Blog / EXECUTION / BEST EXECUTION

Best Execution Is Not the Best Headline Rate

The biggest number on a quote screen is not always what you keep. Best execution ranks routes on net take-home after costs, real slippage floors, and speed when outputs are close.

Headline rate versus net take-home
Headline rate versus net take-home

Quick answer

Best execution is not "the highest number printed next to the destination token." That number is a headline rate. Best execution asks what you keep after gas and related costs, how firm the fill is under your slippage limit, and whether a slightly smaller but faster or gasless path is the rational pick when quotes are close.

RAVN ranks parallel venue quotes with that lens. Marketing screens that only sort by raw output are solving a different problem.

Headline rate vs what you keep

Two quotes for the same pair and size:

Quote AQuote B
Headline outputHigherSlightly lower
Swap gas (user pays)Non-trivial on this chainGasless RFQ (solver covers swap gas)
Price characterOn-chain, market impact possibleFirm RFQ price
What you might keepHeadline minus gas (and approval if needed)Headline close to take-home

If A only wins by less than its own gas bill in output units, B can be the better trade even though A looked "best" on a naive sort. That is the core of net-of-gas ranking.

What RAVN optimizes for

  • NetOutput after estimated swap gas, and approval cost when it is relevant. Gasless quotes are free of swap gas, not free of every possible on-chain cost forever.
  • FloorSlippage and firmness. Quotes that break the user's slippage tolerance are dropped, not surfaced. Firm RFQ quotes report zero slippage on that leg.
  • GaslessPreferred when they still win against a rival's guaranteed floor (rival output discounted by its own slippage), not as a hard override of bad pricing.
  • SpeedWhen net value is close, faster fills beat near-identical but much slower routes.
  • CanonicalDestination is the real asset on the chain you chose. Execution quality includes getting the right asset, not a synthetic stand-in.

How parallel quoting works

  1. You request a quote for a pair, amount, and chains.
  2. RAVN selects eligible venues for that route type (for example native BTC races Garden, Relay, and NEAR Intents when available; Solana can include Jupiter Ultra; EVM same-chain RFQ can include 0x Gasless, Bebop, and CoW).
  3. Venues are queried together. Failures narrow the field instead of killing the whole request when another venue still answers.
  4. Each quote is normalized into one shape: output, fees, execution type, and route metadata.
  5. Ranking applies net-of-gas and slippage rules. The winner becomes the quote you see or the quoteToken integrators pass to execute.

You do not manually pick a venue for best execution mode. You pick the outcome. The router picks the rail.

Net-of-gas ranking (plain language)

For every quote that can be costed, RAVN estimates what the user pays in gas terms (and approval when it matters), converts that cost into units of the output token using a price, then compares output minus those costs.

  • If the output token cannot be priced, or a quote's costs cannot be determined honestly, ranking falls back rather than inventing zeros. Partial netting that only discounts some quotes would favor the ones whose costs were invisible.
  • A gasless RFQ quote is not charged swap gas in that model. A gas-paying rival is. That is why raw output sorts are unfair across those two classes.

This is engineering detail for a user-facing reason: the screen should not celebrate a route that loses to its own fee drag.

Slippage, firm RFQ, and the guaranteed floor

On-chain routes can move between quote and fill. RFQ market makers (0x Gasless, Bebop, CoW on same-chain EVM) quote a firm price: zero slippage on that model, user signs off-chain, solver covers gas on those gasless paths.

When comparing a gasless firm quote to a slipping on-chain quote, RAVN does not treat "gasless" as a free win. The gasless quote has to remain competitive against the rival's guaranteed floor: what that rival still delivers after its own slippage is applied. If the firm quote trails too far, the better net on-chain path can still win.

User slippage tolerance is a hard filter. Quotes outside the limit are not shown as fake options.

Three execution types, one ranking

After ranking, execution collapses to one field:

  • TXTRANSACTION: sign and broadcast the returned transaction (Across, Relay, Mayan, Jupiter Ultra, and similar).
  • SIGSIGNATURE: sign typed data; RAVN submits (0x Gasless, Bebop, CoW).
  • DEPDEPOSIT: send the origin asset to a venue-generated address (NEAR Intents and other deposit rails, including many native BTC flows via Garden-style settlement).

Best execution is decided before you care which type won. The type only tells you what to sign or send.

What best execution is not

  • Not a promise of the global mid-market price. Depth, inventory, and venue coverage set the real band.
  • Not gasless on every route. Gasless is route-dependent. RAVN adds no gas markup; it does not erase network fees on deposit and on-chain paths.
  • Not zero slippage on every path. Firm RFQ is zero slippage. Other models are not.
  • Not a custody product. RAVN does not hold funds to manufacture a rate. Settlement stays on venue infrastructure.

How to use this as a trader or integrator

  • Traders: Read net outcome and execution type, not only the large output figure. Re-quote if the market moved or the quote expired.
  • Integrators: Call POST /quote, pass the opaque quoteToken to execute, branch on executionType. Surface fee fields that already appear in the quote. Do not re-sort by raw output and undo the router.
  • Both: For BTC corridors, native rails and wrap paths are different products. See the native BTC guide linked below.

Try it in the app

Live in beta. Parallel venues, best execution, canonical assets. RAVN never holds your funds.

FAQ

What is best execution in a crypto swap?

Best execution means choosing the route that maximizes what you actually keep after costs and constraints, not the route with the largest raw output number before gas, slippage, and fill quality.

Why can a lower headline rate be better?

If the larger quote burns more gas, carries higher slippage risk, or is much slower for almost the same net result, the smaller headline can leave you with more usable value or a better outcome overall.

Does gasless always win on RAVN?

No. Gasless or firm RFQ routes are preferred when they still beat the rival's guaranteed floor after that rival's slippage. They do not win by default if they trail too far on net value.

Is zero slippage guaranteed on every RAVN swap?

No. Zero slippage applies on firm RFQ routes (for example 0x Gasless, Bebop, CoW) where the price is firm. On-chain and deposit routes use standard market and network properties. RAVN does not claim zero slippage on every path.

How do I see what RAVN picked?

The app quote shows the winning route outcome. Integrators receive a normalized quote with execution type and fee fields. Venue liveness is available via GET /health.

Continue reading