Transcript
Speaker: Welcome to a new episode of the InSilico Terminal podcast. Today you will hear maybe, i hope not, my AC in the background. I'm leaving it on. i don't care. i don't want to die here I hope our audio guy, I have full trust in him that he will fix it. And um yeah, my guest today is Andrew from Blox Golds.
Speaker: A bit of a different guest from the usual type, I guess. And maybe you can start by explaining what what what it is that you do, what is Blockscholes, and what do you do from that, what is your role?
Speaker: Yeah, absolutely. Firstly, thank you for having me. um Excited to get into it. So Blockscholes, the name kind of gives a lot of it away. We do crypto derivatives and ah in a nutshell. So for us, that means research, data, analytics.
Speaker: Mm-hmm. As the name suggests, it started with a big focus on options, ah calibrating implied volatility surfaces, providing options pricing analytics, a lot of pre-trade um analysis data, putting together the tools that you need to make better and decisions as an options trader.
Speaker: This was a big focus for us to begin with. Now we are far more expanded, so we cover research data analytics, yes, for crypto derivatives, but increasingly the influx of TradFi and real-world assets to crypto is also being seen in crypto derivatives.
Speaker: So a big focus of us and our clients at the moment is real-world assets, the derivatives, pricing oil over the weekend, price gold overnight, providing settlement indices for equity perps, solving problems in derivatives that for a long time TradFi had not been interested in solving.
Speaker: So this is a big focus for asking my clients right now. My role is as head of research that is very wide and very all encompassing by research. We mean, yes, the retail facing subscription service where you can sign up to our research.
Speaker: You can read our research reports. We write a lot about decentralized finance and technology in modern tech, intersection between that and TradFi. On the other side, we have very bespoke personal service where we we use our expertise two to advise on treasury management, trading strategies.
Speaker: We can provide backtesting, access to our data. We ask the person sits in the middle to use that data to give better investment decisions. And we also get involved in the design of decentralized exchanges, particularly in prediction markets and in options as well.
Speaker: So yeah, really a little bit of everything so far, all under the hood of crypto derivatives. Interesting. I just realized I'm kind of kind of retarded because I didn't get the name at first.
Speaker: And also I didn't know how to pronounce it properly, but now that you explained it, it it makes sense because it's like the this options pricing model and stuff. I kind of get it now. um how How long has your company been around? So I think we spun out in 2021, April.
Speaker: This was from our founder, Eamon, who was for 15, 16 years a trader in Citi. um He, yeah, at big banks like Credit Suisse and Nomura, trading infrastructure exotics derivatives.
Speaker: So the pedigree of the team is really institutional first. People who have done this at big banks, at places like Bloomberg and Goldman, who are trying to then build the same quality tools for the same users really at an institutional tier in crypto.
Speaker: That was way different back in 2021 where Deribit, I think, was ah was firstly one of the only venues that you could trade anything non-linear. And secondly, it was still tiny, right?
Speaker: You could only trade stuff like Bitcoin and Ethereum. um And even then, they incredibly small volumes and incredibly poor liquidity. It's very different now where we've been tasked with, as I said,
Speaker: building out on-chain liquidity for for a whole bunch different hundle lines. So this was 2021. I joined in November of that year, which was, yeah, for a long time, I could say the last time that Bitcoin was at 69K.
Speaker: Now I've seen Bitcoin hit 69K from above, from below, from sideways, front, back, It's of crazy that we're below that right now actually. I think we're closer to 59k today. Yeah. Which puts a timestamp on when this was recorded, I guess.
Speaker: Yeah. So, um yeah, this was November 2021. And now for the last four and a bit years, trying to build out this company, trying to build out this market of nonlinear instruments. When people say derivatives in crypto, they generally mean Well, firstly, they meant perps, and everyone picked two perps when prediction markets were not yet a thing and options weren't really taking off.
Speaker: um Then they meant liquid staking derivatives, so things like staked Ethereum. Only now are we getting to the ah real meat and potatoes of derivatives, which is structured products, nonlinear payoffs, but stuff that me and my team would get out of bed for, stuff we're excited by.
Speaker: And a big part of that is prediction markets, actually, in a surprising way. who Do you only offer like, um I checked with your website a little bit, I don't know too much about what you do actually, but do you do you only offer research data or do you offer like these products yourself to clients?
Speaker: Good question. um We provide whatever's needed is the cop-out answer. In full, we give um data as ah as a primary product.
Speaker: So we deliver that in a couple different ways. I would say primarily we are a data provider, but we also do that through an Oracle service. We have our own push and pull oracles to deliver that data.
Speaker: We also provide that through WebSocket and REST API. So the types of people who consume that data will go down those paths in different ways. One example of the the Oracle service we provide is to clients like Gravity, the perps exchange. They take not our implied vol data, but settlement indices for perps.
Speaker: Settlement indices for perps and margin mark prices for collateral purposes. Very similar for Derive, another a client of ours, who take implied volatility for liquidation and mark price purposes.
Speaker: We also provide them with collateralization mark prices as well. So if I want to trade options on their hybrid CLOB, I need to deposit some capital, be that stablecoins, be it a liquid staking derivative, be it Bitcoin wrapped with yield.
Speaker: And then I can trade options against that using that collateral. Our job is to sit in the middle and say, this option, its mark price is worth here. You have this amount of collateral in your account, which means you're covered up until this point.
