How Pulse simulates the market
Pulse is a large scale simulation of the financial markets. Its aim is to produce synthetic market data for in depth martket analysis and training of robust trading algorithms.
Your request
[setup]Choose the market you want to study: the symbol, exchange, and trading day, along with the scenario to run and any execution algorithm you want to test against it.
Agent-based simulation
[core engine]A population of rule based trader agents, including background liquidity providers, fundamental traders, momentum traders, and noise traders, interact through a simulated limit order book. The population is calibrated against historical order-book activity, so it behaves like the real market. The agents are defined mathematically, shown below.
qf = κ·(F − Pt)Fundamental trader demand: the gap between fundamental value and priceqm = β·ΔPtMomentum trader demand: the recent price trendqn ~ N(0, σ²)Noise trader demand: random order flowResults & analytics
[output]The full simulated order book, price path, and trade history come back for analysis. If you tested an execution algorithm, its fill schedule and execution are benchmarked against the baseline, to produce market impact, risk and slippage statistics.
Validate (optional)
[pro only]Score how closely the run matches reality with Pulse-Check, comparing your simulation against the same historical window it was calibrated to.
No API key charges, no queue: the Free tier skips configuration and simulation entirely and gives you pre-run results directly.
Browse the library
[no setup]Instantly explore a library of pre-run simulations across a selection of HKEX and LSE symbols, dates, and scenarios: already simulated.
Download & analyze
[output]Pull the same order book history and price data a Pro job would return, as a Polars DataFrame or in bulk as a ZIP, and start analyzing straight away.