Berkshire’s Two-Front AI Strategy: Tokyo Dividends Today, Power Grids Tomorrow

Greg Abel is making clear that Berkshire Hathaway’s future rests on two pillars that, at first glance, have little in common: the trading floors of Tokyo and the transmission lines of the American Midwest. Speaking on Thursday, the CEO reaffirmed the conglomerate’s decades-long commitment to its stakes in five Japanese trading houses, while simultaneously flagging a very different kind of constraint on the artificial-intelligence boom — not semiconductor supply, but the electricity needed to power the data centers that will run the algorithms.

The Japan position is now firmly entrenched. Abel said Berkshire expects rising profits from Mitsubishi, Mitsui, Itochu, Sumitomo and Marubeni, with the potential for additional support through dividends and share buybacks. It’s a bet on patient, predictable income streams — a deliberate contrast to the more volatile growth narratives that dominate technology investing.

That patient approach extends to the energy side of the house. Abel views AI data centers as a meaningful growth opportunity for Berkshire’s utility operations, but he cautioned that permitting processes and grid-connection delays could cap near-term expansion. His warning, first voiced on Wednesday, cuts against the prevailing narrative that chip availability will determine the pace of AI deployment. Berkshire’s vantage point suggests otherwise: power infrastructure is the real bottleneck.

The company is effectively positioned on both sides of the AI trade. Through its expanded Alphabet stake, Berkshire holds a capital bet on the technology itself. Through its own energy infrastructure, it stands to benefit as a supplier of the electricity those technologies will consume. That dual leverage is likely to draw closer analyst scrutiny in the coming weeks as grid-capacity data becomes clearer.

The portfolio shift behind this positioning has been dramatic. Five stocks now account for 63 percent of Berkshire’s invested assets — a concentration rarely seen in the company’s history. Alphabet has climbed to the third-largest equity position, trailing only Apple and American Express while overtaking Coca-Cola and Bank of America. The reordering follows a second quarter in which Berkshire held roughly 106 million Alphabet shares across Classes A and C, valued at approximately $36.6 billion. The publicly traded equity book stood at nearly $360 billion as of September 1.

Should investors sell immediately? Or is it worth buying Berkshire Hathaway?

That Alphabet build was part of a broader reallocation. Berkshire trimmed Bank of America during the quarter while adding to Delta, D.R. Horton, Lennar and Macy’s — moves that analysts have framed as contrarian capital deployment under Abel’s leadership. The shift also marked a turning point: after 14 consecutive quarters of net selling, Berkshire became a net buyer of stocks in the second quarter of 2026, with roughly $20 billion in net purchases. First-half figures showed $39.4 billion in acquisitions against $27.8 billion in disposals.

The operating business, meanwhile, tells a more mixed story. Second-quarter operating earnings rose to $12.98 billion from $11.16 billion a year earlier, lifted by stronger results in manufacturing, services and retail, along with gains at the BHE energy unit and the BNSF railroad. Insurance told a different tale: underwriting results fell 13 percent to $1.73 billion, while investment income from the insurance segment dropped 9 percent to $3.06 billion.

That divergence between a robust core and a softening insurance arm tempers the optimism surrounding the Alphabet purchases. For shareholders, the growing concentration in five names also introduces a clustering risk into what has historically been a broadly diversified conglomerate.

The market has so far taken the strategic pronouncements in stride. Shares closed Friday at EUR 653,500.00, down 0.3 percent on the day and 3.0 percent over the past 30 days. The stock sits 4.7 percent below its 52-week high of EUR 686,000.00, reached on August 10, though it remains up 2.6 percent for the year. With 30-day volatility at 16 percent, the subdued price action suggests investors are reading Abel’s Japan reaffirmation and grid-capacity warning as confirmation of known strategy rather than fresh catalysts.

The open question for Berkshire watchers is whether Abel’s more aggressive capital deployment marks a permanent departure from his predecessor’s cautious allocation — and whether the concentration in five core holdings will prove a source of strength or vulnerability as the quarters ahead unfold.