Arbitrage & HFT

overview

Wick runs its own arbitrage infrastructure. When prices diverge across Wick pools, the Lighter orderbook, and external venues, Wick can execute internally at 0% fee, capture the spread, and keep the profit inside the protocol.

Arbitrage is not optional. It is the mechanism that keeps AMM prices aligned with the rest of the market. When prices move, an external bot captures the spread by default. Wick, the protocol that created the liquidity surface, can capture it instead.

Wick's answer is to capture that arbitrage in-house. Dynamic fees recover most of the value when volatility rises, while internal arbitrage clears residual spreads that would otherwise leak to outside bots. Captured profits are used to buy back WICK, which accrues to sWICK holders, with LP incentives used where applicable.

how it works

Wick captures spreads across its AMM pools and external venues, then routes captured value back in-protocol instead of losing value to outside arbitrageurs.

core AMM–orderbook arbitrage path

Wick's core arbitrage path is the spread between Wick AMM pools and the Lighter orderbook. If the AMM quotes LIT at $1.698 while the orderbook quotes $1.702, an external bot would need to clear pool fees, venue fees, gas, priority costs, and execution risk before that trade becomes worth taking. Wick's internal path is structurally cheaper.

Wick's permissioned bot buys from the cheaper venue and sells on the more expensive one at 0% internal fee, atomically within the same block. External bots still have to clear fees before a spread pays, so smaller price discrepancies they ignore are still worth capturing for Wick.

The result is simple: spread that would have left Wick becomes revenue that buys back WICK instead.

On top of that, wLIT provides a fee discount for Wick's arbitrage accounts, cutting execution costs further.

no-arbitrage band

External bots only trade when the spread is wide enough to clear their costs. This creates a no-arbitrage band where price differences can exist because they are not profitable after fees. Wick's internal arbitrage runs at 0% fees, so the band it needs to clear is effectively zero. That does not mean every possible market inefficiency disappears; it means Wick can capture a much broader set of internal spreads than a normal outside arbitrageur, without getting frontrun.

Drag the slider to see how the band an external bot needs changes with the fee, and how Wick's band is zero.

As volatility rises, dynamic fees widen the no-arbitrage band (shaded region). An unmanaged pool (dashed) drifts and is only corrected once it moves outside the band, causing arbitrage events (red dots). Wick keeps price pinned near fair value (0% band), so it captures spreads across the whole region, not just at the edges.

multi-venue routing

The same logic extends beyond a single AMM/orderbook route. Wick continuously monitors internal pools, Lighter markets, and external venues. When arbitrage is executable, it captures the spread and uses it to buy back WICK, which accrues to sWICK holders.

Wick monitors DEX-DEX, CEX-DEX, cross-chain, and multi-step arb routes and executes them atomically at 0% fee.

path scaling

Wick does not run a single arbitrage loop. It monitors internal pools, the Lighter orderbook, and external venues at once. Each connection is another surface where a spread can be closed at 0%. Every new pool, bridge, asset, and network multiplies the routes it can combine with, and the 0% arb band makes more of those routes worth capturing.

where captured value goes

protocol routing

The baseline is simple: captured arbitrage value goes back to the protocol and LPs, not to outside arbitrageurs. Wick's fees are optimized for maximum revenue with the goal of recapturing LVR as fees, and the internal arbitrage system captures the spread that would otherwise leak out of the pool.

LP incentives

Captured value is used to buy back WICK, as with every other revenue source routed through sWICK; that WICK is then awarded as incentives to the LPs affected by the arbitrage. The arbitrage trade and the buyback are the same either way.

Wick vs fee auctions

Fee-auction designs sell the right to extract arbitrage in a block. They can route some auction revenue back to a protocol, but the liquidity that created the opportunity will still bear the repricing cost while someone else captures the auction value.

Wick takes a different approach: internalize the execution path, capture the spread directly when possible, and route captured value back into the protocol system. The tradeoff is that Wick uses privileged infrastructure; the benefit is that value capture is aligned with the venue and liquidity that created the opportunity. For the deeper LVR and auction math, see concepts.

