Metastability mechanism

This is a write-up of a discussion I had with @hxrts, and only a sketch of a proposal.

This post describes a potential metastability mechanism.

Basic context:

  • Assume a principal token, which we will call A.
  • Assume an infinite set of secondary tokens, which we will refer to as T_0, T_1, etc.
  • Assume that the principal token can be traded against (swapped to and from) any secondary token on a market, with a ratio (price) A : T_{n} varying in time.

The goal of a metastability mechanism in general is to convert short-term volatility of A : T_n into long-term stability of A : T_n, with the following parameters controlled by governance of A:

  • Target “levels of stability” for each T_n, which can vary
    • Corollary: the ability of A governance to exit a “stability relationship” if desired
  • “Acceptable entry” thresholds for each T_n, which can vary

Many potential mechanisms could fit this bill. One could look roughly as following:

  • When the ratio A: T_n is above the current acceptable entry threshold, and the target level of stability for T_n is not yet reached:
    • Anyone can deposit T_n into the mechanism. Governance of A then creates new A in accordance with the current ratio A : T_n. Both tokens are deposited into an xy=k liquidity position. The user receives some corresponding “T_n deposit” token.
    • Fees from this LP are managed as follows:
      • Fees in A are burned
      • Fees in T_n are deposited into the special A-treasury account
    • The user could withdraw the position with their “T_n deposit” token.
      • Alternatively, the user could forfeit the right to withdraw (and perhaps receive some reward from A for forfeiting this right).
    • A might elect to reward the “T_n deposit” token in some fashion.

Holders of A-bonds can then obtain zero-interest loans of T_n, of an amount equal to bond’s proportion of the A-supply of T_0 in the A-treasury, which only need to be repaid if the holder wishes to redeem the bond. A bondholder could also burn the bond and receive all the proportional underlying floor token amounts (which is functionally equivalent to never repaying the zero-interest loans). This establishes a (floating) floor price for A.

Should A wish to exit a particular stability relationship (for T_i), the treasury amount of T_i could be gradually used to buy and burn A (this can be turned into a continuous management mechanism with target stability levels).

This needs a lot more modeling, but I’m pretty sure that the basic mechanism works (modulo users actually choosing to deposit T_n as described).

3 Likes

This is pretty exciting, I’m still chewing it. One question I’m having a hard time with is:

How can you have a stability mechanism if you don’t own the liquidity for all the A governance denominated token pairs. Because I can simply trade on a centralized exchange or in DEX pools not controlled by A governance and move the exchange rate - especially in bull or bear market phases where prices are moving quickly – sometimes prices never truly recover, they stabilize near zero.

3 Likes

In this proposal, the A-system itself (let’s suppose that there’s an A-consensus, for simplicity) can guarantee a floor price in the sense that a user who holds 1 A can always get the floor price’s worth of T_i (for whatever the floor price for T_i is) through this zero-interest loan mechanism.

Strictly speaking, nothing prevents someone from selling 1 A for less than the floor price in T_i, but that doesn’t make sense for them to do – so we can also reasonably expect that actors / arbitrageurs on secondary markets would have good reason to bid at least the sum of the floor prices for 1 A.

1 Like

If we leave aside the time dependency for a moment, this context reminds me of a paper[1] that is in the back of my mind for a long time because if I were to think of tokens as commodities (cf. tokenized xyz), just for the sake of illustration, then there is an obvious relation of the above basic context and the opening of the paper (small text can be skipped on a first reading)[2]

In the standard Shapley–Shubik strategic market game (Shapley, 1976; Shapley and Shubik, 1977), there are two commodities [\scriptstyle B, and \scriptstyle C], and any bid is a quantity physically sent to the market, and to be exchanged for the other commodity. So there is a unique price that will clear the markets—the ratio of the total amount of bids in each of the two commodities. All bids are exchanged according to this price system. When there are more than two commodities, one is reduced either to give a privileged role to one of them (“money” [A]), and set up separate exchanges for each of the other commodities [T_n] in terms of money [A], or to set up a separate exchange for every pair of commodities [T_m:T_n]—but then without any reason to obtain at the end a consistent system of prices.

