Why Crypto’s Growing Everywhere Except the Price | Haseeb Qureshi

You know, nobody's buying or retail has left the building. Whereas crypto slowly goes down. That's not fun. That's not an interesting thing to trade. It's not an interesting thing to own.

The most valuable and important technology created in the last century is probably AI. And yet, it's four years in and most people are paying zero. Something like 30 to 40% of Americans don't use AI even once a month. Why would we think that crypto, oh, it should have happened already. Why isn't it already there? You know, Ethereum is 11 years old.

I think the answer pretty clearly is postquantum. Every single blockchain is going to get hurt by postquantum. There's going to be very, very large changes to Solana and all addresses will have to migrate in a pretty destructive way. I mean, it's pretty clear who's behind, which is Bitcoin. Bitcoin still has not gotten it together and obviously they have the most critical roadmap to get right because of the amount of money that's going to get hacked almost certainly if Bitcoin doesn't transition.

If you come back later when the market is already behind us and things are rip roaring, you're going to regret it.

Thanks for having me, Alessandro. Good to see you again.

Yeah, it's a good question. I think you'll get different answers depending on who you ask. From my perspective, I think this bear market doesn't feel tremendously different than previous bear markets to be honest. If anything, it actually feels better because the fundamentals in crypto actually look really good. Most of what people are complaining about in crypto is price action. Is that Bitcoin's down, ETH is down, alts are down, memecoins are down. And it's like, okay, why won't things go up, which is basically just a function of, okay, you know, nobody's buying or retail has left the building. You can see this very clearly from some of the headline metrics that retail is not very engaged in crypto right now. So, the traditional things you'd look at were, you know, where is Coinbase sitting on the app store? What are we seeing in search volumes? And then of course you know platforms like Robinhood which are traditionally retail dominated. You can see that their crypto volumes have come down tremendously. All of those are indications that retail is not here. Retail is trading meme stocks, they're trading AI, they're fascinated by a lot of other stuff that's going on. You know even the beginning of the year you saw gold was more volatile than Bitcoin was. So we're in this weird regime right now where the volatility in crypto is pretty muted. Price action has been fairly weak and that is resulting in this price bear market. But previous bear markets it wasn't just price. Previous bear markets it was fundamentals that actually looked really terrible. So you know projects were collapsing. Things on chain were just they were pancaking with respect to the TVL or the activity. There were narratives that just turned out to be totally wrong and false. So things like oh you know everyone's going to be in the metaverse, people are going to use crypto games, that was the animating narrative behind the 2021-2022 cycle, turned out to be completely wrong and so anything that went up on that basis had to come back down. Whereas if you think about this cycle, you know, what brought us to where we are today? What are the stories that really animated the cycle? It was the idea that TradFi is coming in. The space is institutionalizing. Bitcoin is not going to go away. It's going to be embraced by more and more institutions as well as potentially even governments are going to start embracing Bitcoin. There's going to be positive regulation in this space. And you know, DeFi and stablecoins and all this stuff is going to continue to grow. Well, here we are. Hyperliquid has become absolutely enormous. You can see DeFi is blowing through previous records with respect to the amount of trading activity and open interest. It's, you know, making front page news almost every single week. You see the same thing happening with respect to stablecoin supply hitting all-time highs. You see regulation continuing to advance. You see governments continue to buy Bitcoin. Like all this stuff is happening. The thing that's missing is that retail is not engaged.

And so, you know, to me, from just having seen crypto for over a long swath of time, I'm just like, look, retail comes and goes. There is a natural pendulum swing. It's kind of sine wave over time of retail engagement with crypto. I find it very difficult to believe that retail is just permanently gone and is never going to come back. Crypto is naturally reflexive. It's the most reflexive asset class I've ever seen. So that means that when retail comes back, retail wants to come back even more and even more and more people want to come because they can see that price action is starting to move. But right now I mean look we're about to have a blockbuster IPO in China. We've got massive volatility. You know what was it? There was what was it IBM that was just down like 25% yesterday going to send it down.

