Epicenter - Learn about Crypto, Blockchain, Ethereum, Bitcoin and Distributed Technologies - Zubin Koticha: Opyn – The Release of V2

Episode Date: April 1, 2021

Zubin Koticha, CEO & Co-founder of Opyn, joined us again on the show to give us an update on Opyn v2.Opyn v1 was focused on insurance. It's successor, Opyn v2, offers a more traditional options infrastructure. We discuss the problems of building trustless options infrastructure: choice of collateral, partial collateralization and efficient trading mechanisms for options.Topics covered in this episode:An overview of Opyn v1 and the different options created by DeFi usersHow options in DeFi workWhat is new in Opyn v2How the oracle specification worksCapital efficiency - nesting options for collateralizationFungibility of different optionsWhat makes Opyn different to other option platforms?Options and the AMM designEpisode links:OpynEpicenter Episode 344 - OpynA beginner's guide to options: Opyn v2YieldSpaceOpyn GitHubOpyn on TwitterZubin on TwitterSponsors:ParaSwap: If you want to make a swap at the best price across the DeFi market, check out paraswap.io/epicenter. ParaSwap’s state-of-the-art algorithm beats the market price across all major DEXs and brings you the most optimized swaps with the best prices, and lowest slippage. *Terms and conditions apply. This episode is hosted by Sunny Aggarwal & Friederike Ernst. Show notes and listening options: epicenter.tv/385

This is Epicenter, Episode 385, with guest Zubin Kotitsya of Open. This is Friedrich Ernsd. And this is Sonny Agarwal. And we're here with Zuban Kutija of Open to talk about Open version 2 today. Hi, Subin. Hey, hey, everyone. How's it going? So happy to be back on Epicenter. Yeah, it's super nice to have you back.

So before we talk about Open, we'd like to thank our sponsor for this week's episode. episode. This week sponsors Paraswap. Paraswap is the fastest and most liquid dex aggregator on the Ethereum blockchain. Its state-of-the-art algorithm beats the market price across all major dexes and brings you the most optimized swaps with the best prices and lowest slippage. Check out paraswop.io slash epicenter and claim 50% gas refunds for a swap of at least one eth traded. So, Subin, we didn't actually have you on that very long ago. I think it was last summer. Since then, you have released Open Version 2. Let's talk about how version 1 fared first. So in a very small nutshell, can you kind of remind us how Open Version 1 worked?

Right. So Open Version 1, which is actually still live because Ethereum, and is still trading for assets like uni and smaller assets was really built for insurance. So the high level is if you are kind of in an insurance mindset, what you want is to ensure that any assets you have on a protocol will be safe. Right. And what that means is that you'll able to get their fair value for them. So if I have a $100, in compound and compound gets hacked, I want my insurance to pay me $100. The way we did this in version 1 was that we said,

what if we give people put options, which are really commonly used as insurance in traditional finance, what if we gave people put options on their C tokens, so on the actual deposits in compound? And so in the normal case, you can take your C tokens, which represent your deposits in compound, and go to compound and say, hey, can I have my $100 spec? Compound will give you $100 USDC back. But if that failed, you had this option to sell your C tokens on Open for exactly $100.

And so even if you couldn't get $100 from compound, you could get $100. from your open option. That effectively was insurance on assets. Insurance is still an extremely interesting and huge use case in crypto and especially on chain. But what we found very quickly was a lot of our users were really excited about being able to hedge financial risk. And that was much better done through a traditional options infrastructure.

And so we moved from using options for insurance to using the financial insurance part of options, along with other options use cases in a much more broad way. And that was what kind of informed and motivated the development of V2. And so V1 was extremely successful. After kind of pivoting from pure insurance to options, we had more than the development. $100 million of volume in the first year. And people traded options on assets like Uni on Wi-Fi and on comp before they were ever tradable anywhere else.

So those were some of the first ever derivatives trades that happened on some of the most important and foundational crypto assets. Was the insurance really going to be your final use case? Or were you kind of always intending to move towards this more financial office? options thing? And was the insurance more of like a way of explaining it like, you know, I feel, you know, if you go rewind like a year, maybe a lot of people in Defi didn't even understand options. And was insurance just a good way of explaining to them how it works? And this was always the long term goal. Or was this something that like evolved as the market evolved as well?

It's a great question. So I think options have always been interesting and they've gotten way more interesting in the last year. I think, I'm not sure that I was on, like, near the time of Black Thursday last time. And since then, options have been, like, such an interesting and foundational part of, like, this new wave of the financial system. You can see things like the GME, like, gamma squeeze as such an instantiation of that. And we can get into that more. at the time and still to this day options are a huge use case and getting bigger and so is insurance.

But at the time we were really focused on insurance. Our first use cases were like compound deposit insurance. And you had Hugh from Nexus Mutual at a similar time. And it's very clear that Nexus Mutual has been doing incredibly well. And there is a big demand for people. to protect their assets from hacks, et cetera, on chain. And so that was really the original use case, but it became very clear, very quickly,

that our infrastructure was even better poised to support options. And V2 is a 10x improvement in that direction built specifically for that use case rather than for the insurance use case. So that use case being hedging financial risks, or something else. Yeah, the use case is hedging financial risks

and being able to make certain bets using options in general rather than like the pure insurance use case. Can you give an example of a primary use case that you see for options right now? So basically where would I as, you know, your very average defy user come to use options and what for? Yeah, this is a great question. So let's start with the use case that most similarly reflects the insurance use case,

which is financial risk. So right now we're at a pretty kind of frothy point for the markets. Things are going pretty well. You know, Bitcoin passed 50K. and it's very kind of there's a lot of optimism. And so we could see Bitcoin 100K, we could see Bitcoin go down. It's very unclear, right? And so is this a top?

