C.R.E.A.M Finance is a permissionless, transparent, and non-custodial decentralized lending platform for both individuals and protocols to gain access to financial services. Currently, C.R.E.A.M. is live on Ethereum, Binance Smart Chain, Fantom and Arbitrum.
C.R.E.A.M Finance is a permissionless, transparent, and non-custodial decentralized lending platform for both individuals and protocols to gain access to financial services. Currently, C.R.E.A.M. is live on Ethereum, Binance Smart Chain, Fantom and Arbitrum. Users can lend any supported assets (including interest-bearing tokens) on the platform and use the assets as collateral to take out a loan in exchange for crTokens. The Iron Bank for protocols is part of the yearn finance ecosystem. C.R.E.A.M. also supports Flash Loans on their money markets, similar to AAVE v1 Flash Loans, except they are implemented on crTokens.
3 Key Highlights
CREAM Finance is blockchain agnostic and allows not only individuals but protocols as well to leverage on its lending and borrowing platform.
The Iron Bank features a smaller selection of assets that are deemed to have higher creditworthiness and funds are in a pool separate from those in CREAM v1.
Introduction: What are Algorithmic Stablecoins (Algo Stablecoin)
Algo stablecoin is a type of stablecoin that uses algorithms to change the circulating supply of an asset so as to get a stable value of the asset. In general, it creates more coins (supply) when the prices are above the pegged value (e.g. $1) and reduces supply when price falls below the pegged value.
In case of a price increase (e.g. >$1) the system is triggered to encourage an increase in the supply of stablecoins. In case of a price drop (e.g. <$1), monetary policy encourages a decrease in the supply of stablecoins. It all happens based on algorithms, which are mathematical models with adjustable parameters, executed by smart contracts. Some of these parameters can be adjusted through governance or other mechanisms, such as the issuance of coupons or bonds.
Types of Algo Stablecoin Mechanisms
There are two main types of mechanism design: single-token or multiple-tokens.
The algo stablecoins following a 1 token model operate with a simple rebase mechanism where tokens are printed and issued by the protocol when the price of the token is above a certain threshold. The undisputed pioneer of the elastic token concept is the original Basis dated 2018. We will see that several tokens in our analysis borrowed some principles from it.
The dynamics work as follows. The token price will be pegged at $1, with three cases of rebase happening every time:
The token price increases by more than $1, the total supply will increase.
The token price drops by more than $1, the total supply will decrease.
The unchanged token price remains around $1, the total supply remains the same.
When the total supply changes, it will directly affect the quantity of tokens that the holders have. The quantity can increase or decrease in proportion to the rebase result.
The standout project and also a leader for the algo stablecoin trend is Ampleforth (AMPL) followed by Yam Finance V1 (YAM).
The mechanism of the multiple-token model uses a similar model to the single-token model since the token price is still pegged at $1. However, it adds another token as part of its stability mechanism. The additional token(s) can be used in many forms. Rebase still has three cases:
Token price per $1, create more tokens.
Token price is less than $1, create debt in the form of bonds.
The token price remains stable at $1, no more tokens minted.
Projects under this model can have up to three tokens including:
Shares tokens, meant as a value accrual tool for shareholders.
Bond (or coupon) tokens representing the debt in the protocol.
Unlike the single-token model, the multiple-token model does not distribute new tokens equally to all holders. Instead, they are distributed at different rates to the different types of stakeholders like shareholders, liquidity providers, bond token holders.
Projects under this model include Basis Cash (BAC, BAS, BAB), Debaseconomics (DEBASE, DEGOV, D-BILL), Empty Set Dollar (ESD) and its fork Dynamic Set Dollar (DSD).
Metric discussion and why these metrics
We used a general economics design model to analyse the fundamental economics between the various token mechanisms and designs. This approach looks at the aspects of market design (digital environment), mechanism design (structural rules of the environment) and token design (mechanism and incentive policies of the token).
We analysed the various metrics within the framework and shortlisted a few metrics. These metrics are deemed the most important metrics that affect the robustness of the design for algo stablecoins.
Market design is the environment in which the participants and tokens exist. It helps to constrain the digital environment. This is important as market design defines the structural limits of the effectiveness of the mechanism.
[Market Design] Number of DeFi Protocols Accepting the Token (Integration)
Integration, and in a broader meaning, adoption, is possibly the most important metric to assess the success of a stablecoin in terms of its assimilation in the overall ecosystem. Digital ecosystems are heavily affected by network effects; the number of DeFi protocols accepting the token is therefore a suitable proxy to determine the network effects of the market.
Based on the research and analysis, Frax has so far accomplished the largest adoption and endorsement from the overall DeFi ecosystem. However, it is also worth noting that many protocols are undergoing upgrades, testing for robustness in their current monetary policy mechanisms and establishing their internal communities before expanding and working with a greater range of other protocols.
