MakerDAO Weekly Governance and Risk, 6 Feb 2020

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Governance and Risk Meeting: Ep. 72

Description

Please join us and help shape the future of the MakerDAO.

Governance Segment
Richard Brown: Introduction
LongForWisdom: Governance at a Glance

Technical Segment
Will Barnes: GSM Dark Fix

Risk Segment
Vishesh: State of the Pegs

Links

- [Video/Voice](https://zoom.us/j/697074715)
- [Dial-in](https://zoom.us/u/acRbIMDvK)
- [Calendar](https://calendar.google.com/calendar/embed?src=makerdao.com_3efhm2ghipksegl009ktniomdk@group.calendar.google.com&ctz=America/Los_Angeles)

A

There we go hello, everyone welcome to nothing, but it's going anyways, alright, so.

B

Welcome to the.

A

February 6 edition of the scientific governance and risk meeting. My name is Richard Brown I'm, the head of Community Development, McCord I, will be talking to risk today, we'll be talking too long for wisdom, and we are going to be getting a presentation because long for wisdom has been bugging me for ages for the opportunity to do even more in these calls, and so occasionally a lot in to do it, isn't that right, Vista, actually, no I sprung on it. This morning we have a new, a new iteration of our governance.

A

Dashboard that's been released recently and it's a valuable resource. I think that it could be an even more valuable resource, and so what I would love is this. This group here is a great cross-section of our primary stakeholders for this tool, and so I would love to have this group watch diligently. Take notes, spend the rest of their day.

A

Thinking deeply about this, the government's dashboard for maker done and give us some suggestions about features that you like to see in or ways that it can be improved or things you like and don't like I'm, not going to do much speech in Phi and today, I think that I'll save any kind of insights or any conversation starters for the Q&A aspect.

A

At the end of the call you know last last, two or three calls we've been using that time fairly effectively and so I want to leave as much content and time as possible for those random questions sessions. Alright. So let's get this thing going, yeah their wisdom blowing up my spot once more in the chat, all right, yeah, I'm, gonna head it off too long and he's going to potentially go inside glance, a very special threat possible and that's we're. Definitely just tour of the governance. Dashboard. Take away.

C

You got any preference on ordering I'll.

A

Leave it up to you, has.

C

That great okay, all right, I, guess I'll start with the dashboard.

C

C

C

Right, can everybody see this thing.

A

Yeah, that's good.

C

Cool okay, um I, don't know how many people have seen this before. But this is the governance. Another success world that we from the pretty far I've been working on and Monica lasts a few months, so we just released a new version a couple of days ago, so this last version was mostly like bug, fixes and stuff for all the new features. There are a couple of cool cool, new things: I guess I'll just go through this journalist I, don't know how many people who have used this regularly.

C

This is the home page. We've got a bunch of metrics that tell you something each tell you something about the mecha d'elegance system, I guess, so they feel like a measure of how much makers staked in the chief. How much is voting, how much non-voting you can also see that supply of maker, which is dips down nicely in the last couple of things.

C

So a couple things tracking the net number of floating addresses God's got the Gini coefficient of the voting maker, which is how we'll just repeated the voting makers across addresses rather than normal maker.

C

It's just a list of how many votes they've been, which kind of shows how much like sort of what the governor's workload has been I guess in terms of how quickly this increases.

C

So in the most recent version, we've added so help text. So all these metrics have like a bunch of paragraphs with caveats and things. It's also got links to the sub graph sub graph queries which power them so you for people interested they can and it's got some more. We've got some problems at loading. I'm, not sure thing moved, that's it and so can.

A

You dig into responsiveness a bit.

B

I think, like the main issue, I have with a lot of these tools is just the loading times on some of the widgets yeah.

C

So that's been a common complaint and it's something I'm aware of as well is I'm gonna keep pushing profile to work on and yeah, there's much to say other than like blockchain is a little bit complicated. I guess I mean it's working progress so hopefully all the improvements.

D

C

So responsiveness is a measure of how quickly maker responds to votes going like executive votes going, live, that's what this is showing and so there's no text which explains all the stuff, but stick over you. So this valley for 0 to 1 days means that on average 25,000 make a response to so votes, I guess on an executive in the first day after it goes live right. This is an average. Obviously, so it doesn't measure like how popular it was. So it's won.

C

Some popular is also gonna do less well than one that is popular, but it gets kind of like a baseline idea of a how much like how much maker we could rely on if there wasn't like a mergency which required response within a certain amount of time.

C

It also gets night like if an executive is popular, you would popular. You would expect you see roughly this amount of maker both on it.

C

So you can sort of see how Jerry like there's more reacts in the first like week, and then it sort of tails off heavily.

D

Does this kind of mean that polls are really yes, so you sector for days.

C

D

C

This is specifically this one, that is, loading, is Motor Show polls and sometimes is.

B

A

A

C

That Richard show someone using after.

B

Is that the responsiveness and in response to the most recent executive, or is it kind of like an average okay? So now it's.

C

An average of like every executive- currently you guys so in the future, I'd like to get it so we can track it over time as well as we can so increases and decreases, but yeah for the time for the, because it's quick, a more complicated metric, we loaded with just a nice.

C

So yes, this is also polls the scaling of obviously wrong here.

C

So so you can see. There's some other thing with polls you expect, like maybe nine thousand, the first date. Five thousand sync. They.

C

Sometimes I mean it's not that's nothing for straight like we could have polls which are longer. We saw the three vehicles like the I agree. I should cap to like the amount of things that's there for.

C

As for splitting to our, we could do that I suppose I'm, not sure how much more useful that would be and I don't know.

B

Yeah also like it's kind of like mixed data right, because if you have, if this is kind of like all of the polls, it's weird cuz, there are some three-day polls or some seven-day polls, so you don't really get accurate sense of what the total responsiveness would be. I'm like longer polls instead of this is still cool.

C

Yeah I mean it doesn't really measure like totally responsive, isn't measures like respond, so this from the start of the poll right.

B

C

So, like yeah, so whether the poll length is three days or seven days, the amount of people that responded within one day should be the same for both of those right.

C

Yes, we talked about.

B

Polls here or executives so.

C

This graph here shows polls, and this girl, fair, shows executives and.

C

No, so it yes, so they're sort of the two systems for running look quite different. So it's it's a little different, difficult to combine them and I'm not sure ever useful. So.

C

Course, if there's not any other questions about those only home to voters, so this is essentially like a leaderboard for how many times and addresses weighted you can see. You can like expand this to two more people.

B

These are both these are. The voted proxy addresses right. Yes,.

C

I believe so I think currently accounts I'm, actually not sure I need to I need to double-check, with with notifier what it's doing, but yeah I, don't think it currently tracks everything correctly. It's I think it just cracks. The proxy addresses and we're like might use, might get the cold wall open all the hot. Well, that's all of something else, whatever seems to be coming a fair approximation. Currently, though, so so you can copy the scan, etc.

C

So there is I'm, not sure it's broke. She dresses appears to be not the proxy, because this is not a contract. So you to double check.

C

Cool, you can also click on these and look at voting history for assigned address. You can do this on most like address like links, you can click on and they will show you the history link to polls there and then once this is loaded. This will shop. I have with first things.

C

This is not currently got a column for date, which is something on ad soon, because it's not great without that yeah.

C

Cool so a couple of other sort of vaguely useful things. We got a top executives, so this is it just a list of all the executives and how much maker they have on them currently listed by an automaker.

C

We go to list the uncursed executives, which is executives that never got cast and just for kind of people's interest in just so. Hopefully, people will see if one of them is gathering a lot of maker, for whatever reason.

C

So these are being an interesting, so distribution by executive shows say the states, so these are like executives, which happened on the dates on each date. On the bottom, see can see the most recent executive has the hats. Just in black, you can see the previous ones too has 57 problem maker on, and some of the older ones have 30 Cohen, 36, K and there's some very old ones still have a little bit.

