Which ASX stocks still have a moat in 2026?

I’ve spoken with enough fundies about AI to know the pattern. Software, productivity, massive addressable markets and the companies supposedly poised to dominate the next decade. It’s a familiar conversation.

The question worth asking in 2026 is a more specific one: which ASX stocks actually still have a competitive moat?

The answer, according to Reece Birtles, Head of Australian Equities at ClearBridge Investments, is not where most investors are looking.

So when I sat down with Birtles, who argued that AI may actually be bad news for parts of SaaS, and potentially great news for businesses like Telstra (ASX: TLS), I paid closer attention.

What followed was one of the more genuinely surprising conversations I’ve had on the AI theme, with some clear implications for how investors are positioning right now.

Why AI is eroding the moats investors thought were strongest

For much of the past decade, investors favoured a very specific kind of company. Businesses with low capital intensity, recurring revenue, annual price increases and large addressable markets were treated as premium assets worthy of premium valuations. Think software, cloud, the names that have dominated every growth portfolio since COVID.

“For the last 10 years there’s been a one-way bet,” Birtles told Livewire. “Quality has been defined as low capital intensity, having a technology solution that you could raise prices on every year and believing there’s a large addressable market.”

Birtles argues AI may fundamentally challenge that playbook – not by stopping these companies from growing, but by eroding the durability of their competitive advantages.

“AI can attack those technology companies,” he says. “Suddenly you can do it cheaper. You can’t raise prices because the next SaaS offering is going to be cheaper. It’s going to be harder to grow your market because there are going to be more competitors, and all your barriers to entry have fallen.”

He points to a simple but powerful valuation exercise. A high-PE SaaS stock on 50x earnings, compared to a traditional business on 15x earnings growing at 5% p.a., needs to sustain earnings growth of 15% p.a. all the way out to 2038 just to break even on valuation. 

That was already a stretch before AI. Now, with AI compressing pricing power and multiplying competition, Birtles argues investors need to shorten the time horizon they are willing to place on these growth runways – which means those stocks have to derate.

"Who's going to believe in a long growth runway with the pace of change?" he asks.

ClearBridge owns none of these names. Birtles specifically flags WiseTech (ASX: WTC) and Xero (ASX: XRO) as businesses where AI could shorten their competitive advantage period and increase the investment required to remain competitive. 

He won't be looking at them "until they're closer to a market multiple." Software need not disappear from portfolios, but investors may need to become much more selective, particularly among expensive growth names relying on long-duration narratives.

The ASX stocks where the moat is still intact

If AI creates more competition in software, who benefits? Birtles believes investors should look toward established businesses with infrastructure, scale and pricing power. His clearest example is Telstra (ASX: TLS).

“If you’re Telstra with a dominant market share in mobile, you’re setting the price,” he says. “If you can get productivity benefits from AI, you can lower your cost base, but you’re not necessarily going to have to lower your price because no one can do it better than you.”

His logic is simple: AI cannot replicate what has taken decades to build. Telstra’s network is not something a competitor can whip up overnight.

This same thinking extends beyond telcos. Birtles argues that any company with established customer bases, dominant market positions and physical assets that took decades to build stands to benefit on the same logic. For these businesses, AI is a productivity tool that flows straight to the bottom line without being competed away.

“For us, all those incumbents look stronger in an AI world,” he says. “Whereas all those SaaS tech stocks look under more pressure to grow, even if they still grow.”

Rather than asking: Who sells AI? Birtles believes investors should increasingly ask: Who becomes more productive because of AI without losing their moat?

The 2 stocks ClearBridge is buying now

Birtles was heading to another event and had about three minutes left when I asked him the question investors love most: what are you actually buying?

Resmed (ASX: RMD)

The first name was perhaps the most surprising coming from a manager known for owning traditional value stocks. Resmed – the global leader in sleep apnoea devices and respiratory care – is not a stock ClearBridge would typically own. However, the global healthcare sector selloff has been broad and indiscriminate, and Birtles sees a clear line between stocks that deserve to fall and those that are simply collateral damage.

