Three months in, the framework is doing its job. The real wage recovery is intact, the dislocation between sentiment and spending is now demonstrated across two consecutive months of SSS data, and the inbound rotation continues to print records. What changed in May is the source of the dominant risk. The Iran shock that broke confidence in April has now broken the BoJ consensus and the fiscal stance with it. The government is preparing emergency energy subsidies for summer; three BoJ board members have already voted to hike to 1.0%. The thesis still holds. The risk no longer sits where it did last month.

The April data confirmed the dislocation we identified last month. Real cash earnings printed +1.0% YoY in March — the third consecutive positive month, decelerating from +1.9% in February but still positive. The relevant figure inside the release is base pay: nominal scheduled earnings rose +3.2%, the fastest pace in 33 years. That is not a cyclical print; it is structural transmission of Shunto into the wage bill.

The April SSS releases removed any remaining doubt about what this means at the till. Fast Retailing posted +10.2% domestic SSS — an acceleration from the +9.2% March print — with traffic at +6.9% and ticket at +3.1%. Three consecutive months of weather-driven full-price sell-through. PPIH printed +5.2% SSS with traffic still positive at +1.4%, now the third consecutive month with both lines green. Zensho continued to take footfall. Skylark printed strong. And on the other side of the ledger, Seven-Eleven Japan's domestic SSS came in at −0.4% — the headline number masking a traffic decline of −2.5% offset by a ticket increase of +2.2%. The price-hike runway in undifferentiated convenience that we flagged in March has now run out: nominal growth has gone flat.

The performance dispersion is no longer cyclical. Across three months and three different macro environments — the calm of March, the sentiment shock of April, and the institutional turbulence of May — the same operators have been printing positive traffic. The same operators have been losing it. Capital allocators can stop treating "Japanese retail" as a single block.

Inbound continued to surprise to the upside. March arrivals printed 3.62 million — a new all-time monthly record — with the first quarter clearing 10 million for the first time. Chinese arrivals are still down 55.9%, and it no longer matters: Korean, Taiwanese, and Western flows have absorbed the gap with room to spare. The structural rotation we called in March is now a settled fact.

What did not hold up is the surface inflation read. Tokyo Core-Core fell to +1.5% in April, the slowest pace since 2022. The number looks like deflation. It is not. The reading is an artefact of the Tokyo metropolitan government's decision to make high school education free citywide — a policy change that mechanically subtracted roughly 0.5 percentage points from the index. National Core-Core remains at +2.4%, services inflation is sticky, and the underlying pressure is intact.

Last month we wrote that the constructive view rested on the BoJ allowing the wage-price loop to run, and that an accelerated rate path was the asymmetric risk we were sized against. That risk materialized faster than we expected.

The April 27–28 BoJ meeting concluded with a 6–3 vote to hold the policy rate at 0.75%. Three members voted for an immediate hike to 1.0%. The Summary of Opinions released in May made the rationale explicit: the dissenters are responding to imported inflation and yen weakness, not to the wage cycle. This is the worst possible configuration for the domestic consumer thesis. A wage-driven hike cycle would be slow and orderly; a yen-defense hike cycle is reactive, faster, and front-loaded.

The fiscal side is moving in the opposite direction. Prime Minister Takaichi's administration, which had been resistant to additional spending, is now actively drafting a supplementary budget — reportedly above ¥1 trillion — to reinstate energy subsidies for the summer. Electricity and gas subsidy funds are running out exactly as oil settles between $104 and $117 and the yen remains weak. The subsidies that have been artificially holding headline CPI below target are about to be restored, then will need to be unwound again. The political logic is simple: a government cannot watch household energy bills rise into the summer Upper House cycle. The economic logic is messier.

The combination is the dominant risk for the next 90 days. The BoJ is being pulled toward a defensive hike to halt yen depreciation. The government is being pulled toward additional fiscal stimulus to ease the same imported inflation. Each move undermines the other. The market has not yet priced the cross. If the June meeting brings a hike or a clear signal of one, the yen rallies, the inbound math compresses, the wealth effect supporting domestic premium spending unwinds. If the supplementary budget passes without a BoJ response, the yen stays weak, imported inflation continues, and the next real wage print risks going negative.

This is not a sentiment shock. The framework cannot trade through it the way it traded through April's CCI collapse, because the channel of damage runs through the cost of capital and the currency, not through household psychology.

The cleaner read on the consumer remains in the company-level data, and the divergence widened materially in April.

PPIH continues to be the cleanest expression of the framework — three consecutive months of positive traffic and positive ticket, capturing both the down-traded domestic shopper and the rotated inbound demographic on the same footprint. Fast Retailing's +6.9% traffic growth in April is the strongest single-month read in the dataset. Zensho and Skylark are converting the Shunto wage gains into casual dining footfall with no resistance to menu pricing. These operators do not look like they are running in the macro environment the surveys are describing.

Seven-Eleven Japan is the diagnostic in the other direction. Two consecutive months of negative traffic, with nominal growth now flat (−0.4% in April). The model has reached the end of the price-hike runway we flagged in March. This matters less for Seven & i's stock — which trades on US 7-Eleven and the activist overhang — than it does for the broader reading on undifferentiated convenience formats. The footprint is too expensive for the daily ticket the consumer is willing to pay, and the alternative formats (PPIH for hard goods, casual dining for prepared food) are absorbing the volume.

The piece that did not exist in last month's note is sensitivity to the policy cross. The names that benefit most from the real wage thesis — discount retail, agile apparel, casual dining — are also the names with the least direct exposure to a BoJ hike. They sell to households whose income is in the payslip, not in the asset price. The names exposed to the wealth effect (urban department stores, prestige beauty, financed durables) face the opposite combination: domestic confidence is already weak, and the BoJ risk is now stacked on top. The hospitality and inbound names sit in the middle — strong fundamentals, direct exposure to a yen rally if the BoJ moves.