Genetically Engineered Quantum Circuits Enhance Fidelity for Financial Market Indicators and Stock Price Data

Quantum computing promises to revolutionise financial modelling, but effectively translating real-world data into a format quantum computers can process remains a significant challenge. Floyd M. Creevey and Lloyd C. L. Hollenberg, both from the University of Melbourne, address this problem by pioneering a new method for encoding stock price data into quantum states. Their research introduces a framework, the Genetic Algorithm for State Preparation, which optimises this encoding process, demonstrably improving both its accuracy and efficiency. By successfully implementing this approach on both simulated and actual quantum computers to calculate a key financial indicator, Singular Value Decomposition entropy, the team reveals a pathway towards more precise financial analysis and unlocks the potential for quantum computers to deliver a genuine advantage in the current era of noisy intermediate-scale quantum technology.

The core finding is that approximate state preparation, achieved with GASP, offers a beneficial trade-off between accuracy and computational cost, potentially making quantum financial algorithms more practical on near-term quantum computers. The study compares GASP’s performance to other methods, demonstrating that GASP can achieve comparable accuracy with potentially lower resource requirements. Quantum state preparation, the process of creating specific quantum states for algorithms, presents a significant challenge, particularly on current, limited-scale quantum computers.

The research focuses on SVD entropy, a measure of information content in correlation matrices used for risk analysis and portfolio optimisation in finance. GASP uses a genetic algorithm to find an approximate quantum circuit that prepares the desired state. Current quantum computers, known as NISQ computers, are limited in size and prone to errors. The central theme of this work is the trade-off between accuracy and cost, arguing that sacrificing some accuracy in state preparation can significantly reduce computational cost and make quantum algorithms feasible on NISQ computers. The research focused on encoding stock price data, specifically logarithmic rates of return calculated from the prices of Exxon Mobile, Walmart, Procter and Gamble, and Microsoft over a defined time period, into quantum states. Calculations revealed that these logarithmic rates of return could be used to generate correlation matrices, essential for financial analysis, and these matrices were successfully encoded. The team measured the correlation between stock prices, calculating a correlation matrix from time series data, and then represented this data as a quantum statevector.

This process involved calculating logarithmic rates of return, averaging these returns, and determining standard deviations to normalise the data. The resulting normalised values were then used to construct the quantum statevector, representing the encoded stock market information. Experiments demonstrated that the GASP framework generates circuits with low gate depth, making them suitable for execution on current quantum hardware. The SVD entropy, calculated from the eigenvalues of the correlation matrix, provides a quantifiable measure of the encoded data’s information content. This work delivers a method for efficiently loading financial data onto a quantum computer, paving the way for more accurate financial predictions and advanced risk management strategies.

Efficient Quantum Encoding of Stock Prices

This research demonstrates a novel approach to encoding financial data for use in quantum computing, specifically focusing on stock price information. Scientists developed and tested the Genetic Algorithm for State Preparation (GASP) framework, which optimises the process of converting stock data into quantum states. Results show that GASP improves both the fidelity and efficiency of this encoding, a crucial step towards leveraging quantum computers for financial analysis. This work addresses a key challenge in quantum finance, namely the efficient loading of real-world data onto quantum hardware.

By enhancing the fidelity of data encoding, the research suggests the potential for more precise financial modelling and decision-making using quantum computers. The scientists acknowledge that the current implementation is limited by the scale of available quantum hardware and the inherent noise present in these systems. Future research will explore the application of GASP to larger and more complex datasets, as well as investigate methods to mitigate the effects of noise on the accuracy of the calculations. The team also intends to refine the algorithm to further improve its efficiency and scalability, paving the way for practical quantum financial applications.