AI Traders, Legal Perps & Tokenized Stocks - Lighter Founder Vlad Novakovski
- Vlad Novakovski’s forecast is one balance spanning perpetuals, options, and spot, with spot positions usable as collateral. That broadens access once reserved for multi-asset hedge funds, but it does not mean everyone should trade continuously: “You shouldn’t spend the whole day in a casino.” US retail will probably remain concentrated in familiar assets, while global market makers follow local retail liquidity across jurisdictions.
- AI is more likely to make traders “superhuman” than replace discretionary judgment. An individual might research 100 ideas a week instead of three, yet humans will still choose themes, venues, and risk limits. Vlad sees a barbell between systematic strategies with Sharpe ratios of five and above and exceptional discretionary investors; the partly systematic middle has historically struggled.
- Tokenized equities developed in the reverse order Lighter expected: liquid perps arrived before spot tokens. Perps need a credible price feed; spot requires inventory, issuers, custody, market makers, and answers around the SEC. Vlad nevertheless calls global asset tokenization a “tens of trillions of dollars” opportunity and expects multiple interoperable issuers rather than one winner.
- Centralized US perpetuals are already arriving, but no decentralized perp DEX yet has the necessary license. Vlad says each venue must independently satisfy CFTC principles and Rule 41 requirements around order-book management and risks such as ADL; copying another exchange’s design is insufficient. Lighter believes Ethereum settlement and transparent smart-contract risk controls can meet—or improve upon—the outcomes of traditional infrastructure.
- Buybacks are Lighter’s present mechanism for passing economic value to LIT holders, not a universal prescription for every token. The host challenged the thesis with Pump.fun, which had reportedly bought roughly $1 million of tokens daily for about a year while its token fell around 85%. Vlad’s answer was narrower: Lighter can buy and burn LIT, or potentially require institutions and market makers to acquire LIT for fees, producing the same economic effect.
- Lighter is probing markets that conventional venues do not serve, including H100 compute exposure, pre-IPO options, and tokenized business-line revenue. The H100 thesis extends from corporate compute-cost hedging to an AI agent autonomously hedging its own expenses. Vlad’s more radical example was isolating one revenue stream from a diversified company so investors could trade that business alone.
- Near-term execution centers on options, new order types, better execution and UX, agent-trading tools, and an even lighter EVM over the next two quarters. Robinhood is Lighter’s exclusive perps partner, but Vlad argues integrations must keep winning on quality rather than lock-in because on-chain execution providers are relatively easy to swap. Lighter has 45 employees and intends to remain selective, while slower cycles have enabled work on partnerships such as Telegram Wallet and Robinhood.
1. One balance turns institutional market access into a retail product
Vlad’s envisioned Lighter account holds spot, perps, and options together, letting a spot position collateralize derivatives.
His democratization claim comes with a warning: broader access improves markets and creates opportunity, but trading still requires skill. Asking whether everyone becomes a trader is like asking whether “everyone will become a poker player.”
Cross-border participation will remain asymmetric. Asian retail has sought US equities and crude oil, while American traders may follow familiar adjacencies—SanDisk or Micron into Korea or Hong Kong—but probably not unfamiliar Indian or Turkish markets. Global market makers will instead meet local retail where local retail already trades.
2. AI expands a trader’s bandwidth without reproducing judgment
Drawing on quantitative finance, Vlad described old quant models as predecessors to agents: trading desks have always tried to encode discretionary decisions, including through neural devices and reverse-engineering track records. “Fundamentally it hasn’t worked.”
His practical forecast is augmentation: a trader without a research staff might evaluate 100 ideas instead of three, while retaining responsibility for which questions to investigate, which exchanges to choose, and how much risk to take.
The historical pattern is a barbell. Mathematically grounded funds such as Renaissance’s Medallion achieved Sharpe ratios of five and above, while Warren Buffett and Paul Tudor Jones succeeded through discretion; strategies stranded between the two “tend not to work.” Future AI might close that gap, but Vlad does not think it has yet.
3. Perps outran tokenized spot because they need less plumbing
Trading everything online gives ordinary users the multi-asset reach once available mainly to elite portfolio managers, with activity “honest, transparent, and verified.” The surprise was sequencing: derivatives achieved substantial volume before tokenized shares.
