Taking Stock 27 November 2025

James Lee writes:

‘‘If you are not sure, sell half.’’

I used to work with someone I respected immensely who had two rules. One was that if you didn’t have conviction, or you were no longer sure, you should take some action, as the simple process of taking action would provide you with new information.

Do you have something better to buy? Are you worried that markets will not keep going? Even if later you bought the same asset back, you would have avoided the worst response, which is being paralysed when things get hard.

So at the end of every year I find it useful to stop and take action, weigh up the year that has been, and consider the year ahead.  The simple process of doing so helps me frame my decisions for the year to come.

In my experience, at this time of year Institutions avoid taking new risks. As they manage the year-end process, some reduce risk, and most go away over Christmas and come back with a new perspective on the year, when the noise has died down.

Therefore, I try to do my thinking now, in case the environment changes.

Looking at 2025 we were focused on a few key themes that would be front and centre of most investors’ minds.

The AI Capital Expenditure (Capex) cycle - would it continue or would the eventual need for value creation from investors give it pause?

The interest rate cycle - how quickly would it turn?

Bitcoin - would the adoption curve continue to support it becoming a real asset class?

Economic cycle - with unemployment increasing, would the consumer stop spending?

Geopolitics - war, tariffs, Trump. How was that going to play out?

After pondering I found many of these things were not easy to answer because, being frank, some things are unforecastable. When that doubt occurs, the question becomes how much risk is the market already pricing in and then, risk adjusted what is the right decision – buy, sell or hold.

Given markets were near all-time highs, and these questions felt difficult to forecast, a 5-10% return was probably going to be a good result.  You should take the right amount of risk to achieve that.

As at today what has happened?

Gold is up 50%

Bitcoin is down 13%

The S&P500 (the US market) is up 10%, (largely driven by Google up 50% and Nvidia up 35%)

The NZX50 Gross is up 2%

ANZ and WBC, the boring old banks, are up 25%

On top of that, volatility is very high. At its worst the markets are down 20% and at their best up 16%.

Therefore, for a 5-10% type year, I would argue that global equity investing wasn’t worth the risk in 2025, and if it wasn’t for Nvidia and Google it would have been a pretty miserable year for most fund managers and investors who took the risk.

What drove that?

The AI capex cycle entered a phase where people started talking about bubbles. Well over $500 billion was invested last year, and trillions more promised in the coming years.  All were funded through capital allocation, not expenses. Capital expenditure from the seven largest technology companies alone exceeded $300bn, and that is before we look at how much x-AI, Open AI and the like raised to develop their models.

What occurred towards the end of the cycle was the circular nature of commitments, with Nvidia investing in partnerships that then purchased Nvidia chips. This raises a fair question about whether there is genuine economic return, and how long the hardware will remain useful. There is nothing inherently wrong with the model, although it does show that only a narrow pool of investors is willing to fund this level of capital expenditure.

I think this last gasp of capital investment is spooking the market, as the main players in this space are the same five names - Open AI, Nvidia, Oracle, MGX and Soft Bank, and they have finite funds to invest.

Interest rates fell pretty much like everyone thought they would. On balance the US fell less and NZ fell more, but everyone got the direction right, so what that meant was interest rate sensitive stocks as a collective didn’t really move a lot, because the market expected that outcome.

Bitcoin - for 12 years bitcoin has been touted as digital currency.  In 2025 the world gave up on that: A - because currency is already digital, B - because it is too volatile and C - the transaction fees are way too high.  In 2025 bitcoin tried to be digital gold. It failed that test as well.

Governments collectively hold around US$4 trillion in gold. By contrast, only a handful of countries including the United States, China and Germany have any meaningful bitcoin holdings, and even these remain modest at roughly US$45 billion in total. Bitcoin is a long way from being a digital equivalent of gold. At best it still operates as a gauge of risk appetite, and at present it shows little correlation to anything beyond online enthusiasm.

