IOSG: An Overview of Algorithmic Stablecoin Mechanism Design, Product Status, and Stability Performance

Original Title: "IOSG Industry Research | Decentralized Central Bank Currency Experiment"

Author: Ray Xiao, Investment Director at IOSG Ventures

1. Background Overview

Stablecoin Market Size

In terms of market size, general stablecoins (both decentralized & centralized collateral) are absolutely mainstream, while algorithmic stablecoins are a very niche product, holding only about 1% market share.

Industry Demand: Why Do We Need Algorithmic Stablecoins?

Collateralized stablecoins have been market-validated, and the model of stablecoins backed by collateral is the simplest and most natural way to achieve a stable peg to the target price. So, is there still demand for algorithmic stablecoins? Is there a future for them?

Many algorithmic stablecoin projects we have discussed mention a story: breaking free from central bank policy influence. However, we believe this is not the actual pain point. We think the underlying demand and potential value of algorithmic stablecoins lie in how to gradually achieve low collateralization rates or even no collateral under the current over-collateralized DeFi conditions, ultimately relying entirely on automated algorithms to change the money supply.

This involves maintaining the relative stability of the pegged price and market price based on supply and demand economic laws and rational economic behavior, making it still a very idealized monetary experiment.

Key Factors for Success of Algorithmic Stablecoins?

After studying the algorithmic stablecoin projects on the market, we believe the following four points are key factors determining whether an algorithmic stablecoin can succeed:

• Is there a collateral mechanism (partial collateral or completely uncollateralized)?
• Global Rebase (smart contract global adjustment of supply).
• Bond/stock issuance (seigniorage mechanism).
• Target price pegging or floating (strictly pegged to $1 or floating target price).

2. Industry Segmentation (Mechanism Design + Product Status)

First, we need to clarify what "algorithm" refers to in algorithmic stablecoins. We believe that "algorithm" in algorithmic stablecoins refers to a set of predetermined rules set by the protocol, adjusting the total supply of stablecoins based on the deviation of the market price of the stablecoins already issued by the protocol from the target price (i.e., Target Price).

This operates without human intervention, with the protocol running automatically, very much like an automated central bank composed of various rules and algorithms.

It is also worth mentioning that many people believe that "algorithmic stablecoins" are the opposite of "collateralized stablecoins (DAI, USDC, etc.)", but from a developmental perspective, there can also be some collateral in the algorithmic stablecoin system. In this thrilling monetary experiment, algorithmic stablecoins without any collateral backing have mostly experienced extreme volatility, making it practically difficult to instill a sense of "stability" in users.

Therefore, after ESD and BASIS, we can see projects like Terra, Frax, Fei, and RAI, which are algorithmic stablecoin projects backed by collateral. Although their stablecoin supply adjustments still rely entirely on automated algorithms, the presence of collateral backing each issued stablecoin provides relatively better system stability.

As mentioned above, we can see that various projects are adjusting the token supply based on the decoupling of stablecoin market prices and target prices, following the economic laws below:

Too much money in the market, oversupply, recover/destroy money in the market.

Too little money in the market, undersupply, increase money in the market.

After the issuance of stablecoins (target price, issuance volume), "market price/target price fluctuations (supply-demand imbalance) > smart contract global regulation/user spontaneous regulation > affect supply-demand, transmit to market price" summarizes almost all algorithmic stablecoins.

Although at a high level, everyone follows this economic law, the specific means of maintaining stablecoin prices vary widely among different projects, and we will categorize them into two major factions: uncollateralized and collateralized.

Uncollateralized Faction

The uncollateralized faction does not require any collateral to be deposited into the protocol during the issuance of currency, so we only discuss how the protocol controls inflation/deflation.

First Generation Rebase ------ Mandatory Increase/Decrease of Stablecoin Supply

When Rebase is mentioned, many people's first reaction is to think of Ampleforth and YAM. Although the concept of Rebase was not originally created by Ampleforth, it was indeed popularized by it (this mechanism was not well-known before), and Ampleforth even calls itself a "Rebasing Cryptocurrency".

Representative Project: Ampleforth

Regulation Principle:

Inflation: When the market price of the stablecoin < target price (currently about $1), the protocol will proportionally reduce the number of stablecoins held by all users in the system.

Deflation: When the market price of the stablecoin > target price (currently about $1), the protocol will proportionally increase the number of stablecoins held by all users in the system.

It is worth mentioning that the number of stablecoins in this system is adjusted based on the degree of decoupling between the market price and the target price (for example, 5%). Proportional increase or decrease means that although the number of stablecoins in users' hands increases or decreases daily, the corresponding proportion of the total issuance remains unchanged.