A perp venue can connect a reliable price feed and begin trading. Spot forces market makers to hold inventory and address the issuer, custody, DTCC, and—where relevant—the implications for an institutional player dealing with the SEC. Vlad believes these frictions help explain why a “tens of trillions of dollars” opportunity is moving surprisingly slowly.
Vlad expects a stablecoin-like issuer landscape: several competitors, not one universal winner. Most share tokenization is based on Ethereum, though other networks such as Solana may participate. Lighter’s answer is neutrality—accept compatible assets from different issuers as collateral rather than choose the winning issuer.
4. Legal decentralized perps require proving outcomes, not copying licenses
Vlad characterized the CFTC under Chairman Mike Selig as operating “a little bit like a startup within the federal bureaucracy.” He joined its innovation advisory committee and cited Kalshi’s centralized contracts as the first fundamentally new market type added to the system in roughly 10 years—a correction from the host’s broader “first new market” formulation.
Centralized perpetual contracts are “happening right now,” but Vlad said nobody yet holds a license for a decentralized perp DEX. Lighter is “pretty confident” in its path, though every applicant must separately work through Rule 41: roughly half concerns order-book management and the other half risks such as ADL.
Blockchain implementation need not mirror rules written 50 years ago. Vlad argued that Ethereum settlement may be more verifiable than an exchange-maintained ledger, while transparent smart-contract ADL can achieve the purpose of traditional insurance-fund rules with less compliance machinery.
The same transparency could have prevented Bernie Madoff’s scheme or exposed it immediately through verification. Customers rarely request hidden orders; if that changed, Lighter’s ZK systems could also support privacy. Vlad suggested that genuine trading alpha may often remain unusable even if outsiders see a sophisticated firm’s position 10 minutes later.
5. Buybacks are a value-transfer mechanism, not a universal prescription
The host pressed Vlad with Pump.fun: approximately $1 million of daily buybacks for around a year alongside an estimated 85% token decline. Vlad did not claim buybacks ensure appreciation; he called them Lighter’s current way to direct created economic value toward LIT holders.
Retail users pay zero fees, while institutions and market makers could be required to pay fees in LIT. Lighter could then burn those tokens in its treasury, or burn an equivalent amount from another pool, creating the same economic effect as treasury-funded open-market purchases.
Vlad stopped short of prescribing buybacks to every project. During regulatory transition they are a transparent mechanism that works for many projects, but clearer rules could permit direct tokenization of businesses or revenue streams, giving holders “the same economic benefit” through another structure.
6. Lighter’s roadmap joins new markets with open distribution
The H100 market began as a hedge for companies whose cost base depends heavily on compute. Its more forward-looking use is an AI agent that “could autonomously hedge its own compute costs,” becoming self-sufficient across changing compute prices.
Other candidate markets include options on pre-IPO shares—even when the underlying is illiquid—and tokens isolating one revenue stream from a company with 10 business lines. Vlad tied the slow experimentation directly to uncertainty over what constitutes a security and how different authorities would view it.
Lighter is Robinhood’s exclusive perps partner, but Vlad rejects permanent lock-in: because on-chain execution platforms are easier to replace, Lighter must repeatedly win through cheaper, higher-quality execution. Prediction markets follow a similar distribution logic; third-party developers, not Lighter’s core team, can build them on Lighter VM and use the same trader balance.
The operating roadmap includes options, more order types, improved UX and execution, agent-trading features, new markets and partnerships, plus an even lighter EVM in progress for the next two quarters. The 45-person team plans to hire selectively, while slower cycles have provided time for big-picture work, including the Telegram Wallet and Robinhood partnerships.
Full transcript
Yo-yo-yo, how is it? It’s Thread Guy, and I’m back with a special taped interview with Vlad, founder and CEO of Lighter Perpex, one of the hottest products in the financial world right now. Vlad is back for part two in this entry. Vlad, nice to see you, dude. Welcome to the stream—or back to the podcast. It seems like an eternity has passed.
That’s right.
Dude, this is funny. I was just watching the interview with Avi, and you said you went to TJ. I’m from Northern Virginia myself. I didn’t know—are you from Virginia, or did you just go to TJ?
I grew up in Northern Virginia.
Okay. Oh, well, this is cool.
Yes, dude. I was just thinking about that. It was there that I met another Vlad. We are both TJ graduates.
Oh, yes, he really studied at TJ. I kind of forgot about that. This is incredible.
