Staking, Token Pricing, and Crypto Carry

The phenomenal rise of cryptocurrencies and decentralized finance have prominently featured “staking”: Besides offering a convenience yield for transactions as digital media of exchange, tokens are frequently staked (and slashed) for base-layer consensus generation or for incentivizing economic activities and platform development, and consequently earn stakers rewards in the same tokens. To provide insights into the economics of staking and its asset pricing implications, we build a continuoustime model of a token-based economy where agents heterogeneous in wealth dynamically solve their wealth allocation (stake-transact-consume) problems. We cast the interactions as a mean field game with stochastic control and systematic shocks, which underscores aggregate staking ratio as a key variable linking staking to token pricing and equilibrium reward rate. Empirical findings on all major stakable tokens corroborate the model predictions. In particular, staking ratio has a positive correlation with reward rates in the cross section and has a negative correlation in the time series. Higher reward rates attract greater future staking, increasing an individuals’ staking allocation and the staking ratio in aggregate, which in turn predicts positive excess returns. Finally, we use transaction convenience to rationalize violations of the uncovered interest rate parity and significant carry premia (e.g., a long-short carry yields a Sharpe ratio of 1.6) in the cryptocurrency data.