So, I would guess that—at least in cases where the direct T_n:T_m-ratios (now) for a given finite (pre-existing) of k tokens \{T_1,...,T_k\}, “almost” give rise to a consistent system of prices, it make sense to consolidate the prices with such a A-token governance system.

So, I guess,[3] as a 10.000 feet above stratosphere TL;DR: A is the :pie: we want to grow while at the same time defragmenting liquidity (divided into T_n). One “just” hast to be careful price floors as described later by @cwgoes in response to @apriori do not cause issues.


  1. Limit-Price-Mechanism.pdf (732.2 KB), see also Jean-François Mertens - Wikipedia ↩︎

  2. Boldface, emphasis, and material in square brackets added by me ↩︎

  3. mental note: “should” read that paper … ↩︎

how can you enforce a floor price? this seems like a fundamental thing that is not possible in a permission less system.

for example, almost every single algo-stable coin that attempted to enforce a floor price of sorts or peg to USD failed.

how does this defragment liquidity?

so, yes, this is the point that may need further attention

That thought was relative to a setting of pre-existing tokens T_1,..., T_k, for which—so far—no A-token is established, and moreover (super strong assumption), there exists already a consistent system of prices.

Now, in this case, we can replace k(k-1)/2 CFMM-pool pairs, one for each unordered pair T_i,T_j, by k pairs, one for each [A,T_k], that’s probably better. Now, k(k-1)/2 > k holds for all values k>3, and for those values we would save on the number of pools. For this I was using a definition of liquidity de-fragmentation as reduction of the number of liquidity pools.

However, happy learn where this argument is broken. The liquidity was an afterthought, quick thinking, replacing a bunch of separate tokens into a single one.

1 Like

If you have tokens that can trade on different venues in a permissionless setting, then liquidity will be fragmented. If the different venues have different controllers, then you cannot get atomic compossibility, which means the assets will remain fragmented.

Imagine Token A is $XAN. If $XAN is listed on a centralized exchange, trades on Uniswap V2, V3, & V4 pools on Ethereum, trades in various AMMs on L2s, and trades on other non-EVM chains, your liquidity will be fragmented. You cannot control this. The same applies to any T_n tokens.

If you’re thinking about Anoma in a way that does not interact with the current blockchain-based system, your argument makes sense. In practice, though, this is not the case if we expect tokens to interact with existing forms of blockchain-based liquidity.

2 Likes

To be clear, this is not a proposal for a “stablecoin” in the sense in which crypto uses the term (which is typically to attempt to guarantee a fixed exchange rate w.r.t. a reference asset or index such as USD). If the word “metastable” is too confusing, we can consider another one, but note for contrast: nowhere in this system is any sort of index or price oracle.

No mechanism in a permissionless system can guarantee that someone holding an A-token will not sell it for less than any particular price, that’s true, but that’s not what I’m proposing – what I’m proposing is that the A-system itself guarantees that someone holding an A-token can always get at least some amount of other tokens, which is a guarantee that the A-system can make if it controls some other tokens, and this mechanism describes a process through which it would acquire that control. Does that make sense?

4 Likes

If you could maybe (briefly) explain by way of an example of what would this look like today if it was live on Ethereum. I have read the post several times, but its not clicking. Seems like a combination of

2 Likes

Thanks for writing this up Chris! I should give credit where credit is due—the mechanism I was describing was from Baseline markets. Definitely worth checking out their project!

@apriori You can think of the floor price as the backing of the paired asset. If that token has zero surplus value then the price of the asset should be equal to the backing. Here you’re enforcing this constraint because the token issuer is the provider of the primary market, where they take fees. The accrued fees increase the backing, at which point the protocol can place incrementally higher floor bids at the pro rata price.

