Markets Fewer Investors are Watching

– David Wanis, September 2025

As small cap investors, we have been boring our clients for the last couple of years with the view that no one cares about small caps, but maybe they should. This seems to have recently changed as outperformance brings renewed interest.

We have also talked about stocks no one thinks about but maybe should (our quiet compounders), who have continued to do well – particularly as an alternative to private equity which seems to be having a few issues at the moment.

This month there are a couple more topics that come to mind which are unfashionable but possibly useful in thinking about potential future sources of positive or negative returns. Positive New Zealand and negative Data Centres.

While everywhere else in the world seems to be celebrating record equity market returns, the New Zealand market has lagged almost everything. In the five years to September 2025, in NZD terms, the NZ50 Index delivered total returns of 2.5% per annum, compared to 10.4% per annum for emerging markets. 14% for the ASX300 and 18% p.a. for the MSCI World (all NZD).

With interest rate cuts to the Overnight Cash Rate from 5.5% in July 2024 to 3% at the end of September 2025 (and likely going lower still) and reforms to government policies, meeting with companies during September 2025 didn’t feel like growth had returned, but it certainly is set up for much better performance over the coming years. Some companies we met with in New Zealand during the month who have a more real-time view of the economy are seeing the first signs in more than two years of sustained – albeit modest – improvement. It seems the experiment of abandoning GDP growth policies in favour of wellbeing may be consigned to the dustbin of history, yet another example of policy enacted by people who don’t live in the real world. As one local put it to us on a recent trip: “the former government had been busy turning off all the taps of economic growth. The new government has been turning them back on, it just takes time for the water to flow”.

We have a few investments in New Zealand – including Mainfreight and Freightways, two of which are more leveraged to a recovery in activity and a less business-hostile government. There are many other companies in New Zealand which we also believe have been overlooked.

In an economy which has averaged real GDP growth of sub 2% p.a. since 2019 and nominal GDP growth of just under 6%, the performance of Mainfreight and Freightways gives a useful indication of what market leaders can do to grow shareholder value in tough times.

Freightways operate in the express package and business mail segment with services including network and point-to-point couriers. They compete with NZ Post for some business lines, a competitor who has had the luxury of government funding to compensate for any difficult trading conditions and financial losses. This competitive dynamic has been in place for many years and does restrain the potential economic returns Freightways can generate for their shareholders. Regardless of a state sponsored competitor, Freightways has delivered strong economic returns from their NZ business and despite the sluggish conditions which directly affect their volumes. Excluding the acquisition of Big Chill they have managed to outgrow nominal GDP (we estimate high single digit revenue growth) and expand margins since 2019.

Mainfreight is New Zealand’s largest logistics and transport company, offering a wide array of services, including less-than-container-load (LCL), full-truck-load (FTL), warehousing and specialised transport services. Despite the global footprint, the stronger market positions and greater network density in New Zealand and Australia mean that New Zealand delivers 31% of EBITDA (F2025) and Australia & New Zealand combined are 65%. Since 2019, New Zealand revenue has grown at 8.3% per annum and EBITDA margins have expanded from 15.4% in F2019 to 20.2% in F2025. For context, the Australian business enjoyed almost twice the revenue growth rate at 13.8% per annum over the same period, with 300bps more margin expansion.

Great business can also be great investments in tough times, and despite the paltry returns available in the New Zealand market, the strong market position – and performance outside the country – has enabled Freightways to deliver +17.7% p.a. total shareholder returns and Mainfreight +9.5% p.a. returns over the past five years.

There is little doubt the utility of LLMs for users is very real and of great value if used properly. It may not be general intelligence, but it is a form of intelligence which is something new and useful. There is also little doubt we are in the grips of a financial market bubble – but knowing when this ends is not our primary focus.

NextDC, Macquarie Technology, DigiCo and Top 20 leader Goodman Group have all tied their fortunes in varying degrees to the AI infrastructure buildout. There is barely any evidence anywhere globally of these data centre assets earning a properly accounted for return on invested capital at acceptable levels. There are a few voices in the market flagging concerns over asset lives, depreciation rates and through the cycle return on capital, but mostly the headlines are on capacity growth, power usage and raw hype. This is a distraction which allows companies with low to mid-single digit return on capital business models to raise equity at 2-3x book value. There are sound economic reasons why Data Centres are the worst performing REITs in the US this year, down 9.6% YTD to September against a REIT sector up 4.5% and the broader U.S equity market up ~15%.

The four Australian data centre operators are also behind the local benchmark (+12.4% YTD for the ASX300 to 30 Sep), either marginally (NXT +12.3%), or more substantially (GMG -7.6%, MAQ -27.9%, DGT -38.6%).

This is the G-rated theatrical trailer for what happens when the AI hype wears off, and reality returns to the market. There is a yawning gap between what AI stocks currently reflect in share prices and reality. We have seen this before, yes through the dot com bubble, but more recently in 2021. This feels similar to both.