Antonio Juliano, founder of dYdX, explains how this protocol for derivatives will enable you to take a short position in a single token, why they are starting with protocols shorting and margin lending and how these protocols work from how shorting and margin lending work today. The former employee of Coinbase also discusses how the protocol determines the prices of the assets involved, how low liquidity affects the trading of the derivatives and who dYdX’s users will be. He also talks about how the company, which has raised money from Andreessen Horowitz, Polychain Capital and others, plans to make money despite not currently having a token. Plus, we discuss dYdX’s plans to create protocols for derivatives that are not fully collateralized and the risks that come with that.
Thank you to our sponsor!
Blockdaemon: http://blockdaemon.com/unchained/
Interested in sponsoring Unchained or Unconfirmed? Reach out to Raelene at [email protected].
Episode links:
dYdX: https://dydx.exchange
Antonio Juliano: https://twitter.com/antoniomjuliano
dYdX white paper: https://whitepaper.dydx.exchange
Expo: https://www.expotrading.com https://medium.com/dydxderivatives/introducing-expo-ffe74a328f85
Investments in dYdX: https://techcrunch.com/2018/08/03/short-ethereum/
Unchained episode about decentralized exchange protocol 0x: http://unchainedpodcast.co/will-warren-of-0x-on-why-decentralized-exchanges-are-the-future
Unchained episode about decentralized debt protocol Dharma: http://unchainedpodcast.co/nadav-hollander-on-how-dharma-could-create-new-forms-of-debt-ep80
Unchained episode about decentralized money market protocol Compound: http://unchainedpodcast.co/how-youll-earn-interest-on-your-crypto-with-compound-ep82
Unchained interview with Josh Stein of Harbor: http://unchainedpodcast.co/harbor-and-trusttoken-on-why-they-dont-mind-being-unsexy-ep77
Unchained interview with CryptoKitties: http://unchainedpodcast.co/what-makes-a-cryptokitty-worth-140000-ep75
Transcript:
Laura Shin:
Hi, everyone. Laura here. Quick request before we get into today’s episode. I’m doing a TEDx talk soon and so I need a short break to give myself enough prep time, but instead of playing a rerun I’m going to do a listener questions episode, so send me your questions. You can email them in written form to [email protected]. If you do that start your subject line with listener queue or, and frankly this is more fun, you can record an audio file, which is super easy to do on your smartphone nowadays. You can also email that to [email protected], and if you choose to send in an audio file please state your name in the recording, and again make the subject line in your email begin with listener queue. The deadline to send me your question is by end of day Tuesday, October 9. Again, send your written questions and your audio files to [email protected] with a subject line that begins listener queue. Thanks, and enjoy the show.
Hi, everyone. Welcome to Unchained, your no-hype resource for all things crypto. I’m your host Laura Shin. If you’ve been enjoying Unchained pop onto iTunes to give us a top rating or review that helps other listeners find the show.
Sponsor:
Want to connect up to the Bitcoin network but are confused by the multitude of choices and intimidated but the complexity of getting started? Come to block blockdemon.com/unchained and launch a Bitcoin node in mere moments. We’ll do all the heavy lifting for you.
Laura Shin:
Looking to advertise your product on Unchained and Unconfirmed? Reach out to Raelene at [email protected] to find out about sponsorship opportunities. Again, that’s [email protected] to find out about sponsorship opportunities on Unchained and Unconfirmed.
My guest today is Antonio Juliano, founder of dYdX. Welcome, Antonio.
Antonio Juliano:
Thanks for having me, Laura.
Laura Shin:
Before we start I just have a funny story to tell, which is that dYdX is definitely one of the more challenging projects I’ve had to wrap my head around, and preparing for this interview was giving me flashbacks to when I was a senior in high school and I took calculus BC, and there’s two kinds of calculus you can take as a senior, and that’s the more difficult one, and math was always more challenging for me than other subjects but I don’t know, I just thought oh, I can do it, but I kind of had too much fun that summer and it took me a little while to get into the school spirit when the fall semester rolled around, so I literally got a D the first quarter, and at the end of it I went to the teacher and I was like how can I change this. Is there something I can do for extra credit, and he said yeah. Well, if you do extra credit, I can give you a D+ and I literally started tearing up and I was like oh my God because I really wanted to go to Stanford. I was like, I’m never going to get into Stanford now, and what ended up happening is he agreed to give me a P for pass and then I got an A- the second quarter and we averaged the grades with my final exam and eventually I ended up with a B- semester so it was all okay.
Antonio Juliano:
Well, I’m glad. Hopefully, the name doesn’t give too many bad memories for too many people.
Laura Shin:
Oh my God, yeah. Going through all this stuff yesterday I was just like okay, it’s not like I can’t get it but it’s taking me so long, and I had done that pre-interview with you before and I wanted to relisten to it to make sure I understood, and I had to listen to it again on .6 or .7 speed because you were speaking so quickly and I could so barely understand what you were saying, so all of this preamble is just to say go slowly, assume you’re talking to a fifth grader because I’m only barely grasping this.
Antonio Juliano:
Yeah. Absolutely. Will do.
Laura Shin:
Okay, so let’s just start with the simplest thing. What is dYdX?
Antonio Juliano:
dYdX is a protocol for decentralized derivatives and margin trading. Basically, the first thing that we’re doing is making a protocol for margin trading that lets users go short or get leverage on any for now Ethereum-based cryptocurrency. Our first product that we’re launching soon is called margin tokens, and with margin tokens, you can basically represent a short or leveraged position as a regular fundable ERC20 token. This is cool because we can basically create the short Ethereum token, let’s say, and now if you want to, say, go short on Ethereum all you have to do is buy the short Ethereum token, which you can do on our product expo or on any other exchange that lists the short or leveraged tokens, but we think of dYdX not just as a margin trading protocol but as a protocol for all types of financial derivatives. We just decided to start with the margin trading because we think that that’s the most applicable to build in the market right now.
Laura Shin:
And why did you decide to tackle derivatives, and just for listeners who aren’t super familiar with the financial services system, which in a way I’m also not. I think also the vast majority of my listeners are…or at least the plurality come from tech, so what are derivatives and why did you decide to tackle this area of finance?