Value: The greatest value of Ampleforth lies in its bold design of Rebase------ "increase issuance when prices rise, destroy when prices fall" to regulate the currency issuance, then relying on arbitrageurs' arbitrage behavior to drive the currency price in the direction desired by the designers.

After Ampleforth, all improvements regarding algorithmic stablecoins have more or less revolved around Rebase (whether through smart contract global regulation or issuing protocol bonds and stocks to incentivize users to complete supply adjustments). Taking issuance increases as an example, after each Rebase increase, the number of stablecoins in each user's hands increases, and short-term arbitrageurs can arbitrage by selling stablecoins, ultimately causing the stablecoin price to drop.

Design Flaw: Ampleforth's method of regulating stablecoin supply is relatively direct and crude------ proportional increase/decrease, ignoring more complex factors (market environment, psychological factors, etc.). Theoretically, proportional splitting is merely cutting a cake and does not change the size of the cake.

However, in actual operation, external market factors and user psychology (the decrease in the number of assets in hand) can negatively impact the system, leading to significant price volatility of the stablecoin.

Second Generation Rebase ------ The First Introduction of Seigniorage Shares Practice

In first-generation Rebase projects, the only means of protocol regulation was the global adjustment of total supply by smart contracts (users would find that the number of stablecoins in their hands increased/decreased daily and could only passively accept it). Therefore, economists, seeing this design flaw, added new means------ seigniorage, marking the beginning of the second generation of Rebase.

The first to think of applying the traditional monetary economic concept of "seigniorage" to the crypto currency system was Robert Sams. In a related paper he wrote in 2014 (A Note on Cryptocurrency Stabilisation: Seigniorage Shares), he described a dual-token stablecoin model: one token is the stablecoin.

The other token is similar to equity that can share the protocol's seigniorage. For example, when the market price of the stablecoin is higher than the target price and the supply of stablecoins needs to be increased, the newly issued stablecoins will be distributed to "protocol shareholders."

Representative Projects: ESD, BASIS

Regulation Principle:

Inflation: When the market price of the stablecoin < target price (currently about $1), the protocol will issue bonds to buy back stablecoins from the market, thereby reducing the liquidity of stablecoins in the market, and users can earn interest by purchasing bonds (the Coupon in ESD and the Basis Bond in BASIS are both bonds issued by the protocol).

Deflation: When the market price of the stablecoin > target price (currently about $1), the protocol will issue more stablecoins, and the newly issued stablecoins will be given to users entitled to seigniorage (ESD is given to stablecoin stakers, BASIS is given to stakers of the protocol stock Basis Shares).

Value: Observant friends may notice the difference between the first and second generations of Rebase. That's right, the most important innovation is the introduction of protocol equity and bonds as two important macro-regulatory means. Although the second generation of Rebase still adjusts the total supply globally through smart contracts, it will only be used during deflation, and the newly issued stablecoins will be distributed to protocol stockholders to enjoy seigniorage.

In the face of inflation, the protocol creates arbitrage opportunities by selling stocks at a discount to incentivize users to actively destroy the stablecoins in their hands, which is a huge difference from the first generation of Rebase, where users passively accepted the increase or decrease in stablecoin supply. Subsequently, many algorithmic stablecoin projects have also introduced protocol bonds, which are used to buy back/destroy stablecoins during inflation, representing a significant improvement over the first generation's simple and crude "increase issuance when prices rise, destroy when prices fall."

Design Flaw: The uncollateralized faction believes that people's optimistic expectations about the system (that it can develop steadily, maintain price stability, and steadily increase supply) are sufficient to support a stable algorithmic system. However, practical situations have proven that expectations alone cannot lead to success. Although projects represented by ESD and BASIS have introduced the design of protocol stocks sharing seigniorage, their performance in terms of stablecoin price stability is far worse than that of collateralized projects.

Due to the lack of collateral backing the system, the net present value of these expectations is what supports the value of protocol stocks. From the current development of "uncollateralized faction" projects, we see that when stablecoin prices drop, users sell stocks, leading to the failure of protocol stock regulation and the collapse of the system.

Collateralized Faction

Currently, collateralized projects outperform uncollateralized projects in terms of both the stability of stablecoin prices and the valuation performance of governance tokens. The main reason is that collateral is deposited during currency issuance as a support for the value of stablecoins.

Regulation Method 1: Elastic Seigniorage Shares

The important aspect of the third generation of Rebased is the new design of the functions and supply changes of protocol stocks, enhancing the protective role of protocol stocks during stablecoin price fluctuations.

Representative Project: Terra (Although Terra states in its official white paper that its stablecoin TerraUSD has no collateral, we believe that LUNA plays a role as a partial collateral behind it, so we temporarily categorize Terra as collateralized.)