I mean look historically why was it that retail investors were so captivated by crypto? The answer is in part because okay crypto is a new technology and it's going to be used by people and blah blah blah but there's a lot of new technologies. There's a lot of things that are potentially going to get used by people. A big part of the reason why crypto was so captivating was because it was the most volatile asset class. I mean, just look at the historical Bitcoin price. Like, that tells you everything you need to know. This is a crazy asset class. It's supposed to be wilder and more exciting than anything else you could put your money into. That is not true right now. Crypto has lower volatility than meme stocks. And meme stocks have the virtue of going up in their volatility whereas crypto slowly goes down. That's not fun. That's not an interesting thing to trade. It's not an interesting thing to own. So I think once you get out of this regime where traditional assets and stocks are more compelling, more exciting and have more potential upside than crypto. Obviously, who in their right mind is going to say, 'You know what? This meme stock, this seems boring. And this doesn't seem like the flavor of the day anymore. I'm going to move on and go back to trading Bitcoin.' Nobody's going to say that. Nobody in their right mind is going to say that. Now, maybe if they're making a long-term investment strategy, okay, they want to have some ownership of Bitcoin in their portfolio. Yeah, fine. But that's not the kind of explosive activity that leads to crypto being exciting, fast-moving, and as a result being the thing that people think is going to drive outsized returns in their portfolio in the short term.

So I think you have to decompose TVL. TVL is an aggregate metric we came up with a long time ago. And I think if you're looking at just this aggregate metric and saying, okay, well, this number hasn't crested over the 2021 high. Obviously last year we had quite a bit more TVL, but and asset prices drew down, which means that TVL also draws down because it's measured in ETH prices and Bitcoin prices are very impactful on what the headline TVL metric is going to be. I think the thing to focus on more is what constitutes that TVL, right? What is the underlying breakdown of what's the TVL and why is it there? In 2021, a large part of the reason why the TVL was there was because of these kind of yield farm type really crazy short-term projects that people were putting money into to go take the token rewards and sell them. And those kinds of farms were very compelling in 2021 because the Fed funds rate was 25 basis points. It was basically there was no yield anywhere in the world. And so anything, no matter how speculative, that could potentially offer you yield was compelling to put your money into. That's why we saw so much TVL in 2021 was not because, okay, we had all these great products in DeFi that people really wanted to use, was just because the idea of yield farming in order to generate some yield in a world that was starved for yield was obviously going to attract a lot of capital. Okay, so after 2021, 2022, FTX collapse, everything renormalized, interest rates get jacked up to 5%. Right? In that world, it's very hard to clear a 5% hurdle rate, especially when crypto prices are plummeting, right? So over 2021, 2022, crypto prices came way down. And that meant that all of these freebies that people were giving out in these yield farms that didn't have any fundamental activity behind them, these all dried up and basically went to zero. So now you're in a world where you have to really compete to find attractive sources of yield that are actually sustainably going to be above the federal funds rate. Right now you're making 5% or just under 5%. Basically doing nothing taking zero risk. So if you're taking zero risk getting paid 4 to 5%. Why would you take significant risk in crypto to get even more yield? So I think what you see on chain is that the yield is increasingly not what is driving all this TVL. Obviously there's still yield. There's still a lot of yield, but the source of those yields are becoming more various. So, one of the things, of course, is staking. Staking yield is kind of the quote-unquote risk-free rate in crypto. It's not really risk-free, but we kind of talk about it because it's the lowest risk, generally speaking, way to get yield on chain. And we consider this in a separate category from TVL. You know, staked assets don't count as TVL. I think in retrospect, this might not be actually the right way to do the accounting because a lot of that capital moved away from the farms into staked assets. And they're still generating yield, but the fundamental activity is yield generation, right? That's what they actually care about. But then you look at to your point, the RWAs, the vaults, you know, vaults, for example, they count as TVL, but something that's an RWA may not necessarily count as TVL because it's not in a protocol. It's just, you know, it's being held as an asset that's tokenized on chain. So all these different measures, I think, result in different ways of decomposing what's actually happening on the blockchain. And then of course you've got things like capital efficiency when DeFi becomes more capital efficient meaning that you know for example take Uniswap v2, Uniswap v3 or v2 in particular, Uniswap v2 was very very capital inefficient required a lot of money in order to be able to create the liquidity in the AMM for people to be able to trade but in Uniswap v3, Uniswap v4 actually it's much more capital efficient you need much less capital to get the same amount of trades now you move from that to Hyperliquid and in Hyperliquid, you need even less capital to be able to trade because all you need is the tight band in the order book, very low latency, which means market makers don't need to have nearly as much inventory to market on something like Hyperliquid. And then you go a step further to something like Variational, which is RFQ based. And if it's RFQ based, then you need even less collateral. You don't need, you know, 15 different market makers all competing to try to be able to fill the same quote at the top of the order book. You only need, you know, a few dealers quoting things just in time with the collateral sitting directly there on chain in some native asset as opposed to in the asset of choice that's being traded. And so all of this points toward DeFi is actually serving more volume. It's actually making more fees. You know, at the end of the day, what we care about DeFi is the economic value, the economic activity. Looking at TVL as the measure of economic activity is actually the wrong metric if capital is becoming more efficient because then it means that oh we don't need to lock up the capital. Locking up the capital was actually a downside. It was not an upside of the previous model of DeFi. And so that's why I think it makes sense in the long run to stop thinking about DeFi in terms of TVL when the composition of DeFi changes in CeFi. The biggest product in all of CeFi is perps. It's where the most money gets made. It's where the most volume gets traded. It's where the most price discovery happens, happens mostly in perps not on spot. Perps is very capital efficient. Perps also but they also make a lot of money. All the big exchanges that do perps most of their profit comes from perp trading. So if that's true we should be compelled by the fact that now DeFi has moved increasingly towards perps away from spot. Spot is not the money maker. Perps is the money maker. Same thing in DeFi as in CeFi, right? And we see that Hyperliquid, you know, lighter, Variational, they're all making tons of money. Whereas, you know, before Uniswap and the previous generation of on-chain trading didn't make that much money. This is a good thing. This shows that DeFi is getting better, not worse. But if you look in TVL terms, you'll say, 'Oh, well, Hyperliquid only has a billion in TVL or a billion five or whatever it is.' And yet, you know, you look at Uniswap and Uniswap had tens of billions of TVL. Why? This is terrible. DeFi is going backwards. It's the opposite. But if you decompose the metrics, you'll be able to see that. But if you look at the aggregate metric of just TVL, it looks like DeFi is losing. I think the opposite is true.

