A Former High-Frequency Quant on Why Markets Shortchange Investors, and Why He Left to Build QFEX
EO Korea
Annanay Kapila spent several years as a quant at high-frequency trading firms, where the strategies he was responsible for traded between 10 and 100 billion dollars a day. He points out that 10 billion dollars a day is roughly the GDP of France. He was well paid, but he says he came to feel he "was wasting my life." In this interview he explains how high-frequency trading makes its money and why he thinks much of that profit comes from flaws in how markets are designed. He also describes why he left to co-found QFEX, which he describes as a 24/7 global stock exchange meant to "revolutionize trading."
From Cambridge to quant trading
Kapila was born in India. His parents later moved to the UK, and he notes that they "left behind a lot in India to kind of start over." He says the family was not very wealthy. He studied math at Cambridge, and near the end of his degree he had to decide what to do next. A PhD was an option and would have been impactful, he says, but quant trading "just paid a lot of money," and the money was too good to turn down.
After leaving Cambridge in 2020, he worked for about a year at Flow Traders, a Dutch high-frequency trading firm. He was then headhunted by Tower Research Capital, an American firm, where he stayed almost three years.
What a quant's day looks like, and where the money comes from
He describes the work as driven entirely by data. Before he even got to the office, he would check on his work laptop how much money the team had made. Quant traders don't trade by talking to people. They look at data and trade from it.
When people ask for a real-world analogy, he compares it to running a car dealership. A dealer keeps an inventory of cars, buys them at slightly lower prices, and sells them at slightly higher ones. Each margin is small. Kapila says quant finance earns so much because so much volume passes through the markets. His example is the S&P 500 future on CME: that one product trades about 500 billion dollars a day, which he says is more than the GDP of any country.
Profiting from design flaws
Kapila's main criticism is that high-frequency trading "is all about exploiting market inefficiencies," and that many of those inefficiencies are structural. His example is futures expiry. According to him, S&P futures expire every three months because the schedule matches the harvest of certain crops in the Midwest, and he argues an equity index future has no need to expire at all. At each expiry, holders have to sell the expiring contract and buy the next one. They pay transaction costs on every trade, so they lose money, and high-frequency traders "make the other side of the money."
He is blunt about the motives in the industry. "No one in quant trading wakes up in the morning and they think, oh, how do I make the markets more efficient today? How do I lower cost of consumers today?" In his view, traders want to make money so they can earn big bonuses. He also describes a "cognitive dissonance" among quants: because they earn large paychecks, he says, they persuade themselves they must be doing something good for the world. He includes himself, since he was earning the same paychecks.
Golden handcuffs and the source of his guilt
Kapila says Tower had very high-quality talent. Many of his colleagues had won International Math Olympiad medals, and many had ranked in roughly the top 50 in India on the entrance exam for the IIT technical universities. He believes they were caught in the same "golden handcuffs": they were paid too much to want to leave.
He argues that quant finance "has sequestered a lot of very talented people in an industry that basically adds no value to the world," and says this was the real source of his guilt. He adds an observation from his own conversations. Almost nobody he spoke to expected to still be in the job in five years, and many said they would stay two or three years and then move on. Yet they are still there. He asks whether they stayed because they wanted to, or because they "failed to re-evaluate."
FTX and the origin of the idea
Toward the end of his time at Tower, several things in his life lined up and pushed him toward starting a company. One of them was FTX. He is careful to say he is not defending Sam Bankman-Fried. He says what SBF did was clearly wrong and that he deserves to be in prison, and he does not accept that the ends justified the means. On the innovation side, though, he considers FTX a very successful company. He says it was profitable, backed by Sequoia, valued very highly, and made many products that were "actually really good." He also finds it interesting that FTX tried to get US licensing and regulation and bought an exchange in the US.
Kapila says he had already had the idea for what became QFEX about six months before. He describes a "burning desire to improve the markets," which came from having worked on the other side of them. His reasoning: if the real goal is efficient markets, the market's design itself should be improved, so that high-frequency firms no longer collect money "which they're just extracting from investors," and the market becomes fairer as a result.
