Brief Description
HSBC and Quantum Dice have been exploring quantum random number generators (QRNGs) for stochastic modelling in insurance, specifically for pricing embedded guarantees. Initial findings, on this particular use case, reveal enhanced model accuracy and efficiency, enabling faster simulations with reduced computational costs or higher confidence outcomes.
Detailed Description
In financial services, stochastic models are essential tools for simulating various scenarios and assessing risks, including pricing embedded guarantees within insurance contracts. These models simulate various financial scenarios, relying on high-quality random inputs to predict future outcomes. Traditionally, these inputs are generated using pseudo-random number generators (PRNGs), which, despite being effective, are ultimately deterministic and can introduce biases.
Quantum random number generators (QRNGs) offer a novel approach by capitalizing on the inherent unpredictability of quantum mechanics to produce true randomness. A collaborative project by HSBC and Quantum Dice explored whether QRNGs could improve the accuracy and efficiency of actuarial models used for insurance modelling.
Central to this effort, Quantum Dice’s QRNG uses a beamsplitter to direct a single photon to one of two detectors with equal probability, harnessing the quantum principle of unpredictability. This technique allows for continuous monitoring of quantum entropy to ensure true randomness. These photonic systems are well-suited for high-demand applications like financial modelling and simulations due to their ability to operate at high speeds.
In the project, PRNG-generated inputs were replaced with QRNGs. This transition led to several significant improvements when testing at 1,000 simulations, a typical production level:
1. Enhanced Model Precision: When examining the moments of the random numbers, the standard error of the standard deviation was reduced by more than 50% with QRNGs. This suggests that QRNGs more consistently represented the underlying distribution across various sample choices.
2. Better Scenario Validation: Financial models often run numerous hypothetical scenarios that test how the model performs under different conditions. With QRNGs, the sample rejection rate decreased to less than 50%, compared to up to 90% with PRNGs, indicating more consistent and reliable modelling within the scope of this project.
3. Reduced Model Leakage: Model leakage refers to the gap between expected and actual results, which can lead to financial discrepancies. The use of QRNGs minimized this gap by more than 50% when compared to PRNGs, resulting in more accurate predictions.
4. Operational Efficiency: By providing more reliable random inputs, QRNGs paved the way for significant cost, time, and energy savings, or the ability to make better-informed decisions.
These advancements highlight the potential of QRNGs to improve the adaptability and responsiveness in financial models, particularly in rapidly changing markets. While quantum computing continues to evolve, QRNG technology is mature and ready for practical applications. Future research could focus on applying QRNGs to more complex economic models, potentially further enhancing their benefits.
Key Highlights
· Objective: The project aimed to assess whether quantum random number generators (QRNGs) could enhance the accuracy and efficiency of actuarial models used for insurance modelling.
· Technology Used: Quantum Dice’s QRNG applies quantum mechanics to generate true randomness, replacing traditional pseudo-random number generators (PRNGs) and providing more reliable inputs for financial simulations.
· Improvements Observed: In this instance, QRNGs significantly reduced model leakage (the gap between expected and actual results) by more than 50%, leading to more accurate financial predictions.
· Benefits: By reducing inconsistencies and improving scenario validation, QRNGs enhanced operational efficiency, allowing for faster, more cost-effective modelling, which could be used to achieve better-informed decision making.
· Practical Application: QRNG technology is a promising candidate for real-world insurance modelling, potentially offering enhanced adaptability and responsiveness to market dynamics.
Customer Testimonials:
· “HSBC has been exploring how quantum technologies could impact real business applications today. We are working with Quantum Dice to address whether quantum random number generators can enhance Monte Carlo simulations. These are important tasks in financial services with applications such as computing risk analytics and pricing complex derivatives. Our joint work with Quantum Dice started under an Innovate UK grant, and continued with a commercial PoC.”
- Philip Intallura, Group Head of Quantum Technologies at HSBC
· “This collaborative study with Quantum Dice allowed us a first look at what might be achieved by applying quantum technologies to our modelling applications and how it may benefit our work." - Michael Taylor, Head or Risk Methodology, HSBC