Did I envision in 2022 that I'd still need belief in 2026? I don't know that I was thinking about it in quite those terms. I would say I'd be surprised if it was all done by 2026. If it was all obvious, if it was basically in the bag. I don't think I look I think the process of what crypto is going to do to traditional finance is going to take a long time. I don't think it happens that fast. I think it might be obvious what's happening pretty fast, but that doesn't mean it happens that fast. I mean, look at AI. People have talked about AI as the most obvious, the most compelling, the most product market fit of any technology that's ever been created in terms of, you know, fastest to a billion users, fastest to revenue, fastest to, you know, whatever, all these things. That said, you know, I was looking at some data recently. Something like 30 to 40% of Americans don't use AI even once a month. It's something like 1% of users of AI are paying anything. Most people are either on the free tier or they're not using AI at all, which means they basically see no value. They don't see it's like worth paying for. And it's not a lot. You know, 20 bucks a month is like, okay, that's like Netflix, you know. But most people say, 'Yeah, that's not even worth it to me. I'm not getting that much value out of AI,' which tells you and you know what are most people using AI for right now? They use it as a glorified search engine. That's primarily what people use AI for today, if they're using it at all. They're not really using it in the more sophisticated ways that people are starting to appreciate that it can be used for. So this is the way that the most probably I think most people would agree with this. The most valuable and important technology created in the last century is probably AI. And yet it's four years in and most people are paying zero. Most people are paying nothing. They don't see any value in it to be willing to pay money for it.

