Float Community Call (Thurs 25th March 2021) — transcript and video
Thanks all those that joined us for the community call with the band on Thursday 25th March.
Agenda:
- Phase 2 update
- Development update
- Questions and answers
Video
Abridged transcript
Phase 2 update
John L: I’ll start with a quick Phase 2 update. Float Protocol started with a democratic launch in Phase 1, with whitelisted addresses and limits per pools, to enable very fair and wide distribution to those who are active and engaged community members in the DeFi ecosystem.
We had a governance vote on the pools that we wanted to include in Phase 2 last week. Those pools were chosen via a new fork of Snapshot Labs, called Scattershot, which allows people to vote with BANK in a proportional manner.
Phase 2 started on Sunday, and now we’re into the third or fourth day. A community member (see here)wrote a forum post, and actually, in fact, many community members raised concerns regarding the state of Phase 2 in terms of distribution and whether it is really in line with what the aims were of the protocol. Either way, we thought it was a reasonable concern and the evidence presented was, I think, very good.
So we took this community forum proposal and put it forth for a governance vote (see here), which is actually underway now and it closes tomorrow evening at 6pm (now finished). And this vote will essentially be deciding whether we end Phase 2 at the end of this week and start something called Phase 3, which will last for six weeks and we’ll distribute the previously allocated rewards to the last week of Phase 2, so 10,500 Bank to this Phase 3.
But the special thing about Phase 3 is, not only will it last only six weeks, but it will also only contain the BANK-ETH SLP. Before that, the vote’s going on. You can never trust the polls, but people have been voting with their BANK in accordance and it’s a strong sentiment for the shortening of Phase 2 and the additional Phase 3.
Further thoughts on changing Phase 2
Paul M: Building on that, I obviously wanted to say that we were very keen on sticking to our guns. We don’t want it to be a fickle, weak promise like, we do one thing, move on to another. But I think it’s also important that we engage community members and make sure we have people who are voters and who are democratically inclined. So I think that it’s also very positive that we had this kind of community engagement. I think the TVL is really cool to have but it’s very much an indicator for us, it’s not something which we gain a lot of value out of, because we don’t utilise that TVL in a very dramatic way.
Phase 2 was always designed to be a distribution point and a way for a bit of price discovery. It’s why the questions about the APY and things like that don’t matter to us, because in our eyes, you know, if one BANK is one BANK, and at the moment one BANK is worth how much it is once the protocol has been fully released. There’s speculation at the moment. Speculation that we think is interesting, but ultimately it will come down to the value of the Float protocol itself, which we are you know, obviously on board and very confident on.
Development Update
Paul M: We’re building towards the final freeze where nothing else can be changed. Literally today, I was looking at the Uniswap v2 libraries, the bits that are open-source, and they’ve got this gas-efficient safe maths which I’m gonna integrate into the protocol, because it’s always nice to get some cost-savings there. Linked to that is the new target price calculation, which we’ve called the sort of monetary policy from the contract perspective. So that’s gonna be deciding what the target prices actually become.
Launch plan
Paul M: So, linking on to that. I’ve seen a question about the launch. The launch’s planned as soon as possible, as soon as the audit is complete. Because when you have this code freeze, it gets handed over to the auditors and the reviewers. And any sort of suggestions we get back, we’re gonna have to sort of consider those, and see whether those are things we’d consider changing, or if they are things which are interesting but so to say, a stylistic concern or if there’s a documentation point which should have made clearer, which is also stuff included in the audits like that. And then we’ll update those if that’s necessary, if it’s a bigger thing then it might take a little bit longer, which is why it’s kind of tricky to pin us down on a hard date. Because we don’t wanna say “Oh, we’ll release it in four weeks” and then the audit takes five weeks, but we decide to release it anyway just because, and that’s exactly what we don’t want. We want to make sure that time that we spend is spent correctly.
