The Yield Protocol: On-Chain Lending with Interest Rate Discovery

ETHCC 3 · March 2020 · 26:33

Daniel Robinson presents Yield Protocol’s approach to onchain lending and interest-rate discovery.

Speakers

Topics

From ETHCC 3, hosted by Ethereum France.

Transcript

hi everyone I'm a den Robinson I'm a research partner at paradigm which is crypto asset investment firm although I'm here representing my own views and not those of my employer I'm here to present this idea the yield protocol I wrote and published a research paper on this last October and basically here I'm gonna be going over the ideas in the paper and I think I can do this pretty simply because it's a very simple idea and once I've explained it or maybe even before I've explained it I think it should be pretty you know I think good the core concept is actually quite simple and nice but I like to go on wrote it so describe essentially how this works so why do we develop this let's push her to the point so the most

popular app on defy right now is secured lending and really I mean like most of the of b-five pulses top protocols most of these protocols are lending protocols which means secured lending you're borrowing one asset usually a stable coin and you with collateral that is a usually eath but you know even the ones on this page by the way that aren't I aren't lending protocols they fit into the same ecosystem so like a lot of what these other protocols like Yunus proper used for is in order to trade assets that have been bought or are being used as collateral on been borrowed or being used as collateral for protocols like

maker and compound so really they're also like almost all the top ten I guess like name is down there but I almost all of these are about defy on almost all the defy ups are really sort of ultimately driven by demand for leveraged trading so go a little more in-depth on what that means again I think most of you probably familiar with make your own compound and the emmys kind of systems but the basic idea is you're borrowing at your bar in an asset and it's secured by some acid some different acid on checking the reason you're doing this is in order to get maybe a levered position on on that asset so if you want to have more exposure to more--the than you actually then you know your net worth because you're crazy you can you can borrow dollars against you if I mentioned I on

maker or buy are buy borrowing CD on compound and you can and you can sell that trade that for more collateral and then leave her up I'm usually up to about like a three exposition so that's what that's what a lot of these verticals are they're wildly popular the ones that exist all those ones that I showed you there are floating rate data and so what that means is perpetual debt you borrow and you're paying an interest rate that's that floats over time so this is the maker stability fee and if people you know the people who sort of like lived through this it's like pretty wild I mean we went up to we went up above 20% for a little while in the middle of in the middle of 2019 sorry that 0.5% in my last january and and you know like the

it would be sort of hard to if you if you want if you just entered a position in april and you wanted to hold it for a while you might be a little non plus that like suddenly the interest rates interest rates have risen so far you might actually have to go close your position now this is great for keeping dye at a at a stable level especially with multi collateral dye which was added with the DSR which was added as part of multi collateral dye where you can actually get paid some of this interest for holding dye so that's very good for stabilizing the value of the currency it's also good because this is a perpetual instrument so the as interest rates change just to keep it stable you kind of have to you have to change the interest rate so it's a really powerful concept floating rate

debt but in when we're talking about lending when people want to borrow and lend often they want to do it at a fixed rate and they want to do for a fixed term you want to lock in a particular rate for a given amount of time now there have been some projects out there that have tried to do fixed-rate lending I don't I had liked examples on this slide but I took them off cuz I get kind of mean about these and you want to be mean to anybody in particular but the general problem that a lot of these that these have where you have fixed-rate lending protocols or interest rates are a lot of would exist in terms of interest rate swap protocols is that they're non fungible in order to take out a loan to in order to borrow dollars secured by eath for some amount of time

