Transcript of the Ampleforth office hours on 30.06.2021

Simon:
Thanks for joining office hours today- it’s an exciting one. Brandon, Evan, Aalavandhan and I are joined today by Mark and Chris of Prometheus Research Labs (PRL). Mark is the head of engineering/CTO at PRL, a smart contracting development company. He was previously the Ethererum dev team lead at Bitgo. Chris is busy building alongside Mark where he is an engineering contributor with PRL. It’s great to be joined by these two, and we have a lot to get into with them but i’d like to start with Mark + Chris saying hello!

Mark:
Hey everyone — really excited to be here! The team at Prometheus is building some very exciting protocols that are enabled by AMPL’s rebasing qualities. We can’t disclose too much at the moment, but happy to answer any questions that I can.

Chris:
Piggybacking on Mark’s comments, happy to meet the community and learn about what y’all are excited about in the DeFi space. I’m really eager to learn the pulse/vibe of the community.

Brandon:
A friendly anonymous person claimed the BAL tokens to the geyser pools already, which is the first step of the process. We’ll complete the remaining steps very soon (probably today) Thanks 0x6d2…!

Question 1:
Curious about how AMPL’s rebase complicates things when integrating with other DeFi protocols, does the team view this as a major hurdle facing AMPL adoption?

Brandon:
I wrote a medium post around how we think about this a while ago: https://medium.com/ampleforth/the-letter-and-spirit-of-composability-9921a65f0502.

But as more pieces are built around AMPL over time, this also changes the landscape (making some things more and less important.)

Evan:
I would say it was a major hurdle for us with centralized exchanges early days, which pushed us to innovate and seed liquidity on DeFi systems (like with the geyser program).

Similarly, the challenge of integration, is encouraging us (and folks like Mark and Chris) to innovate up the stack with lending/borrowing/derivatives. It’s a gift and a curse, but we’re going to see some truly different stuff as a result soon.

Aalavandhan:
I think people do not see the power of Rebasing assets completely yet. I personally don’t view them as “harder” to integrate. It possibly makes integration much easier in some cases. You can count the number of AMPL (or any other rebasing asset) in your wallet/contract and roughly infer its value. If used this way it could actually simplify many smart-contract structures.

Evan replying to Aalavandhan:
Agreed, I think this needs to be shown. Once you’ve seen an application design that assumes rebasing assets as inputs, it’s hard to unsee the simplicity and composability of it.

Simon:
I sometimes view DeFi in this way: Two things make DeFi special as opposed to traditional finance

1. The Decentralized aspect. This is something not available in traditional finance
2. The innovation and new ideas that are enabled through advances in technology not available to traditional finance

The rebase is one of those unique to crypto technological advancements that unlocks a great number of new performance characteristics. Building and experimenting with the uniqueness of the rebase is something I’m quite excited about as it falls firmly in category 2 — and is not just recreating traditional finance on the blockchain.

Question 2:
Is there any information you can give about what PRL is building or resources we can check to learn more? Really interested to see what unique projects the rebase allows.

Chris:
We have a public website with some limited information. Check out our about page: https://www.prl.one/about.

Question 3:
Can you guys provide some examples of AMPL derivatives? Are we talking futures, options?

Mark:
We think that lending/borrowing, options, and a host of other instruments in the financial stack can be built on a foundation of rebasing assets. The rebasing aspect of AMPL actually makes these assets a lot simpler and more composable to build.

Aalavandhan:
Ive floated this idea on the office hours before. I think the first powerful one would be a derivative of AMPL which is supply-stable but moves with AMPL’s price.

Evan replying to Aalavandhan:
Anything separates price volatility from stock volatility would be sweet imho

Question 4:
Has the team ever considered building their own lending/borrowing platform specifically for AMPL since other platforms seem to be facing challenges?

Evan:
What a wonderful idea! would be so great if something like were to come out soon

Question 5:
Any ballpark on the first Forth voting opportunity?

