HashKey Cui Chen: Analyzing the Design Mechanism and Development Prospects of Algorithmic Stablecoins
In addition to mechanism design, the development of algorithmic stablecoins may also require stimulation from other events, such as trust crises in other types of stablecoins or large-scale liquidations.This article is from HashKey Research, authored by Cui Chen.
Stablecoins are more suitable as tools for pricing and payment compared to other crypto assets, playing an indispensable role in the crypto asset market. Currently, stablecoins can be divided into three types: fiat-backed, crypto-backed, and algorithmic stablecoins. The representatives of fiat-backed stablecoins are USDT and USDC, where users exchange fiat currency for stablecoins at a 1:1 ratio through stablecoin issuers, with reserves managed by the issuers or other custodial institutions. Crypto-backed stablecoins are represented by DAI, where users need to over-collateralize crypto assets on-chain to exchange for DAI. If the value of the collateral falls below the specified minimum collateralization ratio, the system will liquidate the collateral to repay the stablecoin debt.
The issuance of fiat-backed stablecoins relies on centralized management institutions and is under strict regulation, thus requiring KYC certification. Compared to the issuance methods of on-chain stablecoins, this is less convenient and carries centralized trust risks. Crypto-backed stablecoins are issued on-chain through users over-collateralizing assets, which leads to low capital utilization. This is because there is no credit rating introduced to prevent the value fluctuations of collateral assets from causing insufficient asset backing. Even with over-collateralization, there is still liquidation risk for users. Algorithmic stablecoins are a method of issuing stablecoins that liberates on-chain liquidity, with no funds backing them, instead adjusting prices through algorithms to maintain their peg.
Algorithmic stablecoins achieve price adjustment by controlling supply and demand. If the stablecoin price is above the benchmark, it indicates that demand exceeds supply, prompting the system to issue more stablecoins. Conversely, if the stablecoin price is below the benchmark, the system will destroy some stablecoins. The methods of issuance and destruction vary among different algorithmic stablecoin projects, but the common challenge is that the regulatory mechanisms struggle to maintain long-term price stability. Algorithmic stablecoins are difficult to use as stablecoins, but why do algorithmic stablecoin projects attract so much attention and discussion each time?
Firstly, for the "natives" of the cryptocurrency world, using an algorithmically controlled token model is a trusted approach. Bitcoin's design utilizes cryptographic algorithms, coupled with economic incentives that enable the trustless functionality of transactions. Compared to fiat-backed stablecoins, trustless financial products in the on-chain DeFi ecosystem align more with blockchain fundamentalism. The issues of low capital utilization and liquidation risk in over-collateralized stablecoins have led to algorithmic stablecoins becoming a new solution. Secondly, algorithmic stablecoins achieve price stability by adjusting supply, creating various arbitrage opportunities during the processes of issuance and destruction. Finally, stablecoins are a crucial component of DeFi; if algorithmic stablecoins can achieve price stability, it will change the current landscape of stablecoins, and a successful anchoring mechanism will allow algorithmic tokens to extend beyond the application of stablecoins.
Based on whether users directly participate in the adjustment process, the design of current algorithmic stablecoin mechanisms can be divided into passive and active types, each with its characteristics.
Stability Mechanism
Passive Mechanism
The passive adjustment mechanism, also known as rebase, adjusts the token price to the target level through continuous issuance and destruction by the system, such as in the Ampleforth project. Its target price fluctuates within 5% of the benchmark price. If the token price exceeds the target price, it indicates that demand exceeds supply, necessitating the issuance of more tokens to balance supply and demand for price stability. When the token price is below the target price, the system will destroy tokens to achieve balance, and the process of issuance or destruction is referred to as rebase. Rebase occurs every 24 hours, directly implemented through addresses, reflected in the increase or decrease of the amount in the holding addresses.
Therefore, regardless of how the total amount is adjusted, the proportion of tokens held by holders remains unchanged. The price return after rebase adjustment of the token supply is achieved through market behavior, so it does not mean that adjusting the quantity will ensure price return. Generally, the assets of project holders are affected by the unit price of the token; in rebase-type algorithmic stablecoins, users' assets are influenced not only by token volatility but also by changes in quantity.
