As we progress to smaller and faster circuits, we are gradually reaching the physical limits of materials and indeed some might say Moore’s law, and beginning to cross the threshold of applying the classical laws of physics. Enter quantum computing power.
The essential elements of a quantum computer come from Paul Benioff, who worked on the concept in 1981 at the Argonne National Labs in Chicago. Quantum computing is a research field focused on the development of computer technologies based on the principles of quantum theory that explain the nature and behaviour of energy and matter at the quantum level (atomic and subatomic).
The main quantum effects are superposition and entanglement. Quantum computers are those that harness quantum physical phenomena not available to conventional computers. Classical computers do their calculations via bits, whereas quantum computers on the other hand do theirs via qubits. This is because bits are formulated in binary form (0 and 1), having only two basis states, whereas qubits are non-binary (floating in-between). This ‘between’ is called quantum superposition and allows for the strange phenomenon of entanglement.
Because of entanglement, although two or more quantum objects are physically separated, their behavior is correlated. This pattern is consistent whether a centimeter or kilometer separates them (this is based on quantum science after all). So, while one qubit is situated in a superposition between two basis states, 10 qubits utilizing entanglement, could be in a superposition of 1024 basis states. Therefore, unlike the linearity of classical computers, quantum performance grows exponentially with the number of qubits. It is this ability that gives quantum computers the extraordinary power of processing a huge number of possible outcomes simultaneously.
A visual example of the difference in the ability of a classical algorithm versus quantum computation power could be that of a problem given to the machine to look at an energy landscape where the solution is to find the location of the lowest point. A typical algorithm would only ‘walk over this landscape’, thereby solving the solution one section of the map at a time, whereas a quantum computer would use quantum effects to go through hills and compute the solution (map) at the same time. No more little slow ‘loading’ line.
The ideal quantum computer is universally applicable and superior to classic machines. There are simply an unlimited number of solutions that a quantum computer could solve once it has been shown to be viable. The solution to these problems is at the heart of virtually every commercial application of the quantum computer.
A realistic mapping of the protein folding sequence would be an important scientific and medical breakthrough for instance, that could make more effective drugs and save lives. Using qubits, a quantum computer can solve several important issues, ranging from the chemical reaction in a molecule to a better understanding of some of the puzzles surrounding the unsolvable issues and costs of climate change.
Quantum computers are good at finding optimal solutions to problems with a seemingly infinite number of variables, protecting sensitive data and communications, and accurately simulating quantum phenomena and molecular behavior. In particular, quantum could also be the key to securing our digital future through secure software development processes. In the distant future, universal quantum computers might also revolutionize the field of artificial intelligence. Recent work has produced algorithms that could function as building blocks of quantum machine learning, but the hardware and software to fully realize quantum artificial intelligence is still as elusive as a general quantum computer.
Competition Amongst Prophets
But other breakthroughs are coming in fast. Google has claimed ‘quantum supremacy’ in October of this year, with their 54-qubit Sycamore processor. Said to be able to process a calculation in 200 seconds, that would otherwise have taken the world’s most powerful supercomputer 10,000 years. However, IBM has disputed this ‘supremacy’, saying that in fact “the same task could be performed on a classical system in just 2.5 days”. These vastly varying numbers demonstrate the fierce competition in this red-hot emerging technology, not to mention the uncertainty.
One case study of a financial institution (FI) and a big tech company working together is the case of J.P. Morgan and IBM, which were apparently able to use a Quantum Machine Learning Algorithm to formulate basic tenants of algorithms. The IBM report Getting your financial institution ready for the quantum computing revolution claims that “Quantum risk analysis is already being tested on European-style options”. Their quantum arm, IBM Q, additionally maintains that they have examples where the “actual implementation of quantum algorithms for European option pricing and portfolio optimization can be found”. The world’s derivatives markets could really always use more computation power, what with the global market cap estimated at between $500 Trillion and $1 Quadrillion.
Another FI looking to experiment in the field is Barclays, which as a large trading and clearing house, needs to ensure that the large volume of trades it conducts on the world’s stock, currency and commodity markets is done in the most optimized sequence and prioritization. Lee Braine, director of research and engineering at Barclays’ chief technology and innovation office said that “If only a small number of trades exist, the calculations can probably be done in your head. Scale up to 10-20 and you end up using paper. Beyond this is the realm of classical computing architectures. When there are hundreds of trades, classical computer algorithms begin to experience limitations.” Take into consideration that picking the optimal order of execution for 5,000 trades contains north of 4.2 x 1016,325 possibilities, this is not your average common algebra.
The reverence runs deep within the industry of quantitative analysts and the like. One Quantum Architect working at the banking giant HSBC told me that he sometimes “feels like a prophet”– in that he is continuously trying to forecast models based on continuously (and quickly) changing parameters of technology and tools. Imagine piloting a flying car whilst replacing different fuel types and altering the body of the vehicle. With $54 Trillion managed by just 28 global banks alone, I really hope that the ‘prophets’ are watched carefully.
One of the most reoccurring models in finance is that of the Mote Carlo Simulation, a ‘forking’ simulation where the algorithm seeks to find the most likely set of end results from a large number of random variables that can change and therefore ‘fork’ to the next step. This step concept is shown in the graph below.
Because the computing power of quantum opens an entire new realm of digital might, it beckons not just new opportunities but also serious dangers. Imagine North Korea’s cyberwarfare hacking branch, Bureau 121, getting their hands on quantum power with applicable uses for targeting foreign energy or financial infrastructure; shutting down the National Grid or the London Stock Exchange. Such systems may be able to be brought on-line quickly, but the outage could cause long-lasting damage, risking trillions in value and countless lives.
A security solution for the financial industry is sorely needed, one that quantum computing power could solve. Forbes crunched the numbers that came to “the typical American financial services firm is attacked a staggering 1 billion times per year”, which is “2,000 attacks per minute or over 30 attacks per second. The rate of breaches, or theft of sensitive data, in the financial services industry has tripled over the past five years.”
But with the range of problems that quantum could pacify, just looking at the financial sector, it is not a technology that an investment bank should overlook. From fraud and money laundering prevention to portfolio optimization and reaping the rewards gained from diversification, a new age in finance is dawning.
By Klaudia Archimowicz, Fintech and Investment Banking Expert at DisruptionBanking