
What was not important over the last week?
· Putin’s photo op with Xi and Trump. Nothing of consequence will emerge.
· The resignation of Nvidia’s Jensen Huang. The biggest invention since, at least, the Guttenberg Press will not stall due to one man getting caught with his pants down.
· Labour’s daily announcements on… nothing in particular.
One snippet worth noting in France. Marine Le Pen is playing a dangerous game with financial markets, announcing she will cut the French retirement age if she wins the Presidential election in 2027. That’s insane, and underscores the existential risks built into the EU structure.
The big investment news continues to revolve around AI. This week it was the Anthropic IPO, along with Nvidia shenanigans. I want to look at their respective roles in this extraordinary cycle of innovation, and how their destinies might be very different in surprising ways.
The AI revolution continues to produce extraordinary numbers. Nvidia's sales reached $115.8 billion in just three months, more than double a year earlier. Meanwhile, Anthropic, developer of Claude AI, is now selling its services at a rate equivalent to $81 billion a year. Yet there is a clear warning from history…
Technology revolutions, from canals to the internet, attract too much money, too much competition and eventually too much capacity. The irony is that the technology can succeed spectacularly even while investors lose fortunes. The internet changed the world, as promised. Yet Cisco, put on a pedestal by investors imagining outsize gains, as one of its great beneficiaries, saw its shares collapse by almost 90% in 2000–2003.
Nvidia and Anthropic face different dangers.
Nvidia supplies the powerful computer chips which are the brains of the machinery in the data centres owned by the likes of Amazon and Google (see 6th Nov on “what is a data centre?”). But what happens when these customers discover they have built too many data centres, or cannot make enough money from them? There are already lively and well-founded discussions on both risks. Orders can dry up very quickly. Nvidia could remain a great company even though its profits fell sharply and its share price collapsed as the future seemed more mundane.
Anthropic effectively rents computing power from cloud providers’ data centres (e.g. Amazon and Google), running its models on a mix of chips, including those provided by Nvidia. Anthropic then makes money by selling AI services ranging from everyday tasks, such as analysing contracts and writing reports, to far more complex work – helping doctors interpret medical information, scientists accelerating their research, and large companies develop software to analyse vast amounts of data and solve difficult business problems.
The problem is not that the demand for such services will diminish – on the contrary as more people adapt. Competition is the problem, forcing prices much lower. Anthropic could serve vastly more customers yet struggle to make the profits investors expect.
That's the great trap of innovation booms. The technology succeeds, the businesses continue to grow, but the returns to investors evaporate.
The lesson? Enjoy the growth, but don't fall in love with the companies delivering it. They can be great, until they aren't. Pay attention to changing trends, and have the discipline to act when those trends change.
It is reasonably clear that the investment cycle for tech is very mature, at the top end. In contrast, the shorter term investment cycle for many commodities is rapidly moving towards a bottom…
Last week I set out some opportunities for 2027, including a number of commodities. Commodities were the most tantalising investment development over the week, as gold and other metals fell sharply. I’ll look at the 2025-26 rollercoaster, and then the juicy opportunities which might lie just ahead.
One of the most remarkable investment stories of 2025 and 2026 was the heady rise and fall of commodities, from gold and gold miners to uranium and agriculture (with oil being the only obvious exception, being driven by Trump’s war).
The upswing began in early 2025. Growing concerns about inflation, government debt, tariffs and the US dollar encouraged investors to look towards real assets. Gold led the way, but enthusiasm gradually spread to other commodities, even though their underlying fundamentals were quite different.
As prices rose the momentum attracted more buyers (it always does). By November and December the headlines were increasingly excited. Retail investors were arriving, including self-directed investors asking how they could join the party, particularly gold. For experienced investors, that was a warning.
I have seen this pattern repeatedly over 40 years, and financial history stretching back 300 years is littered with the same. Early buyers recognise an opportunity. Later buyers see spectacular recent returns and assume they will continue when they join the party. The latter stage of the uptrend is fuelled by greed and investors fearing that they will miss out.
It was at this point in late 2025, recognising the signals, that we took profits in the Dennehy Wealth discretionary portfolios, rather than gamble on capturing the final few percentage points. Then came the sharp reversal from January and February 2026. Different commodities all began to fall together, not because they shared fundamental characteristics, but because the enthusiasm for buying these disparate commodities had evaporated.
The correction in 2026 has been substantial, as you might expect after a rip-roaring 2025 when gains often exceeded 100%. As that correction now moves into its latter stages, the next substantial opportunity appears on the horizon.
Take gold. It has fallen more than 30% since its peak early in 2026. My ideal support is at USD3300-3400, another 8% lower. That’s when it gets interesting, though I would want to see other indicators coinciding e.g. how traders were positioned; despair; negative headlines; gold miners outperforming the gold price. Don’t jump in at that level – look around at the broader picture first.
When these indicators point in the same direction, and prices seem to be holding at these levels over a week or two, it gets exciting because there is considerable upside in gold and beyond. I’ll explain why.
Gold at USD3300 would imply that a correction to the big upward move in 2025 is over – what I call a wave 4 correction. If this analysis proves correct, another substantial rise to new highs will commence, which means a gain of at least 70% in the gold price, probably more in gold miners. The potential could be even greater for two reasons:
Firstly, the 5th wave in commodities is often the longest and most emotionally charged.
Secondly, the opportunity stretches across the commodity complex, it isn’t just about gold.
For example, uranium miners have been smashed (e.g. Global X Uranium ETF), down more than 50% from the peak. This fall was particularly bad because one major company made a schoolboy error on prices at which it would supply uranium in the future. This is slowly being unwound, and this unwinding appears likely to coincide with a turning point in, and leadership from, gold. After a 50% fall a 100% gain is required to get to new highs, which I would expect if that “wave 4” point is valid.
It is a similar argument with copper miners, the Global X Copper Miners being the representative fund, which has endured a horrible week, down 15%.
Fingers crossed those recoveries to new highs are something to look forward to in 2027.
But for now, we want to see a bit of panic in these commodities - doom laden headlines – despairing investors.
(If you're not yet in the FundExpert Facebook group, click here to join and enter our annual FTSE competition which has just gone live)