I have to find somebody who wants to lend for this wants to lend the same asset against the same collateral starting at the same time and ending at the same time and at the same interest rate and so all these protocols had a lot sort of liquidity issues like you know the the trying to actually you know when you when you go online there's not really a single-player mode like there isn't maker we just sort of borrow and you're often sort of trade it go off trade it on you know swap you're off to the races you have to actually find somebody who's who's online you know market making that particular maturity of debt and so that's and then once you've got it it's sort of hard to get out of that position because you have to go find someone say to sell this weird

you know interest particular interest rate particular expiration date loan to someone so it's less fungible and generally and I'm just like anti NFT I think it's a I think it's a it's a mockery of the concept of a token what's really great about a year C 20s we can do these we can interface where you can do all these cool things for them you can you can put them in Eunice Robin there and the fungible and you can you can market make them and that's just true of like any RC twenty it's very hard to market make NF T's you know you can build sort of custom market makers and that's what some of the other it just rate swap type projects have done but what we really nice would be to be able to take all the work that's gone into providing liquidity and price

discovery for ERC xx assets and translate that to loans so like stuff like you know swap protocols like 0x we're actually we're getting better at being able to find liquidity for year C 20 tokens and token pairs and so we want to try to transform the problem of matching lenders and borrowers into the problem of matching buyers and sellers of a fungible token so that's what y tokens are so a Y token it's a synthetic token like dyers or C die or cos you see here anyways and it represents fixed-term expiring secured debt so it's backed by youth just like any of these others there's liquidation I'm not gonna explain I

think too much both the details that are in common among all these lending protocols but very similar but what's different about it is that it is redeemable specifically it's redeemable for a certain amount of face value of the collateral after a particular expiration date and so this eliminates those features those amazing features of stability and perpetuity that you had in systems like maker and compound but what it gets you interest rates that are actually determined by the market so how does that work so suppose i if i want to create tomorrow to borrow asses i want to go leave or long on some long if i can put down collateral and then i'm mint and sell why USD say why USD expiring on on December 31st 2020 then I

just go sell that so I meant it sell it on you know swap and use that to buy more youth on the other side of the trade is somebody who wants to buy and hold this asset until maturity so this is somebody you know a lender what they do is they buy the asset typically at a discount which I'll get into and then at the end of the year they would deem it for $1 so what's great about this is that these white tokens are are fungible with each other so if I borrow an asset today but if I if I I'm sorry actually if I lend I give if I hold by some wide eye that expires at the end of the year and then I want to get rid of it like two months from now I can sell it to

anyone who wants to borrow and till that expiration date and whatever the prevailing prices sets the interest rate so it's not like my by this position that's tied to a particular interest rate so it's much easier to sort of market make this than it is kind of a non fungible token type loan or interest rate swap so the inspiration for why tokens is this concept from traditional finance called the zero coupon bond this is a many coupon bond but if you imagine tearing off all the coupons this is a bond for the for building the Panama Canal and what the coupons are is just like these payments over time so bonds are traditionally sold at like this supposed to be sold at face value the same value is the principal and in order to like trade of that value they have

coupons that get paid over time so you could paid interest you redeem interest over time and traditionally like what you do to actually collect interest on a bond is you could tear off one of these coupons and come in and sell it and redeem it for that interest payment but what some people figured out was if you tear all the coupons off of a bond and go sell them individually each of those coupons and the bond itself is a zero-coupon bond it's a it's an asset that is redeemable for a particular face value at some time in the future with new coupon payments and it's like it's really easy generally to calculate so the interest rate on that and it's this formula down here but what I want to get here is the the reason they did that and this is a fun

whimsical interlude here one great reason to use zero coupon bonds is that until until the IRS fixed this it was a great tax loophole so normally when you receive interest on a bond you have to pay taxes on the interest that your interesting come on it so each of these coupons each time you receive it you're paying you're paying some of that in taxes on a zero coupon bond you buy it at a discount and you never receive interest on it until eventually you sell it or redeem it and the year of its expertise so if it's a thirty year thick hero coupon bond you're binding at a steep discount but then like thirty years from now that's when you realize you realize essentially capital gain on it so it was a way of deferring taxes eventually the IRS you now there's a

