
How to bet on yourself (without venture capital)
What this covers
The conversation is with William Hockey, co-founder of Plaid and founder of Column, who funded Column himself. It opens in emerging markets, including his time in Kinshasa. He points to Kaspi in Kazakhstan as a template: buy a bank, then cross-sell e-commerce, tax payments and licence renewals to customers who already hold accounts. He argues that financial services leapfrog legacy infrastructure in the most constrained countries, such as Iran and Argentina, and that emerging-market management is often stronger than Westerners assume.
The middle of the discussion is a case against the venture model. Hockey describes a venture treadmill in which constant fundraising pushes founders toward trends like stablecoins and AI. He says the model assumes a company behaves like an ATM, and that it needs growth above 30% from a base of a billion dollars or more. He also claims founders lose 50 to 75% of their equity to dilution, plus a share of their upside to the preference stack. His contrarian positions are that founders take too little risk, that early employees carry more risk than founders, and that Silicon Valley's elite consensus encourages drinking your own Kool-Aid.
The last third turns to geopolitics and AI. Hockey treats the dollar as a national security tool: roughly 75% of trade is in dollars, and even Qatari gas sold to Switzerland clears through US institutions. He describes sanctions as a substitute for military force, citing Venezuela. He also defends Federal Reserve rails against crypto. On AI, he forecasts that value will go to incumbents with distribution, as it went to Standard Oil rather than the railroads, with legacy banks as likely winners.
Hockey argues that Column's advantage comes from owning a bank charter while building software, and that founders can build durable companies by staying in boring niches, funding growth from earnings, and sharing ownership with employees rather than following the venture-capital playbook.
- A bank charter creates a regulatory moat that software-first competitors cannot replicate, and roughly 90% of Column's money comes from software per-API-call fees
- Venture funding forces short-term fundraising cycles and trend-chasing, while earnings-funded growth with employee ownership supports long-term building
- Deep study of obscure niches yields disproportionate value, and the US financial system's dollar dominance gives the US unusual national-security leverage that policymakers and technologists underrate
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.
Kaspi in Kazakhstan shows the emerging-market playbook: buy a bank, then use its everyday account relationships to cross-sell e-commerce, tax payments, licence renewals, and other products, because nearly everyone already has a bank account.
“they started out by buying a bank. And then they just kind of built did everything.”
AI's economic value will accrue mainly to companies with massive distribution, brands, and cost bases that harness AI, not to AI labs, similar to how railroad value accrued to Standard Oil rather than to railroads themselves, with the biggest, least efficient brands likely the largest AI beneficiaries.
“the value actually accrued to the people that could like harness that the best... The Standard Oil was the biggest beneficiary”
Dollarized emerging markets, where the US dollar is the main currency officially or unofficially because local central banks are not trusted or hyperinflation has occurred, need US financial services more than countries with strong local currencies such as the UK or France.
“those places tend to actually need US financial services more than, I don't know, you know, UK and the GBP is pretty strong or, you know, France.”
Margin lending against private company stock is a poor business, because lenders end up holding collateral precisely when the private stock is hardest to sell.
“margin lending in private companies is not is is like not a great business because when you want to take collateral that's stock, that's usually the exact time when you do not want to hold that private stock.”
Silicon Valley has become an elite-dominated society resembling 1990s Wall Street more than a 1950s research lab, and elites building software for elites leads founders to drink their own Kool-Aid by talking only to each other.
“elites end up building software for elites... we talk to each other, we build for each other, and we think that like the market is each other, but we don't actually look broader than that.”
Financial services innovate most in the worst-constrained countries, because cut off from global markets they must build bespoke systems from scratch, as in Iran and Argentina, and leapfrog legacy infrastructure.
“financial services tend to be most innovative and most progressive in like the worst countries. You can see this in Argentina, you can see this in Iran”
Brains are distributed equally, so emerging markets have a decent talent pool, and the best talent there goes to where job safety and money are, which often means large banks and brewers, with middle and top management often stronger than in the West.
“if you believe that, you know, brains are distributed equally, you go to Congo there's going to be like some proportion of like equally smart people as there are in France.”
The venture model is built on the assumption that a company works like an ATM, where a dollar invested returns a dollar and thirty cents, which is not true of most markets and businesses.
“the venture model is built off a fact of you can make a company like an ATM, which is you put in a dollar, you get a dollar 30 back. That's not the case with most markets.”