Antonio Juliano:
So, financial derivatives are financial products that are built on top of…on other underlying assets, so they vary a lot in complexity, things like options or swaps and they’re primarily used by bigger institutions or more sophisticated traders to get better risk management or increased speculation on their profile, on their portfolios, so that’s kind of what financial derivatives are. We’re excited about building them because they often times bring a lot of maturity to the underlying markets. The way that most people are trading cryptocurrencies right now is basically just to buy and hold them, but that’s not at all the way that people trade traditional assets. They often use financial derivatives to hedge their bets or get different types of speculation or risk management that they couldn’t otherwise get, so that’s kind of what financial derivates are.
In response to your question about why I decided to tackle the challenge of building financial derivatives on the blockchain, so I really wanted to build something that was actually useful in the market right now, and the way that most people are using cryptocurrencies is for speculation, but like I was saying, a lot of these more advanced trading tools that are traders are accustomed to don’t exist in the market yet so we thought that this would be an interesting thing to build and something that would actually be useful in the market right now.
Laura Shin:
Well. Actually, so one thing that was interesting is in the white paper you said that derivatives will open up new avenues for speculation and yet here your kind of saying oh, we want to do something that’s different, so how do you square those two statements?
Antonio Juliano:
Yeah, so derivatives can be used for a lot of different things. The two main use cases I mentioned are either hedging, something that’s really good to do in bear markets, so a lot of these crypto hedge funds could use financial derivatives to basically hedge their long bets and protect their downside, which is something that probably a lot of people should’ve done more of in the past six months or so. The other application that they’re widely used for is increased speculation or getting leverage. You can also bet on more interesting things like volatility that you couldn’t really bet on with, basically, just buying and holding the underlying assets.
Laura Shin:
Oh, interesting. So, if this protocol had been available as of January 1, or something, then we might see a lot of people making profits off of it this bear market. Is that correct?
Antonio Juliano:
Yeah, absolutely, and actually our first product Short Ethereum, which is the short Ethereum token, would basically be a really simple way for users to bet on Ethereum going down, so a lot of people could’ve used that to actually profit in the bear market of the past six months.
Laura Shin:
Interesting. I recently had a maximalist on the show and it was Tuur Demeester, so maybe I’ll let him know and see if he wants to put his money where his mouth is. Well. Actually, that brings me to this other question because I am curious, so right now if I wanted to short ether how could I do that?
Antonio Juliano:
There a couple different ways you could short ether right now, and they’re pretty much all on centralized exchanges, so probably the easiest way to do it if you’re a big traditional institution is to buy futures on some of these centralized exchanges that offer futures products, like I think CBOE and a few of the other ones do, so a lot of the bigger institutions are using those mechanisms to get short exposure. There are also a number of centralized exchanges that do offer margin trading, such as Poloniex, or BitMEX, so those types of things where you could basically go to take a short position on any cryptocurrency that they offer, and kind of the final way that we’ve seen doing it in a bigger way are through over the counter desks, so there are certain OTC desks that do provide lending liquidity that you can use to short assets, like Genesis, for example, I believe offers OTC lending, so those are kind of the three predominant ways we see institutions getting short exposure right now, but there aren’t really any products that are super available for more individual traders to get short exposure on a number of these assets in a really simple way.
Laura Shin:
And if they’re doing it with dYdX then they can maintain control of their private keys it sounds like.
Antonio Juliano:
Yeah, absolutely. That’s a big selling point of our protocol and it’s similar to a lot of other protocols such as Dharma, or 0x, or similar.
Laura Shin:
Yeah, I’ve had both of them on the show recently. In case people haven’t noticed, I’m trying to a little series on this sort of world of decentralized financed that I am watching being built, which is basically incredibly fascinating to me. So, before we get into how dYdX works I actually also want to bring up your background. How did you get into crypto?
Antonio Juliano:
I got into crypto in 2015 when I started working as a software engineer at Coinbase. I was much different than most other people that worked at Coinbase in that I didn’t know anything about Bitcoin before that.
Laura Shin:
Wow.
Antonio Juliano:
Yeah, so I was at Princeton CS at the time and I was just graduating, and Coinbase was one of the 20 places I applied to as a senior. I kind of only applied because Fred Wilson actually came and gave a talk and mentioned Coinbase in one of the classes I was taking at Princeton, but one of the unique things about Coinbase’s hiring process is that they did a work trial, so basically during my senior year they flew me out to San Francisco and I worked with them for a week and at the time I still didn’t really know anything about Bitcoin, but everybody was just so excited about it and there were so many amazing people at Coinbase at the time so I could just see that I wanted to work with these amazing people, and they were all super excited about this Bitcoin thing, so I just decided to give it a try and jumped onboard full time at Coinbase after graduating, and then once you’re at Coinbase it’s pretty much impossible not to get excited about Bitcoin because everybody just talks about it all the time.
You get to sit down and have lunch with amazing people like Fred, or Brian, or Olaf, who just are far and away the best thinkers in this space, so just kind of having that exposure to the really best people in the space really got me excited about it, and then after I worked at Coinbase for a while I knew that I wanted to build things in this space in the future, which is kind of what led me to dYdX.
Laura Shin:
Yeah, and so fill in that part of the story.
Antonio Juliano:
I actually left Coinbase to work at Uber briefly, so I was at Uber for six months after working at Coinbase, and I knew…even when I went to Uber I knew that I wanted to still work in the cryptocurrency space and that I wanted to start a startup but my plan was to work at Uber for a year, make some more money, give me some time to work on my personal projects, but I only lasted for six months at Uber before leaving to start my own thing, and the first thing that I started was not dYdX it was a search engine for decentralized apps that I called Waypoint.
So, I built this whole thing out and basically was trying to be a Google for a decentralized apps, and I was working on it full time for three or four months, and literally nobody was using it, and that kind of led me to my realization that I really wanted to build something that’s actually applicable in the market right now, and the main problem with eh search engine thing was that just nobody is building any decentralized apps, or especially at the time, a little over a year ago, so there was just nothing to search for and it was too early in the market, so I kind of spent a weekend sitting down and thinking about the types of problems that I could solve with a decentralized protocol that would actually be useful in the market, and that kind of led me to financial derivatives.
