Equity markets remain supported by falling oil prices and associated reduction in bond yields, in anticipation of the peace deal announced this morning.

Last week saw the largest IPO in history as SpaceX listed on the NASDAQ at US$75.1 billion (the prior largest being Aramco at US$29.4 billion). It rose 19% on its first day of trading, putting it on 91x CY26 revenue.

This is an important sentiment indicator and, alongside the recent Alphabet capital raising, provides funding for the roll-out of AI compute infrastructure and Starlink.

An overhang of equity supply is leading to a consolidation in the US, with the S&P 500 up 0.7%.

We are also seeing rotation into some of the lagging sectors – notably REITs and Healthcare – which flowed through to Australia.

The S&P/ASX 300 gained 2.0% last week, led by rate sensitives (REITs +4.9%, consumer discretionary +7.9%) as the federal budget is increasingly seen as a handbrake on economic growth, reducing the need for further rate hikes.

We remain wary of domestic consumer plays as we expect the economy to slow, crimping earnings growth for that sector.

Market outlook

The market has been underpinned since the late-March low by:

1. A resilient global economy despite the Iran war.
2. Oil prices holding lower than feared.
3. The acceleration of revenue growth at Anthropic, signalling that AI business models work.
4. Upgraded investment in AI from hyperscalers.
5. The SpaceX IPO, which highlighted the acceleration of its revenues.
6. The combination of all these reflected in stronger-than-expected earnings growth in the US.

The combination of the continued strength in the market, the largest IPO in history at a punchy valuation, and the growing concentration of the US stock market has rekindled the debate as to whether we are near a market top.

Bears point to the historic precedent of equivalent levels of concentration at previous market peaks.

The 10 largest AI-related companies now comprise 41% of the S&P 500, whereas:

• The Nifty 50 peaked at 40% of the S&P 500 in the 1970s;
• Japan peaked at 44% of the MSCI AC World Index in the late 1980s; and the
• Tech and Telecom sectors reached 41% of the S&P 500 in the early 2000s.

Each ended in significant bear markets.

They also point to warning signs in longer-term valuation metrics. The Schiller P/E (the price over the average real earnings over 10 years) is near its previous high of the tech-wreck – as is the price/book ratio of the MSCI USA index.

The counter to this is that these measures do not capture the shift in earnings and returns on equity, particularly from tech sector.

For example:

• The S&P 500 is up ~9% in US dollar terms this calendar year, but 12-month forward EPS have been revised up 17%, which means the market’s P/E ratio has fallen despite the rise in the index.
• Returns-on-equity (ROE) for the S&P 500 have stepped-up in the last three years to a range of 20-22% (and are currently at the high end), compared to a range of 12 to 20% over the previous 30 years.

That step-up is linked to the mega-cap stocks, which are tapping into global customer bases and have strong market positions.

The factors driving higher ROEs include margin expansion (which reflects franchise quality and industry structure), lower corporate tax rates, increased leverage, and lower interest rates.

There has been some offset from lower asset turnover, however this arguably leads to an understatement of the increase because this is partly driven by the requirement to carry goodwill on balance sheets and higher cash holdings by the megacap stocks.

The key point is returns are higher and when this is mapped over the Schiller P/E it provides greater justification for current valuations than was the case in 1999.

AI return on investment

The key call then becomes the sustainability of these returns, which goes to the current debate about returns on hyperscaler capex. Capex is leading to deteriorating cashflows at the hyperscalers, which in turn drags on returns.

We note current estimates are for AI investment to be US$5.3 trillion from the four largest hyperscalers (Alphabet, Meta, Microsoft, Amazon) between 2025 to 2030 – a number that is trending higher.

The market currently believes the returns on that investment will flow through and cash flows will inflect higher from 2028, dwarfing those previously.

This call is the key to the market. If this plays out it underpins earnings and valuation rating; if not, then capex budgets will have to be cut with knock-on effects to earnings and rating.

The risks are:

1. Use cases for AI not being sufficient to drive revenue expectations.
2. Funding constraints limiting compute availability.
3. Infrastructure constraints limiting compute (power, memory, planning approvals).
4. Regulatory intervention slowing the development of models.

Regulatory intervention had been a lower-order issue but was elevated over the weekend as the US government ordered suspension of the Anthropic Fable 5 model’s availability to all foreign nationals. It is rumoured that Amazon discovered there was a way through the model’s guardrails which enabled its use in cyberattacks. Anthropic dispute that this was as serious as suggested.

SpaceX IPO

The SpaceX IPO is an important part of this debate as it provides further evidence of the acceleration of AI-related revenue growth.

The stock is the all-time largest IPO, both in terms of stock issued and market cap at listing (US$1.8 trillion).

It closed up 19% on its first day of trading, taking it to a US$2.1 trillion market cap and an EV/revenue of 91x.

Revenue is expected to grow 80%+ into 2027 and potentially to increase by 20-25x by 2030; from the current CY26 level of US$28 billion to US$450-500 billion.

The IPO valuation is not reliant on Starlink – although some estimate this will grow 7x between 2026 and 2030.

Instead, the bull case rests predominantly on growth in the terrestrial computing power of the company’s AI division, xAI.

Here, the company is forecasting 2 gigawatts (GW) by the end of 2026, 9-10GW by end-2029 and 13GW by end-2030.

How this gets monetised is an open debate, but the view is that if you build the compute the revenue will come.

In this vein, the implied return of US$50bn/GW of Anthropic’s recent leasing deal with SpaceX is unlikely to be replicated, but even a more sober US$20-30bn/GW across 8GW+, sold by 2028 at 35-40% margins, equates to ~US$100 billion of AI-sector EBIT.