The economic cycle - Unemployment has increased globally, with software, manufacturing, retail, and construction jobs down.  NZ entered into a recession and some economists would argue that the US didn’t grow if you excluded data centre investment.  This overall wasn’t a surprise.

Geopolitics - Enough said. 2025 was like watching my four-year-old play politics with friends at kindy. For those who want some fun, watch Trump Without Trump on social media.

Summing up 2025

2025 was always going to be tough, and a 5-10% return was going to be a good year. On balance it turned out to be a harder year than we thought, with higher volatility. But all of that was masked by the AI Investment cycle holding up both the markets and the US economy. That investment cycle was led by a small group of interrelated companies chasing the dream of genuine artificial intelligence.

As we head into 2026 it’s hard to see how it will be any easier. There are four key questions that we should be asking on a regular basis next year:

How will a cornered animal behave?

We are watching a set of highly volatile leaders engage in high-stakes poker, where logic and emotion often part ways. Whether it is ending wars, threatening new ones, escalating trade disputes, deploying the national guard, or edging towards another government shutdown, it is difficult to predict the next move from the world’s largest economy.

The real question is how these leaders will react when the weaknesses in their economies become harder for their own populations to ignore next year. Will the response be sensible or insane?

Gold investors will follow this.

Will Open AI have any problems raising capital?

Today a small group of investors is pouring billions into making AI a reality, either through data centres, model training or investment in products.  Last year $500 billion-plus was invested. Next year is going to be greater. If at any stage one of the AI companies struggles to raise capital to fund the pursuit of artificial intelligence, then I would expect them all to slow down.

Today Meta, Google, Open AI, xAi, Amazon and Microsoft are all pouring billions into AI.

If Open AI solves artificial intelligence, and regulation doesn’t lead to slower adoption, then maybe the market keeps on going, but as at today what evidence do we have that is happening?

Is it genius or insanity?

For five years, Michael Saylor and his US company MicroStrategy preached to everyone who would listen that bitcoin was an asset, and they needed to buy it. His model was to raise capital to buy bitcoin. In 2025 his stock price has fallen 40% so he now raises debt or offers dividend-paying shares to buy bitcoin.  His theory is that bitcoin will always go up 30% so he can borrow at 10% to buy something that goes up 30%.  Today his leverage ratio is 10-15% but what happens if bitcoin doesn’t go up? MicroStrategy and copycat companies have accumulated 1 million bitcoin using this strategy, but with bitcoin down 13% and Saylor’s obligations now heading towards $1b a year, his is a very high-risk strategy.  I would expect next year the bears to become very vocal on MicroStrategy and therefore bitcoin.

Bitcoin has failed to be a digital currency, and this year has failed to be digital gold. Bitcoin hasn’t become a good proxy on risk. The Nasdaq was up 10% vs Bitcoin down 13%. It remains one the greatest trading assets available because it is driven purely on sentiment. It has zero fundamental drivers. It is just a proxy on global emotion.

As the pension funds and endowment funds finally start to buy, the question must be asked: Has the dumb money finally turned up? Is that why the original investors in bitcoin seem to be selling out?

A crack in bitcoin would impact retail sentiment negatively for stocks, so you can’ t afford to ignore Michael Saylor in 2026. I watch for Saylor quotes, looking for signs of stress, albeit some of his comments make him sound insane, so it’s hard work.

Will NZ recover first because we went into recession first?

Lower rates, record dairy prices, Fonterra payouts, increased investment from the fast-track processes, foreign investors leading to a rebound in housing, even a half decent summer, may produce, finally, a number of potentially positive drivers for the NZ economy.

The last reporting season saw more companies surprise on the positive than miss on the negative, suggesting we have passed the bottom in terms of sentiment.

On the other side of the ledger, increased unemployment in the US, tariffs leading to inflation, and spiralling debt and healthcare costs in the US, beg the question: Will they impact the US consumer? Remember, the US consumer is responsible for more than 25% (some say 30%) of ALL consumption globally for a population equating to circa 4% of the world population.

