Robinhood has transformed from a company focused primarily on options, crypto, and equities into a diversified financial platform. It now operates 11 business lines that each generate at least $100 million in annual revenue. This shift marks a move toward becoming an all weather business that can thrive regardless of market trends or interest rate changes.

Maybe a couple years ago we were a V6, then we went to a V8. There's not a lot of automobiles with 11 cylinder engines, but we're pretty close to a V12. And then maybe one day soon, W16, like a super hypercar.

In 2020 and 2021, the business faced criticism for only doing well when markets were up and interest rates were down. When those trends reversed, the company had to pivot. Today, the platform is much more resilient with multiple business lines at scale, including prediction markets and Bitstamp, and several more growing toward that $100 million milestone.

Robinhood generates revenue through several distinct channels. Transaction revenue comes from trading commissions, rebates, and card interchange fees. Net interest revenue is another significant stream where the company earns a spread on cash held for users and provides margin lending. The high yield cash sweep program has grown significantly since its launch in 2022. It now holds tens of billions of dollars under custody. Vlad notes that the company has also become much more competitive with its margin book by improving rates and user experience.

The cash sweep program, our high yield offering, has gone from basically nothing in 2022 to tens of billions of dollars under custody. That has grown into a big business. We were not really that competitive on margin up until 2024. Then we really started competing on rates and user experience. Now the margin book has well more than doubled in the past year.

Subscription services through Robinhood Gold represent a nine figure business. About 15 percent of all customers are Gold subscribers. That number jumps to 40 percent for new users joining each quarter. Beyond current products, the company is following a three part growth strategy. The first goal is to be the top choice for active traders. The second is to capture the full wallet share of the next generation by handling every financial need on one platform. The final 10 year goal involves expanding globally and moving into institutional business.

I think if you look 10 years from now, we see a world where more than half of our revenue could be outside the US. More than half of our revenue could be business and institutional rather than retail. There are multiple vectors by which we could grow the business by 10x. We have a plan to pursue all of them.

The ultimate vision is to leverage technology to reduce costs while providing a superior customer experience. Vlad wants to create a platform where users have no reason to withdraw money to external bank accounts. The goal is to make it a no brainer for customers to keep their assets on the platform because the services serve them better than any other option.

Robinhood recently launched several updates focused on prediction markets, sports, and artificial intelligence. Prediction markets are now available on the web, moving beyond the previous mobile app restriction. This change supports the idea that these markets serve as a vital source of information and news rather than just being a tool for trading. Making them accessible via browser makes it easier for people to engage with the data as a media product.

The sports experience is also expanding with new features for trading individual player actions and combining multiple trades into one. This is similar to a parlay and points toward a future where contracts become highly personalized. Vlad envisions a path where this technology eventually enables the creation of bespoke insurance products tailored to specific needs.

Perhaps the most significant release is Robinhood Cortex, an AI model integrated directly into the platform. Unlike general purpose AI, Cortex has full access to a user's financial context and real-time market data. It can answer complex questions that would normally require significant manual work to process in other tools.

Cortex has an interface in the app which has full context of all of your financials and also real time market data. It knows about all the tradable stocks and prediction markets and instruments and it can do stuff for you. For example, if you want to scan through the entire universe of traded symbols and place a trade, you can just ask Cortex to do that with your voice.

Vlad explains that Cortex acts as a super intelligent financial assistant. It eliminates the friction of exporting financial data to external AI models to get insights. Users can now use voice commands to scan market symbols and execute trades instantly, allowing for a more seamless management of a financial portfolio.

Robinhood focuses on growing alongside its customers to ensure they do not move to traditional firms like Schwab as their financial lives become more complex. This strategy requires adding new products and features for maturing users while maintaining an interface that appeals to beginners. Vlad notes that many older financial institutions fail because they become tied to a single generation.

A lot of our predecessors get into this trap of being a generational company and I think that's certain death. Eventually nobody's figured out how to live forever, so eventually their kids are going to inherit the money. Their kids are probably less tied to the legacy brokers than their parents and grandparents.

To avoid this trap, the company must serve both established users and the next generation of eighteen to twenty-four-year-olds entering the market for the first time.

Prediction markets are gaining traction among both young people and older generations. The 2024 presidential election acted as a major catalyst for this growth. It served as a turning point that brought in a large number of new customers and started a new market cycle. While many people currently use these platforms for sports, the technology has the potential to disrupt much larger industries.

The presidential election just started it all and created this new prediction market super cycle. And a lot of it now is sports. But we see a world where prediction markets can disrupt multiple gigantic industries, not just sports.

