How AI is reshaping portfolio construction

  0 Be the first to comment

How AI is reshaping portfolio construction

Editorial

This content has been selected, created and edited by the Finextra editorial team based upon its relevance and interest to our community.

Portfolio construction has long been a careful balancing act between risk, return, and investor psychology. Traditionally, it has been siloed, relying on historical data, human judgment, and static models. Classic portfolio construction follows a linear path: where risk tolerance and investment goals are assessed first, assets across predefined categories are allocated, and market shifts dictate how the portfolio is rebalanced.

While this model has worked for decades, in today’s world, it is limited because of its reliance on historical averages and manual oversight. In 2025, portfolios must be capable of rapidly adapting to fast-moving markets or nuanced investor behaviours. Gone are the days where clients can be treated as categories, rather than individuals.

In recent years, namely in 2022 and 2023, retail investors withdrew more money from mixed asset strategies than they invested. This shows that investors have been disappointed in their returns, and this can be put down to insufficient diversification or a lack of evolution of markets amid uncertainty and turbulence. Modernisation of both traditional equity and fixed income portfolios is overdue.

According to T. Rowe Price, the “traditional equity and fixed income ‘balanced’ portfolio needs modernisation.” Further, this portfolio “has not evolved to reflect today’s investment environment.” Beyond diversification, investors must know to make the most of the opportunities available for return enhancement and risk reduction for both equities and fixed income portfolios.

T. Rowe Price reiterates that portfolios should “include a variety of safe-haven assets with a view to increasing portfolio resilience in periods of market turbulence. More attractive yields within fixed income markets, in particular, mean that existing portfolio allocations may need to be revised.

“Finally, as we look ahead to an uncertain path with the possibility of both geopolitical and economic shocks, it is worth reviewing if tools such as volatility management may have a wider role to pay.” It is evident that in 2025 and beyond that there must be a shift from reactive strategy to proactive precision, while also reimagining what it means to invest.

This article will explore what is required for these changes to succeed and how AI can play a role in transforming how portfolios are built, managed and evolved.

Enter AI: Dynamic, data-driven, and personalised

Today, the investment management industry needs a competitive edge. AI can provide this by augmenting human judgement and basic qualitative models by rapidly analysing market data and risk factors through ML and predictive analytics. By picking up on subtle asset correlations within a range of alternative investments, portfolios can be transformed based on predicted market volatility.

Reinforcement learning and neural networks can also help optimise portfolio allocation with risk tolerance, while AI and ML also offers a holistic and real-time view of potential threats to portfolio performance. AI-powered early warning systems can detect subtle signals of impending market stress before they become apparent, ensuring portfolio managers are better prepared for adverse market conditions.

Optimisation algorithms can also enhance portfolio construction by establishing portfolios that are tailored to specific goals and risk profiles, without constant human intervention and in turn, without emotional biases. AI systems can also incorporate spending habits and values to create customised portfolios that extend to tax optimisation.

AI changes the game by introducing: 

  • Real-time data ingestion: Market signals, news sentiment, macroeconomic indicators. 

  • Predictive analytics: Anticipating market movements and investor reactions. 

  • Behavioural modelling: Understanding how individuals respond to volatility, gains, and losses. 

  • Hyper-personalisation: Tailoring portfolios to unique financial personalities, not just demographics.

The rise of alternative data

Alongside AI, alternative data is becoming an essential part of portfolio construction. Investment management leaders should take advantage of any opportunity to invest in new data resources, or risk being disintermediated. According to Deloitte, by 2030, “the global revenue generated by alternative-data providers across all industries could grow by as much as 29x, at a compound annual growth rate of 53%, and possibly reach US$137 billion.”

Where does this alternative data come from? Satellite imagery, social media, geolocation, news feeds, and metadata. These data sources can be leveraged to create additional income streams.

If the number of “alternative data sets applicable to financial services increased by ~36% over the last two years and the number of alternative-data providers increased by ~29% over the same period,” this means that investment firms that make “decisions without incorporating inputs from these alternatives data sets could leave out more information than they include in their investment decisions.”

Despite its promise, AI and alternative data-driven portfolio construction isn’t without risks. As is well known today, AI models are only as good as the data they are trained on. Poor data quality, errors and biases in the training data can lead to incorrect insights and in turn, inaccurate investment decisions.

Alternative data, on the other hand, can include sensitive or personal information that raises concerns about privacy and security, as well as potentially being in breach of data regulations like GDPR or CCPA. In addition to this, there are many transparency and explainability risks that must be addressed to avoid amplifying risks.

The hybrid future

Rather than replacing human advisors, AI is increasingly seen as a strategic partner. Advisors now use AI tools to enhance decision-making, monitor client behaviour and offer more personalised guidance. Leveraging human emotion in combination with computational power can create a hybrid, holistic approach to wealth management.

Moreover, hybrid portfolio construction combines different asset classes to create a diversified portfolio designed to balance risk and return. This allows firms to harness the growth potential of equities while providing the stability and income generation of fixed-income assets.

AI is not just reshaping portfolio construction; it’s redefining the investor experience. By moving beyond static models and embracing dynamic intelligence, wealth platforms can offer strategies that are more responsive, more inclusive, and more aligned with the realities of modern finance.

Comments: (0)

Editorial

This content has been selected, created and edited by the Finextra editorial team based upon its relevance and interest to our community.