"An awakening in animal spirits": Ben Griffiths on the SpaceX effect and what comes next

Locally, the ASX rode a volatile but modestly positive path, with a contentious federal budget and a wave of earnings downgrades adding friction along the way.

In a recent note, Ben Griffiths of Eley Griffiths Group argued, however, that the quarter’s headline act came in the form of the blockbuster listing of SpaceX, which he says has reignited long-dormant retail investor appetite in Australia.

“That this deal can cause an investor stampede in Australia points to a long absent awakening in animal spirits for the local household investor,” Griffith says.

“Whilst overwhelmingly positive in the first instance, it should be viewed cautiously in the context of the evolution of the market cycle and what such euphoria ultimately portends.”

SpaceX and the future of IPOs

In what Griffiths branded “undoubtedly the most extraordinary event of the quarter”, the US$75 billion raise for Elon Musk’s space launch/satellite communications/AI company flagged a more interesting story around the broader interest in IPOs.

“A bespoke Australian prospectus received royal ascent from ASIC with Commsec confirming >28k retail applications for stock following 37k new account openings,” Griffiths says.

“Elon Musk retained a 42% stake plus several venturesome KPI’s, including one predicated on the establishment of a 1 million person Mars colony. The stock debuted at a 19% premium to offer price.”

Retail investors opening new brokerage accounts specifically to chase an offshore listing is certainly not the norm. But is this just the Musk affect or will the appetite remain for other mega IPOs?

“Anthropic (US$1tn valuation), OpenAI (~US$852mn) and Kalshi (US$40bn) are reportedly close to announcing initial public offerings and perhaps the same herd of fevered local investors will look to participate here,” Griffiths says.

“Domestically, we have AI infrastructure group Firmus Technologies tapping capital markets with ease (~US$330m Series 1 raising in September 2025, ~ US$520m series 2 in December 2025 and US$505m series 3 in April 2026). The AFR on July 9 reported that the group is now poised to launch a series 4 round raising of ~$2.9bn ahead of an ASX listing at some point in 2026. This will be a good test of local market vitality.”

Things may not be quite as frenzied as they seem, despite these massive listings, with overall volume down considerably.

AI capex boom

It’s no secret that the AI hyperscalers have been spending big. There’s also no signs of it slowing down in the near term.

“The month of June saw investors aggressively rotate away from AI infrastructure names (those making the AI investments) into semiconductor stocks (those benefitting from the AI spend),” Griffiths notes.

“Focus was trained on ever-increasing sector capex (broad consensus of US$725bn for CY26e/USD$729bn CY27e) abetted with increasing financial leverage and concerns around the prospects of appropriate returns at scale for much of the investment.”

The question is what this massive spending means going forward. If previous capex cycles are anything to go on, then according to Griffiths the current AI capex boom “may contain the seeds of its own demise-in time”.

“Recall the US Shale gas boom of 2008-2014, the Internet boom of 2000, the electrification of the 1920’s and the British Canal mania of the 1790’s where capital supply → excess capacity → poor returns and a contraction in equity valuations,” he says.

“Edward Chancellor in his excellent book Capital Returns: Investing Through The Capital Cycle urges investors to watch rising capex closely as it is frequently a warning signal for future profitability, tending to proliferate at peaks and dissipate at troughs.”

Beyond the mega caps

The market rally through the June quarter was marked by a consistent pattern across major indices, with multiple running gaps, enlarged traded volumes, and old high reclamation, a feature of the S&P 100, S&P 500 and Nasdaq Composite. But it wasn’t confined to just the mega caps.

“The correction to the Bloomberg Mag 7 Price Index presents as a healthy reset in price level with the primary uptrend in tact,” Griffiths says.

“The persistent strength of the S&P 1000, Russell 2000 and the Russell Microcap benchmarks is notable, with these rallies neatly in price/time balance. The outperformance of value stocks versus growth names in recent months needs to be monitored given investors long term disposition (at least since 2013) towards growth names.

“Whilst weaker seasonal effects might impact on stocks through to October, not fighting the tape continues to be the dictum for investors in US stocks.”

Locally, the All Ordinaries “congested into a symmetrical triangle through the period”, and the firm believes the index's February 27 record high "feels safe from retest in the short term".

Be a fox, not a hedgehog

Greek poet Archilochus famously said: “The fox knows many things, but the hedgehog knows one big thing.”

How does this relate to investing? Griffiths and the firm's Head of Quantitative Strategies Pieter Stoltz explain that forecasters that are looking to predict stock market returns in the next one, five or 10 years should aim to be the fox, not the hedgehog.

Running what he framed as a “Bayesian Horserace”, he tests which statistical approach - simple average, median, or compound annual growth rate (CAGR) - best forecasts future returns across 13 asset classes, including the ASX 200, S&P 500 and US Treasuries. In short, the results show the average performs best over a one-year horizon, while CAGR wins convincingly at five and 10 years.

Layering in valuation signals via a Bayesian framework improved on both, outperforming a comparable "Kahneman" adjustment method across most horizons. The conclusion for long-term forecasting: start with historical CAGR, then adjust using valuation data while treating any one-year forecast as “really a wild guess”.

“If a Martian came to Earth seeking to make a long term return forecast, our study suggests they should start with the historical CAGRs. They might then consider incorporating valuation information using a Bayesian approach.

“At shorter horizons like 1y, they should start with the arithmetic average but be aware of large forecasting errors, likely due to variables like investor flows. At any rate, the base rates tested above should be something for ‘hedgehog’ forecasters to think about.”

Outlook

Ultimately, the primary trend for global equities continues upwards, but this looks like a mature cycle, not an early one, and investors should watch for the cracks that typically show up before a top.

“US credit spreads remain benign, inflation break-evens are in retreat and with the real fed funds rate at 0.35%, it’s probable the Federal Reserve will remain on hold with rates for the foreseeable future. Valuations are fulsome but earnings momentum continues to underwrite that confidence,” Griffiths says.

“The case for equities remains intact with a few structural developments, noted above, to keep a weather-eye on.”