[Market Design]Introduction to Quantitative Metrics
As discussed above, market design is the digital environment that the tokens exist in. This environment exists beyond the protocol itself, and includes its acceptance and usability in other protocols. Thus, in the qualitative metrics, we analyse the token with respect to the community holding and using it.
[Market Design] Total supply
The total supply on exchanges includes both centralised exchanges (CEX) and decentralised exchanges (DEX). In general, both types of exchanges help to facilitate the buying and selling of tokens, hence adoption. However, supply available provides insight into the type of users holding the tokens and the liquidity available on decentralised exchanges, required by the AMM mechanism. In general, a high number of token supply available on CEX suggests that token holders are more verified, as KYC verification is required on CEX. A high number of token supply on DEX is not surprising as token deposit is required to provide liquidity for the market. However, it may also vary widely based on the rebalancing mechanism used such as rebasing, coupon model or using DEX as an oracle to rebalance to the peg value, US$1.
[Market Design] % Address with Holdings in the Top 15%
In a decentralised ecosystem, it is important to measure how decentralised the systems are. To do that, we look at wallets holding the top 15% of total token supply. If the top 15% of tokens are distributed within a small percentage of users, the protocol might be less decentralised than is ideal. Most tokens are held in the token contract or decentralised exchanges as liquidity providers. FRAX and Reserve have tokens in 3 Uniswap pools as liquidity providers. Ampleforth and Debaseconomics have tokens in the token contract. ESD and DSD have tokens in DAOs. Basis Cash has tokens mainly in its token contract and Uniswap.
[Market Design] % of Tokens in Daily Transaction
A feature more unique to stablecoins in general is its daily transaction volume. Unlike other tokens, a stable coin’s purpose of stability is to enable p2p trade and transactions. Thus, it is important to understand the average transaction volume by percentage of total available supply. The higher the number, the better the network effects, hence market design, of the token.
[Market Design] Steps to Acquire Stable Token and Other Tokens
When it comes to network effects and the community holding the token, it is important to analyse how new users can join this digital environment. The easier to get hold of the tokens, which qualifies a new positive change in the network, the better it is.
Simply put, mechanism design comprises the rules of the game, which participants have to follow.
[Mechanism Design] How are Parameters Changed? (i.e. Reward Mechanism)
This crucial metric falls under the broader umbrella of governance. Who can change the rules and how. This is important, especially when it comes to algo stablecoins. Whilst there are algorithms and the robustness of mathematical formulas governing it, the responsibility of decision making lies with the people. These people, who are part of governance, define the changing of parameters to alter the outcome of the algorithm
Most of the protocols have endorsed an open and transparent policy around the evolution of their projects. Communities are incentivised to participate in distributed governance, which is, for the vast majority of the protocols in analysis, the tool for introducing changes in the protocols themselves.
[Mechanism Design] What Parameters Can Be Changed?
We analysed what in the system is open to modifications, and why.
The results of this metric are indeed interesting and they vary across the spectrum. Some protocols allow for few changes to be made by governance while others allow maximum flexibility for the owners of the network to implement changes. We appreciate the approach of Frax the most, followed by Reserve and Basis Cash.
[Mechanism Design] Peg Verification and Oracle Used
Algo stablecoins need to retrieve their benchmark value in order to trigger the rebalancing of the system’s mechanisms. This metric studies the different approaches and options used by the protocols.
Protocols do not diverge much across the peg verification methods. Chainlink and Uniswap are the most recurring oracles. We believe Frax has a very comprehensive system in place compared to others.
Protocols often rely on external agents to come and interact with the system’s platforms in order to profit from available and desiderable arbitrages. This metric analyses the different opportunities provided to the so-called “arbitrageurs”.
This metric shows the versatility and complexity across designs. Some protocols suggest easy and intuitive arbitrages, like Luna and Frax. Reserve also offers a pretty simple arbitrage opportunity. Arbitrage on ESD, DSD, Basis Cash and Debase, on the other hand, require much more premeditation and financial acumen. We will dig into the risks involved in these arbitrage opportunities later in this report.
Token design is the rules embedded within the token, that the token has to follow.
[Token Design]Dynamics and Stimulus if Above or Below $1
This is one of the most informative and illuminating metrics of the whole research. For each protocol we study the mechanisms set in place to influence the stability of the system. This is mostly done through incentivising external agents to interact with the protocol in exchange for prospective economic benefits.
These above are the dynamics and the stimuli across the systems.
[Token Design] How Efficient is the Mechanism to be Pegged at $1?
A quantitative perspective to measure the efficacy of the protocol.
More specifically, we looked at a few specific metrics to answer the question.
