The ASX Earnings Season Setup: Can Record Highs Survive August?

ASX earnings season begins with the market pushing into record territory.

Gold is almost 28% below its January high. Silver is about 25% off the boil. Australian interest rates are back at 4.35%.

Labor has taken a wrecking ball to investor confidence by smashing the CGT discount and removing negative gearing.

All while pocketing the extra dollars to fund a bloated NDIS and a ravenous public budget.

House prices are off the boil.

First-home buyers are replacing a fear of missing out (FOMO) with a fear of buying the high (FOBTH).

Inflation levels in the mid-to-high 3% range remain the central-bank equivalent of someone dragging their fingernails across a chalkboard.

It’s unsettling. You want it to stop. But throwing the chalk duster at their head, or in this case raising rates again, comes with consequences.

This is no perfect bull-market setup.

But there are positives.

Employment remains close to — the mythical — full, helped by that bloated NDIS thing. Household debts still look manageable. The RBA remains determined to ‘look through’ temporary price shocks.

That’s central-bank speak for a framework full of rules, forecasts and models, right up until the rules become inconvenient.

ASX earnings season is kicking off. Dividends follow. Before you’ve finished counting the cash, the Santa Claus rally will be on the horizon.

Not a perfect bull-market setup.

Not a horrendous one either.

And yet the S&P/ASX 200 (ASX:XJO) keeps climbing.

So where to next?

Let’s look at the setup as we get stuck into earnings season and what to watch. We’ll throw a few interesting stocks to watch out there as well.

But first, valuation is everything, so let’s check in.

What ASX Earnings Season Must Prove

Are we due for a crash or ready to keep punching out highs?

The Australian market trades on roughly 20-21x trailing earnings and 17-18x forward earnings. Its cyclically adjusted P/E, or CAPE, sits near 18.8x.

That’s higher than about 73% of Australia’s historical observations.

Expensive.

But nothing crazy.

Crazy is out there.

The ASX’s trailing P/E is well above its long-run median of around 15x. Its forward P/E is also above average. Its dividend yield has fallen to roughly 3.3%, about one percentage point below its ten-year norm.

You’re paying more for each dollar of earnings.

You’re getting less income in return.

That doesn’t mean the market must fall. Valuation is a lousy clock. Expensive markets can become more expensive when earnings keep growing.

It means there’s less room for disappointment.

Keep that in mind this earnings season. Misses may be punished harder than usual.

Now look across the Pacific.

The S&P 500 Index (SPX) trades near 40.5x cyclically adjusted earnings, placing its CAPE in the 99th percentile of its own history. On forecast earnings for the next 12 months, the same index trades around 22-23x. That measure looks less extreme, but it still relies on strong corporate earnings growth.

The long-run US CAPE average is around 17x. It reached about 33x before the 1929 crash. The dot-com boom pushed it above 44x.

The Buffett Indicator is flashing the same warning.

This is a different measure. Shiller CAPE compares share prices with ten years of inflation-adjusted earnings. The Buffett Indicator compares the value of the whole US share market with annual GDP.

The classic Buffett Indicator currently sits near 241%. US-listed companies are worth about 2.4 times the annual output of the US economy.

That reading is inflated by globalisation because American companies earn a large share of their revenue overseas. A globalisation-adjusted version sits near 116%. Even with that caveat, the classic measure is ‘strongly overvalued’.

We’re back in that neighbourhood.

The bulls have an answer. Artificial intelligence will transform the economy. Productivity will explode. Margins will widen. Tomorrow’s profits will justify today’s prices.

They may be right.

In February, we made the productivity case in The AI Productivity Superstorm. Intelligent agents may give one worker the output of a small team. Add robotics and the productivity shock escapes the screen.

Factories. Mines. Warehouses.

The opportunity is enormous.

So is the price already being paid for it.

When Good Technology Meets Bad Prices

The internet changed the world.

The dot-com investors understood the technology. They were early, indiscriminate and willing to pay almost anything.

Amazon changed retail. Google reorganised information. Broadband rewired business.

Thousands of other dot-com companies disappeared.

The Nasdaq Composite (IXIC) fell almost 80% between March 2000 and October 2002. Even some survivors took years to recover because their share prices had outrun the businesses underneath them.

A technology can transform the world and still be a terrible investment at the wrong price.

Which brings us to SpaceX (NASDAQ:SPCX).

SpaceX completed the largest IPO in history in June. It raised about US\$75 billion at a valuation near US\$1.77 trillion. Almost three times Saudi Aramco’s previous fundraising record.

The company generated US\$18.7 billion of revenue in 2025.