Recruiting a co-founder and leaving
Around the end of 2024 and early 2025, he pitched the idea to Josh, one of his best friends, whom he has known since they were about 18. Josh worked at Citadel on the engineering side and knew nothing about exchanges. Kapila recalls calling him to say he had "this like crazy idea." He argued it was so obviously a better market design that it had to exist within five or ten years, and that either the incumbents would get their act together or the two of them would do it. Josh agreed and left his job at Citadel, which Kapila calls one of the top hedge funds in the world.
Money mattered to the timing. Kapila says his own job was going well and he had saved enough that he would not be under financial pressure. Once that was solved and he had "some decent money in the bank," he felt ready to work on something he wanted to be his life's work. He left his job in February 2025. His reasoning was that if he spent two to five years on it and it didn't work out, not much could really go wrong.
Getting into Y Combinator
The two applied to Y Combinator. Kapila was in Austria watching an opera when Josh called to say they had an interview, and he flew back to London the same day. He says the interview tested whether they understood the details of where the problem lay and what the "why now" was, meaning why this was the right time to build the idea.
He recalls Paul Graham explaining YC's approach. YC doesn't try to estimate the probability that an idea will succeed. It only wants to know the probability is above zero. Even at 1%, if the idea is huge and the team is good, YC will fund it. Kapila says this fits QFEX, because an exchange "is not like a 10 mil ARR business. Like it never be a 10 mil ARR business. It's either zero or huge." According to him, the questions that mattered were whether it was a huge idea and whether they were the right people to build it. They received the offer the next day.
The night the exchange "blew up"
Kapila stresses that fintech is hard to build because you can't launch something that breaks and then apologize to customers. He calls that "a real breach of trust." His hardest day came during YC. Their partners pushed them to launch early, internally to YC participants only, arguing that otherwise they would never get user feedback.
He was jet-lagged and working at 3 a.m. when his co-founder told him the exchange had "blown up." When they tried to reconcile accounts, balances did not add up. Some users were showing plus or minus amounts that should have been impossible, given that everyone had been given only about $100 to play with. They spent the whole day reconstructing what had happened, working out who owed what, and reimbursing some people. It was the first time the company took a loss. In some cases they told users something like: the system says you're minus $100, we don't think that's right, so we'll just give you $100.
The scale was small, so the financial loss was minor. What stayed with him was how hard it is to build a 24/7, fully available system that keeps accurate track of everything. Running around the clock means anything can go wrong at any time: "Lightning bolt can happen, fire in the data center, whatever." He says they didn't sleep properly for days because they knew that if this happened in a live exchange, "it's game over for us." In hindsight, he thinks the partners made them launch early partly to make them grow up and grasp the gravity of what they were building. He describes that period as both the best and worst days of their lives, working around 100 hours a week, very focused and "very frenetic."
Raising at a $95 million valuation before revenue
Kapila says people have misconceptions about how fast fundraising happens in Silicon Valley. For checks under about $500k, he says, it is usually a 30-minute meeting with a decision on the call or right after. Larger checks took a first meeting and a second in-person meeting, each around 30 minutes. They raised angel money on day one, then stopped accepting angels and focused on larger funds. The round was led by two major funds, General Catalyst and a second firm named in the recording as Nextus Venture Partners. It closed at a $95 million valuation while QFEX was still pre-revenue.
He argues that raising venture money only makes sense if the company can be "truly massive." In that case, he says, a VC finds it very hard to say no.
San Francisco versus London and New York
Kapila lived in San Francisco for about four months and now lives in New York, after years in London. He says that in London and New York people care a great deal about money: how much someone has, earns, keeps in the bank, and will make later. "Everything's about money." In San Francisco, in his experience, people talk and think about it less and care more about impact.
He contrasts this with the HFT mindset he came from. There, he says, strategies aren't supposed to lose money at all. A losing day brings an email from the boss asking what happened. Silicon Valley's approach, as he describes it, is to try to build something that could make a billion dollars in five years without really thinking about the money in the meantime. He found that a new and welcome way of thinking.
What QFEX is trying to change
The stated goal is for everyone to trade "on the same equal playing field." Kapila uses buying Tesla as an example. The stock trades on the Nasdaq exchange, the customer uses Robinhood as the interface, and a separate clearing house handles risk management and settlement. He argues that one trade involving three separate companies is intermediation that should be handled by a single company, as efficiently as possible, and says this is what QFEX offers.
