Quantum Use Case #25: Collateral and liquidity optimisation

Short answer: This use case involves applying quantum computing to the complex mathematical problems of collateral optimization and liquidity management, specifically aiming to minimize funding costs while satisfying regulatory obligations. It is currently a research-stage opportunity where algorithms are being tested in simulation to handle high-dimensional variables, rather than an operational standard in live trading environments.

The Nature of the Optimization Problem

Collateral optimization is defined as the systematic allocation of financial assets to satisfy obligations or secure transactions. The primary objective is to minimize the cost of collateral required to mitigate risk associated with a particular transaction or a portfolio of transactions, while ensuring sufficient protection for the involved parties. This process is not merely a calculation of value; it is a multi-variable optimization challenge. It requires assessing numerous characteristics, including the cost of funding and the quality of the underlying assets, to ascertain the optimal collateral quantity to be posted.

The complexity of this task arises from the interplay of multiple sets of variables. In scenarios such as securities lending, the problem involves cross-optimizing these variables. As the number of variables and constraints increases, the optimization problem becomes increasingly complex. Traditional classical computing approaches may struggle with the combinatorial explosion inherent in such high-dimensional spaces. The proposed quantum approach suggests that certain subparts of the optimization can be outsourced to a quantum computer. These subparts are solved with higher accuracy and then brought back to the overall calculation. This hybrid methodology allows for the estimation of probabilities, such as the probability of default, which consists of many optimization variables, before making a lending decision.

Liquidity and Treasury Operations

Beyond static collateral posting, the use case extends to dynamic treasury operations, specifically liquidity and collateral management. The goal here is to optimize these operations in the presence of data availability and regulatory obligations. This is distinct from simple asset allocation; it involves managing the flow of funds and assets to meet daily margin calls and settlement requirements.

In the context of securities transaction settlement, the post-trade netting process aims to maximize the number of customer transactions settled. Failure to settle net positions efficiently can lead to several risks. First, there is replacement cost risk, characterized by the loss of unrealized gains due to delay in settlement. Second, there is liquidity risk, where liquidity pressures emerge if participants fail to settle net positions. Third, there is credit risk, which arises if there is a decline in the value of securities. Quantum computing is proposed as a tool to navigate these risks by optimizing the settlement process. By improving the efficiency of netting and settlement, the technology aims to reduce the exposure to these specific financial risks.

Research-Stage Potential and Simulation

It is critical to frame the current status of this technology as research-stage. Researchers have proposed and shown in simulation that quantum algorithms can approach these optimization problems. For instance, studies have explored how to approach collateral optimization for Noisy Intermediate-Scale Quantum (NISQ) devices. These simulations indicate that quantum methods can handle the complexity of the problem, but they do not imply operational adoption in live markets.

The potential benefits are often cited in terms of cost reduction. One source suggests a potential daily saving of up to 9% in terms of total collateral cost on a single margin call. However, this figure should be treated as a theoretical upper bound derived from research models, not a guaranteed operational metric. The realization of such savings depends on the ability to integrate quantum solutions into existing financial infrastructure, a challenge that remains largely unresolved in practice.

Limitations and Constraints

The application of quantum computing to collateral and liquidity optimization faces significant limitations. First, the problem requires high data availability. Optimization algorithms, whether classical or quantum, are only as good as the data they process. Regulatory obligations further constrain the solution space, requiring that any optimization result complies with strict legal and compliance frameworks.

Second, the technology is currently limited to NISQ-era devices. These devices are prone to noise and have limited qubit counts, which restricts the size of the problems that can be solved directly. Consequently, the proposed approach often involves a hybrid model, where only specific subparts of the optimization are handled by the quantum computer. This requires robust interfaces between classical and quantum systems, which are still under development.

Third, the transition from simulation to production is non-trivial. While researchers have shown that quantum algorithms can solve certain optimization subproblems with higher accuracy, the integration of these solutions into real-time decision-making processes is not yet established. Claims of automated decisions in real time are aspirational and based on the theoretical capabilities of the technology, not on current deployment.

Strategic Implications for Financial Institutions

For financial institutions, this use case represents an open opportunity to explore cost efficiencies in treasury operations. The ability to optimize collateral costs could lead to significant savings, but only if the technology matures beyond the research stage. Institutions should view this as a long-term R&D investment rather than an immediate operational upgrade.

The focus should be on understanding the mathematical structure of their own optimization problems to determine if they align with the types of problems quantum computers are best suited to solve. This involves identifying subparts of the optimization that can be isolated and solved quantumly. Until such time as robust, error-corrected quantum computers are available, the value of this use case remains theoretical.

FAQ

Is this use case currently deployed in live trading?

No. Most use cases are research-stage. Researchers have proposed or shown in simulation that quantum computing can assist in these optimizations, but there is no evidence of operational adoption in live trading environments.

What specific risks does quantum optimization aim to reduce in settlement?

It aims to reduce replacement cost risk, liquidity risk, and credit risk. These risks arise from delays in settlement, failure to settle net positions, and declines in the value of securities, respectively.

Can a quantum computer solve the entire collateral optimization problem?

Current proposals suggest that certain subparts of the optimization can be outsourced to a quantum computer. These subparts are solved with higher accuracy and then integrated back into the overall classical calculation, rather than the entire problem being solved quantumly.

Are the reported cost savings guaranteed?

No. Figures such as a 9% daily saving in total collateral cost are potential outcomes derived from research models. They are not guaranteed operational metrics and depend on successful integration and scaling of the technology.

What are the primary constraints on implementing this use case?

The primary constraints include data availability, regulatory obligations, and the limitations of current NISQ devices. The technology requires high-quality data and must comply with strict regulatory frameworks, while current hardware limits the size and complexity of problems that can be addressed.