Transcript of the Ampleforth office hours on 24.03.2021
Brandon:
I see a few questions about Uniswap v3. Maybe I can tackle a few of those. It’s a big topic though, so not sure I can really get across all my thoughts on it.
I think it’s cool to see them advance the space and trying to innovate. Especially so with the efficiency of liquidity use. However, I do think it comes with some drawbacks and it will be really interesting to watch how it evolves in the marketplace.
There are two factors that I think were instrumental in Uniswap’s popularity and use: 1) Ease of providing liquidity for yield, and 2) network effects through systems like the geyser and others. I worry that v3 is drawing back from each of these.
With liquidity bands, I think it will force LPs to be more active in managing their positions. Also, fees are not auto-invested back in the liquidity pools and need to be managed as well. This may be more difficulty than small users are willing to deal with. LP activity may start to be dominated by large institutional players and whales, while the larger number of users end up on other easier to use platforms like Sushiswap. So, perhaps Univ3 is more efficient with its liquidity than Sushi, but if Sushi has more TVL it might be more than enough to compensate to generate volume and fees. Which will it be? Hard to say, we’ll just have to wait and see.
As far as network effects go, the NFT representation of liquidity will make these connections in the space much harder. In their paper, they say they expect there to be ERC-20 wrappers from partner protocols, but these will likely be very clunky to deal with. Liquidity mining programs could shift focus to other platforms because of this, especially if that’s where the most LP actually are. We’ll have to see!
In terms of support, there is no official support for ERC-20s except for the most strict kind. While it may work with rebasing or fee tokens unofficially, expect it to be limited. From looking at the code, I know that pools no longer need to be sync’ed as before, but fees are accounted for separately from actual token balances.
I think the v3 design is a natural progression philosophically from where they are, but I don’t think other platforms will necessarily follow them. I expect this is the point were, e.g. Sushiswap and other AMMs start to diverge from each other.
Question 1:
Syntax aside, why use AMPL for DeFi lending?
Evan:
So, I think I’ve said this before but I tend to break “reasons for doing things” in crypto into two categories, long-term ideological and near-term practical. The practical reasons tend to be speculative but ideally they bootstrap the ideological reasons.
Practically:
With AMPL there will be speculative reasons for using it on a lending / borrowing platform near-term because of its cyclical price. When AMPL is trading above the price target, it will make sense to borrow because you know eventually you’ll be able to pay back that debt at a cheaper rate. Similarly when AMPL is trading below the target, it will make sense to pay back any debt obligations denominated in it.
Ideologically:
We ought to have an unbreakable asset (like AMPL) that can be used for lend/borrow. We shouldn’t be satisfied with the current-day dependence on stablecoins like USDC (who have custodial risk) for building sophisticated financial instruments.
Symbiotically:
The practical use of AMPL for speculative reasons, will actually contribute to balance level stickiness. It can and should bootstrap adoption of AMPL as a building block in additional financial instruments.
Question 2:
Does the team have any plans for additional exposure of AMPL? I am a loyal supporter of AMPL but if I had one criticism it would be lack of education and marketing. Exchanges would be great way to get more eyes to see but I know you can’t comment on that, so I won’t ask. But we can do initiatives to learn about AMPL, do NFT giveaways and education courses. Something fun for community while all the developments are happening. It would be a great way to keep attention while we wait on the complicated Aave and Multi Chain projects.
Also check out how some of the large products founders are in community. Like Aave’s Stani and others, they’re quite active and engaging in community matters happening in space. Would love to hear team chiming in with such things as well. The only tweets we get are when the geysers disburse rewards. I think there can be more there. Also I remember Evan used to have videos with the Linked-In co-founder, explaining things. Now it seems absolutely nothing and ultra disappointing.
If this is all purposely done because you’re all waiting for one mega marketing shift that is terrific. But from an outsider it can seem that the team lost the drive and fire from last years massive move. Thank you and looking forward to response.
