Ordinary users on cross-chain bridges are more valuable than those on L2.
Author: Joel John
Compiled by: Luffy, Foresight News
The killer application of cryptocurrencies has emerged, and that is stablecoins. In 2023, Visa's transaction volume approached $15 trillion, while the total transaction volume of stablecoins reached $20.8 trillion. Since 2019, the total amount of stablecoin transfers between wallets has reached $221 trillion.
Over the past few years, capital equivalent to the global GDP has been circulating on the blockchain. Over time, this capital will gradually accumulate in different networks. Users switch between different protocols to obtain better financial opportunities or lower transfer costs; in the era of chain abstraction, users may not even know that they are using a cross-chain bridge.
Cross-chain bridges can be seen as routers for capital. When you access any website on the internet, there is a complex network behind it. The physical router in your home is crucial to the network, as it determines how to route data packets to help you get the data you need in the shortest time.
Today, cross-chain bridges play the same role in on-chain capital. When users want to move from one chain to another, the cross-chain bridge decides how to route the funds to allow the user's capital to obtain the maximum value or transaction speed.
Since 2022, the transaction volume processed through cross-chain bridges has exceeded $100 billion. This is far less than the amount of capital being transferred on-chain by stablecoins. But compared to many other protocols, cross-chain bridges earn more profit per user and per locked dollar.
The Business Model of Cross-Chain Bridges
Since mid-2020, blockchain cross-chain bridges have generated nearly $104 million in cumulative fees. This data has a certain seasonality, as users flock to cross-chain bridges to use new applications or seek economic opportunities. If there are no attractive yield opportunities, Memecoins, or financial primitives, the usage of cross-chain bridges will drop significantly, as users will stick to the chains and protocols they use most often.
A sad (but interesting) way to measure cross-chain revenue is to compare it to Memecoin platforms like Pump.fun. When the fee revenue of Memecoin platforms was $70 million, the fee revenue of cross-chain bridges was $13.8 million.
Although transaction volume has increased, we see fees remain relatively flat, as there is an ongoing price war between blockchains. To understand how they achieve this efficiency, we need to understand how most cross-chain bridges work. One mental model for understanding cross-chain bridges is through the Hawala network from a century ago.
Although today, people's understanding of Hawala is largely around its association with money laundering, a century ago, it was an effective way to transfer funds. For example, if you wanted to transfer $1,000 from Dubai to Bangalore in the 1940s, there were many options.
You could go to a bank, but it might take a few days and require a lot of paperwork. Or you could go to a seller in the Gold Souk, who would take your $1,000 and instruct a merchant in India to pay an equivalent amount to your designated recipient in Bangalore. The currency changed hands in India and Dubai, but it did not cross the border.
How did this work? Hawala is a trust-based system, and it worked because the sellers in the Gold Souk and the merchants in India often had ongoing trade relationships. They did not directly transfer capital, but settled the balance later with commodities like gold. Since these transactions relied on the mutual trust between the participating individuals, there had to be a high degree of confidence in the honesty and cooperation of both parties.
What does this have to do with cross-chain? Many cross-chains adopt the same model. To seek yield, you may want to move capital from Ethereum to Solana. The job of cross-chain bridges like LayerZero is to help convey messages about the user, allowing the user to deposit tokens on one chain and borrow tokens on another.
Imagine these two traders are not locking assets or providing gold bars, but giving you a code that can be redeemed anywhere. This code is a form of sending a message, which on LayerZero are called endpoints. They are smart contracts that exist on different blockchains. The smart contract on Solana may not be able to understand a transaction on Ethereum, so it needs an oracle. LayerZero uses Google Cloud as the verifier for cross-chain transactions, so even on the frontier of Web3, we rely on Web2 giants to help us build a better economic system.
Now, imagine the traders involved don't trust their ability to interpret the code, after all, not everyone can get Google Cloud to verify transactions, so what do they do? Another approach is to lock and mint assets.
In this model, if you use Wormhole, you will lock assets in an Ethereum smart contract and then receive a wrapped asset on Solana. This is akin to depositing US dollars in the UAE and having a Hawala provider give you gold bars in India. You can take the gold, speculate with it, and then return it to withdraw your original capital in Dubai.