Is this not? Well, I think most people in crypto want to continue to hold crypto assets. But they also are aware that there's these cycles that crypto goes through at pretty, you know, continuous intervals of going up and going down. If I bought a put option on Bitcoin or rap Bitcoin with a strike price of $50,000, that means that, and I also hold Bitcoin on the side, that means that if Bitcoin goes up, I have all that upside. And so I'm still like 100% long hoddling.

If Bitcoin goes down below 50K, my option, my put option gives me the right. right to sell Bitcoin at 50K, no matter how low it goes. So if we see 20K bottom, whatever we see, I'm capped in terms of my downside. I will not lose anything for every dollar Bitcoin goes below 50K. And so that's a huge use case people are using options for. The other use case is trading volatility itself. And so what options and generally non-linear instruments, which is what a non-linear derivatives, which is what an option is, since it's non-linear in its payoff, it has convexity, right? What I mean by convexity is that I have infinite upside, but I have capped downside. So an option contract effectively,

is because it's non-linear, it doesn't just matter if, you know, Ethereum or Bitcoin is expected to go up 10, 20% over the next year. That is somewhat relevant, but even more relevant is how volatile will Bitcoin or Ethereum go up? Because with a non-linear instrument like an option, if I make money, if it goes up, and I don't lose money if it goes down,

I want the underlying asset to be as volatile as possible. Because no matter how low it goes, I don't lose anything. And if it goes up insanely much, I gain a lot. And so the wider the distribution of final prices, the more that option is worth. That assumes you have a basket of options, right? If you have one option, why do you want high volatility? So it doesn't matter if you have a basket of options. or even if you have one option,

every option has something called positive vaguely. That means that it pays you money and it goes up in value if there's a lot of volatility in the future. So the way to think about this is like, let's say there's only two prices Bitcoin can take a year from now, right? and one price, I have an option with a strike price of, let's say, $50,000, okay, which is what Bitcoin is at right now.

And Bitcoin at the end of the year can take a price of $100,000 or zero. This is called the binomial option pricing model. We're thinking of it in terms of binomials. There's two possible values it could have in the future. So if Bitcoin ends the year at $100,000, my option gives me the right to buy Bitcoin at 50, which I will obviously exercise. I would love to buy Bitcoin at $50,000 if it's trading for $100,000 on the market. But if Bitcoin goes down to zero, my option is worth zero, right?

And so no matter what, no matter how low it goes below $50,000, it's zero. So I have, let's say, a 50% chance of going up to $100,000 and a 50% chance of going down to zero, my average payoff of this option is $25,000, right? Because 50% chance the option is worth $50k, 50% chance the option is worth zero. So if Bitcoin could take a price of $100,000 or zero, what if Bitcoin had less volatility? say with 50% chance it ends the year at 60,000 and 50% chance it ends the year at 40,000. Well, my Bitcoin option, my call option, still has a strike price of $50,000. So it's either worth $10,000 a year from now or it's worth zero. So basically, its value should be $5,000. So the

wider the final distribution of prices is, the higher priced options are. So any option will gain value as a result of volatility. So this is extremely useful to traders because you can use option prices to figure out what the market expects volatility will be in the future. And this is called implied volatility. And so Wall Street's Fear Index is something called the VIX, right? And so people look at the fear index to figure out how scared is the market, how volatile does the market expect asset prices to be in the future. And in times of distress like Black Thursday in the early kind of days of the COVID crisis, VIX spiked. And so traders that held options or traders that had.

helped VIX, we're able to hedge their risk as a result of future volatility. What does it mean to hold VIX? So VIX is an index that's built from option prices. And so another way of saying hold VIX means that you're kind of betting that volatility is going up in the future. And there's a bunch of ETFs, exchange traded funds that let you do that in traditional finance pretty easily and they're very liquid. But crypto doesn't have something like that. And so that's a huge use case. And you can imagine how volatile crypto is. That's one thing that makes

it so interesting from like a financial market's perspective. So that would be hugely helpful. Can I just put in there? So if you apply this again to me as, you know, your average defy trader, Of course, it's interesting to me to see what the market thinks, how volatile this ecosystem currently is. But how would trading on these financial instruments benefit me personally more than, say, buying futures? Right. So futures are really cool in that they allow you to take very leveraged bets on the price of the underlying. but they don't allow you to trade volatility itself. So what I mean by this is there's billions of dollars of value traded in traditional finance of people who are literally betting volatility is going to be 100% per year next year.

Or like, no, it's going to be 70% and one of those people is going to make money because they're going to go long volatility. And one of those is going to lose money. And so a defy trader that's very savvy will be able to make a lot of alpha and generate returns by betting on volatility and being potentially right. So that's one entire use case. But basically if someone bets on volatility, someone has to bet against volatility, right? So basically this is very much a zero-sum game. So I would expect there not to be a huge market for this.