[Token Design] Number of Days Above and Below $1 (±5%)
Firstly, a measure on the percentage of days the coin is above or below the pegged value within ±5%. Given the boundaries, a 0.95 and 1.05 is considered stable. That allows the calculation of days where the value is at peg. The smaller the number, the better.
[Token Design] Days Off $1 Peg
Whilst the previous metrics are sufficient to understand the effectiveness of peg, they are not representative should the project be very new or have been in existence for a while. Thus, a nominal number adds depth to the percentages above.
[Token Design] Average Days Unpegged
The number of consecutive days unpegged is another measure of robustness of the mechanism. It suggests the average days of “doom spirals”. Due to the nature of a self-sustaining internal economy, the “doom spirals” is a critical measure to understanding the robustness and anti-fragility of a mechanism. The less days, the better.
[Token Design] Average number of consecutive days pegged
If the metrics above describe the measures taken by the protocol to repeg its stablecoin, with this metric we compute the efficacy of these measures. The metric represents the average time length the protocol has historically needed to reestablish its peg. The lower the average time a protocol takes to repeg, the more effective we can consider its mechanisms. Furthermore, a protocol with a lower average time to repeg represents a more usable stablecoin, a better product for the end users.
[Token Design] Where is the Value of Inflation Coming From?
When it comes to the creation of new tokens within the system, that is monetary inflation. Since the tokens are pegged to a fixed rate (e.g. 1 USD), any monetary inflation needs to be substantiated with assets or formulas, to maintain the $1 peg. Since there is no underlying asset backing the system, the main method is to change the token supply.
Our preferences back the results shown above on the quantitative metrics: the category of projects that is backed by stable collateral plus a fraction of the protocol secondary token is the most sustainable model to date.
[Token Design]Risk management level of policies and risk mitigation policies
Here we study the number of levels of risk management methods set in place to decrease the malfunctioning of the protocols. We have tried to qualitatively study the effectiveness of the risk mitigation policies in place, drawing a conclusion on what could be the best approach in terms of safety.
Quantifying a qualitative metric is challenging. Our consensus voting results in Reserve being the most robust when it comes to risk management. At the end, the general framework in risk management is about diversification. Diversifying assets, risks or policies to maintain the peg, should an external activity affect the stability.
A notable mention is LUNA. As long as there is demand in the ecosystem through Luna’s value (as a native token) or Terra’s transactions (as retail users use it), the exchange of both assets preserves the stability. Whilst it is risky, Terra protocols tap into retail demand (pushing Terra’s transactions) to reduce the onchain volatility risk.
[Token Design] Instruments in the Secondary Market and Returns on More Risky Investment Vehicles (e.g. Coupons Tradable in a Secondary Market)
This is an analysis across the additional financial instruments added by the protocols to improve the rebalancing of their systems. This metric also analyses how profitable the additional financial instruments introduced by the protocols have been in the past. It is an additional clarification shedding light on the supplementary incentives for external agents to participate in the systems.
A secondary token is not mandatory for a stable coin token model. It depends very much on how the secondary token is being introduced and how it is part of the overall mechanism design.
[Token Design] New Inflationary Amount – Who is Entitled to it?
Inflationary amounts refers to the monetary inflation by the protocol as a result of introducing new supply into the market. This amount is different when the token has just launched or when the token is more stable and the inflationary amount is used to maintain the peg.
[Token Design] Value Accrual on the Secondary Token
We could not end our metrics without a mention of value accrual. Whilst this is a topic worthy of its own standalone research, the analysis of secondary tokens here is to understand the value capture of the token in a different accounting model. Tokens can be a means of internal accounting and calculation, or a way to distribute governance of the protocol. They can be used in calculation of liquidation priority (via coupons or bonds) or through governance by having the responsibility to change various parameters.
The main goal and effort of this research is to understand which parts of the different designs are most affecting the final scenarios we are seeing playing out in the algo stablecoins in production. To do this we have kept the framework consistent with previous research, splitting the metrics into three main pillars: market design, mechanism design and token design.
As we discuss the economics of how these systems are designed, it is important to begin with the type of risks in the system. These risks are: economic exploits, price volatility and technical risks.
We refer to economic exploits as the misalignment of economic design to allow one party to exploit another.
The most explicit issue that emerged was a misalignment between participants and stability. Often, participants face economic decisions which benefit the individual at the expense of the overall economic proficiency of the protocol affecting the stability of the system. This is a collective behavioural exploit that can be further analysed when building such systems.
