Genetically Engineered Quantum Circuits for Financial Market Indicators

Floyd M. Creevey, Lloyd C. L. Hollenberg · 2025

Read Genetically Engineered Quantum Circuits for Financial Market Indicators by Floyd M. Creevey, Lloyd C. L. Hollenberg (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-2511-15739.txt.

Introduction

Floyd Creevey and Lloyd Hollenberg use genetically engineered quantum circuits to encode financial market indicators. Quantum finance needs fast loading of real data into quantum states. Their approach applies a genetic algorithm framework for state preparation to stock price data. Optimization targets both encoding quality and efficiency on current noisy hardware.

Encoded prices are processed on both a simulator and a real quantum computer. The team compute a singular value decomposition entropy as a market indicator. Results show improved quality over prior encodings and promise for more precise analysis. The demonstration uses real world data rather than toy inputs. Fast state preparation remains a gate to practical quantum advantage in finance.

Catalog record

CREATORFloyd M. Creevey, Lloyd C. L. Hollenberg
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE24 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2511-15739

Browse the full library · About the Tower of Babel

Full text

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

Genetically Engineered Quantum Circuits for Financial Market Indicators Floyd M. Creevey, Lloyd C. L. Hollenberg Abstract Quantum computing holds immense potential for transforming financial analysis and decision-making. Realising this potential necessitates the efficient encoding and processing of financial data on quantum computers. In this study, we propose using the GASP (Genetic Algorithm for State Preparation) framework to optimise the encoding of stock price data into quantum states and show it can enhance both the fidelity and efficiency of the encoding process. We demonstrate the efficacy of our approach by encoding stock price data onto both a simulated and real quantum computer to calculate the Singular Value Decomposition (SVD) entropy. Our results show improvements in fidelity and the potential for more precise financial analysis. This research provides insights into the applicability of GASP for the efficient encoding of real-world data, specifically stock price data, which is crucial for quantum advantage on noisy intermediate-scale quantum (NISQ) era quantum computers. I Introduction The financial sector is always looking for innovative tools and methodologies to address the challenges of risk assessment, portfolio optimisation, and decision-making. Quantum computing has emerged as a promising technology that could impact a range of financial analysis problems. However, the transformative power of quantum computing in the finance industry depends on its ability to efficiently encode and process vast and complex financial datasets. This paper presents the theoretical framework and demonstrates the practical effectiveness of using the genetic algorithm for state preparation (GASP) 4 for encoding stock price data onto quantum states over multiple qubits. To extract meaningful insights from this quantum-encoded data, Quantum Singular Value Decomposition (QSVD) 2 is used to compute the Singular Value Decomposition (SVD) entropy of the correlation matrix. This is an important indicator in various financial analyses 3 . The results show improvements in fidelity, computational efficiency, and the potential for more accurate financial predictions. By showing how GASP is useful for financial data encoding, valuable insights into the future of quantum finance are presented with implications spanning various financial applications, from risk management and asset allocation to market forecasting. This process requires precision and strategies that enhance the fidelity and efficiency of encoding, hence efficient and accurate generation of arbitrary quantum states with low depth and gate count is crucial for the future applications of quantum computing. Previous works have presented methods for state preparation that produce exact states 7 ; 12 ; 8 ; 10 ; 11 , however, the circuits produced by these methods generally have a depth that is infeasible to execute on real hardware. Other techniques reduce the amount of depth required by introducing ancillary qubits 9 ; 13 , another parameter that is limited on current hardware. However, exact state synthesis is not always a requirement for quantum algorithms, as demonstrated in 6 , which gives a further advantage to approximate state preparation techniques that have been developed such as 5 ; 1 . Another prominent method utilises general adversarial networks 14 . This work builds on work presented in 6 , which proposed an approximate amplitude encoding technique for financial market indicators and presented an alternate method for data loading. The current work