Abstract:In this study, the portfolio optimization problem is explored, using a combination of classical and quantum computing techniques. The portfolio optimization problem with specific objectives or constraints is often a quadratic optimization problem, due to the quadratic nature of, for example, risk measures. Quantum computing is a promising solution for quadratic optimization problems, as it can leverage quantum annealing and quantum approximate optimization algorithms, which are expected to tackle these problems more efficiently. Quantum computing takes advantage of quantum phenomena like superposition and entanglement. In this paper, a specific problem is introduced, where a portfolio of loans need to be optimized for 2030, considering `Return on Capital' and `Concentration Risk' objectives, as well as a carbon footprint constraint. This paper introduces the formulation of the problem and how it can be optimized using quantum computing, using a reformulation of the problem as a quadratic unconstrained binary optimization (QUBO) problem. Two QUBO formulations are presented, each addressing different aspects of the problem. The QUBO formulation succeeded in finding solutions that met the emission constraint, although classical simulated annealing still outperformed quantum annealing in solving this QUBO, in terms of solutions close to the Pareto frontier. Overall, this paper provides insights into how quantum computing can address complex optimization problems in the financial sector. It also highlights the potential of quantum computing for providing more efficient and robust solutions for portfolio management.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the multi - objective portfolio optimization problem, especially its application in the financial field. Specifically, the paper explores how to use quantum annealing technology to optimize a loan portfolio in order to balance "Return on Capital (ROC)" and "Concentration Risk", while meeting the goal of carbon footprint reduction. This problem is of great significance in practice because it involves how banks can reasonably allocate limited credit resources before 2030 to meet their missions, visions and sustainable development commitments, especially the requirements for reducing greenhouse gas emissions in the Paris Agreement. ### Problem Background Portfolio management is a classic financial problem, aiming to maximize returns and minimize risks by choosing the optimal asset portfolio. Traditional portfolio management methods mainly focus on a single objective: maximizing expected returns while complying with budget constraints and managing risks. However, in the modern financial industry, investment objectives have become more extensive, including multiple indicators such as environmental, social and governance (ESG) performance in addition to returns and risks. The conflicts between these objectives need to be resolved through multi - objective optimization methods. ### Specific Problem Description This paper studies a specific practical banking problem, that is, determining the optimal allocation of a bank's limited resources, especially the allocation of credit loan portfolios across industries, sub - industries and countries before 2030, to meet the bank's mission, vision and sustainable development commitments, including the low - greenhouse - gas - emission path and climate - adaptive development path specified in the Paris Agreement. ### Optimization Objectives - **Return on Capital (ROC)**: It measures the net income generated by the bank relative to the regulatory capital investment, reflecting the capital utilization efficiency and overall profitability. - **Concentration Risk**: Measured by the Herfindahl - Hirschman Index (HHI). The lower the HHI, the more diversified the portfolio and the smaller the risk. - **Carbon Footprint Reduction**: Ensure that the carbon footprint in 2030 is 24% less than that in 2021, which needs to be achieved by choosing appropriate loan portfolios and more efficient production methods. ### Mathematical Model To achieve the above objectives, the paper proposes the following mathematical model: 1. **Decision Variables**: - \(x_i\): The investment amount for the \(i\) - th loan in 2030. - \(X=\sum_{i = 1}^{N}x_i\): The total investment amount in 2030. 2. **Objective Function**: - Minimize \(\phi\cdot\text{HHI}(x)-(1 - \phi)\cdot\text{ROC}(x)\), where \(\phi\) is a weighting parameter used to balance the relationship between HHI and ROC. 3. **Constraint Conditions**: - The investment amount of each loan must be between the given lower and upper limits: \(LBi\leq x_i\leq UBi\). - Carbon footprint reduction goal: \(\sum_{i = 1}^{N}0.76x_ie_i\leq0.7E\sum_{i = 1}^{N}x_i\). ### Methods The paper uses two methods to solve this problem: 1. **Classical Convex Optimization Method**: Use linear matrix inequalities (LMI) and semidefinite programming (SDP) to solve the problem and generate the Pareto frontier. 2. **Quantum Annealing Method**: Transform the problem into a quadratic unconstrained binary optimization (QUBO) problem and solve it on a quantum annealer. ### Results By comparing the results of classical simulated annealing and quantum annealing, the paper finds that although quantum annealing can find solutions that meet the carbon footprint constraints to a certain extent, classical simulated annealing still performs better in terms of solutions close to the Pareto frontier. Nevertheless, quantum annealing shows its potential in dealing with complex optimization problems, especially when the performance of future quantum computers is further improved. ### Conclusion This paper provides a method for transforming multi - objective portfolio optimization problems into quantum annealing problems and shows the potential of quantum computing in solving complex optimization problems in the financial field. Although currently...

Multi-Objective Portfolio Optimization Using a Quantum Annealer

A real world test of Portfolio Optimization with Quantum Annealing

Comparative analysis of diverse methodologies for portfolio optimization leveraging quantum annealing techniques

Quantifying the advantages of applying quantum approximate algorithms to portfolio optimisation

A quantum online portfolio optimization algorithm

Quantum-Inspired Portfolio Optimization In The QUBO Framework

Reverse quantum annealing approach to portfolio optimization problems

Comparing Classical-Quantum Portfolio Optimization with Enhanced Constraints

PO-QA: A Framework for Portfolio Optimization using Quantum Algorithms

Quantum Portfolio Optimization with Investment Bands and Target Volatility

Quantum computing approach to realistic ESG-friendly stock portfolios

Diversifying Investments and Maximizing Sharpe Ratio: a novel QUBO formulation

Dynamic Portfolio Optimization with Real Datasets Using Quantum Processors and Quantum-Inspired Tensor Networks

Quantum Computing for Multi Period Asset Allocation

Benchmarking the performance of portfolio optimization with QAOA

Hybrid Quantum Annealing Decomposition Framework for Unit Commitment

Portfolio rebalancing experiments using the Quantum Alternating Operator Ansatz

A constrained multi-period portfolio optimization model based on quantum-inspired optimization

Portfolio Optimization of 60 Stocks Using Classical and Quantum Algorithms

Exploring the synergistic potential of quantum annealing and gate model computing for portfolio optimization

Multi-Objective Optimization and Network Routing with Near-Term Quantum Computers