Payment friction in banking, such as slow transfers, is a feature rather than a bug, because it protects the 5 to 10% of consumers vulnerable to fraud, and AI fraud detection could remove that friction for everyone else.
“All the friction is actually built to protect the 5 to 10% of consumers that can't get hurt. ... we can actually take that tail away and it actually is massively beneficial for everybody else.”
The venture model depends on companies that can grow above 30% a year off a billion- or two-billion-dollar revenue base, which is far harder than growing from zero to $100 million, so founders should check whether their margin profile and growth match that model before raising venture money.
“Going like above 30% a year off of a billion-dollar revenue base. That is like 99 times harder than going to zero and 100 million in three years.”
The US financial system's core infrastructure is strong, and the Federal Reserve can move money 24/7 faster than stablecoins; the bottleneck is implementation by business models such as community banks, not missing technology.
“The problem isn't in the fundamental infrastructure. It's in our implementation of it.”
The US financial system is a national security weapon that financial institutions should openly acknowledge their role in, and the Federal Reserve system's infrastructure is far better than commonly believed, able to clear money 24/7 faster than stablecoins or crypto.
“The Fed has a pretty good tech team. ... the US right now through the Fed has a capability to move money and to clear money through all of these institutions 24/7. Faster than stable coins, faster than crypto.”
Software companies often cannot absorb large amounts of capital productively, so raising a billion dollars would not necessarily accelerate growth, and earnings should be split between employees, growth, and a war chest for bad years.
“if you give me like a billion dollars right now, I don't know if I grow it like, you know, a thousand times faster than I'm right now.”
Financial services should be understood as a national security weapon, and banks and fintechs should see themselves as part of that strategy in the same way that defense contractors such as Palantir, Lockheed Martin, and Boeing are.
“financial services is like a key pillar. Like you can argue that financial services is like the first in to wage a country”
Early-stage founders typically lose 50 to 75% of their equity value to dilution and may lose 10 to 80% of their upside to the preference stack, and these economics are poorly understood until founders have gone through several rounds or an exit.
“you're going to lose probably 50 to 75% of your equity value due to dilution alone. You may lose 10 to 80% of your upside by the preference stack.”
Sanctions are a substitute for military force: the US can enforce its will by collapsing a target's trade and oil exports, which reduced the risk and cost of intervention in Venezuela, where the economy was already destroyed.
“they want to use a sanction before they use a missile. The great thing about a sanction... you can enforce American dominance without putting anybody at risk.”
Traditional banks are among the largest inefficiencies in the economy and are hard to take over because they are heavily regulated, so the banks that harness AI effectively will benefit most, since their costs sit in headcount and technology rather than physical assets.
“I think financial services and like legacy banks tend to be like the largest inefficiencies out there. They also are very hard to take over.”
Because venture-funded companies must raise every year or year and a half, they optimize for the next fundraise and chase trends like stablecoins or AI, producing a windy path to the goal.
“you end up optimizing for that next fund raise. And you say, Okay, stablecoins are like cool this year. Let's do a stablecoin strategy.”
Roughly 75% of global trade is still conducted in US dollars, and this dollar dominance is an underappreciated source of American soft and hard power.
“like 75% of global trade still is in the dollar.”
Venture capital works like a dependency: as with heroin, the money feels good but founders must keep raising, and raising a $100M Series A should be enough to build a multi-billion-dollar business without further capital.
“VC money's kind of like heroin. Like it feels good. It's amazing. But like you got to keep shooting up.”
The dollar links the world far more than official statistics show: official data says most economies trade under 10% of GDP with neighbors, but unofficial trade through non-government rails is closer to 90%.
“If you look at the official stats, um like like most economies spend less than 10% of their GDP trading with their neighbors. But when you actually dig into it, like the unofficial trade that's going like not over like government rails, it's actually closer to 90%.”
Qatar gas sold to Switzerland is denominated in dollars and clears through US financial institutions, and China's large purchases of Russian oil and gas are still mostly dollar-denominated, showing how the dollar is embedded even in trade between countries that dislike the US.
“That trade is denominated in dollars. That the money that's moving from Switzerland to Qatar... crosses through US financial institutions.”
Early-stage employees take far more risk than early-stage founders, because founders can usually raise another round or sell secondary, while an employee of a failed startup gains little resume value, so the risk asymmetry should be reduced for founders rather than employees.
“the weird thing is an early-stage employee takes way more risk than an early-stage founder.”
