Modelling Financial Market Imperfection Using Open Quantum Systems

Will Hicks \affilnum 1 · 2025

Read Modelling Financial Market Imperfection Using Open Quantum Systems by Will Hicks \affilnum 1 (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-2505-01284.txt.

Introduction

Will Hicks, writing alone in 2025, models imperfect financial markets as open quantum systems. The analogy treats the market as a quantum system whose state, a density matrix, evolves under influences from an external environment, just as prices respond to outside events. Distinct orbits of that evolution can be characterized, and their long-run behavior compared.

Both classical and non-classical modes of evolution end at the same maximum-entropy states, but classical systems get there faster, which reshapes the final probabilities. The non-classical modes earn their keep by capturing real market flaws. Illiquid trades and imperfect trading mechanisms fit naturally into different flavors of such evolution. The paper offers a physics-flavored language for market imperfection, aimed at researchers in quantum finance.

Catalog record

CREATORWill Hicks \affilnum 1
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE47 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-01284

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Modelling Financial Market Imperfection Using Open Quantum Systems Will Hicks \affilnum 1 Abstract We start with the idea that open quantum systems can be used to represent financial markets by modelling events from the external environment and their impact on the market price. We show how to characterize distinct orbits of the time evolution, and look at the development of the reduced density matrix, that represents the state of the market, over long time frames. In particular we distinguish between classical and non-classical modes of time evolution. We show that whilst both tend to the same set of maximum entropy states, this occurs faster in classical systems, with a knock on effect on the resulting probability distributions. We demonstrate how non-classical modes of time-evolution can be used to incorporate factors such as illiquid trades and imperfect trading mechanisms, and distinguish between different mechanisms of non-classical time evolution. keywords Quantum Finance, Open Quantum Systems, Von-Neumann Entropy, Ergodicity, Self-Referential Market, Endogenous Price Changes † † runninghead: Modelling Financial Market Imperfection Using Open Quantum Systems † † affiliation: \affilnum 1 Centre for Quantum Social and Cognitive Science, Memorial University of Newfoundland † † corresponding: Will Hicks 1 Introduction In [ 10 ] , the author looks at ways in which open quantum systems can be used to model the interaction between the financial market and its’ environment. In particular, the author looks at how a change in the prevailing risk appetite in the external environment can impact prices. For example, positive economic news regarding a sector may lead to an increase in equity prices for that sector. Alternatively a disappointing earnings announcement may lead to a fall in the stock price. In [ 10 ] , and [ 11 ] , the author shows how such events that occur outside the market, can be represented using a simple interaction between one Hilbert space representing the market, and a second representing the environment. In addition to providing new insight into the underlying mechanism for the kind of diffusion based modelling often applied to mathematical finance, the open quantum systems approach easily generalizes to modes of diffusion that have no classical equivalent. In particular the author emphasises the crucial role that the information entropy plays in determining the nature of the time evolution, and the resulting probability distributions. The author gives examples of non-classical modes of diffusion and uses numerical simulation to study the behaviour. In this article we attempt to understand these non-classical modes for diffusion in more depth, and to consider how they can be applied to the modelling of financial market imperfections. In section 3 we show how the time evolution occurs in distinct orbits that are fixed by the initial reduced density matrix. We also look at the stationary points within each of these orbits that act as stable attractors and represent the unique equilibrium point for the relevant system. The discussion and analysis in section 5 shows that whilst the long term equilibrium state does not depend on the nature of the time evolution, the rate at which the system approaches equilibrium is dependent on whether the underlying mode is classical or non-classical. In section 4 we discuss how non-classical modes for time evolution can be used to represent market imperfections, and also consider how non-classical observables can play a role in addition to the non-classical time evolution of the market reduced density matrix. We also look at how to derive Kolmogorov backward equations for the time evolution of the expected value for an observable, and show that systems where at least one of the density matrix or observable is diagonalized relative to the choice of basis that simplifies the presentation of the time evolution, are fundamentally simpler to understand than those where this does not apply. 