Evan:
Thanks for taking the time to write this. I hear and understand your frustration with education and marketing. We’ve been focused on specific major milestone events, and will be rallying our marketing efforts around them. Some of these milestones have path dependencies that rely on external parties like Aave, Crosschain Integrations, and a few others that we can’t yet talk about. I do see your point about parallel education and marketing efforts, however, and appreciate the feedback!
Question 3:
Will Geyser V2 require unstaking from the current version,, and and restaking to the V2 version since it is new code?
Brandon:
Yes it will. HOWEVER, once you get your LP tokens into your universal vault, they don’t need to be moved from there. You can lock those into multiple programs simultaneously from the same place.
Question 4:
Today Tesla announced that they will accept BTC for Tesla and not convert back to USD which is a big deal. Do you think possibly using AMPL and stable contracts will be a way for people to finance cars or real estate in the future?
Brandon:
Yeah, I think the longer the time frame you’re looking at, the more a time-invariant unit of account make sense. This could be payments through time or also denomination of debt. If we agree to transact 500 AMPL every month for 10 years, then everything becomes very simple and safe to think about and use.
Question 5:
I think I read an old article about launching on Curve? Was the team trying to launch on Curve at some point? It seems like that didn’t come to fruition. Will there be a future attempt made to launch on curve?
Brandon:
No, we’ve never talked with Curve about the AMM aspect. Curve is tailored for the stablecoin (or, e.g. BTC flavors) use case (of which ample is not). If there were ever different flavors of AMPL like there are for BTC, then it could make sense. For example, if there were a privacy version of AMPL and a gas-efficient regular ERC-20 version of AMPL, both responding to the same supply policy, then a curve pool between these would make a lot of sense.
Question 6:
I was told that the base for AMPL is adjusted for inflation monthly. But i don’t see how that would still work in a highly inflationary environment. Is it possible to adjust it weekly or even daily in the future?
Evan:
This is an interesting question. When folks think about assets like Bitcoin as a hedge against inflation, they usually mean it’s a good store of value alternative to fiat because it’s non-dilutive. The same is the case with AMPL.
The price-target, however, is meant to preserve a unit of account. Under hyperinflation that might mean the AMPL exchange rate to the USD is behind due to the monthly update cycle. But I think that’s about the rate at which the BEA evaluates inflation altogether (can double check this).
Brandon:
It’s possible, and there’s nothing in the system that stops that from happening. It’s a difficult measurement problem, though. How do you measure inflation on a day-to-day basis? Economists spend their whole career trying to solve this — field of econometrics. We’ve chosen a particular measure that updates on a monthly basis for now (the PCE), and this could be changed in the future through governance.
Question 7:
Can the team comment about the role they see centralized exchanges playing in the future evolution of AMPL use cases, other than increased accessibility for the asset and better price discovery/trading?
Is the brighter future for Ample one that’s primarily in the decentralized world, versus centralized?
Brandon:
Philosophically, we try not to make too strong a distinction between the decentralized finance world and the centralized one. For us, it’s all “Finance”. The DeFi world adds some amazing new options for users, but traditional finance is also not going away any time soon and has many benefits in terms of scale, efficiency, and legal jurisdictions for people who want that. When we talk about “denominating contracts” we mean all kinds of contracts, not just “smart” ones.
We started this project before the word “DeFi” was around. DeFi itself is still finding its use case beyond speculation, btw. DeFi may come and go, but Ample will still be useful, I think. I hope they can both be vibrant ecosystems for Ample to live.
Question 8:
I just read the white paper and I have two questions about your criticism of commodity backed stablecoins. 1) why are they incapable of going off the peg, and 2) why are they incapable of scaling? In the case of MakerDAO demand for DAI is just demand for credit, no? And as long as they offer competitive rates relative to the market there will always be demand for DAI.
Evan:
My criticism of MakerDAO is generally that they tightly couple a credit-market with the issuance of a stablecoin. It basically forces a constant collateralization ratio, which pressures the market to act in an occasionally unsustainable way. This needless complexity can also result in the supply policy failing altogether (as in the case on black thursday).