Right. Exactly. So what does that tell you? That tells you is that like it's pretty obvious to you and me that AI is going to change the world. It's going to change the labor force. It's going to change the way every company operates. But it hasn't done that yet. It's four years in. Not four years in from large language models. Four years in from ChatGPT, four years in from the shot heard around the world. Everybody's heard of this moment. It's been 4 years and still very little has changed. You don't see it in labor metrics. You don't see it in unemployment. You don't see it in layoffs. You don't see it in GDP. You don't see it basically anywhere. And so if it's taking that long for AI, and AI is obviously such a game-changing technology, we can all see it, right? Why would we think that crypto, oh, it should have happened already. Why isn't it already there? You know, Ethereum is 11 years old. It's been about 5 years since anybody who's an institution has even cared to understand what the hell this is. So it hasn't been that long. In absolute terms, it hasn't been that long. Technology takes longer than you think to diffuse and to really make its way through the economy. It's basically 10 minutes ago that crypto became legal in the US, right? Like basically if you go back to January last year, you could be thrown off your bank and potentially, you know, filed a lawsuit against by the SEC to enjoin you from ever running a public company ever again if you were to do anything in crypto. And we went from that to now, okay, fintechs can use it, banks can use it, financial institutions can use it, this just happened. And you go from this just happened to okay, why do I still have to believe anything? Look, why do you have to believe anything about AI like yes, you have to believe that people will use AI for more than just, you know, writing copy for their business emails because right now most people that's the only reason why they can see to use this is to help them write work emails. It's obvious that AI is going to do so much more than that. And to me, it's also obvious that crypto is going to do much more than it has done today. You just have to draw the line out.

Right. I think the same thing you see in AI too. Like it's not unique to crypto is that you know for somebody who's if somebody tells you oh you should set up your Hermes agent it's going to save so much time in your life it's going to be so beneficial to all the stuff that you do and the person says okay well I'll try it and they start doing something and it explodes and it tells them hey I need permission to do this and oh its cron jobs are failing because of some reason and you know if you're a normal person you're just like okay I'm done I'm going back to whatever I was doing before I'm not going to figure out how to solve this problem. What happens over time is that okay managed solutions emerge and people start to remove more and more of the friction right like I mean when you and I started using crypto we didn't have Privy, we didn't have all these API, you know, stablecoin orchestrators, we didn't have all the stuff that now makes crypto so much easier to use shaves off all the foot guns and makes it such that like actually it's really straightforward to take money even in the Philippines right but just even 2-3 years ago that was not true. Even 2-3 years ago like you actually had to be pretty sophisticated to be able to figure out how to take stablecoin payments in the Philippines. And so you know a lot of what we invest into is exactly this is like the people who are solving that last mile to make crypto easy to use in different places around the world whether it's you know for payments or stablecoin orchestration or all the stuff right the reality is I mean take take like B2B payments okay B2B payments are a story the idea that people would pay each other businesses would pay each other using stablecoins like these large international financial transactions I was telling people that story in literally 2017. Okay, I was telling people stablecoins are great for micro payments and macro payments. You want to pay somebody $20 million from Brazil to Russia, best way to do that is with stablecoins, right? I was telling people that and it was not true. It was not happening. No one was doing that for like 7 years. Okay? And then it was something like 2023 that it actually started to happen. 