Latency arbitrage lives and dies on one thing: execution. The signal is almost never the problem — a two-leg engine can spot a price gap between a fast feed and a slow broker thousands of times a month. The problem is what happens in the handful of milliseconds between seeing that gap and actually getting filled. Two forces work against you in that window: slippage, which quietly shaves the edge off every entry, and broker detection, which can freeze, requote or restrict an account that trades too mechanically.
This note looks at both, using two real, publicly verifiable accounts as the reference point, and then shows the exact execution settings that defeat slippage — and the honest trade-off that comes with them.
First, the potential
Before talking about what can go wrong, it is worth seeing what two-leg arbitrage does when it runs. Below are two live FX Blue accounts running the same two-leg approach over the same 27-day window — you can open each one and re-check the figures yourself: 2legs1 and 2legs2 (both open in a new tab). These are the raw numbers, unretouched.
+213%2legs1 · 27 days $6,000 → $18,794
+209%2legs2 · 27 days $5,000 → $15,441
1.3profit factor on both accounts
17 vs 2winning vs losing days
Two independent accounts, same strategy, same month — both roughly tripled their starting balance.
Those headline returns are the product of a very particular shape of trading: a huge number of small trades, where the winners are just a little bigger and a little more frequent than the losers. It is worth breaking that down — because it is exactly this churn that makes execution quality so decisive.
Gross profit and gross loss are both large; the net is the thin margin between them. With a profit factor near 1.3, small execution costs matter enormously.Across roughly 3,000 and 1,800 trades, the win rate sits near 54% with average wins slightly larger than average losses. That edge is real but narrow.
Why this matters. When your profit factor is 1.3 and you win by a few pips on average, a pip of slippage per trade is not a rounding error — it is the business. Protecting the entry price is the single highest-leverage thing you can do.
How slippage eats the edge
Here is the mechanism. The fast feed jumps to a new price. Your engine fires an order expecting to trade at the price you just saw. But a market order does not execute at the price you saw — it executes at whatever price is available a few milliseconds later, by which time the broker has already begun moving toward the new level. You get filled worse than you intended, and that difference is slippage.
The faster and more accurate the signal, the more a delayed market fill costs — because the broker price is already catching up when your order lands.
On a slow, discretionary strategy a pip here or there is noise. On a high-frequency two-leg engine winning by a few pips thousands of times, negative slippage is the difference between the curves you saw above and a flat or losing account.
Fix #1 — open with a limit order (GTC / FOK)
The cleanest way to remove entry slippage is to stop using market orders altogether. A limit order fills only at your chosen price or better. If the edge is still there, you trade at exactly the price that made the signal worthwhile. If the price has already moved, the order simply does not fill — and you take no trade rather than a bad one.
A market order always fills, at any price; a limit order fills at your price or not at all. Negative slippage on entries is removed by design.
SWITCH ON Open with limit — set entries to limit orders (GTC to rest in the book, or FOK to fill-or-kill instantly). This is the one setting that turns entry slippage off.
Fix #2 — close with a stop (S/L)
Entries are only half the round trip. Closing a leg at market re-introduces exactly the slippage you just removed on the way in. Closing with a stop level (Close with S/L) lets you define where the position comes off, rather than accepting whatever the broker offers at the moment you hit the button. Combined with limit entries, the whole round trip is executed on your terms.
SWITCH ON Close with S/L — close positions at a defined stop level instead of a market exit, so the exit price is controlled too. Here alongside limit open/close (FOK) and the arbitrage toggles.
Fix #3 — camouflage, so the account survives
Clean execution is worthless if the account gets flagged. Brokers that dislike arbitrage look for a machine footprint: identical, instant, back-to-back orders with inhuman timing. The answer is not to trade less — it is to trade with a human-looking rhythm. Pauses between locks and orders, a delay before reopening a lock, small random corrections and variable delays spread the same activity into something that looks like ordinary discretionary trading.
Same trades, different footprint. Uniform timing is a giveaway; spacing and randomisation keep the activity under the radar.
TUNE THESE Pauses, reopen-lock and random correction — seconds between closing locks, opening locks and arbitrage orders; reopen a lock after a set delay; add random correction. Together they break the mechanical rhythm that gets accounts restricted.
The honest trade-off
None of this is free, and it would be dishonest to pretend otherwise. When you open with a limit order, some signals will not fill — by the time your price is reachable, the edge has already gone, so the trade is skipped. Fewer trades means the raw, headline profit comes down somewhat compared with grabbing every signal at market.
Illustrative. Market execution grabs every signal but bleeds edge to slippage; limit-open plus stop-close skips a few entries in exchange for fills you can actually trust.
The exchange you are making: a modest reduction in the number of trades and in raw profit, in return for no negative slippage on entries, a controlled exit, and a footprint that keeps the account alive. On a live, slippage-prone broker that is not a downgrade — it is the difference between a backtest-only curve and a strategy that actually survives contact with the market.
Put simply: the accounts above — 2legs1 and 2legs2 — show what the signal is capable of. Limit entries, stop closes and camouflage are what let you keep a realistic share of it on a real broker, month after month, instead of watching slippage and detection hand it back.
Run two-leg arbitrage with execution on your terms
Limit-open (GTC/FOK), close-with-stop and a full set of camouflage controls are built into HFT Arbitrage Platform — the same settings used on the accounts above.
Performance figures are taken from the public FX Blue accounts 2legs1 and 2legs2 over a 27-day period and reflect two-leg arbitrage with standard execution. Past performance does not guarantee future results; trading carries risk. The “trade-off” chart is illustrative and explains the mechanism rather than a measured figure. Always confirm current broker terms and your own account’s arbitrage policy.
Sergey Luts is a highly experienced developer of high-frequency trading systems with more than 20 years of expertise in financial technology and advanced trading infrastructure. Born in Dnipropetrovsk, Ukraine, he graduated from Dnipropetrovsk State Technical University in 1999 and has been specializing in HFT system development since 2000.
Throughout his career, Sergey has developed sophisticated solutions for professional trading environments, including work for a hedge fund where he created a successful trading system. Five years ago, he chose to pursue his own path, assembling a team of mathematicians and traders and founding HFT Software to develop next-generation trading technologies.
Over the years, Sergey has contributed to the development of hundreds of trading strategies, all thoroughly tested and refined, with only the most effective solutions selected for sale and real-world application. His portfolio includes dozens of arbitrage strategies, high-performance execution tools, and an exceptionally fast Forex trade copier built for demanding HFT infrastructure.
With strong expertise in Python, C#, and C++, Sergey is known as an experienced programmer and a skilled architect of advanced trading systems. His work is defined by speed, precision, technical depth, and a practical understanding of what professional traders require in highly competitive markets.
