What Is a Crypto Bridge?
Blockchains cannot see each other, so bridges lock coins on one chain and issue stand-ins on another. That pot of locked coins is why bridges have lost more money than almost anything else in crypto.
TL;DR
- Nothing actually crosses. A bridge locks your coins on one chain and mints a representation of them on another, which you swap back to release the originals.
- The locked pot is a permanent target: Ronin lost 625 million dollars in March 2022 when attackers obtained validator keys.
- Wormhole lost 320 million the month before through a signature verification bug, and Nomad 190 million in August 2022 after an upgrade left everything looking pre-verified.
- Most buyers never need one. Buying directly on the chain you want avoids the entire category of risk.
Blockchains are islands, and Bitcoin has no idea Ethereum exists, cannot read its ledger and has no mechanism for learning about it. There is no shared clock or shared state between them, and no protocol by which one would trust the other's word about anything.
Bridges are the workaround, and they have cost more money than almost any other single category of crypto infrastructure. This is an educational guide, not financial advice.
Nothing crosses anything
The name misleads: your coins do not travel to another chain, because there is nowhere for them to travel to.
What happens is a swap dressed as a journey. You send coins to a contract or a custodian on the first chain, where they are locked and left. The bridge then issues you an equivalent token on the second chain, a representation redeemable for the original. When you want to go back, you hand in the representation, it is destroyed, and the locked coins are released.
Everything therefore depends on one thing: who or what decides that the lock really happened. Answer that and you have described the bridge's entire security model, because the token on the far side is only worth anything while the locked coins remain unclaimed by anybody else.
Why the losses are so large
A bridge accumulates, and every user who crosses adds to a pot that sits in one place, does not move, and is worth exactly as much as everything anyone has ever bridged. It is the most attractive target in crypto by construction, and unlike an exchange it has no staff watching balances hour by hour.
The three canonical failures each broke a different part of the machine, which is instructive.
Ronin, March 2022, 625 million dollars. The bridge relied on a set of validators approving withdrawals. Attackers obtained enough of their keys, partly through a spear-phishing approach dressed as a job offer, and approved their own withdrawal. The cryptography worked perfectly, so the weak point was the people holding the keys, and the US Treasury later attributed the attack to North Korea's Lazarus Group.
Wormhole, February 2022, 320 million dollars. A signature verification bug involving a deprecated Solana function meant the contract could be convinced a deposit had been signed off when it had not. The attacker minted tokens on one side with nothing locked on the other. Jump Crypto, which backed the bridge, replaced the missing funds from its own treasury within a day.
Nomad, August 2022, 190 million dollars. A routine contract upgrade left a trusted root set to zero, which had the effect of making every message look already-verified. The exploit needed no skill to copy, so hundreds of people repeated it, which made it the strangest of the three: a crowd-sourced robbery.
Poly Network in August 2021 was larger still at around 612 million, and almost all of it was returned within a fortnight after the attacker chose to give it back, which is not a security model anybody should rely on.
Kinds of bridge, roughly by how much you must trust
Custodial. A company holds the coins and issues the wrapped version, and you are trusting that company entirely, in the same way you trust an exchange.
Validator or multisig. A defined group signs off on transfers, and security depends on how many keys are needed and how well those holders protect them, which is precisely what Ronin got wrong.
Contract-verified. The receiving chain checks proofs about the sending chain, stronger in principle because it removes the human committee, and it moves the risk into code that must be correct, which is where Wormhole and Nomad failed.
Native issuance. Not a bridge at all: the issuer mints the same asset directly on several chains, which is how USDC works across networks and why moving it that way avoids this whole category.
How to avoid needing one
This is the part most articles skip, and for ordinary buyers it is the answer.
If you want ether on a layer 2, or USDC on Solana, buy it on that network in the first place. On-ramps let you choose the destination network at purchase, so the coins arrive where you want them without a bridge ever entering the picture. You will usually pay less than the bridge would have cost, and you carry none of the risk described above.
Bridging is for coins you already hold in the wrong place. If you must, prefer the official bridge for that network over a third party promising better rates, move a small test amount first, and do not leave value parked in a wrapped token any longer than the task requires. Wrapped assets are a claim on a bridge, and the bridge is the thing that fails.
Frequently Asked Questions
Your coins are locked on the first chain and a representation of them is issued on the second. Nothing moves between chains, because chains cannot see each other. Going back destroys the representation and releases the originals, so the whole arrangement depends on that locked pot staying secure.
Because they accumulate. The locked pot equals everything anyone has ever bridged, sits in one place and does not move, which makes it the most attractive target in crypto. Ronin lost 625 million dollars in March 2022, Wormhole 320 million, Nomad 190 million, each through a different weakness.
Almost never. On-ramps let you pick the destination network at the point of purchase, so buying USDC on Solana or ether on a layer 2 delivers the coins where you want them directly. That is usually cheaper than bridging and avoids the entire risk category.
No. It is a claim on the bridge that issued it, redeemable while that bridge remains solvent and uncompromised. If the bridge is drained, the wrapped token can keep trading on the far side while nothing backs it. Treat wrapped assets as a position in the bridge, not just in the coin.
Prefer the official bridge for the network you are moving to over a third party offering better rates, since official bridges are more heavily reviewed and their failure is the network's problem too. Send a small test amount first, and do not leave funds sitting in wrapped form once the task is done.
About the author — Dan Clarke
Dan Clarke is the author of Bitcoin: The Complete Guide and a former content lead at Binance Academy, where he wrote crypto education for readers arriving with no background in the subject. He has worked in the cryptocurrency industry since 2017. His rule for these guides: plain language first, precision where it matters, no cheerleading.