Question 9:
I’ve been quite interested in the prospect of using universal vaults to obtain rewards from multiple liquidity programs, but… well the point of the reward program is to incentivise people to stake liquidity with the owners of the program.
But you’re not doing that if you stake in one pool and join reward programs with the corresponding LP tokens. Same question, but with brevity: if I stake in a Uniswap LP why would Sushi be inclined to let me claim from their reward program as well?
Brandon:
Wow great questions.
the point of the reward program is to incentivise people to stake liquidity with the owners of the program.
I wouldn’t say that’s true for us. Our geyser program has two primary goals: 1) distribute new tokens into the world in a fair, rules based fashion, and
2) to incentivize activity that beneficial for the network while doing so.
So we naturally want to stimulate liquidity and distribute to those who do. If the tokens physically live in a vault rather than some other contract, it would still actually satisfy both of those two goals. While I can’t speak for other projects, liquidity platforms also already distribute to their LPs where the tokens are other places (and geyser does this for two tokens tokens). An AMPL LP is a natural intersection of multiple platforms’ success, so it seems like a win-win.
Question 10:
In regards to your comments on Hayek Money and Seigniorage Shares, “They both assumed the Quantity Theory of Money to be true in the near term. That is to say, they both rely on supply adjustments perfectly offsetting price deviations. This is not true.” The significance of this isn’t exactly clear to me, could someone expand on this thought?
Evan:
Broadly, the significance is that we noticed two glaringly bad assumptions behind Seigniorage Shares and Hayek Money.
The first is that changes in supply don’t perfectly offset changes in price (ie: the Quantity Theory of Money isn’t true near term). And neither of the systems would be stable based off of supply changes alone.
The second is that, bond/coupon markets tend to have trouble surviving secular decreases in demand without the assistance of the state (ie: Seigniorage Shares is not a reliable system).
Having realized this, we decided that Hayek Money had created something robust (that wasn’t stable) and Seigniorage Shares had created something (more stable on the balance level) but ultimately un-robust.
Prioritizing robustness, we were curious about how Hayek Money would actually behave in the market and what it might be useful for. This led to our synthetic commodity money paper.
Question 11:
I understand scalability with DAI because supply is not a function of demand for DAI, rather demand for credit, but why can’t it go off the USD peg?
Evan:
I think USD collateralized stablecoins are hard to scale. My problem with DAI is that it isn’t robust unless it has stable collateral.
Question 12:
For next generation algorithmic stablecoins that I have read about, they have a market oracle price, and mint new coins until the price is back to the peg. Are you saying that something like this will end up overinflating the supply because traders will not automatically adjust their bids based on a supply increase?
Evan:
Have you seen our movement pattern paper? The underlying assumption there is that supply can respond to demand “algorithmically” but needs to be propagated back into price “behaviorally”
On an AMM these price corrections might be instant, but this is more the exception than the rule. It won’t be instant on matching engines.
I would say that it’s tempting for someone new to economics, to want to simply rebase more frequently, but that can actually be a bad idea and result in over-correction.
Question 13:
With the universal vault strategy, where I can stake LP’s into multiple programs, does this mean I can stake an LP of say $100 of AMPL / 100$ of ETH into the Sushi Geyser and the Beehive Geyser and the same time?
Brandon:
No, but you could stake into two Beehive, two Sushi, or two different programs that use the same LP token at the same time. A program just specifies which token to lock, that’s it. If you were in both Sushi and Beehive, both LPs would live in the same vault. Other programs could incentivize our vaults without even needed permission, since the architecture is general.
We’ll have more documentation that should make this much clearer closer to launch.
Question 14:
Would it be possible to have an office hours on Youtube for more visibility one day?
Brandon:
Yeah, we were talking about that just recently. I think it’s a great idea, hopefully soon
Simon:
Let’s use this form to source questions for a video Q &A. Anyone who has a question feel free to post it here, will work through them on video to supplement office hours sessions: https://forms.gle/hYyrhmga2BhDuuE68 Will release a new form after the session airs.
