WoM = back money / (back money + lay money) across the visible order book. A reading above 60% means significantly more money is queued to back the runner than to lay it — a leading indicator of market sentiment before BSP is set.
The strategy fires a BACK signal when WoM exceeds the threshold for N consecutive ticks, filtered to the 30-minute pre-race window. Multiple consecutive ticks eliminate noise from momentary liquidity spikes. The threshold and N are tunable in real time above.
In the live system, signals feed an Order Service that manages partial fills, hedge exits, and liability limits via Betfair's REST API. This backtest uses the raw streaming data and simple entry-at-signal, hold-to-start semantics for clarity.
Real race data · Wolverhampton, 5 May 2026 · 8 runners · Betfair Streaming API · ← All Demos · Building something similar?
About this demo
This backtest replays real Betfair historical data from a Wolverhampton evening race, applying a Weight of Money signal strategy in post-hoc simulation. Select any runner, set a WoM dominance threshold, and the engine scans the historical price and WoM series to mark the points where the signal would have fired. The LTP chart shows how the price moved after each signal.
Backtesting is the first step in evaluating any algorithmic trading strategy. It answers whether a signal had predictive value in the past — though past performance is no guarantee of future results, particularly on pari-mutuel exchanges where markets are reflexive. A robust backtest accounts for latency between signal detection and order execution, and avoids look-ahead bias by ensuring each signal is calculated only from data available at that point in time.