He says QFEX pricing is fully transparent. It does not earn money through a "profit share agreement," only through fees, which he says are kept as low as possible, and a large share is returned to users, for example through referrals. He compares the company to Stripe: Stripe reduced friction in payments, and QFEX aims to reduce friction in trading. Because so much money moves through financial markets, he argues, even small improvements have "massive downstream impact."
Money, legacy, and advice for people in high-paying jobs
Kapila believes the best founders see their company as their life's work and legacy, as something more than money. He says he had reached a point where he was no longer just thinking about money. He admits part of his motivation for QFEX is still financial, and says there is also the aim to "be the change in the world that we want to see," through a change he thinks few others are positioned to make. His test is what you would want to tell your kids you spent your life doing.
His advice to young people in lucrative jobs is to think carefully about why they are doing the work, to ask honestly whether they want to be doing it in five to seven years, and whether the job will even still exist then. While you are young, he says, you "should be optimizing for learning and optimizing for growth, not optimizing for how much money you're making right now."
I've worked as a quant in high-frequency trading, 10 to 100 billion dollars a day, just, you know, the strategies that I was responsible for. 10 billion dollars a day is roughly the GDP of France. I was earning big fat paychecks as well, but I felt that I was wasting my life.
High-frequency trading is all about exploiting market inefficiencies. I saw how much money we were making due to simple structural design flaws in existing markets. People in trading just want to make money so they can earn big bonuses, right?
I'm an entrepreneur and CEO of QFEX. I used to work in the quant finance sector before I realized a lot of financial markets infrastructure is broken, and now we're building the next 24/7 global stock exchange, which is going to revolutionize trading.
We had two major funds. One was General Catalyst, one was Nexus Venture Partners. Round closed at 95 million valuation. We weren't making any revenues. We're pre-revenue, raising at a 95 million valuation. The only point in raising venture money is if you can be massive, right? Like truly massive. Then a VC will find it very hard to say no.
Yeah, there's certainly a lot of people in the US. I feel a lot of them can be kind of pursuing an entrepreneurial path because their future is so set. And for me, it was really obvious to see given my background, you know, obviously I was born in India and then my parents came to the UK. You know, they left behind a lot in India to kind of start over in the UK.
I studied math at university. I started thinking about what I wanted to do with my life towards the end of my time at Cambridge. We weren't very wealthy growing up and I went into finance, quant trading. It just paid a lot of money. I had an option to do a PhD as well, so it were impactful, but like the money was too good to pass up.
After I left Cambridge in 2020, I worked at a Dutch high-frequency trading firm called Flow Traders for a year, and then I got headhunted to an American firm called Tower Research Capital, where I worked for almost 3 years.
When you work at a trader as a quant, first thing I do when I get in, you know, not even when I get in, when I wake up, I have my work laptop, I check how much money we've made. In quant, everything is data-driven, and you're not doing trading by talking to people, you're just looking at data and you're trading from data.
People ask me for like a real world analogy, I often say it's like running a car dealership. Somebody who runs a car dealership, you can sell your car to them, you can buy a car from them, and their job is basically to kind of have an inventory of cars ready to sell, and they buy a car at a slightly lower price and sell cars at a slightly higher price.
And the reason quant finance is able to make so much money is because so much volume trades in the markets, you know? The S&P 500 future on CME, that's one future, one product, trades 500 billion dollars a day. That's like more than the GDP of any country.
But I think there's a lot of cognitive dissonance amongst quants and traders that they've kind of convinced themselves, because they're earning these big fat paychecks, and you know, I was earning big fat paychecks as well, that hey, you know, we're making a lot of money, we must be doing something good for the world.
High-frequency trading is all about exploiting market inefficiencies, like futures that expire, for example, you know? S&P futures expire every 3 months because that coincides with the time of the harvest, cuz certain crops in the Midwest; there's no need for them to expire. If they expire, what happens is people have to sell the future, buy the next one, they pay transaction costs every time they trade, and they lose money, and high-frequency traders make the other side of the money.
