Should we burn the bridges?

How has the economics of cross-chain bridges changed of late, and what has broadened their functionality? Web3 entrepreneur Vladimir Menaskop shares with ForkLog readers the findings of a large study of more than 50 projects and sets out the authors’ preliminary conclusions.

From the author

ForkLog has revisited this topic in surveys, features and podcasts. Yet bridges are not all alike; time changes everything; and the “bridge-building” of 2024 has moved far beyond the pitfalls of years past.

In the age of LayerZero, we dive into research carried out by the DAO Synergis community and 100DAIDAO to surface several trends that tell an important story.

Special thanks to I. Sandakov (aka Dr. Goldfinger) for help in preparing this piece.

Methodology

“Methodology” may sound dull, but without it any conclusions are merely opinion. A set of scientific methods helps separate wheat from chaff and identify, to a degree, objective vectors.

To learn anything useful one must analyse or synthesise data. Here, analysis is apt. And what is analysis across shared criteria if not classification?

Bridge taxonomy

A non-exhaustive list follows:

• By EVM compatibility:
  • EVM-compatible;
  • non-EVM;
  • hybrid.

• By deal size:
  • small (up to 0.1 ETH);
  • medium (0.1–1 ETH);
  • large (1–10 ETH);
  • very large (over 10 ETH, now rare).

• By product scope:
  • pure bridges (bridging only);
  • mixed (bridge, swap, and so on).

• By specialisation:
  • universal;
  • for gas transactions;
  • NFT bridges;
  • other (including hybrids).

• By network affiliation:
  • native bridges;
  • non-native;
  • other.

Visually, it looks like this:

Any bridge can of course fall into several categories: it might be EVM-compatible, handle small volumes, and try to capture niches beyond pure bridging (Owlto comes close).

What was studied, and how?

The research covered the following components:

1. Extended search, including through non-personalised search engines.
2. Aggregators: CoinMarketCap, CoinGecko, DeFi Llama and others.
3. Ecosystems, especially newer blockchains, where overlaps are hardest to find.
4. Forums: new bridges often appear there first, making it easier to spot strengths and weaknesses.
5. Telegram channels—both Web3-focused and those covering narrower crypto niches.
6. Analytical reviews by Messari, Nansen, Synergis and other agencies and services.
7. Curations by members of the Synergis community and related DAOs.

An extended study with a wealth of empirical material will be published on ForkLog at the end of 2024. For now—some salient observations.

L0, atomic swap or bridge?

Talking about bridges five or six years ago is one thing; discussing them today is quite another. Technology has raced ahead: not only do we have L3, but also a full-fledged L0; L2s have all but swallowed Ethereum, then Bitcoin, and much else besides.

Rather than dwell on minutiae, we look across the full spectrum of bridge operations and segment them along the lines above.

Facts and figures

The study examined 213 bridges; a little over 50 made the final sample. The main reasons for exclusion were:

1. Lack of own infrastructure.
2. Unstable operation or negative user feedback (suspected scams).
3. Bridge hacks with no subsequent recovery.
4. Support only for testnets.
5. Insufficient time to gather initial data.

Further quantitative notes:

• around 10% of the final sample runs a referral programme—a surprise, given that loyalty schemes built on referral links fit neatly into bridge tokenomics, provided volumes are realistic;
• many bridges (over 25%) pay out rewards, primarily funded by ecosystems (Arbitrum, Optimism);
• the vast majority (over 90%) support a limited set of networks, which they expand slowly; only a few, such as Celer, keep broadening chain coverage at pace;
• roughly two or three out of ten bridges (depending on testing time and ticket size) lack sufficient liquidity to complete a swap. Causes range from congestion—eg, after Blast’s 2024 airdrop—to unpopular routes that are not disabled automatically;
• over 20% of all bridges studied, and north of 10% of the final sample, were hacked (in one form or another) yet continued operating—something that typically demands time, loss recovery and capable community management.

The largest hacks include:

• Ronin: losses over $625m;
• BinanceHub: losses over $586m;
• Wormhole: losses over $319m;
• Nomad: losses over $200m;
• Horizon: losses over $100m.

Many other bridges suffered no less from hacks; their liquidity and the attack windows were simply smaller, so the incidents drew less press.

Beyond the numbers, several qualitative conclusions emerge.

First, many bridges have moved away from wrapped assets and now operate with native tokens and coins. Even some leaders still cannot break with wrappers. Use a bridge to swap, say, MATIC, and you might receive an illiquid instrument on a thinly traded network—getting it out becomes a matter of pride rather than practice.

A simple tip: experiment, but start small, scale up gradually, and factor in transaction costs.

Second, because the research employed extended (deep) search, note that in the past three to four years many bridges have adopted the high-frequency keyword “bridge” in a subdomain, domain, or in a base path (/bridge) of the dapp.

This greatly simplifies discovery of decentralised apps and bridges in new networks, including testnets. The method is not for neophytes (given the perils of Web 3.0 and Web3, including phishing), but is acceptable for experienced users.

Third, alongside the blockchain trilemma one can point to a bridge trilemma, which forces a trade-off between speed, security and cross-chain reach.