there's a imputed interest on zero coupon bonds you can't do that anymore but is why I was embedded but I like it is that zero coupon bonds are a lot simpler I don't I don't have a background in finance actually did used to be a lawyer but I wasn't a very good one and I think the great thing about zero coupon bonds is it's just like it's like just sort of almost blazingly blazingly obvious sort of like what's how they should be priced which is basically just like if it's you know here they're treating in a discount and based on just a discount suppose it's expiring if is expiring a year from now and it's worth fifty cents and it's gonna be worth it a dollar a year from now what's the interest rate on it I'm gonna give a prize to anybody who gets

this it's worth fifty cents now to me worth a dollar a year from now what's the interest rate implied yes said 100 percent exactly that's easy math pretty this one's pretty similar you just divide the future value by the present value and put it to the power if you've got if you've got sort of you know I got three month period then that's the power of 1/4 doesn't matter minus one so the math is pretty simple and what's really cool is with these bonds you can you can figure out what the future expected path of interest rates are by looking at the market prices of these different white from different expiration dates so here I've sort of I've drawn a yield curve

for four imaginary yield curve wide I does not exist I would really like it to exist in part because I want to see what what the future path implied path of interest rates is I think right now when I when I published this it was like twenty percent it seemed pretty clear that I was gonna fall I don't know where interest rates are gonna go in but the market I imagine would have some guesses so it'd be cool to see just what people think will happen to that but this implies essentially that the that interest rates are expected to rise over time so this is you know like we view each of these bonds trades at a different price and then I just sort of did the math on this to say what what implied annualized interest rate is expected for like the time period from

now until until like say June 30th and so the idea here is this tells you this is just entirely based on what these like you know different series of white tokens are trading in in the market and you can use that to infer what the market thinks the interest rate is which is very cool no governance know sort of like formula there it's just sort of falls out of trading this as a liquid asset so one application have already talked about a little is the the just borrowing and lending lever trading if you want to lock in a loan but another is if you want to speculate interest rate so if you're there at like the top of the stability fee curve we're back on this on the maybe stability fee and you want to speculate you want to bet that that it's that these rates are

eventually going to fall you can you can do that by by selling by selling white tokens but long-dated y tokens and if you want to bet that I yeah by selling them exactly and if you want to do yeah so the white ovals essentially give you a way to like to bet on whether interest rates are going to go up or down based on trading on these like liquid assets so yes sir look I problems paper around this has more details including about one detail I didn't get into here is how do you actually settle one of these and there's a couple ways to go over them in paper one way is just based on an Oracle you could just say here's $1 worth of

'if this is what you get when you redeem one Y USD upon creation the other is if it's an on chain asset there's a few other things you can do including auction off the collateral to get to get some amount of the target amount of the asset to pay off the debt so more of that in the paper just googled yield protocol I would love it you know if you were canceling anything like this I know a couple people actually who are building ideas that are related to this and I'd be happy to put you in touch or hear your ideas or share or share some additional thoughts we've had since the paper was published imma give you to me to down a paradigm that XYZ or at Dan Robinson on Twitter all right thank you any questions no question sure Hayden Jordan here from synthetics

outside of speculating on future interest rate are there any other sort of implications of having something like this out in marker what other sort of use cases or I guess outcomes can we can we get from having you know this additional tool out there yeah well so it actually one thing you can do with it is use it as use it to make synthetic assets so if you have a long dated one of these you know it's it's kind of like a synthetic acid and that token it's not it's not perfect because it's again the price floats and so like if it's a long dated one it's likely to trade it may be a floating discount to what sort of like the target values there's not

necessarily a fantastic synthetic but it's a really like it's a really easy one it's a straightforward synthetic if you have something that's expiring just like this is just gonna be worth one Bitcoin a year from now say it's like the mechanics are quite simple so it's it's sort of a it's a it's a kind of like jerry-rigged version of a synthetic asset and then yeah interest rate borrowing and lending as I mentioned I think there's really cool things you can do in sort of the interest rate speculation idea with having Y tokens on like seed I for example or unch I and like that allows you to get levered positions and fixed and floating swaps and I have some more ideas and that haven't been published but you coming out it's a great question I think like yeah I don't know if you have any other

ideas on it have you decided with you better [Applause]