<\/p>\n
\n
\n<\/p>\n
BJF TRADING GROUP \u00a0\u00b7\u00a0 OPEN RESEARCH<\/span><\/p>\n A public, browser-based scanner that monitors how far retail forex brokers lag behind a fast reference feed in real time. No download, no signup, no email wall \u2014 open the page and the dashboard loads<\/strong>. Built by BJF Trading Group as a community research asset, like our open-source BEQI broker audit toolkit<\/a> and our existing free crypto arbitrage scanner<\/a>.<\/p>\n Yes \u2014 it’s a real, free, browser-based scanner.<\/strong> The dashboard at \/arbitrage-stats\/<\/a> shows every detected arbitrage opportunity between a fast reference feed and four anonymized retail brokers, across nine of the most-traded forex and index symbols. No software to install. No account to create. No payment. The data is open to anyone with the URL.<\/p>\n The scanner does not<\/strong> execute trades. It is a measurement and discovery tool: it shows you which brokers lag, which symbols have the densest opportunities, what time of day to focus on, and how the lag changes over weeks and months. What you do with that information is up to you.<\/p>\n<\/div>\n The dashboard is organized into six analytical sections, all of which are filterable by date and time range, symbol, broker, minimum differential, and minimum opportunity duration. The default view automatically loads the last seven days of data.<\/p>\n 1 \u00b7 KPIs<\/p>\n Opportunity count, average differential in pips, average duration in milliseconds, single-largest differential, total theoretical pips, and the automatically-detected best time-of-day window for the selected period.<\/p>\n<\/div>\n 2 \u00b7 HEATMAP<\/p>\n A 7\u00d724 grid showing where opportunities cluster across the trading week. London is open, and the London\u2013New York overlap dominates visibly; Asian-session activity is sparser.<\/p>\n<\/div>\n 3 \u00b7 HISTOGRAMS<\/p>\n Two bar charts: one showing the distribution of opportunity sizes in pips, the other showing how long opportunities stayed open in milliseconds. Useful for sizing strategy thresholds.<\/p>\n<\/div>\n 4 \u00b7 SESSIONS<\/p>\n Opportunity count and average characteristics per trading session. Confirms which session is most arbitrageable for the symbol and broker pair you have selected.<\/p>\n<\/div>\n 5 \u00b7 SYMBOL RANKING<\/p>\n Ranking of which symbols generate the most opportunities on the chosen broker, with their average differential. Helps you pick which symbol to focus on.<\/p>\n<\/div>\n 6 \u00b7 EVENTS TABLE<\/p>\n Every detected event: timestamp, symbol, direction (long\/short), differential in pips, duration in milliseconds, broker, session. Paginated, exportable to CSV for offline analysis.<\/p>\n<\/div>\n<\/div>\n Two further charts \u2014 a broker lag profile<\/strong> (average opportunity duration by hour of day, a proxy for the broker’s processing speed at different times) and a cumulative theoretical-pips curve<\/strong> (the running total over the selected period) \u2014 complete the dashboard.<\/p>\n The detection logic is the same one that underlies serious latency arbitrage<\/a> in retail forex. Two price streams run in parallel:<\/p>\n The detection engine walks both streams in chronological order. Whenever the broker’s mid-price diverges from the fast feed by more than a threshold in pips, instrument-adjusted, a state machine opens an “in-opportunity” state and records the timestamp. When the broker’s quote catches up to within a tighter close-threshold of the fast feed, the state closes, and an event is emitted. Events shorter than a minimum duration are discarded as physically unreachable (your order could never have filled inside that window). Single-tick spikes that look like opportunities but are really feed noise are filtered out by a statistical spike filter.<\/p>\n The resulting event is what the scanner stores: timestamp, symbol, broker, direction (which side of the spread the divergence was on), magnitude in pips, and duration in milliseconds. From those four numbers, everything in the dashboard is derived \u2014 the KPIs, the heatmap, the histograms, the sessions, the ranking, the lag profile. For the broader execution-quality measurement framework these events sit inside, see our open-source BEQI broker execution audit toolkit<\/a>.