The best founders find the most boring problem in their field interesting and stay interested for decades, because value sits in extremely boring niches that cannot be researched through an AI deep-research tool, while popular topics like AI and geopolitics are crowded with thousands of people.
“I think one of the best determinants for success of founders is can they find the most boring thing humanly possible interesting? And can they find that interesting over a multi-decade period?”
Community banks cannot offer 24/7 money movement not because the Fed lacks the technology but because staffing a rural bank on weekends is impractical, so the binding constraint is implementation and business model, not underlying infrastructure.
“The reason why community banks can't send money 24/7 isn't because like the technology doesn't exist at the Fed, it's because there were constraints in those business models”
Although Column is a licensed bank, about 90% or more of its money comes from software, priced like SaaS per API call, and it passes most bank-side economics down to its customers.
“unlike banks, we make 90 plus percent of our money off of software. And so, similar to a any kind of SaaS company, it's a per API call.”
Buying a regulated bank during the first Biden administration was a non-consensus bet that required two to three years without a revenue focus, something no venture investor would fund because it is not a software-defined problem.
“we bought a regulated bank during the the first Biden administration and that required us like not grow or like not focus on revenue for like two three years”
Hockey funded the purchase of Column's bank largely with debt: he pledged over $1 billion of Block stock to borrow about $70 million at SOFR plus 10% and roughly 5% loan-to-value, was margin-called three times, and nearly went bankrupt multiple times.
“I pledged over a billion dollars of stock to get $70 million. I bought the bank for $70 million.”
In competitive markets, the path to durable value is to become the world's leading expert in a small, boring niche, because generalist topics like AI disruption of software are crowded with thousands of informed commentators who cannot capture value.
“you can't create value there. You can create value if you're like, I'm the number one person in the entire world at this little niche thing”
Hockey studied a 2,000-page book on 19th-century Chinese banking and a history of Japanese banks, and found a single idea there that became a product, showing that deep study yields rare leverage: one small insight can be worth millions or hundreds of millions of dollars if you own a business in that niche.
“you probably get like one small thing in a 2,000-page book. And so it's probably like not efficient unless you own a thing that happens to be in that space.”
Fear and the absence of a fallback option create creativity and intensity in founders, and removing risk from starting companies, through accelerators and easy seed funding, has produced many companies but few great founders.
“in that fear, in that like in innate desire creates like another part of you. It creates creativity, it creates inspiration.”
Silicon Valley, along with Beijing, is one of the most consensus-driven places in the world, and that consensus environment is both a strength (it lets founders pursue outlandish ideas that insiders already believe in) and a blind spot (it makes the tech industry lose touch with how everyday Americans and the rest of the world live).
“San Francisco is probably the most consensus place I've ever been to. And I think that is both a huge clutch for us, but it's also probably our most valuable asset.”
Emerging-market founders work under material constraints that breed a different kind of creativity and innovation than the abundance of San Francisco, London, or New York.
“they're constrained in a way that like San Francisco and New York isn't. And that breeds a different type of creativity, it breeds a different type of innovation that you really can't get anywhere else.”
Hockey's childhood resilience came from parents who let him take risks and fall, and he worries that over-protecting children with elite schooling produces kids skilled at linear algebra but unprepared to take risks, which may matter more in 20 years.
“we may be creating children that can do linear algebra at seven... do you want kids that are going to be like pretty good at taking risks, that are going to be pretty good at being punched in the face?”
US financial services are relatively decentralized compared with other countries: Canada has four banks serving 95% of Canadians and Australia is even more concentrated, so the US should keep fragmenting financial power rather than concentrating it.
“95% of Canadians have four banks. Take Australia, it's even more concentrated.”
Column is 100% employee and founder owned, grows from earnings, and runs an annual tender in which it spends 25% of earnings buying back shares from employees, giving them yearly liquidity without outside investors.
“we take 25% of our earnings and we just buy back our shares with employees. So, we just run our own tender every single year.”
AI research labs thrive in Silicon Valley because frontier research is a consensus-oriented problem that benefits from smart people sharing ideas freely, but applying AI to everyday life requires building for people outside the valley bubble.
“You look at AI like our research labs are doing fantastic because that's like a consensus oriented problem.”
San Francisco is the most consensus-driven place Hockey has lived or visited, alongside Beijing per Dan Wang, and this consensus is both a clutch and a valuable operating environment for founders building ideas that the rest of the world does not yet believe in.