I love Virginia. I’m going home in 2 weeks to visit my family. I’m really looking forward to this. I haven’t been there in a while. Very excited.
But I was thinking about this. You came to the stream when we were live. I don’t remember the exact date. I think it was before October 10th—maybe not. This was definitely before the TGE for Lighter.
Since then, much has changed in the financial landscape. A lot has changed in my life, because when you were around, I traded almost exclusively on-chain. I didn’t trade perpetual contracts, or perps, at all. I definitely didn’t trade stock perps.
In fact, since then, over the last 6 months, I’ve spent a ton of time on trading volumes. So I feel like I can approach this from a different angle than raising these metrics. Dude, I’m just cranking up the volume like crazy right now. I just raised the volume incredibly. It’s cool.
You’re also on a podcast tour right now. I managed to listen to a few episodes, and I liked them. In any case, thank you for being here with us.
I guess I’ll start with a question. Based on the idea that I’ve been trading a lot of perps lately, could you explain what you think trading will look like in 5–10 years? Do you believe that everyone can become a trader? What will they trade? How will they trade? What will the market look like in 5–10 years?
Yes, I think access will definitely become wider. I think there will be a lot of synergy between the markets. You will be able to trade—or I know for sure you will be able to on Lighter, for example—perpetual contracts, options, and spot from one balance.
I think that’s really cool, right? Because when you trade on a centralized exchange, you can’t just move your position somewhere else. But on a decentralized platform like Lighter, you can have a spot position as collateral for a perpetual one, or trade options.
It’s not about whether everyone will do it. It’s like asking, “Will everyone become a poker player?” I think this requires skills. AI is changing that too, right? The types of trader skills will look a little different with AI agents than without them.
But the main thing is wider access: democratization of finance.
So does this mean that everyone should trade all day?
Probably not, just as you shouldn’t spend the whole day in a casino. But wider access makes markets more efficient and creates opportunities.
On the topic of AI, I think there’s a huge push from many exchanges right now. Coinbase talks a lot about this, and Robinhood does too. The concept is agentic trading.
I integrated AI into my trading from a research perspective to understand why I lost so much on the last 7 trades. What am I doing wrong? I analyze my strategy, but not from the perspective of executing an entry or exit. What do you think about agentic trading?
Yes. When I worked in quantitative finance, a lot of funds, including Citadel, were run like this. You usually have discretionary traders who make decisions based on human judgment, and you also have quantitative models.
The quantitative models that existed 10–20 years ago are, in some ways, the predecessors of AI agents. But it’s kind of a question of whether you can automate these human decision-making processes. That’s always been what trading departments have tried to do.
If you could put what a discretionary trader does into a model, then that obviously creates efficiencies. That’s never been done successfully, and I’m not sure AI agents can do it either.
I think, to answer your question, they’re more likely to make human traders superhuman. For example, instead of researching 3 trade ideas a week, you’ll be able to research, let’s say, 100 ideas a week.
But you’ll still have to make decisions about what topics you’re looking at in general, what type of risk management you want to use, and what exchanges you choose. I still think there’s going to be a lot of room for human decision-making.
Maybe someone who’s really good at risk management but doesn’t have a large research team now gets a superpower that allows them to do a lot of research, for example.
Why has it never been done effectively?
It’s been tried many times. I remember there were even professors at MIT who did this. While someone was trading, they used some kind of neural device to watch what was going on in the person’s brain, and they tried to map it that way.
Things like analyzing a person’s track record and reverse-engineering their actions have been tried a bunch of times, but fundamentally it hasn’t worked. I just think that it’s like a barbell, right?
There are strategies that are very systematic, and that’s what successful quant strategies do. For example, the Renaissance Medallion Fund that I mentioned earlier—Jim Simons was the first to do that. Thorpe did it too. There are a lot of other quant funds out there now.
These strategies are very systematic and mathematically well-founded. They’ve done very well, with Sharpe ratios of 5 and above. On the other hand, you have people like Warren Buffett or Paul Tudor Jones, who are very discretionary, and they’ve done very well.
So it’s like a barbell. If you’re in the middle—partly systematic and partly discretionary—those strategies tend not to work. I think if you try to take successful discretionary traders, what they’re doing is actually not very easy to replicate, even with traditional machine learning.