Whilst the models of coupons and bonds are novel and innovative, the coupon is not a financial instrument that helps with the sustainability of a platform
(seemetric Dynamic and Stimulus if Above or Below $1)
Reallocation of Risks to Other Agents
We also found evidence that risk is often shifted from one segment of participants to another, creating negative vicious cycles. For example, ESD and DSD’s economic models are designed to encourage users to delay their returns. When the protocol needs a contractionary monetary policy it tries to reduce the supply of the stablecoin by issuing coupons, which are bought with the stablecoin token and this reduces their supply in the market. These coupons are risky investment vehicles as they might expire without paying out anything (if the price of the stablecoin remains under 1$ for a sufficiently long time). When the protocol is under distress the riskiness of these instruments might be seen as not worth the premium, rendering monetary policy ineffective.
(seemetric Dynamic and Stimulus if Above or Below $1)
Negative Secondary Feedback Effects
Another crucial highlight of our research is that often participants are indeed not incentivised to maintain the system stability. This is a critical mechanism flaw, as stability should be the central point of incentive engineering. Due to the circularity nature of closed looped systems, secondary feedback effects need to be considered. Some incentives do not lead to stable outcomes, instead shifting the risks from one agent to another, as discussed in the point above.
We noted that in collateral liquidations, some stablecoin holders could be liquidated at par for the collateral asset as opposed to at a floating market price. This eliminates the feedback effect on the stablecoin market price, reducing deleveraging risk on risk absorbers. However, the stablecoin may be less attractive to stablecoin holders as they now take on more liquidation risk. The type of stablecoin structure will also significantly affect incentives. When designs are more agent-based, agents have greater decision flexibility and are more likely to find a profitable participation level. In comparison, when designs are more algorithmic and/or with equity risk absorption, agents are more restricted and may be less likely to participate in the system relative to alternatives.
One interesting and negative consequence resulting from economic misalignment and incentives is the allocation of tokens for early adopters who then amass large quantities of tokens they can later use to rig the market in their favour.
At the launch of the protocol, tokens are issued to the community as part of distribution and initial inflation. The initial inflation asymmetrically benefits some parties more than others, resulting in an unfair distribution of tokens.
Another factor to consider is the voting power of early, large token holders, if the protocol has a DAO structure. They would hold significantly more power than anyone coming on later, which heavily disincentivises active participation and voting by smaller community members. This shuts off the flow of new ideas and could mean trouble for protocols with a quorum requirement. Whilst we do not have enough data to run any empirical analysis on this topic, this is something worthy for further exploration in future studies.
Sustainability of Incentives and Design: Coupon Failures
Misalignment can often trigger economic cascades resulting in negative effects for certain groups of participants.
The incentives designed to improve system stability have often failed, such as in the cases of coupons and bonds. What was intended to be a stability driver became instead a speculation vehicle: we have observed that traders will indeed wait for a price to decrease before making a purchase, trying to maximise their risk/return. This phenomenon, consequently, exacerbates price volatility.
One also cannot rule out the possibility that investors who are large enough to move the market can force the downward price pressure to purchase more coupons, then force it higher to make a profit.
Again, when tokens are above peg, we noticed that incentives are set in a way that favours instability more than consistency. This is particularly true for ESD, DSD, Basis Cash and Debaseconomics. Expiring debt with an often discretionary premium creates unnecessary risk for the agents. We believe the binary outcome is also psychologically detrimental to investors.
(seemetric Dynamic and Stimulus if Above or Below $1)
System Mechanism: Speed of Reactions and Feedback Loops
While reviewing the metrics, we asked ourselves the following question: are these systems reactive enough to market forces? Is the system reacting in a timely manner to price evolutions observed in the markets? And of particular importance: are the incentives designed in a way to boost those reactions?
We found the answers to those questions to be rather disappointing for several projects. Again, solutions provided by the projects fall across a wide spectrum. However, Frax is quite dynamic in these system changes. Ampleforth and Debase introduce a smoothing factor across a period of time. On the opposite side of that spectrum, the incentives introduced by ESD and DSD lag for a conspicuous amount of time.
Price volatility is the risk of price movement. This is most important in stablecoin models, since stablecoins are meant to have zero or minimum price volatility.
Inflation in Tokens
Inflation in crypto is usually referred to as monetary inflation, the increase in token supply. This happens in a situation where the price of the token is above peg.
(seemetrics Quantitative Data, Token design)
In the situation where tokens are less than the pegged value, tokens are either burned or new value is injected into the model to sustain the difference. LUNA does this by having more transactions on the Terra blockchain, bringing in value to the system. Thus, resulting in greater demand for LUNA.
Price Sensitivity to Large Holders
In addition to sustainability of incentive mechanism design mentioned in the Economics Exploit discussion, one also cannot rule out the possibility of large investors or token holders forcing downward price pressure to purchase more coupons, then forcing it higher to make a profit.
Secondary Financial Instruments: are Participants Getting a Good Deal?
From our metrics and analysis, the short answer is no. While coupons and bonds are an intellectually stimulating solution added to the stablecoin realm, they also add tremendous complexity and idiosyncratic risk to any participant willing to interact with the protocol.
