By Andre Costa, CAIA, MCSI, CFP®, ERP®, SCR, FDP & Oswaldo Zapata, PhD
Quantum computing isn’t just a buzzword in the tech world—it’s set to become a strategic differentiator in the investment landscape. This post explores how quantum technologies, particularly quantum computing, could shape the future of finance, from smarter portfolio optimization to breakthroughs in machine learning models.
Before diving in, it’s worth clarifying that “quantum” covers more than one discipline. Broadly speaking, there are three: quantum computation, quantum communication, and quantum sensing. Here, we’ll focus on quantum computation—the idea that quantum processors can tackle certain problems faster, or with more precision, than even the most advanced classical computers. We’ll touch on quantum communication later, as its role in securing financial data is just beginning to emerge. Quantum sensing, on the other hand, is groundbreaking technology with intriguing potential for alternative investments—enough to merit a dedicated post on its own.
In this article, we will cover the fundamentals of quantum computing and where the technology stands today. We’ll also highlight how machine learning already powers much of modern finance. After this, look for a second article on this topic, which covers the important applications investors can utilize for portfolio management and communication.
As noted earlier, our focus here is on quantum computation—the branch of quantum technology concerned with using quantum-mechanical principles to solve computational problems. Traditional computers store information as either 0 or 1. Quantum computers use qubits, which can occupy a complex linear superposition of both states at once—unlocking entirely new computational possibilities. These combinations, manipulated through mathematical tools like unitary operators, and measurements, allow quantum computers to approach problems in fundamentally different ways from their classical counterparts.
Quantum mechanics itself is unlike the physics most of us learned in school—it doesn’t rely on everyday notions like force, velocity, or electric currents. Still, drawing parallels to classical circuits can help. A conventional circuit consists of electrical currents and logic gates (electronic components) arranged to solve a specific problem. These gates process information in a deterministic way: given the same inputs, you always get the same outputs. The efficiency of an algorithm is measured by how many gates and layers the circuit requires—its circuit complexity.
In a quantum circuit, the currents are replaced by qubits and the logical gates by quantum gates. Arrange these gates to tackle a problem, and you have a quantum algorithm. But here’s the key difference: quantum circuits are generally probabilistic. Measuring the same qubit twice may yield different outcomes—even if the circuit hasn’t changed—because quantum measurement “collapses the state” into a particular result.
Mathematically, a one qubit lives in a complex vector space, and quantum gates are unitary transformations that modify its state. n-qubit systems expand this to 2ⁿ possible basis states, enabling phenomena like entanglement, where combined qubits hold more information than they would individually. As in classical computing, there exists a universal gate set—a small set of gates from which any quantum operation can be built.
The case for quantum computing rests on two expectations:
Certain problems could be solved dramatically faster—sometimes exponentially faster—than with classical machines.
Some problems that are practically unsolvable today might become tractable, opening entirely new possibilities.
However, this promise is tempered by significant challenges. The most notorious is maintaining the fragile quantum states of qubits and gates long enough to perform computations. Environmental “noise” and tiny physical imperfections can cause errors, which must be corrected without directly measuring—and thereby destroying—the quantum information.
Error correction in quantum systems borrows from classical techniques, like using multiple redundant bits and majority voting to fix a flip from 0 to 1. But since qubits can’t be cloned, quantum error correction spreads the information across multiple physical qubits to form a logical qubit. Detecting errors often involves indirect checks (e.g., parity measurements), and fixing them without introducing new errors is a balancing act.
While fault-tolerant quantum computers—which can reliably detect and correct errors throughout a computation—remain a long-term goal, current devices operate in the Noisy Intermediate-Scale Quantum (NISQ) era. These machines have limited qubits and moderate number of gates, making fully robust quantum algorithms impractical for now.
To make progress, researchers are turning to hybrid quantum–classical algorithms. In these, a quantum processor handles the computationally hard parts while a classical computer manages the rest. Among the most promising approaches are variational quantum algorithms (VQAs)—a family of flexible, heuristic methods that may offer real-world advantages long before fully fault-tolerant quantum computers are available. Because many problems in finance share mathematical structures with those in physics and chemistry, VQAs hold potential for areas like portfolio optimization, risk modeling, and derivative pricing—even if the full proof of “quantum advantage” is still to come.