I’m not sure it’s easy to do even with this new AI. Maybe someday the next generation of AI will be able to catch up. I don’t rule that out. But so far, human judgment that’s not purely quantitative is not something that AI has been able to really replicate in trading.
It’s a complicated story. Yeah, it’s interesting. So, on this idea of an exchange where you can trade everything, I feel like there are quite a few players who are really keen on that.
Do you think it’s going to become a popular practice for traders from the US to increase their trading volume in emerging markets? Obviously, South Korea is very popular right now because of the memory trade—SK Hynix and Samsung—but they’re a bit of an anomaly, at least as far as I understand how most of these emerging markets are traded.
As we get more access to trading everything, how do you think behavior will change in terms of what people trade?
Yes. What we’ve seen so far has been more of a reverse process. People from places like Hong Kong and other parts of Asia have been getting more access to the US markets, and also to commodities like crude oil. I think we’ve seen that more often than the other way around.
I think you’re right, and for certain markets we’ve seen movement in both directions. But if you think about retail investors versus institutional investors, for institutions like Jane Street—market makers—they want to be active on a global level.
For that to work, on the other hand, there have to be retail investors. So I think you’re going to have that kind of access. Let’s say an institutional firm wants to operate where there’s an exchange with all the necessary licenses. You’d want to be in as many jurisdictions as possible, creating markets.
But retail investors have to come from somewhere. If you’re talking about countries like Turkey or India, is an American retail trader going there if the locals are already excited about those markets? Probably not. They’re probably going to trade what they know.
To a greater extent, if there are related markets—for example, if they trade SanDisk or Micron here, they can trade similar markets in Korea or Hong Kong. But they’re not going to go into markets like India or Turkey.
Market makers are going to go there and trade against local retail investors.
I think that’s a good idea. I actually think that’s a good way to look at things. I like that way of looking at things, and I’ll probably add something.
I never really traded stocks. I learned to trade by playing around with options when I was 16 or 17 in high school, through my parents’ Robinhood account or something. Then you merge 4 or 5 accounts and you’re like, “This sucks,” and then you find NFTs. This is more interesting. I can make more money on NFTs and crypto.
Now it’s all coming back to normal. There’s a certain deep irony that the best crypto product or innovation of the last couple of years is this idea of trading stocks, but I’m just obsessed with it.
It’s the most enjoyable trading experience I’ve had in the last 6 months, and it’s been a great environment for retail investors to trade stocks. The market has been mostly up, except for the last week.
I’m curious as to how you see the landscape of this tokenized stock ecosystem evolving. Who do you think will win? There’s Trade XYZ, some kind of independent company. There are exchanges that are trying to do this.
There are projects like Backpack or, let’s say, Robinhood, that are trying to bring all the stocks online. Is there a major winner in this idea of tokenized stocks? Is it becoming democratized? Who wins the most, and how do you win in this ecosystem?
I think your idea of trading everything is one way to look at it. It goes back to the idea of democratizing finance. If you’re, say, a big hedge fund, you could do that all the time, right? If you’re a portfolio manager at one of the big hedge funds, you could usually trade everything from stocks to commodities to currencies and, in the last 10 years, crypto assets—everything in your portfolio.
Now, the fact that anyone can do it, and that it’s online, honest, transparent, and verified, is a big opportunity. It’s a democratization of what was previously only available to elite traders and portfolio managers. I think that’s one of the points.
With tokenized stocks, we expected there would be tokenization of stocks first, and then derivatives like perpetual futures, or perps, would be built on top of that. But it turned out the other way around: first there were perpetual futures, with high trading volumes, and then gradually tokenized shares began to be introduced.
It seems to me that there will not be just one issuer. I think there will be several players competing. We at Lighter do not choose winners in this matter; everything happens on the network. Most of them are based on Ethereum. Of course, there are others, like Solana or, as you mentioned, other networks, but most of the tokenization of shares happens on Ethereum, where there is also DTCC and the issue of how asset custody works.
We are in a position where all of this can be used as collateral on Lighter and traded on our platform, so we are happy to cooperate with any issuer. In some cases, we already do.
That’s the same behavior as with stablecoins, right? There is Circle, there is Tether. There is no one winner; there are many new players. I think it’s going to look something like this. Some partners might be more aligned with our infrastructure than others, but ultimately, if it’s all on the network, it’s going to be compatible, interoperable, and easier. There’s no one winner in stablecoins, but there’s Tether.