Brandon:
Voting and delegation should be going live any day now. We’re working with the team over at Tally who’s building a very nice, generalized user experience for governance.

Once delegation is easier, we’ll also flip Snapshot signaling over from balance to delegated power.

Question 6:
Any updates on Tron/DOT/Near integrations?

Aalavandhan:
We’re working on Polygon now. It’s almost done actually.

Question 7:
So you are talking about a wrapped AMPL asset (referring to a derivative of AMPL which is supply-stable but moves with AMPL’s price)? Would that remove some incentive to buy/sell the base asset?

Evan:
The tension with wrapped AMPL is that it can dilute price stickiness a little, similarly to how AMM’s with instantaneous price adjustments dilute stickiness relative to price on CEX’s. There are both advantages and disadvantages to this, but for now our view is that it’s strictly better to stimulate liquidity for rebasing AMPL itself than to do so for a wrapped AMPL.

Question 8:
Have you planned the staking for your FORTH token ? If so do you have a date or a time frame ?

Brandon:
It’s possible, but the question I’d ask is, what is the staking for? Staking implies that there’s some work being done, where failure to perform slashes the stake from the user. I’m just not sure what “the work” would be for staking FORTH.

Depositing AMPL into the geyser is “sort of” staking, in the sense that depositing LP tokens means that they’re doing work by providing liquidity. Failing to do this slashes the future claim on the ecosystem distribution.

Liquidity for a monetary asset makes that money a “better monetary asset”… but adding liquidity to a governance tokens doesn’t necessarily make that a better governance token.

So I’m open to any FORTH staking idea that can answer that question above.

Question 9:
For AMPL are we talking futures contracts, yield swaps, options? Derivatives have been mentioned often, but i’m trying to understand what that means specifically.

Mark:
Lending/borrowing, and options are both things that we are thinking about. I unfortunately can’t go into too much specifics about this but I think they are both great opportunities for AMPL.

Question 10:
How easy (or hard) has it been developing around ample’s rebase function? Is there anything else around the dev work you can share? Also what are some previous projects PRL has done the community could look at?

Mark:
It has been very easy, but definitely took a bit of a mindset adjustment (coming from building on non-rebasing assets — I worked on the WBTC protocol at BitGo). I hope that our upcoming project exemplifies this new mindset and shows the community some exciting new ideas for building on top of rebasing assets.

Question 11:
What processes do you find the rebases simplify?

Mark:
For example, a lot of the complexity in the Compound lending contracts revolves around the pricing + liquidation mechanism for collateral. When building lending collateralized by a rebasing asset, to find changes in price, you just need to check the change in balance!

Follow-up remarks:
The rebase uses an averaged, lagging price to determine the rebasing %, not the spot price. Unless I’m missing something else, you really can’t tell the current price from balance change alone.

Mark:
True — any protocol getting price information this way needs to take the lag into account

Follow-up question:
Yes. How would you determine that without, well, an external oracle? The “lag” itself doesn’t have much of any correlation on the current market price, especially when it’s further away from the rebase time

Mark:
One way is to make your protocol not need the exact price, but be flexible toward an analog of the price that gets there eventually. The main reason a lot of lending protocols need exact pricing is to determine if a position is under-collateralized and thus should be liquidated. So, what if the lending platform didn’t have liquidation

Question 12:
I was told to ask Evan about Fragments. Can you give any details on it?

Evan:
Fragments, inc. is actually the delaware c-corporation that we operate. It’s a contractor of the Ampleforth foundation. It was also the placeholding name of our early versions of the protocol.

Question 13:
Does the team think that AMPL won’t stay in negative rebase as long once lending/borrowing is used heavily with AMPL?

Evan:
I think there will be more stickiness in general, by arbitrage, but as for the time it takes to reduce supply (in the event of a secular decrease in demand) things like AIP-5 would have a more direct effect.

Brandon:
First, I think I’d say that decreasing negative rebases also implies decreasing positive rebases, all else being equal. (Unless you want to assume unbounded upward trajectory over time)

I expect that once people are able to “take both sides” of the bet in the marketplace (i.e both long and short), then it will lead to a more efficient market and faster price discovery. Does that mean less negative rebases (and on the flipside positive rebases)? If it means there’s more time in equilibrium, then yes.

