Agent-based Liquidity Risk Modelling for Financial Markets Conference: Working Paper 008; UK; Simudyne 206
Perukrishnen Vytelingum, Rory Baggott, Namid Stillman, Jianfei Zhang, Dingqiu Zhu, Tao Chen, Justin Lyon · 2025
Read Agent-based Liquidity Risk Modelling for Financial Markets Conference: Working Paper 008; UK; Simudyne 206 by Perukrishnen Vytelingum, Rory Baggott, Namid Stillman, Jianfei Zhang, Dingqiu Zhu, Tao Chen, Justin Lyon (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-15296.txt.
Introduction
Perukrishnen Vytelingum and co-authors model the hidden cost of big trades in a 2025 paper. To sell a vast stake fast, a fund must take worse and worse prices as it goes. Each trade also leaks news that can move the price for good.
Their agent model fills a fake market with traders who act by their own rules. The model can then trace both the brief slip in price and the lasting shift as the market digests the news. Risk teams can test sale plans in the fake market first. Such dry runs may save funds from learning the cost of a rushed sale the hard way.
Catalog record
| CREATOR | Perukrishnen Vytelingum, Rory Baggott, Namid Stillman, Jianfei Zhang, Dingqiu Zhu, Tao Chen, Justin Lyon |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 43 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2505-15296
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Agent-based Liquidity Risk Modelling for Financial Markets Conference: Working Paper 008; UK; Simudyne 206 Perukrishnen Vytelingum, Rory Baggott, Namid Stillman, Jianfei Zhang, Dingqiu Zhu, Tao Chen, Justin Lyon Abstract . In this paper, we describe a novel agent-based approach for modelling the transaction cost of buying or selling an asset in financial markets, e.g., to liquidate a large position as a result of a margin call to meet financial obligations. The simple act of buying or selling in the market causes a price impact and there is a cost described as liquidity risk. For example, when selling a large order, there is market slippage – each successive trade will execute at the same or worse price. When the market adjusts to the new information revealed by the execution of such a large order, we observe in the data a permanent price impact that can be attributed to the change in the fundamental value as market participants reassess the value of the asset. In our ABM model, we introduce a novel mechanism where traders assume orderflow is informed and each trade reveals some information about the value of the asset, and traders update their belief of the fundamental value for every trade. The result is emergent, realistic price impact without oversimplifying the problem as most stylised models do, but within a realistic framework that models the exchange with its protocols, its limit orderbook and its auction mechanism and that can calculate the transaction cost of any execution strategy without limitation. Our stochastic ABM model calculates the costs and uncertainties of buying and selling in a market by running Monte-Carlo simulations, for a better understanding of liquidity risk and can be used to optimise for optimal execution under liquidity risk. We demonstrate its practical application in the real world by calculating the liquidity risk for the Hang-Seng Futures Index. 1. Introduction Liquidity risk management assesses the risk that a financial institution will not be able to meet its financial obligations, e.g., if a trader does not have enough cash, collateral or funding. It is an important part of risk management for all financial institutions. For example, banks are required to hold cash or collateral to meet their daily cash flow needs and to understand the liability of large positions in the market. Exchanges need to calculate the margin traders are required to post when trading derivatives and to manage close-out activities in the event of a default. If a position needs to be unwound and liquidated in the market, the cost of execution can be significant, particularly when trying to liquidate large positions within a short horizon. This work takes a novel approach to the difficult problem of evaluating how size and execution horizon affect price impact using an agent-based modelling framework. Market behaviour has been studied from many perspectives, including stochastic theories ( Cont and Mueller 2019 ) and statistical models of limit orderbooks (LOBs), theoretical equilibrium models and so on; however, most of them have limitations and simplifications. More recently, agent-based models have also shown promise at reproducing market behaviour ( Vyetrenko et al. 2020 ; Wang and Wellman 2017 ; Cliff and Bruten 1997 ; Gode and Sunder 1993 ; Bookstaber et al. 2018 ; McGroarty et al. 2019b ) . In an agent-based modelling (ABM) approach, we can model the problem without oversimplifying as a stylised model would. An ABM explicitly models the market as a discrete-step simulation with the mechanics of a Continuous Double Auction (CDA) of an exchange and trading agents with different trading strategies as we would find in a real market as shown in Figure 1 . The ability of ABMs to capture the heterogeneity of trading behaviours and the structural complexity of an exchange has been shown to be effective in recreating the well-known stylized facts in financial markets, such as fat-tailed distribution of returns, volatility clustering or auto-correlation in returns ( Potters and Bouchaud 2003 ) . Yet, a drawback of ABMs is in their modelling complexity which often allows only for a qualitative treatment of the results. Figure 1. The market simulation framework with market data input, virtual trading agents, the exchange with its gateways, engine and protocols and the buy and sell limit orderbooks which are managed by an auction mechanism. This leads to the idea of a minimally complex ABM for explanability while reproducing the liquidity and volatility patterns of real markets. We begin with the so called Zero-Intelligence (ZI) model, which despite its simple and random trader behaviours, is able to produce the certain statistical properties of the market, e.g., fat tailed distributions of returns ( Bookstaber et al. 2018 ) . Here, it is not only the trader behaviours that define market dynamics but the rules of the auction mechanism itself. A basic ZI trader submits limit and market orders at constant rates and orders on the LOB are cancelled at a constant rate. The price of a limit order is drawn from an exponential distribution. Importantly this ZI model can be calibrated to the historic data. While it has a well-understood behaviour and is able to capture some of the market features, the ZI model does not meet the requirement for a high-fidelity market simulator, but our ABM builds on similar principles. We can breakdown our ABM into two parts. First, we have the structure, i.e., the rules of the model, here, a set of known mathematical equations and algorithms that define the mechanics of the exchange matching engine and its protocols. Against these rules, we have the behaviour of the agents with trading and execution strategies. The financial market is a complex system, and the challenge is to build a model with minimally complex behaviours that can be calibrated to market data. Across the different models described in the literature, we observe that value, trend and noise trading strategies are sufficient to capture the stylised behaviours of the market ( McGroarty et al. 2019b ; Farmer et al. 2005 ; Wang and Wellman 2017 ) and we model these behaviours in this paper. Furthermore, we propose a mechanism where traders assume orderflow is informed and update their belief of the fundamental value as new information is revealed whenever a trade occurs. Agent-based modelling allows us to model behaviours of individual agents, including an agent that executes an order given an execution strategy, size and horizon. In the real world, a trader does not execute a large order (also referred to as a meta-order) immediately, but instead slices the order into many smaller ones. Intuitively, we might expect the price impact of the smaller orders to scale linearly, but the data describes a square-root law. In our ABM, as the meta-order is executed, we will empirically demonstrate in Section 5 how we reproduce the square-root law and how transient and permanent market impact emerge rather than be explicitly coded into our model. The outcome is a realistic transaction cost model that can calculate liquidity risk in a novel way by breaking down costs to market risk and market impact. The cost is calculated as an implementation shortfall, i.e., the difference between a reference price, e.g., the current mid-price and the executed trade price. By calculating costs