Making Sense of Artificial Intelligence

This article is part of an educational series designed to explain how we think about investing and client service at Lulworth Investment Partners. It is intended for general information only and should not be viewed as investment advice or a recommendation to buy or sell any asset. Anyone considering an investment should reflect on their own objectives, financial position and tolerance for risk, and seek regulated financial advice if needed.

Artificial Intelligence now features prominently across the wealth and asset management industry. Many firms present it as a way to improve investment decisions, enhance the client experience and operate more efficiently. In some cases, adopting AI is framed as essential to remain competitive.

Our view is more balanced. We see clear potential in Artificial Intelligence and are actively using it where it genuinely adds value. At the same time, we do not regard it as a solution in its own right. The usefulness of AI depends on how it is applied and the role it is asked to play. In our view, technology should support informed decision making, not replace it. This reflects a broader difference between Lulworth and much of the industry. Our business is designed around judgement, clear accountability and long term client relationships, rather than scale.

AI and the Modern Wealth Management Model

Over recent years, wealth management has moved towards larger and more centralised operating models. Consolidation, often through M&A, has helped firms manage costs, standardise investment processes and meet growing regulatory demands. Technology has played an important role in supporting this shift.

AI fits naturally into this framework. It supports model portfolios, automated portfolio adjustments and standardised client communications. From a business perspective, this can bring consistency and efficiency. It can also change the nature of the client relationship.

In large organisations, responsibility is often shared across committees, models and systems. Clients may interact more with processes than people, and decisions can feel remote from the individuals whose capital is at risk. When markets are calm, this structure can work reasonably well. When conditions are more volatile, accountability can be harder to identify and reassurance harder to provide.

At Lulworth, our operating model is deliberately different. We manage a limited number of client relationships so that responsibility remains clear. Each client works directly with an investment manager who is responsible for portfolio construction, ongoing management and communication. That responsibility is not delegated to systems or diluted across layers of decision making.

Personal Service and the Limits of Automation

AI is effective at processing large amounts of data quickly and consistently. This makes it useful for administration, reporting and research support, and we have incorporated it into our systems and processes at Lulworth. Where it is far less effective is in understanding personal circumstances, behavioural responses to risk or how client objectives change over time.

Many firms describe their service as personalised, often with the support of technology. In practice, this usually means segmenting clients into a fixed set of traditional risk profiles. Clients are then matched to models derived from historic asset price relationships, with adjustments only possible at the margins. The same approach often extends to client communications, which can amount to little more than standardised market updates, distributed widely with limited consideration for the individual receiving them.

Genuine personalisation, in our view, looks quite different. It starts with the ability to speak directly to the person responsible for building your portfolio, not a centralised or rotating team of relationship managers. It means having access to that person when it matters most, whether markets are volatile, circumstances have changed or a decision requires careful thought rather than a standard response. Communication about markets and portfolios should also be clear, understandable and genuinely relevant to the person reading it.

These qualities are sometimes described as old fashioned. We would argue the opposite. In a world of increasing automation and commoditised advice, they have become rarer and, as a result, more valuable.

We do not believe that trust or understanding can be automated. People respond differently to uncertainty, losses and change, and those differences are often revealed only through ongoing dialogue. They matter most during periods of market stress, when context, explanation and reassurance play a central role in maintaining confidence.

We use AI to improve efficiency behind the scenes and to support research and information processing. We do not use it to generate portfolios, automate significant decisions or replace direct conversations with clients. Investment decisions are made by people, and each decision can be clearly explained, challenged if necessary and properly owned.

AI as an Investment Theme

Artificial Intelligence is also a prominent investment theme across both public and private markets. Expectations are high. Some will be justified, others will not. Technological progress rarely follows a smooth or predictable path, and market leadership often changes as new opportunities and constraints emerge.

The technology boom of the late 1990s provides a useful reminder. The internet went on to reshape the global economy, but many of the companies that attracted the most attention at the height of the cycle failed to survive. A small number delivered exceptional long term returns, but identifying them in advance was difficult. Valuations at the time left little room for disappointment, and outcomes varied widely.

Today’s AI narrative shares some of these features. Recent market returns have been driven by a relatively small group of large technology companies. Their scale, access to capital and control of data give them clear advantages, but current valuations reflect strong assumptions about future growth and profitability. How this plays out will depend on a range of uncertain factors, including regulation, competition, capital requirements and how the economic benefits of AI are ultimately distributed.

Some investors have chosen to express their views on AI through concentrated positions in a small number of companies or through narrowly defined AI focused strategies. In other cases, concentration arises less deliberately. Many investors hold significant positions in global equity indices that are widely assumed to be well diversified, but in practice are heavily weighted towards a small group of US technology companies that dominate AI related returns. As a result, investors with traditional equity portfolios or ETFs may be making large bets on a single theme, either knowingly or without fully appreciating the extent of that exposure. What we seek to avoid is overreliance on that outcome.

We are in no doubt about the transformative potential of AI. At the same time, the scale of AI related capital expenditure being deployed today raises questions about shareholder value creation. It remains unclear how this investment will ultimately be monetised, which participants across the value chain will benefit most, and whose business models may be most at risk of disruption. Established software companies may adapt and extend their dominance, or they may find themselves disintermediated. Semiconductor manufacturers may offer compelling opportunities, but current valuations already reflect high expectations. Returns may also accrue elsewhere, including across the enabling infrastructure that underpins the ecosystem, from raw materials and energy to engineering firms and data centre operators.

In our view, navigating this uncertainty calls for a more thoughtful and diversified approach to AI exposure than is offered by passive or benchmark driven allocations to the largest US technology companies.

This exemplifies our broader philosophy on diversification. Rather than simply spreading exposure across geographies, we diversify deliberately across factors, sectors, themes and company sizes. Our aim is to ensure portfolio performance is not overly dependent on the fortunes of a single technology, sector or group of companies, however compelling their near term narrative may appear. If the pace of AI adoption proves slower, more uneven or more disruptive to incumbents than markets currently anticipate, a well constructed mix of return drivers can provide a meaningful buffer against disappointment.

Ultimately, our objective is to build portfolios that are robust across a wide range of outcomes, rather than ones implicitly reliant on any single theme unfolding exactly as expected.

A Measured Use of AI

AI will continue to develop and play a larger role in financial services. We do not adopt new technology simply to keep pace with peers. We use tools where they clearly improve efficiency or insight, and we set clear boundaries where their use would dilute judgement or accountability.

Our use of AI supports research, internal processes and information handling. It does not make investment decisions or replace relationships.

Clients come to us for careful thinking, clear explanations and decisions that reflect their individual circumstances. Technology can support that process, but it cannot replace it.

Judgement, responsibility and trust remain central to our approach. Those principles are unlikely to change, regardless of how technology evolves.

Please get in touch if you have any questions or would like to learn more, or subscribe to our newsletter to receive each new article as it is published.