So typically, if when you trade in the market, you expect the entire market to go up over time. So basically there's an upside for everyone, I mean, on average, of course. But for these assets where you have, for someone to be a winner, someone else has to be a loser, those typically don't scale as well, do they? Actually, they do. So the options market right now is a $300 trillion. market and generally derivatives markets are almost all zero sum. So if you think about an interest rate swap, right, one person is literally paying and one person is getting paid. And so it's

exactly zero sum. A futures market is the same where one person is going to lose money at expiry and one person is going to gain money. And these markets are far larger than the underlying markets in traditional finance. The reason why they're so big is because even if someone loses money on the derivatives trade themselves, they might be using that derivatives trade to hedge a bigger position they have elsewhere. And so it's possible that even if it loses money, it's part of a larger portfolio that still gained money due to fundamental value. You know, most people book flights on travel aggregators to get the most options and best

prices for their travel plans. So when you're making DFi swaps, use Pereswap to optimize your trades with low slippage. Paraswap beats market prices across all major dexes. This is thanks to Paraswap pool and network of professional market makers that offer the best prices across Dexas with zero slippage. Paraswap is dead simple to use and it's growing fast. They're constantly adding new features and integrations, and soon they'll release the V4 algorithm, which is faster and cheaper to use.

In fact, you might already be using Paraswap without realizing it, since MetaMask, Argent, and monolith all rely on the pariswap API. So give Paraswap a try at pariswop.io slash epicenter. And when you use that URL, you'll be able to claim a 50% gas refund on your first swap of at least one eath. This offer is available until May 1st, and refunds will be made every Friday starting April 9th. We'd like to thank Paraswap for their supportive epicenter. Okay, so before we actually dive into V2 and how that's different from V1,

maybe I can pull ahead a question that kind of fits in nicely here. So if you look at options in the defy space, they haven't yet found a large amount of traction, right? So basically seeing that you are so bullish on them and seeing that they apparently have a large amount of traction in the legacy financial system, what do you attribute that to? Right.

this is also a very interesting kind of area to dive into. So I think I'll first talk a little bit about options, markets in traditional finance. And there's kind of a third use case we haven't talked about yet, which people will find, I think, the most compelling. And then go a little bit into crypto and then defy. So I think, you know, as I just mentioned, options are a $300 trillion market every single year, which is way bigger than the stock market. And what's really interesting about options is how much they've grown in terms of retail interest. And so users on Wall Street bets, it's very clear that they use options to make these very outsized, highly leveraged position trades, as well as to buy puts to hedge their downside.

And so that's been a very large growth story in traditional finance. And increasingly, if you look in traditional markets, the percentage of options volumes that is traded by retail investors has gone up with time. So it's starting to become more democratized in a sense. And that has lots of kind of different effects on the financial market structure that are interesting and yet to be seen. but it is a bigger and bigger part of the narrative of traditional and legacy finance markets. And what's interesting is that options are even bigger than futures in traditional finance. They're more than seven times larger as a market, but in CFI and in crypto, they're still much smaller, right? They're less than 5% of futures volume.

And the reason is they're just so much more complicated. Kind of, as I described, there's a non-linearity. There's all these Greeks, which means that there's like five different variables that affect an options price, and each one will affect it in different ways. And that's what the Greeks represent. And you use Black Shoals and all these very complicated formulas binomial option pricing model to figure out what an option price is. And so getting market makers to market make,

these in size is increasingly kind of a difficult thing, especially in a market that's as complicated and as volatile as crypto. But we're seeing really big growth. And Deribut now, you know, used to do less than, you know, $10 million a volume day is doing more than a billion dollars of volume a day on good days. So for options. So it's very clear that there's the, the growth story in C-Fi and Bitcoin options are becoming much and much bigger. But I would agree with you. I think in defy, there's still a lot of room to grow, right? And if we look at kind of uniswap volumes and we look at perps and futures volumes,

those are kind of way bigger than option markets in defy so far. If we look at the way that defy trading kind of emerged, though, it started with spot volumes. going up. There was a time where there was almost no dex volume in D5, but then AMMs came about and we started to see so much spot volume. And then we saw perpetuals be huge on BitMex and on C-5, but not yet on D-Fi. And now those are starting to grow a lot. And I think that just because of the complexity of options, they are just so hard to get right. We're seeing a lot of growth now, right? We've gone from zero to $150 million of volume in less than a even. but that's going to continue to kind of compound at a massive rate in the future and going to

eventually eclipse futures volumes if defy goes the way of CFI and legacy financial system. So what's your strategy of how you guys are like, you know, maybe helping accelerate this? Is it, are you trying to educate current crypto and DFI people on options? or are you trying to bring in the expert options traders from the legacy world and bring them into crypto? Right. There are a few challenges that remain for getting either of those user groups onboarded in Defi. And I think that it's more than just a matter of time and getting the people interested. Right now we're in a position where lots and lots of people want to trade options in size.

want to make really big bets, they want to hedge positions, and they want to trade millions of dollars of Notional. And you can imagine especially whales that are more kind of financially savvy want this kind of product. But there just isn't the market maker volume yet, and there aren't kind of the mature AMM infrastructure yet to serve as that kind of demand. And so right now I think it's less of a question of bringing people into the defy options world from like a demand perspective and more bringing liquidity in from the supply perspective. And so that's what we're focusing on right now. And then the other big thing that we're focusing on right now is just the reality of how expensive it is to trade on chain. Right. And so if you're selling options,

you're trying to get yield, right? You're trying to get a premium. for selling that option to someone. And if you're paying $200, $300 to do that, and you would have made $500 to $1,000 of premium for that month, it really eats into your profit significantly. So the really big things to solve in order to get options volumes into a serious kind of part of the defy infrastructure is liquidity and like some of the gas questions.

And so layer 2, et cetera, on that. So that makes sense for a lot of the defy stuff, right? But what about like Deribit? Like you mentioned like even Deribut's volumes are lower than Bitmex, perks and stuff, right? But like as we mentioned, the volatility in crypto is probably higher than traditional financial markets. And shouldn't that be attracting all of the like Tradify options traders to Deribit? But why hasn't that happened yet? Yeah, that's a great question.

