Paper: arXiv 2408.03181
Authors: Dominic Bauer, Derick Diana, Tim Gebbie
Abstract
We use random walks to simulate the fluid limit of two coupled diffusive limit order books to model correlation emergence. The model implements the arrival, cancellation and diffusion of orders coupled by a pairs trader profiting from the mean-reversion between the two order books in the fluid limit for a Lit order book with vanishing boundary conditions and order volume conservation. We are able to demonstrate the recovery of an Epps effect from this. We discuss how various stylised facts depend on the model parameters and the numerical scheme and discuss the various strengths and weaknesses of the approach. We demonstrate how the Epps effect depends on different choices of time and price discretisation. This shows how an Epps effect can emerge without recourse to market microstructure noise relative to a latent model but can rather be viewed as an emergent property arising from trader interactions in a world of asynchronous events.
Complexity vs Empirical Score
- Math Complexity: 8.0/10
- Empirical Rigor: 3.0/10
- Quadrant: Lab Rats — high math complexity, limited empirical testing
Why this score: The paper presents a sophisticated coupled reaction-diffusion PDE model with fractional calculus, requiring advanced mathematics to derive and solve, but it is a theoretical simulation study that lacks real-world backtesting, live data, or implementation-heavy validation.
Research Flowchart
flowchart TD A["Research Goal: Model correlation emergence in coupled limit order books"] --> B["Methodology: Random Walks for Diffusive Limit Order Books"] B --> C["Data/Inputs: Arrival, Cancellation & Diffusion of Orders"] C --> D["Computational Process: Pairs Trader Coupling & Vanishing Boundary Conditions"] D --> E["Key Findings: Recovery of Epps Effect & Emergent Correlation"]
Evidence Card
What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card
Proposes
The paper proposes a coupled reaction‑diffusion limit order book model that can generate the Epps effect as an emergent property of trader interactions.
Source sentence
We use random walks to simulate the fluid limit of two coupled diffusive limit order books to model correlation emergence.
Assumes 6
- A diffusion limit exists for the lattice parameters. model
Source sentence
Lattice parameters We assume that the diffusion limit exists Dα = lim ∆x→0 ∆t→0 r 2 ∆x2 ∆tα , (9) 2 where Dα is some diffusion parameter.
- Annihilation (cancellation) rates are constant across space and time. model
Source sentence
(3) We assume that the annihilation rates are constants: a(j)(x, t) = νj.
- The creation term can be decomposed into source, coupling, and shock components. model
Source sentence
(4) The creation term is separated further into a source terms s(j), coupling terms ℓ(j,k), and shocks δ(j) for the jth assets order book: c(j,k)(x, t) = s(j)(x, t) + ℓ(j,k)(x, t) + δ(j)(x, t).
- All stochastic driving forces share a single Brownian motion source, dropping the asset index. model
Source sentence
Here the in-formation shocks are synchronised and there is only one source for the stochastic force potential Vt, and we drop the j index.
Show 2 more
- The same diffusion parameter and fractional time parameter are used for all coupled order books. model
Source sentence
We will treat this parameter as the same across all the order books, similarly for the fractional time parameter α.
- Time increments are exponentially distributed to allow non‑uniform sampling. model
Source sentence
To simulate non-uniformly sampled grids in time we then assume that the time changes are exponentially dis-tributed: ∆t ∼Exp(λit).
Data
- Synthetic coupled order book price paths synthetic unknown
Source sentence
In Figure 1 we demonstrate price paths gen-erated using our model when considering a two-coupled order book using the parameters from Table 1. Price dynamics In Figure 2 we show snapshots of time increments ∆t of the effect an order book shock has on the mid-price Parameter Description Type Value L System length Fixed 200 M Number of divisions Fixed 400 r Probability of self jump Fixed 0.5 Dα Diffusion rate Free 0.5
Costs
Not applicable
Finds 5
- The model recovers the Epps effect, showing correlations vanish at decreasing time scales. theoretical
Source sentence
We are able to demonstrate the recovery of an Epps effect from this.
- The magnitude of the Epps effect depends on the choice of time and price discretisation. theoretical
Source sentence
We demonstrate how the Epps effect depends on different choices of time and price discretisation.
- There is a positive correlation between the cancellation rate ν and the size of the drop in correlation. statistical
Source sentence
There appears to be a positive correlation between the magnitude of ν and the magnitude of the drop in correlation.
- Increasing the sampling time Δt leads to larger drops in correlation. statistical
Source sentence
We found that increasing the sampling time ∆t leads to an increase in the magnitude of the drops in correlation.
- A cancellation rate ν close to 1 eliminates drops in correlation for both uniform and non‑uniform sampling. statistical
Source sentence
We found that a cancellation rate value of ν close to 1, given the other parameters are held constant, results in no drops in correlation for both cases.