C

So executive time to past is also big interesting, so this shows the average time that an executive it takes to pass from one date to put on chain, and so we can see that we passed most within one day. I mean there's other tails off fairly nice looking curve which I'm expecting to cut it.

C

Look into it and for one day.

C

Straight Falls we've got some the reason. The res recent polls, the minute just happened.

C

So we've also got lists of executive votes, which you can filter and sort so say this is one so one thing which mention is we don't it's going to initially getting the metadata for order? Executive switches, sorry these names and descriptions? So currently we only get the title and description for stuff. That's for executives that are displayed on the the foundation voting dashboard, hopefully I'll change at some point soon. Now this is all we got here. You can see you can solve it.

C

It's you can saw way once it's still open, which ones that never cast.

C

You can search my name as well, not that you've pretty hopes in this case. Could you any other neighbor one got a thing for polls and it takes a while to want to load.

C

Have you have the same sort of things in here? You can sort by you search by name. So if you want to see savings rates, you can find all the savings rate adjustments and see when it happens when they started when they finished all the wedding options were.

C

Yes, it looks a bit weird cuz. We changed to the spread footing, you can see so when the ones open you can like open, open balls.

C

Can see they're waiting prior policy like the total makeup, but situation this is percentage and they've decimals the number for lights a bit wrong, which mean I, hope you fix, but.

C

That's that's not she everything that they details for this particular poll or executive.

B

So this view I know for governance polls. It auto populates pretty well, because all of that data on chain is there any plan to have this kind of view for the executive votes as well, because I know that the description is usually missing and that all that has to be added manually right right now. So.

C

You repeat that you cut out a little bit in the middle there. For me, sorry yeah.

B

So yeah, so is there a similar page for the executive votes, because I know that, yes, the executive votes aren't like Auto populated with this info right? Am I right about yeah.

C

Yeah- that's that's great! So there's there's been no kind of sorry about getting the the data for this. For the for the woods like. No, no, the station lives on chain it lifts in like hubs for polls, there's just a there's, a hash render on your chain in the poll which links back to the köppen yeah me not super one it'll, hopefully get fixed, so no yeah, so certainly poll. You can see some sort of details to start a this latest time, remaining opened, etc. Winning the option.

B

Conceptually is it: is it possible that, in the future, people can probably even vote directly from this site? If you just plugged in yeah I mean there's nothing right.

C

B

C

It's just a just a contract, all right, it would needs like work. Three integration and things were doing like like mathematic integration, which isn't currently half but yeah and there's nothing like stopping as a possibility.

C

So some cool things about this is you can see how boat counts have changed over time, both in terms of count and in terms of maker, and you can see like when people are voting and like when whales waiting and if people change their weights, you can see what they went like the swap happens, which is pretty cool. It's.

A

One of the things I like the best about the graph is occasionally you can see a capitulation which is an interesting people. Yes,.

A

C

Scale forces got any broken 13,000, so so voters, so voters, slash, addresses per option and like address count prop ssin, obviously not simple, safe metrics I think it was like I mean at this stage. I. Don't think me to worry too much about civil issues, especially since there's no monetary like incentive to do that. But we didn't to kind of be careful that that someone isn't start like civil boating on stuff just to scroll. These metrics.

C

As well and again, you kind of got a top voters leaderboard for this, which shows who has contributed the most to the switch, the waste maker, and so you can see this voted for a half percent. It's got 66 percent of the maker on this well and then tails off.

C

So, as I said, we have a similar thing for executive votes.

C

So again we got the details and the description and that only works for, as I said, the current, the current vote, all the weights that appear in the foundation dashboards and again we have the same sort of thing. We have made us make a staked overtime and we have addressed like poachers overtime, and we also have a breakdown of approvals by address size, so we can see again, not so we'll save, but we can see and how many like small country justice, with a smaller maker votives like larger and larger.

C

Once again, you have top supporters thing new server versus current hats, which shows how much make of this proposal has on it compared to the next highest. If it's winning or the hata hits that one and.

A

The proposed by address size is fascinating to write because that's an indicator of the emergence of parties.

C

I think it's been interesting cuz. It just shows like like waiting for each like address noise and, again, not suppose safe they're like it's interesting, because you know a lot of people been sort of saying like okay. Well, only lot is like any large holders right right, but we can see. That's not the case from this like there are like a lot of people have like fairly light amounts, but they're also cooked. Let's walk around sweating, so.

A

You can see the volume, but you can also see the distribution right as long as that is fairly flat. Then you can presume there's no line doesn't say the same.

C

No, it doesn't mean currently at that, and we could look into doing that if we thought it'd be useful, like I, don't know kind of donors commit too much like resource to doing this stuff. Because again it's you kind of you can like break it by just making coatings like dresses, more dresses, so I, don't I, think they're interesting with anything. The things we should like rely on the foot, the data I guess so you can see for further things. We have some details, so we don't have the description you can see. It's through.

C

This example like there are a lot of smaller, just as weighted a nice, large ones and I think that's pretty much the extent of the features at this point there's any questions or comments feel free to chime in. If it wants to see anything more detail, let me explain anything. What do you tell good I.

D

Mean I'm interested in seeing potentially how we could use this to make governance more efficient like what kind of historical data voting data should we go through like what kind of data analysis we can do to maybe help shape future governance, yeah.

C

For sure, like there's, a good question like a lot of the metrics, have been kind of be focused when showing I guess the current state of the system and like like versus the historical state, but we haven't necessarily got anything that sort of tracks, I, guess, efficiency, I, don't know how we would write.

A

C

Measure measure that sort of thing or like some sort of health yeah I mean so so the state. So this thing maker things kind of reflects health right, because the more 'we khuskhus and the kind this like distribution as well, like, I guess, some idea of health and because you've got you can tell like where the makers in the system and like currently overlay for executives. So it's like 62 k is the hat which is less safe than like them all being deflected.

C

Okay right, so you can kind of see that sort of thing, but but yeah in terms of like tracking, what's I guess, increases voting participation or, like that's a little difficult, total votes to some extent like tracks efficiency, we can see like how many votes and polls like we're, creating like or like the rate of change on this on this curve is kind of interesting. You can see you.

B

Could technically also see if, like somebody owns multiple voting, addresses right depending on like the timing of their votes cost you can expect me may be quite.

C

Difficult to do I think yeah good, but yeah do.

E

You have any type of data that might show how long somebody sat on a specific executive vote. I say that, because I go back to the subject matter from I, don't know how many months ago, about the you know, if somebody puts X amount of maker on a given executive and then painfully, unfortunately gets hit by a bus, and it stays there right now how we could track what stuff might be theoretically lost.

C

Yeah, you could say we could only do something which is like so so one of the one of interest I wanted so I've been working on, which is kind of related to that is like a measure of how active maker is, which is maybe related. We had a bit of trouble doing this like sort of making this metric, because there's a lot of cutting edge cases and weird stuff that means take into account.

C

The idea of that metric would be. You know we can sort of there's a number for sort of how much maker has been active like in the last day and in the last week and in the last month in the last three months. So that way we could see like how much expect to react to something that happening like if we know like this much maker is usually active in a day, then maybe we can rely on that in an emergency, and but it also you're like over longer periods.

C

It would show like you know how much maker only tends to vote like maybe once a year or like once, every six months, once we three months, which is interesting.

B

And who the real protocol politicians are gonna end up being the most active voters, the ones who are always participating. You can also see.

C

B

C

I mean so that's one of the things I think the top voters thing will be quite useful for up in the future, because if we well once we have delegation because I'm gonna get at some point like you know, potential and delegates would be able to sort of point at this and saying look. I voted in like like I'm, like totally to what I'm voted more than like anybody else.

C

So you know you can rely on me for consistency or whatever, but yeah I. Think that could be useful.