So I think there's a few reasons for that. The first is we're seeing options volumes grow extremely quickly to the point where, yes, Derivit is still doing less options volume than Bitmex is doing Perps volume, but the growth is very much there and kind of staggering. So I think it's just an eventuality, not necessarily that we're going to eclipse perp volume, because perps are this kind of new and very interesting derivative that doesn't really have an analog in traditional finance. But we're definitely going to see like this compound interest, not compound interest,

but compound rate of growth for options markets that will leave us with really large markets in the long run there. So, but how do we get there? I think that it's going to require a lot of the more professionalized market making infrastructure and firms to become comfortable with some of the risks. So some of the custody-related risk, Oracle-related risks, counterparty related risks. I think another thing to keep in mind is that the markets for derivatives in the U.S., especially are very formalized.

So there's clearinghouses, a bunch of infrastructure for banks to talk to each other, a bunch of infrastructures for like financial insurance in case like an exchange goes under. And a lot of that formalism kind of needs to come into this defy, not even defy, but a lot of this formalism needs to come into CFI before that's attractive for, for traditional and legacy market makers to take it seriously. But I think it's just a matter of time. Cool. So I'm all for talking about caveats and pitfalls,

but I would suggest we kind of put the host before the card. So let's talk about version two and how that's different from version one first. Zubin, could you explain what kind of changed in this upgrade? Right. So the goal of Open version 2 is to be the most capitally efficient and powerful options protocol in DFI. And the way we do that is a few kind of routes we tackled. The first is make user experience way better. And there's a few things we did for that.

The second is pure capital efficiency improvements. And we've gone after that as well. And the third is developer experience. So when it comes to user experience in version one of the protocol, and I think this is something Sonny talked about, they were physically settled options. So if I wanted to exercise my kind of Bitcoin call option that I talked about earlier, let's say Bitcoin is one day trading at $100,000 and I have a call struck at $50,000 with a strike price of $50,000.

that means that I have to the right to buy Bitcoin at $50,000, even though it's trading at $100,000 on the market. And in order to exercise that, there's two ways that I could do it. The first is through physical settlement, which means I literally come with USDC or die or tether or other stable coin. and I exchange that with one Bitcoin that my counterparty gives me. And so there's these physical assets that are delivered and a real trade happens. And you can imagine that this can work in CFI pretty well. But in DFI, there are a couple of problems. The first is it's a huge liveliness requirement for the user.

And so you can imagine a user just wanting to like fire. and forget, and if their option is in the money, they don't have to necessarily pay the gas costs to exercise that. So that's the first thing. And the second thing is it really makes it harder to create perpetual strategies, kind of like the VIX that I talked about, crypto VIX that I talked about, because there needs to be a literal exercise event. And then also, So on top of that, it's a, if someone forgets to exercise, they actually lose all of their money. And so that $50,000 that they would have gained from that option that I talked about, if Bitcoin expires at $100,000 and they have that call at $50,000, they would like have just lost all that money if they weren't able to get their transaction finalized in time. And so it just adds like a real big amount of stress for a user.

So in V2, we've reimagined it and made it cash settled. And so what that means is using an Oracle price, you can figure out what the difference in prices between the underlying asset and the strike price. And you just give that amount to the person who is exercising without them having to exercise, actually. And as you can imagine, that requires an Oracle price for the price of the underlying, but it's also very powerful because it makes user experience way better, especially in DFI.

So the tradeoffs that come with that are that there is some reliance on an Oracle price, right? So that's one potentially point of weakness that it has, but there are many improvements. improvement points as a result. The first is also like this capital efficiency point of view. Earlier, you would have had to probably either have a flash loan or actually find a way to get $50,000 of USC in order to make that position exercise. But here you just need to, you just, when the option expires, you can just withdraw $50,000. So these are some of the kind of tradeoffs that that exists with cash settlement. Is it also now more permission of what types of options you can create?

Like in like V1, because everything is physically settled, let's say I wanted to make an option for like, you know, sunny token and the underlying is in Zubin token, right? These are like random token. Again, because you mentioned earlier that like V1 is still being used for like other assets, like uni and things like that. Because in V1, it's in physical settlement, it seems like, anyone can come along and create an option for whatever they want. In V2, is there like more limitations?

Like, you know, now you have to have Oracle, you have to have running oracles for everything. And like, is it more limited on what people can create options for? Yes, it is a little bit more limited in terms of what people can create options for in the sense that there have to be a very solid Oracle price. But this is actually something we kind of anticipated. And it's one of the reasons why V1 is still chugging along. but it's also something we feel really bullish about is the fact that there's going to be really good Oracle infrastructure for derivatives markets. So you can imagine that most of the derivatives markets that are going to be there in DFI are going to be people trading derivatives,

options, futures, et cetera, on assets that have liquid spot markets. And if they have liquid spot markets, they're probably on either uniswap or some other AMM. And especially with uniswop v3, the ability to create oracles on pretty much any spot market that are relatively secure oracles has gone up massively. And so basically, if there is a liquid spot market for it, then there will be a liquid derivatives market for it. And also a good Oracle. for it. So Subin, how is this checked? So basically, if I want to go ahead and create some sort of options contract, how do you check that there's actually a good Oracle for that? Right. So for right now, it's still relatively permissioned, but the goal is within the next year to get it to a place

where you can have literally thousands of different types of options contracts trading. And I think there's different ways that that can be done and I think the goal would be for some kind of governance framework to start thinking about that but I think you can look at liquidity levels and democratization of liquidity in AMM pools so uniswap pools balancer pools etc and use that as a proxy for a T-WOP Oracle being good. The one caveat there is that you need to make sure that the spot market is really large enough to prevent manipulation. But that's starting to become increasingly true where you have these really, really liquid

and deep spot markets in DFI. You just said that currently is still relatively permissioned. What do you mean by that? So what exactly is the level of... permission that I need. So you don't need any permission to create an option if the Oracle price already exists. So for example, right now we have Ethereum options trading. You can launch any eth option that you want.

