Japan's Rate Reset Is the Real Story — and Sumitomo Mitsui Is Still Priced Like the Old One

Three central banks decide on rates over the next couple of weeks, and the headlines are treating the cluster as one giant, scary macro event. It isn't. Two of the three are noise you cannot do anything with. The third — the Bank of Japan — is different, and it is the one a U.S. investor can actually follow to a company's cash flow, because it is ending a decade of near-zero rates that quietly crushed Japanese bank profits.

Start where the market is, because that is where the fog lives. The Federal Reserve meets Sept. 15–16. Traders now price roughly a coin flip on a quarter-point hike — CME FedWatch put it near 56% (per Yahoo Finance) after Chairman Kevin Warsh's hawkish Jackson Hole speech. But a Reuters poll of economists still has a majority expecting the Fed to hold rates steady through the year (Reuters). That gap is not a signal; it is a coin flip you cannot convert into a position. Norway's Norges Bank, the third in the cluster, is a rounding error for a U.S. portfolio. Each of these, in its own way, is the "which bank holds when" game — fog.

Japan is not a coin flip. The BOJ has already hiked five times, and Governor Kazuo Ueda has said the board will debate the economy and price risks again at this month's meeting (Ueda, via Reuters). The direction is the point: one forecast projects the policy rate at 1.75% by April 2027 (Oxford Economics). More important than any single date, this is a regime change, not a tactical wobble — the same kind of expectations reset that turns a long-suffering business into a surprise machine.

Why a Tokyo hike shows up in a bank's cash

A bank earns the spread between what its loans yield and what it pays on deposits. When rates rise, loans reprice faster than deposits, so the spread widens before costs catch up. At Sumitomo Mitsui (SMFG), the average yield on loans rose 32 basis points to 1.34% while interest paid on deposits rose only 14 basis points — widening the lending spread to 1.14% (International Banker). Japan's megabank net interest margins are projected to climb from 0.85% toward 1.0% (S&P Global). That does not sound like much until you remember how large those yen books are.

The numbers prove the mechanism is working, not just promised. For the fiscal year ended March 2026, SMFG booked a record ¥1.58 trillion in net profit, up 34.4% (SEC filing), with return on equity rising from 8% to 10.4% and fourth-quarter profit more than tripling year over year (Reuters). Japan's three megabanks together earned roughly ¥5.26 trillion ($32 billion), up about a third. And management is guiding to another record — ¥1.7 trillion for the year ending March 2027 — while announcing a ¥180 billion buyback, higher dividends and a stock split.

Here is the honest caveat, because it matters. A bank does not publish clean free cash flow the way an industrial company does, so I cannot hand you the free-cash-flow bridge I normally lean on. The cash-return anchors here are the buyback and the rising dividend — real money flowing back to shareholders — and the net interest income behind them. That is a weaker anchor than free cash flow, and I treat it with the uncertainty it deserves.

Still priced for the old bank

The reason the trade still has room is that the market spent a decade pricing the old story and has only partly let go. Japan's banks traded for years at low multiples of book with mid-single-digit returns — a case for the "uninvestable" label. SMFG's American shares still sell for about 1.7 times book value and roughly 16 times forward earnings, with a dividend yield above 2%. It is the laggard of the three megabank ADRs, up about 36% this year while Mizuho is up closer to 49%. The re-rating has started; it is not finished. The market is still pricing the old risk profile — a low-return ceiling — while the operating setup is already getting cleaner.

The bet breaks if the engine stalls. The whole case rests on the BOJ keeping up its normalization, margins widening faster than deposit costs, and those record numbers not being swamped by credit losses. Watch for the tripwire in three forms: the BOJ abandons the path toward 1.75%, depositors finally demand better rates and eat the spread, or higher refinancing costs push a jump in losses among Japan's weaker corporate borrowers. If the record becomes the last record, cut without ego. This is no longer a fallen knife, so it is not a bargain you buy and forget — it is a re-rating that is real but incomplete, and it pays to follow the spread, not the calendar.