C

So you're just give me the chance, but I, say I think that's yeah yeah! We could that it's great slangs if an address holding maker has other non-voting activity. We could know they're not here by bus, which is true. We could potentially look at that. We could look alike at the lasts operation from an address that holds maker, maybe and then sort of get an average or distribution feel like how long ago, certain make it or like how.

A

C

Or touched proof.

A

C

I could yeah, you could maybe like a heist, is like a sort of like a high school thing of like a high school like let's go river like time since last October, sorting, larger Justin Lodge felt like large make holders. Maybe we could potentially do so. If you have ideas.

A

Well, that was my my hope actually is just get this thing in front of this group, because this group is has no shortage of really cool ideas. There's lots of I'm willing to bet that's a silence, but there's lots of people here that are have experience with mixing and matching data and data, visualization and crunching numbers.

A

Think about this please this is a valuable tool, is going to become more valuable, we're at the second iteration, where it's beginning to show its potential there's a new phase of this dashboard being proposed by the team that built it, and so if people have ideas about improvements or augmentations that they'd like to see now it's time to, let us know, and we can help you either to pass that along or people can get involved as well. It's a community project.

C

It would be really great to gets like your like if anyone's like a data scientist, so it works and, like date to visit and visualization like. Oh, we really they'll be really interesting, potentially useful to get to get someone like that involved, because at the moment, kind of just like winging it and saying like we should display the data this way, but like I'm, not really an expert. So if someone has a board like better ideas about that good to get.

A

The whole world of things to explore about steelmaker and old hats.

A

B

The most interesting like data visualizations and things are gonna, occur when in the community, we're gonna be like hey. How do we figure this out and somebody's gonna be like? Oh, we just need just need to do this. Yeah then it'll exist.

C

Cool all right comments, I think it'll, stop showings done. Does.

E

Anyone know what the status.

A

Is on NPR tools and if Michael McCain's, it's gonna, make a feature. I, don't have I think that people, someone has been reaching out to him in the last month or two, so they might have more up-to-date information but I'm reasonably sure he's waiting for a robust API that he can. He can bounce off and still having to crunch all that data from scratch for MCD, but yeah. It's still kind of up in there I think that there's room for the community to sort of come up with its own solutions. I think.

C

Austin thanks.

A

Which actually yeah, let me take the opportunity to do a little bit of a plug here. So there's no there's a significant overlap with what happens in the governance ecosystem. What happens in the general community development ecosystem we make er down and the models are the same. The domain is different, but we have frameworks. That's incentivize people in the community to build things. We have grants, programs to support them and we have team members available to help administer and project manage.

A

There's very sophisticated mechanisms in you know community development, and so, if anybody in this call or anybody that's called know somebody it's not in this call who's capable of coming up with things like the maker tools for MCD version, 2020 reach out to us and with the proposal- and we can start talking with us these things, that's how this governance dashboard came along. Proto fire came up with an idea, pitched it to us.

A

We went through the evaluation process if you find it I'm stake holder and then easiest why it just magically appeared on the internet right, lawful reason, it's crazy. You see us by Wow and so there's there's mechanisms of working. So if people have ideas, they know how to improve the ecosystem. They know how to fill a hole in our tooling. Let us know because we can make it happen or help you to make it happen.

A

Alright, I'm long did you want to continue to monopolize this meeting this you're insane yeah, I, guess.

C

Yeah I guess I: do you go to Ron's I glanced quickly? I think the feel like discussion has been a little bit lighter this week. There's not a whole lot to go through so in terms of actual discussions.

C

So it's a I mentioned this last week and I'll mention again, there's been a thread about whether we should reduce to the bat suppose you fee, which just had some nice back and forth between Soros and this.

C

This new guy Andy I, was discussing wrist things and like was like fairly premature or like I, feel like that was it's quite an encouraging, encouraging sign in terms of like having people like offering up ideas and things about risk, because, like what the things we've been semi worried about, it's like getting more risk teams, almost more risk personnel, so I think that's been good.

B

I actually have a question about lowering the bat stability feet, but if you guys don't mind so and it's in general, just Lauren stability fees in general, so is it ever possible to have a stability feet lower than the current DSR? My understanding is that no because you'd be able to just issue die and put it in DSR for that spread, so the DSR is like. Does it make sense to say that it's the natural bottom limit for any stability fee we offer on the protocol I? Think.

D

We've talked about this a couple of times in the past right about subsidizing certain collateral types for various reasons um having a lot of people to issue die and lock it up in the DSR if it's in small quantities isn't necessarily the worst thing in the world if it brings other benefits, greater adoption, more diversification or whatever there's kind of like a whole host of benefits, see I. Think it's technically possible.

D

I mean we, we sort of had that situation, sort of right when the size stability fee was lower than the die DSR yeah people were printing people were minting sigh migrating over and locking it up in you. So thankfully, so.

B

Is the idea that, like let's say the bat stability fee set down to five percent, while the DSR is still sitting at, like eight point, seven, five percent, that although people would be able to to mint, died against their bat and earn that spread and basically generate themselves, incur alpha? Some portion of that borrowing would actually still result in and whatever leveraged behavior and like died being out in the world not in DSR. Is that kind of mentality and just also the benefit of having additional collateral? That's not eath.

B

You know more diversified collateral in the portfolio I.

D

I mean if, if we wanted to do that with that, that could be an option, but I don't see why I didn't see why that would happen for this collateral type, but for some other generic one yeah.

C

Like I think, the main benefit is diversity. Versification right, like we're, essentially subsidizing foot diversification because, like the the dye that's created and then like locked up, is like it's like kind of purely bad, like that's, it's like parasitic in the way that it doesn't increase the supply, because the dye this generation will go back into the DSR, so there's actually gained us anything like or doesn't gain us like supply. It just gains us like diversification and and yes, they were paying for that. I think your Nikes yeah.

B

E

And if you lock it up in the DSR you're, exactly what you just commented, you're all you're embracing is just the pure risk. There's no website. It's only downside.

E

It cost it cost the maker community capital in the sense that it's being subsidized and then it that or any collateral that's upside-down, like that, were to have some unexpected enormous drop into a liquidation. There's no there's no benefit.

C

Yes, like risk wise, like yeah a benefit like I, said the I think the only time would be like if we really want a diversification and something that wasn't too correlated.

C

Anyway, O'hara with governor sitting.

C

Cool, so the any other actual discussion was no plug was the was the dashboard but I think I've discussed that enough already so I will skip ahead.

C

Centers of signal requests, I will say so, I create it's. It's a no request. A signal requests regarding the government security module asking whether we want to have another executive vote on it's um like sooner rather than later.

C

So, if you have it our thoughts on that, then for me to go there and express them and, like my only personal own personal opinion on that, was that's kind of one more input before really deciding and because I don't really feel qualified to judge like the relative risk between like technical risk and theoretical crypto like like theoretical government's attack risk, so I think I think we also said that he was going to come and that that the technical team were working on something and they were going to present that at some point soon.

C

So I guess we shall look forward to that.

C

Nice was that so yeah I mean the the sort of tax during a city shut down times. Std shut down in general is still a thing, I kind of moved over to position where I think we should kind of just wait and see mostly for this, but yeah I know. Other people potentially have like more stressed about yeah.

D

Actually I was gonna, I was gonna, say the exact same thing: cuz I'm, the size supply is still is still going down, like I think I had expected it to kind of stop around well a little bit higher. It kind of still keeps going and I feel like until we kind of see where this thing stops and see what kind of what the resulting distribution of CDP holders of unpaid stability fees, size supply. All that I just think it's um it's a it's a bit premature.

D

I mean I feel like we can definitely pick this up over the next couple weeks or in a month or something and that's suggesting any like long delays, but just for now, I just feel like migration occurring every week. I think we just had a big big d5 platform, just migrated like one or two million sigh earlier this week right so I'm not I'm, not really sure what else is kind of still in the works. I think what what is interesting, though, is that the side peg is drifting a bit right. There.