Investors paid close to 94x trailing sales.

SpaceX owns extraordinary launch technology, Starlink and a dominant position in commercial launches. This is no speculative shell with a slide deck and a dream.

But the valuation assumed something beyond extraordinary.

The deal bundled rockets, satellite internet, artificial intelligence and the ambitions of Elon Musk into one US\$1.77 trillion package. The shares surged.

Then they fell.

SpaceX has since reported 92% quarterly revenue growth. Even that has not settled the argument. Investors are starting to ask what they receive after funding the colossal capital bill required to build the promised future.

That’s the question across the AI trade.

Revenue isn’t profit.

Compute isn’t free.

Disruption creates shareholder value only when shareholders capture the economics.

We covered the other side of this in The Great ASX Tech Smackdown. AI will hurt weak software companies. It will pressure prices and force incumbents to adapt.

The market struggled with narrative panic then.

It may be struggling with narrative euphoria now.

The ASX Only Needs America to Hold Together

The US market is the biggest external risk facing Australian investors.

Australia doesn’t need Wall Street to keep rising at this pace. We need it to hold together. Or, at least, avoid coming apart too fast.

If it does, the back end of 2026 could be good for the ASX.

The ASX costs roughly 17-18x forecast earnings, against 22-23x in the US. Australia also offers a higher dividend yield and franking credits.

The discount exists for a reason.

America owns the dominant global technology platforms. Australia owns banks, miners, supermarkets and industrials. US companies tend to offer higher margins and larger markets.

The gap should not close.

It doesn’t have to.

A partial narrowing would support Australian equities, especially if commodity earnings stabilise and the domestic economy holds up.

Australia grew only 0.3% in the March quarter. Soft, but not broken.

Employment is the hinge.

If Australians keep their jobs, service their mortgages and avoid a disorderly housing correction, domestic earnings can hold together.

That’s where the stock picking gets interesting.

ASX Consumer Discretionary Could Hold Some Contrarian Winners

Households are under pressure.

Mortgage rates are high. Energy is expensive. The household saving ratio fell from 7.0% to 6.2% in the March quarter.

Yet household weakness doesn’t hit every consumer company the same way.

If employment holds and house prices avoid a sharp fall, consumer confidence can recover before the RBA cuts. People don’t need to feel rich. They need to stop feeling poorer every month.

That could bring money back into selected consumer discretionary names.

Cettire (ASX:CTT) is the speculative version of that trade.

The online luxury retailer has stumbled. Its trailing sales revenue is about \$731 million. Its market capitalisation has fallen to roughly \$75-80 million.

That’s close to 0.1x sales.

High-growth platforms can trade at 5-10x revenue when the market believes in their margins and runway.

Cettire trades for about ten cents on the revenue dollar.

There are reasons.

Revenue has softened. The company lost money over the latest trailing period. Luxury demand has been weak and investors have lost faith in the bridge from sales to profit.

A low sales multiple doesn’t make a stock cheap if the sales never become profit.

But the distribution engine is still there.

Cettire launched its direct Chinese platform in 2024. It already had a relationship with JD.com. This year it added a flagship store on Alibaba’s Tmall Global, including access to the Luxury Pavilion.

Those are serious channels into the world’s largest luxury market.

Access alone creates nothing. Cettire must acquire customers at rational prices, control returns, defend its delivered margin and turn revenue into free cash flow.

Right now, it appears to be buying market share.

That’s acceptable for a period.

It can’t become the permanent business model.

The next result should tell us more. Watch China sales, repeat purchases, delivered margins and operating cash flow.

The upside potential is large.

So is the burden of proof.

AI’s ASX Picks and Shovels

Australia’s not going to build the next Nvidia overnight.

Our cleaner AI exposure sits underneath the software.

Land. Power. Copper.

Every model needs somewhere to run. Every data centre needs a grid connection. Every connection needs substations, transformers, cables, cooling systems and contractors.

Maas Group Holdings (ASX:MGH) is one of the more interesting examples.

Wes Maas has spent years buying businesses, integrating them and sweating the assets. He isn’t a loud technology evangelist. He’s an operator from Dubbo who knows how to turn separate businesses into one machine.

Earlier this year, Maas agreed to sell its construction materials division to Heidelberg for \$1.7 billion.

That was the crown jewel. Quarries. Concrete. Asphalt.

Then the new strategy emerged.

Maas is moving capital towards electrification, high-density power, AI compute clusters and data-centre infrastructure. It has invested in Firmus Technologies, while its retained electrical and infrastructure capabilities give it a route into the buildout.

