• FIP-0032 introduces changes to gas accounting that affect storage provider proof commits. Though the batch balancer is expected to absorb most of the change, there is a risk of short-term basefee spiking and oscillation. There’s uncertainty in how this plays out due to the behavioral aspect of the batch balancer’s dynamics, which rely on storage providers’ decisions, and how this interacts with the basefee update rule under exogenous shocks.
• We can predict that there will likely be some degree of change to the rational economics of the batch balancer following implementation of FIP-0032. This is because in the benchmark the proportion of gas needed for batched as compared to single commit proofs is altered compared to the status quo. The change to the rational unit economics may however not be large. To give a concrete example, at basefee=0.32 nanoFIL, we estimate the number of proofs to incentivize rational batching increases by four proofs post-FIP-0032 implementation. The quantiles for this change in our analysis range from Q10=-2 to Q90=+11 proofs. The wide range is due to possible configurations of the gas usage multipliers sampled in the FIP-0032 gas usage backtests, which gives some sense of the potential for variation in outcomes.
• A new value for the batch balancer parameter has been generated under the same rational batching assumptions as for the currently used parameter. The new value is not significantly different so we suggest no change.
• It’s important to monitor the relative increase in batch and single sector commit gas usage in case it turns out to differ in production a lot compared to the benchmark. In a scenario where the increase in batching gas is actually x2 larger than the benchmark estimate, this would shift the rational unit economics substantially. For example at basefee = 0.32 nanoFIL, the rational batching crossover increases by 29 proofs instead of by 4, up from a level of 13 currently.
• This summary follows from a Jupyter notebook available in the CryptoEconLab github.