C

Was one like I mean, maybe maybe more but like I, saw one like big trade on yourself which dropped down to like 0.9. So it's it's getting to the point where it's it's getting a bit of a rough to trade in so I. Think well,.

D

It's interesting is that that also kind of provides more incentive for um for CDP holders to uh to repay right, except incan.

F

Well, I mean unless they're waiting for emergency shutdown, in which case they get their fees forgiven like due to peg drift.

D

So it's like almost like those the people who are selling psy right now are are basically helping end up bearing the cost of the forgiven fees. If the peg board adrift loan like if the peg were to trust like ninety cents, then it would maybe somebody what would uh cover their CDP I mean. That's obviously not gonna happen for size, but kind of interesting I mean.

B

Who's, the buyer, if somebody.

D

Seller, why are people selling psy is my question especially.

B

When there is died in the migration contract, it's like yeah.

C

So so this was on you yourself, and someone would have to like I didn't sort of look carefully to see what happened with like, because it's like the that kind of really equilibrium with incentives to get back will appearing with with you know self, like. Presumably someone bought like a bunch and then someone sort of inch which would.

E

Let me I have a question about just the general decision tree. What we're trying to pursue with it goes back to the core question of what is the objective, and you know the ultimate decision tree right now is that if the risk profile stays the same and if we keep lowering the debt ceiling on side is something that's either very, very close to the amount outstanding or even slightly below or slightly above, just let's say, materially close to what is out there. The risk is hedged. We are already exposed to that risk. Today.

E

The system is that risk doesn't go away by doing an emergency shutdown. It just shuts it off and does whatever whatever the stability fees get in effect waived. The question I, guess I, when I've gone through this cycle of thinking about this and talking with some other folks was what is the event?

E

What material advantage do we gain as a community by waiting to do it tomorrow versus waiting a month and then waiting a year if we constant if the risk doesn't change, but the only thing that does change over time is that there is a chance of materially recouping some of the stability fees they get paid back in the form of maker. Why would we rush to shut it down, and you know the peg drifts around more or less no one's really using it anyhow, because it's more or less unusable under.

D

The questions come up a couple times and I think for me, two things one is uh just general brand make her brand right. It's not good if the even if it's the kind of the old version and we've been public about migrations up, if the peg were to drift big time would probably not be a good good look, then.

D

Secondly, I think if there's right, if there's this, even if there's an expectation that emergency shutdown is saying not going to happen or not going to happen anytime soon, then it would, uh it would hurt the peg further right. It would actually do destabilize I even further correct.

E

I agree with all of that and I think this is where it turns into what is our objective and it's a communications issue, I think when we try and find the world's perfect solution where the peg will be perfectly constant, and this will dissolve itself down to zero at some point. There's emergency shutdown, no question.

E

The point, however, is between now and then either a have to accept having some sigh peg volatility that is not friendly or we have to decide to shut it down, or we just decide then, in effect to say we're, gonna market it that this thing is old and legacy and Windows, 95 or XP has viruses but use it. If you still want to use it does.

B

That make sense to say 12 to market, especially the fact that the migration contract still exists and that there's still sufficient liquidity for majority of SIU's errs and sighs CDP positions to migrate so like, although yeah the peg is becoming less and less reliable, we're doing our best still in total good faith. As governors of the protocol to you know to look after the end user.

B

So long as there is my great you know, liquidity in the migration contract.

C

You yeah, that's, actually something I've been wondering as well is like we've kind of we're kind of assuming that, like the city holders are like bad actors to some extent, but it might just be the like. There's never been enough liquidity for some of them to move like, especially like 3088. It's like anything, has ever been like a lump in liquidity in the contract. It's big enough for him to like.

E

But there's Parkman: there are tools out there that allow you to do partial transfers.

C

Yeah yeah, that's true again: I, don't.

B

Think it's our job to assume the the motives or even guess the motives of every individual CDP holder like for perception risk and for like public optics. We have to do. We have to just provide the route right if it, if 3088 wants to take the risk of sitting through emergency shutdown and really testing m'kay US government mm care. Governance is resolved on, like figuring out the tax and the stability fee issue.

B

That's on him like that can't be on us what's on us, is that we provide for the you know the greater population of CDP users and and SCI users the ability to get out and see use, MCD right.

C

Yeah I think I think the students some extent I guess. My kind of my kind of point is that it's not like CERN at 38 is that's his intention. Right like he could be waiting for a global shutdown just because that gives him the best chance of closing without like having to pay like a ridiculous amount of foresight. Cuz like I come really like I forgot what how much, how much he has at this point.

C

But if, just if he trusted like sauce that mostly on the market, then the pig is like he can't sauce that mostly on the market like right now, and it's not a migration contractor right now. What was like a total amount.

B

Like a million in the migration contract right now, I think.

C

It's one point.

C

Yeah, it's like a there's more another spin there's been for awhile because of the next stage it wasn't made. Someone asked the city of you go see to behold. Us have to be paying attention.

C

Yeah, just a just a thought right, that's I wished so see, there's this I stuff. So the only other thing I was gonna say again is so we had the single handling executive bundling with technical changes which should move on shown at some points. Rich. Do you have any update sister timeline on that everything on Jane or any problems or anything.

A

I need to regroup with you to figure out what the.

C

C

I think that's everything yeah.

C

D

Right um should we move on to uh this stuff, yeah?

D

Okay, so this shesh is having some technical difficulties so we'll start off with promotion as uh what he's got and then we'll try to do kind of a brief, maybe just a brief run through dye stats. There's something yes, his battery is a battery died yeah, pretty much want to take it away. Yep.

G

Okay, so the I supply went up about five million in the last week. Actually, net minting wasn't the fight. Only two million Saudis mentioned that there was a lot of site being migrated to who MCD actually last week alone about three million and we're still seeing some amount of that immigration contract.

G

The largest part came from next song two days ago.

G

I think they migrated about one point, one point: four million and out of those three million that migrated about one point, seven or eight currently was actually used by CDP's to migrate to MCD as well to larger CDP's migrated last week with this Mont pretty much one point similar amount is it's what's missing here up to three million and they paid about a hundred thousand of fees alone in the last week, I'm going to show that a bit later, so the total supply is still increasing with pretty much similar rate.

G

Although, as I said last week, only two million made I printed, which is not as much as you would expect. Considering the there isn't it price rally, I, usually when we saw really high increases of either the dye meaning was increasing heavily as well.

G

The thinker note, so this is MCD by the way. This is I believe currently the only website where you can pretty much see what aggregate, what would the aggregate repayments or drawers or collateral being added in MCD such as, what's able to to get it at maker tools, so this is from if I explore the team behind if I sailor to can basically checked I, checked how much they were generated in Amsterdam so on I. Believe, that's! That's!

G

Currently, the only platform somebody was asking before so just wanted to mention I notice, a large withdrawal of eater named Sydney four days ago, and this actually pushed the amount of hitter locked in MCD below two million that we saw in the past I'm guessing this is happening because collateralization ratios are really high. No that's because hitter pricing, resistant, CD burners are just withdrawing the unnecessary capital from collateral, which was usually the case. What actually surprised me how high the collateral ratio is in a CD? Usually it was lower than in MCD.

G

Now it's it's almost at four hundred thirty percent and I'm guessing this may be due to due to general inactivity, we're seeing right in inner city. There's about half of depth is a short like two weeks ago. I think those CDs were not really interacting during the migration and on average, the interact the last time they interact. It was more than six months ago.

G

So it's not only this thing that these cities are not migrating they're, not even you know withdrawing some of the collateral which you, which usually see that ceiling is now it's 30 million profess today. So this should leave us with about 7 million buffer, actually 8 million. If you, if this size burnt that's in migration contracts, so 8 million is pretty I mean it's quite a low tech.

D

Sure, sorry I think that's a really interesting point actually about the collateral in ratio. Cuz. Wouldn't that be one of the one of the few indicators we have for if the CDP owners are active or kind of in touch with. What's going on right, because that behavior shouldn't really differ much from MCD to SCD.