3 Likes

As per request here are some resources on protocol owned liquidity. Including

  • relevant historical context,
  • current examples live in production,
  • as well as theory.

Driving Long-Term Protocol Growth Through Liquidity Mining and Protocol-Owned Liquidity (POL)

Protocol-Owned Liquidity (POL)

Protocol-Owned Liquidity, or POL, is a native derivative of Liquidity Mining and refers to the strategic acquisition and management of onchain liquidity by a protocol itself rather than relying solely on external liquidity providers. This approach offers protocols more direct control over a portion of their market’s liquidity, ensuring stability, reducing slippage, and generating revenue from trading fees.

Similar to Liquidity Mining, POL enables a protocol to kickstart trading activity or maintain a baseline level of liquidity without depending on external incentives to attract LPs. Unlike Liquidity Mining, where users are motivated to contribute liquidity in exchange for rewards, POL places the responsibility on the protocol itself to act as the LP, utilizing its own resources to ensure market liquidity.

DeFi liquidity management via Optimal Control: Ohm as a case study

As decentralized finance grows to autonomously managing hundreds of billions of
dollars of assets, capital efficiency has become an ever increasing component of protocol design. Recently, the Olympus protocol (also known as Ω) has utilized a novel liquidity provisioning mechanism that improves capital efficiency. This system introduces the concepts of a decentralized protocol renting, leasing, and buying liquidity when it is required for protocol functioning. In this note, we formalize the notions used by Ohm smart contracts in probabilistic and control theoretic terms. In particular, one can view the Ohm system as a stochastic non-linear control system. We show that the non-linear control mechanism is actually approximating the behavior of a simpler stochastic linear-quadratic regulator. We construct an associated Hamilton-Jacobi-Bellman equation for a mean-variance portfolio optimization problem, and show that the protocol can stabilize price by choosing appropriate portfolios. Our main result shows that the Ω protocol enjoys increasing ability to control price as the number of bond durations increases, but that this ability has diminishing marginal returns. Therefore, using this formalism, we show that with proper dynamic tuning and adjustments, the Ohm protocol can both improve capital efficiency and reduce risk to protocol users. We conclude by generalizing the Ohm controller model to a generic mechanism for optimizing risk and incentives in decentralized protocols, which includes other mechanisms like Tokemak and ve.

4.2 Interchain Allocator: Economic Coordination and LongTerm Alignment

Rebalancer

The Rebalancer system is a tool for the execution of third-party capital allocation strategies for liquid assets that takes the current portfolio, a target portfolio, and exchange policy as inputs. The Rebalancer periodically computes assets to be sold or acquired to incrementally move toward the target portfolio. The assets are sold or acquired according to the user’s desired policy, which could include gradual dutch auction or direct trade on a whitelisted exchange. Users of the Rebalancer can set parameters to strike the desired balance between urgency and slippage minimization on a portfolio or per-asset basis.

Field Guide to the Curve Wars: DeFi’s Fight for Liquidity

The “Curve Wars” centers around the fight for liquidity in DeFi, primarily focused on Curve Finance’s platform. At its core, protocols compete for veCRV (vote-escrowed CRV tokens) to direct liquidity to their pools, which was largely won by Convex Finance who “controls more liquidity than any other DEX.” Convex created “a liquid version of veCRV” called cvxCRV that lets users “earn Curve platform fees without having to lock your tokens for 4 years.” This led to the “Convex Wars” where protocols now either buy CVX tokens or “bribe” CVX holders to vote for their pools, with platforms like REDACTED Cartel and Tokemak emerging as new players in this “fight for liquidity” that is "just getting started.

Aerodrome

Aerodrome was originally a fork of solidly, the OG ve(3,3) token model.

What is Olympus?

The Olympus protocol is a decentralized financial (DeFi) system that supports OHM, a treasury backed, liquidity-enabling token on the Ethereum network. Olympus leverages the mechanisms of Protocol Owned Liquidity (POL), Range Bound Stability (RBS) and Cooler Loans to create a robust, flexible, censorship-resistant, and smart money.