If US consumers crack, that would have global ramifications. That said, interest rates still have room to go lower in the US which would likely protect this very important part of the world economy.

If you compared the two situations, the NZ economy is likely to begin its recovery in 2026, as the market was flat for 2025 and expectations aren’t high.

Compare this to the US economy, which has more potential to show cracks, the market is already 10% higher than a year ago and currently the market remains very optimistic.

My risk tolerance would make that a simple decision.

Summing up what 2026 could look like

The risks that geopolitics, the AI super cycle, and the US economy all have nothing go wrong, given they are all priced for perfection, makes the risk-free assessment hard to swallow. On balance the notion NZ will be in a better space is easier to believe.

To me it seems likely that 2026 will have more volatility than 2025 and that 5-10% would be another pretty good year, so building a low-risk way to earn that will be my primary focus for me and our company. I think geopolitics will still create volatility in gold but its price has already factored in increased geopolitical risk.

While there are many people calling the US market a bubble, on the other side of the debate some commentators highlight that some of these companies are growing at a rate that suggests the market could carry on. I don’t think that is easy to assess. I do however think the risk return isn’t weighed in investors’ favour, but equally if there are four things that we are watching on the negative, there are two things I am watching on the positive that could change this.

Energy and portability of energy

The average household, once it combines heating, transport and electricity bills, will find that somewhere between 5% and 10% of their income is spent on energy.  The cost of gas and huge energy requirements for data centres has been increasing energy prices since 2020.  So, when we look at areas that could provide relief long term (not short term) the opportunities relate to the capture and storage of energy.

Battery technology - currently there are lots of advances in battery technology. If the world can solve high density safe batteries that extend current range by 2-3 times, then it would be a new world - flying cars, drones at scale, who knows what else? - but the market would rally.

Fusion - Fusion energy is the idea that you can cause very light atoms to join (fuse) through extreme heat, and this creates more energy than is used in creating the heat.  You contain this energy within a giant magnetic field so it doesn’t expire and continuously feed it new atoms to make it a sustainable energy source like the sun.

There are many people trying to solve the fusion equation. Two companies have advanced projects that in 2026 we should watch. While neither will produce energy this year, they have important milestones to gauge as this technology would literally change the world. It would not be the free energy of science fiction, as the capital cost would be very high to begin with, but it would dramatically lower energy costs, allowing households of the future to spend elsewhere.

Bringing this all together

2025 was a tough year to get right, 2026 isn’t looking any easier. It is not doom and gloom by any stretch of the imagination, rather another year where 5-10% returns on a risk adjusted basis would be acceptable, given there are so many things around the world that can go wrong.

It isn’t sensible advice to parrot the notion that markets always go up, or just hold on in the expectation that investing is a long-term journey, sometimes requiring you to accept the slow road. Today Warren Buffett has $380bn in cash-like assets - that is 31% of his portfolio.  Therefore, you are in good company if you accept that it is pretty hard to see low-risk opportunities in the market.

_ _ _ _ _ _ _ _ _ _ _ _

Kiwibank Tier 2 Note Offer

Kiwibank has announced that it is considering an offer of Tier 2 Notes, which will carry an investment grade credit rating.

The notes have a final maturity date of 12 March 2036 but are likely to be repaid at the first reset date on 12 March 2031. Similar notes from major banks, including Kiwibank, are typically repaid on the reset date.

Based on current conditions, we expect an interest rate above 4.50 percent. Investors are unlikely to be charged brokerage, as Kiwibank is expected to cover these costs.

Further details will be released later this month.

If you would like to be added to the list for this offer, pending further details, please email us with your CSN and an indicative investment amount, and we will contact you once the details are confirmed.

We are expecting this issue to open on 1 December, with payment due around 10 December.

_ _ _ _ _ _ _ _ _ _ _ _

Travel

5 December – Christchurch – Chris Lee

8 December – Christchurch – Chris Lee

James Lee

Chairman

Chris Lee & Partners Ltd