We recently broke down the playbook in our MAAS Group synergies case study.

The risk is clear.

Maas is selling a proven division and leaning into a theme full of hype. The transaction is expected to complete in the second half, so investors are still waiting to see the final balance sheet and the quality of the replacement earnings.

But Wes Maas has built enough businesses to deserve attention.

He flies under the radar.

That probably suits him.

SKS Technologies Group (ASX:SKS) offers a more direct contractor exposure. In April, SKS expanded a hyperscale data-centre electrical contract from \$130 million to \$210 million after the facility grew from 90 megawatts to 126.

One project.

That’s the scale of electrical work underneath the AI story.

Then there’s Goodman Group (ASX:GMG), NEXTDC (ASX:NXT), Megaport (ASX:MP1) and Infratil (ASX:IFT). Several featured in The AI Productivity Superstorm.

The opportunity is substantial.

So is the capital bill.

Watch contracted returns, power availability and gearing. A billion-dollar pipeline looks impressive. It means less if the capital is expensive and the eventual return is mediocre.

Again, revenue isn’t profit.

The Metals Behind the Machine

Data centres are buildings filled with machines.

Machines need metal.

Copper carries the electricity. Silver sits inside high-performance electrical systems. Lithium and other battery materials support storage. Rare-earth magnets power motors, cooling systems and robots.

This is where AI, electrification and supply-chain security collide.

Copper is the clearest structural exposure. The world needs more generation, transmission and computing capacity. New mines take years to permit and build.

Silver offers industrial demand and monetary sensitivity, but it is volatile. The metal broke above US\$100 an ounce in January before retreating about 25%.

Gold has cooled too. It finished July near US\$4,043 an ounce, almost 28% below its January high.

Yet the long-term monetary case remains.

Government debt keeps growing. Currency intervention is back. Central banks remain trapped between inflation and weak growth.

Rare earths matter for the next stage.

In June, we explained how China controls close to 90% of rare-earth processing and why humanoid robots may require two to four kilograms of permanent magnets each. Read the full argument in Rare Earth Elements: Seventeen Metals, One Chokehold.

That brings us to robotics.

The Mobile Brains Have Arrived

Robotics has been waiting for intelligence.

The hardware has been competent for years. Industrial robots can weld, lift and repeat the same movement thousands of times.

What they could not do well was adapt.

Move the object. Change the lighting. Put the robot in a room it has never seen.

It struggled.

Modern AI changes the equation. Vision models improve perception. Language models allow natural instructions. Reinforcement learning helps machines adapt through experience.

The robot finally gets a mobile brain.

This could become the next great wave on AI’s coattails.

The first gains will come in controlled environments. Warehouses. Factories. Mines.

Then the machines move into hospitals, construction sites, shops and homes.

The investment chain reaches far beyond robot manufacturers. Sensors. Semiconductors. Batteries. Motors. Rare-earth magnets. Electricity.

We will expand on this theme soon.

For now, remember the central point.

AI escapes the screen when it enters a machine.

That’s when the productivity story gets much larger.

The Edge

The case for a good finish to 2026 is simple.

Wall Street avoids a valuation collapse. Australian employment holds up. Commodity and infrastructure earnings support the index.

Three things can break it.

First, US valuations.

America doesn’t need a recession to fall. Investors only need to decide that a CAPE above 40x and a market worth 241% of GDP is too much.

Second, Australian housing.

A gradual decline would help affordability. A fast one would hit confidence, construction and bank credit quality.

Australians spend differently when the value of their home is falling.

Third, inflation.

Yes. Inflation again.

We’ve discussed it almost every month this year. Back in January, Brace for Runaway Inflation argued that markets were treating inflation like a post-pandemic fade when the risk looked more persistent.

The RBA has since lifted the cash rate back to 4.35%.

June headline inflation eased to 3.8%. Trimmed-mean inflation held at 3.6%.

The panic eased.

The problem didn’t disappear.

Another energy shock, stronger wages or renewed housing inflation could force the RBA to tighten again. That would hit discretionary spending, housing, leveraged companies and long-duration growth stocks.

It’s not panic stations.

Stay vigilant.

The ASX is expensive. Wall Street is expensive on another scale.

If the US holds together, Australia has room for a good finish to 2026. Our valuation discount, resource base and growing exposure to AI infrastructure give us several ways to participate.

But this won’t be an easy index trade.

Cettire needs margins.

Maas needs execution.

SKS needs delivery.

Wall Street needs AI to live up to one of the most expensive promises markets have ever made.

There’s opportunity here.

There’s also more risk than the record-high index wants you to believe.