What Are the Market Dynamics of Sui?
Sui strives to keep the good actors and bad actors in balance with (dis)incentives.
By Eric Nemeth, investment associate
Introduction
Sui recently unveiled their Testnet Wave 2, following the success of the Testnet Wave 1 run in December 2022. This wave plans to increase the number of validators and users able to run a full node as well as test the market mechanics for Sui’s tokenomics.
The purpose of this test is to validate that the core functions, including staking, unstaking, and reward distributions, are operating as intended. The wave will also provide an opportunity for validators to become familiar with the new mechanisms implemented by the Sui network. Sui will conduct a gas price survey to determine the quotes offered to users, and the wave will include testing of payment transactions for the storage fund.
Sui Tokenomics Design and Market Mechanisms
Tokenomics encompasses a few areas, including proof of stake rewards, market-based gas pricing, and the cost of storage and the storage fund itself. Sui's proof of stake is standard to how it implements incentives into the economy, but it differs from other Layer 1’s in that it parallelizes execution of transactions.
The security model assumes that one-third of validators may act maliciously, and this assumption is reflected in the market-based gas pricing mechanism, which has three components: a gas price survey, tallying rule, and a reward distribution system that incentivizes staking.
Gas Prices
The determination of gas prices for the next epoch in Sui is based on a survey of the current epoch's validators, who attest to the prices they are committed to for processing transactions.
Tallying Rule
A tallying rule is then applied at the end of the epoch, once transactions have been processed and before stake rewards are distributed. The idea is for each validator to subjectively evaluate how others have honored the gas prices they committed to the survey. This test assigns a score to validators based on their performance in upholding prices they committed to in the survey. The community subjectively grades each other and submits their evaluation of each member’s performance.
This market-based approach assumes that, even if one third of validators are Byzantine, the honest actors will strive to maximize their rewards and submit accurate gas prices that they will uphold. If there are over one third of bad actors, the network would not be Byzantine tolerant anyways and would fail consensus.
Validators with good performance can get a boost in the rewards, indicating that their model for forecasting gas prices was accurate and they were able to process transactions at the prices they committed to.
In short, validators are incentivized to maximize their rewards by committing to realistic prices that they will uphold in the next block. Dishonest or malicious behavior is disincentivized by the final test as follows:
Stake Reward Distribution Rule
With the tallying rule used to determine gas prices validators will honor, the ‘stake reward distribution rule’ creates another multiplier metric to incentivize low gas prices. This rule takes a weighted median of the multipliers from the tallying rule and rewards the validators who submit low gas prices below the two thirds percentile of all validators' submissions.
The objective of this rule is to discourage validators from attempting to gain an excessive share of rewards by giving each other excessively high tallying multiples.
In theory, validators will optimize their performance to provide pricing which grants them rewards proportional to the amount they have staked.
In practice, the validators will try to align their honesty with the rest of the network as they are penalized for being dishonest through the tallying and reward distribution rule.
From this high-level overview, it becomes clear that the tokenomics for determining gas prices are driven by market mechanics to incentivize optimal behavior which validators can develop business models for.
As the network grows, the validator market is expected to become competitive as validators strive to measure market metrics accurately and find ways to evaluate their peers' performance.
However, it appears there is a lot more game theory to understand here. It will be challenging to determine gas prices within the early days as there will be limited data and the early growth of the network will be volatile. It will be hard to understand the level of elasticity to gas prices as users will accept any costs to execute transactions either from a surge in demand for Sui apps, or because validators are unable to determine if gas prices are high given the lack of historical data.
Storage Fund
Lastly, there is the Sui Storage Fund which is used to offset costs from storing large amounts of on-chain data. When users transact on the network, they will pay the fees for computation and incur fees for storing the data. The fees are distributed to the storage fund and to validators in order to mitigate excessive data stored on Sui.
The unique features of the fund allow future validators to be compensated for historical data stored by users, which is paid out by the storage fund.
Users can also delete data from the on-chain storage for a storage fee rebate. This helps incentivize users to reduce the bloat of storage on-chain by effectively setting a price floor for data on Sui. What this could mean for data is that there is an intrinsic value for data created on Sui’s network.
The storage fund also offers a loan function for validators to borrow a stake to bootstrap their node. The validator will then repay the funds by returning a percentage of rewards they earned from processing transactions. The storage fund will only perform functions like this using returns on capital and will not dip into the principal reserves deposited in the storage fund.
Conclusion
Overall, the testwave will introduce new market dynamics to Sui’s network. To read more about the test wave, visit Sui’s Wave 2 announcement post. There are a number of engaging ways to get involved to help stress test the network, like running your own node or getting involved with the games to test tokenomics design and implementation.