2024 is when it started to really grow. And now it's growing like crazy. B2B payments using stablecoins. It's a big part of the reason why so many financial institutions are now starting to support stablecoins is because they see this use case is growing like gangbusters. Now why did it take so long? The stablecoins were always there, right? Like the wallets were always there. The functionality was always there. Why did it take so long for businesses to start paying each other with stablecoins? And the answer is that it's this sort of death by a thousand cuts thing and it's coming to life by fixing these thousand paper cuts. So what are those paper cuts if you're a business? So one, it's like okay, well I don't know how this thing gets taxed. Second, how do I KYC this person? How do I ensure that this is not criminally sourced stablecoins? Third, how do I do on-ramps and off-ramps in all the different countries I might be sending to or receiving from? Fourth, what's the tax regime for this? Do I pay taxes on it? Or have there been clarifications in my local jurisdiction? The fifth, how does it integrate with my payment system, with my ERP system, with my accounting system? Is it acknowledged? Sixth, what about my bank? Will my bank see that, oh, you seem to have gotten some stablecoins. I'm going to shut you off because this is risky and I don't deal with high-risk clients, right? All these different things, they all add up. So yeah, in principle I could always just take a USDC or USDT payment 7 years ago, but why would I if I was a business until all these things get cleared away? And they eventually do, but it takes a while. In the same way, you know, with AI, it's obvious that somebody is going to make a UX for AI that's going to become easier and easier and easier and easier. It's going to paper over all these little paper cuts that you and I deal with when we deal with Frontier AI or, you know, an AI agent or whatever. Right now, we're dealing with those paper cuts. Learning how to paper them over, but eventually they will become so seamless and automated that everybody will say, 'Oh, yeah, obviously I'm going to use this now because it's just, you know, it'd be crazy not to.' That is where we are going with any technology that is midway through its adoption curve. That's where we are right now with crypto. I think that's pretty clear is that even the businesses that we look at that are starting to adopt stablecoins for international payments, most of their counterparties don't use stablecoins yet. Now, they can convince some of them to and they can say, 'Hey, look, we actually will you'll be able to take home more if you use stablecoins just the way that we're trying to pay you with stablecoins.' And they prefer it. Both the senders and the recipients prefer it. You know, there's a story about Meta is starting to pay out creators using stablecoins in emerging markets because they're realizing that look, we can offer to pay people in, you know, let's say I don't know where they're doing these payouts, but let's say in Nigeria, okay, I can pay you out in Naira if you want Naira. Obviously, there's a lot of creators in every country in the world. If you're YouTube, you're paying people in Nigeria as well. If you pay them out in Naira, okay, it takes 2-3 weeks to settle and you take home, you know, 93% of your royalties. But if I pay you in stablecoins, you get it instantly and you keep 99%. Which do you want? It's up to you. I'm not going to tell you. I'm not going to force you to take stablecoins, but if you want the stablecoins, you can take them. Well, it turns out almost everybody quickly realizes, oh yeah, I'll take the stablecoins. Thank you. And that process is only at its very beginning. It's only at its very beginning. Meta is only doing this in a few corridors. Once they see how much this works and how good this is for Meta, they are going to ramp this up almost everywhere in the world. So that's why I'm saying we are in the early innings of this. And when all those people have stablecoins and all of them come on chain, they're like, 'Oh, now they have stablecoins. What can I do with my stablecoins?' Guess what? Those are new entrants into the on-chain economy. That's where this is going. That's why I say it's still early. It's still early.

