MakerDAO Know Your MIP, 10 Feb 2021

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Know Your MIP #05: Term Lending Module - February 10, 2021

Description

Know Your MIP #05: Term Lending Module

Intro: @juanjuan
Presentation: @niemerg and @equivrel

Agenda and Discussion:
https://forum.makerdao.com/t/know-your-mip-05-term-lending-module-wednesday-february-10th-18-00-utc/6297

Governance Forum:
https://forum.makerdao.com/

Disclaimer: These calls and the summaries are produced and hosted by MakerDAO community members. Content produced by the community are not the statements or views of the Maker Foundation.

A

Hi, everyone welcome to another know your mip session. In this case this new map number five and we're going to discuss a term lending module. The tlm map, which is map number 43.. uh The authors were what are alberto cuesta, canada, uh lovely and alan nimberg, nimr. Sorry for your your last name: okay, um but yeah before we. We start, I don't know alan and lavish. We want to give a quick uh update on on who you are for those that don't know you.

B

Sure uh so I'm uh alan emerick, I'm the founder of yield protocol, uh started it last year around this time uh we launched version one um in october um and it's been live on the mainnet uh ever since um and uh uh yeah. That's me.

A

I guess you don't need.

B

Love did you wanna? I think everybody knows who you are, but you wanna say something.

C

Sure uh yeah, I'm love, so I'm a maker contributor for some sometimes uh and I'm. I should also mention that I'm an advisor to yield so I've been helping beyond just go through this um of this specific myth- um and I you know, helped help some of the myths, especially on the technical implementation side.

A

Great so yeah jumping a bit into the the action. Can you guys tell us a bit more about this map and what.

B

Yeah sure so I mean the purpose of this. This mip is simple: it's it's to permit maker to land die at a fixed rate by by buying fixed rate, uh fixed term loans, um and it does this by buying a tokenized version of a loan. um We call these these these loans, uh these tokenized loans, uh phi die, that's short for fixed yield die and um what they permit. Users to do is is like maker to put in eth into our to the yield protocol.

B

uh Mint these fidei and then sell them for die, um locking in a fixed rate, loan and the properties of the phi die are, are pretty simple that you, um if you, if you have a fidei token, um you know and there's several different kinds. They each have. uh We call them series, they each have a maturity date and on that maturity date, you can redeem that token for for die one to one. So if I have 10 um december 2021 uh fidei come december, 31st 2020 uh one I can.

B

um I can just you know, reading those for for 10 die um and the way that you you, um you use these for for fixed rate loans. Is you know when you put in the eth and mint phy die? You can then sell that five dolla at a discount um and that discount represents an implied rate of interest because you you've you've received an amount of die, but you have some future obligation repaid that die um and um you can you can. Basically that implies an interest rate so um yeah it's it.

B

We think that this is a very elegant mechanism. This friday mechanism is a very elegant way to tokenize uh loans um in die at a fixed rate for fixed terms um and the term lending module. Then, let's, let's make her use this this. This primitive we've created for your protocol um to lend uh die loans um and what it does is pretty simple it it maker governance can set an interest rate, say five percent or whatever on on a particular series of these loans.

B

So if it's a december 20-21 uh loan, they can, uh you know, make or could set a five percent interest rate, and then users can sell uh these fidei tokens to make or directly um uh at that interest rate. So essentially that interest rate uh is a is the discount that maker is willing to pay to buy these loans. um The loans are, then, uh the the term lending module then just holds these loans to maturity, at which point they can be redeemed or sold.

B

A

Happy to answer particular questions. Yeah, I don't know if you were posting for questions, um so the one of the questions that came up is: how do you determine the the well the interest rate or the implicit interest rate.

B

Right, um do you mean how does one calculate it or how does how does would maker determine how to.

A

Well, more: how to sell.

B

A

I guess more: how maker determined we leave the calculation for the smart guys. You know I.

B

Mean yeah so so.

B

uh um You know, there's a there's a s, I think that's a governance question. Ultimately, um the rate that that maker would choose to pay is is, you know, is something that the governance would would decide, and so it could be changed like the stability fee has changed. um You know it's like okay. If the stability fee is four percent, you know you might decide. Okay, we're willing to buy um a three month. You know three month uh to the maturity.

B

If I die for maybe 4.25 or something like that, um and that and that interest rate then is just you know- plugged into the the yield to maturity uh formula when when a user is uh selling the fi die to determine the actual price that they get.

C

So, basically like for the for the kind of side, guy issues that are out today, which are like quarterly um quarterly uh quarterly maturities that are backed by yeast, a heat that has a um mineral cultural ratio of 150 percent, which, incidentally, is what our eat a like our biggest classical type uses.

C

We would view those as analogous to the sympathetic disability fee for ea, except with this additional uh quirk of having a term, so we would have to basically the most likely thing we would do is like. If we were doing this today, we would look at the first quarter.