I didn’t really know anything about derivatives at the time, but I spent a lot of time talking to a lot of good friends who had gone on from Princeton to work in the finance world, and quickly came up to speed on what derivatives are. Kind of the hard part about derivatives is not understanding how they work but more understanding how people trade them, like how an option works is kind of fundamentally fairly simple, but how people trade options is very, very complicated, so I definitely did not understand all of that at the time and that’s still something I’m working on, but I did kind of come up with basically a protocol that supports fully collateralized options, which is the first thing that I came up with, and then soon after that came up with the beginnings of the margin trading protocol, which is what we decided to start with and are launching now, so that’s kind of what led me to building dYdX and the derivative protocol.
Laura Shin:
Yeah, I want to get into that a little bit more later about how what’s more complicated is the way people trade them, because would imagine maybe part of the complication comes from the complexity of the existing financial services system itself, but…so before we get there why don’t we just go into how dYdX works. You started by talking about the margin tokens, but I actually want you to walk me through how a short sale works in dYdX on the back end. Not through the token but just step by step, and for listeners who don’t have the background on how short cell works, traditionally it would be you borrow the asset that you want to short, you immediately sell it so you kind of know how much your sale price is, and then after the price of that asset falls at that point you buy the shares at, obviously, this lower price and return them back to the lender, so in dYdX it’s a little bit different, so why don’t you walk me through how that goes.
Antonio Juliano:
Yeah, that’s a good walkthrough of how normal short sales work, and it’s actually quite similar in dYdX. At the beginning of the show I did mention or short and leveraged tokens, which are the simplest way to get short or leveraged exposure on dYdX, but actually backing every short or leveraged token is a short or leveraged position, so for taking a short position on dYdX it basically works the same way as a normal short sale in that you have to borrow the assets that you want to short and then sell it to a trader on the spot market and then lock up some collateral in the position, and then that collateral basically just sits there for the duration of the trade, and then you use…whenever you want o close your position you use some of that collateral to buy back the tokens that you borrowed from the lender and then you get whatever’s left, so let’s just walk through an example of how shorting Ethereum on dYdX works.
If you want to short Ethereum you need to borrow Ethereum from somebody. dYdX has its own lending protocol built in and basically, it’s kind of works in a couple ways. Lenders can basically offer to make loans for Ethereum by signing an off chain message, and this message contains details such as how much they want to lend, the interest fee, the maximum duration, etcetera, and then these loans are held on a relayer, which is very similar to the concept of a relayer used by many other protocols, again such as the 0x or Dharma, and then…so now if you want to short Ethereum you can go to the relayer and find an offer to lend you Ethereum. Similarly, you need somebody to buy Ethereum from you on the other side. You also need something to short Ethereum against, so say we’re shorting Ethereum against the Dai stablecoin, which is Maker Dallas stablecoin.
Then you basically need an offer to buy Ethereum and sell Dai, which you can use that buy order to basically sell the Ethereum that you just borrowed from the lender. dYdX plugs into a number of different decentralized exchanges. The primary one being 0x, so normally people will use a 0x order that’s basically offering to buy Ethereum and sell Dai in this case, so you the trader just want to short Ethereum, so you go to the relayer and the relayer finds you this loan offering, offering to lend you Ethereum, and similarly this buy order offering but Ethereum and sell Dai. Then you take those two things and send them in a transaction to our smart contract, and our smart contract does a few things. It verifies the signature on the loan offering to make sure it’s legitimate, and if it is legitimate then it basically pulls Ethereum from the lender, so now our smart contract has some Ethereum. Then it sells it to the external buyer and gets some Dai from that trade.
An interesting thing about 0x and other decentralized exchanges is normally they’re used for just peer to peer trading, like I’m a human and I trade with some other human, but a really interesting property of them is that smart contracts can also trade with them and that’s kind of what we utilize to enable our smart contract to basically trade with an external 0x maker, so at this point our smart contract only has Dai. It basically sold all the Ethereum that it got from the borrower, and then it takes some margin deposit from the short seller proportional to…and the amount of the margin deposit is basically specified on the loan offering that was offered by the lender, so now it has a Dai equal to whatever it got from the cell plus your margin deposit. Let’s just say, putting concrete numbers to this, you borrowed one Ethereum and then you sold that Ethereum for 100 Dai and then you put up 50 Dai as margin deposit, so after all of this stuff has happened then the smart contract ends up with 150 Dai locked in a position.
It’s also important to note that all of this stuff that I just talked about happens atomically in one transaction on Ethereum. A really interesting property of Ethereum that we heavily make use of is that smart contracts can call other smart contracts within one transaction, so basically, we use that to kind of do the borrowing and the selling and the locking up of collateral all in one Ethereum transaction, so it makes the process really simple. So, now kind of…
Laura Shin:
Wait. Just to clarify, so this part that you…what you described, it all happens atomically. Is that with the margin token?
Antonio Juliano:
So, this doesn’t have anything to do with the margin token quite yet, but it basically has to do with just how Ethereum transactions work, and it’s a really nice property, because say for example we do the borrow and then for some reason the selling fails it’ll kind of roll back the whole thing. That’s basically what atomicity means, so kind of all of it will happen or none of it will happen. You can’t get into a weird state where you have borrowed some of the funds but haven’t actually sold them or things like that.
Laura Shin:
Okay, so keep going.
Antonio Juliano:
Okay, so now the state of the world is that you…there’s a position and you are the short seller and you’ve shorted Ethereum. On the other side there’s a lender that you owe Ethereum back to and there is 150 Dai locked up in the position, so we use continuously compounding interest to basically determine the amount of interest that’s owed to the lender at any given time, so you the short seller can decide to come back and close your position whenever you want to, so say you wait a week and now the price of Ethereum has dropped to 50 Dai and you owe a little bit of interest, so say you owe 55 Dai back to the lender.
Now what you can do is basically you need to find somebody that’s willing to sell, this time, Ethereum for some Dai, and you’ll go to a relayer to find somebody to do this, and normal this will be through a 0x order, but I do want to emphasize that our protocol will work with a bunch of different decentralized exchanges not just 0x, so you go to a relayer and find, let’s say, a 0x order that’s offering to sell back the Ethereum that you owe to the lender for, say, 55 Dai so you send that to the smart contract and the smart contract trades some of its own Dai that it has locked up in collateral for that Ethereum, so it trades 55 our of the 150 Dai that it has locked up in collateral and buys back the Ethereum that it owes to the lender, gives the Ethereum back to the lender so the lender is happy.
