He discussed among other things his investments in 2020 in the five major Japanese conglomerates - Mitsubishi, Sumitomo, Mitsui, Itochu and Marubeni - usually referred to as "trading houses".

These five groups and their thousands of subsidiaries around the world are characterized by hyper-complex, even indecipherable structures - which does not seem to have put Buffett off. What they have in common is that they benefit directly from high commodity prices.

They serve as suppliers to Japanese industry, which is entirely dependent on imports. By pooling their purchases - hence the title of "trading houses" - and by taking foreign holdings, they secure the flow of commodities and ensure advantageous pricing conditions.

Publicly, Buffett justified his investments in the five groups in the classic way: at the time, they were quoting x7 their profits, i.e. a return on earnings of 14% in a context of long rates at less than 1%. So it made a lot of sense to borrow in yen to finance these acquisitions.

But there may be something else under the surface. Buffett, notwithstanding the good-father approach he usually preaches, is also accustomed to idiosyncratic bets on commodities. Fanatics of the Oracle of Omaha will think of his cocoa or silver trades, or his recent investments in Chevron and Occidental Petroleum.

The Japanese operations may well be part of the same trick. On paper, it is an extraordinarily well-sequenced "short bonds/long commodities" trade: rates in Japan have quadrupled in the last 18 months, while commodity prices have rebounded considerably since the lows of the pandemic.

It is true that the Japanese central bank indicated a few days ago that it was not planning to raise its key rate, but it is certain that it will not go any lower in the current inflationary context - on this side, optionality is almost guaranteed.

In parallel, on the commodities side, there are multiple signs that a new "super-cycle" is beginning after ten years of stress and under-investment.