Dear investors and friends of the fund,

The third quarter felt like a turning point for the portfolio after a tough first half of 2026. The conflict in Iran and fears that artificial intelligence would wipe out software businesses led to indiscriminate selling across the sector. Good companies were sold alongside weak ones. In the third quarter, many of those good companies began to recover as their results came through.

I am more enthusiastic about the opportunity in global smaller companies today than at any point since we launched. That isn't because I can predict the next quarter, but because of what we own. Across the portfolio, we target organic sales growth of around 14% a year, M&A growth on top, margin expansion and return on capital (CFROI) of c. 25%. All this with the portfolio trading on just 15x PE. Owning that combination of quality, growth and value at that price is rare. This low multiple is an important metric for our multi bagger framework, but importantly it also materially reduces a key risk that we feel the market is underappreciating; multiple risk (only limited debasement of valuations post the end of ZIRP (zero interest rates), despite interest rates returning to more normalized levels).

This letter covers why we think smaller companies are well-placed, what we bought and sold in the quarter, the themes we're following and how we manage risk.

Richard Tennant, Partner and Fund Manager, Goodhart Global Smaller Companies Fund.

Why smaller companies, and why now

Global smaller companies have trailed large caps for most of the past decade, and today they trade at a clear discount. We think that gap is an opportunity for patient investors.

The long period of underperformance has little to do with how smaller businesses have actually performed. Most of it reflects money flowing into a handful of very large technology companies. The ten largest stocks now make up almost 28% of the MSCI World index. In the S&P 500, the top ten account for about 40%. Owning a global index today means owning a concentrated bet on one theme.

Smaller companies are priced very differently.

In the US the gap is wider - at the end of 2025 the Russell 2500 traded on 18.5 times forward earnings against 23.0 times for the S&P 500. Before the pandemic, smaller companies usually traded at a small premium to large ones.

History suggests these periods don't last forever. Leadership between small and large companies tends to rotate over multiyear cycles, and the current run is long by historical standards. Whilst we aren't trying to time the turn, it feels like it’s close. Just the concentration of the top 10 might suggest this trend is peaking (see below). Either way, we can buy good businesses at sensible prices and a change in sentiment would offer a meaningful tailwind on top.

What we did in the quarter

• We added seven new holdings – 3 serial industrial acquirers with MSD organic growth, DD contribution to growth from M&A, margin expansion, attractive returns and acquiring at 3-6x EBIT multiples. 2 undervalued defence related companies with 20%+ growth, a unique product and valuations that significantly undervalue the quality and growth profile. 1 UK software business with attractive growth, FCF and which is really a US business (just at a much lower multiple because of its listing geography). Finally, a very exciting Japanese software company in a specific niche, growing 20% pa, with M&A consolidation potential and really leaning into their flywheel.

• Added to several of the businesses the market had sold indiscriminately around AI and which I believe will likely be AI beneficiaries, crucially around the application layer rather than the picks and shovels semi space. This is where all the money was made from the dotcom boom (Netflix, Spotify, Amazon, Meta etc) and I expect the same here.

• Notable sells - Angi, Intermap and Lumine Group. After reassessing each moat against our AI framework (described in the next section). We sold GQG Partners because the investment thesis changed and my waypoints weren’t hit. We also sold Pinewood Technologies, which received a takeover bid.

Takeovers

Two holdings received takeover approaches in the quarter, Pinewood Technologies and Ashtead Technology. Both are listed in the UK but sell almost entirely abroad. Around 50 UK-listed companies have received formal bids in the last three years and on average 77% of their sales came from outside the UK. About 9% of our portfolio is listed in the UK, but those companies earn less than 2% of the fund's sales in the UK. Patient private equity and trade buyers value the qualities we look for, and we expect more takeovers across the portfolio over time.

Themes we are following

AI: owning the application layer:

AI has driven much of the dispersion in markets this year. Software and internet shares were sold heavily on fears that large language models would replace the applications built on top of them. We think the market has drawn the line in the wrong place. The infrastructure names are priced for strong growth and no cycle. Many of the specialist software businesses that run complex, regulated workflows have been priced as if AI will make them obsolete.

The businesses we want to own share five characteristics, and AI makes each of them more valuable.

1. Mission critical. If the product fails, the customer cannot bill, get paid or stay compliant.

2. Deep domain knowledge. Decades of rules and edge cases that aren't on the open internet, so a general model cannot learn them.

3. Proprietary data. Data built up over years of running the product which is not available to large language models.

4. High return for the customer. The product is a small part of the customer's costs compared with the value it protects, so there is little reason to switch.

5. AI as an accelerant. AI can be layered onto the data and workflow to raise productivity and pricing, rather than giving customers a cheaper way around the incumbent.

   
   

We held or added to companies that meet all five tests. Where a business relied on discovery, matching or generic information, we sold. The fund has no direct exposure to AI infrastructure, which leaves us positioned on the opposite side of where the market's attention is concentrated. We may be early in the short term, but our longer-term conviction remains. In our view the scale of the current build-out has pushed earnings above trend, while valuations appear to imply that these companies have become structurally less cyclical, with current levels of returns sustained despite the potential for increased competition.

