Latency Arbitrage Tutorial: How WP7 Detects a Price Gap

Latency arbitrage, step by step. In this tutorial, you'll see how Westernpips Private 7 compares a fast price feed with a slower broker quote, how a price gap is measured, and which settings decide whether that gap is worth trading. By the end, you'll know the four checks every setup needs before a single order goes out.

First, the building blocks. A latency arbitrage setup has three parts. One: a fast market-data source, usually an institutional feed connected over FIX. Two: a slower trading account at a broker, often an MT4 or MT5 terminal. Three: the software in the middle, in this case Westernpips Private 7, running on a server. The software doesn't predict the market. It compares two prices that should be the same, and reacts when one falls behind.

Step one: find out which source is actually slow. Don't assume it. Westernpips includes an Analyser tool that records ticks from different sources at the same time. Let it run during active hours, like the London and New York sessions, then compare. The question is simple: when the fast feed moves, how long does the broker take to follow, and how often? A broker that lags once a day gives you nothing to work with. A broker that lags consistently is your slow leg. The Analyser also helps you spot artificial spikes: prices that jump and come back in a way the real market never did.

Step two: map the symbols. Your feed may call gold XAUUSD. Your broker may call it XAUUSD.m, or simply GOLD. Decimal places and contract sizes can differ too. If the mapping is wrong, the software compares two things that aren't the same, and every gap it sees is just a data error. So check each symbol pair, one by one, and confirm the prices line up on a quiet market.

Step three: read the gap. Here you see the fast price and the slow price side by side. The difference between them, measured in points, is the gap. Most of the time it sits close to zero. Then the fast feed moves, the broker hasn't updated yet, and the gap opens. In this example, gold moves up on the fast feed, while the broker still shows the old price for a few milliseconds. That's the window. But remember: a gap on screen is only a potential opportunity. It is not a profit.

Step four: decide when a gap is big enough. That threshold is the OpenGap setting: the minimum difference that triggers an order. Here's the rule that matters most. Every trade pays a spread, often a commission, and sometimes slippage. If your OpenGap is smaller than those costs combined, the trade is born losing, mathematically. Set it too low, and costs eat everything. Set it too high, and the software almost never trades. Start clearly above your total costs, then refine it with recorded data. That's exactly what we cover in the backtesting tutorial.

Step five: how a OneLeg strategy uses that signal. OneLeg is the classic structure. The fast feed is the reference. When the gap crosses your threshold, the software sends a single order on the slow broker, in the direction the fast price already moved. When the broker catches up, the position is closed. One leg means one position, with no hedge on the other side. So if execution is slow, or the price slips, that position is fully exposed to the market. That's why OneLeg setups rely on protective settings like a stop, a trailing stop or break-even, depending on your configuration.

Step six: the safety filters. Three settings protect you from the worst moments. A maximum spread filter: if the broker's spread widens, during news for example, the software doesn't trade. A slippage limit: the largest difference you accept between the expected price and the filled price. And time management: trading only in active sessions, when gaps come from real market moves, instead of quiet hours where a gap is more likely a technical glitch.

Step seven: check what really happened. Every trade leaves a log, and two columns tell you a lot. The close reason shows why the trade was closed. Execution time shows how many milliseconds the order took. As a rule of thumb, under 150 milliseconds is healthy. If execution regularly climbs above 500, the problem isn't the strategy. Look at your server location or your provider.

So, before any latency arbitrage setup goes live, run these four checks. One: is the fast source really faster, consistently? Two: are the symbols mapped correctly? Three: is your OpenGap above your total costs? Four: does your broker execute fast enough, without heavy slippage? Miss just one, and the edge disappears.

One honest reminder. Brokers watch for this kind of trading. Many restrict accounts that repeatedly profit from stale prices, so read your broker's terms before you connect anything. Results always depend on the broker, market conditions, infrastructure and costs.

In the next tutorial, we connect the pieces: FIX, TCP, and how WP7 links a price feed to a broker. To learn more about Westernpips Private 7, compare licence plans or read the documentation, use the links in the description.