They basically earned interest on their ether and were paid back, and then you the short seller get whatever is left, and in this example you actually made money because the price of Ethereum goes down, and you made, basically, the amount of collateral, which is 150, minus what it cost you to buy back the tokens that are owed to the lender, which is 55 so in this example you get 95 Dai and you initially put up a margin deposit of 50 Dai, so in this example you make a profit of 45 Dai on the trade.
Laura Shin:
Interesting. So, a couple of things. First of all, I spoke with…shoot. I think it was Robert Leshner of Compound, and they used something called collateral ratio to determine how much collateral you put up. Is that just determined on an individual one to one basis in dYdX?
Antonio Juliano:
Yes, it is determined on an individual one on one basis, so a lot of the parameters for loans are determined by lenders in dYdX. dYdX is a very open and general protocol so we let lenders specify pretty much any detail about the trade from the maximum duration of the trade to even the interest rate that they’re offering, and other things like that, so it’s more just specified by the lender and our protocol, whereas for some other protocols, like Compound, for example, is what you were mentioning, a lot of the parameters are specified globally on their smart contracts.
Laura Shin:
And the other thing I was wondering is in your example you used ether and Dai, and Dai obviously is a stablecoin that’s currently pegged to the price of one dollar, but what if, let’s say, I want to short REP with ether and both of those prices are fluctuating in relationship to the dollar then how does the collateral work in all the other aspects?
Antonio Juliano:
Yeah, absolutely. You can use any pair, basically, for what you borrow and what you put up as collateral, so you could absolutely use dYdX to basically short REP for Ethereum, and we kind of anticipate that that will be a really big use case. That’s basically just taking a bet that you think that, say, REP or whatever other token, it doesn’t has to be REP, is overvalued as compared to ether, and basically using dYdX to short that as compared to ether would be a really good way to take that bet without actually even exposing yourself to the overall risk of cryptocurrency, so even if, say, Ethereum went down by a ton but REP went down even more, so you were actually right on your bet on that Ethereum went up as compared to REP. You’d still make money on that trade, so there are a lot of use cases for margin trading where it does make sense to actually use, as collateral, something that’s not just backed to the dollar, or some other fiat currency like that, and it would make a lot of sense to use underlying cryptocurrencies such as Ethereum, or really whatever you want, as collateral as well.
Laura Shin:
That’s interesting. I actually would think it would be the opposite because ether is down so much this year, so you would have to be…well, I guess…I mean there are a lot of not so good tokens that have plummeted even more, so maybe actually it does make sense, but I just think it would be trickier. It’s like you have to sort of make more than one calculation in a way.
Antonio Juliano:
Yeah, absolutely, and you really do. Financial derivatives and margin trading are definitely financial products that do require a higher level of sophistication to trade, so you really have to understand exactly what you’re buying or what trade you’re making before you make it, but these types of trades are commonly employed by more sophisticated traders and that’s more of the target use case that we’re going after.
Laura Shin:
All right. So, let’s then describe the short token and how it works.
Antonio Juliano:
Yea, absolutely. So, in my example basically, it described, basically, just an individual taking out a short position on ether. The way that the short token works is very interesting and it’s actually fairly simple. All you have to do is basically there’s this concept where positions and loans cannot just be owned by individuals, like humans, they can also be owned by smart contracts, and this is really important because it allows you to unlock a lot of other functionality on top of the protocol, so the way the short token works is it works exactly the same as in my first example except now the owner of the position is not a human but a smart contract, and this smart contract basically is an ERC20 token, and the ERC20 tokens basically just represent fractional ownership of the backing position, so say I take out a position in exactly the way I described in the first example and I take out…I’m now short, say, 100 Ethereum so the protocol will basically mint me 100 short Ethereum tokens.
Now if say you want to short 50 Ethereum you don’t actually have to do all the complicated stuff that I described in my initial example. All you have to do is buy 50 short tokens for me and then you’ll basically own half of the short position because there are 100 short tokens and you own 50 and I own 50, and basically just the fact that the position is owned by an ERC20 token allows you to do really interesting things like splitting up ownership of it, and also due to the economics of it the price to buy a short token will always be equal to the price to mint a new short token, and minting a new short token is just done by the process of opening a position that I described before, so it’s really interesting because now only a really small number of actors need to actually do the complexity of opening and closing positions in the vast majority of users and just live in this simple world where if they want to get short exposure they can just buy the short token, or if they want to, say, close their position or exit then they can just sell the short token to somebody else and somebody else can deal with the complexities of opening or closing the position for them.
Laura Shin:
And so, with the short token then there’s not a duration baked in, right? It’s just like when I want to sell that that’s when I sell it, and it’s at that moment that it determines whether or not I profit or loss. It’s not like I’m betting on, I think the price will fall to this price by this time, and if it does then it cashes out for me, and if it doesn’t then it goes out to the full duration and I lode money. How does that part work?
Antonio Juliano:
So, there is a maximum duration, both to positions and the short and leveraged tokens, but you can also close your position before that maximum duration, so say the maximum duration is a month and you buy a short token, there are two ways you could profit on it. First is, say, 15 days through the month you decide you’ve either made or lost enough money and you sell your short token, and you basically just make whatever profit that the short token made during that first 15 days, or you could basically wait the entire duration of the short token, and we have automated mechanisms in our protocol that will basically close the short token at the correct price without you having to take any action, so you could simply…the simplest way to get short exposure on the dYdX protocol will be to just buy a short token and then wait, say, the maximum duration of the short token and then let the automated mechanisms the token for you, and then you can just come back and withdraw the payout that you would’ve gotten, and then payout that you would’ve gotten is basically just the profit that the short token made during the entire duration.
Laura Shin:
Okay, we’re going to keep discussing your other products, but first I’d like to take a quick break for our fabulous sponsors.