Regulation Principle: First, it is important to clarify that Terra is a Zone sidechain on Cosmos, which uses PoS consensus. Miners need to stake $LUNA to obtain the right to package transactions that occur on the chain. The Terra algorithmic stablecoin operates on the Terra chain, and $LUNA also serves as the equity token in the algorithmic stablecoin, having dual functions.

Inflation: When the market price of the stablecoin < target price (currently about $1), the system prints stocks & recovers money, allowing arbitrageurs to exchange stablecoins for newly issued stocks (for example, if the current market price of Terra stablecoin is $0.9, the system allows arbitrageurs to exchange a stablecoin worth $1 for a newly issued LUNA, effectively obtaining a new LUNA at a price $0.1 cheaper than the market price).

Deflation: When the market price of the stablecoin > target price (currently about $1), the system prints new money & recovers stocks, allowing arbitrageurs to exchange equity token LUNA for new money (for example, if the current market price of Terra stablecoin is $1.1, the system allows arbitrageurs to exchange a LUNA worth $1 for a Terra stablecoin, effectively obtaining a new Terra stablecoin at a price $0.1 cheaper than the market price).

Value: First, the design of Elastic Seigniorage Shares involves Rebase of the supply of protocol stocks and stablecoins during inflation and deflation, relying on arbitrageurs' behavior to pull back the decoupled stablecoin market price, effectively redesigning the supply changes of protocol stocks to provide better absorption capacity for stablecoin price fluctuations.

Second, in this case, $LUNA itself is also the native token of the sidechain, and the PoS staking scenario of the Terra chain actually endows $LUNA with new functional value. Therefore, if the Terra blockchain can develop, we believe that the algorithmic stablecoin project of Terra will be more stable in theory.

Design Flaw: Miners on the Terra chain need to bear the risk of stablecoin price fluctuations. During deflation, the system printing new stocks will dilute the computing power of Terra chain staking miners. However, currently, the price stability of Terra stablecoin ranks among the best in stablecoin projects (see section 3.1 later).

Regulation Method 2: Protocol Controlled Value (PCV)

Representative Project: Fei

Issuance: Stablecoin FEI is minted based on Bonding Curve, where users deposit ETH to obtain FEI. The ETH deposited by users actually enters the PCV-controlled treasury as an asset supporting the value of FEI.

Regulation:

PCV Regulation: The core idea is that the smart contract continuously adjusts the liquidity of the stablecoin trading pool issued by the protocol using the collateral in the treasury (Reweight), keeping the price in the stablecoin trading pool stable at the target price of $1.

User Penalty/Incentive: When the market price of the stablecoin < target price of $1, selling stablecoins will automatically incur a discount; buying stablecoins can earn additional stablecoin rewards.

Value: Using the treasury composed of collateral assets to control the liquidity pool of stablecoins in the market, PCV can be considered an interesting idea related to algorithmic stablecoins that emerged after the advent of AMM liquidity pools.

Design Flaw: The project team uses smart contracts to automate the regulation of collateral in the protocol against market prices, which is effective in most cases. However, if faced with extreme market conditions where the value of collateral drops significantly (currently, FEI's collateral is ETH), this regulatory mechanism will face serious challenges, which is a common issue for all low-collateral DeFi products.

Regulation Method 3: Introducing Engineering PID Control System to Adjust the Ratio of Target Price and Market Price

Representative Project: Reflexer

Issuance: Users need to over-collateralize ETH (collateralization rate not less than 145%), and the system also has a liquidation mechanism.

Regulation:

When the market price (assumed to be 4) > target price (currently $3.14), the control mechanism is activated (using PID control system), continuously increasing the redemption rate of RAI (the ratio of target price to market price), effectively raising the target price. The increase in redemption rate will incentivize those locking ETH (RAI borrowers) to buy RAI now because their liquidation costs have increased. Theoretically, this will align the market price of RAI with the redemption price.

Value: Reflexer’s system is relatively robust, inheriting the collateral and liquidation mechanisms of MakerDAO, and the collateral can be redeemed at any time (unlike Fei, which cannot directly redeem collateral). The novelty of this project lies in using the PID algorithm to control the target price and interest, thereby changing the market price based on arbitrageurs' arbitrage behavior. We believe Reflexer is one of the most stable projects in terms of price among all algorithmic stablecoins, possessing better resistance to death spirals in extreme situations where collateral prices plummet.

Design Flaw: No obvious flaws; over-collateralization also means lower capital efficiency, and compared to MakerDAO, it is not a disruptive innovation (is the stability of the stablecoin price mainly due to following MakerDAO's mechanism, and is the PID algorithm not the main reason?).

Regulation Method 4: Control of Collateralization Rates of Composite Collateral

Representative Project: Frax

Issuance: Issuing stablecoins requires two types of collateral------ external assets (currently USDC) and protocol stock FXS. The ratio of these two assets (collateralization ratio) is controlled by the protocol algorithm. For example, a collateralization ratio of 50% means that to mint $1 of stablecoin, $0.5 of USDC and $0.5 of FXS must be deposited.