D

So if we're trying to like gauge active CPI elders right because you can't go by the dye supply because they may not be adding or or wiping debt, obviously because that's risky cuz of kind of the peg and the the potential tax and stuff, but the collateralization ratio should theoretically stay in line.

G

Yeah I mean exactly this was my point actually, but you could think about it. We've seen even higher fertilization ratio, so the ratio itself is, you know the is related to leverage. So what is that yeah? They're they're, not active, obviously, but the other could be that you know they're, just okay with clique and heated in collateral and not really removing it and converting it into something stable. So this might also tell you that they are a really bullish right, but I'm, leaning towards the first right. It's more related to general inactivity, we're seeing yeah.

D

I mean because if they were, if they I mean, if they just wanted to leave it in deliberately, then so, with the MCD see degree holders right, why would there be a difference? True, okay, sorry, yeah I just thought it was interesting. Yeah.

G

So yeah there's eight medium, but for now or available simon thing actually sound as a set, but it if the remaining side get burnt it if CDP's migrate. This should be, of course enough. If there's a sudden salinity, dre shortage and new sign needs to be minted other hand. We had debates that we shouldn't, or we wouldn't want to have too high buffer, because we want we don't want to see unnecessary side printing, so I basically checked.

G

Was there any side printing in the last two weeks, basically I mean I, have my own tables but I'm, just showing you this Thunder ccp's data from a curse. Can you can check for the ones who were throwing sigh, how much they actually, how much demented it and the numbers are actually really low. I haven't actually noticed any new SD CD piece trying to mean significant amount of sigh at most.

G

You could see some CDP still using Cassidy and I, don't know meaning 10,000 sigh or filters inside, but I don't see any any any side minting activity to be honest, any higher one.

G

The next think is time. Dsr, it's all time. Fine! Fifty percent of all days locked in DSR, probably because the side siren savings rate is, of course at Oldham high as well, but also what I notice is that on compound, the die landing market actually surpassed us DC, landok markets. So, just just in the last month, the dye supply at compound went from 20 to 30 million and I'm. Also guessing this ratio will increase in favor of the I versus us DC.

G

The last time we sold Islanders dominating us this, yet compound was I think more than five months ago, even before the migration. What why I believe the dye will dominate is because, of course, the supply rate is now much higher for die, because the unutilized part goes to DSR it compound this, of course, isn't the case.

G

Sigh still casts some amount of liquidity. There will not sure why this is not reducing at all the last few weeks. It just became a bit sticky and I. Don't know. Perhaps just users are unaware or some size, speculators, I, don't know just deposited they're. Turning some extra fees next thing appear here is the CI supply. You can see how it's how it rosette compound, so this is dye landing at compound and the last month or two three thirty million I want to focus a bit on remaining psyche holders.

G

So here we have.

G

So we have grouped addresses by amount of side the court and you conceded basically three categories, so the ones holding from ten thousand side two hundred thousand sigh and then the ones from hundred thousand two million and million plus they hold I- think twenty twenty two million, or even more of the supply, which is pretty much all of the supply. That's out there I think ninety percent, and if you want to see how many addresses are represented by by this wallets, it's almost three hundred so two hundred it is something.

G

What was also interesting was that you know knowing that only one hundred hundred psych holders they represent 80% of the supply. I thought the number will be much bigger, but it's fine if we compare it with top CDP holders at a CD, pretty much the same, the same percentage, so you could say basically that we are waiting for. One hundred side raises to match with hundred top size CD piece and if course they would migrate all at once. You would be left with only about 10 to 15% of available supply, currently, which is few million.

G

So you could say that we are basically waiting for two hundred entities to you know finish this migration action I mean finish to really be left with only few million side. It's not entirely correct, though, because sigh addresses all represented by more people right compound again represents hundred people there, but I'm still estimating, like there's 200 to 300 people, who we are waiting for really only few minutes. I left right to perform this migration.

G

Just lost think I've checked, which CDP's migrated. So this is the table. I showed two weeks or three weeks ago when I wanted to estimate what type of sorry how how the depth is structured audacity.

G

The inactivity who is who is fighting the death, who is minting more left and I, just wanted to see what changed during this last two weeks or I. Think it's even more like 17 days and I've noticed that emigration is still going on. Cdp's are migrating in those days. Eight CDP's larger ones migrated under paid four million side and they paid more than 300 thousand fists, so last week alone, hundred thousand fists. What's so the red the red ones are marked, which migrated in this period, so they should be actually excluded from this table.

G

But, what's what's interesting, though, is that these City, please, which migrated are still the ones who were active during during the whole during the whole period right we're still not seeing any activity of inactive City piece at all, even in the last two weeks and another thing, I wanted to show that migration is still alive.

G

I'm using this nice word from doing analytics, you can basically check how many CDP's migrate daily, so the average is still around 10 I'm, not sure if this page actually shows the the ones who are migrating partially I'm guessing that not this is only the one. This shows only the ones that are using migration tool, but still it's a good thing to see, and here I notice, the large increase of sign mentorian migration contracts.

G

So this is this is the spike two days ago when, when EXO migrated- and that's I, think pretty much from my site today,.

D

Awesome thanks that was ton a ton of data nice, though.

D

See I think that's it for for risk. We can open it up for Q&A and discuss anything. Anybody has any yes. So.

B

So when you know that there was a thread talking about whether we should revisit the government security module, delay vote, so I was curious about maybe like Derek's perspective on the whole trade-off, whether it's appropriate to start bringing that up again or whether we want some sort of like process defined inspected out for solving bugs that can maybe figure out a way to get around a delay. If one existed or something like that, but curious to revisit that conversation, David.

A

I think the latest on that was the ladder was having.

G

A conversation with the smart contracts team yesterday evening.

A

And he was going to come back to us next week if he doesn't I'm more than happy to, but I know that they were having some discussions in that space from a technical perspective, that's.

B

Awesome, that's good to know.

D

Yeah, if I understood it correctly, they're they're working on kind of a middle-of-the-road hybrid that cannot get us a little bit of the best of both worlds. For the short term, yeah I think we still have to be a little bit more patient for waiting on the on the technical team for an update on it.

F

Yeah, he also said that they have a positive result, so seems promising.

F

One other thing that was missed in the forum I started to reduce the frequency of the DSR spread governance poll, so go vote and talk on that. If you want well.

A

I'd like to hear it look at the recap, a bit Sam over your thinking here and the reason I'm asking it is we.

B

A

Data points to know whether it's time to start thinking about this or is what what time, once you think you would.

F

Yeah so I mean this was kind of pointed out. Two weeks ago, I was just kind of waiting to like double check and make sure that the votes would kind of coalesce around zero ish, and it seems that's the case this week. Actually it's gone up. The 0.5 has actually garnered the most votes, but regardless of like the specifics of 0.5 or 0.25, I feel that it's weekly is just excessive, and so yes I think.

A

That we well I'm the horse app that has to take these things in every week. So I agree with you just on that standpoint, um but the thing that I'm curious about the is do we feel that the ecosystem understands what a spread is as opposed to the way that we're doing things before have they internalized the idea of what this thing actually implies? It's it's a back and forth between maker and MKR, maximalists and DSR maximalist almost sword, Ison interest seekers and whether we have seen the system sort of settle out on this.

A

But we've done it for two weeks and we have feedback based on current market conditions, I'm just I'm playing that I just want to get a sense of whether I have.

D

A few thoughts on that um I mean the, and if the community is not kind of cognizant of this, then then we should kind of we should keep repeating it, but the the point of the spread and there's no point before we had a DSR instability fief, uh we haven't really made any efficiencies.

D

If we just now have a stability fee and a spread, it's still two poles on the idea is that the the spread which is kind of more reflective of the underlying collateral should be a slow-moving slow-moving number it should it shouldn't be changing week to week, presumably because the quality of the collateral shouldn't be changing week to week. So in that sense you could reduce that frequency, so the community should vote on what they want the spread to be, and then I mean barring any significant events.