LP protection

An LP position leaks value when the pool is slow to update its price. External arbitrageurs buy the underpriced side, sell the overpriced side, and keep the spread. Wick cannot remove the directional risk of holding a two-sided LP position, but it can reduce how much value leaks away during repricing.

internal arbitrage

When prices move, someone captures the spread between a stale pool quote and the rest of the market. Wick runs that capture in-house: internal trades on Wick pools execute at 0% fee, so spreads too thin for outside bots still get closed before value leaves the protocol. The chart below compares LP outcomes with and without that protection.

Dynamic fees are the first line of defense: they raise the cost of trading against the pool when volatility rises, reclaiming most of what arbitrage would take from LPs through the fee channel before the trade even happens.

liquidity asymmetry

The CEX vs DEX depth gap is why Wick internalizes arbitrage instead of simply widening fees. The arbitrage will happen either way; the real question is which venue absorbs the price impact and which side keeps the profit, and the math heavily favors the deeper venue.

In cross-venue arbitrage, what the arbitrageur gains equals what both venues lose combined. Moving liquidity between venues changes who pays, not the total.

The pool with less liquidity absorbs most of the price impact. At a CEX / DEX liquidity ratio of 100:1 (a conservative estimate for major markets), the DEX bears 99.1% of the total loss and the CEX bears 0.9%. As the CEX side grows, the DEX bears an even larger share of the loss.

This is the structural reason DEX LPs need protection. Absent any internalization, CEX-driven arbitrage is a direct subsidy from DEX LPs to CEX market makers; the deeper venue suffers almost none of the loss while capturing almost all of the profit. Running arbitrage in-house reverses that: Wick captures the same spread the external bot would have taken, and routes it back to the protocol rather than to an outside venue.

Together, dynamic fees and internal arbitrage keep more of the repricing value inside the system instead of losing it to outside bots.

For the underlying math (path costs vs endpoint costs and fee-arb decomposition), see concepts.

protocol-owned ALM

Wick runs protocol-owned market making on its core pools, starting with LIT/USDC, so the exchange always has active, tight liquidity where it matters most. Better execution brings traders in; deeper main-pool flow generates swap fees and repricing activity that Wick's own fee and arbitrage systems can capture instead of losing to outside bots. As Wick lists more pools and routes more volume, that revenue surface grows with the exchange.

Most venues cannot run this in-house. Tight market making creates constant repricing flow, and in a normal setup that value leaks to outside arbitrageurs. Wick already operates the fee system and the internal arbitrage path described above, including lower-cost execution provided by wLIT.

LIT market making

Protocol-owned liquidity on LIT/USDC stays always active through continuous rebalancing. Much of the repricing flow that would leak out in a standard tight ALM is captured in-protocol, so the position can compound instead of lose value.

On LIT/USDC, the diagram below maps the same repricing event two ways: a normal vault on the right (value leakage, position bleeding) versus Wick's ALM (dynamic fees & arbitrage, position growth).

Standard ALMs can earn strong fees and still lose. Tight depth increases repricing flow; without dynamic fees, arbitrageurs take most of the spread on a normal vault. Wick's ALM works because Wick is both the liquidity provider and the arbitrageur: dynamic fees capture most of the repricing revenue on each move, and internal arbitrage keeps the residual spread in-protocol.

why it works: capturable vs structural

Compare how each side accounts for the same repricing flow. A standard tight ALM:

value extracted = fees kept + value lost to arbitrage

When fees fall short of impermanent loss and what leaks to arbitrage, the position bleeds even with strong fee income. Wick changes the second term by internalizing the repricing flow:

value growth = fees + value kept by protocol - IL

Standard ALMs lose both layers: the capturable repricing cost and the structural IL. Wick recovers the capturable layer through dynamic fees and internal arbitrage; what's left is impermanent loss, the minimum cost of two-sided liquidity.

scaling with range

A tighter range makes the strategy more productive when markets are active, but increases how often the position must be repriced. Tighter positioning raises both fee generation and arbitrage pressure, so the strategy only works if Wick captures enough of that arbitrage to outweigh the added impermanent loss. Because Wick charges adaptive fees and internalizes arbitrage at once, tighter liquidity magnifies the rebalancing flow it can recycle rather than only magnifying losses.