2 Modelling Exogenous Prices Changes Using Open Quantum Systems: In this section, we give a brief overview of the analysis presented in [ 10 ] and [ 11 ] . This helps to provide context for the discussion that follows. 2.1 Hilbert Space Setup: In this article, for simplicity, we work in the discrete case, and represent the market Hilbert space using: \displaystyle\mathcal{H}_{mkt} \displaystyle=\mathbb{C}^{N} (1) The price operator acting on this Hilbert space is taken to be: \displaystyle X \displaystyle=\sum_{i=1}^{N}x_{i}|f_{i}\rangle\langle f_{i}| (2) Where |f_{i}\rangle represent the orthonormal basis for \mathbb{C}^{N} , and x_{i} represent the measured price values ( x_{i}>x_{j} for i>j ). The environment Hilbert space is given by: \displaystyle\mathcal{H}_{env} \displaystyle=\mathbb{C}^{K}\otimes L^{2}[\mathcal{K}] (3) Here the eigenvectors in \mathbb{C}^{K} represent different levels of environment risk appetite, and L^{2}[\mathcal{K}] represents the Hilbert space of square integrable functions over a compact space \mathcal{K} . The system Hamiltonian is given by: \displaystyle H_{sys} \displaystyle=\mathbb{I}\otimes H_{env}+H_{I} (4) \displaystyle H_{I} \displaystyle=\sqrt{\kappa\gamma}\sum_{\alpha\in\{u,d\}}A_{\alpha}\otimes B_{\alpha} \displaystyle\mathbb{I}\otimes H_{env} \displaystyle=\mathbb{I}\otimes\gamma\sum_{l=1}^{K}l|e_{l}\rangle\langle e_{l}|\otimes H^{\prime} Where H^{\prime} acts on L^{2}[\mathcal{K}] . B_{u}/B_{d} and A_{u}/A_{d} are described in section 2.2 . 2.1.1 Choice of the System Hamiltonian: The action of the system Hamiltonian on the market Hilbert space: \mathcal{H}_{mkt} defines the drift. For example, if we set \mathcal{H}_{mkt}=L^{2}(\mathbb{R}) , we could write: \displaystyle H_{sys} \displaystyle=-ir\frac{\partial}{\partial x}\otimes H_{env}+H_{I} to give the wave function a constant drift at rate r . However, for the sake of simplicity, and since in most cases we are working with Martingale probability measures, we assume zero drift. As described in [ 10 ] section 5.4, the choice of the environment space is partly pragmatic. Ie to ensure the convergence of the necessary time integrals. However, from a financial perspective the operator: \displaystyle\sum_{l=1}^{K}l|e_{l}\rangle\langle e_{l}| Acts on the different levels of risk appetite, returning an escalating scalar value to the different risk appetite levels. The operator: H^{\prime} acts on the space: L^{2}[\mathcal{K}] to configure the ease with which the system switches between these levels. 2.2 Exogenous Price Moves: The interaction Hamiltonian given by H_{I} in 4 is given by: \displaystyle H_{I} \displaystyle=A_{u}\otimes B_{d}+A_{d}\otimes B_{d} Here B_{u} and B_{d} are intended as operators that represent a jump up or down in the environment risk appetite (or pieces of good/bad news respectively). We write: \displaystyle B_{u} \displaystyle=\sum_{i=1}^{K-1}|e_{i+1}\rangle\langle e_{i}|\otimes\mathbb{I}\text{, }B_{d}=\sum_{i=1}^{K-1}|e_{i}\rangle\langle e_{i+1}|\otimes\mathbb{I} (5) The operators A_{u} and A_{d} represent the market reaction. For example the reaction to a piece of good news/an increase in the overall environment risk appetite (denoted A_{u} ) might be a jump in the underlying price. Similarly the reaction to bad news, might be a fall in the underlying price: \displaystyle A_{u} \displaystyle=\sum_{i=1}^{N-1}|f_{i+1}\rangle\langle f_{i}|\text{, }A_{d}=\sum_{i=1}^{N-1}|f_{i}\rangle\langle f_{i+1}| (6) Since, we can make observations only on the market Hilbert space, and not the external environment, in the Open Quantum Systems approach we look at the evolution of the reduced density matrix that is obtained after taking the partial trace over the environment space: \displaystyle\rho_{mkt} \displaystyle=Tr_{env}[\rho] 2.3 Markovian Approximation: In [ 10 ] proposition 5.6.1, and [ 11 ] proposition 6.1, the author derives a Markovian approximation to the dynamics of the reduced density matrix given by: \displaystyle\frac{d\rho_{mkt}(t)}{dt} \displaystyle=-Tr_{env}[H_{I}(t),\rho^{I}(0)] (7) \displaystyle+\sigma^{2}\Big(A_{u}\rho_{mkt}(t)A_{d}+A_{d}\rho_{mkt}(t)A_{u}-\frac{1}{2}\{A_{u}A_{d}+A_{d}A_{u},\rho_{mkt}(t)\}\Big) \displaystyle+\nu_{u}^{2}\Big(A_{u}\rho_{mkt}(t)A_{u}-\frac{1}{2}\{A_{u}A_{u},\rho_{mkt}(t)\}\Big) \displaystyle+\nu_{d}^{2}\Big(A_{d}\rho_{mkt}(t)A_{d}-\frac{1}{2}\{A_{d}A_{d},\rho_{mkt}(t)\}\Big) Here \sigma^{2} , \nu_{u}^{2} and \nu_{d}^{2} are determined by the environment state. We write: \displaystyle\rho(t) \displaystyle=\sum_{l,m=1}^{K}\rho^{lm}_{mkt}(t)\otimes r_{lm}|e_{l}\rangle\langle e_{m}|\otimes\rho_{B} (8) Where: \displaystyle\sigma^{2} \displaystyle=\kappa Tr\Big[B_{u}B_{d}\sum_{l,m=1}^{K}r_{lm}|e_{l}\rangle\langle e_{m}|\otimes\rho_{B}\Big]=\kappa\sum_{l=1}^{K-1}r_{ll} \displaystyle\nu_{u}^{2} \displaystyle=\kappa Tr\Big[B_{u}B_{u}\sum_{l,m=1}^{K}r_{lm}|e_{l}\rangle\langle e_{m}|\otimes\rho_{B}\Big]=2\kappa\sum_{l=1}^{K-2}r_{l(l+2)} \displaystyle\nu_{d}^{2} \displaystyle=\kappa Tr\Big[B_{d}B_{d}\sum_{l,m=1}^{K}r_{lm}|e_{l}\rangle\langle e_{m}|\otimes\rho_{B}\Big]=2\kappa\sum_{l=1}^{K-2}r_{(l+2)l} Taking N\rightarrow\infty , so that we can ignore boundary conditions, we can re-write equation 7 : Proposition 2.1 . Assume N\rightarrow\infty . Then

Read the complete text