C

uh The second quarter, maybe the third quarter um issues and we would uh decide basically on a on essentially a stability fee for um for those three um based on um things similar to the considerations that we use for a stability fee so like risk, but also like market factors like what is. What is a? um What is the competitive rate, and you know potentially.

D

C

uh Monetary policy factors right which, which obviously play a lot into our into our uh setting of stability and kind of incidentally, this would be the first time that we actually have the ability to influence um monetary policy.

C

Beyond the um ultra short-term uh stability rates, we can actually influence longer-term borrowing rates in a way, that's kind of analogous to um central banks doing that, and we do that and also in a very similar way, which is by buying things essentially on the open market at a certain price.

A

So stepping back a little bit potentially yield would take care of locking the heath and minting the phy die. Is that how it works, and then the fight that goes into maker.

B

That's right, it's uh you know, uh yield protocol uh is, you know, essentially a separate uh system that is, that is very much like maker and in fact we looked a lot to make her when we designed it um and it does a lot of the same things. It takes an eighth collateral, um you mint a new token um and that just like you could meant die um that and the amount you commit is based on the clatterization ratio. In fact, we use we don't just it's not just it's the same as a's collateralization ratio.

B

We actually actually use each a's, collateralization ratio directly um and makers. Oracles um to to you know, make sure that we're fully collateralized uh that a particular bonus, fully collateralized um and then yeah that the phytax token is then minted um and then can be sold. uh uh So it's locked in a die rate.

A

Okay and and the fiber is then locked into another vote within maker. Is that how how it works or.

B

Yeah so in in this case, with the term lending module, what would happen is um there would be a well, I mean they're called uh join contracts, but there would be a special purpose. Permission join contract um in the in the maker system that would receive this. uh This spy die when users sell it to the term lending module.

A

Okay and uh well maybe this is a a very simplistic question, but uh usually maker is more about generating fees, so why would maker want to like pay fees in this case? What's the advantage? Oh.

B

Oh, it doesn't pay. A fee um maker is earning that interest. So it's it's more like uh it's. It's it's buying the loan by by you know, um creating, I guess, new die or borrowing your diet using that loan um and then giving that to the seller. um But it's it's effectively.

B

What maker is doing is locking in a rate of interest rate return, so if maker is buying a loan at five percent interest, it's going to earn five percent interest during the life of that loan um on the principal um and those and then the what the term lending module does is, of course, when, when that income is realized at the end of the loan it put in the surplus buffer, just as um I think, stability fees are.

A

Can you maybe walk us through an example with numbers, even if it's a very rough math.

B

Well, um okay, so uh yeah sure, like let's say, let's say you were um it was january 1 and you were uh maker. Had you know you were, you were wanted to sell a one-year fidei, um the and maker said five percent interest rate right. um You could then effectively what you would do is.

B

You would sell this fidei token to make here at something like something a little bit less than you know, uh 0.95 die um and uh you know, uh and so you'd receive the 0.95 die maker would receive one phi die, um and then uh you know come one year later that that that that one fidei that maker purchased could be redeemed for for one die um and it would be so. um uh What would happen then?

B

Was you know maker would redeem it for, for one dire would get that one die, would uh essentially pay off the the 95 or 0.95 die uh principle um uh in the term lending module, and then the remaining you know five percent interest. You know which is like point: zero five die uh would be, would be placed into the surplus buffer.

A

Okay, make sense, so it's a, can you potentially redeem it within the? I guess not right, that's that's! The the difference is that you can only re-limit at the end of the year, so it's kind of like a zero coupon bond.

B

Exactly these are, these are exactly like: zero coupon bonds. um In theory, you could, you know, make or could choose. If, if you wanted, you know we could have a more sophisticated system where these bonds are sold at some time before maturity. um I I think that that would be something best left for the future. um I think that, like really the you know, the the simplest and easiest path forward, at least to begin with, is to just buy and hold the loans to maturity.

B

um Then you're, not then, then maker is not exposed to any interest rate risk. You know so um or or you know, has to manage that that um aspect of it just you know, locks in a fixed rate of income. You know when it buys the loan um and and receives that income at that maturity sounds.

A

C

Idea there is that we would basically one of the main parameters that we would be setting is obviously the debt ceiling for every issue, and we would be setting the debt ceiling somewhere, where we were actually hoping that it would be used, and we were actually comfortable carrying that um that position through to maturity at that at that debt ceiling.

C

um And then it has has various benefits to us compared to debt that we issue on our usual, like cdp, um open term basis, um which is that we have that um income locked in for the duration of the of the loan right. So because, obviously, the borrower has interest rate risk where their interest rates might go up on a cdp. But we also have interest rate risk in the sense that if we go into a market where we have to you know lower all of our interest rates.