One major area for growth is the insurance sector. New platforms now offer weather prediction markets. These allow people to trade and hedge against risks like fires or hurricanes. This process is often more efficient than traditional insurance. Standard insurance can be expensive and requires talking to brokers on the phone. These new contracts are already becoming competitive with traditional products. Vlad believes this is only the beginning of institutional adoption.

Molly notes that products often change once they are available to the public. Users may adopt them in ways that create a speculative or high risk reputation. Maintaining market integrity is the primary way to manage this perception. If users feel a market is being manipulated, they will stop participating. This leads to lower liquidity and worse pricing for everyone.

If people are concerned about market integrity, then they stop participating in the markets. And that makes everything downstream a little bit more challenging. It's harder to pull liquidity, prices get worse.

The industry is using tools similar to those in the stock market to monitor for issues like insider trading. Establishing common standards will help prediction markets move toward mainstream institutional use.

Retail trading has changed significantly since the volatile periods of the COVID pandemic. During the initial surge, activity was often driven by idiosyncratic events. Some traders focused on companies that had filed for bankruptcy, like Hertz. Vlad notes that while Robinhood wants to make it easy for people to do what they want, the platform also prioritized giving users information to ensure they understood the risks of investing in bankrupt stocks. Other trends were driven by a sense of nostalgia.

Retail investors got money in the form of stimulus checks. A lot of that went into brick and mortar retailers like GameStop and AMC. There was a component of nostalgia associated with it. People were familiar with the names. Millennials remember a world where they would go to the movies and go into GameStop. They missed that because everyone was locked up at home.

The market environment today feels much more stable than the ephemeral atmosphere of 2021. The current wave of interest is centered on artificial intelligence, which is supported by real revenue and substance. Unlike the meme stock craze, the companies retail investors are targeting now have actual growth and products, such as chips and AI services. This shift suggests a more meaningful foundation for market activity.

Robinhood serves as a bellwether for retail behavior, and its data shows a move toward diversification. In 2021, activity was highly concentrated in specific meme stocks and crypto assets like Dogecoin. Today, the platform manages about 350 billion dollars in assets under custody with trading volumes spread across many different stocks. This change reflects a broader growth in the market.

The maturation of the retail investors mirrored the maturation of Robinhood. We have gotten more diversified, they have gotten more diversified, and there is a much more stable foundation underneath the activity.

Being a public company often gets a bad rap, with many CEOs viewing earnings calls as a chore. However, earnings calls are a massive opportunity to engage shareholders if approached with the right mindset. Vlad explains that Robinhood previously conducted earnings the old-fashioned way, using scripts and Polycom phones where only sell-side analysts could ask questions.

I was just sitting there looking at it and I said to myself, man, I am bored. And if I am bored, the people watching this must be really bored.

To fix this, Vlad looked to the NBA playoffs for inspiration. Post-game interviews are entertaining regardless of whether a player wins or loses because they show real emotion and personality. By applying first principles, Robinhood rebuilt the experience to include live video and Q&A sessions with retail investors. Many people warned that sell-side analysts would be upset, but these constraints turned out to be imaginary. This shift increased their audience from hundreds of listeners to nearly 100,000.

Retail investors often become vocal defenders of a company during rocky times. This grassroots support is something currently missing in the AI sector. While crypto has a massive retail army that acts as a voting bloc to fight regulatory overreach, major AI startups like OpenAI and Anthropic remain out of reach for the general public. Vlad suggests that this lack of retail ownership might contribute to public fear and negative sentiment toward the technology.

Retail shareholders become kind of defenders of the company when things get a little bit rocky too. I think that one thing that worked very well for crypto was that so many retail investors had access from the very beginning that there was like this army that became a significant voting bloc.

Molly mentions that Vlad has been called the LeBron James of earnings calls. Vlad prefers a comparison to Anthony Edwards because he wants his career to have more longevity. He notes that while LeBron is one of his favorite players and he has met him once, the player is currently at the tail end of his career.

I'd like to have a little bit more longevity than that.

Vlad points out that LeBron might be the only NBA player older than him. He jokes that his own dream of playing in the NBA will be over once LeBron retires. He says the dream is already unrealistic, but the retirement of the last player older than him makes it official.

If he retires, my dream of being able to play in the NBA will be officially unrealistic at this point. It's over. I can't make it anymore. It's done.

Vlad explains that tokenization involves taking a real asset, holding it in custody, and then minting digital tokens against it. These tokens can trade 24/7 and interact with decentralized finance tools for lending and borrowing. This process mimics how stablecoins work with traditional currency. Robinhood has already tokenized over 1,500 stocks in Europe and plans to integrate them with Bitstamp to allow for continuous trading and self-custody.