follows a similar approach and demonstrates that GASP can generate circuits with sufficient accuracy to encode the stock market data required for computing the SVD entropy of stock price dynamics. Moreover, the circuits produced by GASP have a low gate depth, making them feasible to run on actual hardware. This enables analysis of the QSVD algorithm on real-world hardware. The results presented will use the methods from 6 and 14 for comparison. The rest of this paper is organised as follows; Section II will describe how the stock market data statevector to be loaded is generated. Section III will describe the variational quantum singular value decomposition (VQSVD) in detail, and Section IV will discuss current methods for data loading in VQSVD. Section V will explain the methodology of using GASP for data loading in VQSVD, and Section VI conclusions and potential future work. II Stock Price Data, Returns, and Correlation Matrices For some number of stocks, n=1,2,...,N_{s} , over some time period, t=1,2,...,T , the logarithmic rate of return can be written, r_{nt}=\ln(s_{n, t})-\ln(s_{n, t - 1}), (1) where s_{n,t} is the price of the n th stock at time t 3 . This logarithmic rate of return can be calculated between stock time points, meaning there are T logarithmic rates of return for each of the N stocks. The logarithmic rates of return can then be used to calculate the correlation matrix, C_{nm}=\sum_{t=1}^{T}a_{nt}a_{mt}, (2) where, a_{nt}=\frac{r_{nt}-\langle r_{n}\rangle}{\sigma_{n}\sqrt{N_{s}T}}, (3) and \langle r_{n}\rangle and \sigma_{n} are the average and standard deviation, defined as, \langle r_{n}\rangle=\frac{1}{T}\sum_{t=1}^{T}r_{nt},\;\sigma_{n}^{2}=\frac{1}{T}\sum_{t=1}^{T}(r_{nt}-\langle r_{n}\rangle)^{2}. (4) In Figure 1 (a) we show the time series data for stocks for Exxon Mobile (XOM), Walmart (WMT), Procter and Gamble (PG), and Microsoft (MSFT), and the corresponding log rates of return in Figure 1 (b). In Figure 2 we give the associated correlation matrices for these time series stock data. Figure 1: (a) Shows monthly stock opening price data for XOM, WMT, PG, and MSFT, for the period between March 2008 to March 2009, retrieved from Yahoo Finance. (b) Shows the calculated logarithmic rates of return r_{nt}=\ln(s_{n, t})-\ln(s_{n, t - 1}) for each n stock at time t . Figure 2: Shows the generated correlation matrices for XOM, WMT, PG, and MSFT, for the period between March 2008 to March 2009, retrieved from Yahoo Finance, T=5 , C_{nm}=\sum_{t=1}^{T}a_{nt}a_{mt} . The correlation matrix, C , is positive semidefinite, so has nonnegative eigenvalues and satisfies, \text{Tr}(C)=\sum_{n=1}^{N_{s}}\sum_{t=1}^{T}a_{nt}^{2}=1, (5) so for the positive eigenvalues \lambda_{1} , \lambda_{2} , …, \lambda_{K} , which satisfy \sum_{k=1}^{K}\lambda_{k}=1 , the SVD entropy is defined, S=-\sum_{k=1}^{K}\lambda_{k}\ln(\lambda_k). (6) These a_{nt} values can be loaded into a |\rm data\rangle state vector for use on a quantum computer 6 , |{\rm data}\rangle=\sum_{n=1}^{N_{s}}\sum_{t=1}^{T}a_{nt}|n\rangle_{\rm{stock}}|t\rangle_{\rm{time}}, (7) where \{|n\rangle\}_{j=1,2,\ldots,N_{s}} is the basis set in the stock index Hilbert space \mathcal{H}_{\text{stock}} and \{|t\rangle\}_{t=1,2,\ldots,T} is the basis set in the time index Hilbert space \mathcal{H}_{\text{time}} . Figure 3 displays the resultant statevector encoding of the generated correlation matrices. Note the |\rm{data}\rangle state defined by Equation 7 is normalised as a result of Equation 5 . Figure 3: Shows the generated statevectors from the correlation matrices of XOM, WMT, PG, and MSFT, for the period between March 2008 to March 2009, retrieved from Yahoo Finance, T=5 , C_{nm}=\sum_{t=1}^{T}a_{nt}a_{mt} . Taking the partial trace over \mathcal{H}_{\text{time}} allows the extraction of \rho_{\text{stock}} , \displaystyle\rho_{\text{stock}} \displaystyle=\text{Tr}_{\mathcal{H}_{\text{time}}}(|\text{data}\rangle\langle\text{data}|) (8) \displaystyle=\sum_{nm}C_{nm}|n\rangle_{\text{stock}}\langle m|_{\text{stock}}, (9) where C_{nm} is the (n,m) element of the correlation matrix given in Equation 2 , allowing the state \rho_{\text{stock}}(C) to be calculated of a quantum computer. Diagonalising \rho_{\text{stock}} allows the entropy to be calculated via VQSVD described in Section III . To understand how VQSVD is used to diagonalise \rho_{\text{stock}} , the Schmidt decomposition must be described. The Schmidt decomposition is used to represent a bipartite quantum state as a sum of products of two orthogonal states. Consider a composite quantum system composed of two subsystems, denoted as A and B . The Schmidt decomposition of a joint state |\psi\rangle_{AB} is given by: |\psi\rangle_{AB}=\sum_{i=1}^{N}\sqrt{\lambda_{i}}|a_{i}\rangle_{A}\otimes|b_{i}\rangle_{B}, (10) where \lambda_{i} are non-negative singular values, and |a_{i}\rangle_{A} and |b_{i}\rangle_{B} are orthonormal states in the

Read the complete text