Artificial intelligence (AI) has already reshaped many aspects of the financial industry. From fraud detection to credit scoring to portfolio optimization, AI’s ability to process vast amounts of data and uncover patterns is helping institutions make faster, more informed decisions. Here, we focus on machine learning (ML)—the branch of AI concerned with algorithms that learn from data to generate actionable insights.
In finance, data is everywhere: market prices, trading volumes, corporate filings, economic indicators, customer transactions, sentiment feeds—the list goes on. But “data” is not all created equal. Structured data fits neatly into tables (think balance sheets or transaction logs), while unstructured data—such as news articles, audio transcripts, or images—requires more preparation before it can be analyzed. In both cases, quality matters. Poorly cleaned or incomplete data leads to weak models, and in finance, weak models lead to costly decisions. That’s why data preprocessing—cleaning, normalizing, and structuring data—is a critical first step in the ML pipeline.
Two of the most common ML approaches are:
Supervised learning: The algorithm learns from labeled data, where the correct answers are known in advance. This is widely used in credit scoring, where historical borrower data is used to predict whether new applicants are likely to repay. Supervised tasks include:
Regression (predicting continuous values, e.g., next quarter’s earnings)
Classification (assigning discrete labels, e.g., “fraud” or “no fraud”).
Unsupervised learning: The algorithm looks for patterns in unlabeled data. This is valuable when we don’t know what to look for—such as detecting new forms of fraud or finding hidden relationships between securities. A key challenge is dealing with high-dimensional data, where many features make analysis unwieldy. Principal Component Analysis (PCA) is a go-to dimensionality reduction technique that distills the data to its most important drivers.
Neural Networks and Sequential Data
Neural networks (NNs), inspired by the human brain, consist of interconnected nodes that process information in layers. They can be used for both supervised and unsupervised tasks. For time-series data, such as stock prices or macroeconomic indicators, recurrent neural networks (RNNs) and their improved variant, Long Short-Term Memory networks (LSTMs), are well-suited to capture patterns that evolve over time.
Applications in Finance
The range of ML applications in finance is broad:
Credit Scoring: Supervised models predict the probability of loan repayment using borrower demographics, credit histories, and financial metrics.
Risk Assessment: Beyond credit risk, ML can evaluate market, operational, and geopolitical risks by finding relationships between risk factors and outcomes like price volatility.
Fraud Detection: Algorithms like k-Nearest Neighbors (kNN) compare new transactions to historical data, flagging those most similar to known fraud cases.
Portfolio Management: PCA helps portfolio managers identify the most influential risk factors and remove redundant assets from the covariance structure, improving diversification.
Market Analysis: ML can detect shifts in investor sentiment or uncover hidden correlations between assets, commodities, and currencies.
Anti-Money Laundering (AML): Unsupervised clustering methods, such as k-Means, group similar transactions together, flagging outliers that may signal illicit activity.
Why This Matters for Quantum Computing
While classical ML methods already deliver value, many financial datasets are too large, complex, or computationally intensive for today’s processors to handle efficiently. That’s where quantum-enhanced machine learning could come in—accelerating pattern recognition, improving optimization, and enabling real-time analysis on scales that are currently impractical.
About the Contributors
André Costa, CAIA and Oswaldo Zapata, PhD in Theoretical Physics are the Founders of The Quantum Finance Boardroom (TQFB Community). André Costa, CAIA, is the co-founder of The Quantum Finance Boardroom (TQFB Community), as well as the founder of Free Enterprise Advisors in Brazil. Born in 1968, in Brazil, Costa graduated in veterinary medicine from Universidade Federal de Goiás in Goiânia, GO, in 1992. He is a CTA - Commodity Trading Advisor, registered with the CFTC and a member of the NFA, is registered as Investment Advisor with both the SEC in the U.S. and the CVM in Brazil.
Dr. Oswaldo Zapata is a specialist in quantum computing for finance, passionate about bridging the gap between theoretical financial solutions and their practical applications for finance professionals. This passion has led him to author several eBooks on the topic and educational materials, from introductory to advanced levels. He actively engages with professionals worldwide on LinkedIn and co-founded The Quantum Finance Boardroom, an online community where leading finance and quantum experts exchange ideas, collaborate, and explore new business opportunities. He holds a PhD in theoretical physics and lives in Dubai.
Learn more about CAIA Association and how to become part of a professional network that is shaping the future of investing, by visiting https://caia.org/