Quantum Leap in Finance — Day 22
Day22 of #Quantum30 Challenge
Hello, readers! Day 22 of #Quantum30 Challenge was one of the very informative and interesting day. Today I dug into another interesting domain of Quantum Computing applications, and that is Quantum Finance. Let’s go!
The first resource is “Quantum for Finance” from the YouTube channel of Dwave by speaker Steve Flinter from MasterCard.
In this talk by Steve Flinter from MasterCard’s innovation division, he introduces the MasterCard Foundry as their research and development group focusing on new technologies. He discusses how they navigate new and emerging technologies, collaborate with partners, and share a demo of a Quantum Annealer application.
MasterCard Foundry’s interests include various emerging technologies such as cybersecurity, blockchain, and Quantum Computing. Steve outlines their focus on Quantum technology within three domains: Quantum Computing, Quantum Security, and Quantum Communications.
- Quantum Computing: Steve explains two different quantum computing models: annealing and gate-based. MasterCard believes quantum annealing will provide near-term results, while gate-based approaches will deliver long-term commercial value.
- Quantum Security: He highlights the concern around quantum technology breaking cryptographic systems and explains their work in post-quantum cryptography and quantum key distribution. The goal is to address quantum threats using quantum solutions.
- Quantum Communications: Steve discusses the potential impact of quantum communications on the industry and explores its future adoption.
He further elaborates on some specific applications they are exploring:
1. Offer Allocation: Steve introduces the concept of optimizing offers for cardholders and merchants, ensuring relevant and non-spam offers are assigned through quantum solutions.
2. Hidden Flow Discovery: This application involves uncovering obscured financial flows, particularly in the context of cryptocurrency transactions, by leveraging quantum annealing to improve tracking and tracing.
3. Fraud Detection: He describes their approach to optimizing machine learning models for fraud detection using quantum annealers to enhance the feature selection process.
4. Net Settlement: Steve explains their work on optimizing the settlement process for cross-border transactions involving various currencies. Quantum annealing is utilized to find the most efficient path for currency exchange and settlement.
He concludes by emphasizing that Quantum computing can bring significant efficiencies and advantages, especially in complex scenarios involving multiple currencies and transactions. The demonstrated applications show how quantum technology is becoming an integral part of MasterCard’s approach to innovation in the financial sector.
The second video is “Quantum Computing for Finance” from the YouTube channel Center for Quantum Technologies by the speaker Román Orús from the Institute of Physics, Johannes Gutenberg University.
The speaker begins by introducing the guest, Professor Roman Oroz from Spain, who has a background in quantum physics. They mention that quantum processors are becoming more powerful and could have applications in the financial sector due to their ability to solve complex mathematical problems.
The speaker gives a brief overview of the International Physics Center in San Sebastian, Spain, where they work, and how the center is involved in various quantum science activities and quantum startups, including applying quantum computing to finance.
They discuss the potential applications of quantum computers, emphasizing that while certain applications like material science, chemistry, and machine learning are known, the most important ones might still be undiscovered.
The main focus of the talk is on applications of quantum computing in finance. The speaker emphasizes that finance is a field full of complex mathematical problems and that quantum computers can offer speedups and accuracy in solving these problems. They mention that many entities, including banks, central banks, finance departments, regulators, and more, are interested in solving these financial problems using quantum computers.
The speaker delves into the types of problems in finance, categorized into optimization problems, machine learning problems, and Monte Carlo simulations. They explain that quantum computers can offer advantages in all of these categories. The speaker describes the International Physics Center in San Sebastian, its collaboration with startups and universities, and its active involvement in quantum science research.
Moving on, the speaker explains the concept of quantum optimization using various quantum algorithms, such as quantum annealing, the Quantum Approximate Optimization Algorithm (QAOA), and variational quantum eigensolvers. They discuss how these algorithms can be used to solve optimization problems in finance, like portfolio optimization, arbitrage opportunities, credit scoring, and prediction of financial crashes.
In portfolio optimization, the speaker explains how quantum algorithms can help determine the optimal composition of a portfolio by maximizing returns and minimizing risks. They discuss the inclusion of various constraints, such as transaction costs and investment limits.
Regarding arbitrage opportunities, the speaker explains how quantum algorithms can be used to identify profitable cycles in financial markets, taking advantage of price discrepancies between assets.
The speaker moves on to credit scoring, highlighting how quantum algorithms can assist in assessing an applicant’s creditworthiness by optimizing relevant features and minimizing correlations among them.
