Why searchers and solvers are a temporary phenomenon…

Let’s start this article off with a simple analysis of what solvers and searchers do today, and where they’ll have to go, or else be like the Dinosaurs….

Solvers/Searchers Solving Solutions Simultaneously

Solvers today fall into two general categories.

1. Same chain auction fulfillers

2. Cross chain auction fulfillers

Searchers fall into the following categories:

Do you notice something? It took me a long time to notice, but don’t these things look awfully similar?

• Solving = Atomic arb between incoming orderflow to an intents dex, and some dex routes

• Searching = Atomic arb between incoming orderflow/liqudiation… and some other dexes?

Now for the serious analysis. The real distinction between these two is just the simple fact that when you send your orders to a solver platform, you’re letting yourself get arbitraged against other dexes with the idea that you’re getting fucked less.

While it’s a small distinction, it’s an important one none the less. Solving generally is a net zero sum game, as is searching. If you do any systematic analysis on dex-dex arbitrage and “searching,” you’ll find that the profits have continually been declining. The validator’s eat all the profit in bribes, and the searchers generally make under 10% of the value of an arbitrage, sandwich, or liquidation.

Another interesting piece of anecdotal evidence from my time speaking with solvers is that they are bifurcating into similar market segments that we saw in the early days of searching. Searchers generally found edges in two ways:

1. Being faster and smarter about bidding on short tail volume, i.e. ETH/USDC, WBTC/USDC, etc. etc.

2. Being faster to integrate in long tail sources of MEV. Liquidations, things that look more like hacking, etc. etc.

Searchers today are doing the same. They either win trades on intents platforms by either:

1. Being faster to route trades through short tail dex’s

2. Being faster to integrate more sources of liquidity in that other solvers might not have yet.

Given that we see such a distinct overlap between these two very differently marketed types of “DeFi Market Makers,” what does that mean for the future of searchers?

Let’s find out by looking at how the searchers have done over the years.

Shockingly low, right? And it’s only been going down. All the edge in atomic arbitrage has declined and nearly to zero. Look, 13M isn’t nothing, but it’s not the billions made in HFT market making annually from the likes of Citadel, Jump, Wintermute, Jane St, etc. etc. This doesn’t exactly bode well for solvers. But they have one advantage that searchers didn’t.

Hindsight.

Is There Hope For Solvers?

I’d say yes. That edge of hindsight should not be ignored. It’s how we got Facebook after all, given that Zuck has even said he learned what not to do by watching MySpace and Friendster.

Searchers and solvers are different, yes. But they’re alike enough that I see their life cycle’s being quite similar in the long run. I think solver’s if they had been around several years ago, would have found themselves spending most of their time (if they already didn’t) building optimized searcher algorithms that look almost identical to their existing codebases. I think solver’s are the next generation of this type of startup, and it’s important for them to analyze those who came before them and learn from their mistakes.

Back to the point of this section. If searchers are the MySpace of this analogy, how can solvers hope to avoid that fate?

It’s very simple. They have to take more risk.

That may seem like an odd recommendation but if you look at what we said earlier it is that solvers do, it seems more like an interesting information edge than temporary phenomenon. It makes more sense as well if you look at the history of trading in general. In crypto and in traditional finance, the way to build a successful trading firm is shockingly similar to what Web2 investors have told their portfolio companies to do: find an overlooked market that people underestimate, become incredible at that market, and expand outwards.

Citadel started with a massive focus on convertible bonds, the founders of SIG began their trading journey’s betting on horse races, and several (unnamed) shops in crypto made it big trading the most illiquid shitcoins they could find with options overlays.

Solver’s have the opportunity to do the same, but their edge is DeFi routing. It’s my belief that the solvers that don’t innovate and expand into other markets, outside of atomic arbitrage/routing will be left behind and find their margins continually driven to zero. It’ll just take someone open sourcing their solving algorithm to wipe out the majority of the edge there. With the creation of platforms like Across, Everclear, and other Cross-Chain auctions, we’re beginning to see some solvers actually holding inventory for more than one block. I think this is a great first step, but it’ll need to go much further.

So let’s look at their options:

1. Expand to more markets, take more risk, do more market making

2. Build their own orderflow sourcing platforms

They either must become more sophisticated quantitative traders, or get their sales rolodex’s ready by my estimation. I’m happy to be proven wrong here, but I don’t really see how that skillset of highly optimized routing can be applied in other ways. You can see the second category of solvers already existing, they just call themselves “Aggregators” rather than solvers.

An aggregator runs the same algorithm to a T that a solver runs, they just run try to sell it direct to consumer rather than through an intermediary platform. We’re already seeing a shift today away from AMM’s to more intent centric platforms, so what pools are these solvers or aggregators even going to be routing through in the next 3-5 years?

The role of CEX-DEX arbitrage as a way for market makers to eat more and more of the solver market share shouldn’t be under-estimated either. This is a major way we could see solvers get out competed before they can even leave the cradle. For example, the number of red circles is growing, and that doesn't bode well for solvers:

If intents go the way I expect them to, namely becoming entirely swaps or bundles of swaps, the more liquidity sources they can integrate the larger the moat they can maintain. Solver’s don’t need to become market makers specifically. They could fulfill a role of a smart order router as well; which admittedly many market makers also have as a product (they usually market it as an RFQ system or OTC desk backend).

While I make many predictions in this article, one thing is for certain: it’s up to solvers to determine their own fate.