The goal of Olympus is to build a programmatic policy-controlled money that:

  • Preserves purchasing power via long-term price predictability.
  • Maintains reliable liquidity across decentralized exchanges.
  • Is used as a unit of account (e.g., by being paired against many other decentralized assets)
  • Is utilized as a trusted asset (e.g., to collateralize other assets or deposited into protocols’ treasuries).
  • Is fully decentralized and controlled by the community
  • Is financially flexible, allowing users to borrow the backing against their money

Introducing: Tokemak | The Utility for Sustainable Liquidity

Tokemak creates sustainable DeFi liquidity and capital efficient markets through a convenient decentralized market making protocol." It works through “token reactors” where “Liquidity Providers (LPs) deposit assets” and “Liquidity Directors (LDs) utilize TOKE to control liquidity direction.” The protocol aims to replace traditional solutions like “engaging centralized market makers” and “yield farming” with a more sustainable model where “TOKE is the native network token that is earned through participation in the protocol” and can be used for “directing liquidity and governance.” The system operates in “Cycles” where assets are deployed and withdrawn, creating what they call a “black hole effect” that “will pull in value” over time

FRAX AMO Overview

The second (V2) expansion of FRAX focused on the idea of fractional-algorithmic stability by introducing the idea of the “Algorithmic Market Operations Controller” (AMO). An AMO module is an autonomous contract(s) that enacts arbitrary monetary policy so long as it does not change the FRAX price off its peg. This means that AMO controllers can perform open market operations algorithmically (as in the name), but they cannot arbitrarily mint FRAX out of thin air and break the peg. This keeps FRAX’s base layer stability mechanism pure and untouched, which has been the core of what makes our protocol special and inspired other smaller projects.

Berachain Proof of Liquidity

Proof-of-Liquidity (PoL) is a novel economic mechanism that uses network incentives to align the interests of ecosystem participants and bolster both application-layer and chain security.

3 Likes

Thanks @apriori.

I just want to note (as was orally discussed) for the sake of future readers that it should be possible to combine this long-term “metastability” mechanism with a short-term “volatility dampening” mechanism elegantly, where:

  • The short-term “volatility-dampening” mechanism determines how the A-governance manages its liquidity positions (POL), which could e.g. be concentrated more or less, placed at different price levels, etc. in order to encourage certain volatility dampening dynamics (and collect more fees). Perhaps this is not entirely dissimilar to how Tokemak works (but with different specific strategies).
  • The long-term “metastability” mechanism converts fees earned from those liquidity positions (and perhaps other sources) to create a floor price, zero-interest loans, etc. over time (as already described here).
1 Like

Just to clarify:
If anyone wants to take out a loan against A they get a portfolio containing proportional amounts of \{T_1, \dots, T_n\}?
If someone wants to have only certain T_i, they would need to swap the other tokens current exchange rates, but this should probably not happen internally if we want to maximize for stability.

If liquidity can be withdrawn per T_n by redeeming deposit tokens, liquidity providers can influence the shapes of portfolios anyone would receive in the loans (or upon exit).
If we want to prevent this, liquidity providers could just receive the A created for their deposit, exchanging their liquidity positions for shares of the pool proportional to the exchange rate of T_i : A at the time of liquidity provision.

We should definitely model/simulate this in more detail, since the multi asset case could very well introduce unwanted behaviors if we’re not careful.

I’m particularly interested in understanding the choice of DEX for this mechanism. Could you clarify whether the DEX will be native to Anoma, or if it will leverage the existing Interchain ecosystem, specifically the Interchain intent framework?

2 Likes

Welcome to the Anoma research forums.

The mechanism could be thought of as native to Anoma. In general, the mechanism could be used in the various ecosystems where Anoma is deployed.

Our first deployment will be on Ethereum. Eventually, we do intend to have a cosmos SDK protocol adaptor. We’ll update our roadmap in the near future, but for now, take a look at this post.

1 Like