For example, the use of coupons in ESD and DSD has often incentivised the purchase of coupons once the price re-approaches peg and not when it deviates away, which is when contraction is most needed. We then repeatedly have the two extremes of a situation happening: traders either wait for the price to be closer to peg to enter a “safer trade” or wait until the price is very far away from peg, increasing their risk/return profile of the trade. Intuitively enough, both those phenomena increase volatility, jeopardising stability.
DSD for example, introduced bonds in case the token price is below the peg. The protocol aims to incentivise token holders to contract the token supply as doing so will help to get the spot price back towards the peg. The DAO then issues so-called coupons (debt) which are purchasable by users who burn their DSD (de facto reducing circulating supply). The incentivisation takes the form of the heavy discount you always purchase Coupons with.
The actual discount depends on the debt ratio (Debt/Circulating supply) in the system.
Once there is a positive rebase event, the DAO mints a programmatically sufficient amount of new DSD. In this case, an equal amount of coupons will be redeemable for DSD. Extremely important to note is that the coupons face an expiry of certain epochs after purchase. There is also a debt ratio currently capped 35% at the time of writing, which implies a max. premium of ≈46%. The risk for the coupon (bond) buyer is to buy coupons and never be able to redeem because the peg is below 0 throughout the life of the instrument.
Secondary Assets and Value Accrual
Substantial discussions have been entertained across the participants of this research on if and why an analysis of secondary assets (and their consequential value accrual) would be purely beneficial and not an additional, external risk to the stability of the system. In other words: is a dual token system actually valuable for the aim of reaching optimal system functionality?
Our research shows that a secondary asset alone is not good enough. However, a well designed and decentralised coordination tool (token) accruing all or some of the value created by the participants of the network would be extremely useful and additive to the design.
Comparing between these groups of protocols A) Luna, Reserve, Frax and B) ESD, DSD – the former has a secondary token to maintain the peg as a booster to the system whilst the latter accelerates the model to downward negative spirals.
(seemetric Dual Token, Speculation and Value Accrual)
Lastly, risk management is a very important topic that is often neglected. The truth is that there is no “end all and be all” economic model for stablecoin protocol. Thus, other than considering the internal model and primary market, it is essential to also consider different external risks and mitigate them from within the system.
Our research shows that not many protocols consider the risks beyond immediate risks like price volatility of the secondary token and price volatility of the pegged asset.
We refer to technical risks as smart contract bugs and conceivable hacks.
Oracles are a weak point in the systems, becoming a possible target for attacks. All protocols rely on oracles to execute rebalancing code. Since most rebalancing creates at least a small arbitrage opportunity, predicting the timing (if it is variable), direction and scale of the adjustment offers a profit motive for a malicious actor to game the oracle in some way.
AMM oracles, like the UNISWAP TWAP (time weighted average price), rely on high liquidity and an active market to keep people from artificially changing the price, combined with a delay in price transmission that is averaged over many periods. Despite all these safety measures, one could manipulate an oracle on a small token pair with enough capital and time. It is dubious, however, how profitable this attack would be, since a small token pair that is easier to manipulate usually also means a small protocol. The arbitrage opportunity generated from the attack wouldn’t yield major profit opportunities. In essence, the attack is likely “not worth it”.
If a protocol uses an off chain oracle (like their own website in the case of Ampleforth), they face normal cyber risks of a malicious actor taking control of the price feed. This would not require any major funds to execute, just a skilful hacker.
The discussion and analysis by the data led to a few recommendations in the three categories of risks.
Balancing Early Adopters and Price Stability
One method of bootstrapping a stablecoin ecosystem is to attract investors and users into the system. Unlike other token-based ecosystems where early users are rewarded with potential high price inflation, stable coins do not have this luxury. Thus, it is more challenging to bootstrap the community.
Whilst it is attractive to sell initial stable coins at a discount, it is important to balance between attracting early adopters and having a discount that attracts high speculation. The greater the discount, the more speculators the community attracts. This outcome of participants might do more harm than good in the system.
A good way is to focus on defining the right parties to be part of the community and targeting them specifically.
Dynamic Response to Price Actions
A big problem with central banks is time lag in response to market events. Algorithmic stablecoins have the advantage of being able to immediately, predictably and precisely react to changes in demand and other possible events. Due to the predictability and codification of rules such actions by the algorithm would be priced in immediately and would not yield the market shocks we often see with real central banks.
For this approach to work, the moving parts of algorithms have to be fine tuned to the actual market they operate in which requires more fluidity and dynamism on epochs and lockups. This is done in the early days of the protocol to reduce volatility and avoid massive “free” token dumps which destabilise the protocol for years to come. Supply changes are often disconnected from price movements and have proved to be too slow as a solution.