Auctions
Paul M: So linked on to that, the UI/UX design for the auctions, those are progressing nicely. We’ve got this sort of design sketched out for how we want those to look, how people are gonna participate. Obviously too, an element of that is going to become sort of very behind the scenes, so it’s, we expect there to be lots of arbitrage bots or things that don’t use our user interface to be participating in these auctions. That’s to be expected and I think that’s actually good. We just wanna make sure that the auction also displays that clearly, as well as also allowing people who are at the human rate to participate, especially early on where those bots won’t be as active.
Timeline for audit
Paul M: So yeah, code freeze, the time where no one can touch the code any more, is gonna be in a week. So that’s the point where I’m determined to say “no more code touches”, unless it’s a documentation thing, and then it gets handed off for everyone to look at and review.
Day in the life of Abbey Road
Paul M: In terms of the day in the life of the Abbey Road hacker, the relevance to that is basically I tend to work late, it’s sort of my work pattern, whereas John likes working early. So it tends to be a sort of hand-off, where I will sort of go to sleep and John will start working, so that’s kind of how we always manage to be a community presence there. And I normally sort of start off work by seeing if there’s anything relevant from the channels, then move on to — at the moment I’ve been very smart contract-focused, did a bit of Scattershot updating for the strategies yesterday night, faffing around with those, making sure the LPs can vote, trying to optimize the earned calculations. Because in general, it’s very expensive to compute for the two and a half thousand voters we were having before. So we needed to do a bit of optimization there to get it into the running stage.
More on audits
Paul M: Linked to the reviewing, the audit stuff, it’s worth mentioning that we’re doing a split between peer reviews and official audits. The reason we do that is because we find that sometimes peer reviews or reviews which are less formal actually bring up more stuff than the official audits do. We run a lot of, you’ll notice if you look at the GitHub or contracts in depth, that we already run a lot of like, automated tools to test the security of the contracts. And often that can pick up a lot of stuff by default. But most of the logic style bugs, or the logic sort of flaws, tend to be picked up by someone who can spend a fair amount of time, and he really knows that stuff inside out. Yeah, and we’re aiming for two professional audits alongside those.
Hiring
Paul M: In terms of adding people to the team, yeah, I’m kind of on board for it. The team, I think, is ready to sort of move faster and expand if necessary. I think the plan is as a kind of group is you get a pseudonym, everyone will be someone who will come into the team under that. And that will be your representation to the Float community. And maybe these kind of roles will swap out as and when. So if someone else becomes sort of head dev, which is kind of the role I’ve been taking over, then maybe there’ll be a new pool that I’ll move off to a different character. I’m not planning that anytime soon. But I think that’s kind of the way we imagined the team structure kind of flowing.
John L: And there was always a fifth Beatle, right?
Adoption of FLOAT
Paul M: So there are two main avenues for approach here we see. Avenue one is collateral and becoming the new form of collateral for various protocols. The reason that’s useful is because of the dampening low volatility effects. The second avenue is becoming the unit of currency for certain applications on-chain, specifically related to crypto applications. With the reason we expect that is, because it’s more natural to price things in FLOAT than it would be to price things in dollars, because the price of ETH say, increases so dramatically. That becomes very volatile and hard to price things in, but the price of Float will increase enough. So your purchasing power within the crypto sort of purchasing world is protected. Yeah. But the dollar value is more, so you get these kind of digital goods are more easily priced in float than they would be in any other system. So that’s the sort of second avenue for adoption.
