Author: Le Ming
May 2, 2026,Omaha. Eighteen thousand seats were only slightly more than half full—in the past, tickets to the Berkshire Hathaway shareholder meeting were in high demand, with lines forming to get in and hotels surrounding the venue fully booked.
This time, 95-year-old Buffett did not host on stage as in previous years; instead, new CEO Greg Abel stood in front of the main screen, answering investors’ questions about why Berkshire holds nearly $400 billion in cash.
In the same week, six thousand miles away in Tokyo, Son Masayoshi’s team was doing something else: bundling SoftBank’s unprofitable AI assets into a new company called Roze AI, targeting a $100 billion valuation and planning to list in the United States in the second half of 2026.
The reason is simple: SoftBank must continue raising and investing money to cover OpenAI’s $64.6 billion check, which could eventually grow to nearly $100 billion.
One holds $397.4 billion in cash, buys nothing, and waits for the market to crash; the other carries $100 billion+ in interest-bearing parent company debt, betting the market won’t crash.
Berkshire: Having too much money is itself a problem
What does $397.4 billion mean?
It is nearly 40% of Berkshire Hathaway's total market capitalization and twice the average cash level Warren Buffett has maintained over the past two decades.
Of this $339.3 billion, it is directly held U.S. Treasury bills, and Berkshire Hathaway has also become one of the U.S. Treasury’s largest non-government creditors.
This pile of money wasn't passively accumulated—it was actively built.
Over the past fourteen quarters, Berkshire has been a net seller of stocks in every quarter. Apple, once its largest holding, has been continuously reduced over four consecutive quarters starting from the third quarter of 2024, with nearly 688 million shares sold, generating over $100 billion in proceeds.
Buffett's explanation has always been the same: he can't find anything cheap.
In his 2024 letter to shareholders, he wrote: "Generally speaking, nothing looks attractive." During a recent off-site interview at the shareholder meeting,he compared today’s market to "a church next to a casino," saying that of all the market sentiments he has experienced, this one is the most like gambling.,
The problem is that he has already been making this judgment for more than a year.
Over the past twelve months, Berkshire's stock performance has lagged the S&P 500 by approximately 40 percentage points. This is no small margin—it is one of the largest relative drawdowns since Buffett took over Berkshire in 1965, with the last time such a significant underperformance occurring during the final stages of the dot-com bubble in 1999.
At the time, Buffett said he would only buy "bricks, carpets, insulation, and paint"—industries at the cutting edge. Two years later, the Nasdaq dropped 78%, proving him right.
But this time, investors have waited for more than two years. The market rose, but Berkshire remained unchanged; the market rose again, and Berkshire still didn’t move. On January 1st of this year, the day Buffett officially stepped down as CEO, Berkshire’s stock slightly declined—the market expressed a restrained disappointment in the most restrained way possible.
This was the situation Abel inherited. He was a Canadian accountant who had spent his entire career rising to vice chairman of Berkshire Hathaway, working in heavy-asset, regulated, slow-moving industries such as utilities, railroads, and energy.
He is not Buffett, and he knows he is not Buffett.In his first letter to shareholders, he repeatedly emphasized "continuity" and "decentralization"; at the first shareholders' meeting, his absolute response to every suggestion of splitting the group was "impossible."
Abel's dilemma is that he cannot deploy this cash (because the market is too expensive), nor can he continue to pretend it doesn't exist (because investors are voting with their feet).
If the market continues to rise over the next five to ten years, he will eventually have to confront a question that has never been seriously discussed in Berkshire’s history—give the money away.Either return it to shareholders as a special dividend, or truly break up and sell off the monster stitched together from more than 60 subsidiaries.
Will Berkshire die? It won’t die suddenly.Its assets are too diversified, its cash reserves too abundant, and its liabilities too low for any external shock to truly break it. But it will slowly and gracefully transform into something else.
SoftBank's problem: Not enough money, but still betting on.
Sun Zhengyi's situation is the mirror opposite of Abel's.
On February 27, 2026, SoftBank released an announcement. The most critical sentence translates to: "SoftBank Group's cumulative investment in OpenAI is expected to reach $64.6 billion, representing approximately a 13% stake."
$64.6 billion, 13%. This is the most expensive single bet of this era.
To understand the insanity of this number, look at how SoftBank could afford it.
The interest-bearing debt at the SoftBank parent company level rose from ¥12.14 trillion in March 2025 to ¥16.34 trillion in December 2025. The parent company’s so-called cash reserves amount to only about ¥3.8 trillion, nearly one-third of which consists of unused committed credit lines, not actual cash.
Where did this money come from? SoftBank raised $20 billion by pledging its shares in Arm and approximately $7.7 billion by pledging its shares in its Japanese telecom subsidiary, SBKK.
