Kioxia's Profit Jump, Marubeni's 15% Target and the 5% Yield Squeeze
Japan's Kioxia expects net profit of 1.27 trillion yen ($7.91 billion) for the July-September quarter, up from 40.6 billion yen.
The forecast was published on July 31, 2026. Nikkei Asia
Marubeni is aiming for a 15% return on equity by putting its cash to work, which would lead Japan's trading houses. ROE is net income divided by average equity, or profit per yen of shareholder funds.
Its ROE has fallen for two straight years from 14.9% in the year ended March 2024. Nikkei Asia
AI stocks fell as calls for a global slowdown grew and the U.S. 10-year yield hit 5%. Yield is the annual return for lending to the U.S. government for 10 years, a benchmark for mortgages and business borrowing.
Rising oil prices added to expectations that the Federal Reserve could raise interest rates. Nikkei Asia
The broader context here is that the 31-fold jump is mostly arithmetic. Starting from 40.6 billion yen makes a strong quarter look enormous, like measuring growth from a low base. For savers and investors, what matters is lasting earning power through the cycle. That means checking product mix, price versus volume, inventory values, and whether profit turns into cash. Accrual profit can run ahead of operating cash flow when money is tied up in working capital, and the tax line can shift the result. Normalize the base before stretching one quarter into a full year.
In my view, the Marubeni target is about the denominator as much as the numerator. Dividends and buybacks shrink equity, while fresh investment aims for higher marginal returns. Both can lift ROE. More leverage and one-off gains can also lift it without lasting strength. Durable gain needs repeatable earnings per unit of equity, after impairments and cyclical provisions. After two years down from 14.9%, 15% is a test of capital allocation, not just earnings recovery.
What this means for markets is that the link from yields and oil to growth stocks is mechanical. A 5% 10-year discounts distant cash flows harder, which hurts AI stocks priced on long-term growth. Rate expectations pass through short rates, term premium and credit spreads. Oil lifts input costs and inflation expectations together, squeezing real margins while lifting discount rates. De-risking then widens spreads, tighter credit slows activity, and slower activity confirms the slowdown worry.
The risk to watch is that errors cut both ways. With high operating leverage, guides can overshoot up or down. ROE can improve on paper while economic value added lags. Yield-led selloffs can reverse fast if policy expectations settle. These figures are the starting point. Cash flows, payouts and forward rates will decide whether they hold.
