lot easier.”
Being a great investor requires patience, which younger investors typically lack: young investors want deal experience and notches on their belt, but older investors have seen more, can wait for the big fish, and feel calm because a big opportunity will come.
“The hardest thing when you're young is developing patience and the conviction that you can wait a little longer for the big fish.”
The most impactful time to mentor an entrepreneur is at their darkest moment, because a mentor who has been through a crisis still has enough light to help others and that gives the mentor strength to get through it.
“it is precisely at your darkest moment that you mentor because it's a sign of you have something to give”
Escobari's own risk orientation comes from childhood trauma: growing up in Bolivia in the 1980s with seven presidents in ten years, four coup attempts, and inflation of 35,000%, plus a genetic bruising disorder since age five, made every decision a risk-reward tradeoff.
“Bolivia in the 80s seven presidents in 10 years including four coup detas. We had inflation now you got 5% inflation. We had 35,000% inflation.”
Being an older investor shifts the role from front-line dealmaking to coaching young partners, which can be demotivating until the investor reframes success and enjoys colleagues' wins almost as much as their own.
“you are less in the front lines and more as a coach player helping you, training young partners”
Escobari recommends that people in their 20s or early 30s (or anyone with a young mental age) go work in AI now, because it compresses roughly seven years of learning into one year, and such compressed learning happens only about once every 20 years.
“Unambiguous recommendation. If you're in your 20s or early 30s, go work at AI because you're going to live through dog years.”
Competition in private markets has intensified: the number of $100 million-plus revenue businesses growing 40-50% has grown tenfold, and GA competes against roughly 19,000 GPs, so edge now requires clarity and specialization in named 'power alleys' with scale and case studies.
“we compete against 19,000 GPS”
Heart is essential for leading large organizations because leaders must move the hearts of hundreds or thousands of people to row in the same direction, and heart can override gut and brain, but few investors combine the checklist (mind), gut, and heart, and the best leaders are rare wizards of the trade.
“It's really hard to do all three.”
GA has invested in over 500 companies over 45 years, with half of those investments outside the United States, and it is a platform for global innovation rather than a US or European privilege.
“We've invested in over 500 companies over the last 45 years. Half of our investments have been in the United outside of the United States.”
Investors should not fall in love with deals because they have a fiduciary duty to produce returns and must make tough calls, and love is a treacherous thing; Escobari's marriage is the one case where he ignored his checklist, and he says that in matters of the heart you should forget the checklist.
“part of being a good investor is to not fall in love because at the end of the day you have a fiduciary duty to produce returns”
Few venture and growth equity investors remain active past age 50 because their brains age and stop being plastic, not just because they get rich or tired; Escobari argues that staying young requires refusing to think like an old man, continuing to wonder, and treating learning as play.
“I think our brain ages and stops being plastic.”
China has undergone the fastest change in per-capita GDP in modern history at scale, and GA has been underweight China for five years but recently did two deals and plans to pick it up, because geopolitical risk is binary but Chinese innovation and entrepreneurial zeal are very strong.
“It's the fastest change in terms of per capita GDP in modern history or any country at at scale.”
GA's two structural trade-offs are that its unusual dual-series fund structure takes a very long time to explain to investors, and that fundraising is a perpetual activity rather than a periodic sprint; Escobari says this is still a no-brainer because it leads to more productive capital deployment.
“fundraising is a perpetual activity. Whereas for uh a lot of my competitors every five years it's a six-month sprint”
Escobari argues that the new AI coding reality means human programmers become hyperproductive while working alongside agentic programmers that lack a moral north star and do not sleep, so the central challenge is getting them to produce a coherent output for clients.
“human programmers are hyperproductive and they're working alongside agentic programmers who have no moral northstar and do not sleep”
Bill Ford rejected Escobari's 1998 pitch to invest in Brazil with 'we're not ready for Brazil', and ten years later, after Escobari had sold his business, Ford reached out to try again; Escobari notes that GA would have made 18 times its money had it said yes earlier.
“you would have made 18 times your money if you had said yes”
Escobari keeps roughly 95% of his net worth in General Atlantic and holds treasuries for the rest, which he describes as a deliberately undiversified portfolio that still lets him sleep well at night.
“I have 95% of my net worth in that product and I sleep well at night.”
Managing emotions productively requires therapy, journaling, or meditation, and one should do two of the three.
“Managing one's emotions productive requires either therapy, writing a journal, or meditating. You should do two of the three.”
The most powerful way to start a conversation with someone you want to know is silence: a therapist who stays silent prompts the other person to speak, and asking someone what the most important question they need answered from the universe is liberating.
“The most powerful way to Start a conversation with someone you're trying to get to know. Silence.”
To stay current, Escobari recommends talking to young people, surfing TikTok, trying different apps, and going to places without older people, because value lies in the new even though most experiments turn out to be dead ends.
“Talk to young people... Surf Tik Tok. Try different apps.”
The US higher education system has lost its way, and Escobari wants to help restore it because the Harvard scholarship he received as a Bolivian kid had a transformative impact on him.
“I think we've lost our way.”
During the dot-com bubble, a 'bubble' can feel better from inside than outside, so Escobari left his fund quickly and raised $80 million within about three months to launch Submarino.com, an e-commerce venture that merged with Amazon and Alibaba's efforts to compete in Brazil, showing that bubble periods reward moving very fast.
“within I think three months we had raised $80 million to launch subarino.com”
Escobari's investment selection criteria include huge TAMs, business models that create economic value, strong teams with forward capabilities, inorganic growth opportunities, and strategic value where acquirers would overpay; he also requires that companies make the world better.
“Huge TAMs, business models that create economic value and have mode teams with the right go forward capabilities”
Over the next 10 years, investors who achieve some level of diversification will be rewarded, because US markets have some froth while many emerging markets are at the opposite extreme.
“those who achieve some level of a diversification will be rewarded because I do think there's a little bit of froth in the US market”