Overview
Chaos Labs recommends increasing the rate limit and per-second refill rate of the GHO bridge on Base. These changes will be implemented via the direct-to-AIP process.
Motivation
Base’s main GHO liquidity pools have rapidly accrued tens of millions in TVL, but users deploying large amounts of GHO have encountered significant bridging friction under the 300,000 GHO cap and 60 GHO/sec rate limit enforced by Chainlink CCIP. Transfers of high volume have required splitting into many smaller transactions, taking up to ~12 hours to complete in some cases. While these rate limits serve as a safety mechanism to protect GHO’s peg and security across chains, current market conditions indicate they may now be overly restrictive. This report evaluates the depeg risk implications of raising the bridge limits against the benefits of improved liquidity flow and adoption. We find that moderate increases in the bridging limits are justified by GHO’s growing maturity, robust peg stability to date, and the need to reduce barriers to cross-chain usage.
Analysis
GHO’s cross-chain transfers are facilitated by CCIP. When GHO is bridged from Ethereum to an L2 like Base, tokens are locked or burned on the source chain and minted on the destination, keeping total GHO supply constant. Crucially, CCIP gives Aave governance fine-grained control over bridging via configurable rate limits. The current limits – 300,000 GHO “bucket” capacity and 60 GHO/second refill rate per bridge transfer lane – were set conservatively at launch as a risk mitigation mechanism. These act as a token bucket: at most 300k GHO can be moved at once (if capacity is full), and capacity regenerates at 60 GHO per second. At maximum throughput, this equates to roughly 216,000 GHO per hour (≈5.2 million per day) transferable cross-chain under current settings. Any transfer larger than the available capacity must wait as the bucket refills, effectively throttling large or rapid movements.
Following a successful Aave governance vote in early 2025, GHO was officially launched on Base. GHO’s adoption on Base has been evident, as the one period of quick bridging off of Base was quickly reversed.
Major GHO liquidity pools on Base have quickly grown to substantial size. For example, by late April 2025, Base’s primary Balancer v3 GHO/USDC pool reached about $11.57 million in TVL, and a GHO/USR pool accumulated roughly $10.5 million. These rapid inflows (e.g. ≈$6.5 M added within 12 hours of the GHO/USR pool launch) underscore the strong demand for GHO on Base. Users are actively bridging GHO from Ethereum to deploy into these pools, motivated by attractive yields.
However, the cross-chain transfer limits have become a bottleneck in this expansion. Because the CCIP bridge only permits 300k GHO per transfer with a 60 GHO/sec refill, larger entities moving millions of GHO to Base experienced delays. In practice, bridging substantial amounts (e.g. 1–2M+ GHO) can take many hours. This operational friction means capital is tied up awaiting bridge completion, and users must manage multiple transactions. By comparison, competing stablecoins can often be moved across chains with less friction.
Risks
Maintaining GHO’s $1.00 peg across all markets is the foremost priority. Any change to bridging limits must ensure that neither global nor localized peg stability is compromised.
On Arbitrum and Base, new pool mechanisms such as Balancer’s “StableSurge” dynamic fee (which raises swap fees if the pool’s price diverges from peg) provide additional defense, automatically increasing swap fees as GHO moves off-peg. No significant depeg events have been observed to date this year, even during periods of stress in broader markets.
The CCIP parameters were originally established as a conservative measure, ensuring that bridging was orderly. This prevented users from rapidly bridging GHO to Base, where there was limited liquidity upon launch. However, given the substantial increase in liquidity, it can be relaxed.
The rate-limited bridge influences arbitrage efficiency between chains. If GHO’s price deviates on Base vs Ethereum, arbitrageurs rely on the bridge to transfer GHO or reserves until prices realign. Tighter limits (status quo) mean large price discrepancies cannot be immediately equalized by a single trader moving funds; arbitrage must occur in chunks over time. In contrast, higher bridge throughput would let arbitrage capital flow faster, likely narrowing cross-chain price gaps more quickly.
Other Considerations
As noted, the current bridging delays impose an operational burden on users and make GHO less convenient relative to alternatives. Over time, this friction could slow GHO’s adoption on L2s or limit its market share in yield farms and trading, especially as new opportunities arise.
Higher bridge limits mean liquidity can flow to where it’s needed more freely. For example, if Base offers a better yield for GHO holders, more GHO can migrate there until yields equilibrate, benefiting GHO holders and ensuring no one chain has an oversupply or undersupply. This dynamic supports a more balanced multi-chain ecosystem for GHO. As GHO’s circulating market cap increases, having a bridge that scales accordingly will be important.
We must ensure, however, that liquidity remains robust on Ethereum (the source of GHO’s collateral backing). If limits are too high, there’s a theoretical risk that too much GHO could migrate out to side chains chasing yield, leaving shallow liquidity on the mainnet and making peg management there harder. In practice, this seems unlikely in the near term given Ethereum’s deep markets (GHO has >$85M liquidity on Ethereum DEXs).
Recommendation
Overall, depeg risk from a moderate increase in bridge limits is low – higher throughput should actually enhance cross-chain peg arbitrage. With GHO now proven on Base, it appears the benefits to liquidity and adoption outweigh the marginal increase in tail risk, especially if we opt for a conservative-to-moderate increase rather than an unbounded one.
Specifically, we recommend values 5x higher than the initial values, reflecting the vastly improved liquidity on Base.
Specification
| Parameter | Current | Recommended |
|---|---|---|
| Rate Limit | 300,000 | 1,500,000 |
| Refill Rate per Second | 60 | 300 |
Next Steps
We propose making these updates via the direct-to-AIP process.
Disclaimer
Chaos Labs has not been compensated by any third party for publishing this recommendation.
Copyright
Copyright and related rights waived via CC0