But tell me, why has it gone in the opposite direction? Why is there a proliferation of perpetuals, with tokenized spot assets clearly taking a back seat? Why is that? Are there any technical limitations?
I think it’s more that smart markets are slower to develop, because you need all the plumbing to work, and so do market makers, right? If they actually have to maintain inventory, it’s different than if you have perpetuals. With perpetuals, if you have a good pricing channel, you can just plug in and start trading.
With spot, you have to worry about inventory, you have to worry about the issuer, and, in some cases, what this means for an institutional player dealing with the SEC. There are a lot more things to think about.
It’s surprising to me that this is moving so slowly, given how big this opportunity is. The tokenization of global assets is massive. This is a tens-of-trillions-of-dollars opportunity, right? Despite some friction points, it’s surprising that more players aren’t acting more aggressively, even though it’s already happening.
We’ll be there when this is ramping up, and it’s going to be a consolidated trading experience on Lighter.
The big topic of discussion right now in the crypto community is the regulatory landscape. Of course, clarity is getting a lot of coverage, and so is the legalization of perpetuals in the US. That’s a topic where I think there’s a lot of confusion. Even I feel a bit confused about the regulatory landscape in the context of legal perpetuals.
There’s Koshi and Coinbase, which have legal perpetuals; they essentially act as perpetuals. And there’s also Hyperliquid, which is very active on Capitol Hill trying to get results. Can you outline the situation with legal perpetual contracts in the US? Who is allowed to do it, who is not, and what is needed to make progress?
Perpetual contracts and derivatives in general are regulated by the CFTC, right? We spent a lot of time with Chairman Mike Selig and his team. I also joined their innovation advisory committee, helping them think through some of the larger issues.
The CFTC moves quickly. In that sense, they operate a little bit like a startup within the federal bureaucracy. They’ve already approved centralized contracts in the context of Kalshi, which was the first time in 10 years that a fundamentally new market had been added to their system.
They’re moving quickly, but the rules are still there. There are basic principles of the CFTC that need to be complied with, either literally or in some other way, and there has to be a rationale for why it works.
For example, if settlement happens on Ethereum, which is a very reliable settlement layer that has been around for a long time and is very decentralized, that could be even more transparent and more compliant than a traditional ledger, right? The exchange doesn’t have to maintain that on its own if it relies on Ethereum for settlement.
Another point is something like ADL. If it’s implemented correctly, where the risk management is transparent, you actually need less compliance, because in traditional finance you would have to maintain all these different insurance funds, and that requires rules. Whereas if you implement it through smart contracts or mechanisms like ADL, those rules can be implemented in a way that gets the same or even better results using blockchain technology.
The bottom line is that centralized perpetual contracts are here. It’s happening right now; there are people in the US trading perpetual contracts on decentralized exchanges, specifically Cali. The next step is going to be decentralized perpetual contracts. I think that’s the area we’re most interested in: the decentralized perpetual DEX space.
Right now, nobody has a license to do that. We’re pretty confident moving down that path. We think the technical solutions we’ve implemented at Lighter make it easier to comply with a lot of these rules.
We’re very impressed with the leadership and the CFTC team, who are really working to innovate while staying true to the core principles.
The first new market in 10 years—that’s just wild. I didn’t know that.
Well, not the first new market, but the first new type of market.
The first new type of market. And what was it 10 years ago?
I think there was some new type of swap. After the financial crisis, they added certain markets at the time. Early 2020s.
The first new type of market in 10 years is crazy. So what do you think the landscape of regulated decentralized perpetual DEXs will look like? Will it be the type of exchange with this infrastructure where everyone gets a license? Will only certain companies get a license and others don’t? How big is the moat for those who get a license compared to others who can get a license?
I think a license is needed, like in traditional finance, right? Let’s say the CME has a license, just for the sake of analogy. A new exchange can’t just come in and say, “We’re doing the same thing as the CME, so give us a license.” No, you still have to go through the process and comply with Rule 41, where half of it is about order book management and the other half is about risks like ADL and things like that.
Some of these rules could be implemented on the blockchain—not exactly the way they were written 50 years ago, but actually in a better way. You have to explain why. But that doesn’t mean you can just say, “We have to go through the whole process with the agency and convince everyone.”