Tokenization is the act of taking a real asset, an asset that has some utility. And you take that asset and you custody it. So you hold it somewhere and then you mint and burn tokens against it. And then those tokens can trade 24/7 and have access to the world of decentralized finance.

The real power of tokenization lies in its ability to make illiquid assets tradable. Traditional finance makes it easy to hold assets but difficult to trade them because building exchanges and attracting liquidity is complicated. Crypto solves this by providing trading, fractionalization, and liquidity out of the box. This could eventually apply to everything from private companies to art and unique collectibles.

Vlad describes how this could specifically unlock private markets. While it is not yet legal to tokenize individual private companies in the US, Robinhood is using a closed-end fund model called Robinhood Ventures Fund 1. This fund allows investors to buy into a diversified basket of private companies. Outside the US, Robinhood has already experimented with tokenized shares of companies like SpaceX and OpenAI. This shift aims to give regular investors access to venture-style returns that were previously reserved for a small group.

Robinhood started by tokenizing shares like OpenAI and SpaceX using mechanisms like LP interests in special purpose vehicles. However, the preferred approach is direct investment through Robinhood Ventures, where companies openly welcome the participation. Vlad notes a massive shift in how companies perceive retail investors compared to just a few years ago. In 2021, securing retail allocations for IPOs felt like a constant battle. Executives often questioned why they should bother with retail shareholders when they could just go public the traditional way.

This year, basically every consequential company going public is reaching out to us beforehand, asking us for advice on their retail strategy, how they can engage retail. There's been a shift from me having to really argue and advocate and work hard to get into these deals to now it's just much more organic and companies see the value.

This shift is accelerating for private companies as well. Vlad envisions a future where retail exposure is a standard part of every capital raise. He wants to push this access to even earlier stages to support entrepreneurship. Currently, founders often have to choose between building their product and fundraising full time. Making early stage capital easier to access would not only help startups grow but also provide retail investors with higher potential returns despite the increased risk.

Harmonic is building mathematical superintelligence. The goal is to create an artificial intelligence that can solve math problems better than any human. Vlad describes the ultimate objective as solving the Millennium Prize problems. These include famous unsolved challenges like the Riemann Hypothesis and the Hodge Conjecture. These problems have been open for hundreds of years. Solving them would unlock everything downstream of math, especially theoretical physics.

The North Star was can we actually solve really, really important math problems like the Riemann Hypothesis or Hodge Conjecture? If we could solve those problems well, maybe everything downstream of math, like theoretical physics, becomes unlocked.

One of the biggest goals in physics is finding a theory of everything. This would unify the force of gravity with the three other fundamental forces. Currently, we have no idea how gravity fits into the existing theories. If math can help bridge this gap, it could lead to incredible engineering feats like faster than light travel. Harmonic recently reached a major milestone by getting a gold medal on the International Math Olympiad. This is the hardest math competition in the world. However, Vlad notes that there is still a massive gap between winning competitions and solving math problems that have never been solved before.

Aristotle, a specialized math model, recently made significant strides by assisting in or solving 11 Erdős problems. These problems are named after Paul Erdős, a legendary mathematician who was so prolific that people still track their connection to him through an Erdős number. While about half of the 1,100 problems Erdős left behind remain unsolved, researchers are now using AI to work through them systematically. Aristotle even solved one of these complex problems entirely on its own.

Aristotle, which is the name of the math model that the company builds in a consumer form which people can use freely, got 10 out of the 12 problems of the Putnam correct, which is a lot more than I got on the Putnam. It's like the child has surpassed the parent in mathematical ability, and then things get weird.

The model performance is particularly impressive given the difficulty of the William Lowell Putnam math competition. Even Nobel Prize winners like John Nash felt the weight of its challenge. Vlad admits to scoring a zero when he attempted it, highlighting how the AI capabilities have already overtaken those of highly skilled humans. This rapid advancement raises questions about what else such models might eventually solve, from advanced physics to theoretical concepts like faster than light travel.

Vlad shares the story behind his appearance on the cover of Men's Health magazine. The feature linked the financial health of Robinhood to his own physical fitness. As the company reached profitability, Vlad was focused on reducing body fat and increasing his strength in the gym.

I was deadlifting 400 pounds. I worked really hard to lift that weight. It was a lot of weight. And in the article she puts my warm up set so she has me looking really intense but lifting 135 pounds.

While a journalist shadowed him for a 90-minute session, Vlad found that the extra pressure served as motivation. He ended up setting personal records during the recorded workout. Despite the heavy lifting, the final article featured a photo of him using much lighter weights from his warmup set.