Protocols invented various ways to add value when the price of their token falls, some more successfully than others. The most likely scenario of peg deviation is the price being too low, meaning that either supply is too high or demand too low. Manipulating supply is certainly easier for protocols but, as we can see from the results of our peg analysis, it often does not work.
The alternative of increasing demand could be explored but there are more unknowns and more variables to account for. One way to increase demand is to obtain new partnerships or integrations that either drive demand on their own through added coin utility or just provide additional exposure to new users. While we may think that this strategy is unsustainable in the long run, once protocols reach a critical mass these external actions will be needed less and less.
Another strategy to increase demand is providing higher return securities or better yield farming opportunities. This strategy is risky and should be applied only if absolutely necessary. It has been very effective in the AMM space, even in the long run, where some protocols skyrocketed in popularity solely due to yield farming opportunities. The rewards should be locked for a certain period so as to not exacerbate the problem of high supply. This decreases the present value of the investment so the rate of return should be adjusted and based to reach a deflation-inflation equilibrium.
Black swan events can put the protocol under a lot of distress by inducing fire sales of the stablecoin. With an extremely high sell pressure the coin might start to lose its peg and create panic in the system, which in turn would create even more sell pressure.
In times when there is a lot of panic in the traditional financial system, markets are closed for some time in order to let the authorities intervene in the economy, and to let the investors “regain rationality”. Some alternative solutions in the crypto markets could be emergency shutdowns of the protocol through governance mechanisms. The emergency shutdown could be triggered by some prespecified events (like a price fall) or voted for by a quorum of the protocol token holders.
Another strategy to absorb losses from black swan events is to issue some investment instruments divided in seigniorage tranches. The funds from the sale of these instruments could be collected in a treasury for recovery during difficult situations. Losses would be absorbed by investors in order of juniority, reducing the value of their investment instruments.
Secondary Token to Promote Stability
Secondary tokens are mainly used for protocol governance and in some cases for active price stabilisation as well (Terra). There is, however, a risk of correlation between the secondary asset and stable asset.
To prevent this, the governance token should provide a return (under conditions) regardless of the accuracy of the peg at that moment in time, for example as compensation for voting on the protocol or stabilising it. The two should also have a strong token pair on AMMs and other exchanges to arbitrage away any inefficiencies (as these secondary tokens are often also bought directly from the protocol in ICOs or seasoned offerings).
Most of the discussion thus far focuses on supply side economics. For example, the technical risk and economic exploits are initiated from the mechanisms underlying the monetary policy regarding how token supply is determined. The price of a token is defined when the token supply meets the token demand. The token demand is dependent on the usability of the token.
Other than focusing on the token monetary policy, managing token demand is crucial. As evident in the market design discussed in the metric Integrations, having a robust demand is important. This can result from more protocols accepting the token and having high acceptance or from low correlated activities like non-crypto related demand. For instance, e-commerce demand in Terra. This reduces the risk of lowered demand, in case of black swan on-chain events.
This could be in terms of protocol to protocol partnerships, acceptance of stable tokens as collaterals in other protocols or stronger community acceptance in using the tokens. There needs to be a strong reason to hold the tokens: and since it is less of an investment opportunity due to the stability in price, it needs to come from the utility and usability of the token itself.
The oracle problem has been discussed extensively by the community and there are a few general pointers every protocol should follow to maximise safety, stability, accuracy and decentralisation. Most protocols use just a single AMM integrated oracle, namely the UNISWAP TWAP oracle. This does make sense, given the limited representation on centralised exchanges but it does present a single point of failure. Oracles are so integral to the functioning of stablecoin protocols that they cannot afford such a risk.
To counteract this there are a few possible strategies they could follow:
Establish token pairs on other AMMs (SUSHI, 1INCH) and get price data from there as well, using Chainlink.
Establish more variety of token pairs, not just with ETH. They would be balanced by arbitrage and reduce gamability, easy to integrate.
Use innovative oracle solutions like UMA priceless oracle, especially for protocols with longer epoch durations (24 hours +).
Secondary tokens can be used to validate data for external oracle information. Tokens are staked and can be removed when data is inaccurate.
Lots of people are not keen on adopting algo stablecoins. But we are, hence this report. However, why should we be bullish on algo stablecoins? With algo stable coins, we can experiment with these designs and study the empirical evidence of their efficiency. Hence, we can build better monetary models in the future.
According to our research based on the current public plans as of March 2021, FRAX comes out as the best protocol with the highest overall score.
Our findings stated clearly that so far, purely algorithmic stablecoins have failed to deliver the bare minimum solidity and efficiency needed to attract meaningful adoption. Frax, in adopting its hybrid approach, has been able to maintain a solid peg (seemetric Max Number of Consecutive days pegged) devising a clever and effective system to stabilise the protocol (seemetric Dynamics and Stimulus if Above or Below Peg) and providing a compelling arbitrage opportunity for participants (seemetric Stabilisation Method). The results are stronger adoption compared to peers (seemetric Number of DeFi Protocols Accepting the Token).