Speaker: If at any point your collateral goes below some some value where you pose a risk to the exchange, it's our job to provide that mark price and it's the right job to then make great. this is the service we provide via Oracle. Now, yeah the data that we built was not originally intended for liquidation of mark prices.
Speaker: We stumbled into this business very early on because we were the only ones providing it. I think lot Lyra, who are now derived, formerly known as Lyra, are pivoting to their v2, which required them to to take an Oracle price for for options. and I think we were the only people in the market that had gone into such depth.
Speaker: like Many had done some level of implied vol. We had focused more on the quality, the frequency, the robustness of that data, which meant that we were prime to take that market. so We've now developed that for DEXs and really pushed into the quality of that. and Then at the same time, we have that same data be consumed for other purposes by other types of entities like banks or asset managers, hedge funds.
Speaker: Anyone who needs that data for an independent reference price, yes, but also pre-trade analytics. I need to this data to know what is rich, what is cheap, yeah what the market is pricing for now and how I can best take the advantage of that.
Speaker: and for back testing. So if I have an option strategy and I want to see how it performed over the last three or five years, you come to us. We can provide that data across all of the exchanges that we have. I think it's Garibit, Bybit, OKX, and very soon to be prediction markets.
Speaker: We calibrate volatility services to that data and we make that available via either API or through our in-house backtester, which is where I step in, I'm a first user of that backtester. I'm developing quant strategies in in volatility, in options, using that real wealth of data that we have.
Speaker: up I would say i'm I'm user zero of that data. okay I find all the bugs first and go through the pain of checking to the tech team to fix them. What makes your data better than like other data providers?
Speaker: It's good question. um It depends on the the data that we provide. So for stuff like settlement indices, for spots, for example, we don't really claim to have much more of an edge, and I don't think many places will claim to have much an edge.
Speaker: Everyone can see the Binance MID, a very well-broke data point, right? Neither do we claim to have an edge on on-chain data. We're not an aggregator of on-chain metrics, be it raw point-in-time values or derived aggregations on derived metrics. This isn't our edge either.
Speaker: Where we have an edge is on the stuff that requires financial engineering and a bit of computation. So there's two parts to where our edge lies there. One is in the delivery. I'll come to that later.
Speaker: The other is in the design and creation. So Yes, things like volatility surface calibrations are well known. This maths has been around Black Shoals, the original pricing formula we mentioned was from 1980, April, the SVI calibration we provide from a paper produced of maral lint in 1999.
Speaker: All of this is well known. The bit that we add is in the secret source, cleaning, aggregation, robustness, Yeah, there is a lot of very idiosyncratic issues with crypto data.
Speaker: Yeah. Not just because it's crypto in a very intense 24-7 volatile market, but because the places historically where that where those instruments have traded have been shakier off-chain centralized exchanges, offshore exchanges, whose APIs come and go like the wind. They drop in and out.
Speaker: And their data is not formatted in the same way across many different venues. So on the cleaning side, we put together um all of that data and clean it.
Speaker: We provide no arbitrage checks as well, which is another big thing for options services. surface and furring And then we shut it out. Now on the delivery side, um we are better and then we can do it faster and more robust with higher uptime than all the others. So that part is so simpler. And so it's everything that you need from an Oracle and you need to rely on for delivery.
Speaker: We're thinking about and we are we're laboring over in the background. Do you have an example maybe of like ah a simple trade that you can like walk us through or something that someone is using or doing on a DEX and then they use your for that?
Speaker: Yeah, um I can give a really relevant example to today's market as well, actually. um So for a long time, prediction markets were not around and retail were not trading anything with convexity.
Speaker: They were getting all of their Convexity, all of their margin and leverage through, well, leverage on perpetual swaps that happens to pay that that margin. And if they get wiped out, it's nothing against the 50X that they would get on the the long on the short, right?
Speaker: So prediction markets come along and suddenly retail loves trading crypto options. As long as you just don't tell them that they're crypto options, they're very happy to punt on whether it's up or down of the next five minutes, right?
Speaker: So it's really solving a UI issue there. However, we are options guys. We recognize prediction markets on financial assets for what they are, which is a binary option.
Speaker: Yes, with a very short tenor, five minute minute market. But ultimately, you need to be able to price that using implied volatility. That is the key input to to pricing those the those instruments.
Speaker: Where our data is incredibly useful for that is forming a market view that is calibrated to vanilla markets like on Deribit, vanilla calls and vanilla puts. Use that to calibrate a volatility surface that can then be used to give the fair mark price or fair mid price of a binary option on, let's say, Bitcoin, for example.
Speaker: um This is very different, by the way, to a binary market on the World Cup, right? Where the idea of volatility surface for France doesn't really make sense.
Speaker: And if it does, then I would not be giving away that secret sauce on this podcast. Yeah. um So you need our volatility services and our data and the stuff we calibrate in order to market make on a prediction market.
Speaker: That is how you get a fair price for a binary option. So it's relevant for Bitcoin and Ethereum, which are very popular. But increasingly, we are seeing altcoin binary options really take up a lot of the...
Speaker: a lot of the liquidity. um Where we have been tasked with solving that problem is by creating volatility services to maybe markets that do not have options markets already.
Speaker: Hyperliquid and Zcash, two great examples. Derive came to us and said, look, we want to launch Hyperliquid options. They don't trade anywhere else. We can't calibrate a volatility service to the market because there is no market.