Sponsor:
Interested in becoming a blockchain citizen, but unsure of how to get started? Come be a part of the decentralized revolution and head over to blockdaemon.com/unchained to launch your own node on a public blockchain network or join a private network within minutes. Blockdaemon is the only multi-cloud platform that supports the leading blockchain protocols, such as Bitcoin, Ethereum, Stellar, Quorum, Fabric, Aon, and more. Our goal is to offer all blockchain projects, a global decentralized network management tool connected to multiple infrastructure providers. Go to blockdaemon.com/unchained and start a free 30-day trial today.
This ad spot could be yours. Got a great product or service for Unchained or Unconfirmed listeners? Reach out to Raelene at [email protected] to find out about sponsorship opportunities on Unchained and Unconfirmed. Again, that’s [email protected].
Laura Shin:
I’m speaking with Antonio Juliano of dYdX. So, your other product right now, I think, is a leveraged long token. What is that and how does it work?
Antonio Juliano:
It works very similarly to how the short tokens work. The only difference is that you flip the token that you borrow and the token that you put up as collateral, so actually, the protocol doesn’t really even know what is a short versus what is a leveraged long. All it knows are what the tokens that you borrowed are and then what the token that you put up as a collateral is, so let’s go through an example of how getting leverage on dYdX would work.
So, in our first example we borrowed Ethereum from the lender and then we put up Dai as collateral. If you want to use dYdX to instead get leverage on Ethereum you’d, in this case, borrow Dai from the lender and you’d put up Ethereum as collateral, so you could basically…so if you borrow Dai from the lender you’re basically increasing your buying power, so say you put up…you have 100 Dai with which you want to buy Ethereum you could just buy 100 dollars’, say, worth of Ethereum with your 100 Dai, or you could borrow, say, another 100 dollars from a lender and then use your 100 Dai plus the 100 Dai that you borrowed from the lender to buy 200 Dai worth of Ethereum with your 1090 Dai, and the way that this works is say in our example you’ve put up 100 Dai and then you borrow another 100 Dai, so now the protocol has 200 Dai locked up in the position. Then it sells that 200 Dai for Ethereum.
Let’s say in this case the price of one ether is 100 Dai, so it sells that 200 Dai for two Ethereum and then locks that two ether up in the position. Now that two ether basically just sits there for the duration of the position, but you still owe the lender back an amount denominated in Dai, not in ether, so in this example let’s say the price of Ethereum goes up to 200 Dai, whereas it started out at 100 Dai, so now the two ether that are locked in the position are actually worth 400 Dai and you still owe, say ignoring interest, 100 Dai to the lender, so you could close your position by basically buying back 100 Dai and giving that to the lender, and then basically there’s 300 Dai left and you could basically take that as your profit, so in this example you’d end up with 300 Dai, whereas you only put up a margin deposit of 100 Dai, so in this example you’d make a profit of 200 Dai because, basically, you were leveraged 2x on Ethereum whereas if you had only, say, bought 100 Dai worth of Ethereum with your 100 Dai you would have initially bought one ether with that, and then ether went up by 100 Dai so you would’ve made 100 Dai on the trade, whereas being leveraged 2x with dYdX you would’ve made 200 Dai on the trade, so our examples works very similarly to how shorting on the dYdX works except you basically just flip the token that you’re borrowing with the token that’s held as collateral.
Laura Shin:
Super interesting. One other thing I was wondering about is how do you guys determine the price of the assets because obviously, all these exchanges have different prices?
Antonio Juliano:
The way that prices are determined on dYdX is that each trader can select their own spot order that they want to use to open or close a position, so say, going back to our initial example of shorting ether, I mentioned that the price for ether in that example was 100 Dai for one ether, and basically the reason that that price is what it is, is because that’s the best price that is offered by a 0x order to basically trade Ethereum for Dai on 0x, so basically the trader has an economic incentive to pick the best order for themselves because they wouldn’t…there would also be orders on the book for, say, trading ether at 90 Dai, but you want to open your short position at the highest price possible, so traders will always pick the order with the highest price for themselves.
So, it’s very similar to how just limit orders work on exchanges, and traders always have economic incentives to basically pick the best order for themselves, or similarly say 0x was offering a price of 100 Dai was the best price you could find on 0x, but Harbor or Oasis, or something else, like some other decentralized exchange like that, were offering a better price then you could open your position with Oasis instead and basically get a better price for your trade.
Laura Shin:
That’s interesting. As far as I understand, for margin tokens when you close your position those tokens get burned. Why is that?
Antonio Juliano:
The margin tokens basically represent your ownership in the position, and when you use…basically the reason that margin tokens have value is because they allow you to close a fraction of the backing position and receive the appropriate payout, but when you basically use your margin tokens to close part of the position and get the appropriate payout you no longer own a fraction of that position so we basically just have to burn the tokens because the tokens only represent ownership in the backing position, and if you close your ownership and that backing position then you should no longer have a margin token to represent that.
Laura Shin:
Okay. Something that I was curious about is you were talking about how these orders will be filled on these different decentralized exchanges and how you can set the parameters with your lender and stuff like that, but how well do you think margin tokens will work with low liquidity tokens?
Antonio Juliano:
Yeah, that’s a great point and we do think that they will work a lot better with tokens that have more liquidity, and that’s one of the main reasons why we’re starting with short and leveraged Ethereum, because the Ethereum/Dai market is by far the most liquid on decentralized exchanges, so that’s actually one of the main criteria we’re considering when deciding when to offer different assets for short and leveraged tokens is how much liquidity there is in the underlying spot market, so it’s kind of that combined with how much lending liquidity that we think there is for each asset, so if you want to short something, like we’ve been saying, you need to borrow it, so we actually need to go out and find lenders that are willing to lend all these different assets that people want to short, so kind of the two things that you need are the underlying spot liquidity and the underlying lending liquidity in order to trade a margin token on dYdX.
Laura Shin:
Yeah, I imagine that’s the case now, but I could sort of picture this going into some kind of longtail thing where really, really far in the future you’ll be able to do all kinds of different combinations, including of somewhat obscure tokens if this all works.
Antonio Juliano:
Yeah.
Laura Shin:
But yeah. One other thing that I was curious about is…so, is the person creating each margin token the trader him or herself or do you imagine there are going to be companies that crop up to feature different kinds of smart contract-based derivatives that you can buy, or how will that part work?