<\/p>\n The scanner is a measurement<\/strong> tool, not an execution tool. It identifies windows where arbitrage was theoretically possible against the monitored brokers; it does not place trades, manage positions, or guarantee that the same windows will be available to you in live conditions on your own account. Execution requires its own infrastructure \u2014 a co-located VPS, a fast feed subscription, and automated entry logic \u2014 which is a separate decision. The scanner tells you whether the opportunity exists; our software products<\/a> handle execution.<\/p>\n<\/div>\n The scanner currently monitors four retail forex brokers across the most-traded forex pairs, gold, and three index CFDs.\u00a0Brokers are\u00a0anonymized<\/strong>\u00a0in the public dashboard (Broker A, B, C, D) for two reasons: to limit the legal exposure that comes with naming regulated entities in a comparative ranking, and to prevent broker-side optimization of the audit signal \u2014 a Goodhart’s law concern we discuss in our\u00a0BEQI methodology paper<\/a>.<\/span>\u00a0The methodology is open; downstream researchers are free to compile their own named leaderboards from the data.<\/p>\n The list is a starting set chosen for liquidity and arbitrage activity. We will extend it as the dataset matures \u2014 a tracker for adding new instruments is on our roadmap. If you have a specific symbol you’d like to add, the contact form at the bottom is the right place to suggest it.<\/p>\n Retail forex execution quality is the most opaque layer of the trading stack. Brokers compete on advertised spread; their actual execution behavior \u2014 matching latency, slippage symmetry, spread response to order flow \u2014 is invisible to the trader and unpublished by the broker. Third-party broker reviews score on user experience and withdrawal speed; nobody measures the variables that actually decide whether an arbitrage strategy survives in production.<\/p>\n We have been measuring this internally for years to build broker-dialect support into our commercial software. The free scanner, the open-source BEQI methodology<\/a>, and our research on the anti-arbitrage plugins retail brokers use<\/a> are all part of the same project: bring the measurements out into the open so the trader community has shared evidence, instead of forum-anecdote evidence, about what the broker layer is actually doing.<\/p>\n The scanner is therefore offered in the same way as an academic dataset: free, public, and reproducible, with deliberate choices about anonymization and methodology that we expect to defend in public. If you find a problem, the methodology page on our BEQI repository<\/a> is where to open the discussion.<\/p>\n Four practical uses for the scanner output, ranked by effort.<\/p>\n There are paid forex arbitrage scanners on the market, and there are free pseudo-scanners (correlation indicators marketed as arbitrage tools, GitHub demos based on triangular-rate Bellman-Ford). The free BJF scanner sits in a different category from both.<\/p>\nFree Forex Arbitrage Scanner \u2014 Live Broker Lag Detection<\/span><\/h1>\n
\nHow it works<\/a><\/div>\n
\n9<\/strong> symbols scanned
\nLive<\/strong> data, UTC
\n365-day<\/strong> history<\/div>\n<\/div>\n
\n
\n<\/p>\nThe short answer<\/h2>\n
\n
\n<\/p>\nWhat you see in the scanner<\/h2>\n
The headline numbers<\/h3>\n
Hour-of-day \u00d7 weekday<\/h3>\n
Differential and duration distributions<\/h3>\n
Asian \/ London \/ NY breakdown<\/h3>\n
Top instruments by opportunity count<\/h3>\n
Raw events with CSV export<\/h3>\n
\n
\n<\/p>\nHow the scanner detects opportunities<\/h2>\n
\n
What the scanner is not<\/h3>\n
\n
\n<\/p>\nWhich brokers and symbols are monitored<\/h2>\n
Symbols monitored<\/h3>\n
\n
\n
\n<\/p>\nWhy we made the scanner free and public<\/h2>\n
\n
\n<\/p>\nHow to use the data<\/h2>\n
\n
\n
\n<\/p>\nFree scanner vs paid alternatives<\/h2>\n