The image below shows the wrapped Bitcoin assets today. Most of these were minted during the DeFi summer to generate yield on Bitcoin on Ethereum.
The key points of the cross-chain bridge business model are:
- TVL: When users come to deposit funds, these funds can be used to generate yield. Today, most cross-chain bridges do not absorb idle capital and lend it out, but rather charge a small transaction fee when users move capital from one chain to another.
- Relay fees: These are fees charged by third parties (such as Google Cloud in the case of LayerZero) that charge a small fee per transaction to verify transactions across multiple chains.
- Liquidity provider fees: Paid to those who deposit funds into the cross-chain bridge smart contract. Imagine you are running a Hawala network, and now someone wants to move $100 million from one chain to another. You personally may not have that much capital. Liquidity providers are the individuals who pool these funds to help facilitate the transaction. In return, each liquidity provider gets a small cut of the fees generated.
- Minting fees: Cross-chain bridges can charge a small fee when minting assets. For example, WBTC charges a 10 basis point fee for each Bitcoin minted.
The expenses of cross-chain bridges go towards maintaining relayers and paying liquidity providers, while the revenue comes from transaction fees and the value they create for themselves by minting assets for the transacting parties.
The Economic Value of Cross-Chain Bridges
The data below is a bit messy, as not all fees flow to the protocol. Sometimes, fees depend on the protocol and the assets involved. If cross-chain bridges are primarily used for low-liquidity long-tail assets, it may also result in users bearing slippage. So the following does not necessarily reflect which cross-chain bridges are better than others in terms of unit economics. We are interested in seeing how much value is generated across the entire supply chain during cross-chain events.
First, we look at 90-day transaction volume and the fees generated across protocols. The data is as of August 2024, so these numbers are for the subsequent 90 days. We assume Across has higher transaction volume because it has lower fees.
This roughly indicates how much capital flowed through cross-chain bridges in a given quarter and the types of fees they generated in the same period. We can use this data to calculate the fees generated per $10,000 transferred through these cross-chains.
Before we begin, I want to clarify that this does not mean that the fees charged by Hop are ten times higher than Axelar. Rather, it means that on a cross-chain bridge like Hop, a $10,000 transfer can create $29.2 in value across the entire value chain (for LPs, relayers, etc.). These metrics differ because their transfer nature and types are different.
The most interesting part for us is to compare it with the value captured by the protocol and the amount of cross-chain capital.
To benchmark, we looked at the transfer costs on Ethereum. As of the time of writing, the transfer cost on Ethereum is around $0.0009179 with low gas fees, and the transfer cost on Solana is around $0.0000193. Comparing a cross-chain bridge to L1 is a bit like comparing a router to a computer. The cost of storing a file on a computer will be significantly lower. But what we're trying to solve here is whether cross-chain bridges capture more value from an investment perspective than L1.
From this perspective, and referring to the metrics above, one way to compare the two is to look at the dollar fees charged per transaction by each cross-chain bridge and compare them to Ethereum and Solana.
The reason why several cross-chain bridges capture fees lower than Ethereum is due to the Gas costs incurred when making cross-chain transactions from Ethereum.
Someone might say that the Hop protocol captures 120 times more value than Solana. But that misses the point, as the fee models of the two networks are completely different. What we're interested in is the difference between value capture and valuation.
Out of the top 7 cross-chain protocols, 5 have fees cheaper than Ethereum L1. Axelar is the cheapest, with an average fee over the past 90 days of only 32% of Ethereum. Hop Protocol and Synapse are more expensive than Ethereum. Compared to Solana, we can see that the L1 settlement fees on high-throughput chains are orders of magnitude cheaper than today's cross-chain protocols.
One way to further strengthen this data is to compare the cost of transacting on L2s within the EVM ecosystem. Generally, Solana's fees are around 2% of Ethereum's. For comparison, we'll use Arbitrum and Base. Since L2s are designed for low fees, we'll use a different metric to measure economic value - the average daily fee per active user.
