It used to be a showdown between quants and fundamentals. But today, much like a rally race, portfolio managers are seeking a co-pilot to navigate volatility. Some find this in new AI platforms, providing modern portfolio managers with an extra gear to discover new opportunities and leverage current markets.
For every portfolio manager, the arrival of new Artificial Intelligence (AI) platforms at the right place and time is nothing short of luck. In fact, ignoring the benefits or choosing not to explore AI's use in the investment process seems almost risky. In a period where geopolitical risks, inconsistent monetary policies, and factor rotations are influencing all asset classes, it's crucial to leverage every available tool to make sense of it all and handle volatility. It's no wonder that many are combining traditional methods with AI to seize opportunities that previously went unnoticed. Quantitative or fundamental management? It might no longer be a matter of choosing sides. The flexibility and adaptability of AI has brought in a hybrid approach—sometimes referred to as quantamental—where human intelligence merges with the computational power of algorithms to design robust asset allocations. What will it be like to be a portfolio manager in the era of Artificial Intelligence?
Portfolio Construction as a Science
The idea that rewards are proportional to uncertainty is as old as commerce itself. Yet, turning this concept into a coherent and scientific system took centuries of research. Today, we can explain the forces that affect the expected returns of each security and asset class. However, the academic debate unfolded systematically, granting quantitative finance its first Nobel Prizes. But behind the scenes, the duel between skill and discipline has never ceased to fuel discussions. The more time passed, the more the ability to beat the market was understood in its basic elements. As art transformed into science, it suggested the existence of a systematic process for managing portfolios. This led to the development of the risk premium theory, a coherent system for categorising rewards related to specific risks such as size, indebtedness, or relative volatility to the market.
After discovering the equity risk premium in the mid-1960s, many other premiums emerged, giving birth to the family of so-called investment factors—investment strategies capable of consistently beating the market and the precursors to modern smart beta products. The traditional approach that relied on intuition, insight, and experience gave way to a different interpretation: good forecasts weren't the only essential element. They had to be combined with a solid, calculated, and practically engineered asset allocation to create performance and diversification. If predicting expected returns, volatility, and correlations relied on classical econometrics, quantitative finance reinforced it with a dose of engineering. The definition of risk-return targets, weights, constraints, and rebalancing frequencies laid the foundations for the theoretical framework known today as the science of portfolio construction. It was only by combining these two efforts that two key concepts emerged over time, both theoretically and practically: Forecast Alpha and Operational Alpha, related to portfolio optimisation.
In a sense, the first inherited the classical interpretation of Alpha: the ability to beat the market with more accurate forecasts of volatility, returns, and expected correlations between different assets. The true innovation, from an analytical perspective, came with the introduction of the latter. Under this light, Operational Alpha, also known as Portfolio Construction Alpha, can be defined as the ability to construct a portfolio over time (and at each rebalancing) that remains efficient as time progresses. For portfolio managers, this is an extremely delicate matter. Operational Alpha represents the benefit of investing in a portfolio that positions itself at the efficient frontier at every turn.
Over time, the gap between fully quantitative and fundamental portfolio managers is converging into a mixed and synthetic approach, often referred to as quantamental. Given these premises, it's no wonder that the excitement is palpable. Alpha, in a way, resembles Lavoisier's famous law: it doesn't create or destroy but transforms, changing in form. Once seen in forecasts, it is now a mix of forecasts and portfolio engineering. The questions may change, certainly, but the questions themselves hardly change. Whether it's AI platforms or capturing the new nuances of Alpha, every portfolio manager needs to equip themselves with the tools and intelligence to seize opportunities in any market environment.
The Modern Portfolio Manager: Bridging Theory and Practice
When historians trace the market's course, cyclicality and unpredictability are often presented as opposing concepts. Some see the peaks and troughs as the predictable result of invisible market cycles. Others interpret numbers as the reflection of probabilistic structures deduced through statistical approaches. But if recent market developments haven't been challenging even the most experienced managers, an additional signal is now asking an AI to analyse impartially how markets have changed in the past 15 years. Regimes, cycles, and phases of alternation undoubtedly exist. Yet, their alternation isn't regular or uniformly distributed over time. This unexpected complexity of markets highlights the need to bridge the gap between theory and practice continuously.