D

They shouldn't really need to think about that again for say a month or a quarter or something and then all they would have to vote on is uh is the other component, so in theory it should be able to reduce the number of polls that have to occur every week. Yeah.

B

And also to add on to that, maybe a way to see if people understand or are engaging with the spread question versus, like the manual rate setting a good thing to do, might be to compare the polls for the for the spread as a recent as compared to like the polls for DSR adjustments and just to compare how many unique addresses if there are similar addresses like that, would be an interesting little piece of research facilities.

D

F

Spread actually I think it had 72k maker voting. I can't recall a governance poll that actually had higher maker voting. So that's a pretty high number. It's.

C

One of the top ones yeah like yeah nice.

A

Yes, this discussion is an about is a valuable one, and it's one that I've raised this. The specter of this topic I think maybe it was last year towards the end of the year and I want to do it again. That was on my topic of conversation. Starters, for this call is cadence these.

A

These boring conversations we've been having for the longest time but I think they're, fairly significant and the operational overhead cognitive overhead that we impose on the ecosystem every week is an issue that I think directly relates to apathy and engagements our desire and.

B

A

Potentially, an inflammatory discussion I want to have is I'm not entirely convinced that we've seen we have conclusive evidence that the decisions we make them a week-to-week basis, inverts the rate, have the impact on the ecosystem or on the peg, specifically that we assume that they do think there's compounding factors, there's a significant amount of leaver pulling that goes on and I, maybe I'm, just naturally skeptical to a fault but I'm, not convinced that we pull a lever and we see numbers go up and numbers go down.

A

We assume it was because we pulled this lever, I'm, not sure. That's the case of all of the time. One of the example of what maybe this week would be a decent example: we've seen a fairly unusual reaction or activity in the market in that East has going up. Surprisingly.

B

A

A cool activity event has significant impacts on the way that the market behaves, but we didn't pass an executive and talked just recently. So presumably our Peck didn't move in response to all this newfound euphoria, and so the question I have in my mind, is the peg seemed fine regardless, so do we, the bids, all anecdotal at this point, I'm looking at a graph of a good number, I think huh and then I move on with my life. So obviously that's not what scientific governance is all about.

A

What scientific government should be about, though, is us knowing that the things that we do on a week-to-week basis have some effect on the market or on the peg and I'm, not convinced that we're at that point yet.

A

So that's one of the things that I want to just put sort of a bug in people's ear that let's not lose sight of some sort of an initiative to figure out whether what we're doing actually makes sense and it has the effects that we think it does and then does that tie back into the cadence question, which Lincoln expects. What you're talking about saying. Like obviously asking everybody to like pick point two-five or 0.5 or 0 every week might not be an effective use of time and we're not we're not giving it enough.

A

Didn't ton of lead time for us to even figure out whether it did anything and that's a larger issue that I want our ecosystem to start thinking about she, maybe Cyrus I'm gonna, throw you under the bus here, um yeah I can see you like starting to sweat and there in the corner. So how do we figure this out? Man where's the magic graph that shows us whether what we did makes any difference.

D

I, don't think you can definitively prove causation, I mean if you arrow down I mean, can we narrow it down.

B

D

B

Decadence, help with.

A

The sufficient data set, though we can say that, like okay, like over the course of the last year and a half, we've moved its ability up, but a noticeable margin and all other things being equal on a relatively steady value of ether. This is what happened to the type you can start, making some guesses at that one right, I, just don't feel like we have that data set.

D

Right but I mean intuitively the regardless of if it was regardless of it truly had an effect or not. The data is gonna show that it did work right, because the peg is still exactly where it is you're not going to find what.

A

You're really interested in it is stable, I think the data will also show that significant or not anything, a sizable amount of time. Where we didn't do anything. The market sort of homeostasis kicked in right. The peg managed itself week to week. We don't know whether that would be the case over six weeks of no voting, but it would just be obviously you're right, like nobody can predict the market. This is an irrational space. We all exist in its entrance country, but still I think that it's worth devoting some thoughts to.

A

How do we figure out whether what we do on a week to week basis is actually having an effect.

D

Yeah well well, we'll put some Oh put some resources into that since you.

C

D

Mean I I can tell you right now, like you're you're, you can you're not gonna, be able to prove a definitive, no.

A

D

Relationship, that's so I mean I think we can all agree on that. I can't prove anything in.

A

This face, but like a good faith effort to just say, like oh look at this, whenever we do this, the DSR, this thing seems to happen. What do we do this ability? That thing seems to happen. Fifty percent of time so may believe, that's good in you or we can find out that, like 90% of the time, there's not even a blip, and that might be a cause for some reflection on how often and.

B

Some richest point and I think rush definitely brought it up in the past as well. It's that if the cadence is too high, it doesn't give enough of like a time period right so so do we lose anything from reducing the cadence, even experimentally, so you can bi-weekly something like that right.

A

Yeah, nothing, we have anecdotal evidence, so it's obviously we need to have some kind of concern. It reasonably thought-out effort in order to test this in this idea,.

E

If the community.

A

Is interested in but I've seen us like a week executives go through well a couple times and.

H

A

Little thought as well, in fact, in the back, it's a to make it I just look over on grass. Maybe I wasn't pay attention. Maybe that was tough that week. So we don't. We don't know whether there's something happening here, but I think that it proves us to actually start. We have what we have data now that we should be looking at to try to figure out whether what it is that we're doing.

B

A

The effect that we assume that it does I.

C

Think I think my testing different Cadence's is potentially a good idea like like we could. Let have this vote right and like this, the signal threat and then the end thing we agree to could be.

C

We will try like the consensus option for like six months and then revisit this and determine whether we want to like stick with that or try something else lay doesn't have to we could we could define that directly in the poll is to say like this is a trial for X state, at which point we will confirm or like reset and try something else.

A

Yeah yeah well, this yeah I'm concerned that we I have tons and tons and tons of years decades almost experience watching graphs and then changing something and waiting for that graph to change. I, admittedly had nothing to do with money, but still I know their cognitive pitfalls that can descend on people. Were you looking for you look at graph and you try to find a narrative. You do one thing and then you look over there and see something happen. You know, I'll get these two things are obviously related or you don't do something.

B

A

Here something over there happens and you assume that whatever that was because you didn't do anything so that we have some human failings. So we need to compensate for it, particularly in the crypto space. People are desperate to see, like oh grab, that graph just changed. So obviously we know why, because somebody posted something in China on a blog, we grasped for correlations, and sometimes it would be interesting to start reducing our reliance on those correlations as much as we possibly can and.

H

Just to sense here, various I saw him tag. You get to be be able to hear me yep, okay, yeah David, so this is not a new conversation.

H

We have a little bit additional complexity this time because of the DSR and so sifting out. What's actually going on is, as Cyrus just said, not easy. It's not gonna be clean and it's gonna be hard for us to reuse data from the SCD world for in the MCD world, because there are some fundamental differences. The other point which I think is well brought up is the stability of the peg. I am NOT attributing to changes that have been made in the stability fee.

H

The stability of the peg I'm a tribute ngey, yes, maybe due to effects from the DSR, but definitely due to differences in the eath price ecosystem due to increased aggregate amount of dyes to find diadem and compared to what had been seen in the past. I mean it's unless there's a tremendous amount of oversupply and we're just not seeing any indicator of that.

H

It's significantly grown in just the aggregate amount of demand and the matching amount of supply, plus a lot more maturity in the liquidity of dye and people's sort of desire and willingness to market make dye. And things like that. So there are a lot of factors at play and I. Think many of them contribute to the stability of the peg I. Don't think that what's driving, that is, you know the stability fee per se, and that's that's something. That's harder to measure. I!

