4 megatrends you need to know about (and how to play them)

There are plenty of reasons for investors to feel uncomfortable right now. Artificial intelligence valuations have surged, geopolitical tensions are rising, governments are spending heavily on defence, energy systems are being rebuilt, and the race for critical minerals is intensifying.

Yet step back from the daily noise and these apparently separate trends have something important in common - they require an enormous amount of physical investment.

At Livewire Live, four investors approached that opportunity from very different directions. 

• EQT Partners’ Sam Franklin focused on the energy infrastructure needed to power the digital economy. 
• Munro Partners’ Nick Griffin argued the AI boom is only getting started. 
• Janus Henderson Investors’ Daniel Sullivan made the case for a new resources supercycle. 
• VanEck’s Jamie Hannah highlighted a transformation in global defence spending.

Taken together, the four presentations pointed to a much bigger investment story. The next decade could require extraordinary amounts of capital, infrastructure and raw materials. 

For investors, the opportunity may be less about predicting the next killer app and more about owning the businesses supplying the picks, shovels and infrastructure behind it.

The energy system needs to get much bigger

For Franklin, managing director in EQT’s infrastructure team, one of the most important changes underway is that electricity demand is growing again; in fact, it's surging. 

After roughly two decades of stagnant demand across developed markets, AI, data centres, electrification and rising cooling requirements are changing the equation.

In Australia alone, Franklin expects data centres to grow from around 3% to 13% of grid capacity, while East Coast electricity consumption could rise from roughly 170 terawatt hours annually to 400 terawatt hours by 2040.

At the same time, the way that electricity is produced is changing.

Solar is winning on cost and speed, batteries are becoming cheaper and more effective, while gas remains critical to providing firm generation. The bottleneck is increasingly the ability to build enough infrastructure quickly enough.

Franklin pointed to extraordinary order books for gas turbines, including 116 gigawatts at GE, 69 gigawatts at Siemens and 35 gigawatts at Mitsubishi.

“Gas generation turbines are sold out for the next decade. So the future output of gas is solely capped by technology.”

EQT estimates capital investment across the energy transition subsectors it targets could increase from around US$100 billion globally to US$425 billion, representing annual growth of roughly 20%.

Rather than simply owning anything exposed to rising electricity demand, however, Franklin believes the winners will be businesses capable of solving the bottlenecks.

“We're looking for businesses which can provide speed to power. So businesses that can deliver firm electricity to customers faster than the grid can provide it and can provide it directly to the customer and increase their energy security are going to have a structural advantage.”

One example is Scale Microgrids, an EQT-owned US business that develops on-site solar, battery and gas generation for data centres and heavy industrial customers. Long-term contracts can provide infrastructure-like revenues while helping customers secure electricity without waiting for grid connections.

“The sustainability trends will become security trends, they'll become energy sovereignty trends, they'll become something else, but the underlying demand thematics we see continuing over the very long term.”

AI is a boom, not a bubble

If energy demand is one consequence of AI, Griffin believes investors still underestimate the scale of the underlying cause.

His central argument was unequivocal: AI represents the next major technology platform shift and the infrastructure build-out supporting it is closer to the beginning than the end.

“This is a boom, not a bubble. This is a boom, this is a CapEx boom. It's an infrastructure boom.”

Griffin compared the change with the transition from newspapers and free-to-air television to the internet. The “top of the funnel” moved to Google, Facebook and other digital platforms. Now it is shifting again as consumers increasingly interact directly with AI platforms and agents.

Importantly, those agents consume dramatically more computing power. Griffin said a basic AI query might once have required the equivalent of one unit of compute, reasoning models increased that dramatically, and AI agents can now use thousands of tokens as they complete increasingly complex tasks.

That is driving an extraordinary infrastructure cycle.

Griffin said hyperscaler capital expenditure had risen from roughly US$30 billion annually several years ago to an expected US$800 billion this year, with spending potentially reaching US$1.3 trillion next year and US$1.6 trillion the year after.

For investors worried about whether that spending can generate a return, he pointed to rapidly rising hyperscaler revenues and demand that continues to exceed available computing capacity.

“You're three years into a CapEx build-out that's probably going to last a decade. Eventually it will end. It makes sense today. In five years it might not, but you're very much at the start of this.”

That leads Griffin towards the “shovels” supplying the boom - cloud infrastructure, semiconductors and power.