Speaker: Our job then is to create and defensible synthetic volatility surface used to bootstrap liquidity. Something that can get the market off the ground, that the market makers can then and start to provide.
Speaker: um that then allows the market to go around it. So give me one second this is a very similar problem to um oil markets, let's say, volatility surfaces for oil over the weekend.
Speaker: You are tasked with creating something that can keep a stable and robust market that isn't necessarily tradable prices, but allows a mark price to exist such that the market can can trade over the weekend.
Speaker: who This is where, yeah, I guess where we have an edge is that we're very comfortable, very happy doing that financial engineering around the the obvious stuff. So we have competitors who can calibrate volatility services. As I said, but that tech is known.
Speaker: The tough part is then extending that to to things um either outside market hours or outside reasonable levels of liquidity. and Can you maybe explain a bit um what a volatility service is and how people or market makers or whatever use it to stuff for the retarded people like me?
Speaker: Of course, yeah. I think maybe you're sandbagging the question a little bit. um But very good question. So way back in 1980 or thereabouts, we discovered or invented or created the Blackstrolls formula for options pricing.
Speaker: you What this does is it allows you to express the price of a vanilla call or put as a function of the volatility of the underlying asset. Now, the volatility of the underlying asset is just how much it's wiggling around.
Speaker: It turns out that an auction an option's price can be conducted decomposed into its intrinsic value. So if I was to expire today, how much in the money would I be?
Speaker: Let's say I've got a call option struck at like $10 and we're trading at $100. got about of intrinsic value there This is one part. The other part is extrinsic value.
Speaker: So intuitively, you can think of it as um the dispersion of the terminal price distribution, which gives you extra chance of expiring further in the money than you are.
Speaker: Maybe you're at about $100 right now, but volatility is very high. I don't lose anything if I'm below my strike, but I gain a lot more if I'm even further above my threat. who So volatility is a way of expressing that spread of the distribution. You think of it as a vol curve, right? Yeah. As a bell curve.
Speaker: By thinking of it as a bell curve, we are explicitly making the assumption that the underlying asset is moving with a normal distribution or technically a log normal distribution. The volatility level is the standard deviation of that log normal distribution. Very simple. So this is great as long as you're only looking at one option price.
Speaker: Excuse me. um If you then look at a market of other options prices, what you find is that options prices at different strikes will imply a different level of volatility than each other.
Speaker: That's because one of the fundamental assumptions of the Black-Scholes equation is that we are log normally distributed, and in reality, stock prices, bond prices, and definite are the crypto prices are not log normally distributed. They have what we call far fatter tails.
Speaker: So the chance or the probability of extreme large moves up and down are far more likely than you would guess from a normal distribution. So that means that by expressing it as a single number, the standard deviation of um of a normal distribution That violates reality.
Speaker: So when you look at the price of options deep in the money, it turns out that they trade at a higher price than you'd expect, given the volatility implied by an ABBA money option. a That means you need to turn up the volatility that's implied by those other strikes in order to compensate.
Speaker: So ultimately the volatility smile is a curve that arises as a function of the strike. gives you the volatility implied by options at each strike. And it's generally curved upwards in the wings because the wings are pricing in a higher probability than you'd expect given normal distribution.
Speaker: So the volatility surface is that volatility smile expressed at different tenors. What the volatility surface ends up being is a convenient way to quote the price of an option at any strike in every tenor.
Speaker: It's not necessarily um
Speaker: agreeing with reality. Now, the volatility surface is just a concise way of pricing an option at every strike and every tenor. doesn't necessarily mean the market is correct about the way they're priced.
Speaker: If they were, then markets would be fundamentally completely efficient and you could never make all these any money on average because the market charged you how much you'd expect to then to make all those.
Speaker: So what we find, at least in crypto and from my analysis, is that the volatility surface is incredibly reactive. And therefore market expectations are incredibly reactive to what's just happened.
Speaker: So different metrics that you can pull from the volatility surface, like term structure, like volatility smile skew, where you look at the the relative volatility implied by put options and call options, quite a key directional metric.
Speaker: All of these things move quite responsively to spot over the previous three months, six months or so. And then very often you see the market get caught off guard with, um let's say, a huge skew towards puts. Spot has just fallen. Let's say it's, I don't know, 11th of November, 2022. FTX has just imploded.
Speaker: Bitcoin has gone down to something like 16K. We see the strongest skew towards put auctions that we've ever seen because people want to protect against worse. so um Thereafter, the year after you got was a massive rally in the year of 2020 and 2023 and then a recovery into 2024.
Speaker: So this is a the quoting mechanism. It shows you what the market is pricing for today. It's not necessarily a forecasted model. So the volatility surface is incredibly useful. That same volatility surface can be used to derive the prices of Vanilla options of prediction markets of a whole range of structured products. It is not um it's sufficient for every type of product. There are quite a lot of exotic pricings that require more complex models.
Speaker: Even something like an American option versus a European option requires something like a Monte Carlo to do RAM just because there is path dependency in the value of the An American option versus a and ah vanilla European option allows you to exercise early.
Speaker: So you need to look at the value over time and see whether it's worth you exercising early. The volatility surface is still a fundamental um building block of those other prices.