You can’t just say, “We’re doing the same thing as another exchange. Take our word for it,” and that’s it.
By the way, a little digression: you told a story, I think, on Aya’s podcast about Bernie Madoff and how, if there had been on-chain settlement or something like that, the whole operation would have been exposed instantly. You would have known about it.
That’s right. It’s an amazing story. I think he wouldn’t have even been able to start that pyramid scheme in the first place. But even if he had somehow done it, the verification process would have exposed it right away, as opposed to having to pick up the phone and call someone, and they just forgot to do it and it never got exposed.
Another thing that interested me was this narrative that people don’t want their trades to be public, right? Everybody wants a dark pool. They want to hide their orders. You also mentioned that the complaint, “I want to hide my orders,” is not a common one among customers. I think it’s a misconception in the crypto space that people hate public orders.
For us, it would be a competitive advantage if that were a customer request, because we use ZK. We use ZK for scaling now, and our ZK schemes can also be used for privacy. If that were a customer request, we would actually be in an even stronger market position.
But we have to meet customer requests where they are, if it’s not already a request. Maybe that will change as more traditional players come into this space, after CFTC approval and all that. But we’ll see.
Why doesn’t it matter that much?
I think it’s because a lot of the alpha investors—it says something about where this alpha is coming from. Maybe it’s not coming from where people might assume, right? Maybe the source of the alpha is not such that, even if you knew about a high-yield trading firm’s trades 10 minutes after they were made, you couldn’t really do much about it.
That's interesting. I saw Multicoin had a little bit of a commotion today because they took Hyperliquid off staking, and they were like, “This is why I'm bullish on Zcash. We need private orders.” It seems like that was on Hyperliquid. They de-staked Hyperliquid, and they were upset. People were watching the wallet.
Yeah. I think it's a little different. It's a long-term investment. It's not a firm that's trading for “alpha.”
I guess these long-term rates, like in traditional finance, are usually publicly known. These short-term rates aren't.
Got it. Yeah, that makes sense.
Okay. Here's another question for you. Has there been a historical discussion in the crypto debate, if you will, about the concept of buybacks? There's a classic example: Rollbit, of course, in its failure or lack thereof. There's the story of Hyperliquid, which has been a huge success lately.
I've talked a lot about Pump.fun and its buyback mechanism, and on crypto Twitter there were discussions about whether buybacks matter, because Pump.fun has been buying back about $1 million worth of tokens every day for the last year, give or take, and the price of the token has gone down—I don't want to be wrong—by about 85%.
Lighter has this mechanism where, at this point, you're essentially buying back the Lighter token with all the revenue, at least for now, and then in the future you're going to split that just with the revenue. But the point is, you guys are taking buybacks pretty seriously. Do they matter? Are buybacks important? What's the purpose of them?
I think right now it's the primary mechanism through which the economic value that Lighter creates goes to LIT holders. It's really important to do.
In the long run, we want to get to a state where any business can be tokenized and revenue streams can be tokenized, and then you could potentially achieve the same thing that way. I think buybacks are a good way to do it right now.
Another way to think about it is that it makes sense for the exchange. For example, we could have users pay fees. Lighter has zero fees for retail users, but institutions and market makers could have to pay fees in LIT. Then they buy LIT on the open market to pay those fees, so you achieve the same thing that way.
In any case, we in our treasury would like to burn that LIT, or the equivalent amount from another pool. If our fees were paid in LIT, we wouldn't want to sell them; we would want to burn them. I think it makes sense when we achieve the same economic effect through a different mechanism.
I think it works for businesses like Lighter. Broadly speaking, is this right for every project?
In the current environment, where we're in a state of transition and there's still no clarity, there's a question about which tokens are securities. In an environment where a lot is still uncertain, I think this is a transparent mechanism that works for a lot of projects.
In the long term, whether every project will have to do a buyback, I'm not sure. There may be other ways to get the same economic benefit to token holders. But in the case of Lighter, it's very appropriate, and we're going to continue to do that.
You mentioned regulatory clarity a few times. Do you think achieving that will have a significant impact on the crypto industry as a whole?
I think so. It's actually going to have a smaller impact on Lighter directly and a bigger impact on the rest of the industry: how tokenization works, how new DeFi protocols work, KYC, and all that.