Over the 90 days we collected data for this article, Arbitrum had an average of 581,000 users per day, generating $82,000 in fees per day. Similarly, Base had 564,000 users and generated $120,000 in fees per day.
In comparison, cross-chain users are fewer and fees are lower. The highest is Across, with 4,400 users generating $12,000 in fees. From this, we estimate Across generates $2.4 in fees per user per day. This metric can then be compared to the fees generated per active user on Arbitrum or Base to measure the economic value per user.
Today, the average user on a cross-chain bridge is more valuable than the average user on an L2. Connext's average user creates 90 times the value of an Arbitrum user.
- Cross-chain bridges, as monetary routers, may be one of the few product categories in crypto that can generate meaningful economic value.
- As long as transaction fees remain high, we may not see users migrate to L1s like Ethereum or Bitcoin. Users may directly join L2s like Base, or we may see users switching only between low-cost networks.
Another way to measure the economic value of cross-chain protocols is to compare them to decentralized exchanges. Think about it - these two primitives have similar functionality. They allow tokens to be transferred from one form to another. Exchanges allow the transfer of assets between assets, while cross-chain bridges allow the transfer of assets between blockchains.
The data above is only for decentralized exchanges on Ethereum.
Here I avoid comparing fees or revenues. Instead, I'm interested in capital turnover rate. It can be defined as the number of times capital flows between the smart contracts held by the cross-chain bridge or decentralized exchange. To calculate it, I'll divide the daily transfer volume of the cross-chain bridge and decentralized exchange by their TVL.
As expected, for decentralized exchanges, the velocity of money is much higher, as users trade assets back and forth multiple times in a day.
However, interestingly, when you exclude large L2 native bridges (like Arbitrum or Optimism's native bridges), the velocity of money is not too far off from that of decentralized exchanges.
Perhaps, in the future, we'll see cross-chain bridges that limit capital allocations and instead focus on maximizing returns by increasing the capital turnover rate. That is, if a cross-chain bridge can transfer capital multiple times in a day, it will be able to generate higher returns.
Are Cross-Chain Bridges Routers?
Source: The Wall Street Journal
If you think the influx of venture capital into the "infrastructure" space is a new phenomenon, let's take a trip down memory lane. Back in the 2000s, when I was just a young boy, many in Silicon Valley were raving about Cisco. The logic then was that as internet traffic increased, routers would capture a significant portion of that value. Much like NVIDIA today, Cisco was a high-priced stock at the time because they were building the physical infrastructure that supported the internet.
Cisco's stock peaked at $80 on March 24, 2000. As of the time of writing, it trades at $52, never having recovered to its heyday. Writing this article in the midst of the Memecoin frenzy has made me ponder how much value cross-chain protocols can capture. They have network effects, but it may be a winner-take-all market. This market is increasingly trending towards intent and solutions, with centralized market makers completing orders on the backend.
At the end of the day, most users don't care about the decentralization of the cross-chain bridge they use, they care about cost and speed.
The cross-chain bridge space has seen some early maturation, with multiple approaches emerging to solve the age-old problem of cross-chain asset transfer. The key driver of change is chain abstraction: a mechanism for cross-chain asset transfer where the user is completely unaware that they have transferred assets.
Another factor driving volume growth is innovation in product distribution or positioning. Last night, while exploring Memecoins, I noticed that IntentX is using intent to package Binance's perpetual contract market into a decentralized exchange product. We're also seeing cross-chain bridges specific to certain chains evolve to strengthen the competitiveness of their product.
Whichever approach is taken, it's clear that, like decentralized exchanges, cross-chain bridges are hubs through which massive amounts of monetary value flow. As a primitive technology, they will continue to exist and evolve. We believe that specialized cross-chain bridges (like IntentX) or user-specific cross-chain bridges (like those enabled by chain abstraction) will be the main drivers of growth in this industry.
Shlok added a nuance in discussing this article, which is that past routers never captured economic value based on the data they transmitted. You could download TBs or GBs, and Cisco would make the same money. In contrast, cross-chain bridges make money based on the number of transactions. So their fate may not be the same.
For now, it can be said with certainty that the situation we see with cross-chain bridges is analogous to the physical infrastructure of routing data on the internet.