Speaker: So this is why we start here. This is why we see the most demand for it. So that is what you can use then to give market makers a way to to market and make these markets that don't exist yet anywhere else.
Speaker: Exactly. um more More relevant to, let's say, on-chain stuff is providing a mark price. So what i have here is a volatility surface. It is smooth in tenor and it is smooth in striped.
Speaker: And that means that I can price an option or give you the fair mid price of an option anywhere on that. That is incredibly useful for for markets where I need to know how much your position is worth in order to liquidate you or not. I want to make sure that liquidation is fair, that if you have posted margin to your position, that it is being considered against the value of your position.
Speaker: Is there any... um From what you've explained so far, yeah like you you provide data and you work with like ah bigger, more institutional clients or like exchanges and stuff like that. Is there any way in which retail traders would be able to like benefit from you without just like using the products or whatever? or that Is there any reason for them to like be interested in you in wu or what to do?
Speaker: Yeah, absolutely. um I would say for retail traders, the biggest benefit over the last six months, yeah getting on for six months now, is the advent of AI tooling.
Speaker: So no longer do you need to go to a
Speaker: a software as a service company to go and provide you something like an analytics dashboard. right So big pattern we've seen over the last six months, speaking to people at conferences, is that the the pitch that many of these companies give has gone from their usual platform to data is our moat.
Speaker: Well, I think that has a big shift over the last six months because it means that execution on building these dashboards is no longer tough. right the The execution cost for building that has gone to zero.
Speaker: Very easily ask someone like Claude, like Gemini ChatGBT, as long as it has the data, to go and visualize it and show you that data in a way that makes sense and aggregates that information.
Speaker: For now, the real premium, the real value is on data. This was a decision that we made very early for ultimately different reasons. We were not, well, we are an oracle, but we didn't necessarily see the future.
Speaker: We didn't see AI coming and and eating up so much of that workforce. um we had to pivot to data just because we weren't competing on platform, on pre-trade analytics. right So we focused on data because, as I said, with DEXs, that was where our business that's where our business was. That's the pie that we could eat up.
Speaker: As a result, it means that um our data product is far more higher quality. I mean, we've been a great position now, having not sunk summing resources into something that can be very easily one-shot with X number of tokens, wherever is, A, right?
Speaker: yeah So we got a little bit lucky, um but we also made that decision for good reasons at the time. um Where people can use all our services and our data is on the research side.
Speaker: Me and my team are writing high quality, institutional grade research that can be accessed by anyone, there is no KYC. You don't need to be part of the bank to to tap into this research. And we link that to our data as well, which using services like Claude, which now everyone's got a $20 subscription,
Speaker: um is far easier to integrate our stuff into your systems. So a big um a big way we anticipate people using the DeepBook predict product that we seen ah we're building with the DeepBook team this is the prediction market built on SWE.
Speaker: A big way we see people using that product is through market making bots, um built with Claude. People are already doing this in prediction markets. yeah As I said before, we need um high quality data in order to feed into those such that those bots can then market link, really know where their edge is such that they're not gambling.
Speaker: They're actually market making. um We expect people to be doing that exact same stuff on prediction markets like on SWE, definitely doing it on Kelshi. um For that, you need our data.
Speaker: right Strange, because had you asked me this maybe six, 12 months ago, I would have said, yeah, traders need this data to make better decisions. yeah Traders aren't making decisions anymore, right?
Speaker: I mean, it depends, I guess, but... Yeah, I don't know whether it's just I'm in a bubble and I see this. um Yeah, it's it very tough to see us going back to a world where that's not the case now.
Speaker: we sort of We've seen the Promethean fire of what AI can do. yeah Why not apply that everywhere? You can't go back. Do you think um like these market-making bots, could they be created by pretty much anyone or is there like a certain level of knowledge required? Because usually market-making is like quite complex and competitive.
Speaker: I think it's a self-fulfilling prophecy. I think it is... um Firstly, you can learn anything now far easily. The access to information, firstly of the internet, or access information to zero.
Speaker: AI tooling gave interacting with that information far less of a benchmark. um So now it's it's far easier to get into everything. But that is a self-defeating curse because it means that exercising that edge that you have over less informed traders, well, you don't have less informed traders now. Everyone is poured in their back pocket on their trading suite. So...
Speaker: You are now trading against people with equivalent information to you, and therefore you would expect to have far less of an edge against them. um So I think this is playing out most interestingly and most um playing out first in places like prediction markets, where it's very easy for ProTel, Retail to connect to the API and just start market making, start punting.
Speaker: um I think it will be very interesting once large institutions start exercising that same edge against each other on traditional markets. Now, does that then lead to a more efficient market?
Speaker: You find out. I don't know. you You start to remove a lot of the edge that humans had and replace it with the edge that edge and biases, by the way, that machines have.
Speaker: They no longer have that human edge of... Humans are well known for long shot bias, us for example. yeah I want to go and bet on a Morocco winning the World Cup, which, okay, maybe in one month I'll look silly and that's the the obvious buy, but right now they are something like 5%, right?
Speaker: um The long shot bias tells you that that probability will trade slightly higher because people love a dark horse. People love to see an underdog win.
Speaker: Well, will markets look the same and will machines make the same mistakes? Probably not. Will they make different mistakes? Maybe. I don't know. Do we see that in options markets and crypto?
Speaker: I would say probably not yet. We definitely see that in prediction markets where I expect market maker bots to be quite a lot of the liquidity at five, 15 minute markets.