C

Our uh revenue right over the protocol goes down. um To the extent that we have locked in loans at a fixed rate of interest. Then those will actually continue, and you can't um those can't be like repaid earlier. There's no way that.

D

C

Lose out on that interest run suite once we've locked it in so that's kind of a nice way of also thinking about it from a kind of diversifying the interest rate risk in maker, which is basically not something that we've really um at all been able to address before, which is why it's never even been brought up. As far as I know like we're, generally, not thinking that that is one of the risks of the of the dev portfolio.

A

Just curious who's, the like.

E

A

Mentioned the user who's, the the perfect user, the core use it that the core user that would potentially be interested in using this uh well this product or this solution.

C

I mean it's basically like quite similar to it's essentially like what they've built is like a system, that's really similar to cdp's, except with a fixed um with a fixed term, um so in general, like you're, coming with your 150 minimum 150 percent, these collateral um you're you're getting you're getting die at the end of it right because you're um you're you're, creating this fy die, which he then maker then buys from you. um So it's kind of a similar experience, except for the rate being guaranteed and the term being guaranteed.

C

So you also can't close this thing early. um So I.

D

C

You could imagine it being used by you know a lot of the same kinds of um users, but there's potentially a chance here of reaching a market that currently isn't accessing them because of rate risk. um So that's kind of interesting so like another, like a perspective that I thought was pretty interesting. That um seb uh wrote up in a post a few months ago that um kind of served to inspire this proposal.

C

um He was describing basically like kind of the entry level perspective on maker. Is that it's like cdp's right and then the more like abstract perspective maker, is that it's just like machine for for issuing debt rather than the cdps are just a um adjusted like a just like a delivery mechanism for that or like they're, just a um they're, just a front-end product.

C

Basically, that's like is a way of originating that debt, but in reality, like maker, has this power to issue debt that doesn't necessarily have to come from this sort of cvp and it can have different terms like you can have um different kind of collateral terms, which is something that we've been exploring on the real world asset side.

C

But it can also have fixed fixed rates and fixed maturities and stuff like that, um and it can also just you know, just buy assets like we've done with the psm right, so the psm is is implemented technically as a cvp, just like everything else, but really it's not really a cdp. It's like the thing. That's just buying selling stuff on the market. um So it's basically this this. um This proposal is kind of taking that perspective as well, which is that um maker could reach new markets by having a new debt origination um tool.

C

That is, you know, basically distinct from this kind of cdp, which has been our only um origination tool um to date, uh and I think that's like that's like a nice way of thinking about this from a sort of uh market um in business point of view, which is that we should be looking at other ways of originating that.

A

And potentially- and I don't know- maybe pretty much can collaborate with this one, but potentially the the rate would be similar to a bank which, in which fixed the fixed rate, is something slightly higher than the than the variable rate.

A

Is that how you guys thought about it?.

B

I I think you certainly could do that, um um and, and that makes sense um you know, setting it slightly higher. um You know a little bit of uh of a premium for for a longer term, but I could I could imagine situations where, where maker might actually choose to do the opposite, where you know you might have longer term loans at a lower rate um than than short-term rates. um uh You know not.

B

Unlike a central bank, you know inverting the curve um and uh you know, and that may be valuable in times where there's just really really high um die demand. um So you know you essentially um essentially charge people for for the liquidity premium of okay, borrowing on the very very short term, um while well, not uh well letting you know um long-term borrowers get a lower rate, um but yeah, I think, in the normal situation you would expect that there would probably be a premium, and that would be.

B

That would be a sensible thing to do.

E

Right is, is, um you might have asked a one.

A

Yeah, please go ahead by all means. That's I.

E

Hate no, of course, of course, um for those who don't know me, I'm kenton, I work for the foundation on the intercreations team. I have a question on um yield, yield curves, um so in so you detailed the example of short-term interest rates um being much higher than long-term interest rates and and the the tlm module um purchases these uh fidei and and effectively creates a price floor and therefore sealing on the borrowing costs for for a given series, I'm wondering if the if the opposite can work as well.

E

So if the dss tlm is not going to be selling maturing assets, how how do you see it um creating a? um How do you see it creating a floor on interest rates for for for long-term loans.

B

Well, I think it uh I'm not sure if I've quite got your question, so just give me a little rope here, but uh you know, I do think that you're right that when you set an interest rate, say with the the tlm um you've, I think you've set a ceiling right like if um you know it's, it's like five percent uh that that maker's willing to pay. Why would anyone you know? um Why would anyone take out a loan higher than that right?

B

You can always just sell it to maker, assuming the debt limit isn't full. um So I'm not sure uh uh what what you mean by a floor or a shirt.

E

To get the question wrong, yeah! No, no close close! It is in the example when stability fees are zero percent and we want- and we want to encourage, like a regular sloping, yield curve where long-term rates are going to be above some x percent.

