Quantum Computing: The Future of the Financial Industry
Imagine the computer and all of the electronics you use on a daily basis, whether it’s to get a project done, delve deep into a podcast you’ve been catching up on, or even monitoring your personal stock portfolio. All of these applications are governed by two specific values: 0’s and 1’s. These are the fundamental building blocks of any digital electronic device and computer algorithm either being on or off in massive amounts and rapid succession to complete even the most complex tasks. Now imagine a computer that is not restricted by the limit of only two binary states and can operate by using orders of magnitude and more states as its fundamental building block with an ability to process and change these states at the same time. This is what quantum computers have to offer, but what would that mean for governments, institutions, or even you? Instead of using binary to encode and transfer information, quantum computers utilize quantum and physical phenomenon such as superposition, interference, entanglement and annealing to manipulate information and solve computational problems using these so called quantum bits or qubits. While traditional systems can only encode data in binary with two definite states, quantum mechanics allows for information in qubits to be in superpositions of states which allows for these systems to solve current problems that are unsolvable with current technology because of the lack in computational power traditional systems have. Even though quantum systems are still in the early stages of their development with adoption within the financial industry in closer to ten or more years, that doesn’t stop us from predetermining useful applications of these systems once they are more mainstream.
Capital markets become an interesting topic when discussing the potential of quantum computing because part of it are individuals seeking to minimize their overall risk and maximize their returns. As individuals, we all want computers to be able to speed up the analysis of these situations. This has created a computational race within the industry in which quantum computing will revolutionize, especially in topics such as the optimization of portfolios, quantum annealing and a more effective and efficient method of pricing derivatives.
Portfolio Optimization
One current issue in the world of the capital markets is the proper optimization of a portfolio to mitigate the most risk while retaining the greatest possible amount of return. With current technology, it is very difficult to efficiently determine an optimal portfolio due to so many parameters and constraints changing frequently over time and the many potential investment strategies which classifies it as an NP-hard problem. In a simple terms, an NP-hard problem is known as a non-deterministic polynomial time, where many non-linear optimization and search algorithms fall under. In NP-hard algorithms the execution time cannot be predicted using polynomial or exponential laws because of the very large size and the diversity of the input parameters which makes such algorithms very computationally expensive. Considering the fact that the markets change every second and can many times be unpredictable, quantum algorithms will allow us to properly optimize for the future through a more effective quantitative method.
Another known problem in optimization is finding global solution versus local solution. To put this concept of optimization into more perspective, think of a 3D landscape with many valleys and mountains. If a regular computer wanted to find the solution in to the problem of locating the lowest point in the landscape, it would have to test each point against each other individually until it found the optimal values. Now imagine how much more difficult it would become when we cross into multidimensional math problems. While many methods and algorithms have been developed — all of them require using different degrees of “brute force” search that is very computationally expensive process in multidimensional space.
Quantum Annealing Applications
Quantum annealing uses various phenomena like quantum qubits, energy states, state superpositions and quantum entanglement by integrating them into a single computational unit. This allows the problem to be formulated by configuring qubits and their interconnections, using quantum annealing to converge the entire quantum network to its lowest energy state. When the annealer finishes and all qubits in the quantum network settle, the final states of each quantum qubit will represent the answer to the initial problem in its lowest energy point — the global minimum (our desired answer). While we already mentioned how standard computers would handle the 3D landscape, an annealer would create a direct tunnel that cuts through all mountains and valleys directly to the best answer. For a bulge bracket bank like Barclays, they already have begun experimenting with quantum programs on IBM’s quantum cloud to gain a sustainable competitive advantage. Barclays hopes to see improvements in the optimization of large transactions that have varying levels of collateral, credit and constraints when it comes to asset liquidity. These large transactions are usually never instantaneous which creates inefficiencies in trades that need to go through in the moment which is where quantum computing hopes to step in.
Someone may argue that similar approach can be programmed using traditional methods and using traditional computers. While somewhat true for simple problems, this process becomes significantly more time consuming and difficult when you introduce functions with thousands or more highly dynamic and random parameters that constantly change over a short period of time, which typically represents how the financial markets behave. Quantum annealing will allow researchers to delve deeper into potential applications within the finance industry, such as creating more efficient arbitrage opportunities and finding the best way to execute large trades against the market through a more precise way of finding most optimal opportunities and optimizing for them.
Monte Carlo Methods and Derivatives
Quantum computing also has future promise in utilizing Monte Carlo methods to handle the pricing of derivatives with complexities and many uncertainties through a varied risk valuation. Traditionally, derivatives have been priced through the Black-Scholes equation, a model which makes the assumption that certain subclasses of assets such as derivatives follow geometric Brownian motion with a constant volatility and drift. This changes with Monte Carlo methods in mathematics, using random numeric sampling to map out many possible future price points with unique inputs.
Monte Carlo methods reduce the uncertainty of an outcome by allowing the individual controlling the algorithm to visualize many risk assumptions under different parameters and therefore using the model to predict potential future price points. The downside to the Monte Carlo model and why we aren’t currently using it is because of the amount of time it takes to generate and test all of these random values against each other, as well as potential issues in sampling bias. With current technology, Monte Carlo comes nowhere near as fast or convenient compared to the former model because the Black-Scholes equation utilizes predetermined parameters with set values compared to a fully randomized sample.
Quantum computers help solve this problem by computing random sampling calculations at an astronomically faster rate than current state of the art technology. To give an example, D-Wave is a leading Canadian company in this space with its own quantum computer named 2000Q. As of right now, 2000Q is currently millions of times faster than most personal computers in solving specific problems like search, random sampling and global optimization. It holds the power of 2000 qubits, a feat of accomplishment Google and IBM are struggling to keep up with. Although in early development and not being specifically used for the field of finance, the 2000Q is already being used by companies such as Google, Lockheed Martin and NASA for practical applications such as machine learning, code optimization and financial simulations.
Overall, I have a firm belief that quantum computing will be the future of the financial industry after ten or more years. Even though current application has not been fully realized yet due to the infancy of the industry and because it is only feasible for smaller calculations that can properly be expressed, that shouldn’t stop people from coming up with new solutions to currently existing problems through the promising future technology. While institutions that race to invest the most money and time into developing this technology may reap its future benefits among the industry, it still comes to one interesting question: will average consumers like me and you ever transition to quantum computers as personal devices in the future? We’ll have to see.