On March 27, 2026, SoftBank signed an unprecedented $40 billion bridge loan, led by JPMorgan Chase, Goldman Sachs, Mizuho, Sumitomo Mitsui, and Mitsubishi UFJ, later expanding to eight institutions—among the largest bridge loans in Asian history. Of this amount, $30 billion was directly allocated to co-invest in OpenAI. The term is 12 months, meaning SoftBank must repay $40 billion by March 2027.
This is why Son Masayoshi seemed a bit out of character this year: he liquidated all of SoftBank’s NVIDIA shares, withdrawing $5.8 billion in a single transaction in October 2025. At a Tokyo speech in early December 2025, he admitted: "I didn’t want to sell a single share of NVIDIA, but I needed more money to invest in OpenAI. I sold NVIDIA through tears."
Son Masayoshi went to great lengths to raise funds for his investment in OpenAI: SoftBank sold its stake in T-Mobile—569 million shares in the first three quarters of fiscal year 2025, raising $12.7 billion; another 125 million shares in the fourth quarter, netting $2.3 billion. It also fully divested its holdings in Deutsche Telekom and Alibaba. At the end of April this year, SoftBank initiated a $10 billion margin loan secured by its OpenAI equity, with an interest rate nearing 8%.
At the same time, SoftBank has been issuing bonds everywhere.In November 2025, SoftBank issued a bond worth 500 billion yen with a coupon rate of 3.98%; in April 2026, it followed up with a subordinated bond worth 418 billion yen, offering a coupon rate of 4.97% for the first five years—this is the most expensive retail bond in SoftBank’s history and the highest coupon rate ever recorded for a Japanese non-financial corporate yen retail bond—indicating thatthe market has begun to harbor doubts about SoftBank’s debt.
The credit market reacted immediately: SoftBank's 5-year credit default swap surged to 355 basis points in early March, reaching an 11-month high.
Son Masayoshi's latest "relief" is hoping that OpenAI will go public as soon as possible; otherwise, if the debt pressure becomes prolonged, it could truly trigger a crisis at SoftBank.
However, while OpenAI’s CEO Sam Altman advocates for an IPO in the fourth quarter of 2026, CFO Sarah Friar argues for delaying it until 2027—the public split between the CEO and CFO itself signals to the market that the company is uncertain about whether it’s ready.
One must die
Berkshire's demise is gentle.
It won’t go bankrupt—its subsidiaries are all high-quality cash cows, its debt levels are extremely low, and even if the AI capital expenditure bubble bursts, data center demand halves, or the S&P 500 drops 50%, Berkshire’s cash reserves are more than enough to devour any bargains for the next decade.
Its death is the death of identity— the compounding myth of Buffett’s approach to "buying great businesses at bargain prices" may no longer hold in a world where valuations consistently exceed 30 times earnings and wherethe 10-year Treasury yieldhas shattered all traditional valuation models.
Abel may execute well as a "rational CEO"—continuing operations, making modest buybacks, and pursuing small acquisitions on the margins—but Berkshire Hathaway as a disciplined narrative of capitalism died the moment Buffett stopped writing letters. Its shell remains, but its soul has departed.
SoftBank's death sentence could be severe.It has three death triggers; any one of them being pulled could set off a chain reaction:
The first trigger is OpenAI.If its IPO is delayed until 2027 or even 2028, if Amazon’s $35 billion commitment tied to the IPO ultimately fails to materialize, or if OpenAI’s revenue growth stalls in even a single quarter, OpenAI’s 13% stake on SoftBank’s balance sheet would be revalued downward.
The second trigger is Arm.Arm is SoftBank’s only truly liquid asset that is still highly valued by the market—with a market cap of approximately $200 billion, of which SoftBank holds 87%.
Arm's royalty revenue increased by 26% year-over-year in the third quarter of fiscal year 2026, with data center-related royalties doubling, serving as one of the pillars supporting SoftBank's overall valuation narrative.
But Arm is also the most easily re-priced asset today—if its forward P/E ratio drops from the current 70x to a "normal" semiconductor valuation, the coverage ratio for SoftBank’s $20 billion Arm-backed loan would collapse.
The third trigger is the refinancing itself.The $40 billion bridge loan matures in March 2027, and before then, SoftBank must achieve at least one of the following: take OpenAI public, take Roze AI public, sell another batch of assets, or issue another debt of similar size to roll it over.
But each path is more expensive than the previous year—the coupon rate on SoftBank’s retail bonds has risen from 3.98% in 2025 to 4.97% in 2026.
Individually, the probability of each of these three triggers is not high—OpenAI is likely to go public, Arm is unlikely to collapse immediately, and the credit market is unlikely to shut its doors to SoftBank overnight.But they share one critical characteristic: they are highly correlated, not three independent events.
If the bubble doesn't burst, Son Masayoshi will ultimately be deified:Roze AI goes public at a $100 billion valuation, OpenAI completes its IPO, and by the time he turns 70, he will realize the ASI（Artificial Superintelligence）narrative he has championed.
Meanwhile, Berkshire will be gently, continuously, and irreversibly marginalized under Abel’s steady management—until one day, a successor is forced to do what Buffett never would.