Generating Intraday Price Paths from Synthetic Order Flow

Summary

The paper models high-frequency order flow as a source of intraday price changes and applies Sequence Generative Adversarial Networks to generate synthetic order sequences. It uses those sequences to simulate price paths, then compares the generated behavior with that of real market data. A parametric model from quantitative finance serves as the benchmark, with both models fitted before random order-flow paths are sampled.

Evaluation focuses on whether simulated prices reproduce empirical features of real price movements, including log-return distributions, volatility, and heavy tails. The authors report that the generative model produces order sequences whose resulting price variations better match observed statistical behavior than the benchmark. The excerpt does not identify the market, sample period, model details, or quantitative size of the improvement. Its evidence concerns statistical simulation fidelity, so it does not establish that the synthetic paths support profitable trading or capture every relevant market feature.

Key ideas

• The method uses a sequence GAN to generate synthetic high-frequency order flow.
• Generated order sequences are converted into simulated intraday price variations.
• A parametric quantitative finance model is used as a benchmark.
• Evaluation compares return distributions, volatility, and heavy-tail behavior with real data.
• The reported improvement concerns statistical resemblance, not trading profitability.

Tags

# Intra-Day Price Simulation with Generative Adversarial Modelling of the Order Flow


# Intra-Day Price Simulation with Generative Adversarial Modelling of the Order Flow









Intra-day price variations in financial markets are driven by the sequence of orders, called the order flow, that is submitted at high frequency by traders. This paper introduces a novel application of the Sequence Generative Adversarial Networks framework to model the order flow, such that random sequences of the order flow can then be generated to simulate the intra-day variation of prices. As a benchmark, a well-known parametric model from the quantitative finance literature is selected. The models are fitted, and then multiple random paths of the order flow sequences are sampled from each model. Model performances are then evaluated by using the generated sequences to simulate price variations, and we compare the empirical regularities between the price variations produced by the generated and real sequences. The empirical regularities considered include the distribution of the price log-returns, the price volatility, and the heavy-tail of the log-returns distributions. The results show that the order sequences from the generative model are better able to reproduce the statistical behaviour of real price variations than the sequences from the benchmark.

Shown in full with attribution under the source's licence. Licence: abstract CC0

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.