“San Francisco is probably the most consensus place I've ever been to. And I think that is both a huge clutch for us, but it's also probably our most valuable asset.”
Second-time founders with liquidity should put their own money into the company, because asking employees to bet everything while founders hold diversified liquidity makes startups too safe and attracts founders who want to be employees.
“If you're going to be asking this of employees, you should ask it for yourself.”
Mission should be near the bottom of a list of five priorities for a company, because customers buy a product only if it creates value for them, and an over-focus on mission and values can become distracting and leads to drinking one's own Kool-Aid.
“I think it's like on a list of five things most important, I think it's probably on the bottom end of that five list.”
Software companies that want to offer payments, deposits, and credit must control the underlying money, because enterprise software that touches money requires owning the dollar; Brex and Ramp show this, and Column exposes primitives and APIs so others can do it without becoming banks.
“But in order to do that, you actually have to control the dollar. You have to control the money, whether it's the lending money, holding credit, etc.”
Column is a software company that also owns a bank, using its regulatory position to build software that non-bank competitors cannot easily replicate.
“we have this interesting regulatory moat with a bank that most other people don't have, and we're going to build this incredible software behind it that nobody else can build because they're not a bank”
Compounding on yourself through your own company is often the best investment, and builders should specialize deeply because the alternative of being both builder and investor creates a conflict of interest in which you end up trading against yourself.
“there's like a me on the other side of that trade. Like I don't want to do that. I'm a builder. I think sometimes there's a trap where builders ain't their investors, and investors ain't their builders.”
The best determinant of founder success is whether they can find the most boring thing imaginable interesting and sustain that interest for decades.
“one of the best determinants for success of founders is can they find the most boring thing humanly possible interesting? And can they find that interesting over a multi-decade period?”
Second companies are easier to build than first companies because founders have already made costly mistakes and learned which activities to spend time on; first-time founders would have been better off working at a company for about nine months.
“if we would have just like worked at a company for like 9 months, we would have learned a lot. Like we would have always saved like 3 years.”
Employees fall into three categories: mercenaries who use a company as a launchpad for options at other firms, missionaries who will go to the ends of the earth for the mission but may rebel when commercial trade-offs arrive, and a third group who want upside plus stability, warmth, and near-term financial value and are usually the best hires.
“There's like the mercenary type, which is, okay, super smart, probably super pedigreed.”
Employees join companies mainly for near-term financial and life needs like children's schooling, housing, and avoiding roommates in their thirties, not for billionaire-scale equity, so illiquid stock that cannot fund those needs is a poor fit for many people.
“the majority of people join companies because, you know, they hit into like their late 20s or their early 30s and like what are they optimizing for?”
Column's customers pay for safety and longevity, because financial services switching costs are high, so Column plays a longevity and risk game and profitability is essential for sustaining customer trust.
“Our customers pay us for safety and our companies pay us for longevity. It's a unique thing about financial services”
Hockey owns only Column and Plaid and has almost no diversification, and he argues that building a company is special because concentrated ownership drives daily motivation and compounds your best investment, despite conventional advice to diversify.
“I own two things. I own Column and Plaid. That's it.”
US-style banks could not offer super-app experiences like Rawbank's in Congo because there is no market for them in the US, so emerging-market fintechs can land and expand differently, and their dollar-based currency makes innovation there look more like US fintech than like a traditional bank.
“Even if they could build that, there's no market for that. And so their ability to land and expand is fundamentally different.”
It is a good time to start a non-AI company, because the best startups are created in bad environments, starting is cheaper and less risky than ever, and most founders crowd into AI, leaving less competition in other sectors.
“it's probably a pretty good time to be a founder in a non AI related place right now. Because there's like probably less competition, less smart people.”
Column targets second-time founders and experienced operators rather than new graduates, because they evaluate total compensation and liquidity more carefully, and Column's pitch of yearly liquidity plus basic dilution and preference math resonates with people who have seen several rounds.
“It's a hard story for a 21-year-old. It's actually a pretty easy story for a 25-year-old that's been through like six rounds and four pivots”
Column is funded through profitability rather than venture capital, and the company commits to keeping 100% of employees and the founder as owners for the foreseeable future, which makes growth harder and constrains the business but improves long-term culture and effectiveness.
“We're going to grow by our earnings. We're going to make sure that 100% of employees and myself like like own the company for the foreseeable future.”
