The number is $800 billion. The other number is zero.
1inch has routed roughly $800 billion in cumulative trading volume across its aggregator stack, and its co-founder โ unnamed in the coverage, which itself is a signal โ recently confirmed the protocol still cannot turn a profit. DeFi, he says, is still too small.
Stop there. Two data points, and they contradict the standard growth narrative every treasury deck has been selling for four years. If $800 billion in routed flow cannot produce a positive margin, the problem is not user demand. The problem is architecture. Volatility is noise. Architecture is the signal.
I have spent most of 2024 inside aggregator contracts โ reading routing logic, mapping fee hooks, tracing where value actually settles after a swap executes. The 1inch admission is not a surprise to anyone who has opened the bytecode. It is a confirmation of something the aggregator model has been quietly hiding behind volume charts.
Let me be precise about what a DEX aggregator is, because the mechanics determine the economics, and the economics determine the profit answer. An aggregator does not hold liquidity. It holds a router. A user submits a swap, the router queries multiple underlying venues โ Uniswap, Curve, Balancer, hundreds of long-tail AMMs โ splits the order across paths, and settles the fill. The product is execution quality. The unit of value is basis points of price improvement.
The business model is a fee on top of that improvement. Usually 0.1% to 0.875% depending on chain and configuration. On a $10,000 swap, that is $10 to $87. Sounds healthy. It is not, because of what sits underneath.
When I decompiled the router logic in 2019 โ sitting with Ethervm.io and a partially reconstructed interface โ the thing that struck me was not the pathfinding. It was the absence of a value-retention layer. The router was a passthrough. It touched liquidity, moved it, and exited. Every dollar of profit in the transaction went to the liquidity provider who supplied the pool, or to the arbitrageur who closed the price gap after the fill. The aggregator collected a toll and immediately paid it back out as operational cost: gas rebates, keeper incentives, referral fees to integrating wallets.
That structure has not fundamentally changed. It has only scaled.
Here is the arithmetic the $800 billion headline hides. Aggregators route, they do not custody. Routing generates gross fee revenue, but the cost base is brutal and fixed. You need continuous quoting infrastructure across every supported chain. You need an intent-solver network or a routing mesh that runs 24/7. You need contract deployments and audits on a dozen networks. You need frontend maintenance, indexer infrastructure, RPC bills that scale with query volume, and a security team that never sleeps. None of that cost curves down when volume spikes. It curves up.
Now add the competitive ceiling. Aggregator fees are visible, comparable, and elastic. If 1inch charges 0.5% on a route and a wallet-native swap charges 0.25% silently, the user does not see the difference โ the wallet embeds the router, and the router competes on a single variable: does the fill land cheaper. That means the aggregator cannot raise its take rate without losing the route. The route is the product. The take rate is set by the market, not by the protocol.
This is the structural trap. Cumulative volume is not revenue. Revenue is volume multiplied by capture rate, and the capture rate in aggregation is structurally near zero because the aggregator sits between two parties who both extract more than it does.
I ran the pattern against a smaller experiment in 2020 during the liquidity mining chaos. I had a Python monitor running against Balancer V2 vaults, watching gas patterns and pool rebalancing. What I found then generalizes now: in high-velocity environments, the entities that capture real value are the ones holding the inventory or the information asymmetry. Market makers capture spread. Arbitrageurs capture latency. Liquidity providers capture fees. Aggregators capture flow โ and flow is a liability, not an asset, because you have to keep paying to serve it.
So when the 1inch co-founder says DeFi is 'too small to turn a profit,' read the sentence carefully. He is not saying there are not enough users. He is saying the value available to be captured at the aggregation layer is thinner than the cost of capturing it. That is an architectural statement, not a market-cycle statement.
We didn't see this because the industry spent four years measuring the wrong variable. TVL. Volume. Users. None of those are profit. They are throughput. And throughput without margin is a public utility running on venture capital and token emissions.
Now the contrarian angle, and this is where the standard coverage misses it entirely. The consensus read on 1inch is that it is a category leader facing a temporary profitability gap โ a good product in an immature market, waiting for scale to fix the math. That read is wrong, and it is wrong for a technical reason, not a market one.
Aggregators face a security and design blind spot that directly caps their upside: they have no proprietary liquidity, therefore no proprietary pricing power, therefore no defensible margin. Their only moat is routing efficiency, and routing efficiency is asymptotically close to free. Any competent team can replicate pathfinding. The intents-based solvers โ UniswapX, CoW Protocol's batch auctions, and the emerging solver networks โ are already commoditizing the exact function 1inch built its brand on. When the routing layer becomes a public good, the toll booth has nothing to tax.
Worse, in a bull market โ which is exactly where we are sitting right now โ this problem gets masked. Volume swells. Headlines print bigger cumulative numbers. Everyone assumes the missing profitability is a timing issue. It is not. It is a capture issue. And in the next drawdown, when volume compresses by 70% and the fixed cost base of multi-chain routing infrastructure does not, the gap widens rather than closes. That is the part nobody prices in, because bull-market narratives do not model cost rigidity.
There is a second blind spot, and it is regulatory rather than economic. If 1inch is positioned as a user-facing swap gateway โ the front door where retail flows enter DeFi โ it is functionally a trading service provider. MiCA, and the various national implementations now tightening across the EU and UK, do not care that the matching happens on-chain. They care who the user interfaces with. A compliant aggregator front-end will eventually need KYC hooks, geo-fencing, and monitoring โ all of which are fixed costs that further compress an already thin margin. The teams that survive this will be the ones whose value capture sits at the API layer, embedded invisibly into wallets and apps, not at the branded consumer front door.
Which brings me to what actually matters. The interesting question is not whether 1inch can turn profitable next quarter. It is whether the aggregator category itself has a viable margin structure, or whether it is a transitional layer that gets absorbed into wallets and intent solvers. My read, based on the code and the cost curves, is the latter. Aggregation is becoming a commodity feature. The value is migrating to whoever owns the user relationship and the order flow โ and increasingly, that is the wallet, not the router.
Watch for three signals over the next two quarters. First, whether the protocol's public financial disclosures shift from cumulative volume to revenue per routed dollar โ if they stay on volume, the story is unchanged. Second, whether the team pivots toward B2B or API monetization, which would indicate they have privately conceded the consumer front-end margin is unrecoverable. Third, whether token unlock schedules or treasury spend accelerates as a share of protocol revenue, which would tell you the runway is being burned to hold market position rather than to build margin.
The bytecode didn't lie in 2019, and it isn't lying now. The router is a passthrough. The volume is real. The profit is not there because the architecture was never designed to hold it.
$800 billion moved. Almost none of it stayed.
That is not a bull-market problem to be fixed by more users. That is a design constraint to be engineered around โ or designed away from entirely. The next team that solves aggregator economics will not do it by routing more trades. They will do it by owning the flow before it ever reaches a router.