To prevent significant fluctuations in supply, Ampleforth introduces a parameter k during adjustments, where the number of tokens issued or destroyed will be based on k days. If the token quantity exceeds the target by a%, then in the next rebase, a/k% of tokens will be issued, and the value will be recalculated the next day. For example, if the token is valued at $1.5 before rebase and k is 10, then after rebase, the number of tokens held by users will increase by 5%. In this case, the optimal strategy for holders is to sell the newly issued tokens, especially when the token value is still at $1.5. Selling when the price is above the target price can help bring the price back to the target. Conversely, if the token price falls below the target price, holders have no incentive to sell, as selling below the target price would incur losses. Instead, purchasing at a price lower than the target would yield profits when the price returns to the benchmark, and such buying behavior would push the price up to the target.
Although the rebase mechanism theoretically adjusts supply and demand to stabilize the token, it is challenging to achieve a balance in practice, especially with many speculators involved. When rebase occurs with issuance, if no one sells, everyone profits at the current price. Even if the price gradually declines, only later buyers incur losses. Therefore, in a rebase where the price is consistently above the benchmark, early buyers' profits will always exceed those of later buyers. If early buyers do not sell, they will continue to accumulate the benefits of issuance, leading speculators to seek early entry for greater profits, creating a situation where apparent demand exceeds actual demand, forming a structure akin to a Ponzi scheme.
If the token price is below the benchmark and needs to be destroyed through rebase, the reduction in quantity and the drop in price will cause holders to suffer double losses. If the token cannot return to the target price, the earliest sellers will incur the least loss. In the absence of confidence, this leads to a downward spiral of competing to sell. These two abnormal trends can easily intersect when supply and demand balance is disrupted, until a new equilibrium point is reached. The following figure also shows that since its inception, Ample's price has fluctuated between being too low and too high.
Figure 1: Ample's Price and Supply
In the rebase mechanism, aside from the difficulty in maintaining stability due to the mechanism's setup, the 5% fluctuation range makes it challenging to use as a stablecoin tool. Frequent rebase can affect usability due to the constant changes in the quantity held in wallets. Currently, besides the passive adjustment stability mechanism of rebase, active mechanisms requiring user participation have also emerged.
Active Mechanism
The active mechanism similarly controls the token supply to achieve supply-demand balance, stabilizing the token around the benchmark price. Unlike the rebase method, the active mechanism introduces more user behaviors by issuing multiple tokens, incentivizing users to participate in the destruction and issuance of tokens. Changes in supply only relate to the holders of the other issued tokens and no longer affect all holders, so for users not participating in supply adjustment, the total value of their assets is only influenced by the unit price. The active adjustment method is easier to understand in practice, reducing the impact of quantity fluctuations in wallets on actual applications.
Taking Basis as an example to illustrate the adjustment method of the active mechanism, when the price of the stablecoin Basis Cash (BAC) is below the target price, users can destroy BAC to exchange for another certificate, Basis Bond (BAB). The price of BAB is the square of BAC, so the lower BAC is below $1, the lower the price of BAB. Users purchasing BAB is essentially destroying BAC to reduce supply. If the price of BAC exceeds $1, the system will issue more BAC, allowing users to exchange their low-priced BAB back for BAC at a 1:1 ratio. When the system issues more, it will prioritize repaying previously issued debts, making BAB purchasers effectively engage in arbitrage. If after repaying all debts, the price of BAC has not returned to normal, the system will continue to issue and distribute to staked Basis Share (BAS) holders until supply-demand balance is achieved and BAC returns to the target price.