He is more cautious about trying to identify the ultimate application winners. Purely digital businesses with high margins could themselves be vulnerable to AI disruption, while businesses combining digital capabilities with difficult-to-replicate physical networks may be better protected.

“The applications are probably not where you think, and the biggest ones are actually in front of us.”

The resources supercycle could be back

Sullivan sees another beneficiary of this investment boom - natural resources - which are being reshaped by the collision of three forces. 

• Deglobalisation is pushing countries to reshore manufacturing and secure supply chains. 
• Decarbonisation continues to require enormous investment in grids, renewables, batteries and electric vehicles. 
• Defence spending is creating another source of demand for strategically important materials.

Sullivan believes these forces could produce something reminiscent of the China-driven resources supercycle. During that period, he said natural resource equities outperformed MSCI World for seven consecutive years on the way up, with some years delivering double-digit relative gains.

“We think that's happening again now. We're there again now. We've had a couple of good beats last year and this year, and we expect to have a lot more.”

The reason is that all three forces are colliding with the same constraint: the world needs substantially more critical minerals, while bringing new supply online can take years, or even decades.

Copper sits at the centre of Sullivan's thesis because of its importance to electrification. He estimates demand could grow by close to 30%, while the mining industry is not currently positioned to provide all of the required supply.

That helps explain why companies such as BHP Group (ASX: BHP) are committing enormous amounts of capital to expanding copper production.

Lithium provides another example. Sullivan believes the industry is already entering its third boom, with EV penetration offering substantial remaining growth.

The attraction is not simply higher commodity prices. New mines can take years, sometimes decades, to develop, creating the potential for prolonged tension between demand and supply.

“We're very heavily leaning into mining. That's our heritage, our specialty. There's so much action here in individual companies growing. They will grow. 

They won't care what the interest rate is. They don't care if we have a recession. They don't really care what happens in the wars. These are real assets. They're building stuff that's going to last for 20 to 100 years.”

Sullivan's portfolios are particularly exposed to copper, alongside silver and gold, while maintaining diversification across commodities rather than betting heavily on individual specialty materials.

Defence enters the software era

The final megatrend is being driven by a less welcome development: a more dangerous world.

VanEck's Hannah said global defence expenditure reached US$2.89 trillion last year, after increasing 41% over the preceding decade in inflation-adjusted terms. Australia alone is expected to allocate $887 billion to defence over the next decade.

The important investment question is where that money goes. Previous defence cycles were dominated by traditional hardware including fighter jets, tanks and radar. The next one is increasingly about drones, autonomous systems, AI, satellites, cyber capabilities and electronic warfare.

“The key takeaway is that this is a long-term structural trend. This is not a small sum of money. These are significant sums of money allocated by governments, which will be pushed out over the next 10 years.”

Hannah highlighted France's Thales, which operates across radar, sonar and cyber capabilities, as one example of the integration layer increasingly important to modern defence.

Palantir Technologies (NASDAQ: PLTR) demonstrates the growing role of software and AI, using fragmented datasets to assist operational decision-making, while RTX Corporation (NYSE: RTX) provides exposure to missiles and missile defence.

There are risks. Valuations matter, government announcements do not automatically translate into revenue, political decisions can alter spending priorities and technological disruption will create losers alongside winners.

For investors wanting diversified exposure, VanEck's Global Defence ETF (ASX: DFND) holds 45 global defence companies, with constituents required to generate more than 50% of their revenue from defence technology and be based in NATO or NATO-aligned countries.

Four trends, one enormous investment cycle

Energy, AI, resources and defence may appear to be separate themes, but the four presentations revealed how tightly connected they have become.

AI needs semiconductors, data centres and enormous amounts of electricity. Expanding the electricity system requires grids, batteries, gas generation and critical minerals. Reshoring and defence require additional industrial capacity and secure supplies of many of the same resources. Defence itself is becoming increasingly dependent on AI, software and advanced computing.

The common denominator is capital. After decades in which many of the world's biggest investment winners were capital-light digital businesses, the coming decade could look very different. Trillions of dollars may need to be spent on data centres, power generation, electricity networks, mines, manufacturing capacity and defence infrastructure.

Investors will still need to separate genuine structural winners from businesses simply attaching themselves to fashionable themes. Yet the message from Livewire Live was clear: these forces are already underway, and some of the biggest opportunities may lie in the physical infrastructure required to make them possible.