No one in quant trading wakes up in the morning and they think, oh, how do I make the markets more efficient today? How do I lower cost for consumers today? People in trading just want to make money so they can earn big bonuses.
And I think Tower has really, really high quality of talent. So, you know, a lot of my colleagues were International Math Olympiad medal winners. There's this entrance exam in India for like the top technical universities called IIT. A lot of them were ranked in kind of the top 50 in India when they did this test. They're stuck in the same golden handcuffs, you know, they're getting paid too much basically and they don't want to leave.
Quant finance has sequestered a lot of very talented people in an industry that basically adds no value to the world. And that was really the source of the guilt. I felt that I was wasting my life.
Think carefully why you're doing it. Think about, honestly, is this what you want to be doing in 5, 7 years' time? Almost nobody I spoke to said they would still be in the job in 5 years' time. A lot of them were like, 2, 3 years and I'll go do something else. Like, you know, they're still there. So, are they still there because they want to be there, or are they still there because they failed to re-evaluate?
Towards the end of my time at Tower, a bunch of things happened in life that kind of aligned that made me think that maybe I should leave this and build a startup.
FTX was a very profitable business. They were backed by Sequoia, very high valuation. I'm not going to defend SBF and say what he did was a good thing or the end justifies the means. I think what he did was clearly wrong. He does deserve to be in prison, but I think on the innovation side it's a very successful company. A lot of the products that they made were actually really good.
I think another really interesting thing is that FTX did try to go down the path of US licensing and US regulation. FTX purchased an exchange in the US. Interestingly, I'd already had the idea to do this exchange, what is now QFEX, like about 6 months before.
I have like a burning desire to improve the markets because I've worked on the other side of the market as a quant in high-frequency trading. And I thought if the aim really is to make markets more efficient, why don't you just improve the nature of the market, improve the nature of the design, so that high-frequency trading firms don't make all this money which they're just extracting from investors, and the market just becomes fairer that way.
Towards the end of 2024, early 2025, I kind of pitched this idea to my co-founder, one of my best friends. We've known each other for a very long time, I guess like since we were 18 pretty much. Josh worked at Citadel but on the engineering side. He didn't know anything about the exchange side of things, right?
He just called me and said like, "Hey, I have this like crazy idea." I was like, you know, this is so obviously like a better market design that there's no way it doesn't exist in 5 or 10 years' time, right? And either the incumbents get their act together or we do it. He was like, "Yeah, like let's do it." He left his job at Citadel, which is one of the top hedge funds in the world.
Job was going pretty well and you had a lot of money saved up. I wouldn't be under any financial pressure. Once that side of the equation was solved for and I had some decent money in the bank, then I thought, okay, it's time to do something that, you know, I want to make my life's work.
I left my job in February 2025. You know, I was really getting to the point where I was thinking, hey, you know, can anything really go wrong if I spend, you know, two, three, four, five years of my life, let's say, and things don't really work out.
You know, we applied for funding from Y Combinator. We applied, you know, Josh calls me and I'm like in Austria watching some opera show. He's like, "Oh, we got an interview. We got an interview. You have to come." I flew the same day back to London.
They ask questions that really test whether you've got into the details of understanding where the problem lies and what the why-now moment is. Like, why is now the right time to build this idea?
PG explained this to us. They don't think about things like, what's the probability of this idea succeeding? They just want to know the probability is bigger than zero. Even if it's 1%, but it's a huge idea and you're a good team, they'll fund you.
I'm aware that the company we're doing right now, right? It's either going to make me like $50 or zero. Cuz like an exchange is not like a 10 mil ARR business. Like it'll never be a 10 mil ARR business. It's either zero or huge. And they said, "Is this a huge idea, and are these guys the right people to do it?"
The next day, we got the offer, and then we're like, okay. And now that we have funding, we can probably build it. But building in fintech is always tough because you can't launch something if it breaks, to say sorry to your customers, right? That's like a real breach of trust.
The hardest day was during YC actually. We'd actually launched the exchange internally just to YC. We were kind of forced by our partners to launch early cuz they were like, "You need to do this, otherwise you'll never get user feedback. Just do it, right now."