Antonio Juliano:
Yeah, so dYdX is a very general and unopinionated protocol so anybody can use it to take out any position or margin token that they want to, so you can use dYdX just to trade peer to peer if you want to. The first use case that we’re really focusing on is the margin tokens, and the standard margin tokens, which are the ones that we’re offering on expo are the ones that are created by dYdX, but anybody else could come along and create their own margin token on dYdX, and all that creating the margin tokens means is that you kind of specify the parameters for the position, so you specify things, like what are the tokens being traded, what’s the interest fee being paid to the lenders, what’s the maximum duration, other things like that, but dYdX is a very general protocol so we do anticipate that other people will use dYdX to build all kinds of different applications on top of our margin trading protocol and other protocols that we offer in the future, but we did think that it’s important to offer the first and easiest way to trade dYdX margin tokens, which is what we’re launching with Expo.
Laura Shin:
And actually, just before we continue on, you mentioned Expo before, but I don’t know what Expo is.
Antonio Juliano:
Yeah, so Expo is an end user application that we’re building on top of the dYdX protocol, and Expo will be the simplest way to buy and sell dYdX margin tokens. This is actually an interesting point and something that we’re doing differently than a lot of the other protocols that I’ve seen is that we are building a lot of these end-user facing applications and supporting services on top of our protocol, so what we’re doing is a lot more than just building smart contracts on Ethereum, but Expo is basically just a regular website similar to any other DApp that you might use, and basically it allows you to go and see what the different short or leveraged tokens that are available to trade are and gives you a really easy interface to buy and sell them, and kind of how it works is it plugs into other decentralized exchanges on the back end to get the buy/sell liquidity that you need to create…to buy or sell the short tokens.
Laura Shin:
Wow. This is really, really interesting. I know I keep saying that. I’ll try to stop. So, earlier you mentioned that as you’ve been going along you’ve realized that what’s easier, in a way, to comprehend is the actual derivatives themselves, but what’s more complicated is to understand how traders are trading them, so what are the differences, and what are you having to learn in that regard?
Antonio Juliano:
Yeah, so just the way that margin trades or options work aren’t really that complicated from a technical perspective. For margin trading it’s basically what we just went through, like you borrow something and then you sell it to somebody else and then you lock up collateral, and for options, which we haven’t really talked about but they’re fairly simple, basically you lock up collateral and give somebody else an option to, basically, buy that collateral at a certain price, so it’s not too difficult to make them from a technical perspective.
Really, the complicated thing is to understand how traders actually use them in trading and what trading strategies they use with options, or with margin trading. This is something we’ve been doing a lot of, is reaching out to customers to understand how they want to use, to start with margin trading and then in the future options or other types of derivatives on cryptocurrencies, and what strategies they want to employ with them, like interesting things. Margin trades are probably simplest type of more advanced financial product in my mind, but once you get to more sophisticated types of derivatives like options, or swaps, or similar then you can kind of do a lot of really interesting but more sophisticated things, like for example, with an option you can lock in your price for Ethereum, or really limit your downside, so say Ethereum was at 1300 dollars before and you were like wow, this is great. I still think Ethereum will go up, but I want to protect myself if the price of Ethereum goes down.
You could basically use a put option, which is basically an option to sell Ethereum at a certain price to lock in your price, so you could buy a put option with basically a strike price of, say, 1000 dollars when Ethereum is at 1300 dollars, and then if Ethereum goes up you don’t have to exercise that option because the price of Ethereum would’ve been higher, but if the price of Ethereum, say as it did, went down to 200 dollars, or whatever it is now, then you could’ve used that put option to instead sell your Ethereum at 1000 dollars, so that’s kind of one example of a really simple use case where these more advanced financial products can be used to really protect the downside and hedge against extreme volatility, so we think that they’re a really important part of the overall trading market and something that we haven’t really seen used very widely in cryptocurrency, but traditionally we see the derivatives markets being on the order of 10x having more…at like 10x more volume than the underlying spot markets and in cryptocurrency that isn’t the case at all yet, so we really think that derivatives are going to be really widely used in the future, especially given the underlying volatility of cryptocurrencies so that’s kind of dYdX is something we’re excited to build in the short term.
Laura Shin:
Before we move on do you want to just walk me through your version of an option briefly?
Antonio Juliano:
Yeah, absolutely. Our current options protocol, and to be clear this isn’t something that we’re launching right now but something that we’re planning to launch in the future, but our current options protocol is for what’s called a covered option and covered basically just means a full collateralized option. So, there are two different types of options, and I mentioned one of them before. There is a call option, which is an option to buy an asset at a certain price, so you could buy a call option on Ethereum and it would have some strike price, or a put option, which is an option to sell an asset at a certain price, so the way our options protocol works is that there are these people called writers, which basically are the people that create an offer to sell options, so if you want to, say, buy a call option, which is an option to buy Ethereum on Ethereum then you’d go to a writer, and when you buy an option basically you have to pay the writer a premium, which is money that you pay the writer upfront in order to basically own the option, so let’s go through a concrete example.
Say the price of Ethereum is at 200 dollars right now, you could basically buy a call option on Ethereum at, say, a strike price of 300 dollars, so if the price of Ethereum goes down or stays the same then that option will expire worthless because why would you ever buy Ethereum for 300 dollars when you could just buy it on the market price of 200 dollars, but say the price Ethereum goes up by a lot. Say it goes up to 1,000 dollars or something like that and now you still have this call option to buy Ethereum at 300 dollars then you would’ve made 700 dollars on that trade, so that’s an example of kind of how options can be used to increase speculation. Going into the specifics of how our options protocol works, it’s actually a good bit simpler than the margin trading protocol, especially for the covered options, and the way it works is that a writer will basically lock up Ethereum in a smart contract and will immediately get paid a premium, and what the owner of the option gets is a brand new ERC20 token very similar to how the margin token works. That basically represents the option, and what this ERC20 token gives you is the right to exercise the option at a specified date or strike price in the future.