E

um I I was wondering if that could also be done um with the current implementation, so.

B

Like a um so like, uh uh not a price ceiling, but a price floor by selling the fidec, yes, um the the existing implementation, I don't believe, has a way to to sell fidei um and um it's something that could be added when lev and I discussed this and and we were sort of putting this proposal together, um the the one thing I think we sort of thought about is okay. Would that even be useful? Is that is that a common thing that would come up?

B

um I'm not sure that it is, I mean, maybe maybe there's something I'm missing, but also. I think it does raise some some more complicated questions, because then you have to um if you're doing that you could, in theory like lose money on on doing this right, because um when interest rates uh change, uh you know the the fide assets could go down in value right and um if you're both buying and selling. If this module's, both buying and selling it could in theory, you could set rates in such a way.

B

You know having a bought at one point and then selling at another point where you've you've locked in a loss, um so the module would have to account for that or at least deal with that situation, um and it just raises the complexity to a considerable degree um that didn't really make sense. It didn't seem like we had a motivating reason to add when creating this, so the tlm as it's currently engineered just buys and holds um so there's, never there's, never a a loss that it can incur.

B

At least um uh you know absent some sort of catastrophical failure of yield or something like that.

E

Right right, right, yeah, no, it was just. um um I realized that that that the interest rates can be set only one way, so I I just thought that I would. I would ask that, um and it would ultimately like again. I think I think long term. We should have that ability um so that it gives us more flexibility as we as we interact with the with the yield curve. Eventually.

E

B

uh Yeah, I think that makes a lot of sense. um I suspect you know, I think would be it would probably work better after seeing this, this whole market mature, a lot more and and having a lot more understanding of what fixed rate, uh lending and d5 really looks like right.

C

Right, there's there's basically like two ways to come to mind for like make you get more complex in the way that you're you're describing and firstly, I just want to say that I agree with alan that, like basically, the priority here should be to make some sort of mvp that lets make her access longer term lending, without necessarily having to become an expert at trading um fixed income products, which is essentially what we would have to do.

C

If we're willing to buy and sell these, because now we suddenly have great risk and we all need to really think very hard about where we can set these rates to divide, and so on um I mean the other thing to say: is that just like um you know maker is a natural buyer of this because being a buyer, if this makes you a lender maker is naturally a lender maker is naturally not a borrower, which is what you're doing essentially when you're selling this now.

C

Obviously, if you, if you bought it before so you were lender and then you sold it back and you're sort of unlending, um but essentially it makes sense that it goes only in one direction, because maker is naturally a lender right, you could. The other way you can make this more complex is that you can obviously build a thing that will actually allow maker to borrow using using yield.

C

So baker can actually mint dye to post this collateral to um to borrow fy die from yield, which is one of the ways one of the things that yields can do and then try to sell that fy die on the open market for like usdc or something so like that would be like a bank issuing a bond or something so like because banks, all banks, are obviously lenders but they're. Also borrowers sometimes right. So so you could.

C

You could imagine makers, starting to borrow for some reason like if they thought that the market conditions for that were were attractive, um but it's it's obviously sort of it's going to some next level of uh of complexity, but it's definitely possible to do to do stuff like that right right.

E

Right I see that I see that that one use case for having higher rates long term in a market where, where rates are suppressed due to the demand of of borrowing um is, is basically for fixed term lenders, fixed term lenders who or or fixed rate lenders that are just more conservative in general.

E

It wouldn't benefit borrowers, of course, but but lenders for sure.

E

So um another question I had actually, if or actually I'll, throw it over to preemos.

E

If, if you have a few questions as well.

F

Yeah I'll go back after you thanks, um so I just have a question about: uh what's the current liquidity of fy diet, um I've been trying to because, as I imagine, if you want to borrow fixed term, you need to find the lander and there needs to be some kind of market for fyi die, um so people actually buy it um because the zero bond uh uh zero coupon bond, I imagine like the biggest benefits for you guys as well- is that you're gonna get a nice liquid market if this gets implemented.

F

That was just curious. The current state, how liquid is it if, if someone wants to borrow what's the other side of the market? Currently, how does it look like.

B

Yeah um uh yeah, I definitely could answer that so for, like um you know, yield is pretty new, um so we don't have like the tvl of the the largest protocols and we're still trying to make the case for this. This whole kind of borrowing, um but for the march 2021 um liquidity pool that we have it has about 1.5 million dollars or 1.5 million, die worth of liquidity uh at the moment. um To give you some sort of uh to give you a sense of scale of it. So it's not it's not huge.

B

I mean, especially when you've got maker out lending a 1.5 billion die. um It's it's quite small, um uh but um you know we've, uh you know it's not zero, and um you know. I think that uh we're we're learning all the time, how that's growing and how that can grow right.

