Why does bridging matter?

By now, it’s clear that the future is multichain and omnichain. For example, L2 Beat now tracks nearly 60 active networks, with TVL having increased 5x over the past year alone.

As more chains emerge, the need for seamless communication between networks becomes more important. From a user’s perspective, they shouldn’t have to waste time and energy moving assets from chainA to chainB by navigating through several apps, signing a handful of transactions and clicking far too many buttons in the process. So how do we fix this?

Eventually, users won’t know which chain they are transacting on as DApps and other platforms abstract these layers away. While in the short term, protocols like Stargate and Across are making it easier, cheaper, and faster for users to actively bridge assets from one network to another.

Stargate vs. Across

Among today’s leading bridges, both Stargate and Across stand out in terms of transaction volume. Last week, Stargate V2 facilitated the transfer of over $160M in assets according to their dashboard. Whereas during the same period, Across saw nearly $205M in bridging volume according to DeFi Llama.

With the recent launch of Stargate V2, a few of us have taken to Twitter to debate over which of the two bridges is cheaper and faster from the user’s perspective.

While it may be too soon to arrive at any sort of definitive conclusion, data has started to roll in. In the meantime, I am hoping to contribute to this discussion, with the remainder of this post dedicated to analyzing which bridge is cheaper for smaller-sized stablecoin transfers between Ethereum L2s.

Test 1: moving USDC from Base to Arbitrum

Over the weekend, I conducted a series of L2-to-L2 transactions on both Stargate V2 and Across V3. My objective was straightforward - to determine which platform offered lower transaction costs for stablecoin transfers ranging from $25 to $200.

Before we discuss the results, let’s get one thing out of the way. With a sample size of only four, my observations are not conclusive. My aim is simply to share some basic data to foster discussion. That said, here are the results, followed by a few takeaways that I have observed. You can also view detailed calculations and transaction hashes here.

1. Onchain > Offchain

First, it’s worth noting that its extremely cheap to bridge assets from one L2 to another using both Stargate and Across – the average transaction fee across the sample was only 0.13% or 13 basis points.

Compared to sending a wire transfer of $1,000 from one financial institution to another, L2-to-L2 bridges are approximately 20x cheaper.

2. Stargate Charges a Lower Protocol Fee

By protocol fee, I am referring to the transaction cost charged by the protocol and split between stakers / relayers and the DAO. I calculated these amounts by comparing the quantity of USDC deposited on the source chain to the amount that eventually arrived on the destination chain. For example, if I transferred 25 USDC from Base and received 24.9 USDC on Arbitrum, the difference is the protocol fee.

In all cases, transfers initiated via Stargate left more residual USDC on the destination chain compared to those initiated via Across. I think we can attribute this to Stargate simply charging a lower fee than Across for its bridging services.

3. Bridging with Across Consumes Less Gas

As part of Stargate V2, new measures were introduced to reduce the amount of gas required at both the Stargate and LayerZero levels of the transaction. This included migrating Stargate’s liquidity pool rebalancing actions offchain, while also enabling users to split gas costs through a new transaction batching mechanism.

However, across all test transactions, bridges initaited via Across still consumed less gas than those that passed through Stargate. For example, when I transferred 50 USDC using both platforms, Stargate consumed 121,396 units of gas while Across required only 78,856.

4. There’s Some Sort of “Extra” Fee on Stargate

I noticed something interesting when looking further into the transaction data on Etherscan. When bridging via Stargate, not only did I need to pay gas on the source chain, but I also had to send a small amount of ETH along with my USDC to the destination chain. See the image below for an example.

The quantity of ETH that was included in the transfer was generally consistent across all Stargate-based transactions. Whether I sent 25 USDC or 200 USDC, I also had to send approximately 0.0000006 ETH along with the stablecoin.

5. For Transfers Over 100 USDC, Stargate Was Cheaper

While Across offered a lower total transaction cost for transfers of 25 and 50 USDC, Stargate was cheaper for both 100 and 200 USDC transfers. Can you guess why?

Well, I believe it has to do with the fixed cost component that was discussed above. For both transfers in which Stargate was more expensive, the dollar value of the 0.0000006 ETH was materially higher than the actual gas cost required to process the transaction.

Closing Thoughts

As mentioned earlier, it’s likely still too early to determine which platform is more cost-effective for bridging transactions.

For example, as more users start to bridge on Stargate V2, their new transaction batching mechanism has the potential to lower costs even further. Essentially, the Stargate V2 flywheel could function like this:

• More users bridge via Stargate, choosing a bus over a taxi for their transactions. This results in transactions being batched, with messaging costs spread across the group. Consequently, this reduces the gas costs associated with bridging on Stargate, making it cheaper to use than competitors, and the cycle continues.

Could a similar dynamic unfold on Across? If not, it’s hard to see how Stargate won’t increasingly dominate the market over time.