Oh, Australia

Second Level Thinking · Issue 47

A single theme has split the globe into two economies, and why, lately, I’ve found myself genuinely worried about us.

I’ve been turning this over for months now, and underneath all the charts and the data sits something simpler and more uncomfortable: I’m worried about where Australia is heading. Not in a vague sense, but in a specific, structural, look-at-the-numbers way.

If you’d spent this year watching only the ASX, you’d think very little had happened; the index has drifted around breakeven for months, flat, quiet and dull. No story. But that flat line is one of the most misleading things in global markets right now, because what sits above it, beyond our shores, is the largest and most concentrated creation of wealth in living memory. And it is happening almost entirely without us.

Let me show you what I’m seeing, and you can tell me if I’ve got it wrong.

Two worlds, and the line between them

The defining market story of the decade is the AI trade, and this year it changed character. For three years it was a narrow affair — a handful of American mega-caps doing the heavy lifting. In recent months it has widened into something closer to a global industrial arms race, a scramble to build the physical machinery of artificial intelligence. This sea change has sorted the world’s markets into two camps with a clarity I’ve rarely seen: the economies with genuine exposure to it, and everyone else.

Rank the world’s major share markets by their returns this year and the pattern barely needs commentary. Sitting at the very top, in a league of their own, are Korea, Taiwan and Japan - the former up over 80% and comfortably ahead of the American indices, which themselves have posted double-digit gains. The common thread among the leaders isn’t clever monetary policy or luck. It’s silicon. These are the economies that spent decades building real positions in semiconductors and the hardware AI runs on — Samsung and SK Hynix in Korea, TSMC and MediaTek in Taiwan, the equipment and materials makers feeding the whole chain. They planted those trees thirty years ago, and this is the harvest.

Now run your eye down to the bottom of that same list. There, keeping company with Switzerland and trailing every significant European market, sits Australia — the ASX 200 fractionally negative for the year, the broader All Ordinaries worse still. We are not mid-table. We are at the back, in the company of the few developed markets that, like us, simply have nothing meaningful to sell into this boom.

I don’t think most Australians have grasped how cleanly we’ve been sorted. We have no semiconductor industry to speak of and no real AI exposure. What technology we do have lives almost entirely at the software and services layer — the very layer AI is busy compressing rather than enriching. Our two biggest listed tech names have fallen roughly a quarter this year and only a couple of smaller players touching data infrastructure have caught any of the updraft at all. It isn’t that we backed the wrong horse. It’s that we never entered the race.

It always comes back to compute

In our last issue I went deep on compute — the rise of decentralised compute networks, and why the ability to access raw processing power is becoming one of the defining strategic questions of the age. I want to draw a line between that issue and this one, because they’re really the same story told from two ends.

Everything above — the chips, the memory, the hardware arms race, the national indices going vertical — is, at its core, a global scramble for compute. The reason Korea and Taiwan are winning is that the world cannot get enough of the thing they make, and the thing they make is compute in physical form. Whether you approach it from the frontier (the decentralised networks we discussed last time) or from the foundations (the silicon and data centres driving these markets), the conclusion converges: compute is the strategic resource of this era, the way oil was of the last one. And Australia, at every layer of that stack — the chips, the hardware, the infrastructure, even the energy to power it — has chosen to be a spectator.

Compounding our woes: our headline index is propped up by its largest members, which masks what’s happening underneath. Strip out the size-weighting and look at the typical Australian stock — give every company equal say — and the market isn’t flat at all. It’s down around five per cent, and broader still it’s closer to minus six. The “steady” ASX is, beneath the surface, a market where most companies are falling.

The cleanest way to see it is to simply count. In the United States, more than half of all listed companies are up on the year. On the ASX 200, only about a third are — meaning roughly two in every three Australian stocks are down in 2026. That is not stability. It is a thin crust of a few large names stretched over a hollow. For anyone trying to build momentum in this market — traders chasing breakouts, investors hunting growth — it has been quietly brutal. The flat line isn’t peace. It’s a market holding its breath.

I don’t want to pin all of this on AI, because that would be too easy and not quite honest. Some of our weakness is homegrown. Healthcare — long one of the proudest corners of the ASX — has fallen by about a third this year, against declines of just four or five per cent for the equivalent US and UK sectors. A stronger Australian dollar eating into offshore earnings and a higher-for-longer rate environment punishing growth stocks explain much of it. These are our own headwinds, layered on top of the global one. The point isn’t that everything is someone else’s fault. It’s that even our home-built strengths are wobbling at the worst possible moment.

The one thing holding us up — and how we’re treating it

So what is keeping the ASX from simply falling over, if most of its stocks are sinking? You already know the answer, because it’s the same answer it always is: our rocks.

Energy and materials have been the only reliably green part of the market this year, and because the big miners carry such enormous weight in the index, they have almost single-handedly held the headline number above water. I want to be clear-eyed and grateful about this. Take BHP, Rio, Fortescue and the energy names out of the picture and we aren’t flat — we are falling, plainly and visibly, to the bottom of the developed world. Our endowment of dirt, and the companies who pull it out of the ground, are the only reason this is a disappointing year rather than an alarming one.

But I want to sit with this, because there are two hard truths inside it.

The first is that this commodity strength is a bid, not a boom. The buying is being driven by geopolitical supply disruption, the energy transition, and a flight into precious metals — not the deep, structural, decades-long demand of the 2000s, when the industrialisation of the developing world swallowed every tonne of iron ore and coal we could ship. And here is the part people miss: the AI build-out will not refill that well the way China once did. Modern AI hardware is value-dense but material-light. A leading-edge chip is one of the most valuable objects humanity makes, and it’s fashioned from a trivial amount of physical matter. A trillion dollars poured into AI infrastructure moves a tiny fraction of the raw tonnage that a trillion dollars of roads, rail and apartment towers once did. We cannot assume the dirt will save us next time.