But if you want to be trading, uni, et cetera, V1 is still better for now. But so in order to add a new Oracle for a new asset, that right. now is a little bit of a process for V2. So you're curating the oracles as open? Yeah. Okay. I have a couple of questions around oracles, but maybe we can answer them in the pitfalls section that I look forward to.

So can tell us a little bit about some of the other methods of capital efficiency you have. So there's one of the other pieces I know is you now have the ability to collateralize, positions using other options themselves. So can you talk a little bit about how that works and what that even means? So there's two kind of goals that we have when it comes to capital efficiency. The first is to allow people to use options as collateral for other options. And in doing so, build this gigantic like kind of network of O tokens, which all can interact in these interesting ways.

And that's really weird and interesting because you could have. have three different types of long options you're using to collateralize a vault that you have for a different type of option. Then the second part is for naked options positions, which means that options that don't have any collateral, which is O tokens, can we get the collateral down to a place where we'd say we're partially collateralized? So right now, if an option requires 100 U.S.TC of collateral, can we get it down to things that would be reasonable in C-Fi? So can we get it down to like $30 or $20 or $10? And so the first part of that is through something called Options Spread Trading.

So a spread in the options world is when you're buying one option and you're selling a different option that's very similar, the only difference is the strike price. And so let's talk about selling options for a second because selling options is a little bit kind of more involved than buying options. Let's say I wanted to sell this call option we've talked about where Bitcoin with a strike price of $50,000, right? Now, if Bitcoin goes to infinity, if it moons, then my loss is potentially also. infinity, right, in dollar terms, but my loss is capped at actually exactly one Bitcoin, right? And so if my loss is capped at exactly one Bitcoin, that's still a lot of risk, right? And as a result, the amount of collateral I need is like $50,000, which is exactly one Bitcoin.

But I could use an option that I bought, a different call option, to collateralize my $50,000 option. So let's say I'm selling this Bitcoin call option with a strike price of $50,000, and I need one Bitcoin in order to collateralize it. If instead I tried to collateralize it with a Bitcoin call option, that has a strike price of $60,000, right? That means the difference between the option I sold and the option I bought, the difference in strikes is $10,000. And it turns out that $10,000 is actually the maximum

I could lose from this position. I can no longer lose an infinite amount of money. If Bitcoin keeps going up, keeps going up, I'll keep losing money from the call option that I shorted, the call option that I sold, but I'll also gain the exact same amount of money with the option that I bought. So when Bitcoin goes from $50,000 to $60,000, I will lose $10,000, but after $60,000, I own a call option, and so I won't lose any more money. And so I can collateralize that position with just $10,000 rather than $50,000.

So I've unlocked, you know, 5x capital efficiency. And so in a way that means like technically everything is fully collateralized. It's just not being collateralized via underlying assets, right? It is still 100% collaredized just with other options. Exactly, exactly. So V2 launched with full collateralization plus spread trading. And spread trading is using options positions that you have. in order to create a more capital efficient portfolio on the portfolio level.

But it doesn't mean that you're ever like in a position of margining or liquidation. So the goal is to become, to kind of be at the forefront of defy capital efficiency and have positions that are margined and liquidatable. And that's something we're releasing in an upgrade very soon, which is our partial collateralization upgrade. which I can get into as well. Well, why do you have to build that into open rather than like saying, relying on other D5 protocols to give people margin on their collateral and then allow

them to deposit that? Well, you don't have to, but it's not like a have to or not. It's like this is not doable if it's not built into open itself, right? If you want to unlock system level capital efficiency, you need to actually build this to open itself. So I think maybe to make that more clear, it might be helpful to explain exactly how this upgrade we're releasing. And it's going to be sent to audit actually in the next few days as we're speaking right now.

But how this upgrade works. So basically, in the options world, you have strike prices, right? And as I talked about with like this Bitcoin option, let's say you have a strike price of $10,000 or sorry, $50,000. you need one Bitcoin in order to collateralize it, right? Because the maximum it could lose is one Bitcoin. Actually, it might be easier to talk in terms of puts as well, just to kind of add to it. If you have a Bitcoin option that's a put option with a $50,000 strike price,

you're essentially providing insurance to someone and you're saying, like, I will give you $50,000 in exchange for Bitcoin. Right now in every single options protocol on chain in DFI, you need to put down $50,000 as collateral. And that's what we call full collateralization. But the price of the option is not going to be $50,000. In fact, it's not even going to be close to $50,000. Let's say the option expires tomorrow.