Speaker: Do you think that's like the way that or like the the world that we are moving towards where it's just going to be AIs trading against each other and then every market is going to become efficient?
Speaker: Because it's kind of interesting to think about this. the Like the the traditional models always like it assume this ah efficiency of markets, which is obviously not true because humans are idiosocratic and then do stupid stuff. But um so it would seem kind of natural if we just moved to a world where AI is just like trade everything amongst themselves and that's how we achieve all the perfect pricing. But I guess then the problem is kind of like that they're not actually conscious so that people still have to make the inputs so it's like still people at the end that make the decision somehow like they give the their AI the inputs and then some are better than others because they still are more informed in a way because people you can give someone all the information you can give every retail trader all the information but retail flow will probably always exist because people are like stupid and stupid people will have money and use it in stupid ways
Speaker: I mean, smart people use it in good way. Yeah. um I think I disagree with like the fundamental layer of the question, though, is that, firstly, that markets are necessarily well-informed and efficient and bar makers or efficient information discovery, um and that the only thing stopping them is human biases.
Speaker: I'm not sure that that's necessarily the case. Yes, there are a whole bunch of ways about how humans trade and how humans exercise their will in markets that cause a lot of these dislocations and seasonality, um inefficiencies.
Speaker: like If I'm a trader and I'm at a bank, then I have a P&L target that I need to make, um and I need to make that in a certain period of time. I don't have time to be correct and to be right, um which will change the way that I trade, and it will change the trades that I end up making.
Speaker: I have a risk manager that I need to satisfy. um And even beyond all the structures, even if I can fu fulfill all those in a very inefficient in a very efficient way, um we still see humans make that mistake. right That, okay, I have a risk budget, but I will not have a fillet because I'm risk averse.
Speaker: However, I'm not sure I necessarily agree that humans are the only thing standing in the way of efficient RP pricing. um I think it is a very interesting experiment and I don't know which way it will end up falling but if machines are trading against each other, yes, to your point, you cannot get rid of that human input.
Speaker: but Machines will come with their own inefficiencies, their own biases. Ultimately, these machines are only as good as the context that is fed to them.
Speaker: yeah um The only way that you will ah independently X interaction with human gets a a system which trains itself and reinforces itself to to be a better trader is by adding some sort of selection bias to these models, I think.
Speaker: So you let those that make money continue to trade. and In some Darwinian way, you allow these machines to to evolve and to pass on their quote-unquote genetics.
Speaker: yeah With trading, you have a very clear or it's a very human way of putting it. But um I think to us and other humans would agree, P&L is a good survival metric. It's what we pay our bills in.
Speaker: But it's not necessarily clear to me that um that is the only way in which you can motivate efficient pricing of assets. and It's a very self-centered way of doing it. right I am motivated to give a better price for oil, for example, because I can make money out of it.
Speaker: And I need to make money because I need to pay my bills. I need to control resources. If we have some hive-mined machine um trading against itself, well, maybe it's happy to let the other machine win because he doesn't necessarily need to make the P&L on his book because he's not going get sacked.
Speaker: Maybe he's happy to let the other guy win because it leads to a lot more efficient market. so All this to say, there are a lot of unanswered, open-ended questions.
Speaker: um And the influx of machines into this market or into this industry challenges a lot of the assumptions that we make. We're used to humans sitting at desks trading, not even twenty four seven What happens when you do away with a lot of those fundamental assumptions and not just the ones that we can think about today? What happens when we change those assumptions um even for the motivation of a machine?
Speaker: and Is it necessarily going to be the same? I don't know. Yeah, in short answer to your question, I'm not sure. i didn't want to like drift off too much into philosophical debates here, but I just thought it's kind of interesting to to go into the thought experiment maybe a little bit. But um I'm curious, can you expand a bit more? I think you'll do it a little bit earlier, but on what you concretely do in your role, like what what does a head of research do in your company?
Speaker: Yeah. so I would say two to three things i I'm mainly responsible for. So on the one hand, I'm managing a team of research analysts.
Speaker: We are collecting information. We're speaking to DeFi teams. We are reading documentation. We are using our data to provide research articles. And those research articles span um a lot of tech.
Speaker: So we are understanding V-Fi protocols, how they work under the hood, how the design decisions change the financial engineering, how those products should be priced, is there an analogue to a product that we know from traditional finance, can this be priced with no arbitrage argument, this kind stuff.
Speaker: In doing so, we have access to a lot of DeFi teams. So this is one example um is how we met the DeepBook team. We wrote to research report just understanding how DeepBook, their liquidity layer on SWE works.
Speaker: And in and the process, we got to know their team. Conversations span out about building options markets on chain. This then led to the other side of our business, which is more advisory for bespoke, more consulting level decisions, financial engineering. So i would say I would say I personally spend more of my time on the financial engineering problems, trying to design. Right now we're trying to design a prediction market on-chain and how we do that in an a efficient way that is well priced, that is capital efficient.
Speaker: I spend a lot of time prompting Claude on one of these two tasks. um Yeah. This kind of stuff. So partly management of the the research team and the research reports were writing. The other side, thinking more deeply about trading and trading problems.
Speaker: And what does ah the more institutional side of of the company look like? You're like providing structured products to institutional clients that are interested? or So we're not providing a balance sheet for those structured products to be issued.