H

Think people tend to cling to the stability fee as an easy answer, because it's something that's under your control and it's something that's easy to measure. And so the human, like psychological bias, is to jump at that as an answer and explaining as a driver of whatever it is you're measuring, but I, don't think it's quite as simple as that and I do think. It's a lot of external forces that aren't being measured. You know explicitly in this whole conversation and are driving a lot more of that pegged civility lately. Yeah.

A

I think that we can't underestimate yeah I agree was you know. We can't underestimate the impact of just this dawning awareness in the ecosystem, that this peg is being actively managed, and so tom foolery really no longer be tolerated, and so that that encourages external actors to take it seriously to hedge to look for arbitrage opportunities to get going anyway.

A

So the confidence that's created by actually having things happen potentially and I'm, not saying this is the case, but could potentially have more active value on stabilizing effect and actually moving things about the couple bits every every single week. But I don't know so that's just another s, I mean what I would dearly love those to make some effort to try and figure that it I mean.

D

Is it as a counterpoint and I mean I agree just generally that it's super complex question but I think empirically, maybe anecdotally. We have just plenty of evidence that that shows it does have an effect right when the stability, if he goes up the wider community, is up in arms right about how high the fees are and over the past few weeks seem, but absolutely tremendous tremendous interest in the DSR from.

A

The kind of the far those are different things all right, like I'm.

D

A

G

A

That what we do here is a complete waste of time. We should all pack it in right, like.

D

This mess obliterate. That's literally what you've been saying, though, that's.

A

Literally, not what I've been saying, absolutely not what I've been saying. That's what you've been hearing Cyrus I, want you to pay attention. What I'm saying is that this week to week tweaking of of numbers, we we on Monday we come all. We ask you on Friday, we move a number and then you know seven short days later. We do it again. We do it again.

A

We do it again and then we have no idea how long it takes those things to come into effect, what impact it might have on the market, whether there's compounding factors that are negating or or amplifying the things that we're doing we're. Assuming that, because we're furiously active and the peg is stable, that those two things are directly connected. What I'm curious about is, do we need to do it every week? Do we need? Oh they do it every two weeks, maybe three weeks, okay,.

D

So the question the question is: more broadly, it works, but do we actually have any increased signal by by tweaking it every week versus every month, like? Is the increase in his efficiency increased worth the overhead? Yes, that's a that's a good question for sure.

F

Yeah um one issue I've been seeing a this like this week's governance poll, is that the parameters were set to move from the week before, like the executive vote, because the one that went in last Friday didn't ask and me you know, I'm fairly active on the system, I accidentally, I kind of missed that and voted for the eight on the dying stability fee thinking. It's no change, I'm voting for no change but and then I realized.

F

They came back on like Tuesday or Wednesday and I'm like oh oops, I meant to keep no change at 9:00. So we're wondering if some of the boats were mistakenly put there now.

A

I'm the course log it's like has to submit these things on Sunday at like 11:00 p.m. it's just in order to give the executive a maximal amount of time to pass before we need to get the next batch of polls in and for the last couple weeks there's been, it's been a bit slightly slightly nerve-racking because there's been almost a 50/50 split on the executives when I'm attempting to anticipate what an accurate poll range would be for the following week, and so do I need to guess what the picture is previously.

A

You don't need to guess what the weather this executives going to pass. So there's that confusion, because it's in the intervening six hours, maybe the executive passes and now these polls are slightly skewed because they don't represent the correct range they should and then there's the other thing too, that these polls are going in that reflect a picture of the system of our ecosystem. There's only existed for three hours, so how?

A

How could anybody make any reasonable guesses about the impacts of the previous numbers if they've only been operating for three hours and we're about to change them again, based on what we seems to the impact on the peg has been it just. It feels like we're chasing outputs, and we don't know whether the inputs had a direct effect well.

D

We could I mean we can. um We can acknowledge that it definitely has some positive impact, because if it was like completely useless doing it every week, instead of say every month, then the the poles, the executive boats, would just be no change for across the board to be a.

A

Semi-Rational laughing right, like assuming that everybody understood that there is no point in voting every week. They would just presumably do all the research okay. Well then, it's been three and a half weeks. I haven't seen personally a direct correlation, so I am going to ensure that the system doesn't change at all. We do we don't see that we see these executives. These polls are reflecting whatever the current state of the system might be. We've seen it over and over again and read it in the chat and all the rest that people do cynical like.

A

Oh, my god, the peg is off by a percent. Why aren't people raising the feet we have to raise the fee like based on what happened yesterday, yeah.

H

I agree, rich I think, usually that's the type of reaction that you see of this community is like when the peg is off slightly below, then there's a call to racist ability when the peg is slightly both there's, you know a call to lowest and that that kind of reaction seemed fairly predictable. At this point, I'm not saying with the impact, the actual quantitative impact of making those changes is predictable, but the call for the reaction is, and so the there is a loaded statement.

H

I think something you said: seven squared you're saying we know it has a positive impact, but we know it has an impact. I. Don't think anybody has a clear picture that that impact in any given moment is positive or negative, and that would sort of necessitate having an idea of like okay. Well, the is currently lower by this amount for this period of time.

H

Because of this reason, and because the stability fee will have this impact on that underlying reason, we know it will solve this problem right and like that kind of end-to-end process is not as clear and predictable as I.

H

Think people like it to people so there's a point to consider, which is what is the impact of your actions and, what's the most direct way, to solve the problem that using your solving and the point is like if the problem is that the peg is low, because there's too much dye floating out there, then what's the most direct way to solve that problem. Well, your expectation is by raising the civility fee you're going to get people to pay back some of their debt, and you know, as a consequence, try to purchase up some diet.

H

But that is a really dependent question on is: if what are they going to do that, for? Why would they do that right? Well, they would do that if there's enough of a Delta between those rates, because they want to be engaging this leveraging anyway.

H

So now, what you're saying is that that increased stability to you would have enough of an impact to get them to migrate, to say compound, for example, and because you're not likely to lift a one or two percent change in the stability fee get people to change their overall behavior of what they're trying to do on earth. It would primarily be your trying to get them to change how they're engaging in those behaviors and theoretically soak up some dye on the open market to theoretically fix the price.

H

Now with the DSR, though, there's an additional lever, which is something to consider like, is that a more direct way of soaking up dye. That's sitting out there on the open market, and the answer is potentially yes, I mean if you have a nine percent yield floating out there. That's a pretty significant hope, I would say compared to the traditional world compared to other options in crypto, and it's seeing really heavy utilization there's something to consider there.

H

Where do you want the DSR to be the kind of function where it's just a flat, 9% sitting out there that people can take advantage of and we'll get in, to die, and simply for the purpose of chasing that you or do you want to use that as something that you can move up when you need to as a lever for offering more yield to more directly pull die off the open market, whereas the stability fee is obviously an indirect lever, and so like?

H

These are questions to consider I, don't really hear anybody having this kind of discussion kind of.

A

Yeah I think those are excellent points possession. That's so there's there's two avenues of discussion here. So how do we? How do we determine what is the we we are doing? Has the impact we think it's having. So where does that information come from what mechanisms are in place? That could potentially explain some of those impacts, though, is it because the system is actively managed that we have all these calls? If there's a sufficient level of background noise, an impressive grasp floating around that people are inclined to.

A

You know are of the system because they think somebody's driving the bus is have we seen some new mechanisms arrive. I I agree this yesh that this the DSR has a significant impact. That's obviously I think far more predictable than with stability might be because stability fee is more of a nebulous stick that might have long-term effects for maker holders, as opposed to the DSR being more of a benefit of the people immediately.

A

And so people see the number go up on the way so safe and that's that's an enormous motivator to get involved and pick a number. But- and we have to think about this cognitive overheads. So if we are, we hurting ourselves by by on a weekly basis, asking people to spend hours researching the system to figure out whether I never.