I'm very jet-lagged. So, I wake at 3:00 a.m. My co-founder is there. He's like, "Oh man, the exchange has blown up." We looked at everyone to resolve, like somebody was plus $1, somebody was minus $1. You're like, "Oh man, we only gave them like $100 to play with. Like, how has this happened?"
So, we had to basically spend all day reconstructing what had happened, figuring out how much money everyone owed and how much they didn't owe, reimbursing some people. That was the first time we took a loss as a company where we had to say, "Oh, we're sorry. It says that you're minus $100, but we don't think this is right. We'll just give you $100." Luckily, it was still small scale. It wasn't like a big loss for us.
The issue was it really hit home how difficult it is to build a 24/7 perfect, fully available system that keeps track of... If you run something 24/7, like anything can go wrong anytime, right? Lightning bolt can happen, fire in the data center, whatever. We didn't sleep properly for days after that cuz we were like, "We don't want this to happen in life, right?" Cuz if it does, it's game over for us as an exchange.
I think the reason the partners made us do it is to basically make us grow up and realize the gravity of the situation we were in. That was a really tough time. Both the best days of our life and the worst days of our life, we were working all the time, like 100 hours a week, very focused, very frenetic.
You know, there's some misconceptions maybe about how fundraising happens in Silicon Valley. It's very quick. If the check size is less than 500k, typically it's a 30-minute meeting and you get the decision like on the call or just after. If it's kind of a bigger check, then, you know, we had a first meeting and then a second meeting in person, which were both about 30 minutes, let's say, roughly.
And day one, we had a bunch of angels, we got some money. And then we stopped taking angel money. Yeah, then it was about the bigger funds. So, you know, we had two major funds. One was General Catalyst, one was Nexus Venture Partners. Round closed at 95 million valuation. We weren't making any revenues, we were pre-revenue. The only point in raising venture money is if you can be massive. You have to be like huge scale. Then a VC will find it very hard to say no.
And I think you guys are based in San Francisco. It was very interesting to me, kind of being in London for a while, now we're in New York. I think London, New York, people are very concerned with how much money someone has, how much money they make, how much money they have in their bank account, how much money they're going to make in the future. Everything's about money.
San Francisco, people really, I think, don't talk about that as much or don't think about it as much. They're much more concerned with, you know, impact, that kind of thing. And it was really useful to live in San Francisco for like 4 months.
HFT strategies don't lose money, right? Like you can't even lose money doing HFT. If you lose money on a particular day, you get like an email from your boss like, what the hell happened? Silicon Valley approaches it like, you know, let's try and make something that's going to make like a billion dollars in 5 years' time, but don't even think about the money, right? It's like not even on the scale. Which is like a new way of thinking. That was a nice part of being in San Francisco.
You know, what we wanted to do is make sure everyone trades on the same equal playing field, level terms. If you want to buy Tesla, Tesla trades on the Nasdaq exchange, but the interface is through Robinhood, and there's another intermediary, which is called a clearing house, which basically does the risk management and settlement, right? It's that intermediation, three separate companies involved in doing one trade, that should really just be done by one company that's as efficient as possible. That's really what we're offering.
On QFEX, our price is completely transparent, right? So, we don't make money from this like profit share agreement. We just make money from fees. Our fees are as low as possible. We actually give a lot of our fees back to users if you kind of refer your friends or whatever.
It's similar to Stripe if you want like a direct comparison. Stripe really reduced the frictions when it comes to doing payments. We're reducing the frictions when it comes to trading. And so much money trades through financial markets in ways that people don't even understand that even if you make small improvements, they have massive downstream impact.
I think the best founders always have that, you know, that the company they're working on is their life's work, right? Like they want it to be the legacy. It's more than just money for them. And I was ready to reach that stage where I was no longer just thinking about the money.
You know, there's an element of me that's, you know, doing the current company for the money, but there's also an element of, you know, we want to go out there and be the change in the world that we want to see. And here's a very obvious change now that I think not a lot of other people can work on. It's what you want to tell your kids that you spent your life doing.
Do you think this job will still be around in 5-7 years' time? If you're still a young person, you should be optimizing for learning and optimizing for growth, not optimizing for how much money you're making right now.
Article published