Financial Risk Management on a Neutral Atom Quantum Processor

TL;DR

This work tackles fallen angels forecasting in credit risk by marrying quantum-enhanced ensemble learning with a neutral-atom quantum processor. It introduces a QBoost-inspired classifier implemented via Random Graph Sampling and Tensor Network simulations, achieving competitive precision ($P \approx 0.28$) at a fixed recall ($R \approx 0.83$) using far fewer learners than a Random Forest and with faster runtimes. The study demonstrates hardware-tailored algorithms on near-term quantum hardware and offers a clear path to improved performance as qubit counts grow and negative couplings become available, supported by Tensor Network results showing potential superiority. Overall, the results indicate that quantum-assisted, interpretable ensemble methods can match strong classical baselines in financial risk tasks and pave the way for scalable quantum-accelerated decision-support tools in finance.

Abstract

Machine Learning models capable of handling the large datasets collected in the financial world can often become black boxes expensive to run. The quantum computing paradigm suggests new optimization techniques, that combined with classical algorithms, may deliver competitive, faster and more interpretable models. In this work we propose a quantum-enhanced machine learning solution for the prediction of credit rating downgrades, also known as fallen-angels forecasting in the financial risk management field. We implement this solution on a neutral atom Quantum Processing Unit with up to 60 qubits on a real-life dataset. We report competitive performances against the state-of-the-art Random Forest benchmark whilst our model achieves better interpretability and comparable training times. We examine how to improve performance in the near-term validating our ideas with Tensor Networks-based numerical simulations.

Figures (11)

• Figure 1: a. A binary classification problem deals with two classes (e.g. squares and triangles). The classifier is trained on the labeled training data and tasked to classify the instances, i.e. to find a decision boundary (black line) separating the test data. Considering that the test data comprising 2 features $k$ and $l$, the model either predicts correctly the class of an instance, thus increasing $T_{p/n}$ (green squares/blue triangles) or incorrectly, thus increasing $F_{p/n}$ (green triangles/blue squares). b. Classification performance of the classical solution on the test set shown as a confusion matrix. From the proportion of $T_p$ and $F_{p/n}$ obtained by the Random Forest model described in section \ref{['ssec:classical_solution']}, both recall $R$ and precision $P$ scores can be derived.
• Figure 2: a Precision-Recall curve (blue) calculated using different decision thresholds for the predictions. A close-up of the region of interest is shown in the following (dashed-red). b Precision value corresponding to the recall of 83% obtained through linear interpolation (green) between the neighboring points $R=82\%$ and $85\%$ on the PR curve.
• Figure 3: Comparison of QBoost performances, i.e. recall (blue) and precision (orange), with different base learners including decision trees, k-nearest neighbors, gaussian naive bayes and logistic regression.
• Figure 4: Random Graph Sampling pipeline for solving a QUBO $Q$ on a neutral atom based QPU. First, a QUBO, here with negative weights on the diagonal (red) and positive weights outside (green scale), is taken as input. From this QUBO, a trap pattern is devised and sent to the QPU, as well as the wanted number of repetitions and the pulse sequence used. At the beginning of each cycle, a first fluorescence picture enables to identify which traps were filled by an atom (bright spot). The system evolves according to $\hat{H}(t)$ and a final picture is taken to measure the collapsed state of the system, outputting a bitstring $w$. Using $Q$, we select $w_Q$ among $D$, the distributions of bitstrings obtained from repeating this process several times.
• Figure 5: Gap convergence obtained with classical uniform sampling (black), Simulated Annealing sampling (blue) and RGS sampling with optimized relabeling (red) for increasing size of QUBOs. The best gap found after some cycle repetitions is averaged over sets of $5$ QUBOs (plain line). b. Scaling analysis of the number of repetitions needed to reach a gap below a threshold of $1\%$ with respect to problem size. The results obtained by the three mentioned methods at sizes $N= 12$, $20$, $32$, $40$, $50$ and $60$ (dots) are fitted either exponentially or polynomially (line) depending on the best match.