Plaid's attempted sale to Visa for about $5 billion was blocked by the DOJ, so Hockey left without a liquidity event and his paper wealth far exceeded his actual cash.
“we attempted to sell it to Visa for like $5 billion, and that's kind of the point where I left to go start something. It didn't go through. We got blocked by the DOJ.”
Plaid would not exist without venture capital, but venture capital is not a perfect model for every company, and founders should choose investors with timelines that match their margin and growth profile.
“I like Plaid would not exist without VCs straight up. I just don't necessarily think like the VC model is perfect for every single type of company.”
Kinshasa, the capital of the Democratic Republic of Congo, will become the world's largest city in 5 to 10 years, and most Silicon Valley people could not name what it is the capital of.
“it's going to be the largest city in the world in in probably 5 to 10 years.”
YC's list of startups to start should instead be a list of startups not to start, since once an area is consensus, capital and smart people flood in and the competition becomes intense.
“My recommendation is like that should be a list of startups you should not start. Because by the time that it's so consensus that this is a good area”
Silicon Valley's YC-style playbook has made founding a company too safe, producing founders who want to be employees and companies that merely pivot to AI rather than take bold, bet-the-life risks, and this is unhealthy.
“starting companies is just it's too it's too [ __ ] safe. Yeah. And that's caused a lot of companies to be just super safe companies.”
Hockey's approach to hiring means avoiding new grads and targeting people in their second company, since a 21-year-old is pulled by prestige and peer signaling, while a 25-year-old who has been through several rounds and pivots understands the value of a low-dilution, yearly-liquidity structure.
“It's a hard story for a 21-year-old. It's actually a pretty easy story for a 25-year-old that's been through like six rounds and four pivots”
Column's competitors are not other Silicon Valley startups but banks running patchworks of vendors, and the company must prove it is more sustainable than those alternatives, which is why earnings are critical.
“we aren't competing with other Silicon Valley companies. We're competing against maybe them doing it themselves or maybe competing with them trying to like build a patchwork of software vendors on top of like a legacy bank”
Column's card and account products are built on its bank infrastructure, so customer-facing fintechs like Built can offer banking while Column handles network connections, Federal Reserve and card-network complexity, and regulatory obligations behind the scenes.
“they build the application, they build the website, they build the consumer marketing, and we're going to handle everything kind of behind the scenes that has to deal with the Federal Reserve or TCH or the card networks or Swift”
Silicon Valley's aspirational consumer-product pattern means elites build products first, which then move down-market, but this has caused companies to ignore the everyday needs of the broader population.
“people buy something that's aspirational then it moves down market.”
In the DRC, mobile phone penetration is under 25% and banking penetration is under 5%, which means there is a large unmet need for basic financial infrastructure that few in Silicon Valley are paying attention to.
“Like, take DRC, like mobile phone penetration is still less than 25%. Your banking penetration is like still less than 5%.”
Fintech is probably the last area of software to be disrupted by AI, and the likely outcome is a collapse into domestic fintech and enterprise software, as companies like Ramp move deeper into workflow management.
“fintech is probably going to be like the last area that is somewhat maybe disrupted by AI. I think what you'll see is you'll see a collapse into like domestic fintech and enterprise software.”
Dan Wang's observation that San Francisco and Beijing are the two most consensus societies he has visited is accurate, according to Hockey, and it is an important lens for understanding why Silicon Valley tech looks alike.
“the two the two most consensus societies he's ever been to is San Francisco and Beijing. And I think that's like quite accurate, actually.”
Silicon Valley's approval ratings are probably very low and its ability to build resonant software is at a low point, because the people building software do not live the lives of the Americans they target.
“our ability to actually build software or have ideas or perspectives that really resonate is probably at the low point in the entire time I've been here.”
Column has almost no regretted attrition, meaning people the company wants to keep tend to stay, which Hockey attributes to a more mature company, better hiring, and an understanding of what matters to employees.
“We have almost close to no no regretted attrition, which means kind of people that we'd like to stay end up end up staying.”
Hockey is intentionally concentrated, owning only Column and Plaid, and plans to diversify at some point, especially now that he has a young child.
“I own two things. I own Column and Plaid. That's it.”
Founders in Silicon Valley should expect that venture-backed companies will look for scale-driven growth that neglects broader markets, and Silicon Valley's approval ratings are probably low, suggesting that the tech industry underestimates how unpopular it has become.
“Silicon Valley is not a popular place and I think we we tend to forget that.”