F

And, as I understand that's, it's not really an issue right because we would be holding it until mature maturity, um so the liquidity. um The the other question I had was uh you know. You said that fy die is basically mimicking e tablet, um but in some sense the you know these loans, that maker would would be buying their originated your protocol, so it does have a risk profile of your protocol.

F

Although it's mimic cta, so I'm just curious, you know, do you have any kinds of because we would be buying a pool of loans right um as much as I understand, and uh you know, what's your debt ceiling, how do you print losses? uh What kind of auction system do you use? This is all relevant because even if you set an interest rate higher long term, you know there's also this risk aspect and it's not. It's not makers risk. In this sense, it's yield protocols risk that we need to assess right.

B

That's right, that's right! Absolutely um there. It certainly is the case that the the spy die exposed would expose maker to to uh to yield um we've when, when building yield protocol, we took we keyed off of a lot of the things that maker has done and we've tried to, um I think, be as safe with maker as as maker's own heath. A collateral is, I mean, and so so that's directly influenced.

B

You know um the fact that um you know we're using the same oracle we're using the same I mean effectively read the value directly out of maker, uh the same uh collateralization ratio um on the auction in terms of liquidations, which I think is a primary area of risk. um I think we're using a similar system that would that's close to to the uh maker liquidations 2.0, which is something I'm not very familiar with.

B

So I'm told that, but I'm not uh I'm not 100 sure about that, but but generally the way that it works is uh you know, there's some there's some debt that our system would have. um It's essentially the debt right and we have some collateral um and we hold an auction where we we raise the amount of collateral we're willing to to. um I guess this uh sell to to potential uh auction buyers uh to to buy off that debt, um and we have uh that auction runs for about an hour.

B

So um uh you know it's it's a liquidation system, um much like many others in in the eco system,.

F

Thank you thanks and, and if it comes to losses, um do you have any buffer to cover them.

B

So, if, if the system uh so the way that yield protocol version, one is set up, is it's it's uh in entirely uh it's deployed and and there's uh uh no no further control by us. So like we created the system, we set it set the parameters and we put it out into the into the or onto the main net. um So and there's no tokens.

B

So we don't have a a um you know, a buffer, an economic buffer in that case, um uh and we decided to do that because we knew we weren't going to do it. We weren't going to do a token for this, uh at least for now, and we don't have current plans to do one. So we wanted to create something that uh we that we could try to minimize risk. So we created like sort of a self-contained governance.

B

Less system that you know is built on top of maker and actually all the collateral that's placed in our system is placed in maker um in a cdp, and you know we just wanted to make it. You know, as I think, safe and secure as we could and sort of self-contained, um but there is, but that that sort of comes with the the cost of there's there's not really a an economic buffer uh that that same mkr provides to the maker system.

F

Okay, thanks for my questions,.

E

Yeah, I actually have a couple questions on the liquidations system, um so I've I've been involved with um the the development of liquidations 2.0 in maker, and we chose a a a touch auction, a standard touch auction where the price starts high and and then it's dropped. I'm not as I'm not like an auction expert, and so I'm wondering if, uh if, if your team had a rationale for choosing a reverse touch auction over a dutch auction or yes,.

E

So it's the same type of mechanism, but just it's done with collateral going up rather than price, going down.

B

C

E

I mean it's just a different question.

C

Collateral going up is price going down, so in that sense, it's really the same as as how is the liquidation, 2.0 mechanism, where um we have a fixed amount of debt that we want to um uh to to to cover and we're reducing the amount of cholesterol that you need to get to? um Okay, uh to cover.

F

E

C

The whole reverse uh terminology is just confusing, because the reverse part refers to the fact that we're changing the price by um changing the amount that you get as proceeds rather than the amount as you pay, which is kind of usually how you bid for things. uh But this, but we are. We also have that in liquidations 1.0 right, but that's an english auction in the sense that the price is increasing, um so it reverse doesn't mean that it's like price increasing in english right. So it's right.

D

E

Basically, the.

C

E

Right right, right and then, and then, uh where does like relative to the to the osm price or the price oracle at which the auction starts. Where does the price of the collateral end at the top? At the top of the hour.

B

Right so uh it starts at offering 50 of the collateral for the entire debt and it ends at um 100 of the collateral for the entire debt um and what uh so so and after that, it's just held open. So that's it's sort of different. It doesn't go to it. It doesn't go to like a price of zero effectively where uh you know your your somebody could buy the entirety of the collateral by paying just a small amount of the debt, um and we chose to do that because.

B

The the alternative where we you know the price are the the amount of. uh I guess you know debt, that's paid off. You know we go down over time, and um I mean after after at that point you start to you're, locking in a loss right um and the the trade-off we wanted to make, because there's no backstop because there's no, um uh you know say mkr token or whatever that would fill in that loss. um The we felt like it would be.