Paths to monetisation include:

1. Software (reinvesting in building out the Cursor coding product, plus healthcare and robotics tools enabled by Grok),
2. Further neocloud-type deals, or
3. A better consumer business.

In terms of a broadband product, the key is delivery of the first Starship orbital flight (previous flights have been suborbital). This can in turn facilitate bigger satellites, meaning materially faster speeds and a path from ~17 million to 30-40 million subscribers.

A direct-to-device service is the harder leg. It needs one of the top three telcos to fold (which seems unlikely) or purchase of a large swathe of spectrum. The latter is more likely as there is plenty of spectrum coming online via the One Big Beautiful Bill and SpaceX have said they’ll pay whatever it takes.

Testing of the Starship V3 rocket (the rapidly reusable spaceship) later in the year is likely to be a defining event. This is at the core of the move to build out Starlink and orbital compute.

Elon Musk has floated the idea of hourly launches – in the order of 8-9,000 per year. The market is assuming this is more likely to be 100-200 per annum by 2030.

We note the smaller Falcon 9 rocket is estimated to have 140 launches this year, while the new starship is said to have a payload of 10x the current Falcon.

The upshot is that the buy-side is getting to a US$2.5-3.0 trillion valuation, versus the current day one value of US$2.1 trillion.

This belief in the revenue trajectory highlights the market’s growth optimism, which underpins the broader market given 1) the overall weight of earnings relating to AI, and 2) the flow-on effects through investment spend on industrial companies and resources.

Iran conflict

Reporting this morning suggests a deal has been agreed on US President Donald Trump’s 80th birthday and 107 days after the conflict began.

The market appears confident that it can happen, noting statements from other Gulf countries and reports of Qatari officials going to Tehran to finalise details.

There have also been reports of loading in Iraqi ports for the first time in weeks, suggesting some production is slowly ramping up.

That said, we have been here before and there were a series of attacks through last week.

Oil continues to fall on the expectation of a resolution plus reports that the US has been covertly getting vessels through the Strait, at the rate of around 2-3 million barrels-per-day (bpd) for the last couple of weeks.

Reiterating previous notes, the market has been able to offset the loss of 20 million bpd via a combination of lower Chinese imports, inventory releases, some alternative routing, some flow through the Strait, and some demand destruction.

The latter has been disproportionately met through petrochemicals and some gasoil/ diesel reduction. Regionally, it has been in Asia and the Middle East.

Overall, the tail risk around this issue continues to fall.

There will likely be a long process to begin to restock the supply chain. It will take one to two months to ramp up production and we will see an extended period of strategic stock rebuilding.

All this is likely to hold oil prices around current levels for some time – but the ability to plan for this will improve confidence and likely be supportive for growth.

US macro and policy

May inflation data was largely in-line with expectations, with headline consumer price index (CPI) at 4.2% year/year and core CPI +2.9%.

Producer price index (PPI) was a bit higher than expected and will underpin higher personal consumption expenditures (PCE), likely around 3.5%.

There are three drivers for higher inflation:

1. Tariffs – though this wave is rolling over.
2. Energy prices – this is playing out as expected but will take time to wash through as companies catch up on passing on costs.
3. AI investment-related inflation – relating to the sheer size of spend, the ongoing effect of this is uncertain.

The Fed has its first meeting under Chairman Kevin Warsh. With labour markets in decent shape and inflation not deteriorating – and with potential relief on energy prices – the Fed has time to wait to see how trends play out.

The key judgment on a six-month view is whether the economy can maintain good growth at current interest rates, which would suggest we are nearer to neutral than previously believed.

Markets

It is interesting to note the Mag-7 has been underperforming year-to-date. It is down ~2% versus the S&P 500 up ~9%, the equal weight S&P 500 up ~10%, and the small cap Russell 2000 index ~+18%.

This may reflect both the capital intensity increase as they spend more on capex and also the need to fund the SpaceX IPO and secondary issuance.

Total CYTD issuance in equities has accelerated, leading to a draw on the market.

If we factor in debt issuance, which is also up due to AI funding needs, 2026 has already surpassed 2025 levels.

Sentiment on software – which had been improving – has reversed significantly in the last two weeks driven by the combination of a return to semis (with software used as a funding source) and the launch of Anthropic’s Fable 5 model which renewed concerns around sustainability of traditional franchises.

We have seen early signs of a recovery in lagging sectors of the S&P 500 (REITs and Healthcare), perhaps reflecting a combination of a stabilisation in the outlook for rates, positioning (underheld) and valuation proving supportive.

This is important for our market, and we saw early signs of life in these sectors last week in Australia.

Australia had a good week led by REITs, staples and discretionary. This would indicate the market is getting more comfortable that, with the apparent softness in the economy, the RBA may not need to raise again, which would help rate sensitive sectors.

We think this may be premature optimism, as the RBA needs the consumer to slow meaningfully, to offset strong investment spending and population growth and bring inflation down.

So whether this needs more rate increases or not, the outcome would be for domestic earnings to be under some pressure.

Tech (-4.5%) was the underperformer, mirroring the drop in US software.

IMPORTANT INFORMATION: This document has been prepared by Enhance Financial Partners, ABN 45 146 707 173 AFSL 515518, based on our understanding of the relevant legislation at the time of writing. While every care has been taken, Enhance Financial Partners makes no representations as to the accuracy or completeness of the contents. The information is of a general nature only and has been prepared without consideration of your individual objectives, financial situation or needs. Before making any decisions, you should consider the appropriateness for your personal investment objectives, financial situation or individual needs. We recommend you see a financial adviser, registered tax agent or legal adviser before making any decisions based on this information.