The Paradox of Algorithmic Stablecoin Investment: I like you because you're a stablecoin, but I invest in you because you're unstable.
Stablecoins, as the foundation of blockchain decentralized finance (DeFi), have always been a focal point in the industry. By definition, they are coins with stable value, such as the well-known USDT. However, a new type of stablecoin has recently gained significant attention—algorithmic stablecoins. Their emergence has sparked widespread interest but also considerable controversy. So, what exactly are algorithmic stablecoins, and how do they work?
Before diving into algorithmic stablecoins, let’s first clarify what stablecoins are. Stablecoins are primarily divided into three categories: fiat-collateralized, crypto-collateralized (over-collateralized), and algorithmic.
The first category (e.g., USDT and USDC, including exchange-based tokens like BUSD) is centrally managed, backed by fiat currency, and redeemable on a 1:1 basis. These stablecoins offer the advantages of maintaining a peg and capital efficiency (i.e., no over-collateralization), but their centralized nature means users can be blacklisted, and the peg itself relies on the trustworthy behavior of a central entity.
The second category, crypto-collateralized stablecoins, includes MakerDAO’s DAI and Synthetix’s sUSD. These stablecoins are over-collateralized by crypto assets and rely on price oracles to maintain their peg to the dollar. Unlike centralized tokens like USDT and USDC, they can be generated permissionlessly. However, in the case of DAI, it’s worth noting that centralized assets like USDC can also be used as collateral. Additionally, their over-collateralized nature makes them highly capital-intensive, and the high volatility and correlation of crypto assets have historically made these stablecoins vulnerable to crypto-wide shocks.
The third category, algorithmic stablecoins, adjusts the total supply algorithmically. When the stablecoin’s price exceeds the peg, the supply increases; when the price falls below the peg, the supply is reduced, or arbitrage opportunities are provided to balance the price. This model doesn’t peg to fiat currency or require collateral, relying solely on market dynamics and algorithms for regulation—hence the term "elastic money." Currently, there are two mainstream algorithmic models: AMPL and the ESDBAC hybrid stablecoin.
To better understand, let’s take AMPL’s Rebase mechanism as an example. Ampleforth recalibrates the peg at a fixed time each day, and its smart contract adjusts the total supply of AMPL accordingly. This daily adjustment is called a "Rebase," where all wallet balances are proportionally updated. After Rebase, each holder’s share of the total supply remains unchanged. Importantly, this is not dilution, as all account balances adjust proportionally—whether positively or negatively—without any airdrops or transactions. It’s purely a function of the AMPL smart contract.
However, this logic has flaws. When the token price rises, AMPL’s mechanism increases the circulating supply, which theoretically should lower the price as holders sell expecting depreciation. But because the increased supply boosts holders’ token quantities before any selling occurs, the price remains unchanged in the short term, leading to an increase in holders’ total asset value (more tokens at the same price).
In this scenario, would holders still sell AMPL to drive the price down? Savvy whales would likely hoard tokens to reduce supply, further pushing up the price and waiting for the next Rebase to mint more AMPL. This creates a vicious cycle where algorithmic stablecoins lose their stability.
