Google’s parent company Alphabet, Inc. Engages in some unusual and creative equity-raising practices for a large, publicly traded company. Long story short: It’s raising large amounts of investor money that it can put into expanding its footprint in AI hardware.
It plans to raise $80 billion through new “equity offerings,” according to Bloomberg. This includes $40 billion of new stock that came into the stock market in the third quarter (meaning probably starting in July). For high rollers, Alphabet is offering $30 billion in special underwritten shares with “mandatory convertible preferred stock” backed by Goldman Sachs, JPMorgan and Morgan Stanley.
And for good measure, Berkshire Hathaway is committing $10 billion to shares. Hopefully an SEC filing soon will clarify what kind of terms they are getting.
And it’s coming at an absolutely inopportune time for anyone else competing for investors hungry for AI exposure. SpaceX, Anthropic, and possibly OpenAI have applied for initial public offerings, either confidentially in the case of Anthropic, or by issuing a larger old prospectus like SpaceX.
In other words, if you take a peek, it looks like Google is playing strictly with SpaceX and others. Public offerings tend to benefit investors with deep pockets, and Google is also coming along and reducing the bulge in their pockets a bit. As Bloomberg’s in-house analyst Mandeep Singh says in a Bloomberg article, “There’s only so much capital you can allocate, even in the public markets.”
it is In fact Zero sum? Does competing on such a large scale in the equity market at such an auspicious moment reduce the enthusiasm for IPOs of other companies? When you consider that investors (Robinhood users, say) can and do regularly take money out of other stocks in the stock market to invest in an IPO, and the price of this kind of thing is typically fixed (not to mention the fact that SpaceX is increasingly being tracked in indexes like the Nasdaq), then maybe not that much.
And I also suspect that large institutional investors snapping up bank-backed preferred stocks are putting all their eggs in one basket. They can potentially even invest in all of the above if they wish.
However, the Berkshire part of the deal is a single, huge, concrete decision that actually removes $10 billion of capital investment that could have gone into these IPOs.
And according to Bloomberg, it’s all an effort to turbocharge Google’s chip-manufacturing operations. Its tensor processing units, or TPUs, have been competing internally with Nvidia’s GPUs for some time now, and could soon be in for a piece of the AI processing pie if a critical mass of AI companies take interest. The $80 billion raise is an effort to double or perhaps triple or quadruple Google’s AI silicon ambitions.
<a href