There are many ways to frame the issue; four stand out:

1. Native bridges are safer but slow.
2. Fast bridges usually specialise (eg, L2 EVM).
3. Native tokens are fine but demand deep liquidity—often the very cause of hacks.
4. Users stop noticing differences between bridges, with the unhappy outcomes noted above.

Fourth, more bridges are adding specialised tracks; three are notable:

1. NFT transfers (mostly within EVM and for limited collections so far).
2. Gas exchanges.
3. Swaps routed through a project’s own token (often far from the most liquid).

Fifth, amid the restaking trend many bridges added ETH wrappers (stETH, wstETH, rsETH) to pairs—yet the exchange ratio is not always what you expect.

Another tip: handle such routes carefully and in stages, comparing the cross-rate to the underlying projects’ prices.

Bridges here fall into the following groups:

• gas exchangers;
• bridges specialising in L1–L2 on/off-ramps (rHino, Orbiter);
• connectors across disparate chains (Wormhole, Core);
• native;
• for NFTs (solutions by NFTscan, Envelop);
• for derivatives;
• other.

Sixth, the number of native bridges has risen over the past three to four years because:

1. There is a profusion of L2—and even L3—solutions.
2. Large DAOs are moving to their own networks. Beyond the obvious Arbitrum and Optimism, think Mantle, Blast and many others.
3. Non-EVM networks, and their bridges, are multiplying. As Solana grows, another rival to Ethereum—Avalanche with its core.app—has carved out a decent niche in DeFi.
4. Many bridges now handle transfers of native bitcoin as well as wrappers (wBTC and the like).
5. Web3 entrepreneurs see bridges as a clear, accessible niche as OTC and other p2p venues close worldwide and exchangers and CEX face regulatory pressure. Bridges can also issue a token to attract liquidity, alongside standard staking and farming.

The core development window for today’s bridges is 2021–2024; the formative years were roughly 2018–2020. We now see direct competition between innovative (LayerZero) and quasi-classical approaches simply because the industry is young and has yet to settle on stable models.

Bridges look like a transitional species in this hierarchy—albeit a vital one. Hence the headline’s answer is simple: no, it is too soon to burn bridges; given Web3’s defining property, that time is unlikely ever to come.

Outlook

A few prospectively important points stand out.

First—tightening cross-chain compatibility within Web3 projects. Look closely at the TON hype and the playbook is clear from the start: 1) introduce stablecoins into the ecosystem; 2) bridge to EVM; 3) lure developers with grants.

EVM is present in almost every ecosystem: Cosmos, Polkadot, EOS, XRP Ledger, Solana, Near, Bitcoin (bEVM) and many others.

That mattered when liquidity lived either in Bitcoin or in Ethereum. Now Solana has climbed to fourth place, TON has long sat in the top ten, and the top 30 counts Cosmos, Cardano and Polkadot.

Cross-chain compatibility now resembles a kind of quantum entanglement—everywhere at once, and nowhere in particular. Expect more complex solutions, including in non-EVM settings—complex on the back end, simpler for users.

Second. Talk of bridges usually means bridging native tokens or fungible tokens (ERC-20 and the like), less often NFTs (typically ERC-721 and similar), and only exceptionally anything else.

Even in EVM land, new “standards” like ERC-404 and new approaches (SFT, omnichain tokens) have arrived—needed also for segments such as RWA.

Third. Account abstraction is charting new directions: TON is pushing gasless transactions, while Rabby makes them a reality in EVM. Bridges will bend further that way under the wind of mass adoption.

Fourth. We are about to see what giants look like on this market—and how they coexist. On one side, a fierce fight is raging in zkEVM (zkSync, Scroll, StarkNet). On the other, few know how to take on Wormhole, LayerZero and their ilk. Niching is the obvious answer, though not one everyone will like.

Supermarkets once squeezed kiosks but did not kill them; markets remained their own kind of commerce. Expect much the same over the next three to five years in crypto.

Fifth. Another important vector: bridges in private DeFi. Interface-level account blocks are growing more common, and rising Tornado Cash liquidity suggests not everyone accepts a regime where “all are equal, but some are more equal than others.” This is a separate issue—fascinating for research and for DeFi practice alike.

Sixth. Bridges face many direct and indirect competitors:

• OTC venues and P2P deals generally;
• exchangers;
• CEX—many newcomers use them for bridging;
• multichains—Polkadot’s parachains, Cosmos hubs, Avalanche subnets and others are built to retain value inside their ecosystems;
• various innovative L0 solutions.

Bridges can be used inside those services—or standalone.

Seventh. Bridges ultimately compete with wallets. Gnosis Safe and Rabby, and many others after them, have introduced gasless transactions and gas payment in liquid tokens. Swaps inside MetaMask and peers are the next rung, with cross-chain only a short reach away. Several wallets already offer it, mostly via API aggregators.

Whether “bridging” remains a niche or becomes an inseparable part of broader infrastructure is a fork the industry will take in the next few years.

Read the remaining conclusions in the final study at the end of 2024.