Regulation: The specific regulatory mechanism has been analyzed in an easy-to-understand manner by Haseeb Qureshi and Alexis Direr, among others.

Overall, Frax has designed a dual-collateral mechanism, and stablecoins can always be redeemed at $1, thus relying on arbitrage behavior to pull the market price of stablecoins back to a certain range.

Inflation: When the market price of the stablecoin < target price (currently $1), arbitrageurs will buy stablecoins in the market and redeem collateral from the Frax protocol, ultimately selling the protocol equity (FXS) in the collateral for profit. The increased demand for stablecoins from arbitrageurs will raise the market price of the stablecoin.

Deflation: When the market price of the stablecoin > target price, arbitrageurs will deposit collateral into the Frax protocol to mint stablecoins, ultimately selling stablecoins in the market for profit. The increased supply of stablecoins in the market will lower the market price of the stablecoin.

Value: Frax is one of the projects we are currently most optimistic about in this field.

Its design seems to draw on the strengths of various previous algorithmic stablecoins while learning from their failures, such as: using stablecoins as collateral rather than other volatile assets in the early stages; abandoning the global supply increase/decrease Rebase > arbitrage behavior > price increase/correction, and instead designing a fixed exchange rate for two-way minting/redeeming at any time > arbitrage behavior > price increase/correction.

From the current system asset collateralization rate, it is also the first algorithmic stablecoin project that can maintain stablecoin price stability without requiring over-collateralization.

Design Flaw: No obvious flaws; the initial stability of the system is also largely related to the collateral being USDC. Whether it can withstand significant drops in crypto assets when the collateral is changed to other volatile assets will require time to prove.

Regulation Method 5: Protocol Treasury Dutch Auction

Representative Project: Float

Issuance: Similar to Frax, issuing stablecoins requires two types of collateral------ external assets (currently ETH) and protocol stock BANK.

Regulation:

Method One: Protocol Treasury Regulation

Method Two: User Arbitrage Behavior Regulation

Value: Float combines multiple factors that can help the algorithmic stablecoin system better resist death spirals: composite collateral, seigniorage tokens, the use of bonds, protocol treasury regulation + Rebase issuance relying on user arbitrage behavior for regulation.

At the same time, the protocol also has its innovations in the fluctuation of target prices, so I believe this project possesses better robustness and innovation. Although liquidity is currently low, we believe it has good development potential.

Design Flaw: Due to the lack of a liquidation mechanism and over-collateralization, this DeFi system also faces the challenge of collateral in the treasury dropping during extreme market conditions.

3. Actual Performance of Algorithmic Stablecoin Projects and Price Stability

For the current mainstream algorithmic stablecoins' price performance, we compared their daily price fluctuation standard deviations and found that the Frax and Terra projects performed relatively better:

How to Face the Death Spiral

The death spiral is a problem that all algorithmic stablecoins must face (or more broadly, all synthetic asset DeFi protocols face this issue, as stablecoins are also a type of synthetic asset). We have observed the performance of current market projects and summarized several possible concurrent factors that can help projects better cope with extreme market conditions:

Speculation certainly exists in the early stages, but the key is that speculators ultimately bring a large amount of funds to the protocol, driving the economic flywheel.

Collateral can provide users with greater confidence. Greater over-collateralization, as adopted by MakerDAO and Synthetix, provides a protective layer for the system's price. Therefore, from the perspective of current algorithmic stablecoin price stability, some collateralized projects overall perform better than uncollateralized projects.

The design of protocol equity and seigniorage is crucial. Here we see that Terra's elastic protocol equity and Frax's equity minting/redeeming arbitrage design give their economic systems greater robustness.

The use of bonds during deflation can also absorb a lot of selling pressure.

4. Investment Logic------ What Might a Successful Algorithmic Stablecoin Look Like?

Based on the diverse designs of current algorithmic stablecoins and their actual market performance, we believe the following points will be essential for supporting the steady growth of algorithmic stablecoin projects:

Almost essential:

• Composite collateral: external collateral such as ETH, other stablecoins; protocol equity tokens as collateral.
• Elastic Seigniorage Shares, designing the supply of protocol equity tokens to absorb the volatility of stablecoins.
• Protocol equity tokens should have additional functional design value beyond governance and seigniorage (like the protocol equity token $LUNA in Terra stablecoin, which is also the native token of the Terra blockchain).

Potentially beneficial:

• Use of bonds.
• Target price not pegged to $1, redesigning a floating target price mechanism that automatically adjusts the target price based on market supply and demand to create arbitrage opportunities.
• Using the system treasury (collateral) to control liquidity in the stablecoin trading pool in the market.