Discussing the prediction of financial crashes, the speaker explains how quantum algorithms can be used to analyze the stability of financial networks and identify potential instabilities that could lead to crashes. They mention the concept of first-order quantum phase transitions in financial crashes and how this analogy can help predict such events.
The speaker continues by mentioning collaborations with institutions such as D-Wave and IBM to implement these quantum algorithms and solve financial problems. They emphasize that, while there’s still much to explore, quantum computing holds great promise for addressing complex financial challenges.
The speaker’s presentation was a comprehensive exploration of the ways in which quantum computing could revolutionize the financial industry. They delved into the details of various applications, elucidating the potential advantages and challenges associated with each. One of the key takeaways was the idea that quantum computers could offer a substantial speed-up in solving optimization problems, which are prevalent in finance. The discussion revolved around portfolio optimization, where the goal is to find the best combination of assets to achieve a desired outcome while minimizing risk. Classical computers struggle with the sheer complexity of such calculations, but quantum computers, with their inherent parallelism, hold the promise of tackling these problems more efficiently.
The speaker also delved into the domain of quantum machine learning, explaining that quantum classifiers could play a pivotal role in fraud detection. In financial systems, where even a slight improvement in accuracy could translate to substantial monetary gains, quantum classifiers showcased their potential by outperforming classical counterparts in certain scenarios. They illustrated their point through a practical example involving credit card fraud detection. This case demonstrated that quantum support vector machines could yield improved accuracy, making them a promising tool for enhancing fraud prevention strategies.
Another fascinating topic covered was quantum amplitude estimation. The speaker introduced this algorithm as a hybrid of Grover’s and Shor’s algorithms, highlighting its efficiency in estimating various moments of probability distributions. The practical applications in finance, particularly in tasks such as pricing financial derivatives and risk computation, were underlined. It was emphasized that while there are challenges, quantum amplitude estimation could provide a quadratic speed-up over classical Monte Carlo methods, offering a potential leap in computational efficiency for such tasks.
However, the presentation was well-balanced, acknowledging potential pitfalls and hurdles. The speaker brought up the issue of classical data preprocessing, which could potentially overshadow the computational benefits of quantum algorithms. They also acknowledged that the accuracy of quantum calculations would be critical in finance, where precise predictions are essential. The talk prompted discussion about the practicality of these applications in the real world, acknowledging that while the concepts are exciting, there are still technological and implementation challenges to overcome.
In addition, the speaker painted a comprehensive picture of the intersection between quantum computing and finance. The talk was a balanced blend of theoretical insights and practical considerations, leaving the audience with a clearer understanding of both the potential and the limitations of quantum computing in the financial realm.
During a discussion following a presentation on quantum computing’s applications in finance, several questions and points were raised. One participant inquired about the advantages of tensor networks in optimization over existing tools. The speaker explained that tensor networks are efficient for computing the ground states of Hamiltonians due to their ability to handle highly correlated variables. In finance, where problems often involve almost random connections between variables, tensor networks proved effective and efficient. Moreover, different tensor network configurations could be explored to improve results.
A question arose about the comparison between quantum SVM and classical SVM. The quantum SVM showcased better accuracy, possibly due to the entanglement of variables. The speaker also clarified that the number of features and qubits used influenced the comparison. Another participant inquired about combining metrics to evaluate quantum and classical models’ expressiveness and complexity. The discussion centered on the accuracy of predictions and the comparison of quantum-inspired methods, with a focus on false positives.
The discussion shifted to dynamic portfolio optimization, where questions emerged about linear constraints and the optimization variables. The speaker clarified the optimization variables as omega vectors representing asset percentages at different times. The discussion also delved into financial cross-asset prediction, which involves predicting returns based on a financial network’s interactions and asset quantities.
A participant raised the point that quantum RAM could be difficult to practically implement, and Patrick, one of the authors, discussed the challenges and potential improvements in algorithms, including the applicability of HHL for specific cases. The topic of probabilistic quantum algorithms and the potential for specialized devices that replace Monte Carlo simulations was explored, with the suggestion that such devices might benefit from quantum phase estimation.
Overall, the discussion encompassed the technical nuances of various quantum finance algorithms, challenges related to quantum RAM, the practicality of certain approaches, and potential avenues for improved quantum algorithms in finance. Patrick emphasized that there is still much work to be done to fully harness the power of quantum computing in financial applications.
Thank you, readers! QuantumComputingIndia #Quantum30