Speaker: We are providing services, ancillary services around that. So anyone that's building our prediction markets, sorry, structured products business, like a bank, like a hedge fund,
Speaker: um they need high quality data and they need good pricing models. Ultimately, if they're going to take this risk under their balance sheet so that they can distribute it out to issuers, they need to know where the fair price is and they need to know how how an independent reference point for that.
Speaker: So, okay, this is great. Banks know how to price options. They don't need to come to us. Banks are just getting into this business now. They do not have a long multi-year history of the data that they need.
Speaker: Even Ibitop ships have been trading since somewhere in 2025. ETS on Bitcoin, for example, launched in Jan 2024. yeah They come to us because they need that data going back 2020, where Deribit had liquid auctions.
Speaker: As I said before, for Affinity Surfaces used as the fundamental building block for pricing a whole range of more exotic stuff, they come to us and say, we'd like to start selling those more exotic stuff to our clients or allowing our clients to trade those products.
Speaker: We need you for the data that we've read.
Speaker: Since you've been around for quite a while and you have maybe bit of a broader view on crypto and also the institutional side and people are just like on Twitter and looking at the price or whatever, but everyone can see, as you and me and everyone, that the price is down, we're like in the bear market, things aren't looking too good. Have you seen any, how has it changed since October 10th, I guess, the side of your business? What are institutions thinking? Are they interested? Do they still want onboard? Are they still thinking that there's a future here that more stuff will be developed?
Speaker: Very good question. And I would say since maybe not exactly October 10th, but during the fallout afterwards, we've seen in our business a very similar bifurcation of projects as the wider crypto market has. So our view going into 2026 was that If we are to get a bear market, um it is likely to compound the effect of favorable US regulation on crypto in that it is projects with fundamental value and um a clear link between the project's value and let's say their token price that will outperform in this market. yeah
Speaker: Hyperliquid is um an example I keep coming back to because they've done exactly those two things. They have managed to, in a really clever way, navigate the pivot in attention from retail traders, from other people in crypto, away from altcoins and towards other shiny new toys.
Speaker: Firstly, gold, then oil, and then even pre-IPO stocks. So perpetuals on SpaceX before SpaceX launched, for example. Hyperliquid has done a really good job of capturing that attention and making it a core part of that product.
Speaker: um At the same time, Hyperliquid's token is directly linked to the revenues of the project. je Fundamentally, that means that yeah the project that's doing well, that is providing a use case and providing value to traders, and they are reflecting that in the token value.
Speaker: Therefore, it's no surprise that Hyperliquid is one of the only coins to be up in the green years bit. yeah um We've seen a very similar bifurcation in the way that we work.
Speaker: um Those projects that are providing value, you are still surviving. um Those are the ones that arere doing well. Those are the clients that we still see and we're still serving.
Speaker: um And yeah, we're very proud to now, I think, be one of the last remaining independent providers of volatility after many of the competitors either got acquired or had to shut down.
Speaker: So crypto in these, yeah I think we can call it a bear market now. um Is it a bear market yet? I guess so. I think it's pretty consensus at this point, I guess. Yeah.
Speaker: Yeah, this is has always been the way in previous bear markets that this is a time for building, this is a time for focusing on quality, um and the real game is surviving to the the next one. This is what we do, this is what we've done in previous bear markets.
Speaker: um And it means that we are now uniquely placed to take advantage of what if you look back through our history has always been a ah key goal of ours and that's to provide institutional grade quality research data analytics um for a long time we would say this and you say well which institutions Who is here to buy your services? Well, now we can name them, right?
Speaker: Now we have large US banks, institutions, the biggest in the world, um taking our data. Now is the time for us to capitalize on having reached this point. Now is the time for us to capitalize on on the people that are buying that data now. So as an example, we are the sole provider of Bitcoin and Ethereum and applied volatility to the Bloomberg terminal.
Speaker: Huge coup for us um makes conversations with other institutions far easier. And at the same time, when we speak to those institutions at conferences, at meetups, at whatever, they are not talking about price.
Speaker: um They are not concerned about where price is going. They are far more interested in who you are, what you're building, um and how it is useful and valuable to them. yeah They are slow to move, but that works in both directions. They're slow to enter the space, but once they're in the space, they are not there as tourists.
Speaker: They are there to build and to make money out of the space, which means the horizon over which large US banks are looking to develop products in crypto is multi-year. It is not multi-month.
Speaker: They are far stickier clients and they are far more interested in how you can help them build out desks over the next year, 18 months. So huge validation for us, having survived long enough to just see them in flux of institutions.
Speaker: Yes, prices are down. um We see the marginal buyer and the marginal entrance to the space as far less interested in headline value.
Speaker: um The retail, I think, is still in crypto. They're just trading other things like gold, equities, oil. There are far more shiny toys to play with now than there have ever been in crypto.
Speaker: Why would I go to a meme coin when I can take a view on prediction markets? yeah and stocks have probably taken a lot of mine share as well since they kind of move like coin sit in 2021 so even if you're just in a normal stock market as a retailer it's kind of more interesting than buying coins that do nothing exactly but i think i think this was very like this is a very uh i kind of appreciate you sharing this to like broaden our perspective a bit because i I don't take myself out of this, but like we on this podcast kind of fall into doomers sometimes a little bit.
