Trusts Tested, Shareholders Push Back, and AI Tackles FX

Retail Investors Should Not Lose Sight of Trusts

Long-term asset funds, or LTAFs, in the UK have received a boost with Hargreaves Lansdown agreeing to add two Schroders funds to its platform – a move that should boost retail investment in private markets.

LTAFs are described as an opportunity for individual investors to diversify their portfolios, access investments beyond the stock market with less correlated returns, and support the wider economy.

Discover how neo-banks become wealthtech in London at the FMLS25!

Inevitably, other players will enter this space. For example, Scottish Widows received FCA approval for a fund in August and is expected to bring it to market before the end of the year.

Widening access to LTAFs has not met with universal approval, with some observers warning of the danger that retail investors will be pushed towards private market holdings that are complex and opaque.

There is also the potential for confusion among investors unclear about what exactly constitutes a private asset. Then there are issues around liquidity, with these funds typically coming with 90-day minimum notice periods and limits on redemption volumes.

Read more: FCA to Allow Retail Investors to Gain Access to Crypto ETNs Starting October

One fund manager referred to giving retail investors access to LTAFs as the equivalent of taking a sledgehammer to crack a walnut. He noted that retail investors have been able to access private markets via investment trusts – an established trading structure that has proved its resilience during testing market conditions – for some time.

Another expressed surprise at the underplaying of investment trusts and reiterated the adage that liquidity comes at a price.

Private assets are ideally suited for closed-end structures and, generally, trusts have been inexpensive and liquid. Currently, some trusts are trading at significant discounts, leading to a situation where they are assigning no value to the private exposure they provide.

Although it is fair to say that there will be additional volatility due to the premium or discount fluctuations, there is also the potential for increased returns if discounts decrease, and the industry is actively working towards this goal.

You may also like: Around 20% FCA-Regulated CFD Brokers Are Inactive

Not all investment trusts have produced the desired results, particularly some of those focused on infrastructure. But the private equity investment trust sector has been a high performer over the last decade, which should at least encourage retail investors eyeing LTAFs to pause for thought.

Shareholder Dissent Sharpens Focus on Capital Decisions

A new report indicates shareholder opposition to executive pay rose sharply during the 2025 proxy season compared to last year, while opposition to share issuance resolutions also rose.

Georgeson’s European AGM season review – based on results from annual general meetings across nine key European markets – found that the proportion of contested remuneration reports increased from 29.9% in 2024 to 31.1%.

Resolutions are defined as ‘contested’ if 10% or more of shareholder votes are cast against them.

The proportion of resolutions on share issuance that were contested also increased from 13.4% to 18.9%. This is important because companies in most European markets are required to gain shareholder approval via a share issuance resolution before they can issue new shares.

Read more: IG CEO Made $4.5 Million in FY25, but Still Earns Less Than Top Plus500 Executives

Opposition to such resolutions increased in all markets except the UK (which experienced a slight decline) and Switzerland, which had no contested share issuance votes.

The rise in opposition to share issuance has been interpreted in some quarters as a push from the market for stronger oversight of capital decisions. Shareholders may also be increasingly concerned about shareholder dilution and how companies manage capital allocation.

One of the reasons why Warren Buffett overcame his aversion to tech stocks to invest in Apple was the company’s aggressive share buybacks, which have increased his ownership percentage over the last decade. The head of Berkshire Hathaway has a penchant for investing in companies that return capital to shareholders.

Some of the comments made in Georgeson’s global institutional investor survey published earlier this year underlined the importance investors such as Buffett place on engagement beyond just meeting numbers or sharing documents.

Discussions that align with investment strategies (with a focus on financial materiality and long-term value creation) are valued, and when companies are slow to respond to engagement, shareholders are increasingly willing to take collaborative action.

Language Matters When It Comes to FX

Swiss National Bank research suggests large language models (LLMs) outperform traditional AI when it comes to predicting FX market moves – which will leave many traders asking how they can use this to improve their forecasting.

The research authors suggest that fine-tuned LLMs capture the nuances of news articles more effectively – particularly when trained on a combination of human-labelled and distant-labelled datasets – and that domain-adapted LLMs offer an advantage over pre-trained financial models that have not been fine-tuned for the FX market’s unique linguistic and structural challenges.

You may also like: Forget Prompts—It’s Context Engineering That Matters

They concluded that domain-specific adaptation of LLMs improves sentiment classification accuracy and enhances the predictive utility of sentiment signals in trading applications.

Large language model co-pilots, which allow traders to use natural language to query data, were in vogue last year, although there was a sense that banks would want to explore the technology and look for potential issues before deploying it.

But increasing competition from non-bank market makers has accelerated banks’ machine learning and AI strategies.

While the SNB model proved effective, it has some limitations. In an international market where large volumes are traded in places where English is not the first language, it does not take account of news published in a number of major languages, which affects the movement of currency pairs where the dollar is not on one side of the trade.

There is also some concern that the model reflects analyst consensus instead of predictive sentiment, as the former is frequently already priced in.

To their credit, the research authors acknowledge that the scarcity of labelled FX sentiment datasets remains a constraint and that additional research is needed to assess the model’s applicability in more complex trading frameworks, such as high-frequency trading.