B

What might be a better risk is that if we just hold the auction open at that price, if these prices rise, there comes a point at which um you know somebody will buy that that loan. You know some, maybe at some future point. Of course, that assumes that that actually happens that youth prices, um you know, continue to generally go up, um but we felt like that was the safe.

B

That was a safe um sort of bet relative to like okay, we lock in a loss and there's no way to fill in that hole like that. Just hold it that just exists forever. um So.

B

Yeah I mean there's there's you know, pros and cons of that approach, but oh yeah yeah.

E

It makes sense it makes sense why um the the the auction timeline or the or the price decreasing function is held within an hour and and the the rate of t increases. I think it's if you start at half and then go to um probably the the full liquidation price is. That is isn't that, like a floor of 33 or something like, if eath falls less than 33 within the hour, then um it would be, it would be profitable to to purchase it or like or like is like. Like.

E

Are you familiar of the threshold at which.

C

I think that's the drop, I think that's right, because it's 150 to 100 percent collateral right.

E

Right right, so so in the event like, of course, it like this is a an extreme like almost black swan-like market condition, but in the event that eath drops over 33 in the hour, then, is that the is that the failure mode.

B

Right I mean so in that failure mode. Then what happens is the auction just doesn't doesn't complete and it's held open at that at that last price, which is, you could buy the entirety of the collateral for paying off the entirety of the debt, um and so if, if you do have a black swan event like that, where it's down 33 and then you know a few days later, it goes back up to that price.

B

You know, you know the system may um work its way back out, um uh and so I guess in some sense this is sort of an optimistic auction. In that um it's it it. It works better under a situation where you know over longer periods of time. Eth appreciates right.

E

Right right right, which is currently with the market uh that we're in now, which is uh very fortunate, of course, um okay got it got it. um So it looks like the.

C

It's actually working, it's actually worth pointing out something which is maybe a bit weird but like if we, if we somehow did end up in this worst case scenario of having um you know essentially, maker dao holding all of these fy die bags that got um that had these auctions that basically got stuck at um at 100 and um can't complete and there's basically no mechanism internal to fy die for, like actually realizing that loss.

C

um If we wanted to close those out um and basically get something for it, it is, it is actually possible to um basically for maker to make like a huge bid on those losing auctions, and then um you basically use that to pay itself back with the other wide eye for like basically whatever the fraction of value that you were able to um recover. So this would mean that you don't have to write this down. You don't have to write them down to zero, um just because the auctions are stuck.

C

You can kind of use this as a trick to um to get whatever is left in there. um You kind of have a problem with like holdouts, who are also holding fy die, so maybe it would be better to like basically try to sell this somehow, but this is all in the domain of like having to do extraordinary governance actions to like do some sort of manual operation with this stuff. But um but that is like what would happen in the worst case scenario,.

E

Yeah yeah yeah.

C

B

Again because, oh sorry, I'm just.

E

Going to say it and again.

B

Because uh because you know all the collateral in yield is held in maker, currently um and and perpetually um the uh you know there, there are a lot of different options that maker can have to deal with sort of really weird scenarios um and of course, actually maker can liquidate the entire yield system. If there was some sort of maybe systematic risk right right.

E

So it looks like the um I was. I was looking at the at the smart contracts earlier and it looks like the maturing gem. Abstract is pretty general and could possibly be adapted to other fixed term lending facilities. um So how do you obviously yield is one of them? Notional is another, then I think there's a there's. Another protocol called mainframe.

E

How? How do you think maker governance should assess these different zero coupon lending facilities, as it considers different maturing assets to purchase.

B

I feel, like that's, really a sort of maker question on how it sort of looks at this whole ecosystem.

B

um I think what we've tried to achieve with with this uh the tlm is, um you know it's opinionated about how about how these loans are weakly opinionated, perhaps about how these loans are structured and that they're erc 20 assets, um but it's not it's not yield dependent. So nothing about this, this, the tlm locks uh or like um integrates with with yield on any sort of fundamental level. All it does is just take in these assets and hold them. um It should be possible.

B

I think, in a very straightforward way, to extend that um if you know not not directly, maybe it's like notionals case, although actually my understanding is that they could, they could decide to to add the relevant feature with relatively easy to to integrate with the tlm. um If not not directly, then you know, at least by analogy it would be a straightforward um extension um uh to to you know, incorporate or to do the same thing.

B

I guess with other protocols um and we're seeing I mean one of the things we tried to do with with uh the structure of phi die is actually create, something that that um other protocols could use. I mean it's a relatively light um extension of erc20, that's you know, set up for for fixed rate uh assets.