Algo stablecoin is still a new mechanism that is being explored. We have experimented with coupon based mechanisms that do not seem to work in holding its peg. Pure algo stablecoin with rebasing might have some potential. But it seems like the best bet so far is fractional collateralisation with algo mechanisms to maintain the peg.
Appendix A: Update APRIL 2021
Ampleforth is integrating with AAVE to offer lending and borrowing in AMPL. It also launched a token, $FORTH, a new governance token.
Empty Set Dollar
ESD is transitioning from V1 to V1.5. First and foremost, the protocol is changing from functioning with just one token, to having a two token system (ESD as the actual stablecoin and ESDS as a seigniorage token).
The stablecoin will be backed 100% by usdc at the start. The team plans to include other available assets to back the coin and/or change the backing requirements. The DAO will play a major role in stabilising ESD through market operations involving both USDC and ESDS.
ESD is transitioning away from an algorithmic model to a reserve model.
Debase is changing the model drastically to create a dynamic monetary policy with fractional reserve, and supporting prices through profits earned rather than supply changes. This mimics the way traditional central banks work. The difference is that time lag is now reduced. Stabiliser pool 4 is under development to create a general monetary policy and future improvements can be voted on to propose new models for monetary policy. It is dynamically managed, with a focus on risk management. A new stable coin will also be introduced.
Dynamic Set Dollar
DSD publicly released the details of a V2 in early April. The new version re-designs the whole incentives scheme, replacing the debt and coupon system with a surrogate set of contraction incentives. The new introduction is a token called CDSD which will be obtainable by burning free-floating DSD. CDSD will be freely tradable and transferable without expiry. Additionally, users can bond CDSD into the DAO, in order to receive contraction rewards.
During contractions, rewards will be capped on a per-account basis (i.e. 100% of the DSD amount burned by a single participant). The aim is to have participants looking to maximise their contraction rewards, which will happen by continuously buying and burning more DSD. Finally an incentivised pool will converge liquidity towards the pair CDSD-USDC, avoiding DSD being the CDSD exit route in times of contraction.
Basis cash is undergoing a solid migration towards V2 aimed to dynamically target the emissions and quickly respond to changes in market conditions. The emission is proposed to change in contraction times to 60%, dynamically allocated between BAC/DAI and the BAC Vault, 15% to BAS Boardroom stakers, 10% to BAS/DAI, 5% to a BAB staking pool, 5% to Strategic Pairs and 5% to the CDF/Vision Fund. This latest percentage aims to build a reserve for future buyback and burn/incentives and for new partnerships. In expansion times, 60% will be dynamically allocated between BAC/DAI and the BAC Vault, 15% to the CDF/Vision Fund, 10% to BAS/DAI, 10% to Strategic Pairs, and 5% to BAS Boardroom stakers.
Terra, Reserve, Frax
There are no known updates as of April 2021.
Appendix B: Intro to Protocols
In this report, we analysed 8 different types of algorithmic stable coins.
Empty set dollar
Dynamic Set Dollar
What is Terra?
Terra is a blockchain project for the e-commerce sector. The project creates their own stablecoin to pay and reduce transaction fees.
TERRA Token used for:
Pay for goods or services provided by e-commerce floors and pay rewards for miners
LUNA Token used for:
Using LUNA to stabilise the price of Terra Stable Coin.
TERRA stablecoins will peg 1:1 to the world’s top currencies
Can be swapped between Terra Stablecoins
The value of each type of Terra stablecoin is determined by miners
Using algorithms to change the supply of cryptocurrencies in the Terra Family. At the same time, the reserve fund from LUNA Token will also be used to stabilise the price.
Terra Protocol uses Tendermint’s PoS consensus mechanism. In this mechanism, miners must own LUNA Token and stake in the system which presents The power of Producers Block. In addition, LUNA has the function of stabilising prices for the Terra Family.
What is the Reserve?
Reserve is a Blockchain project that provides a payment solution by issuing Stablecoin.
Reserve Protocol can be deployed on the Top Smart Contract of any Blockchain.
Three token mechanisms help stabilise the price of RSV Stablecoin.
RSV is the Stablecoin of Reserve used in daily payment.
RSR is a token used with the purpose of helping RSV stabilise the price.
Collateral Tokens are tokens of tokenised assets. Collateral Tokens are responsible for backing up the value of the RSV Token.
Reserve Protocol consists of two pools:
Pool 1 is a Reserve worth $1.
Pool 2 is the Vault corresponding to collateral tokens. The aim is to always ensure (at least) that Reserve is 100% collateralised.