This is accentuated by the increasingly circular nature of AI revenue. For example: a chipmaker invests in an AI lab. The lab uses that capital, or commitments backed by it, to sign multi-year compute contracts with a cloud provider. The cloud provider borrows to build data centres and fills them with the chipmaker's GPUs. Each step is booked as revenue or backlog by someone, and each headline deal lifts the share prices that make the next round of funding possible. In real life this looks like Nvidia investing in OpenAI, OpenAI committing to Oracle and CoreWeave, and AMD handing OpenAI warrants alongside a chip order are all versions of the same loop[1]. Money is moving round the circle, but it isn't clear how much new money is entering it from end customers paying for AI products.

We saw this vendor financing in the late 1990s e.g. Lucent and Nortel lent customers the money to buy their own equipment. Capex is being spent and depreciated over time to buy sales / profits today as a one-year boost, which flatters near term earnings, accentuated by the hyperscalers lengthening the assumed useful life of their servers and GPUs to five or six years, even though Nvidia now refreshes its chip architecture roughly every year[2]. Stretching depreciation over a longer life cuts the annual expense, so profits look higher than the economics justify. This is not to say that AI isn’t going to be huge. I believe it is, but where the profit and returns accrue is crucial in positioning the fund to benefit from this theme. Today I believe we have a real benefit, with the real application winners trading at very low multiples, and the cyclical picks & shovels that have almost never built a structural USP (and whose returns will undoubtedly be competed down over time), trading at very high multiples. Whilst this is anti-consensus, I feel the risk-reward is likely skewed in the strategy’s favour.

When investors focus on a single theme, everything else gets neglected. Smaller companies feel this most because they have less analyst coverage and fewer long-term owners. The last time technology dominated the market this heavily was the dotcom era. When that bubble burst, the parts of the market that had been ignored did very well.

European defence

We have been adding to defence while sentiment towards the sector has been strangely weak given the global backdrop. The 3 new investments (from 0) give us exposure to the long-term European rearmament trend through specialist suppliers, with a strong moat and with multi-year order books.

Reshoring and deglobalisation

More generally, many of our industrial holdings will likely benefit from manufacturing returning closer to home, as well as from infrastructure investment. Industrials were the largest positive contributor to the fund's returns in the period.

How we manage risk

The portfolio aims to combine quality, growth and value while staying diversified and idiosyncratic with limited correlation. This is exemplified by the fund’s beta, which remains consistent at around 0.8, and the incredibly low correlation of the fund to the index (portfolio construction and risk is a key focus for the strategy given the upside potential). The table below highlights the correlation of the various holdings against each other (lots of green / yellow means little correlation and red would suggest high correlation). This is unusual for a global small cap fund, but I think really important, especially given we believe we will be entering a world with more cycles and volatility.

Linked to this risk management, we are preparing for a world with more cycles. At Goodhart we believe the investment regime has changed. For most of the past fifteen years, near zero interest rates and plentiful liquidity smoothed out the economic cycle and rewarded owning the index. We expect the next decade to look very different. Ageing populations, rising geopolitical & security pressures, resource constraints and rapid technological change all point to shorter and more frequent economic cycles, with more volatility along the way. Broad equity indices, which are now heavily concentrated in a few large companies, may find that environment harder.

We see this as an opportunity rather than a threat. Volatility creates mispricing, and smaller companies, with less coverage and fewer long-term owners, tend to feel it most. We have built the portfolio with this in mind. Most of our holdings depend on their own decisions more than on the economy and the portfolio's beta is around 0.75 to 0.85[3]. Valuations are low enough to give us some protection if multiples fall. Our aim is to own businesses that can keep compounding through the cycle and to use periods of stress to add to them at better prices.

Process updates

Waypoints and Sentinel

As communicated previously, we underwrite every holding against five or six measurable milestones, which we call waypoints. They cover things like organic growth, margin potential, M&A contribution, new products and cash conversion. To track them objectively, we built Sentinel, our own AI platform. It reads earnings calls, results and filings. It also picks up mentions of our companies by their competitors, customers, suppliers around the world and in trade press. It then scores whether each investment case is on track, improving or deteriorating in relation these waypoints.

Sell discipline

Investors tend to talk extensively about their process for buying a company, but much less about when to sell. In our experience, selling a long-term winner too early can be one of the more costly investment mistakes. I saw this in my own record at Fidelity. Looking back, there were occasions when I sold long-term winners too early and missed a meaningful part of the subsequent upside. Exits from the fund now follow three set rules tied to the waypoints as below.

Looking ahead

We think the best opportunities in global smaller companies don't come from hidden or complicated situations. They come from finding those special businesses with huge scalability, and, crucially, being willing to be patient. This is one of the few areas of investing where the value of doing this has risen over time as the market has become increasingly impatient.

In our experience, impatience creates a recurring pattern. For example: a good business has a clear short-term problem, such as a delayed product, a lumpy quarter or a period of heavy investment. Its shares fall much further than the long-term value has changed. Earnings downgrades and a lower valuation multiple compound each other and a long-term owner can buy a better future at a lower price.

Goodhart is a partnership, and we invest alongside our clients. We don't need each stock to “work” next quarter to justify owning it, and we aren't forced to sell a good business after a bad six months. We believe that patience is becoming a real competitive advantage.

I'm grateful to our investors for their support and I'm excited about what the portfolio can achieve over the coming years. If you would like to discuss the fund or anything else, please get in touch.

[1] Source: Reuters, Coreweave, Investing.com and Tom's Hardware

[2] Source: Bloomberg, Alphabet and Amazon 10k filings, CNBC

[3] Source: Goodhart Partners, as at 30 September 2026