That option might be only worth $100. So you're putting down $50,000 of collateral for a position that's worth $100. It's way more than 120% or 130% collateralization as you think about it when you think about it in Maker or Compat. It's like a thousand percent plus collateralization sometimes for these positions. So what you ideally want is an Oracle price for the price of an option and you want to make sure that you have some buffer above,

that. And so let's say instead of wanting like $50,000 to mint this Bitcoin option that's worth $100, you want $200, you want $130, $140, whatever is kind of reasonable. The problem with doing the naive approach of using an Oracle price for the price of the option is that there is no really reliable Oracle price for. every single option that could exist out there in Defi so far. It's kind of a recursive problem of you need this partial collateralization in order to have the liquid market that gives you the Oracle price and then you need the Oracle price in order to have the partial collateralization. So what we need instead is we need to figure out a way to have a upper bound estimate on what the options price could be without ever really knowing what the options price is. And the way we do this is by using an Oracle for the price of the underlying asset and applying some basic principles

of black shoals to figure out a upper bound that we're comfortable with, but also an upper upper bound that is liquidatable. So let's say there's a $100 option. We're able to calculate that upper bound price of this option is $500. We're able to calculate that on chain without knowing actually what the price of the option is. And then we say you need 2x of that upper bound in order to capitalize it. And so you still need $1,000 rather than $50,000. to collateralize that option. Does that make sense? Yeah, but there's a lot of assumptions that go into this.

I would say this because I'm a physicist, but basically all of these financial market models, they're governed by parameters that are kind of just set by experience, right? So there can always be a black swan event where the model is actually wrong because the predictive power is low. Right. So this is a great point, right? And so the idea is you need to take into account exactly what Frederica just said.

You need to have the built-in assumptions very clearly outlined, but they also need to be assumptions that are very clear and are very kind of accepted to a certain extent. So in Defi, there are a few assumptions that exist. The first assumption that exists is that it really depends which market. you look at, but it's something like, ETH cannot move more than 50% in the time it takes to do a liquidation. So this is kind of maker's assumption, which is actually 33%, not 50%, makers, maker's 1.5x collateralization ratio that is required means that a liquidation can, the assumption

they're making is that a liquidation can happen before the, you know, before the, time, ETH moves 33%, right? And so we use that as the basis for our liquidation methodology and our partial collateralization. But on Black Thursday, the prices fell so fast that that assumption was kind of turned on its head, no? Exactly, right? So you need to have a strictly higher kind of amount than that, but you also need to have the fundamental truth that if you want more capital efficiency improvements, you kind of need to make some assumptions about the model. And so you can get comfortable with the notion that there's like maybe 50%,

maybe 60%, that ETH can move before a liquidation happens. But you kind of need some level of assumption in order to make that possible. And of course, we looked at the past and we looked at Black Thursday, etc to form some of these models but there is the only way to really have no risk is to never put your money on chain or to put like 100% of the strike price on chain but that's kind of untenable to have like if we want there to be a liquid options marketplace we need to make a set of assumptions that that we're comfortable with in defy are there going to be different tranches of options that form based off what level of risk counterparties are willing to take?

I think that might become true in the long run, but that's not the goal for the medium term. That is really interesting that you could potentially have two counterparties that trust each other entering into like a bespoke contract using Open with a lower collateralization ratio. But the problem is that really hurts fungibility. And so these like multiple different tranches of collateralization, they also hurt kind of transparency to a certain extent because if you look at 2007-2008 financial crisis and you look at Big Short,

it was different tranches of collateralization and trust, not just collateralization, but also in terms of creditworthiness, that made it hard to follow exactly the amount of risk a user is taking when they enter into like a CDO, for example. example. And so we want to keep it as possible, as simple as possible, make the assumptions very clear. The assumptions of our model is that basically, ETH cannot move more than 33% without a liquidation, which is exactly like makers assumption. And everything else in DeFi is more aggressive than that.

Basically, it assumes that ETH can move only a smaller amount than that, which is even more aggressive than us. And then also implied volatility at the same time can spike to no more than 400% or 300% per year. That's still a parameter that we're going to have to set. And then that's something that will be like set by decentralized governance in the long run. But yes, it is one of the core assumptions of any defy protocol that does any sort of margining that there are some parameters set by governance that that are like that. Cool. Another improvement that would be huge for DFI users is allowing users to use interest earning collateral, right? Because opportunity costs currently are so big of, you know, setting capital aside and have it be idle.

So is that coming to open as well? Right. Yeah. So that's something we built V2 with the ability to do. So re-hypothecation is like a fundamental part of legacy financial system that also unlocks capital efficiency. Because in order to sell options, if you're still getting the interest, you would have by putting your money on compound or somewhere else, then your opportunity cost is way lower. And so that's something we built V2 with the ability to do. The thing there is that governance needs to decide. And right now, we are kind of still in the process of decentralizing, and that's going to be something we're doing over like the next year. But it's up to governance when we give up governance to the community to decide what specific interest-bearing assets they want to allow.

And so, you know, because every single interest-bearing asset has a different risk profile. And so that's something that's really exciting about V2. But also, again, there starts to be these risk metrics that the system in a whole has to govern and take care of, which will be interesting to see. Cool. And as how much will Open v2 be, will kind of determine how the option Lego bricks are shaped, right? Because so basically if you say, well, a Lego brick has to have like six pins and four holes, whatever, I mean, that's kind of not nearly specified enough. So basically, what I'm getting to is if you want to achieve fungibility of different options, so basically being able to collateralize

something with a put or a call option after the same shape or form, however you want, you kind of need to ensure some basic interoperability, right? So basically things like, which Oracles are you allowed to use? How collateralized do things have to be and so on? So basically, there's a trade-off in giving the user's freedom in the first place and then kind of ending up with all these disparate options that you can't really use together. So how are you moving in that field? That's a great point. And when we originally wrote like the Convexity Protocol paper, that was something we thought about really deeply is you want to maximize standardization fungibility in order to make sure there's like a liquid and fungible options, a marketplace