Checked by others
Contradicted by
- NYSE Price Correlations Are Abitrageable Over Hours and Predictable Over Years
Compare the findings
This paper: Increasing the sampling time Δt leads to larger drops in correlation.
That paper: Cross‑stock correlations increase with time lag over the studied range.
Code
No repository by the authors found in the paper.
Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.
Cited References (61)
- Anomalous diffusion and price impact in the fluid-limit of an order book — → 2310.06079
- Simulating Anomalous Diffusion Models and Calculating Price Impacts of The Lit And Latent Order-Book in Julia; code (2023) — link
- Calibrating a Latent Order Book Model to Market Data (2022) — link
- High-frequency correlation dynamics: Is the Epps effect a bias? (2021) — link
- Code: Calibrating a Latent Order Book Model to Market Data (2022)
- Scaling in stock market data: stable laws and beyond (1997) — link
- A simple learning agent interacting with an agent-based market model (2022) — link
- A simple learning agent interacting with an agent-based market model (2023) — link
and 53 more
- New Equity Market Trading and Information Solution: Volume 00 – Trading and Information Overview (2013)
- JSE Matching Engine Simulator (2017) — link
- CoinTossX: An open-source low-latency high-throughput matching engine (2022) — link
- A global optimization heuristic for estimating agent based models (2003) — link
- Comovements in stock prices in the very short run (1979) — link
- Optimal execution with nonlinear impact functions and trading-enhanced risk (2003)
- Optimal execution of portfolio transactions (2001)
- Empirical market microstructure: The institutions, economics, and econometrics of securities trading (2007)
- Continuous auctions and insider trading (1985)
- Exponential resilience and decay of market impact (2011)
- Trading mechanisms and stock returns: An empirical investigation (1987)
- Price discovery and learning during the preopening period in the Paris Bourse (1999)
- Anomalous Price Impact and the Critical Nature of Liquidity in Financial Markets (2011) — link
- A fully consistent, minimal model for non-linear market impact (2015) — link
- From stochastic processes to numerical methods: A new scheme for solving reaction subdiffusion fractional partial differential equations (2016) — link
- Numeric Solution of Advection-Diffusion Equations by a Discrete Time Random Walk Scheme (2016) — link
- On the problem of calibrating an agent based model for financial markets (2013) — link
- The Problem of Calibrating an Agent-Based Model of High-Frequency Trading (2017) — link
- A comparison of economic agent-based model calibration methods (2020) — link
- Simulation and estimation of an agent-based market-model with a matching engine (2021) — link
- The Kolmogorov-Smirnov Test for Goodness of Fit (1951) — link
- THE ESTIMATION AND APPLICATION OF LONG MEMORY TIME SERIES MODELS (1983) — link
- Distribution of the Estimators for Autoregressive Time Series with a Unit Root (1979) — link
- An objective function for simulation based inference on exchange rate data (2007) — link
- Inference about the Tail of a Distribution: Improvementon the Hill Estimator (2010) — link
- Long-Term Storage Capacity of Reservoirs (1951) — link
- Robustness of the rescaled range R/S in the measurement of noncyclic long run statistical dependence (1969) — link
- A Simplex Method for Function Minimization (1965) — link
- Malliavin–Mancino Estimators Implemented with Nonuniform Fast Fourier Transforms (2020) — link
- Financial correlations at ultra-high frequency: Theoretical models and empirical estimation (2010) — link
- Market Microstructure Theory (2015)
- The High-Frequency Trading Arms Race: Frequent Batch Auctions as a Market Design Response (2015)
- High Frequency Trading and the New-Market Makers (2013)
- A fractional reaction–diffusion description of supply and demand (2018) — link
- From stochastic processes to numerical methods: A new scheme for solving reaction subdiffusion fractional partial differential equations (2016) — link
- The Epps effect under alternative sampling schemes: Dataset (2020) — link
- The Epps effect under alternative sampling schemes (2021) — link
- Using the Epps effect to detect discrete processes (2021) — link
- High-frequency Cross-correlation in a Set of Stocks (2000) — link
- A closer look at co-movements among stock returns (2009) — link
- Correlation based networks of equity returns sampled at different time horizons (2006) — link
- Time variation in the correlation structure of exchange rates: high-frequency analyses (2001) — link
- Impact of the tick-size on financial returns and correlations (2010) — link
- A closer look at the Epps effect (2001) — link
- Cross-correlation Measures in the High-frequency Domain (2007) — link
- Modeling the Epps effect of cross correlations in asset prices (2007) — link
- The Epps effect revisited (2009) — link
- A simple learning agent interacting with an agent-based market model: Julia code (2022) — link
- The Epps effect under alternative sampling schemes: Julia code (2020) — link
- CoinTossX (2021) — link
- Correlation emergence in two coupled limit order books in the fluid limit - Julia Code (2024) — link
- InteractingLOBs.jl (2024) — link
- Correlation emergence in two coupled limit order books in the fluid limit (2024) — link
1 of 61 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.