Market Maker Arbitrage — Complete Guide

1. The idea: two exchanges, two different roles

Market Maker Arbitrage is a cross-exchange strategy built around two exchanges playing deliberately different roles:

Target Exchange — usually less liquid, with wider spreads. This is where you place a maker (limit) order and wait to get filled at a favorable price.
Reference Exchange — highly liquid, tight spreads, fast execution. This is where you immediately fire the opposite order as a taker once the maker leg fills, to lock in the hedge.

The whole point is capturing the maker rebate/lower fee and a better entry price on the illiquid side, while relying on the liquid side to guarantee you can always close the position instantly. It's a step up in complexity from Simple Arbitrage because the first leg is not guaranteed to fill — you're queuing in the order book, not taking existing liquidity.

2. The math: how the maker price is actually set

A limit order needs to sit just ahead of the current best price to have a realistic chance of filling, without giving away the edge by overpaying. ArbiMaster uses a small offset above/below the best bid/ask as a tick-size proxy:

Maker pricing formulas

MAKER_PRICE_OFFSET_PCT = 0.0001 (0.01%, used as a tick-size proxy across exchanges with different tick sizes)

Buy-First scenario (you enter as a maker buyer on the Target exchange):
makerPrice = targetBid × (1 + MAKER_PRICE_OFFSET_PCT)
→ just above the best bid, to jump the queue without paying the ask.

Sell-First scenario (you enter as a maker seller on the Target exchange):
makerPrice = targetAsk × (1 − MAKER_PRICE_OFFSET_PCT)
→ just below the best ask, for the same reason on the sell side.

Tradeable volume is always computed from the taker side only (the Reference exchange's bid or ask quantity) — because that's the side that must absorb the hedge instantly and therefore the real bottleneck for how much size you can actually run.

This maker/taker asymmetry is the core mechanical difference from Simple Arbitrage: there you take two existing prices, here you deliberately create a new best price on one side to improve your entry.

3. The two execution scenarios in detail

Scenario A — Sell-First (you sell on Target, buy on Reference)

Analyze the spread, liquidity and price correlation between the two exchanges.
Place a maker sell order on the Target exchange just below the best ask.
Wait for a fill confirmation (real-time monitoring, not polling).
The instant the sell fills, immediately buy the same quantity on the Reference exchange as a taker.

Example (BTC)

Target Ask = $41,400 → maker sell at $41,390.86 (fee ~0.05%, or a rebate depending on the exchange)
Reference Ask = $40,010 → immediate taker buy (fee 0.10%)
Gross profit: ≈ 3.45% ($1,380 per BTC) — Fees: ≈ 0.15% — Net: ≈ 3.30% (≈ $1,320 per BTC)

Scenario B — Buy-First (you buy on Target, sell on Reference)

Place a maker buy order on the Target exchange just above the best bid.
Wait for the fill.
The instant it fills, sell the same quantity on the Reference exchange as a taker.

Example (BTC)

Target Bid = $40,000 → maker buy (fee ~0.05%)
Reference Bid = $41,200 → immediate taker sell (fee 0.10%)
Gross profit: ≈ 3.0% ($1,200 per BTC) — Fees: ≈ 0.15% — Net: ≈ 2.85% (≈ $1,140 per BTC)

4. The main risk: your maker order might not fill

Unlike Simple Arbitrage, the maker leg here is not guaranteed. While it sits in the queue, the price can move away, meaning it never fills — or worse, it fills after the reference price has already shifted, so the hedge you execute a moment later locks in a smaller (or negative) spread than the one you calculated. This directional exposure window is why Market Maker Arbitrage carries more risk than Simple Arbitrage, even though a single leg pays lower fees.

5. Setting it up on ArbiMaster

From Settings you choose your MM reference exchange — the highly liquid venue ArbiMaster will always treat as the hedge side. The Market Maker scanner then continuously compares every other monitored exchange against it, computing both the Buy-First and Sell-First maker price for each candidate pair, and reports only opportunities above your configured minimum spread (default ≥ 0.8% — deliberately higher than Simple Arbitrage, to compensate for the non-fill risk).

Telegram notifications respect the same per-strategy cooldown as the other strategies, so you won't get repeat alerts for a spread that's still open but hasn't meaningfully changed.

6. Best practices

Capital allocation: a common split is 40% on the Target exchange, 40% on the Reference exchange, 20% held back as a rebalancing buffer.

Exchange tiering: Tier 1 venues (Binance, OKX, Kraken) make good Reference exchanges thanks to deep, fast books; Tier 2 / regional exchanges more often produce the wider spreads that make good Target candidates.

Monitor fills in real time. Don't poll — use WebSocket order updates so the hedge fires within milliseconds of the maker fill, not seconds later.

Set a cancel timeout. If the maker order hasn't filled after a reasonable window, cancel it — a stale limit order sitting in a moved market is pure directional risk with no upside.

FAQ

Why is the minimum spread higher than for Simple Arbitrage?

Because the maker leg isn't guaranteed to fill, and when it does, the market may have already moved. A wider required spread (≥ 0.8%) gives you a cushion against that adverse-fill risk.

What happens if my maker order never fills?

Nothing is lost except opportunity cost — you simply cancel and look for the next candidate. The risk only materializes if it fills late, after the reference price has already moved against you.