Engagement with technology
Interdisciplinary approach
Historical perspective
Picture this: you’re sitting in a café on Montparnasse, sipping your morning coffee, and reading the news that IBM has taught its quantum computers to assemble investment portfolios. The first thought – «Finally! Machines will replace financial advisors.» The second, more cynical – «Or is this just another bubble, only this time a quantum one?»
Markowitz meets Schrödinger
Let’s revisit the classics. In the 1950s, Harry Markowitz proposed an elegant formula: find the balance between risk and return. His «modern portfolio theory» ticks like a Swiss watch – at least in theory. In practice, the tiniest change in input data can flip the entire portfolio upside down. It’s like trying to predict the weather in Paris based on the temperature in Marseille two winters ago.
The problem of scale makes the challenge even more fascinating. When you’re choosing from thousands of bonds for an ETF, even the most powerful classical computers start to «gasp for air.» This is where quantum computers step onto the stage – like a deus ex machina of modern finance.
Quantum magic or an expensive trick?
IBM researchers decided to apply the so-called CVaR Variational Quantum Algorithm (CVaR-VQA) – a rather pompous name for a method that, in essence, makes a quantum computer «guess» good investment choices. Imagine a very expensive roulette wheel, but instead of a ball, qubits spin around – the quantum analogues of bits.
What’s especially intriguing is their take on risk. Instead of focusing on the average outcome (the usual investor’s approach), the algorithm zooms in on worst-case scenarios – Conditional Value at Risk, or CVaR. It’s as if you planned your vacation not around the average temperature, but around the coldest days of the year. Pessimistic? Perhaps. Rational? Absolutely.
Experiment: 109 qubits vs. Wall Street
The real fun began when the researchers ran their algorithm on IBM’s Heron quantum processor with 109 qubits. To grasp the scale: imagine trying to conduct an orchestra of 109 musicians, each capable of playing in two states at once – all while dealing with constant interference.
The results were surprising. The quantum algorithm managed to find a solution with only 0.49% deviation from the optimal. For comparison: a typical portfolio manager would be delighted to achieve that level of precision just in choosing their morning coffee – let alone billion-dollar investment decisions.
But here’s the paradox: the task assigned to the quantum computer was artificially simplified to just 109 bonds. A classical computer could have solved it in a few seconds. It’s like using a Formula 1 car to fetch bread from the bakery around the corner – technically impressive, economically dubious.
A hybrid reality
The most surprising discovery was that the best results came not from a purely quantum approach, but a hybrid one: the quantum algorithm generated candidate solutions, and the classical computer refined them using local search. It’s akin to the work of a seasoned sommelier who first picks a few wines by instinct, then compares their qualities analytically.
Researchers also noticed that more complex quantum circuits – the ones harder to simulate classically – delivered better results. Perhaps this is the key to future quantum advantage: not in the simple tasks we already know how to solve, but in those that demand entirely new approaches.
The human factor in quantum decisions
Beneath all this technological grandeur lies a deeper psychological truth. People don’t fail at portfolio optimization because they lack computing power, but because we are irrational. We sell in panic and buy in euphoria. Quantum computers are free of these flaws – yet inherit new ones.
Quantum «noise» – the inevitable errors in computation – can be seen as the technological twin of human irrationality. The difference is, at least this noise is predictable and steadily reduced as technology advances.
The economics of quantum investing
Let’s talk money. IBM’s Heron quantum computer costs millions of dollars. A typical portfolio optimization requires thousands of algorithm runs. Even if we assume the quantum solution is 1% more accurate than the classical one – will such an investment ever pay off?
For a hedge fund managing tens of billions, a 1% improvement means hundreds of millions in extra profit. But for most asset managers, a good classical solution – found in minutes on a regular computer – is more than enough.
Paradoxes of quantum finance
The study revealed several paradoxes. First: quantum algorithms perform better in combination with classical methods than in isolation. It’s like the most advanced GPS working best when paired with a paper map.
Second: the more «quantum» the algorithm (the harder it is to simulate classically), the better the results. This suggests the true quantum edge may lie not where we’re looking now.
Third: the problems that genuinely require quantum computers are still too big for today’s devices. We’re stuck in a peculiar in-between – advanced enough to see the potential, yet too limited to realize it.
The future: evolution or revolution?
The history of financial innovation is full of false prophecies. In the 1960s, people believed computers would make markets fully predictable. In the 1990s – that the internet would eliminate all intermediaries. In the 2000s – that mathematical models would eradicate risk. Each time, reality proved more complex than expected.
Quantum computers in finance may follow a similar path. Perhaps their real value won’t be in replacing classical methods, but in tackling entirely new challenges – like modeling complex market interactions or optimizing in real time.
Philosophical reflections
In the end, portfolio optimization is an attempt to predict the future using the past. Quantum computers may do it more efficiently, but fundamental uncertainty doesn’t go away. Markets remain a system where millions of irrational individuals make decisions on incomplete information.
Perhaps the true value of quantum finance lies not in sharper forecasts, but in a new understanding of market uncertainty itself. As Niels Bohr famously put it: «Prediction is very difficult, especially if it’s about the future.»
For now, quantum computers in finance look more like very expensive calculators than revolutionary tools. Yet history shows: the most important discoveries often begin as «costly toys.» Time will tell whether they become the new standard – or remain a curious technical footnote in the chronicles of financial technology.