Price change of FLOAT over time
Paul M: In terms of another question that’s been dropped in the chat is, how much should we expect Float to change in price. This is generally something we’ve covered in the monetary policy, and we’re modelling the target price changes to it. What that means is we want this to have a target price change that’s in line with how the crypto market moves, we don’t want it to be a completely static “stays as 1.618”. The reason we think this is important is because real currencies, like the dollar is the classical example, do inflate in the amount of like, the goods you can buy with them. So if you get paid $10 ten years ago, you’re going to be quite upset if you’ve got paid $10 today. It’s that kind of inflationary model for how much you can buy with stuff, we think is quite important. So we have our first assumption of how much that target price will move, our very first one had a much more dramatic move than we think we are going for now. So we’re aiming for about 30%, for 2020 as opposed to 50%, which was like our complete original one. We also want to make sure that those policies are sort of malleable enough that we can adapt those. And it’s a bit like something that governance will adjust as the project comes out, depending on what the desire and the tolerance for float adjustment is.
Death spiral case
Paul M: Which brings me on something I wanted to bring up is the sort of, I think, the death spiral kind of discussion is quite interesting. Because I like to think of it for Float as more of a twirl. And that because it’s a floating currency, it means that the target price can renormalize. So we never have a complete collapse of the protocol caused by these kind of classical scenarios of people trying to sell off, like FLOAT and BANK too quickly. So if there’s like a death spiral without any death, it is kind of a more of a twirl in my eyes. But it’s interesting because the absolute worst case is a 30 day decrease in price, based on what I was talking about that target price stuff early. And it’s likely to be much less than that, in order to find that new normal or that new price. And the link to that is because there’s never a price collapse, it means you get that confidence that the protocol will come back and that BANK will start making a profit.
And if there’s a panic selling, ie there’s sales while we still have the collateral to support the Float price, which is code for if the basket factor is above one, then BANK holders actually make a profit on that panic selling. Which is why I think it’s quite an interesting concept that with this kind of floating currency, that you can gain that sort of death spiral to make a more profit-based scenario for people who stick with the protocol. And because they gain if they stick with the protocol, it kind of curtails that sort of game theory effect, which causes this spiral to get really bad. So it’s kind of two aspects. One, it won’t ever mean the protocol collapses completely. And two, there’s this game theory sort of positive side for sticking with as the price gets production, because any panic selling turns into profit.
Comparisons to Reflexer
Paul M: So Reflexer is a really cool model. Like, I think the PID model is really interesting. It’s not the model that I think is the most effective at capturing this idea of purchasing power. Because it’s very, it can be very tightly constrained to a very narrow band. But that narrow band also harms it because it means if ETH does a 20, 30x, or more likely a 5x, as it has done in the last year, if you hold just right, it becomes a very similar scenario to holding just DAI, for instance. It becomes this kind of, you’ve missed out a lot on that opportunity.
Whereas if you hold FLOAT, you will still be able to buy the same amount of goods with that. Whereas if you hold RAI, and then the crypto price moves up, or the crypto sort of, think about gas prices. It’s a great example of like gas prices now compared to what they were two years ago, if you were trying to do this all in dollar terms, you would never be able to afford the same operations. Whereas if you are holding that money in FLOAT instead, because the purchasing price or the target price will adjust up in correspondence, it’s just a dampened version of that, it means that you still gain some value.
Linked to that is that RAI has a quite interesting property of the redemption price moving down if there’s more demand for RAI. So if, if people really like RAI for whatever reason, then it actually means the price means down, in a sort of opposite way to the way you’d expect. Whereas in Float, if there is more demand, then the price moves up or down or gets adjusted back to what you’d expect, depending on whether ETH is stable.
BANK holder incentives
John L: First of all, if people aren’t aware: the BANK token serves three purposes in the protocol. One is it takes profit in times of excess demand for FLOAT , which you can argue actually is one incentive right there. And two, it sometimes supports the price of FLOAT in times of low demand, especially if the basket is not well supported. And three, it governs the Float protocol. And I think, you know, one is an obvious incentive.
On three, I think governance has proven itself to be very key, even in the short term. So far, BANK holders have been able to literally decide which pools they want to see in phase two. And furthermore, they are currently deciding whether we’re even going to continue with phase two, in the long run, or in the short run. So there’s that.
