Tokenomics: Optimal monetary and fee policies
What this paper finds — and why it matters
The rapid proliferation of cryptocurrency tokens—roughly 10,000 outstanding with a total market capitalization around $3 trillion as of early 2024—raises new questions about the design of token monetary policy and fee structures. This paper makes two contributions. Empirically, using supply histories for approximately 2,000 tokens, it documents three systematic patterns: average token money growth rates decline with age and stabilize at about 0.2% per month; long-run money growth rates and convergence speeds are positively correlated across tokens in the cross-section; and tokens more widely held by retail investors have relatively lower long-run money growth rates and convergence speeds. Theoretically, the paper derives optimal issuance and fee policies for a profit-maximizing issuer in a dynamic model where commitment matters, showing that a fully committed (Ramsey) issuer who maximizes profits after the initial period makes choices that maximize the total utility value of all tokens, that without any commitment no equilibrium with a positive token price exists unless fees are charged, and that under partial commitment issuers with higher commitment credibility optimally choose lower long-run money growth rates and fee ratios.
Summary of a forthcoming paper, AI-assisted and human-reviewed. See the linked original for the authoritative claims and full conditions.
Q1. What are the three empirical facts about crypto monetary policies?
Fact 1: the average money growth rate across the ~2,000 tokens declines with cohort age and stabilizes at approximately 0.2% per month, with more recent cohorts converging faster to their long-run growth rates. Fact 2: in the cross-section of tokens, the estimated long-run money growth rate and the speed of convergence to it are positively correlated—tokens that will eventually have higher money growth also converge more quickly to that level. Fact 3: tokens whose circulating supply is widely distributed among retail investors (measured by the proportion of wallet addresses holding less than 0.1% of total supply) have both lower long-run money growth rates and lower convergence speeds. Together these facts suggest that the degree of decentralization of token ownership systematically influences the choice of monetary policy.
Q2. What are the main theoretical results about commitment?
Without any commitment, no equilibrium with a positive token price can be sustained if the issuer cannot charge user fees; fees create credibility by giving a Markov-perfect issuer an incentive to restrict future token issuance (to collect fees on legacy token holders), which supports a positive price. At the other extreme, with full commitment the Ramsey issuer maximizes profits subject to the full sequence of resource constraints, and a central analytical result is that at steady state the Ramsey issuer’s profit-maximizing choices are equivalent to maximizing the total utility value of all tokens outstanding. This equivalence—profit maximization = welfare maximization for token holders—holds because commitment prevents the issuer from diluting legacy holders; under full commitment, the issuer has no incentive to deviate from the policy that maximizes total token value.
Q3. What does partial commitment imply for policy design?
Under partial commitment (modeled as a probability of sticking to a given policy rather than re-optimizing), issuers with higher commitment probability optimally choose lower long-run money growth rates and reduce both the money growth rate and the fee ratio more slowly over time. This is because higher commitment credibility raises the present value of legacy tokens (users expect the issuer to honor its policy), which makes it optimal to extract less seigniorage and fewer fees in each period while sustaining a higher token price. The model is analytically solvable under partial commitment despite rich economic mechanisms, and the results help interpret Fact 3: tokens with more decentralized ownership (higher effective commitment due to governance constraints) exhibit lower money growth.
Q4. How do the model’s predictions compare with the Bitcoin and Ethereum cases?
Bitcoin exemplifies near-full commitment: its supply schedule is essentially deterministic and hard-coded, resembling the Ramsey policy, though the paper notes a non-zero probability of a fork that could alter the policy. Ethereum exemplifies frequent policy changes (the transition from proof-of-work to proof-of-stake altered issuance and burn rules significantly). Discretionary platforms like MakerDAO, which allow issuer discretion over token supply, more closely resemble the Markov-perfect benchmark. The paper’s framework covers this entire spectrum and predicts that tokens with more commitment (closer to Bitcoin) should have lower long-run money growth and slower convergence—consistent with the empirical facts.
Key concepts
token monetary policy : the issuer’s choice of the rate at which new tokens are minted over time (the money growth rate) and the fees charged per transaction; the paper shows these jointly determine the value of a token and the issuer’s profit.
Ramsey policy for tokens : the full-commitment profit-maximizing policy for a token issuer; a key result is that the Ramsey issuer’s profit-maximizing choices, after the initial period, are equivalent to maximizing the total utility value of all tokens.
partial commitment : a probabilistic commitment technology in which the issuer maintains a given policy with some probability and re-optimizes with the complementary probability; the paper uses this to model the spectrum between Bitcoin (high commitment) and more discretionary platforms.