So, if you, say, own a call option token on Ethereum and the strike price is 300 dollars whoever owns this new option ERC20 token will get the right to buy the Ethereum, which remember was locked up as collateral in the smart contract by the writer at, say, 300 Dai in the future, so that’s pretty much all there is to it, so it’s basically just the writers lock up collateral in a smart contract, the smart contract mints a new ERC20 token which represents the option, and whoever owns that option token can buy the collateral that’s locked up in the smart contract for a specified price.
Laura Shin:
That’s so fascinating because I imagine that in existing financial services everything that you just described takes a lot more paper and a lot more intermediaries, so I just find it really mind-blowing, frankly, when I learn about how this can be done with technology. One thing I wanted to ask about was that in a previous interview with me you said that you’re “working on things where you can take margin positions for other types of derivatives with not full collateralization,” and then you said that you’ll be putting all of that in a pool of collateral that will back many different positions, so I just was curious to know how would that work if there’s a market crash?
Antonio Juliano:
Yeah, absolutely, so kind of just taking a step back and I can describe a little bit what I was talking about there.
Laura Shin:
Yeah.
Antonio Juliano:
Our options in margin trading protocols are fully collateralized but this isn’t the way that most people trade derivatives in more advanced financial products in the future. They usually trade them with fractional backing. I mean, you can kind of see this by the sum of the total notional values of all the derivatives in the world somewhere in the range of a quadrillion dollars, which is clearly way more than all the money that exists in the world, and the reason for that is because all these people are trading derivates and advanced financial products with fractional backing, so that’s kind of what we’re trying to solve with this future protocol that will enable us to back options or margin trades with not full collateralization.
I want to caveat all of this with saying that this is very, very much in the research phase and we strongly believe that launching the fully collateralized margin trades and options are the most impactful thing we could do in the short term, but we are thinking about ways that we can kind of solve this problem of not full collateralization in the future, and what you were eluding to is one of the ways that we’re thinking about solving that is with a pool of capital, and basically the pool of capital will back a number of different derivatives and financial products together, and will kind of diversify the risk across a number of positions, so thinking about some easy things that you could do with this.
On dYdX with options. Say you have a put option and a call option for Ethereum, so that’s kind of one’s betting that the price will go up and one’s betting that the price will go down. With the version one options protocol you’d have to put up full collateralization on both sides but you don’t really need it because it’s either the price will go up or the price will go down, so in that example you’d kind of back the position with half the amount of collateral, and that’s kind of the general trend of how you can back a number of different positions with less capital by diversifying the risk across a number of positions, so that’s kind of the general theme of the protocol that we’re in the process of designing for the future, and something that we’re really excited about, and we think solves a really massive need for these not fully collateralized derivatives, but before you have not fully collateralized derivatives or a pool of collateral backing them you need to literally make a margin trade or an option, so that’s what we’re focused on in the short term is basically just making the easiest possible experience to do a margin trade, or an option without trusting your tokens to anybody else.
Laura Shin:
Yeah, I imagine also for that to work that you would need to mix in other assets classes besides crypto, like security tokens, because at least for the moment crypto markets…the crypto market tends to move in sync, like those days when you opened coin market cap, and everything is red, or everything is green so that actually leads me to my next question. I know that Harbor, which does security tokens, that their R-token standard is ERC20 compatible, so does that mean that you could use a security token built with Harbor on dYdX?
Antonio Juliano:
Yeah, absolutely. You can use any ERC20 token on dYdX, so that opens up a lot of really interesting possibilities, because once you start tokenizing real-world assets with Harbor or similar you can immediately tap into this whole ecosystem of decentralized financial applications, such as you could trade it with 0x, you can take loans out on it with Dharma. You could take margin trades or options out on it with dYdX, so it’s really interesting once you get all these open standards working together, and that’s something that we’re just starting to see the power of that I think, and I think dYdX is a really good example of different protocols working together because we are kind of built on top of different decentralized exchanges such as 0x, and it’s a pretty interesting example of how you can combine lending with 0z to basically make margin trading, so I think we’ll see a lot of really interesting use cases pop up really quickly once a lot of these protocols are actually live and start working together like that.
Laura Shin:
Just out of curiosity, obviously, this is a different token standard, but do you think eventually ERC721 tokens, which are crypto goods like crypto kitties, will work on dYdX?
Antonio Juliano:
It’s possible. It makes a little bit less sense with the way that we’re doing margin trading or options to start with because you need an underlying spot price for the asset, but some things that people are thinking of doing are basically fractionalizing ownership of an ERC721, like if it’s a really expensive crypto kitty or something like that then you can kind of have an ERC20 token, which represents ownership of that backing ERC721 token, and then you could kind of use that to say short a crypto kitty, or something like that, but the concept of just shorting the one crypto kitty doesn’t make as much sense because there’s not a spot price for it, but those are the types of things that I think we’ll be exploring more in the future.
Laura Shin:
You have two million dollars in seed investment from Andreessen Horowitz, Polychain Capital, Brian Armstrong, and others, but you don’t have a token, so how does dYdX plan to make money?
Antonio Juliano:
There are a couple of different ways we could capture value, and the biggest ways I think about it, or the biggest question for us as a company is how we capture value, and we could either do it at the protocol level or more at the application level, and what I mean by that, so at the protocol level we could actually do an actually useful token and I do think that there are a number of tokens that actually do have value, so that’s something that we might consider in the future. We don’t think there’s a strong need for a token in the margin trading or the options protocol, and my thesis on tokens is that they’re not something that you need to rush into and it’s worth thinking pretty deeply about.
First of all, whether you even need a token, and then if you do what the best way is to implement that token in your protocol, so that’s something where we really just wanted to come out with a protocol that works without injecting a useless token into it, first of all, and then kind of thinking about how we can add a token later, and the way we’re thinking about tokens right now is we may introduce one with some of the future protocols that we’re building and we briefly touched on the collateral pool protocol, which supports not full collateralization, so that’s something that might have a strong need for a token based on how we design the protocol.
The other way that people are thinking about capturing value at the protocol level is through just straight up charging a fee through their smart contracts. I’m not as much a fan of this but I could see a world in which this emerges as the standard for how protocols capture value. I think it works in applications. Like crypto kitties, for example, charges a 3 point something percent fee whenever you trade a crypto kitty, so I think it works for games and things like that, but I don’t really think it works as well for serious financial applications, like derivatives or lending because people really care about fees and middlemen sitting in the middle, but that could potentially be an option if that emerges as the de facto way that protocols are capturing value, so those are the ways we could potentially capture value at the protocol level.