And furthermore, this element of governance is so important for a protocol which is, you know, been raised by the community and as you know, has this democratic launch aspect to it. It’s something that you as a Bank holder, you kind of really, I don’t like the word, but you know you buy into this, in this sense.
Then the way the auctions are designed, to go back to point one, is that BANKers actually benefit in many cases. And that’s in these expansionary cases, where there’s high demand for FLOAT.
And even in the case where the basket is very well supported, but the demand for FLOAT is low so we need to contract supply, we’re looking at essentially situations where we’re willing to actually buy that BANK off the market and actually capture the excess basket factor and pass it over to BANK holders, as well. Because this excess basket factor isn’t necessary for the protocol, we’re here to build a capital efficient system.
Participating in the stabilisation auctions
Paul M: Yeah. There’s been another question in the chat about what degree of technical competency will auction participation require? It’s a good question. And I think it’s gonna become progressively more difficult to make money on these auctions as time goes on. Because there’ll be, the efficiency will increase, and hence, the opportunities for making a huge amount of FLOAT or BANK is going to be trickier.
That being said, yeah, so there will be an element of buyer beware where the auctions come in, because you can make trades that are very unprofitable. That’s kind of the point of having these Dutch auctions where you descend from a price that isn’t great to a price that is great. So if you jump in too early, there is a chance that you are gifting the protocol money, in the sense you are giving it more than you will get back. That’s something which you should be aware of going in, and we’ll make that very clear. It is also true though, that we will do our best to make it clear in the UI when we think you will make money.
Paul M: So we have to do this kind of modelling when we’re doing the simulations, is when arbitrageurs will take a certain offer. So that is based on what the market price is at the moment, what the target price is and kind of working, given the amount of FLOAT we’re going to get out, or given the amount of bank in ETH we have to put in, you can work out what that profit ratio is. Most likely, because we’ve already implemented that code, we will probably just drop that into the auction as well. And it will tell you what the likely profit opportunity is. If someone is really on it, then they will beat you to the punch in the sense they will take a smaller, more narrow profit opportunity because we will likely give some tolerance ranges. And hence, we’ll be able to gain profit from the auction before you can. The auctions are kind of a beginning early on and more of an interesting, fun mechanic. Because we expect the kind of UI side to disappear slowly over time and it become more of a monitoring system, in like you can see what happened in the auctions, you can see how the target price adjusts, and the market price adjusts. Yeah, the answer is we’ll try and make it very easy to to participate without the technical competency, whether you will make money using that logic, I won’t be able to say. It ultimately comes down to how quickly everyone reacts, and how quickly people who are willing to take more risk than the UI will present to you comes into it.
John L: Yeah, it’s this amazing thing about Dutch auctions, which is why we’re kind of using them. They really allow this capital efficient system for us to move, for us to try and move this market price to target price in the simplest way possible.
Paul M: And always have that reliance on not having completely up to date prices. Because we use an average price for the past sort of period, we can actually get away with that, because these Dutch auctions start with a sort of bad trade from the perspective of someone coming in but a good trade for the protocol. Which means that if the prices turn out to be stale, all it means is that the clearing price or the price where we think profitability will occur, actually just shifts earlier in the auction. And if we’re wrong in the other way, it just shifts later in the auction, which means you just have this dynamic of shifting about when this sort of profit point occurs, depends on how accurate the TWAP is or the time weighted average, in case I didn’t explain that, for that price period.
Socials
Website — https://www.floatprotocol.com/
Docs — https://docs-float.gitbook.io/docs/
Twitter — https://twitter.com/FloatProtocol
Telegram — https://t.me/officialfloatprotocol
Medium — https://medium.com/@floatprotocol
Github — https://github.com/FloatProtocol/
Discord — https://discord.gg/nVCZacJJqM
Forum — https://forum.floatprotocol.com
Scattershot (our snapshot fork) — https://scattershot.page/#/