that forms. And so the way that we're thinking about that is, and it's kind of the reason that we are still focusing on just like ETH options that are like at the money or near the money. and with a few specified kind of expiry, that's the first way you ensure fungibility. But the real way to ensure fungibility is to create a single asset that represents a perpetual position on options that someone wants to take. And that one asset that is fungible completely is one that is like continuously rolling over,

and rebalancing to create some kind of desired property. And so something I talked about was VIX, right? And so VIX is in the legacy financial system an example of taking a highly disparate set of options. I think that they do like 30-day options and they have both puts and calls straddles that are like at different price. And in doing so, they have like, functionally thousands, if not tens of thousands of options that have contributed to a single

asset called the VIX. And that's one way to create fungibility in the options marketplaces is to create kind of strategies, systematic strategies that buy options or sell options at pre-specified intervals at pre-specified prices near the spot price, away from the spot price, if you can reduce that massively complicated problem space into a single asset, that would be really cool. And so that's something that we built V2 with that in mind. We were thinking about how do we create a way for developers to create these systematic fungible strategies very easily?

and the way that you do that is through something called an operator. So an operator is a kind of special permissioned actor in the open ecosystem. A user can decide whether or not to give an operator access. And when I say permissioned, I mean anyone in the world can be an operator and any developer in the world can create an operator smart contract. But a user decides whether that operator smart contract can essentially roll over their funds, can rebalance their portfolio, can deal with their margin after an expiry. And so if someone wanted right now to create VIX for crypto and make it liquid for the first time in D5, they could build an operator smart contract that did exactly that for Open. So that's how you create fungibility out of like this really complicated problem space of options.

Obviously, there's a lot of like really cool features that went into Open v2. How does this compare to like, you know, I was joking earlier with someone that like seems like there's a new options protocol on Ethereum every week. And like building an options protocol seems to be like the hello world starter kit for like how to learn how to do solidity development. And so like what makes. open different than, you know, we have all the other platforms like, you know, some of the popular ones are like hedgic and stuff, but we're seeing a lot more and more come out over time. What makes open differentiates it? Yeah, I think that's a testament to just how big options are going to be, is seeing like

the amount of interest that's coming on from the dev side in terms of options. At the end of the day, no one has really solved this yet. And so what I mean by that is until someone can reasonably make a really, really large trade, like a $100 million size trade, which is very routine in both CFI and in traditional financial system, or until, you know, literally tens of thousands of smaller users can make trades. Until that's like a reality in Defi, I think it's still like erased a who's crass. tax the coat. And I think there are a lot of very interesting different models out there. And I think that the more different models there are, we actually welcome it because fundamentally we've been in DFI for so long. We just want to see the success of DFI and the success of really cool new derivatives markets that are enabled by like a programmatic developer first financial

system, but no one has solved the problem yet. And so the ways that you could solve the problem are in many different ways, and we're seeing the different philosophies there. Our philosophy is that people need to be able to buy and sell options in size at a good price. And a lot of, and that the best way to do that is by making trades really cheap. So that's kind of moving to L2 in the next few months, which is something we're starting to do really seriously. And then the next thing is by having an AMM, where users can both buy and sell from the AMM also very easily,

which is something that no one has really done in a liquid and successful way yet. And so the reason we feel that we're kind of ahead of the pack when it comes to that, especially that second point, is just the kind of ideas that we're forming and some of the technology that we have and the people we're working with. It's just really cool to see. And there's like a lot of optimism internally that we're going to be kind of the first to crack that extremely difficult puzzle. And I can talk about a little bit some of the ideas that I have there as well. Yeah, absolutely. I have so many questions about the AMM, but before that, I have a short one. So which L2 are you looking at? Yeah, we're looking at tons of things. I think that when it comes to just a trading infrastructure

for people to buy, sell options and exercise options, we're looking at, you know, ZK roll-ups are very promising from that point of view. But when we talk about like having the entire infrastructure, the margining system, partial collateralization and AMM trading, all of that stuff looks like it needs to be done on some kind of optimistic roll-up. I'm pretty bullish on the idea of being able to take something that looks very close to solidity code, modifying it very slightly and having a very similar experience for users. And I think like everyone else who's building a defy would be feeling similar.

And that's one reason that something like optimism where the OVM, is kind of based on the EVM is really compelling, but we're still in a space in March 2021, in March 2021, where no L2 has won out clearly. So super open-minded when it comes to that. Cool. I look forward to seeing what option actually wins. Talking of AMMs.

So actually building an AMM for an asset that has a correct price at a certain point in time is actually, I would have said impossible, but basically what you said earlier makes me think maybe you've, maybe you have thought about something that I haven't thought about yet. So I mean, basically the thing about AMMs is impermanent loss. And in the case of options or other things that clearly have a right price at a certain time, this impermanent loss becomes very much permanent loss. So basically, because there's an information asymmetry between a dumb, bot and a sometimes not as dumb trader, the AMM liquidity provider should lose. But it sounds like

you kind of figured something out here. So let us in on that. To put that into simple terms that maybe people can understand, I think that it made it obvious for me was like, you know, you can take something like an option. And if it expires out of the money, the value is zero. And on Uniswap, if you have a, you know, Eath versus these like options and the value of the options, and the value of the options go to zero, then everyone will just take their worthest options and sell them to the pool and withdraw all the ETH from the system. And the liquidity providers are left holding the bag of these zero value options. Yeah, exactly.

