A Theoretical Approach to Stablecoin Design via Price Windows

1. Problem Statement

The paper addresses a fundamental open question in cryptocurrency: Can a fully decentralized, stablecoin exist that is provably stable in both the short and long term?

While many stablecoins exist, they generally fall into two categories with distinct failure modes:

• Fiat-backed (e.g., USDT, USDC): Stable but require trust in off-chain custodians, violating the "trustless" ethos of crypto.
• Crypto-backed/Algorithmic (e.g., DAI, Terra/Luna): Decentralized but prone to volatility and catastrophic failure (e.g., the Terra collapse).

The authors focus specifically on "Price Window" designs. These are mechanisms where the protocol offers fixed mint and redeem prices to all agents regardless of their identity or transaction history. Examples include Tether, USDC, and crypto-backed designs like Djed and USDN. The core problem is whether such designs can maintain stability over volatile backing assets (like ETH or BTC) without secondary stabilization mechanisms (like governance tokens or equity).

2. Methodology

The authors employ a game-theoretic and mathematical modeling approach to analyze the emergent properties of stablecoin systems.

• The Model:

  • Time: Discrete, infinite timesteps $T$.
  • Assets: "Tokens" (stablecoins) and "Backing Coins" (volatile assets like BTC/ETH).
  • Mechanism: A stablecoin system $M$ with reserves $R_t$. It mints tokens at price $\alpha$ and redeems them at price $\beta$.
  • Price Window: The mechanism offers a fixed price window $[1-\epsilon_\beta, 1+\epsilon_\alpha]$ in USD terms. The market price $d_t$ is constrained to lie within this window as long as reserves are non-zero.
  • Interchangeability: A key axiom of the model is that $\alpha$ and $\beta$ depend only on the system state, not on the specific agent or transaction history. This makes all agents and tokens interchangeable.

• The Adversary (The Speculator):

  • The authors introduce a "sensitive speculator" who possesses perfect foresight (or near-perfect foresight) of the backing coin's price sequence.
  • The speculator's goal is to maximize profit by arbitraging the difference between the backing coin's market price and the stablecoin's mint/redeem prices.
  • Strategy: Buy backing coins when the price is low (minting stablecoins cheaply) and sell them when the price is high (redeeming stablecoins for more backing coins than they cost).

• Stability Definitions:

  • $\epsilon$-Stability: The market price stays within $[1-\epsilon, 1+\epsilon]$.
  • Weak $\epsilon$-Stability: The mechanism maintains this stability over a non-convergent (volatile) sequence of prices indefinitely.

3. Key Contributions

A. Theoretical Framework for Price Windows

The authors formalize the "Price Window Model," distinguishing it from personalized redemption models (like DAI, which can block specific agents). They argue that the lack of agent-specific data makes price window designs inherently vulnerable to systematic arbitrage.

B. The Fundamental Trade-off (Theorem 3.1)

The paper's central theoretical result establishes a necessary condition for long-term sustainability.

• The Result: A price window mechanism with constant fees ($\epsilon_\alpha, \epsilon_\beta$) is weakly $\epsilon$-stable against a sensitive speculator if and only if the size of the price window scales with the asymptotic volatility of the backing asset.
• The Formula: Stability requires:
  $$(1+\epsilon_\alpha) \liminf \frac{1}{p_t} - (1-\epsilon_\beta) \limsup \frac{1}{p_t} \geq 0$$
  Where $p_t$ is the price of the backing coin.
• Implication: If the backing asset is volatile (i.e., $\limsup \frac{1}{p_t} > \liminf \frac{1}{p_t}$), the transaction fees ($\epsilon$) must be large enough to cover the spread. If fees are small (to maintain short-term stability), the system is mathematically guaranteed to suffer reserve depletion eventually.

C. Quantitative Analysis of Depletion Time

Moving from asymptotic theory to practical modeling, the authors:

1. Modeled a Rational Speculator: Defined a speculator with risk tolerance ($\Lambda$) and patience ($\delta$) who optimizes utility over a sequence of independent, identically distributed (i.i.d.) prices.
2. Derived Depletion Formulas: Calculated the expected time to reserve depletion based on the volatility ($\sigma^2$) of the backing asset.
3. Simulated Real-World Data: Ran simulations using 3 years of historical Bitcoin and Ethereum price data.

4. Key Results

1. Inherited Volatility: For price window designs, the stability of the stablecoin is an "inherited" property. If the backing asset is volatile, the stablecoin's market price will inevitably fluctuate proportionally to the backing asset's volatility unless the price window is widened significantly.

2. Inevitable Depletion: Without secondary mechanisms, a price window mechanism backed by a volatile asset will eventually be drained by a speculator. The paper proves that for small transaction fees, the total profit a speculator can extract is infinite over a non-convergent price sequence.

3. Volatility vs. Time-to-Depletion:

   • There is an inverse relationship between the volatility of the backing coin and the time to reserve depletion.
   • Simulation Findings: Using historical ETH and BTC data, the authors found that a speculator with a modest initial endowment ($n_0$) could deplete a system with $R_0 = 1000$ reserves within three years.
   • ETH (higher volatility) led to faster depletion than BTC.
   • Even with "patient" speculators, high volatility drastically reduces the time to failure.

4. The "Wait" Interval: The model shows that speculators do not trade on every price fluctuation. They wait for prices to fall outside a specific "waiting interval" $[y_1, y_2]$. However, as long as the backing asset's price distribution has non-zero variance, the expected profit per trade cycle remains positive, leading to eventual depletion.

5. Significance and Implications

• Design Taxonomy: The paper provides a rigorous mathematical taxonomy distinguishing "interchangeable" (price window) designs from "non-interchangeable" (personalized) designs. It explains why crypto-backed stablecoins often avoid pure price window models in favor of over-collateralization or algorithmic supply adjustments.
• Theoretical Limits: It proves that price windows alone are insufficient for long-term stability over volatile assets. This challenges the viability of designs that rely solely on mint/redeem arbitrage for stabilization without additional "secondary" mechanisms (e.g., seigniorage shares, governance tokens, or dynamic fee adjustments).
• Practical Guidance:
  • For system designers, the paper suggests that transaction fees must be set proportional to the volatility of the backing asset to prevent depletion.
  • For highly volatile assets (like ETH), the required fees to ensure long-term safety might be so high that they destroy the utility of the stablecoin (making it too expensive to use), effectively rendering the design impractical.
• Policy and Regulation: The findings highlight that "decentralized" does not automatically mean "stable." The paper suggests that true stability in a decentralized environment may require mechanisms that break the "interchangeability" assumption (e.g., tracking user history to prevent arbitrage) or accepting that the stablecoin's value will fluctuate with its collateral.

Conclusion

The paper concludes that price window-based stablecoins cannot achieve both short-term stability (tight peg) and long-term sustainability (infinite reserves) when backed by volatile assets. The mechanism faces a hard trade-off: either accept a wide price window (high volatility) to prevent arbitrage, or accept a narrow window (tight peg) that guarantees eventual reserve depletion by a rational speculator. This theoretical insight explains the fragility observed in several real-world stablecoin failures.