In the active adjustment mechanism, the economic incentives stimulate users to participate in the destruction and issuance processes, but like the passive mechanism, there are still long-term stability issues. When the price of stablecoin BAC is above the target price, users want to hold the equity token BAS to receive BAC dividends. Besides direct purchases, BAS can also be earned through liquidity mining of BAC. To obtain BAC dividends, users need to buy BAC for staking and mining, thus driving up the demand for BAC. The demand for BAC raises its price, and the higher the BAC price, the greater the motivation for users to obtain BAS, creating an unhealthy model. Once the stablecoin BAC loses the expectation of dividends due to high prices, the demand for BAS will quickly decrease, which will also affect the demand for BAC, causing BAC's price to drop rapidly. When the price is below the target price, arbitrageurs buying BAB to destroy BAC is a crucial part of restoring its price to normal levels. Only when BAC can rise above the target price can arbitrageurs profit, which requires sufficient confidence in the stablecoin's price increase. Without this, a lack of sufficient supply may keep the stablecoin below the target price for an extended period, potentially leading to project failure. Although BAB plays the role of a bond in the active adjustment mechanism, debt repayment depends on the price of BAC, and only when BAC's price is above the benchmark can BAB be exchanged for BAC. Thus, the destruction mechanism only takes effect when there is confidence in BAC.
The advantage of the active adjustment mechanism lies in the use of multiple token models, issuing bonds BAB to destroy BAC when the price is overly depressed, restoring it to normal levels, and repaying debts with the newly issued BAC when the price is overly inflated. This distribution method is easier to understand and engages users, making supply adjustments more flexible and aligned with market dynamics. Currently, the active mechanism also cannot maintain the stability of stablecoins, as BAC's price has been below the target for the past two months.
Figure 2: BAC Price Trend (Image Source: CoinGecko)
Other Asset-Related Algorithmic Tokens
In addition to algorithmic stablecoins, there are also algorithmic tokens pegged to other assets, such as the Bitcoin algorithmic token DIGG. DIGG's mechanism is rebase, adjusting the total supply elastically to peg DIGG's price to Bitcoin's price. For tokens pegged to other assets, maintaining a complete price peg through the rebase mechanism is similarly challenging. However, unlike stablecoins, the payment demand is not significant; rather, there is more demand for spatial transfer of assets. For example, issuing a Bitcoin-pegged algorithmic token on Ethereum allows a "Bitcoin-like" entity to participate in the Ethereum ecosystem without cross-chain transactions. Pegged tokens can only passively reflect Bitcoin's price through continuous rebase, unable to influence the actual asset price.
Synthetix, Mirror Protocol, and UMA in DeFi are synthetic asset platforms that have garnered much attention. Through over-collateralization, they allow other assets to be traded within existing networks, such as stocks. Although the prices of synthetic assets cannot influence real assets and do not enjoy other rights, for users, having price exposure is sufficient, aiming to gain market access opportunities. Unlike synthetic assets, algorithmically pegged assets have no physical backing; if there are few participants and poor liquidity, algorithmically pegged assets can easily experience long-term price distortions.
Reflection and Conclusion
Currently, no stable algorithmic stablecoin mechanism has emerged, and both popular stability mechanisms face stability issues. The problems stem from two sources: the mechanism design does not fully align with human nature, and the presence of numerous speculators has not found a balance point. Bitcoin's design to prevent double-spending attacks considers both mechanism and human nature's dual regulation. The proof-of-work mechanism sets a threshold for accounting nodes, requiring 51% of the network's computing power to execute a double-spending attack. Currently, the mechanism design of algorithmic stablecoins has many theoretical flaws and does not account for the psychological impacts of speculators. Stablecoins without physical backing make it challenging for people to maintain long-term confidence. This lack of confidence and the alternating expectation of profit make it difficult for algorithmic tokens to remain near the benchmark price. The participation of speculators prevents token prices from reflecting real supply and demand relationships, thus affecting the mechanism of algorithmic stablecoins. For algorithmic stablecoins, in addition to mechanism design issues, there are also regulatory risks. Stablecoins are the crypto assets that have the most significant impact on the existing financial system, and future regulatory measures concerning stablecoins in the DeFi ecosystem may affect their development.
People's expectations for algorithmic stablecoins will continue to drive their development, and updates on model design will receive particular attention. Besides mechanism design, the development of algorithmic stablecoins may also require stimulation from other events, such as trust crises or large-scale liquidations of other types of stablecoins. Once the stability mechanisms mature, algorithmically pegged assets will become an important development direction for the future, primarily aimed at achieving cross-space market access needs.
