Solver economics, in plain English

Solvers are not altruists. They are not running a public service. They are profit-maximizing firms — sometimes literal firms, sometimes one person with capital and an algorithm — that have decided being a solver is a better use of their balance sheet than the alternatives. The intent settlement architecture works because that calculation pencils out. If it stopped penciling out, the solvers would leave, and the system would stop working.

This post is the economic engine, in plain English. What a solver is, how they make money, what their risks are, and why the structure of the auction keeps the prices users see roughly honest. Companion pieces: intent settlement vs bridges and how 1Click works.

The TradFi analogue (for orientation)

Before walking through what a solver is, it helps to point at what they are not new. Solvers are the latest entry in a long lineage of professional intermediaries who hold inventory and quote prices to people who want to trade.

An OTC FX dealer at a bank in London does this. They hold EUR and USD and JPY in their book; a corporate treasurer phones in to swap fifty million of one for the other; the dealer quotes a price slightly worse than the wholesale mid for the corporate, slightly better than the mid for their own next hedge trade, and pockets the difference. The spread is the rent on inventory and balance sheet. Multiply by millions of trades a day across thousands of dealers and you have a functioning FX market.

A NYSE specialist did the same thing for stocks until about 2008. A market maker on a modern CEX does the same thing for crypto. A solver in an intent settlement protocol is doing the same thing across chains. The mechanism is new. The economics are old.

What a solver actually owns

A solver running on NEAR Intents needs inventory on every chain and in every asset they want to quote prices for. If they intend to fill BTC-to-USDC-on-Arbitrum trades, they need to hold native USDC on Arbitrum (to pay the user) and be willing to accept native BTC (which they will receive from the user's deposit). They also need gas tokens on every chain they operate on — ETH on Arbitrum, ATOMs or NEAR or whatever, BTC for the change output.

This is the first thing to internalize: solver balance sheets are not virtual. They are real positions in real assets on real chains, accumulated by deploying real capital. A solver with a hundred million dollar book has a hundred million dollars of risk distributed across some portfolio of chains and assets. That portfolio is constantly rebalancing as trades come in.

How they make money

Two ways, in roughly equal measure.

The spread. When a solver quotes a price on BTC-to-USDC, they quote it slightly worse than what they could get by selling the BTC immediately on a deep CEX. The user pays this premium implicitly; it is the price of getting the trade done atomically and non-custodially. On a deep, liquid pair like BTC-to-USDC, the spread is small — single-digit basis points. On a thinner pair — some obscure L2 stable to a small-cap token — the spread can be much larger.

The hedge. A sophisticated solver does not sit on the BTC they just received. They simultaneously hedge it — sell BTC futures on a CEX, or instantly trade out of it on a DEX, or earmark it as inventory for the next opposite-direction trade. Done well, the hedge locks in the spread and offloads the price risk. Done badly, the solver gets stuck holding inventory that moves against them.

What their risks are

Three risks dominate, and a solver who does not manage them goes out of business inside a year.

Inventory risk. Between the moment a solver commits to a quote and the moment they finish hedging it, the underlying price can move. If a solver quotes a BTC-to-USDC trade at a market price of $100,000 and BTC drops to $99,500 before they can hedge, the spread they earned is now negative. Sophisticated solvers manage this with very tight hedge latency and inventory limits; less sophisticated ones blow up in fast markets.

Route risk.A solver who has quoted a cross-chain trade is exposed to the source chain's confirmation behavior. If a Bitcoin trade is quoted and the user's deposit takes thirty minutes to confirm (slow blocks, fee market spike), the solver's hedge window is much longer than they wanted. Solvers price Bitcoin trades wider than EVM-to-EVM trades for exactly this reason.

Default risk on themselves.A solver who cannot deliver — software bug, inventory exhausted, gas wallet drained — gets slashed. The slashing mechanism is the protocol's way of making solver default expensive. The solver posts collateral upfront; if they win an auction and fail to settle, the collateral is partly or fully taken. This is the economic glue that keeps the system honest. A solver who can default cheaply will default whenever it is convenient; a solver whose default costs them ten times their expected profit on a trade will not default unless something is genuinely broken.

Why competition produces honest prices

The structural reason intent settlement works at all is the auction. Every trade is quoted by multiple solvers simultaneously. The protocol picks the best price for the user. Solvers therefore compete on price, not on relationship, not on UI placement, not on which Telegram group the user is in.

Auction theory has a result that is relevant here, due roughly to Wilson and Milgrom: in a sealed-bid auction with many participants and similar information, the winning bid converges to the bidder's true cost plus the minimum sustainable spread. The intent protocol's quote auction approximates this. As long as there are enough solvers and they have roughly similar costs (which they do, because they are all running similar hedge infrastructure against the same CEXes and DEXes), the user receives a price close to the true wholesale price plus the rent on capital and the cost of the protocol's operational overhead.

The interesting failure mode is not collusion; collusion is hard to maintain in an anonymous, programmatic auction where new entrants can join cheaply. The interesting failure mode is too few solvers on a given corridor. If only one solver is willing to quote BTC-to-some-obscure-L2, that solver is a monopolist and can charge accordingly. This is why solver count per corridor is a metric worth watching.

Why solvers bother with cross-chain at all

A sophisticated trading firm has many places to deploy capital. Why intent settlement instead of, say, market making on Binance?

Two reasons, in current market conditions.

First, the spreads are wider than on a deep CEX, because the corridors are thinner and the operational complexity is higher. A firm with cross-chain inventory and good hedge infrastructure can earn more per unit of capital deployed in intent solving than in vanilla CEX market making. This will compress over time as more solvers enter, but the premium is real today.

Second, intent solving captures a flow type that CEX market making does not — the user who refuses to deposit to a CEX. That flow is qualitatively different (smaller, more episodic, less informed about short-term price action) and is therefore more profitable per dollar of notional. Solvers are happy to pay for access to it.

The honest takeaway

The case for intent settlement, economically, comes down to one observation. There exist professional traders with cross-chain capital who are happy to be your counterparty on a swap, take their basis-point spread, and bear the inventory risk you don't want to bear. The architecture gives them a way to do this without holding your funds and without you holding theirs. The protocol takes a small piece. Everyone is better off than under the bridge-and-wrap status quo, where the same flow was intermediated by a multisig that kept getting hacked.

Solvers are the unglamorous, load-bearing piece of all this. The user experience is "type, click, receive." The economic substrate is "competitive market making across chains with on-chain enforced atomicity." The gap between those two sentences is the entire product.