Another promising angle is AIP-5 here: https://aips.ampleforth.org/AIPs/aip-5

Question 14:
Could the team give more details about AIP-5? My initial thoughts are that it adds complexity. While that’s not inherently bad, the good thing about AMPL is that I can explain it to a three-year old.

Evan:
Right now the AMPL supply policy is like a kite. A gust of wind (ie: influx of demand) can cause supply to exponentially increase. But then it kinda gradually floats back down slowly. This means, for a given quantity of tokens and price deviation, it tends to take longer to remove supply than to increase it.

AIP-5 brings more symmetry to the system. The disadvantage to it, in my opinion, is that you can’t easily calculate supply changes in your head.

Brandon:
I actually think it’s pretty easy. It essentially replaces a straight line that goes up to infinity (and down to zero), with a curve that has bounds on each end. Everything else is the same.

The benefit is that it decreases sensitivity of the system in extreme scenarios like black swan events, software bugs, or an oracle attack.

I can say that at Google, this is basically how they’d design the system from day one. The magic of their search algorithm is basically “curves that look like this”

/

me making hand wavy motions

Question 15:
Has the team discovered any other new use cases for AMPL since the EFi roadmap was announced?

Evan:
YES — can’t wait for more to come out on this.

Question 16:
I’ve a broad interest in all of the PRL ideas, but I think price-stabilisation wrappers would be one that captures the widest attention. Time and time again we see ill-fated attempts at stablecoins…

Mark:
Ah — what we mean on the website by “price stabilization wrapper” is not a stablecoin, but a way to create price-stable, elastic-supply tokens derived from the assets you know and love (ETH, WBTC etc). Since we are building rebasing-native protocols, this helps extend the utility of our projects.

Question 17:
Along with these questions, has the team considered the use of CAPs (or in Ampleforth’s case AAPs) for proposal submission? https://discord.com/channels/613841245969580143/613841246607376405/859892272626401291

Brandon:
I generally like CAP. Copying from a previous response in the #governance channel: If I delegate to an immutable smart contract, then I know exactly what it will and won’t do. That makes me feel safer, and I may delegate more to form a proposal because of it.

I wouldn’t say it’s hugely transformative, but it would be a nice to have once everything else is in place.

Follow-up remarks:
Absolutely — I recall you mentioning this. What about these:

1. What are some mechanisms being considered to improve/stimulate governance participation (outside of delegation)?
2. What are some considered implications for this regarding $AMPL? (as per the on-going derivatives conversation)
3. As delegation rolls out, how will potential proposals integrated affect cross-chain Ampleforth extensions?

Brandon:
Hmm, good questions. I think the goal of governance is to maximize its effectiveness in shepherding and securing the protocol. This doesn’t /necessarily/ mean maximizing participation. I’m fearful of any direct incentives to governance participants because it risks distorting the outcomes via farmers, etc.

I’d want to maximize participation of informed people who have skin in the game. One thought we had was a voting boost to people who lock up AMPL. This is similar to Curve vcCRV approach but would require some work to make happen.

Question 18:
Brandon and Evan agreed on multiple occassions on the priority of enabling lending/borrowing for AMPL on a large platform, preferably AAVE. With the AAVE team stalling progress on that front for an unknown amount of time; do you think the time is right to start work on integration on one of AAVEs competitors like Compound or Maker? This may serve as an incentive for the AAVE team to get work done first and also as a hedge against further delays.

Aalavandhan:
We made some code changes to aAMPL after the first audit based on feedback from the AAVE team. This triggered a follow-on audit with peck shield. It completed last week and they didn’t find anything. So from where I stand the code looks good.

Brandon:
I want AMPL to be everywhere, myself.

Aalavandhan:
AMPL today vs AMPL when on AAVE:

Link to the start of the office hours on Discord