The second truth is the one that genuinely frustrates me: we are making life harder for the very sector keeping us afloat. Approval timelines for major resource projects under our federal environmental laws blew out by around sixty per cent in recent years, to an average of nearly four years per major decision. The estimated cost to national output from these delays runs into the tens of billions of dollars. Projects, transmission lines, and the critical-minerals developments the whole world is begging for sit tangled in process while the window of opportunity narrows.

And capital, unlike governments, doesn’t wait. Watch what the majors actually do rather than what they say. BHP — our own national champion — has thrown in with a partner to develop one of the largest undeveloped copper discoveries on the planet. Not here. In Argentina, high in the Andes, where it is roughly doubling its spend to around eight hundred million US dollars this year alone, with talk of a total build measured in the many billions. Why Argentina? Partly the geology, yes — but listen to the executives and the word that keeps surfacing is the incentive regime that Milei’s government has rolled out to court exactly this kind of investment: tax certainty, customs relief, a clear path. A more pro-investment government has lured one of the great Australian miners to pour billions into copper on the other side of the world. That is what voting with your feet looks like. That is capital telling you, in the only language it speaks, where it feels wanted.

Meanwhile we are sitting on one of the richest critical-minerals and rare-earths endowments anywhere — the neodymium, dysprosium, lithium and the rest that the energy transition and the defence world are scrambling to secure. We are the most-explored rare-earths destination in the world. And yet we extract and, more importantly, process it far too slowly, hamstrung by an approvals system widely described even by its reformers as slow, complex and inconsistent, and by a near-total absence of domestic refining.

The same paralysis shows up in compute itself. The world is in a frantic race to build data centres — the physical homes of the AI economy — and Australia has genuine natural advantages for it: stable ground, vast cheap land, the potential for abundant energy. We could be a regional hub for the very compute everyone is fighting over. Instead we have a couple of small listed names in the game and a build-out that is nowhere near the pace the moment demands. Strong resources sector we won’t let grow; a compute opportunity we won’t move on. The pattern repeats.

The own goal

You’d think a country structurally short of high-growth, wealth-creating enterprise — a country watching the future get built elsewhere — would be doing everything imaginable to encourage the founders and the capital who might build us a seat at the table. Instead we are doing close to the opposite.

This year’s federal budget announced the deepest overhaul of capital gains tax since 1999 — scrapping the long-standing discount in favour of indexation, and imposing a thirty per cent minimum tax rate on gains from 2027. The structure we’ve just adopted is close to unique among advanced economies, and for a great many ordinary investors and founders, that floor now sits above the headline rate they’d pay in the United States, Britain, Germany, Spain or Italy. Not to mention fast-growing areas like Dubai and Singapore. We haven’t built the world’s highest wall, we’ve built one of its least welcoming ones, aimed squarely at the people we most need to stay.

Pile on the new superannuation taxes for larger balances, and a coming minimum tax on the discretionary trusts that ordinary family businesses actually use, and the direction of travel is unmistakable - We are choosing all of this precisely when our productivity can least bear it. Australian multifactor productivity actually went backwards in the most recent full year. Labour productivity growth has crawled along at under a third of a per cent annually for a decade — a fraction of our long-run average — to the point where our own Productivity Commission’s gentlest framing is that we are merely back to where we stood two years ago. Standing still, in other words, while the countries we measure ourselves against compound away from us.

So where does that actually leave us?

We throttle the resources sector that is the only thing holding our market up. We move slowly on the critical minerals and the data-centre build-outs that could define our next thirty years. Our productivity is flat to falling. We have almost no exposure to the AI and compute supercycle reshaping global wealth. And, faced with all of that, our policy response is to raise the tax on precisely the entrepreneurs and builders who might have dug us out.

Stifle the builders, slow the build-outs, tax the risk-takers, and watch the productivity numbers and the capital both quietly leave. Where does that leave a country? I’ll tell you what I fear it leaves: a beautiful, lucky, well-educated nation slowly converting a once-in-history resource inheritance into a comfortable, managed decline — taxing capital like a mature welfare state while generating the growth of a stalled one. Not a high-growth innovation economy, because we’ve chosen through policy not to become one. But not a lean, low-tax, capital-friendly haven that pulls builders in from elsewhere either. Stranded in between. A resource economy with a thin software veneer and no clear bet on what comes next. No identity. Caught flat-footed in the gap between what we were and what we’ve declined to become.

Now, almost none of this is permanent. The cyclical hand is survivable: trades rotate, commodity cycles turn, strong currencies soften, high rates fall. We have not been left behind because something is broken in us. We’ve been left behind because the theme driving the world right now is one we happen to have no stake in, and themes change. The endowment is still in the ground. The talent is still here. The advantages — the land, the stability, the minerals, the educated people — haven’t gone anywhere.

But the structural choices are ours, and ours alone. Whether to free the resources sector or strangle it. Whether to move on compute and critical minerals or dither. Whether to welcome the builders or tax them toward the exits. Those aren’t the weather. Those are decisions. And right now, decision by decision, I think we are quietly choosing the smaller version of ourselves.

That’s what I’m worried about. I’d genuinely love to be talked out of it.

Parting Shot

People think good decision-making is about being right all the time. It’s not. It’s about lowering the cost of being wrong and changing your mind.

When the cost of mistakes is high, we’re paralyzed with fear. When the cost of mistakes is low, we can move fast and adapt.

Make mistakes cheap, not rare.

Shane Parrish