So on Uniswap, which was like the first place where open O tokens were trading, where options were trading, you got really, really bad impermanent, basically permanent loss from drift due to the. option marching forward towards its expiry, towards its maturity, right? And as it marches towards expiry, especially if it's out of the money, that means that it's going to be completely worthless. And so you have this drift that is very clear that's going to just destroy the LPs, as Sonny just said. The thing that is interesting, though, is that Uniswap empirically works very well for a spot market where drift is minimal. So if you have an ETHUSD option, you find that the drift between ETH and USDA,

yes, there's like tons of volatility, but it's not like ETH is always like going to expire at like zero or something the way that you would have for an option. And so in fact, if a liquidity provider in the idea, deal case is able to provide liquidity at a specific price and take out liquidity at the same price, then they're almost guaranteed to have made a profit. So that's how we started thinking about options and AMM design. We were like, hmm, yes, the price of an option goes down over time. This is called theta decay, and it's the Greek that represents how much an option goes down in price

over time. So that's a drift we know about ahead of time. Is there another parameter, which actually for an option is expected to be constant over time in the absence of any kind of new information? And options are actually not really priced in terms of dollars. When traders price options and think about options and say, like, can I buy an option from you? Here's the price that I'll pay. they actually talk about it in terms of implied volatility or vols, right? And implied volatility is like basically when you use black shoals, you back it out from black shoals, but it's the way that you compare options against each other, and it's the way that you really inherently price an option in terms of vols.

So the idea is, could we say that the implied volatility of the option is staying constant? And that's how we're pricing the option in terms of its implied volatility. And so over time, even if Theta Decay kills the value of the option, implied volatility, it's constant. And so the price for offer in dollar terms should go down with time. So this is very similar to kind of the yield space idea. yield space is a paper from Dan Robinson and a few other folks, Lev Nav, about kind of how to build an AMM when you have interest rate drift. So if you have a Y token, which is a zero coupon bond,

that's drift you know about ahead of time. What they did is they took the idea of an AMM and they flipped it on, on its head and its head and said, wait, we're not trying to figure out what the price the zero coupon bond is. In dollar terms, we're trying to figure out what the price the zero coupon bond is. In interest rate terms,

we're going to hold interest rates constant in the absence of trades. Any trades that buy the zero coupon bond will make the interest rate go down. And any trades that sell the zero coupon bond to the pool will make the interest rate go up. And in doing so, we'll have this predetermined drift that's nicely taken care of. For us, we, instead of trying to price, you know, like bonds are priced, their inherent way to price and compare bonds is in terms of

interest rates and not in terms of dollars. Similarly, in the options world, the way to compare options is in terms of vols, and implied volatility and not in dollars. And so you can hold that constant and have that go up and down based on a constant product. And that's kind of an idea that we came up with towards like late summer last year. Turns out it's like really complicated doing blackshoulds on chain, blacking it out, backing out volatility from black shoals, figuring out exactly what slippage function is, how steep it is. But we've made significantly significant headway when it comes to that and have some really kind of exciting ideas there.

as well. So I know currently you guys are actually not using an AMM at all. You guys are actually using X to do a lot of your market making. Is that because, what's the reason for doing that? Right. It's because the kind of current AMM infrastructure that's not purpose built for options has the exact problem, Federica mentioned, where you have permanent loss,

and the options are just so leveraged and so complicated that a very kind of naive spot AMM is almost certain to lose a lot of money. And also, fees are not high enough as well as a percentage of premium. So the idea is the only tenable way to have like a large market maker volume on chain is to have a centralized order book model until you have. until you have like the AMM. And so that's the stopgap measure on the way to the AMM. And I think in the long run, as we see market participants get more sophisticated,

as we figure out L2 solutions, we're going to have lots of order book volume in DFI for options. It's something that we see in traditional markets. And so there's no reason we wouldn't have it be also like a large part in addition to the AMM on chain. So what's the, what's the timeframe for the AMM? When can we expect to use this? Yeah, I think AMM, that's a really good question and really hard.

I can't really promise any dates because it seems like we really want to, want to crack it in a very elegant and easy to understand way. Just because the complexity of something like an options AMM means that if there's any sort of issues, it would be catastrophic for a liquidity provider, which is one thing that gives me pause with a lot of the other AMM designs for options that are out there. It's very clear that there's catastrophic flaws in them. And I don't mean that in like, it is an insulting way. I think they're really cool ideas out there.

It's just such a hard problem. So I can't give you an exact date, but let's just say, in the next, we will be able to, we'll definitely be able to have an AMM before L2s are big in Ethereum and way, way before ETH 2.0 is a thing. So it's definitely like way before you have the ability to have massive order books dealing with a majority of volume, it you will have,

is my estimate of when we're going to have super liquid AMMs for options. I feel like you're doing this nesting collateral thing here where you answer my question by, you know, like inputting another parameter and saying, I don't know, but before that parameter. Okay, anyways, I think this is probably a good way to wrap up. Thank you, Zubin, for coming on. And I look forward to using the new AMM. I think this is going to be a huge thing if it actually works out. Yeah, so awesome to be here and so excited to be kind of working closer with the Every Center team when it comes to doing my own interviews. So fun to be on the other side of it as well.

And as always, it's really fun to chat with you, Sonny. Thank you for joining us on this week's episode. We release new episodes every week. You can find and subscribe to the show on iTunes, Spotify, YouTube, SoundCloud, or wherever you listen to podcasts. And if you have a Google home or Alexa device, you can tell it to listen to the latest episode of the Epicenter podcast. Go to epicenter.tv slash subscribe for a full list of places where you can watch and listen. And while you're there, be sure to sign up for the newsletter,

so you get new episodes in your inbox as they're released. If you want to interact with us, guests, or other podcast listeners, you can follow us on Twitter. And please leave us a review on iTunes. It helps people find the show, and we're always happy to read them. So thanks so much, and we look forward to being back next week.