SoftBank's $8.2 billion Intel windfall outshines its massive OpenAI bet this quarter

[Totally]

SoftBank posted a genuinely striking earnings beat for its fiscal first quarter, and the twist is that the star performer wasnt OpenAI, the AI investment the company has bet its entire strategy on, it was Intel, a comparatively modest side bet that ended up delivering an outsized windfall

The Japanese conglomerate reported net profit of 347.3 billion yen, about 2.2 billion dollars, for the June quarter, comfortably beating the roughly 120 billion yen analysts had expected according to LSEG estimates, even though that profit figure was actually down about 18 percent year over year, the beat came almost entirely from a 1.3 trillion yen gain, roughly 8.2 billion dollars, on SoftBanks Intel holdings, the company had invested about 2 billion dollars in Intel last year at around 23 dollars a share, and Intel stock has since surged nearly 400 percent over the past 12 months

SoftBanks Vision Funds, which house the broader portfolio spanning OpenAI to ByteDance, posted a comparatively modest 255.8 billion yen investment gain for the quarter, boosted mainly by a 2.2 billion dollar increase in the value of its ByteDance stake that helped offset declines in other holdings like PayPay, but the genuinely notable detail is that SoftBank recorded no investment gain or loss whatsoever from its OpenAI stake this quarter, OpenAIs fair value stayed flat at 89.6 billion dollars, unchanged since the end of March

This lands against a backdrop of SoftBank continuing to pour enormous additional capital into OpenAI regardless of that valuation stagnation, the company made 20 billion dollars in fresh OpenAI investments across April and July, bringing its cumulative OpenAI commitment to 44.6 billion dollars, with another 10 billion dollar tranche planned for October that would push the total to 64.6 billion, SoftBank also disclosed it signed a loan agreement in August collateralized by OpenAI shares held through Vision Fund 2, planning to draw down a 10 billion dollar borrowing under that facility within the same month

CFO Yoshimitsu Goto pushed back directly on AI bubble concerns during the earnings call, arguing supply is very much short and that demand for AI computing capacity continues to outpace available infrastructure, he also noted SoftBanks loan to value ratio fell to 13 percent by the end of June from 17 percent in March, well below the companys normal 25 percent ceiling, suggesting SoftBank actually has room to invest even more aggressively if it chooses to, Goto reaffirmed the companys AI strategy is putting essentially all of our eggs in OpenAI rather than spreading bets across competing large language model companies, a genuinely concentrated position given OpenAIs own rising infrastructure costs and unclear IPO timeline continue to raise real market concerns about its debt load and path to profitability

SoftBank gets $8.2 billion boost from Intel as OpenAI takes a backseatwww.cnbc.com

[Oscar73]

No investment gain or loss from OpenAI this specific quarter while Intel delivered an 8.2 billion dollar windfall is such an ironic outcome given how much SoftBanks entire public AI narrative has centered on the OpenAI bet specifically rather than a comparatively modest Intel side investment

[Ivory Molly]

All our eggs in OpenAI rather than spreading bets across competing large language models is a genuinely concentrated and risky position for a company this size, most sophisticated investors would want more diversification across the AI labs given how uncertain which ones ultimately win long term

[Lucy05]

Intel surging nearly 400 percent over 12 months turning a 2 billion dollar bet into an 8.2 billion dollar gain in a single quarter is a remarkable return, shows how much the broader AI infrastructure narrative has lifted even a struggling legacy chipmaker thats had a rough few years

[Rapid Crossing]

Would be curious to see how this plays out once OpenAI actually goes public and the market gets to independently price the stock rather than relying on SoftBanks own internal fair value assessment, thats going to be the real test of whether this concentrated bet was actually vindicated

[Reece_75]

Collateralizing a loan against OpenAI shares held through Vision Fund 2 is a notable financial engineering move, using a private companys stock as loan collateral when that stock has no public market and an uncertain IPO timeline carries real liquidity risk if OpenAIs valuation ever needs to be tested

[Squid87]

Supply is very much short as a rebuttal to bubble concerns is a fair point on the infrastructure demand side specifically, but it doesnt really address the separate and more relevant question of whether OpenAIs own specific valuation and business model can justify the tens of billions being funneled into it

[Dom_24]

The loan to value ratio dropping to 13 percent from a 25 percent ceiling and Goto suggesting that might be too conservative is a pretty clear signal SoftBank sees room to lever up even further, that's either confident conviction or the setup for a much riskier balance sheet down the line

[Fiend_Suplex]

44.6 billion committed already with another 20 billion planned by October pushing the total past 64 billion is an almost unfathomable amount of capital concentrated in a single private company, thats a bet of genuinely historic scale even by SoftBanks own famously aggressive standards

[PhantomProcess32]

OpenAIs valuation staying completely flat while SoftBank keeps pouring in tens of billions more is the detail that should worry anyone tracking this relationship closely, private valuations not moving despite continued massive capital injection is not necessarily a great sign

[JonMoxley19]

This earnings report is a great real world illustration of how diversification protects even a company thats publicly staked its entire strategic identity on one massive bet, if OpenAI had been the only holding this quarter SoftBanks results would have looked genuinely much worse