Highlight (Stabilisation Mechanism)
RSV < $1: Reserve Manager (RM) will use Vault to buy and burn RSV
RSV >$1: Auction RSV which is priced $1.
If there is still RSV, RM will use RSR to buy RSV and put it into circulation.
If not, RM will print more RSV and sell them to Vault, then put RSV into circulation.
Dynamic Set Dollar
What is Dynamic Set Dollar?
Dynamic Set Dollar is a fully decentralised stablecoin designed to be highly capital efficient and more dynamic than Empty Set Dollar, with faster epochs, extended supply caps and modified extension/contraction formula.
Basically this project is similar to ESD, and has changed the following parameters:
Short epoch duration (2hrs)
Not limit rebase amount
DSD uses an elastic supply mechanic:
DSD < $1: DAO issues debt and can be bought by DSD. This decreases supply.
DSD > $1: DAO mints new DSD and remaining debt will be cleared. If DSD remains after paying all debt, the DSD bond holders and LPs will be rewarded.
What is Ampleforth?
Ampleforth is an algorithmic stablecoin functioning on the rebase mechanism. It uses a native Ampleforth oracle and Chainlink for price data and then actively adjusts the number of AMPL inside users’ wallets to keep up with demand and supply and maintain the peg. While they do maintain some peg stability, the accuracy of the peg is not exact.
Rebase mechanism entirely done through smart contracts:
24 hour epochs
Insurance for LPs
Single token design with no coupons
Every 24h a rebase is executed, the price is then returned to peg by arbitrageurs on DEXes.
Supply changes will be gradual.
Governance by foundation, registered with SEC.
With Ampleforth Foundation multisig all the code can be updated.
Empty Set Dollar
What is ESD?
Empty Set Dollar (ESD) is an algorithmic stablecoin which has been developed with the goal of becoming a reserve currency in DeFi.
ESD has three properties:
Stability: ESD maintains price stability around $1 through supply and demand elasticity with Rebase algorithm.
Composability: Compatibility with other DEFI protocols integrating ESD.
Decentralisation: ESD is fully concentrated from the first day of launch.
Highlight (Rebase Mechanism)
The amount of ESD is expanded or contracted through the process called Rebase based on the TWAP.
TWAP price > $1.05, ESD will be released further.
TWAP price < $0.95, ESD will issue debt as a coupon.
What is Debaseconomics?
Debase is an algorithmic stablecoin similar to Ampleforth. It stabilises the currency by proposing stabilisation groups that can be programmed to try to encourage DEBASE holders to stabilise token prices over a number of cycles, in an intermediary managed process.
Assuming the initial parameters, the rebase function of Uniswap’s Oracle Debaseconomics protocol queries the price of Debase against the DAI every 24 hours.
Assuming the current parameters for $DEBASE, if the price difference between the original DAI and DEBASE is more than (+/-) 5% in both directions, this will re-enable the base.
Rebase only is executed one time every 24hrs or the peg price is deviated by 5% (whichever comes first).
Via governancer, holders can adjust parameters.
Supply changes will be gradual to manage behavioural lag in participants.
What is Basis Cash?
BAC is an algorithmic stablecoin where its value is designed to cling to the value of the stablecoin DAI, equivalent to $1.
Basis Cash uses the central bank model to be operated by an algorithm to balance stablecoin supply compared to the change of demand to ensure that the value of Token is stable.
The price of BAC is maintained thanks to the protocol which operates two policies to contract supply and expand the supply.
BAC < $1, users can use BAC to buy BAB whose price is (price of BAC)^2 (cheaper)
BAC > $1, protocol will issue more BAC via BAS holders.
What is Frax?
Frax is the first segmented algorithmic stablecoin protocol with the goal of providing highly scalable and decentralised cryptocurrencies.
In the Frax protocol, there are two tokens, FRAX (stablecoin) and FXS (Frax Shares).
FRAX: there is price stability through collateral (both assets and FXS). The collateral rate changes per hour according to the value of FRAX with a step of 0.25%. Currently, the asset used as collateral to create FRAX only has stablecoins such as USDT, USDC to create stability for the protocol before allowing collateral with many other types of assets. The total value of the FRAX minted must be equal to the total value of the collateral at the collateral rate and the value of the FXS (will be burned). When FRAX is redeemed, users will receive back the collateral and the amount of FXS at equal value (this FXS is mined).
FXS is a shared token in the Frax protocol, has unlimited aggregate supply and is primarily used for protocol administration.
FRAX is always minted / issued with value at $1. FRAX > $1, it creates benefits that encourage users to mint out more FRAX. FRAX < $1, there is benefit to FRAX redeem.
FXS token holders when staking will have the right to vote on decisions in the protocol such as:
Adjustment of collateral types.
Changes to the adjustment of fees for minting / redeeming FRAX.