Speaker: Because if you talk to crypto participants and um like people that are on Twitter every day and like trade the market and all of that stuff, when the price is down, it's always kind of difficult to find the the good cases for why it's worth to stick around, especially if you see like first it was metals and then oil and then now stocks. everything else is going up and we're just kind of here and nothing is really happening and obviously hyperliquid is one thing that everyone can kind of cling to and it's great and it's doing well but the space is still centered around bitcoin and bitcoin is kind of like not that easy to make a bull case for anymore and ethereum
Speaker: at this point at least failed pretty strongly in that and Solana is also not doing so well. So I think it's very important that you also keep in mind that the institutions are here and they still like do stuff, they use stuff. like Just the other day I think BlackRock did some announcement with Athena and they're still like into tokenization, real-world assets and building stuff using the rails even if Ethereum maybe won't go up again. It's still like being used and stuff. so I think it's it's very good, especially during the bear market and when prices are low to keep this in mind because I think crypto is mostly based on reflexivity and then as prices are down everything feels kind of bad but if the prices were higher now we would all feel a lot better about what's going on because there are a lot of good things even if there are a lot of bad things but this is also like a good time to like wash out all the the trash that we don't really need and then
Speaker: as you said, like keep building and and the stuff that will survive and is positioned well will do well in the next cycle, whatever that may be. 100%. I'm glad to be able to provide a slightly more rosy picture.
Speaker: I think me and my team are in a very privileged position in that we get to see behind the scenes what people are building, what people are talking about. um And really, I think the the benefit of talking on podcasts like this is expressing that view, right?
Speaker: are I'm very privileged in that I get to speak to guys building this. From what I see, those guys are still building, still nose to the grind, putting together projects with value.
Speaker: So for me personally, it's still an incredibly exciting place to be working and to be building. um Yeah, I'm not checking the price that often, to be honest. I am, but it's still an exciting place. I want to ask a bit more about some of your insights into prediction markets because I'm not too deeply into that. But many people have like talked about it for last couple of months, so because it's one of the last interesting things left apart from perps and all of that stuff. and Options are always interesting. I talked a lot about options ah on on the last couple of podcasts as well, but I think most people don't really like, they're just like not used that much and maybe prediction markets are a way for that to to be brought a bit more into retail. So what what's your view kind of on on the prediction market space, HIP4 and all of that stuff? Where is that going?
Speaker: um I think it is an incredibly interesting way for the markets who have evolved, right? At least a lot of it is. is a timing thing um because Polymarket, CalSHI, these were not the first prediction markets.
Speaker: fine We have had prediction markets in some level in crypto for a long time, even four years before. I think the timing ah really allowed them to take off, both in terms of capturing the attention from the US election and then the subsequent results of the US election, Polymarket had an implicit bet on Trump winning the election because they would then result in favorable regulation. Yeah.
Speaker: And we see quite a lot of that playing out even now in the battle between the federal um jurisdiction and states control over, let's say, gambling markets or sports books.
Speaker: That, I think, would be a much harder conversation for prediction markets had they not had a um a favorable presidential administration. So at least part of its timing.
Speaker: um' On the other side, it's bizarre to me that retail markets have have needed this shift in order to take something with convexity.
Speaker: So for a long time, our business and even our clients business was looking for ways to educate and to grow the market in retail for options in crypto. If you look at the way retail trade options in equities and in traditional markets, S&P, zero data expiry options are a massive part of the volumes of options, right?
Speaker: Not just that, options markets in traditional finance are many multiples the volumes of the underlying market. So much more is traded on those derivatives than in spot or in linear instruments or in futures.
Speaker: So it's strange to me that that pattern in behavior doesn't pass over to crypto. Instead you have people trading, it it may just be an access thing, you have people trading perps which fulfill that goal or that need for them for for high leverage.
Speaker: um and so far and for for such a long period of time that was the case so when prediction markets come along and you find people are willing to trade really short five minute 15 minute markets um with convexity this is great news for us right because now we have a product in the hands of retail that they like um um Now, the tough part for us is then how do you use that liquidity and that attention and that demand to bootstrap liquidity, attention and demand further out with the curve?
Speaker: So for other assets, for altcoins, for others, how do you use that those prices to reflect um bigger liquidity in vanilla markets? How do you get more of that attention on vanilla calls and vanilla puts?
Speaker: And then how do you get it fundamentally um the demand for optionality away from 5-minute and 15-minute markets to real tenors, real maturities.
Speaker: yeah way Traditional finance um trade options, they are looking at a far different part of the curve. Ibit options, for example.
Speaker: um Options that settle to the BlackRock ETF. i If you look at the Bloomberg terminal and you look at an implied volatility curve for that, you see implied volatility at tenors one month, two months, all the way out 10 years and 30 years. right Oh, really?
Speaker: That's real, real vega, we would say. yeah um Vega being sensitivity to implied volatility, much larger at longer tenors. On Deribate, for example, the longest tenor that's listed is at most 12 18 months.
Speaker: So the real disconnect between the way that crypto native people trade options and the way that traditional finance participants institutions are used to trading auctions on the same underline You have people trading mainly in different directions. I think you have demand for volatility at the front end and you have supply of volatility through things like structured products at the back end.
Speaker: I'm a bank. I want to sell optionality on Bitcoin. So you have these two parts of the curve pushed in different directions by different people that are not meeting each other in the middle.
