“ The Risk of Staying the Course

Holding $365 billion in T-bills looks safe in nominal terms. In real terms it is a slow bleed. T-bills pay what the Fed decides in an era of trillion-dollar deficits and relentless money printing. T-bills yield 4.2% while inflation runs 3.4%. Hence, the pile shrinks in real terms while appearing to earn a decent return. In 2021 and 2022 T-bills paid near zero vs. 7% to 9% inflation, the period when the cash was piled up the fastest. Berkshire's interest income fell in the first half of 2026 with falling short rates. The T-bill portfolio earned $6.1 billion in the first six months on an average balance of ca. $323 billion. That is ca. 3.8% annualized. After corporate tax it is ca. 3.0%, which is below the 3.4% inflation rate.

The world's central banks have reached their own conclusion. They are buying gold at ca. 50 metric tonnes per month in 2026, triple the pre-2022 pace. Goldman Sachs forecasts $4,900 per ounce by year-end vs. today's $4,160. At current prices that runs ca. $80 billion a year; a fifth of Berkshire’s cash pile spent on gold per year. Since the end of 2021, gold has beaten both Treasury bills and inflation by a wide margin.

Buffett and the late Charlie Munger dismissed gold for decades, stating that it produces nothing. Nevertheless, a T-bill in a debasing currency produces nothing either. It just does so with a coupon attached. At minimum, Berkshire should diversify its reserve like the central banks are doing.

Inflation is the slow leak. Complacency is the fast one: While Berkshire waited for a deal that fit their strict criteria, Westinghouse got away. Brookfield and Cameco bought it in 2023 for $8.2 billion including debt. The $50 billion IPO has been filed and they float it at more than six times what they paid. Cash does not just lose purchasing power against goods. It loses purchasing power against the assets Berkshire wants to buy.“