If you're running a derivatives exchange, regardless of whether there's clear regulation, you need some form of regulation anyway. But for a lot of other projects that are going to come out, it's important to have that clarity. I think it's going to be pretty important.
You recently launched a market on H100, which is pretty exciting. I think it's also the first market on H100, at least from what I've seen. It's the first decentralized perpetual exchange. Some centralized exchanges added it later. The first perpetual exchange to launch an H100 market. What was the thesis behind that move?
The initial thesis was that there would be demand for this from companies that are heavily dependent on computing power, like scalable services or other AI companies.
But I also think the AI world and the cryptocurrency world are still a little bit far apart. There needs to be some convergence, because most people in AI right now are asking, “What is this?” They don't know exactly how it works, and vice versa.
Ultimately, if you're running a company where compute is a big part of the cost structure, hedging those risks would be very valuable. But what's even more forward-looking is the idea that an AI agent could autonomously hedge its own compute costs. That opens up a lot of interesting possibilities, where an AI agent can be self-sufficient regardless of what's happening with compute. That's an interesting concept.
As a follow-up to that question, what else isn't there a market for right now that you think there should be?
If Vitalik were here, he would say leasing. That's something he likes to write about. But I think there are markets like options; that's something we're working on.
An interesting concept would be: What would an option on a pre-IPO stock look like? Even if the underlying asset is not liquidly traded, that doesn't mean the option can't be liquid. There are some interesting markets of that kind.
Another interesting market is the tokenization of income streams. Imagine if you could have a token that captures, let's say, one of the 10 different lines of business a company operates. You want to bet on just one of them, and you could somehow tokenize the income stream of that line of business. I think that kind of thing could be interesting.
Wow. I've never thought of that concept before. It's crazy. Income isolation is really good.
Yeah, that's right. That goes back to your previous question. Part of the reason there hasn't been enough work done on these ideas is because there's a lack of clarity about what is a security and what isn't, and how it's viewed by different authorities.
For example, perps are pretty straightforward because they're not that different from traditional futures. But with some of the other things, there's still a lot of room for financial innovation.
What are you spending the most time on right now? Are you thinking about integrating something new? What's taking up the majority of your day at Lighter right now?
That's a tough question.
It's interesting, because I feel like when you're in a good cycle, you have less control over your time, which is kind of funny. When you're on the upswing, you're seeing good numbers, and a lot of people come to you with ideas that they want to work on together.
I don't just mean internal things. I'm talking about potential partners. A lot of it is saying, “Okay, spend time with this group because they're thinking about this new market,” or, “Maybe we could work together on a wallet integration.”
When you're going through a more challenging period, nobody really calls you. You can step back a little bit and think about what the future of the world looks like, analyze your roadmap, and figure out what technology you need to create. It's interesting how that works. It's a bit countercyclical.
I can’t talk about the price of the token, but in terms of the price, you had a lull period where everyone was asking, “What happened? What happened to Lighter?” It was a couple of months, at least, if I remember correctly, sometime in the spring. Then, just recently, everyone was saying, “Wow, Lighter's numbers are really crazy, and it's a really good trading platform.”
Do you change what you do when the numbers are down compared to when they're up? What do you focus on? How do you deal with the downturn—numbers down, marketing down, crypto token price down? How do you work with that?
There's just less incoming requests during those times. You have more time to take a step back and think about the big picture, and maybe work on 2 or 3 really big partnerships.
That's when we worked on the Telegram Wallet and Robinhood partnerships. I think that's something every founder has to learn. There are cycles in any business, and in crypto too.
You're building more publicly, and because of that, you're getting a little more attention. I think that's mostly good. We're trying to be transparent. There's a bit of a toxic atmosphere on Twitter, but you have to take the good and the bad.
You like that we're still always on?
Yeah, I like that. Although I think the token holders don't like that.
Some of your tweets really stand out. Some of them are iconic. You have a few posts like that every month.
Yeah. Of course, it's interesting. Communication is another thing that we're improving on.
One thing we’ve noticed is that every time we talk about technology, we win. For example, we had this event with Vitalik, where we took a deep dive into technology, and I think that was kind of a turning point for us. I saw you were having a fireside chat with him, like an interview.
Yeah, that was really cool.