Speaker: I think it is the job of people like us and for the people that we work with to find a way or a product that marries those two. How do you get the buyers of volatility at the front end to meet against the sellers of volatility at the back end?
Speaker: This is an interesting problem. This is an unsolved problem, I think. um And it's one that's going to become more and more extreme. And i think it's going to be exacerbated a little bit by massive stablecoin balances, which right now are struggling to find ways to to earn a yield.
Speaker: Genius Act, as an example, Genius Act plus Clarity Act. um going through the US Congress now, this explicitly prohibits the earning of bank-like interest on stablecoin balances on check.
Speaker: Well, structured products and options trading, this is another way to get yield um by taking on risk, sure. I think if that big stablecoin balance that's sitting on chain and otherwise idle is able to be directed towards options, then you may be able to find more of a supply at the the volatility curve.
Speaker: Is that going to happen anytime soon? No. I think this is a longer term problem for us to solve. However, it is going to, in the meantime, have a very interesting impact on the volatility surface.
Speaker: The shape of that volatility surface will change. We've already seen that happen. Volatility is far lower now, now that we have institutional sellers of volatility selling call options, covered calls, selling straddles um against IBIO options. We're already seeing that impact on firstly Bitcoin and then Ethereum.
Speaker: Bitcoin's volatility, especially over the summer, has reached all-time lows. um Can't quite trade negative, but I think it's going to try. Ethereum um volatility reaching just above Bitcoin levels, which is bizarre. If you think of 18 months, two years ago, Ethereum's implied volatility was twice that of Bitcoin.
Speaker: yeah Now it's touching. Very strange. Options on Ethereum have never been this cheap on a relative basis to Bitcoin um in the last two years.
Speaker: who Where does that go? I don't know. Do we see a rejection of that support level and it bounce back up? Or is this the new regime? Can you do a technical analysis on on two different implied volatilities?
Speaker: Now we are getting into human biases and where machines might be able to perform. Yeah. How do you see... Do you have maybe a bit more insight into... Because that's something I don't really think about at all because I'm kind of like in in the anti-government camp and whatever. But obviously ah we all have to like contend with regulation and all of that stuff. And I think a little bit of the consensus view right now is that it's smooth sailing as long as Trump is still in office. like We're all gonna make it. Crime is legal. You can do anything. the president is a scammer. We can just like continue. But it is quite likely that the Democrats will come into power in 2028 and then everything will unravel and Hyperliquid will have KVC and and where it's all over. and um But how how does that mix with like you saying that institutions who there for the long term, they're like building for years in the future, because they must also kind of like have a view on this risk?
Speaker: like what What would they think about it? Very good question. I think the risk is far shorter term than 2028. I think we have midterms now, this year in November.
Speaker: um And that is incredibly important for the passing of legislation that is not just created and signed by the presidential administration, we're getting it through the House and the Senate. um Democrats are in with a good chance of changing a lot of the majority that the the Republicans had from um from before the US presidential election.
Speaker: So the risk is shorter term than 2028. um However, the way the institutions are thinking about it, I think these guys do have more of a sway um on legislation than we think.
Speaker: And I think we are past the high watermark of regulation against or enforcement um of punitive regulation against crypto. I see.
Speaker: I think it's an incredibly good sign that people like JP Morgan, like Standard Chartered, like proper stalwarts of the TradFi world who are getting involved in quite detailed, um quite high risk stuff in crypto. It's because of building stuff not on private blockchains anymore, but on chain. but They would not be doing that with a little bit more confidence. however Do I have an insight into regulation and how that shifts? No.
Speaker: um If I had told you all of this before the the US s election, I think my view would have been far different. We have had a fundamental shift in the way regulation is being approached, at least in the US.
Speaker: And i was as a result, we're seeing that send ripples through the way that regulation is approached in other jurisdictions. We almost needed the U.S. to move first for the rest of the world to follow. Always seems to be the case in traditional finance.
Speaker: That has happened. I think we we are past a point of no return in terms of adoption of technology. What that adoption looks like, I think, is not nailed on. I think that part is less clear. So, yes, these institutions will continue to to build in this space.
Speaker: In what capacity? Far less clear to me.
Speaker: I like this. There's a lot of lot of good hope here. It's quite a positive podcast episode. we've Since we've covered ah lots of topics until now already, is there anything else that you would you would like to share? Anything if you think we should go over it that's like interesting we haven't spoken about yet?
Speaker: I think we've we've touched on most of the things that I have salient thoughts on, yeah, which is ah usually doable in and just under an hour. umm I think this has been a whistle-stop tour of all of the stuff that Block Shoals does.
Speaker: um yeah We are a a an incredibly privileged position in that we're speaking to people. We are builders in this space. I would say yeah most people spend less time on Twitter, more time building real projects, speaking to people who are building stuff.
Speaker: ah Despite us being in what is ostensibly a bear market, this has been one of the more productive um periods of time for me and the rest the team at block shoals not just because of help that we have poured but the conferences that we've been to the people we've spoken to um in that six month period have been incredibly positive so yeah little extra hopium at the end there um we we are far less focused on headline price levels and i think far more focused on fundamental value and what you're building i think that's the same for the people we speak to
Speaker: I think that's a very good way way to close it. Very, very hopeful. I definitely found this very interesting and I think the listeners will agree. So thank you very much for coming on. Thank you very much for having me. your pleasure.
