June 2026 Update

June, 2026

Haven't you sold and gone away?!

May was another strong month for equities apart from those sensitive to interest rates which rose particularly at the long end of the curve. The global indices rose over 4% in A$ terms but the infrastructure complex (adversely affected by interest rates) fell almost 3%.

The global trust performed (gross) in line with the benchmark rising over 4% in A$ terms and the Global 30, a Value biased portfolio, approximately 60bps under its relevant hurdle. Over the last year both strategies have outperformed their respective benchmarks by over 18% and over 16%.

Notable price moves in global equity holdings were seen in Japanese companies NGK +35%, Hirose +29%, Stanley Electric + 17%, Orix + 16%, Amada +15%, T&D +11%. In the USA we enjoyed returns of 60% in Qualcomm.

This can't and won't continue. We sold at month end and into early June some long held I.T. companies such as Teradyne, Hirose, Akamai, and significantly reduced Qualcomm and Arrow Electronics. We signalled at the end of last year we would be looking to take the Momentum price exposure risk out of the portfolio as much as possible and this last switching and selling has probably achieved that. We lucked out since we managed to complete orders before the sell-off on Friday 5th June in the USA which continues into Asia as we commit to this update.

Euphoria surrounding A.I. and its required capital expenditures, and SpaceX and Anthropic IPOs at these levels, is irrational. Growth requires capital investment and the capital has to come from somewhere. We have pointed out that the trilogy of strong employment, executive share option support programmes (aka buy-backs) and a significant increase in capital expenditures or drain on market liquidity, cannot all be met at the same time. One or more, of these must give. We guessed that executive option plans (buy backs) would take precedence over headcount and thus we expected lay-offs - tick:- see Meta, Atlassian, Oracle, Amazon amongst others.

We now must gauge whether the market can accept the increase in capital expenditures promised by technology companies, and discounted in P/E ratings, or will baulk and question the likely return on capital if the companies remain committed and keep spending. If the latter, then any reduction in capital expenditures will be felt not just in the I.T./A.I. sectors but in other sectors too. Currently we anticipate that commitments will continue and thus favour, as we have for a while, the picks and shovels providers of the frenzy. This is because regardless of final power demand, the truth is that power generation, transmission, and storage, will require higher investment even if the promised A.I. data centre spend fails to live up to the highest expectations. The "Tech Bros" psychology, testosterone, and experiences in the 'internet bubble' where they probably watched as juniors while their bosses gave up too early in the fight for the internet dollars and left it all to Meta, Google et al, mean they won't make the same mistake twice. This one will be a different mistake. As we write this, we learn that the index agencies probably won't give full weight to SpaceX, Anthropic, such that the passive index source of capital will not be available. Given that no one wants to be out-invested, the forthcoming capital calls may well be a surprise to a market addicted to buy backs, the exact opposite?

SpaceX and the Equity Issuance Signal

Enter SpaceX, which filed to list on Nasdaq on 12 June at a fixed price of $135 per share, implying a valuation of approximately $1.77 trillion. Let's add some context: the company generated $18.7 billion in revenue in 2025, making the implied price-to-sales ratio somewhere above 90 times. Losses have ballooned - from a $791 million profit in 2024, the company turned to a $4.9 billion loss last year, largely because 76% of its capital expenditure (roughly $40 billion per annum) is now directed toward AI infrastructure rather than rockets. SpaceX is raising $75 billion. The valuation has more than doubled since December.

The fixed pricing structure is itself intriguing and worth examining. Traditional book-building allows price discovery; the market tells the company what it is worth. A fixed price does not. On the contrary, it tells the market what the company has decided it is worth and invites the public to agree. If the equity market deteriorates, a fixed-price deal carries more risk than its advocates typically acknowledge. There is no downward safety valve for adjustment.

More broadly, SpaceX is part of a pattern of IPOs. Anthropic has filed confidentially for a Q4 IPO at a valuation reportedly approaching $965 billion, having doubled its valuation in roughly two months. OpenAI recently raised $122 billion at a valuation of $852 billion. Combined, the three mega listings could require the market to absorb approximately $200 billion of new equity - a figure that exceeds the combined US IPO proceeds from all listings above $50 million between 2022 and the first quarter of this year. When the most valuable companies in the world are raising capital at this pace through equity issuance, it is reasonable to wonder about the founders' assessment of their own valuations. They are, in aggregate, selling, and the market is buying. History suggests this asymmetry rarely ends well for the side with less information.

From our March ramblings

"A.I. companies continue to insist they will spend the money required and this will have implications for share buyback programmes which have supported share prices. Be wary of high levels of capital investment combined with continued share buybacks because good balance sheets will quickly become levered. Be also wary of the off-balance sheet financing of these committed investment programmes, and believing lazy analysis which will look at cash flow and say "no problem', while failing to account for the real cash cost of share option programmes. Meta anyone?"

And from April

"Our central case remains that a structural dash for energy and material security and a re-industrialisation will drive a re-appraisal of the very stocks which have been out of favour for many years. Add in the balance sheet constraints (and increasingly obvious creative accounting?) of the tech darlings such as Meta and Oracle and the 'maximum pain to the maximum number of investors' role of the market will have been again realised. We have pointed out that with $1 trillion of buybacks supporting share prices, particularly of the heavily favoured stocks, but now a promised surge in capital investment, that the "free" cashflow of many of these largest companies will become er …'constrained'. If buybacks cease, then share market support is removed. If capital investment fails to meet promises, then analysts will question the growth rate and there will be downstream implications for the A.I. ecosystem. No surprise then that head count is being rapidly and massively reduced in many companies? Way back we did use headcount changes in a quantitative model as indicative of management confidence in their companies' prospects. Time to revisit?"

Finally, just because the index moves downwards or sideways doesn't mean you can't make a decent return. The concentration of the index in the largest stocks means that any serious fall in these stocks will be reflected in the index due to their weighting, but not necessarily in all stocks. It remains our central case that the best hedge for your hard earned against rapacious, fiscally incontinent, governments is to hold dividend paying equities with sensible multiples, with sensible management, operating in essential areas of the economy.

Delft Partners June 2026