From this perspective, it's clear why the mentality of decision-makers in portfolio management activities has evolved more than market movements. Along with concepts such as risk-return, correlation, variance-covariance matrices, regression, beta, Sharpe ratios, and drawdown, which have officially become part of the modern portfolio manager's vocabulary, several conceptual rigidities have been overcome. The division between absolute predictability and unpredictability has been abandoned. Instead, the modern approach seems to gather the best of both worlds: the structure and theoretical framework of tradition combined with the adaptability of new tools, allowing for real-time recalibration of decisions as market conditions change.
The use of technology has led to a collective update of the decision-making process. In particular, Artificial Intelligence (AI), often integrated through specialised platforms or APIs, has become a common tool for analysing vast amounts of data and identifying potential investment opportunities. According to Forrester, large corporations, including major banks and financial institutions, have already invested around 64 billion dollars until 2025 to update their processes with AI software. However, this evolution doesn't mean the traditional approach has retired. What emerges is a landscape where market practice acts as a bridge between classical theory and emerging trends. Consequently, more and more managers and wealth advisors see this new technology as a means to surpass the traditional approach and build portfolios tailored to specific objectives. Notably, this is an area where technology has significantly accelerated the transition from frequently using model portfolios to a more modern, elastic approach. Thanks to new AI software and platforms, it's now possible to build investment portfolios (and adjust their positions, rebalance, and perform real-time risk management) based on specific client preferences and market views.
How to Leverage AI in the Investment Process
Born about twenty years ago to explore the mechanisms of our decision-making process, behavioural finance has gradually revealed two significant insights that portfolio managers should ponder. First, both too little and too much information have negative effects on our decision-making ability. Second, the human mind reasons quite differently from a computer. When faced with a choice, people think synthetically and nonlinearly, interpreting data within a mental framework. In contrast, a computer interprets data according to the patterns it represents, without emotions or the ability to assess the accuracy of the initial data. Hence, the challenge remains open: portfolio managers must understand and utilise new technologies to distinguish relevant information from irrelevant information effectively.
Integrating AI into the investment process becomes a valuable tool, provided it is guided and integrated in support of specific activities with awareness. While AI can provide valuable inputs—both predictive and operational—the portfolio manager will always remain the conductor of an orchestra, with the ultimate responsibility for investment decisions. This is evident when observing the adoption rate of AI among financial institutions, as shown by a recent study conducted by Stanford University in collaboration with Google, Bloomberg, and McKinsey. According to the study, the sectors where AI has been most successfully adopted are processes, risk management, and product innovation in the financial industry.
Among the most appreciated applications by wealth managers and portfolio managers are asset class and sector positioning indicators, scenario and regime analyses that leverage AI to discover new diversification opportunities in each market phase. As Tommaso Migliore, co-founder of MDOTM, the Fintech company that develops AI investment solutions for institutional investors, notes, "In an era where technological innovation becomes ubiquitous and AI is adopted by more professionals, financial institutions finally enter an era where fundamental and quantitative approaches can converge into a new investment paradigm. On one side, the intuition, imagination, and foresight of human intelligence, on the other, execution, calculation, and learning from data. The alphabet of investments will be a language that unites human intuition and experience with the power and precision of machine learning techniques. With the growing prevalence of Artificial Intelligence benefits, portfolio managers must grasp it adequately and utilise it effectively: the race has just begun."
Quantamental, the New Paradigm?
While financial institutions' footsteps indicate an increasingly intimate integration of humans and AI, the latter arrives as the latest chapter in a process where banks, insurance companies, and funds have used new technologies for over fifty years to improve processes and enhance service quality. These moves undoubtedly represent a new strategic asset, but they all share an extremely pragmatic foundation: will AI truly replace portfolio managers, or will it be a new generation of managers who use AI appropriately to replace the previous generation? The entire financial world is betting on the latter.
In this regard, the opportunity to restore or renew the investment process with new technologies leads to a fluid scenario capable of reshuffling the deck. Through collaborations, partnerships, and joint development of new investment solutions, investment houses are achieving more efficient resource use, productivity advantages, and better investment results. Portfolio management is one of the areas most involved and interested in this transition, where the role of AI platforms is similar to that of a co-pilot in rally races: providing continuous support for quickly deciphering and directly integrating rapidly evolving market scenarios into investment portfolios. From designing investable universes to market regime analysis and tactical and strategic asset allocation decisions, the new quantamental approach transcends the traditional gap between fundamentals and quantitative methods. The new investment language will blend human intuition and experience with the power and precision of machine learning techniques. As the benefits of Artificial Intelligence become increasingly widespread, portfolio managers must seize them adequately. The race has just begun.