“ResMed has double-digit earnings growth, strong penetration, is high quality, going great, and hasn’t had an earnings downgrade,” he says. “Healthcare stocks worldwide are selling off on a thematic and indiscriminately and ResMed has not had one problem.”

For context, Commonwealth Bank (ASX: CBA) trades on approximately 25x earnings with limited growth. ResMed is on 16x and growing double digits.

He contrasts this explicitly with Cochlear (ASX: COH), which he says does have fundamental issues. Birtles emphasised that separating businesses with genuine operational issues from those caught in sector-wide pessimism is exactly where active management earns its keep.

"That's the type of thing where we now see value in a quality growth name," Birtles says. "It's a very different sort of exposure for us."

Light & Wonder (ASX: LNW)

The second name was Light & Wonder, the gaming technology company that Birtles says was trading on less than 10x earnings at the time of purchase – well below what he would normally consider for a business delivering double-digit earnings growth. The attraction to him was not only valuation, but what the stock’s behaviour revealed about today’s market.

After a softer quarter that had already been pre-flagged and despite reiterating full-year guidance, the stock fell 9%, only to rebound strongly up 13% the following day.

“You can’t tell me there’s not crazy things going on in terms of flows,” he says.

In Birtles' view, that kind of price movement isn't rational, it's what happens when flows replace fundamentals. For active managers willing to do the work, it keeps creating opportunities.

Why energy and resources still matter in an AI world

Energy and resources tell the same story. The market just hasn't worked it out yet.

AI is power hungry and data centres require enormous amounts of electricity. Birtles estimates they are currently adding around five terawatts of demand per annum to a global grid of approximately 180 terawatts. That is a new and meaningful demand driver for energy businesses that had largely been written off.

The supply side hasn't kept pace – two oil refineries, decades of underinvestment, demand is only going one way. In his own published commentary on Livewire, Birtles expands on the specific names he sees as best positioned.

Santos (ASX: STO), Ampol (ASX: ALD) and AGL Energy (ASX: AGL) sit within this broader view on energy. On the resources side, BHP (ASX: BHP) and South32 (ASX: S32) offer copper exposure, while Lynas Rare Earths (ASX: LYC) – arguably the leading rare earths producer globally – and Independence Group (ASX: IGO), owner of the world’s highest-grade hard rock lithium mine, round out the critical minerals picture. 

These are the commodities, Birtles argues, that will be critical in powering electrification and AI infrastructure over the coming years.

The core argument is consistent with his broader framework: find the mispricing, get in early, let the business do the rest.

The bigger picture: a market still mispricing the obvious

Both stock picks, and the broader energy thesis, sit within a larger argument Birtles has been making since late 2024 – that the Australian market is in a period of extreme valuation dispersion comparable to the GFC and the peak of the tech bubble.

Cheap stocks currently trade approximately 38% below the broader market. The historical norm is closer to 20 to 25%. The gap has narrowed slightly from its peak of 40%, but remains historically extreme.

He attributes much of the distortion to the rise of passive investing. Daily trading volumes in names like Commonwealth Bank and Wesfarmers (ASX: WES) are down approximately 40% from pre-COVID levels, as passive flows effectively turn large-cap index stocks into permanent, illiquid holdings. 

In practice that means stocks drifting for no reason, then moving violently when there is one. That is more or less exactly what happened to LNW.

"The market is becoming inefficient," Birtles says. "And it's becoming better for an active manager who can identify that mispricing."

I walked into that conversation expecting another predictable AI take. I didn't get one. Whether you agree with Birtles or not, the question he is asking is one worth sitting with: in a world where AI lowers every software barrier to entry, what actually constitutes a durable competitive moat?

Managed Fund

ClearBridge Select Opportunities Fund

Australian Shares

View