What’s moving in the markets


The AI bubble debate just got more interesting

Companies are using the public markets to raise equity again.

For the past two years, the standard rebuttal to AI bubble concerns went something like this: yes, the capital being deployed into data centres is extraordinary (and reminiscent, in scale, of the fibre-optic frenzy of the late 1990s), but the companies writing the cheques are among the most profitable enterprises in human history. These are cash-flow machines funding the next leg of growth from their own earnings. If the returns disappoint, you take the write-down and move on. Painful, but survivable. Debt-fuelled speculation, this is not.

But that narrative is quietly shifting.

SpaceX is heading for a public listing. Rockets and satellites are the headline, but Elon Musk has been unambiguous that AI is the real prize… and few would bet against a meaningful share of that $80 billion IPO finding its way to xAI. OpenAI and Anthropic are also eyeing the public markets, likely before year-end. Meanwhile, last week Google announced it was raising $80 billion through a share issuance. Days later, rumours surfaced that Meta was contemplating something similar.

Alphabet is raising $85 billion from equity markets to fund AI infrastructure projects

A pattern is emerging. The industry appears to be entering a second phase: one where even the most cash-generative companies on earth are choosing to dilute shareholders rather than slow down. The ROI on all of this remains, to put it charitably, uncertain.

Our take: This is not a cause for alarm; it is capital markets functioning as designed. The public equity model exists for exactly this reason, to allow corporations to fund investments of a scale that internal cash flows alone cannot justify.

AI fits that description. The opportunity is large enough, and the cost of underinvesting high enough, that the major players have little rational choice but to stay in the race. As early as 2024, CEOs were saying as much publicly: better to over-invest and absorb some pain than to cede ground and spend a decade regretting it.

As shareholders in Alphabet, we are not troubled by the share issuance. If anything, with the stock trading at a full valuation, this is precisely the moment to be raising equity: more capital per share diluted.

Broadcom: A masterclass in ignoring the noise

Volatility is the price to pay for outperformance

Broadcom reported last week, and the business is compounding at a remarkable pace. Custom ASICs (chips Broadcom designs exclusively for AI workloads) alongside AI networking hardware remain the primary engine, with overall revenues up 48% year-over-year and AI semiconductor revenue up a whopping 143%. The fundamentals, in other words, are exceptional.

Yet, the stock fell 20% in two days.

This is a valuation story, not a business story. After a 44% rally in April alone, the shares had been priced to reflect not just strong results but an upgrade to guidance. When CEO Hock Tan stepped to the microphone and simply reiterated the company's 2027 targets rather than raising them, the market sold first and asked questions later. At close to 100x free cash flow, there is no margin for anything less than perfection.

Broadcom dropped 20% after it reported Q2 earnings

A second cloud came from the company’s marquee client, Google, which indicated it intends to spread its chip procurement across multiple suppliers. It is a useful reminder that customer concentration is a real risk, even for the best businesses.

That said, the concern should not be overstated: Broadcom's roster of custom chip customers now includes Anthropic, Meta, and OpenAI… a client list that reads like a who's who of the AI capex cycle. Losing some Google wallet share is manageable when the rest of the table looks like that.

Our take: This is a near-perfect illustration of what it means to hold a winner through volatility. Up 44% in a month, down 20% in two days. That is the price of admission for owning a business growing this fast, in a sector this contested. It is uncomfortable, especially when the valuation becomes extremely demanding.

The right response is to look through the share price and focus on what actually matters: will the business earn a lot more money five years from now? In Broadcom's case, the answer is almost certainly yes. As they say, the market is a voting machine in the short run and a weighing machine in the long run.

Other Updates


Ratings

  • Accenture (ACN) got a rating initiation

    “Like many of its peers, Accenture’s valuation has come under pressure in 2026, reflecting a broader market anxiety about the long-term implications of AI for knowledge-based work(…) We think the case for Accenture’s continued relevance remains compelling, particularly over the medium term. The global pivot to AI might be one of the largest sources of demand the firm has ever seen. Every corporation in the world is trying to figure out how to adapt, and the complexity of doing so at enterprise scale plays directly to Accenture’s strengths. At a valuation of 10x free cash flow, the shares looks relatively attractive today.”

    Click here to view the full update.

Upcoming Earnings

Wednesday

  • Oracle (Q4)

Thursday

  • Adobe (Q2)

Want to see what we’re actually doing with real money?

Upgrade to unlock full access to the RatedA Portfolio (with weekly updates on every buy/sell and rationale), and the complete database of companies + ratings.

Keep Reading