Olympus DAO's Governance System Didn't Work Until the Third Try
Olympus DAO has already cycled through three different governance frameworks before it finally hit on the ampleforth governance token mechanism it has in place today to control its now $196 million treasury. It was not a smooth process. Olympus' first system of governance was plagued by voter apathy, rebase-based voting power dilution, and misaligned token mechanics that miscalibrated proposal thresholds. To answer the question "what are governance tokens?" and why they fail, you need to go deep into protocol data, and Olympus is one of the most illuminating case studies in DeFi of iterative design in action.
The current system in which the proposer must own at least 0.017% of the total gOHM supply, and the proposal must receive 20% of supply voting "FOR" for it to pass didn't just come from a whitepaper. It evolved out of two failures that preceded it.
The Participation Crisis and the Rebase Quorum Problem
Olympus first launched all the way back in early 2021. At launch, there was 1 governance token called OHM. Holders would stake their OHM, in exchange for sOHM (staked OHM) and rebase rewards. At Olympus's height, this meant that users could earn rebase APYs of over 7,000%. The issue was that almost nobody voted. Snapshot proposals on Olympus during v1 would regularly have less than 200 unique wallets voting. This was a minuscule number relative to the ~5% participation rate of holders at the time. The DeFi Pulse Top 100 even included protocols like Uniswap, who would regularly see 2,000-5,000 unique voters during the same time period.
The issue was simple. OHM's staking mechanism created a prisoner's dilemma. If you unstaked to vote in a decentralized governance process, you lost compounding rebases. The relative position of unstakers is reduced with each rebase. Financially, voters were punished for interacting with the governance system of the protocol.
Snapshot proposal pass rates during v1 tell a story. 14 Snapshot proposals were submitted between May and October 2021, 11 passed. 78% pass rate sounds healthy on the surface. But due to low absolute participation numbers, proposals were passing with support from a minuscule part of the community. Several treasury allocation votes attracted fewer than 150 wallets total. For a protocol with hundreds of millions in reserves, that degree of centralization of power among so few people was an affront to the very idea of crypto governance.
The disconnect between Olympus's monetary policy and blockchain governance ran deeper than indifference. Rebases altered the total supply of OHM every eight hours. The quorum threshold, expressed as a fraction of raw OHM supply, was a moving target as a result. A proposal with a 50,000 OHM voting requirement could represent 3% of total supply one day, and 2.8% three days later. Governance mechanisms pegged to a fixed number of tokens cannot function if supply is inflationary. At least three different v1 proposals had their effective quorum ratio modified between the time of submission and the close of voting.
As a second experiment, the protocol introduced an sOHM-based voting mechanism. Votes were weighted by the number of tokens staked, but sOHM could be used towards governance even when unstaked. This solved the economic penalty problem. It did not resolve the math problem. sOHM balances were still rebased every time a rebase occurred. The sOHM voting experiment lasted for roughly four months. Participation did improve somewhat (median number of unique voters per proposal went from ~180 to ~260), but the fragility of quorum calculations remained. Two proposals had to be manually extended during this period because supply changes due to rebase reduced participation ratios below quorum after votes had already been submitted to pass the threshold.
How gOHM Changed the Math Overnight
Near the end of 2021, a clean slate was decided on by the creation of the Governance OHM token (gOHM). gOHM is an ERC-20 wrapper on sOHM which automatically accounts for rebase rewards in its exchange rate instead of a token balance. 1 gOHM will always represent an increasing amount of underlying OHM while the holder's gOHM balance remains the same. This single change eliminated the moving-target quorum problem entirely.
The beauty was in the simplicity. Fixing the token balance and floating the exchange rate allowed Olympus to design a governance token with predictable and stable voting power on the individual level and predictable macro-level behavior on the protocol level. The 0.017% supply threshold to submit a proposal and the 20% quorum requirement to pass it were both meaningful, enforceable numbers.
The gOHM token also provided a solution to the cross-chain governance problem. gOHM has been deployed on Ethereum, Arbitrum, Polygon, Avalanche, and Optimism, so holders across multiple networks can vote without having to bridge their assets back to mainnet. The protocol was cross-chain deployed on Arbitrum, Base, Berachain, and Solana following the integration of Chainlink CCIP in June 2025. The August 2025 Coinbase listing on Base network further enhanced the token and the governance process it underpins.
What Actually Changed After the Switch
Over a 12-month period, the network had 23 formal governance proposals voted on after gOHM compared to 14 in a comparably longer (16 months) v1/sOHM era. The number of proposals nearly doubled with a 64% increase in the volume of activity. Unique voter turnout has also improved. The median number of unique voters per proposal increased from the sOHM-era median of 260 to more than 400 for the first half of 2022.
Types of proposals put to vote have shifted as well. Most v1-era proposals were one-dimensional parameter changes (bond terms, reward rates, etc.). After gOHM, the community began passing more nuanced treasury management proposals like the Yield Repurchase Facility and the Cooler Loans program. The latter (borrow up to 95% of OHM's liquid backing at 0.5% interest against gOHM collateral with no price-based liquidations) wouldn't have been possible under the old governance system due to the complexity involved. It required weeks of discussion across multiple forum threads and on-chain votes, which itself required the stable participation metrics that gOHM made possible.
The catch is, pass rates also declined, from 78% (v1) to just over 70% (gOHM). That's a good thing. This represents a marked shift from the prior system, where it was often rubber-stamped by a small group of people. Comparisons to other DeFi governance tokens are also informative here. Protocols with fixed-supply governance tokens like Aave and Compound have also not faced the rebase issue that plagued Olympus, but their participation rates (on average 2% to 8% of the token supply voting on a particular proposal) aren't meaningfully better than gOHM's after the transition. Olympus solved a more difficult problem, and the result is a governance system on par with protocols that never had to contend with elastic supply.
What the Treasury Votes Revealed Under Stress
The real measure of a crypto governance framework isn't voter turnout. It is whether the governance process consistently yields reliable decisions under stress-testing. Olympus's treasury votes provide direct evidence. During the January 2026 crash, OHM price fell by 15% while Ethereum fell by 26% over the same period. The yield repurchase facility, range-bound stability, and Cooler Loans that helped mitigate the crash were all systems implemented via gOHM-era on-chain governance votes. OHM had no liquidations during the crash because the Cooler Loans system does not have a price-based liquidation trigger. This decision to opt out of a liquidation trigger was made by gOHM voters and the protocol protected capital at a time when other lenders were forced to liquidate.
The protocol's reserves are largely in sDAI and are distributed and rebalanced according to policies approved by governance, rather than at the discretion of multisig. Every significant treasury decision since gOHM has been made via the on-chain proposal process. The price of gOHM at about $4,621 is still 85% below its all-time high of $31,351. The figure also suggests a project's advancement in governance does not protect its underlying crypto asset from market cycles. gOHM's market cap of $196 million is less than half the size of the protocol at its 2021 peak. The ecosystem has undergone a contraction and the governance framework which matured during this period is now processing an actual treasury with measurable results.
The evolution from OHM to sOHM to gOHM wasn't 3 different experiments. It was 1 refinement process, based on 1 technical limitation: how to construct a workable governance framework for an elastic-supply asset. The solution demanded decoupling of value accrual from voting power, quorum mechanics stabilization, and cross-chain deployment. Other projects developing blockchain governance solutions for rebasing or inflationary tokens encounter different facets of this problem. Olympus's 3-iteration path to gOHM is a data-driven blueprint not because gOHM is infallible but because the failure modes of its predecessors are exactly documented in on-chain proposal records.