Sounds right.

Yeah. So okay, two things that I'd say in response to that question. So, first, in the spirit of doing these reframes, I'm going to do one more reframe. Please, if you look back to the railway mania, right? So, you were alluding to these historical manias where there was a lot of overinvestment. What do we mean by overinvestment when we say that, okay, you know, in the internet age or in the railway age there was too much investment. What that generally means is that the ROI on all these investments was negative, right? In aggregate. Obviously, what we don't mean is that somebody built a railway once and no one used it. Obviously, that of course that happens. That happens even in a railway boom, right? Even if things are, even if people have great returns, there's going to be people who make bad investments. So what we mean is we're talking about aggregates, which is that for all the dollars that were put into the railways, there was negative return on all those dollars or returns below the risk-free rate. That's probably the right way to think about the barometer. So if that's what we mean for railway mania because obviously railways were used and you know there were very profitable businesses. But in aggregate people lost money, right? Okay. In blockchain investments into infrastructure. The question is not are there bad chains and are there scams and are there, you know, rugs or whatever. Of course there are. They're always going to be that. The question is in aggregate was the dollars that were invested into this sector a waste of money or not.

Well, okay. So how do you tally that up? Well, so okay, obviously Ethereum didn't really have any of this investment, right? That predated VCs. But everything past Ethereum is what exactly? It's, you know, it's Solana, it's Avalanche, it's, you know, BNB, it's, there's, you know, there's a lot of infrastructure that falls under the banner of infrastructure. And includes things of course like Monad and Mega and so on. All of that goes in the same bucket, right? And so that whole bucket, the question is, did it produce positive returns or did the investors in all those things in aggregate lose money? Not were there individual chains that lost money but the whole sector did it make money or lose money? I think the answer is pretty clear that it made money if you aggregate over everything including the winners, including Solana, including Sui, including all these things did all the dollars go into there return capital or not the answer is very clear that they returned capital they made money that's part of the reason why people are like, oh there's so much, you know, why there's so much investment in infrastructure the reason why is because those investments have actually been pretty tremendous for the people who made them. Now, if you made the wrong investments, obviously you lost money. Same thing is true if you made the wrong investments into railways, right? Yeah. Obviously, bad railways lost money, good railways made money. But the question of was the entire boom a mistake is a question of aggregates. And I think in the aggregates, if you add up across all the funds that invested money into all the infrastructure, they made money on average. They didn't lose money. So, that's the first point I want to make.

Now, if you cut out by vintage too narrowly, right? You say, 'Okay, but yeah, fine, but what about the 2022 to 2024 vintage? Okay, yeah, that vintage of infrastructure probably lost money, right?' But that's basically saying like, look, the late infrastructure boom, the late infrastructure investment cycle, did that part make money? And that part, yeah, probably didn't make money. But that's not what people were talking about with railways, right? With railways, they didn't say, 'Oh, well, the last two years of it, people lost money.' They're saying in aggregate, over the entire period of people investing in railways. In aggregate, they lost money. That's what a bubble is. So, that's my first defense of infrastructure investing. Now, I think part of the reason why you see infrastructure investment declining, right? When's the last time you saw a big L1 round? When's the last time you saw a round that looked like one of these rounds that were, you know, getting a lot of attention in 2022, 2023? It's been a while. And part of the reason why it's been a while is that it's pretty clear, yes, we are at the phase now that we have a lot of infrastructure, that there's not that many gaps left in the market to try to address, but that's not zero. And what do I think is going to be the next place where infrastructure potentially is going to matter? I think the answer pretty clearly is post-quantum.

One thing that we don't fully appreciate is that we don't actually know how much post-quantum cryptography is going to hurt blockchain performance. Every single blockchain is going to get hurt by post-quantum. And the hurt is meaning not just oh they're going to get hacked by such, you know, whatever. Not this kind of stuff. I'm talking about when they make the post-quantum transition. Post-quantum cryptography is much slower. It requires a lot more bandwidth and it potentially requires you to remove a lot of optimizations that these blockchains have made in a pre-quantum world. It turns out pre-quantum cryptography is great. Pre-quantum cryptography is so fast, it's so efficient, it's so there's so many wonderful things you can do that don't work in a post-quantum environment. So, for example, with Solana, Solana addresses are not post-quantum resistant. And this is actually really helpful. It allows Solana to move faster because it has non-post-quantum resistant algorithms. Their transaction size is small enough that it fits within one packet. That doesn't work in a post-quantum world because the signatures are too big. They can't fit in one packet anymore. Now you need multi-packet stuff. Now you need to change the compute unit model for Solana. A lot of stuff has to get changed in order for these things to become post-quantum which means the big numbers we're talking about will go down. All these big numbers will go down. Now, how much will they go down? We don't know yet. We don't have architectures that were built from the ground up to be super high performance and post-quantum. So, we're going to see what the post-quantum performance numbers look like. And we're not going to be talking about Visa numbers anymore most likely after we do the post-quantum transition. So, all this stuff, the answer, look, we don't know. Obviously, it's very easy to draw a design document and say, 'Oh, you know, according to my design document, my blockchain is going to do blah blah blah.' This is a big engineering problem that nobody has done before and so once we see these post-quantum blockchains we may have to go back to the drawing board now not to say we will look we don't know but the reality is that we don't know and of course it's going to be a long time that there probably there's going to be challenges some of these post-quantum algorithms and there's going to be new ones that emerge and new constraints that we realize that we have to satisfy in order to make post-quantum work. But this is usually where new architectures come from is a change in the design surface, a change in the set of constraints. So I'm genuinely uncertain as to what that transition is going to look like and what it's going to do to blockchain performance. And I'm not saying that we're going to get, you know, 20 TPS blockchains after post-quantum, but I think the idea that, okay, we're going to have, you know, 10,000 TPS for free, because we had that pre-quantum, I think that's unlikely.