B

um You know that act like zero, coupon bonds. um We've, you know, sort of abstracted it into like a very basic interface. I mean the tlm. Doesn't even use need to use all that, but there are other protocols now that are that are starting to um add zero coupon bonds with the same kind of interface, for example, um empty set dollar further version, two is adopting um a token. That's that's essentially identical to what fidei is, but for their for their yes, empty set dollar.

B

um So you know it's definitely the case that that you know this is this is weakly opinionated, but it's not. We meant to be nonpartisan, it's not supposed to be um like locking maker. In any way towards towards yield.

E

Great to here, thanks.

C

I mean on a technical level, it's basically very similar to the psm and just that it's got like a price and it buys stuff at that price. um So that's.

D

The most generic.

C

Version of lending you can, you can imagine right, it's just buying stuff. Basically, so it's it's really. It's really not like a tight integration with um with yield, which is, uh which is, I think, pretty important, because we like basically in the vision of like feb's uh idea about originating um dead in new ways like basically, we should use this as an opportunity to like kick off a new, um a new direction and and- um and you know, finding assets that we can um originate in this way by just buying them on the open market.

C

A

I have a question regarding the: why not to call it maturity, but are you guys regarding the time or the horizon? Is it like around one year always or what's the plan behind that.

B

Yeah absolutely so currently, uh the yield protocol has four outstanding maturities. um Basically every three months from now, until the end of this year december, 20 december 31st, 2021. um so um yeah I mean that's sort of the the general term of things we've. You know we've spaced about three months apart, um we've found that most interest has been. You know, generally on the the three month, the closest maturity, which I think is pretty common in in all kinds of bond markets that the the more recent maturities tend to have a lot of liquidity.

B

um Yeah. So uh does that answer your question.

A

Certainly, I was actually curious about the the liquidity um and and yeah. I was checking the apr as well on the well it's for the borrowers right, but uh they appear on your on your website um and yeah. The one from from march is actually lower than the one from fortune, so you have like an interesting curve going on there.

A

I don't know. I'm not expert in economics.

B

So yeah, I think the curve is a little out of whack right now, just because um gas is so high. It changes how people are looking at interest rates at the moment.

G

Yes, there is not much liquidity and gas is high; it doesn't make much chance to reduce it. But if you we look at uh national, as I guess, a bit more liquidity or mature. uh The rates for july are six seven percent.

G

So that's the kind of things you can lock in which is higher than the is a volt. So it's more fees from us and the interesting part is that it will decrease exposure to usdc, because the current psn currently is buying usdc and we might want to buy something else that is not related to to usdc and fydi is related mainly to maker and another protocol, which is decentralized at all, because the smart contract is immitable.

G

So there is trough between those two. So only difference is that if there is a drop in the demand of dye, we can sell the usdc. But if there is a drop of the demand of dye, we cannot sell the fy day, at least in this current implementation.

B

That's correct: the current implementation doesn't allow you to sell the phy diet, although I think that that that's a let's see in the case where demand for die drops what that means. The uh the price would be be dropping what you would be raising interest rates um right so yeah in that environment in that environment. um Well,.

C

It may be a lot, it's it's basically, it's it might, it might not literally be a loss, but it's it's in the direction of being a loss right if you, if you were bar you're lending out at low rates and then rates went up. So that's exactly the scenario that alan was talking about before, where, if you want to get really serious about this, you have to think about the the rate exposure in the in the portfolio.

B

That's right but holding it to maturity. You you know you, don't you don't realize that loss you're you're, just you have a mark to market loss but you're, not realizing it.

A

Pretty much asked on the chat: are you also planning on offering non-eth based fy die? How will you call it.

B

Yeah, so um uh definitely um in the future, um future versions of the yo protocol would would. I would expect to have.

D

That alan uh in following up on that, because I do wanna, I wanna first of all yeah- this is a great module. In my opinion, I think it's gonna be really beneficial for uh maker, but uh focusing on yield protocol right.

D

Do you see a future where maybe yo protocol is also going to focus on some of the new players? Maybe like compound cache or maybe even usdc and there'll, be a fy cache, fy fyusdc.

B

I I think that, uh yes, in the sense that I don't know about about compound cash, in particular, excuse me, but I mean I think, that you know we. We started this uh yield with with dyed, because you know it's the in my mind. It was the most interesting uh decentralized protocol um and had I think um we felt like it was doing it here could make an impact, um and we hope that it will um and we think we're moving that direction.

B

But um I guess the the the vision is to extend uh fixed rates to to other sorts of borrowable assets in the future, um which ones in particular. uh You know that would be uh decided when we're releasing a a future version.

D

And I think for me uh for me, um seeing someone like linen, which is an app for like the everyday, jane and joe to have access to a fixed rate via fy die. It's. I think it would be just such a win for them um to introduce and write it to allow someone like the average joe just to get a six percent for a six sorry six month, uh zero coupon, the equivalent of that would be it's a win for me. So yeah yeah.

