Datt is ditching retail and refiners, unsold on AI - What he's buying instead

It remains hard to ignore the shadow that the AI supercycle is casting over equities markets, but that doesn't mean there aren't bright spots elsewhere worth exploring.

Where others are bullish, Datt Capital's Chief Investment Officer Emanuel Datt says it remains cautious on AI, given how tenuous the technology's bull case is in reality.  

"Investors are being asked to underwrite long-duration capital commitments against short-duration technology," said Datt. Too much is still unquantified, and there's no shortage of attractive opportunities on offer elsewhere, he says.  

"We would rather deploy capital into high-quality businesses whose competitive position does not depend on the specific configuration of the next silicon cycle."

Here, Datt talks through the big recent changes to the Datt Capital Absolute Return Fund, including why it's moving away from oil names and is instead overweight a commodity that has frequently been overlooked in the recent discourse: coal. 

He also details the notable recent additions to the portfolio and gives us the state of play on the ASX medtech company backed by Pro Medicus that has been a recent standout performer and is using AI the right way. 

What was the most notable addition to the portfolio recently and why?

We recently established positions in HUB24 (ASX: HUB) and Netwealth (ASX: NWL), Australia's two leading specialist investment platforms.

The thesis is fairly straightforward where a decade long migration of adviser flows away from legacy bank and institutional platforms toward superior technology and service is well underway, but far from complete. Combined, HUB and NWL still administer only a minority share of the Australian platform market, yet continue to consistently capture the lion's share of net industry inflows.

Valuations have reduced significantly below historical norms, but paying a fair multiple for durable compounders has, in our experience, proven cheaper than buying mediocre businesses at bargain prices.

What was the most notable sell or downsize in the portfolio recently and why?

First, Asian refining margins have compressed as new capacity met softer regional demand. Singapore cracks, which anchor local domestic refiner margins, have re-rated lower over time.

Second, China's decision to relax refined product export quotas is exporting the excess directly into Ampol and Viva's price-setting benchmark. Government fuel security payments cushion the downside potential.

Third, and less appreciated, both companies' convenience retail arms are being structurally impaired by Australia's illicit tobacco crisis. With excise driven price gaps now enormous, credible estimates put illicit share of tobacco consumption above one-third of the market. Tobacco has been the anchor category for foot traffic and margin however, this potential is diminishing.

Both companies are quality franchises, but the earnings base is at risk outside of improved refining operations.

What’s your most notable overweight and why?

Our most notable overweight is Australian coal — specifically Whitehaven Coal (ASX: WHC) and New Hope Corporation (ASX: NHC).

The immediate driver is a severe dislocation in global gas markets. Qatar Energy's force majeure on LNG cargoes since March, following Iranian strikes on Ras Laffan and the effective closure of the Strait of Hormuz, has taken roughly one-fifth of global LNG supply off the market. European buyers, entering summer with storage at the lowest level since 2021, are now bidding hard against Japanese, Korean and Chinese utilities for LNG cargoes ahead of the northern hemisphere winter.

Thermal coal is the natural switching fuel, and Asian utilities are already dispatching more of it. Whitehaven and New Hope sit precisely where that demand lands, high-quality Newcastle-benchmark thermal into the Japan-Korea-Taiwan market, whilst returning cash to shareholders.

The market is still pricing terminal decline whilst we think the cyclical setup could not be more favourable.

What’s your most notable underweight and why?

Our most notable underweight is Australian discretionary retail.

The consumer setup is deteriorating on multiple fronts simultaneously. 

The RBA has hiked three times in 2026, taking the cash rate back to 4.35%, with market expectations of further rises. Capital city dwelling values are now falling with Sydney and Melbourne both posted roughly 3% declines in the June quarter reversing the household wealth effect that has underwritten Australian discretionary spending for many years.

In addition, the May Federal Budget delivered the most significant tax overhaul in three decades. CBA estimates the negative gearing and CGT changes are equivalent to a 90–155 basis point rate hike for property investors, compounding higher mortgage rates. 

The domestic addressable market is being increasingly contested by global firms with Temu, Shein and Amazon continuing to take share.

What’s been one of your most notable performers over the quarter?

One of the strongest contributors this quarter has been EchoIQ (ASX: EIQ), an Australian medical AI company whose software analyses echocardiogram data to detect structural heart disease: aortic stenosis and heart failure (under FDA review) via its software EchoSolv.

Our long-standing question was never whether the technology worked, the Mayo Clinic and Mount Sinai deployments had partially validated that, but whether a small Australian company could distribute an enterprise clinical AI product across the fragmented US hospital system at speed.

The 25 June binding agreement with Pro Medicus (ASX: PME) solves both problems elegantly. Pro Medicus is investing up to A$20m via convertible notes and, more importantly, becomes a US reseller of EchoSolv through its enterprise imaging platform running at many of the largest US health systems and academic medical centres.

A key part of the distribution question has been answered. The credibility that comes attached with this deal is arguably the bigger prize and we anticipate the company can scale rapidly.

What themes and trends are dominating discussions right now?

Data centres, compute and all things AI have dominated conversation, and the ASX pipeline reflects it. 

Firmus Technologies is expected to seek an ASX listing at a A$6 billion-plus valuation, Sharon AI is pursuing a secondary Australian listing, and a widening cast of existing small-caps are pivoting their narratives toward AI infrastructure. Every second broker note is now anchored on hyperscaler capex.

We remain cautious. The bull case rests on hyperscaler spending, Microsoft, Alphabet, Amazon, Meta and Oracle collectively committing US$700 billion in 2026, roughly double last year's outlay being sustained indefinitely, and on end-user demand that remains genuinely unquantified. Nvidia's annual architecture cadence renders each generation of installed silicon partially obsolete within 18–24 months. 

Investors are being asked to underwrite long-duration capital commitments against short-duration technology. At this juncture we would rather deploy capital into high-quality businesses whose competitive position does not depend on the specific configuration of the next silicon cycle.