B

A

Go up, our number should go down. Are we getting apathy exhaustion there, or are we just maintaining a certain level of momentum where people are staying engaged and that there's value in that I don't know? Or do we consider a system where we try to wait for cause and effect to be a bit clearer and potentially lower the weekly overhead, we're asking from people and from ourselves I? Don't I was trying to phrase this whole hypothetical question in a way that wasn't colouring potential outcomes.

A

I don't know if I did a good job at it, so it's entirely possible and I'm willing to easily concede the fact that weekly is definitely the way to go and we it will take whatever amount of effort to prove it with a series of graphs that I'm sure fish and pretty much you can just whip together in a moment's notice to prove to us that you know we do this, and that happens great job done.

A

Let's move on to our weekly kids but I think it's a good, it's good to diligence for us to figure out whether that's actually the case or not, because if it's not, we could save ourselves a tremendous amount of operational and cognitive overhead by adjusting the cadences I. Think.

H

The the a good way to frame that would be what what does it take right. So what would it take to justify making weekly changes? Well, a you would have to be confident that the changes you're gonna make have an impact that takes less than seven days to properly. That's.

A

The feel sure, and also like a direct relationship right, wouldn't it be like the stability fee- has going up this much and then the peg moved that much and that's the same behavior that we see week over week. Right, it's and it just like. There should be a direct correlation, or at least somewhat of a direct correlation with some standard deviation and there's.

H

A discussion as well as like how you choose to measure the peg right, because it's a stylistic choice to sit there and say you want to look at the 24 hour weighted average. But if your volume is very low and you had a large quantity of trades outside of a 24 hour period that were you know well above the peg and then just so happens that that semi arbitrary time window has moved past those trades and now you're.

H

Looking at a tiny overall volume of trades that are below the peg, you, theoretically, your average is now showing below the peg. Even though you know, 25 hours was a large quantity of trades, and so there may be a discussion about like a better methodology for looking at this. If the purpose of that price is to drive decisions on what actions you choose to take monetary policy.

C

Yeah because that's.

D

C

D

About on I mean it's just a test idea like on Mondays rich, you look at the the recent 24-hour waited price and then, if it's within the quarter plus or minus a quarter of a percent, we skip the stability fee poll. That way.

H

Well, and so my point is sort of maybe looking at just 24 hour. Slices is too naive for the tradition. What you're trying to do here and it's more of like if you've seen a certain you know they call it over a seven day, time frame or just an infinite time frame with like a time-weighted decay of like how much you're waiting trades. So it would be like a different methodology entirely like a 24/7.

A

Day thirty day, average of average of the peg, and then people can say, oh well, look at this. We haven't crushed the threshold that requires us to do any voting this week. The peg is fine. If you look at it from a seven-day threshold, organ right.

H

Different, like momentum ranges, media a better way of thinking about this, and then sure somebody can come along and say well in the last 24 hours. The price is petered and that will take time to propagate into longer moving average, but it will- and that's also a reasonable point to consider so. There's like a human element to this ball. Yeah.

A

Another mechanism we can add on to this, though, is how long historically has it taken for a corrective measure in the governance ecosystem to fix the peg. So if it turns out that sure had a crater and then 90% time its recovers the day later or something, but if there's a significant deviation, does it take seven days five days, whatever 30 days for governance activity, to correct the peg? And that's that's another thing. Well,.

H

So and the presupposition there is that that governance activity is what corrected the peg, where we did that time. Magic problems well, I'm, not even a hundred percent, convinced that, like that's quite simply how that works like there's, there's time involved with changes in each price, there's time involved with people, you know observing and recognizing opportunities for arbitrage. There's differences in in order books over time as well in, like that's, not really something. We've had the time, or at least I personally had the time to like look at in depth.

A

H

A

Perspective comes in so maybe I'm just asking dumb questions, but Here I am I, barely understand what kryptos all about so I. Look at this number 212! Oh my god, we're mooning I, wonder what's happening to the peg and I look at the pen like almost nothing. So how do I? How can I reconcile those two activities right? So what.

F

A

Should have furious levels of governance activity to do an emergency peg intervention, but that hasn't been required. So do we what's going on my question? I mean.

H

In my mind, I see that, like I see the peg being stable at a train call, each price is nowhere near the same, the peg being stable at 160-foot they're, two very different things in my mind and I. Think like talking about these things outside of just Diet price, is I. Think a more holistic discussion like there's a number of things suppose I place.

H

But you know this is primarily a conversation about timeframes and I have said before I'll say again: I would be shocked if you're able to take government's actions on you know a faster than seven day but seven-day timescale. Even two correct. You know economic instabilities in the parent clique, I, don't know if I know that it would be very unlikely or I think it was really I agree.

A

I think I have the same skepticism that I know I'm happy to be proven wrong, but I don't believe that when we said ok, whatever its abilities going to accent, therefore, the direct result is that the numbers have changed. I. Think that frequently it provides sort of a forward guidance there. Social and cultural implications that are farmed have far more impact in indirect labor plug, but maybe I'm wrong pattern. I would.

B

A

To see some kind of proof either with or evidence he broke either way. I.

H

Think it would be interesting to see the problem. Is you don't think it really allows it to happen, PETA on when peg first slightly and then recovers on its own without any sort of governance? Action, though, especially in the SCD world sort of t, believe that's something that we will have observed in the MCD world. That was not really something that was happening, because people would make changes faster than you would see what would happen in the system.

H

But that's an important data point like one way or another to learn. What does the system do if died just below the peg and to no governance action is taken. How long does it take to recover? Does it recover it? Does it just get worse.

A

Yeah it'll be a fascinating experiment. I don't know if we need to play chicken, though, with like just stick your hands off the wheel and see if we crash or not, but at least playing with cadence a bit or experimenting like do we let's get establish a baseline, let's figure out what we're, as we did in the past, and what effect I had come up some best guesses, and then we can potentially decide where the community can just find any weeds dangers. The community can decide if they want to like.

A

Let's just do two weeks and see what happens then and then, if there's and then get another baseline and then, if you know other, it supports or contradicts the those assumptions and then potentially goes to three weeks or maybe goes back to one week. Two notes.

A

So I aren't even suspicious decline. Give us a ruling.

E

A

Uncertain all right.

C

Hiv the story, just summarize it quickly yeah.

A

There's something interesting going on. Please.

C

Cool I was saying shesh. We should look at larger time frames than just 24 hours. Sam suggested a seven-day exponential decay version. I suggest it might be good to have multiple moving average. A certain English like 24 hours, seven day three day you 90 day, would be interesting just to see Sam sort of I. Could it might be more important to wait. Today is data more than yesterday's data, which I'm not really convinced out, but I. We would need to I, don't know what the pros and cons of doing that would be so.

H

Thing was food.

H

It makes sense to weigh today's data more than yesterday's, but the question is: does it make sense to way larger trades more or less than recent trades, and that's now a different question right? So if something is, if a trade is ten times larger, do you weight it?

H

You know ten times more and then, if a trade is one day more recent, do you weight it twice as much three times as much cetera so like determining what the relative scales are between waiting size of trades and volume versus recency is like a super nerdy quantitative discussion like and I, don't know maybe Sam and I can take. Those could.

C

You like make a cool girl. You had, let's light us to change the the waiting's problem. Yeah.

F

Yeah I can add that to dye peg. Oh.

C

Really obviously, yeah.

A

We can okay, you can all come up with our own custom pictures that support our assumption. No, but yeah.

H

A

Be an amazing addition to type thing I'm going to ruin the party, so it's we've gone ten minutes over the Q&A session and I think that people have work to do not have meetings coming up, so this I would like to maybe call to action so maybe Sam and Fischer. You guys gonna toss this idea around a bit and continue the discussion next week.

F

Yeah sure there's tons got.

A

Delegation a little bit all right thanks everybody, there was a super spicy discussion super happy with the way these governs falsity regards. Please join us in the forums reasonably sure that this chat we just had will surface there soon enough, and thanks for joining us and I'll see you next week.