On the application level we could capture value through actually just making money, and as I was mentioning this is something that we’re doing differently than a lot of other protocols in that we’re building end user facing applications on top of our own protocol, and a good example of that is Expo, so Expo won’t be charging fees to start because what we’re looking to do is just to get as many users of our protocol as possible, but if it does come out that most people are using Expo to interact with the dYdX protocol that could be a really good opportunity for us to charge fees like you would see on a centralized exchange at that level, so those are the three different ways we’re thinking about capturing value.
We have a lot of really excellent backers who are really, really long-term aligned in this space, and also raising equity has given us a lot of flexibility to basically just build out something that’s actually useful in the short term, and think about capturing value later, but we are thinking about it a lot, and we’re very much a for-profit business, not just an open protocol or anything like that that doesn’t make money, so that’s kind of how we’re thinking about it for the long term.
Laura Shin:
Do you imagine your target users will be…do you see Wall Street using this technology?
Antonio Juliano:
Eventually. Definitely not to start. So, to start with we’re thinking our target customers are smaller crypto hedge funds. There are a ton of these crypto hedge funds these days, or more sophisticated individual traders, so that’s kind of who we’re targeting with the launch of the margin trading protocol and Expo. Once we start getting into more sophisticated types of derivatives, like options, or anything else that we might add that’s kind of when you need more institutional type traders or really sophisticated individuals to be able to trade those, so I think we’ll kind of move up the spectrum from…like I think about customers as a spectrum from full retail customers to full institutions.
I think we’re starting somewhere in the middle, but we’ll shade more towards the institutional customers over time. A really interesting thing that having these tokenized positions offers you is that the positions, like the short tokens, for example, are just regular cryptocurrencies so they can be traded anywhere, even on centralized exchanges very much in the same way that Bitcoin is decentralized but you can trade Bitcoin on a centralized exchange if you want to, so I imagine that this will be one of the main ways that users interact with our protocol, basically just buying the dYdX short or leveraged, or in the future maybe options tokens on an exchange without ever even having to interact with a decentralized application at all, so that allows us to cater to a much more traditional audience.
There are a number of centralized exchanges, as I mentioned at the start, that are offering margin trading in certain derivatives products, which those are kind of the…like CVOE, for example, will be where likely most of the traditional financial people whoa re trading Bitcoin will trade derivatives for the short term, but I think in the long term there are a ton of advantages to using an open protocol like dYdX to do margin trading or derivatives, and kind of a number of those are that they’re fully global. You can trade with anybody anywhere in the world, which is kind of like something you see come up a lot for why you would use a decentralized application. It’s more secure, there are less middlemen sitting in the way.
You can share liquidity across a number of exchanges, which is actually a really important point, because for all of these exchanges that offer margin trading you can only borrow from people that are on the same platform, like for example if you’re margin trading on Poloniex you can only borrow from people who are on Poloniex, but with dYdX the protocol doesn’t care where you get the lending liquidity from, so you can share lending liquidity and spot liquidity across a number of different exchanges, and I think that we’ll really see that pay off in the long term, but that’s how we see the evolution of our customers, starting with people whoa re already trading and using decentralized apps right now and eventually getting to the more mainstream over time.
Laura Shin:
When you mentioned that you can trade with anyone anywhere in the world it just made me think, so obviously there’s an advantage of that in terms of liquidity but it just sounds like a regulatory nightmare, and I just wonder what are the regulatory issues and is it simply incumbent on the trader themselves to make sure that they’re complying with the laws of their jurisdiction or does that burden lay with dYdX, or how does that work out?
Antonio Juliano:
Yeah, so a lot of the time it will fall onto the centralized exchanges or relayers on which traders are trading. I think we see this a lot with different protocols. If they have any centralized components those are often times the most highly regulated, so I think if there’s any regulatory burden it will fall with those players. We have been thinking about this a lot. We’re working with one of the very best cryptocurrency law firms Perkins Coie, or are actively engaging with the CFTC, which is the regulatory body that regulates derivatives, and also other regulatory bodies to try to understand and educate them on how dYdX works and what the regulations surrounding it are.
You’re absolutely right that there’s a lot of regulatory uncertainty and this has been said over and over, but especially once you get into some of these more heavily regulated spaces such as financial derivatives. A lot of the laws were literally written for physical commodities, like you could sell an option on corn or something like that, and now the CFTC has classified Bitcoin or Ethereum also as commodities and are trying to apply the same regulations to it, so we’re definitely taking a very compliant approach to this. This is something I think I really get from Coinbase and we really saw work well there, in terms of we’re being built in the US. We’re not incorporated overseas or anything like that. We’re actively engaging with regulators. We’re really trying to provide support as much as we can to other relayers and exchanges building on top of the protocol, but it’s definitely something that I think the regulation will have to evolve with the technology over time, and that’s kind of how we think about it.
Laura Shin:
This has been an incredibly fascinating discussion and I don’t think I nearly failed as I was having flashbacks to last night, so thank you, so much. Where can people learn more about you and dYdX?
Antonio Juliano:
Yeah, so you can go to our website dydx.exchange or if you’d like to try our application head over to expotrading.com. We are hiring for a number of engineering, design, and recruiter positions, so if you’d like to work with us check out our job postings on our website.
Laura Shin:
Great. Well, thanks for coming on Unchained.
Antonio Juliano:
Thanks, so much, for having me.
Laura Shin:
Thanks, so much, for joining us today. To learn more about Antonio and dYdX check out the show notes inside your podcast episode. New Episodes of Unchained come out every Tuesday. If you haven’t already, rate, review, and subscribe on Apple podcasts. If you liked this episode share it with your friends on Facebook, Twitter, or LinkedIn, and if you’re not yet subscribed to my other podcast, Unconfirmed, I highly recommend you check it out and subscribe now.
Unchained is produced by me, Laura Shin, with help from Raelene Gullapalli, Fractal Recording, Jennie Josephson, Russel Zingaretti, and Daniel Nuss. Thanks for listening.
