For advisers who manage their own model portfolios, fund selection and ongoing oversight can be one of the most time-consuming parts of the investment process. The investable universe continues to grow, data is scattered across multiple sources, and the pace of change in markets shows no sign of slowing.
Advisers therefore face a constant trade-off. Time spent on fund research, portfolio monitoring and documentation is time not spent with clients. Yet these activities are essential for delivering consistent results and meeting regulatory expectations.
Artificial intelligence (AI), and more specifically machine learning, is beginning to play a meaningful role in addressing this challenge. Not by replacing, but by supporting adviser businesses through — improved consistency, strengthened governance, and reducing operational burdens through automation. More broadly, AI has the potential to support advisers across four key areas of fund research and portfolio oversight.
- Clearer, more consistent client communication
One of the most important parts of managing portfolios is not making changes but explaining them. Whether it’s a fund switch, a rebalance, or a period of underperformance, advisers are expected to articulate their decisions clearly and confidently, often with time constraints.
AI can assist advisers by structuring trade rationales and portfolio commentary based on objective data. By analysing changes in asset allocation, risk metrics, and performance drivers, AI-supported tools can structure and create the initial communication which helps explain what changed, why it changed, and how it aligns with the client’s objectives and risk tolerance.
- Supporting asset allocation decisions
Asset allocation decisions require an interpretation of vast amounts of information: performance data, volatility metrics, correlations, macro signals, fund-level characteristics – often across disconnected systems, applications and data providers. Machine learning excels at aggregating and analysing this data holistically. By identifying patterns across asset classes and funds, AI can assist advisers with assessing portfolio positioning more objectively, identify emerging risks and concentrations, or even unintended exposures earlier.
AI also plays an important role in countering behavioural biases. Anchoring to familiar funds, reacting to recent performance, or delaying necessary changes are common challenges, even for experienced professionals. Data-driven signals provide a valuable counterbalance, supporting more consistent and repeatable decision-making.
- Proactive risk management through monitoring and alerts
Risk management can sometimes be treated as a periodic exercise, with checks carried out quarterly or bi-annually for investment committee reviews. In reality, risk evolves continuously – sometimes quietly, or sometimes very suddenly.
AI enables a more proactive approach. Machine learning models can monitor portfolios and underlying funds for all forms of changes – in volatility, correlations, drawdowns, or factor exposures – alerting advisers when something deviates from expectations. These alerts do not dictate action but help focus attention on areas where deeper analysis is genuinely required, improving both efficiency and oversight.
- Stronger due diligence and ongoing fund oversight
Initial fund selection is only part of the job. Ongoing due diligence – monitoring manager behaviour, philosophy, and consistency – is critical, yet often constrained by time.
AI can assist by reviewing manager materials at scale, analysing fund reports, commentaries, and disclosures to highlight changes in language, emphasis, or stated approach. These insights will help prompt timely follow-up and deeper qualitative assessment, without the need for an initial manual review.
AI agents and the future of fund selection
In the near future, advisers are likely to be supported by AI agents that assist with portfolio construction and fund selection – operating within clearly defined parameters, risk tolerances, and investment beliefs.
Rather than filtering funds through rigid screens, advisers will be able to search for funds in the same way we search the internet today: by describing the characteristics that matter to them. The following is an example:
Search for a global equity fund with a quality bias, downside resilience, low turnover, experienced management, and a clear valuation discipline.
An AI-driven system would then analyse the entire fund universe against this criteria – drawing on performance data, risk characteristics, factor exposures, manager behaviour, and qualitative disclosures – returning a shortlist of funds that best align with the adviser’s requirements. Importantly, those requirements will not be generic. The AI agent will operate within the adviser’s defined parameters: their client objectives, constraints, and – just as important – their core beliefs around investing. Those beliefs become embedded into the search itself.
This represents a real step forward in fund research. Instead of advisers adapting their process to fit the limitations of tools, the tools adapt to the adviser. Over time, these AI agents will also learn from decisions, understanding which recommendations were accepted or rejected, how portfolios evolve through different market environments, and how risk is managed across cycles. The result is a more personalised, consistent, and scalable investment process.
From a governance perspective, this future is equally powerful. Every search, shortlist, and selection can be recorded and explained, with fund choices evidenced against clearly articulated criteria at the point of decision-making.
At Collidr, our view is that this evolution is not about replacing advisers’ judgement, but augmenting it. AI will increasingly handle the heavy lifting of data analysis and pattern recognition, allowing advisers to focus on what machines cannot replicate: context, client understanding, and client service.
For advisers who continue to manage their own portfolios, the future of fund selection is likely to be faster, more intuitive, and more aligned to individual investment philosophies.
AI will not tell advisers what to think, but it will dramatically reduce the time spent on analytics and deliver more actionable intelligence.
Portfolio Adviser Article – 11 February 2028
The AI advantage in fund selection: Four ways it can support advisers