And congratulations on the success of Robinhood and Telegram TON. I know you’ve talked a lot about Robinhood. I want to ask the question a little differently. I’m curious about how the partnership with Robinhood is working and what else they can do with other infrastructure providers, because there’s been an interesting evolution. Robinhood has integrated KHI as a prediction market partner, and it’s incredible for both parties. It’s also incredible for KHI’s business.
I think that’s given KHI a big boost in volume. It’s really strengthened their position in the US and helped their business a lot. And then Robinhood is trying to build its own prediction market, and maybe that’s hurt them a little bit, right? It’s kind of slowed things down a bit. Maybe they would have been better off starting with their own product from the beginning.
So I’m curious: have you learned anything from that dynamic with Lighter and Robinhood? Have you learned anything from that?
I think, first of all, in the DeFi context, it’s different, right? Because you can have all the orders on the blockchain. And so if you had a centralized execution platform, you would use one or the other, I think.
From our perspective, first of all, I think the efficiency that Lighter offers is a win-win for the customer, for Robinhood, and for us. And we are the exclusive partner for perps, but even if there was another collaboration where that wasn’t the case, and we had to prove ourselves every time, we would still win on quality, right? Because it’s a cheaper structure, and it’s very easy to swap one for another. On-chain, swapping one execution platform for another is much easier than in a centralized way.
So ultimately, we have to provide value to the partner and their customers. And I think we really do that with our technology. But it would be unfair to anyone if it was like, “Okay, you’re locked into this platform forever, regardless of whether it brings value.” That would be wrong.
Yeah. Yeah, I agree. Cool. I have one or two more questions for you. I’ll let you go soon. One, maybe one more clarification on prediction markets: how do you feel about prediction markets and perps being somewhat mixed up—maybe figuratively in the way people perceive trading, but also literally, because every exchange that’s desperate to add perps is also desperate to add prediction markets?
Do you have any plans to do something like HIPP4? What do you think about the overall evolution of markets?
Yeah. I mean, I think it’s all about distribution, right? Every platform, like every blog site, wanted to add videos, and every video service wanted to add blogs. You know what I mean? It’s probably a distribution issue. I think they’re pretty different markets.
In terms of outcomes, our view is that the infrastructure that we’ve built, and especially with the advent of the Lighter VM, will allow developers to implement prediction markets on Lighter using the Lighter balance, so a trader can trade prediction markets with the same balance they use for other markets. That’s great. We’re fully supportive of that.
I don’t think our core team wants to be in the prediction markets business. I think we want to let other developers who are passionate about it do that. So that brings us back to the democratization of finance, right? Because democratization means, by definition, that not every market is going to be amazing for every participant.
There are going to be some markets where people are going to lose money if they bet on the wrong outcome. But we don’t think we want to be in the business of tracking who did what in a particular baseball game or anything like that, right?
Yeah, that makes sense. But it’s a great vision—the democratization of finance. I’ll give you this question to finish off. What’s coming up on the Lighter roadmap, and also on the user experience side of trading? I know you’re working on a lot of cool things, like new order types and things like that. What’s the near future, and maybe the longer-term perspective, for Lighter and the use of the platform?
Yeah. Well, I think there’s a lot more to come in terms of new order types and a greater variety of features that, as I said, make every trader a superhuman trader. That includes things like improved UX, better execution, and AI—adding agent-trading elements to the platform and adding ways to implement these kinds of strategies. That’s definitely part of it.
I think fundamentally, options are a big part of our roadmap. We’ve talked about that a few times, and an even lighter EVM is already in the works for the next 2 quarters. So yeah, that’s happening.
There are lots of good things. Of course, as always, we’re adding new markets and new partnerships. The team is very inspired. We just had a general meeting, the recording is on Wednesday, like every Wednesday, and each team had a dozen ideas that they wanted to work on. So I think it’s time for us to act.
How big is your team?
There are 45 people now.
Wow. Wow. Are you hiring new people quickly? I guess, you know, there’s an old saying: hire slowly, fire quickly.
I don’t think so. We don’t want to hire too many people, as you understand. But we’re very interested in talking to talented people who share our vision.
I love it. Vlad, thank you very much for coming. A lot has happened since our first meeting. That was before TGE, and a lot has happened since TGE. You’re in a great position now, and I appreciate everything you’re doing for the industry. We’re excited to see you and discuss what you’re working on. Thanks again, man.
Thanks for the invite.
Sure. Okay. Have a nice day.