B

I think that that's I think that that's a that's it. You know, I mean that's what excites me about about yield as well. I mean that's why I started this. Is I think that, like the um fixed rates just make a lot of sense in a lot of contexts- and um you know it's really cool that you can go yield farm and make eye popping apys, but you know um they require a lot of management, and then you know two weeks later, you've got nothing most people can't do that.

B

I mean they would be happy to lock in six percent on on. You know a savings account, um and um you know we think that, like that, we hope that that yield can help. People do that and that's our goal.

H

Hey hey guys, um do you mind if I I say a few words here? This is teddy from notional.

A

Go ahead teddy.

H

Thanks thanks um yeah, so I I just just wanted to chime in really quickly. I I think it's a um you know that this is like a super interesting proposal, um and you know I I feel the same way as alan in in that uh uh you know we're obviously very interested in kind of um expanding the ability for maker to offer fixed rates to their users. Just because you know sort of fixed rates are very important in um eventually taking d5 mainstream right, and uh so you know.

H

I think that although this this proposal was sort of tailored for the yield protocol, which I think is understandable, given that you guys wrote it, um you know I'm supportive of this and and uh just wanted to sort of make clear that um the way notion was written. It's it's. uh Our fcash tokens are erc 1155 based, but that is a flexible standard and we would have no problem issuing erc20 wrappers. That would be totally erc20 compatible and could be eligible for purchase by the tlm.

H

So you know, I think this is a really cool thing. I'd love to see it happen.

A

Thank you teddy thanks.

C

Daddy, I mean it sounds like if you guys make those rappers, then potentially um notional,.

C

Will just work out of the box with with the tlm as proposed? If I understand correctly, which sounds like really good news.

H

uh Yeah, I think that's correct.

C

It would be great to uh have you guys, provide feedback on the implementation of this map, just so that, um if there's like some perspective that we're missing to make that integration um possible like we hope that it's really generic. um But I think it would be good to uh hear your guys thoughts.

A

No pressure teddy.

H

um Yeah, so I I sort of uh so I put you know, I commented a little bit on the sort of um on the forum. uh What I think is is important here is that uh and alan responded, and I think he sort of answered what I was saying, um but I think what's important here is that uh the tlm be able to lend directly to users uh in the way that um maker currently does.

H

uh Just because, like you, don't want to introduce an unnecessary dependency on a liquidity pool um when you know maker could just lend directly to borrowers.

H

um So I think that, like that's, that's one thing that I think is would be pretty good to have in here um and uh but you know just from some from sort of like uh the technical perspective lev um now I am not the the the cto. My my partner, jeff wrote all right.

C

Just to be clear, I didn't mean to put you on the spot here when I said it would be nice to have feedback. I didn't mean necessarily right this moment. I meant that as we're working on this myth and getting it to getting its main, that you know if your team could uh sorry interrupted, but just let you finish, I wasn't saying that we needed to hear all the feedback right now. Yeah.

H

No no worries man, I I um yeah we'll we'll take a look. I I'm very confident you know we can. We can do whatever. We need to do here that, like again, the erc 1155 is a very flexible standard and there's no problem. uh um You know putting some sort of erc20 wrapper on top, but we'll take a look and make sure.

A

Sounds good um yeah we're almost at the top of the hour. I don't know if there are any any parting comments or questions before we.

E

Yeah, you should trust them. I I was trying to to to find an information on uh the the yield series, such as like volume and or tdl, I'm wondering if, if you guys have like an analytics dashboard, that's that's currently public.

B

uh We we do not. um I can get you some numbers offline if you'd like yeah, please and and happy to follow up on that.

A

Thanks um so alan, when, where can people find more about yield and more about yourself.

B

Yeah, absolutely I mean uh our website is yale.is. Where you can view our app, you can see our links to our documentation. You can join our discard cord and uh see our twitter, um so those are those that's a great place to go yield. That is.

A

Thank you love. I don't know if you want people following you.

C

Sorry what I missed, what was the question.

A

No, if you wanted to like give out your twitter handle or something for people to to follow your thoughts or write to you. But.

C

Oh um well yeah. Please please write to me uh on uh maker chat or the forum or something like that.

A

Sounds good teddy.

H

Oh um yeah, you know notion about finance, is the website uh you can follow me on twitter. If you really want to, uh I don't I, but it's probably just better to follow, follow notional, uh notional, finance um but yeah. You know, if also we have um uh so. I miss you guys uh asking about the sort of the dune analytics um sort of like just get some information on the system. uh We have. We have a dashboard up there um and uh you know I can.

H

I can send that information over to you guys kind of after this meeting, yeah.

A

Upgrade yeah you can post it in the forum, maybe so everyone can see it and it's there for posterity all right thanks everyone for coming. It was really nice.

B

Thank you for having us.

E