Japan’s Business Recovery Splits by Company Size as Large Firms Turn Positive

Japan’s latest official corporate survey describes two recoveries happening at different speeds. In the July–September 2026 Business Outlook Survey released by the Cabinet Office and Ministry of Finance on September 11, the business-conditions BSI for large corporations rose to +5.3. The equivalent reading for mid-sized corporations was +0.8. For the survey’s smallest capital group, it was still -10.9.^[1]

That gap is the headline, but the movement underneath it matters just as much. Large-company sentiment had been -0.5 in April–June. The smaller-company reading had been a much weaker -17.6. Both groups improved. Only one crossed zero. The quarter therefore looks less like a simple split between winners and losers than a test of how quickly different kinds of companies can absorb the same pressures.^{[1] [3]}

Positive does not mean “booming”

The Business Survey Index used here is easy to misread. It does not ask whether conditions are simply good or bad. It asks whether a company’s business conditions are rising, unchanged or falling compared with the previous quarter. The BSI subtracts the share reporting “down” from the share reporting “up.” A +5.3 reading therefore means the weighted share seeing improvement exceeded the share seeing deterioration by 5.3 percentage points. A -10.9 reading means the reverse.^{[1] [2]}

This is different from the Bank of Japan’s Tankan diffusion index, which is designed to describe the level of business conditions—good versus poor—at a point in time. The Ministry of Finance itself warns users to distinguish the two concepts. That distinction is essential to interpreting the current result: the smaller-company group remains negative, but its reading is materially less negative than three months ago.^[2]

Manufacturing did the heavy lifting for large firms

The rebound among large corporations was strongest in manufacturing. The large-manufacturer BSI jumped from -1.8 to +7.6, while large non-manufacturers moved from essentially flat to +4.2. The government identified production machinery at +23.4 and information and communications equipment at +19.9 as major positive contributors. Food manufacturing, at -6.3, and petroleum and coal products, at -14.3, pulled the other way.^[1]

That pattern fits a separate Reuters corporate poll conducted around the same time, in which large Japanese manufacturers reported strong semiconductor and data-center-related demand, especially in electronics. The surveys should not be conflated: the official Business Outlook Survey does not itself say that AI or data centers caused the +7.6 manufacturing result. But the parallel evidence helps explain why machinery and electronics-linked industries are currently doing more of the lifting than some consumer- and energy-sensitive sectors.^[7]

The view of Japan is darker than the view from the company

The sharpest size divide appears when companies are asked about the domestic economy rather than their own business. The domestic-economic-conditions BSI stood at +2.9 for large corporations, -4.3 for mid-sized corporations and -19.2 for the smaller-capital group. In other words, many companies that can see improvement in their own operations remain cautious about Japan as a whole.^[1]

The outlook does not erase that divide. The smaller-company business-conditions BSI is forecast to improve to -2.8 in October–December, then slip to -6.4 in January–March 2027. Large corporations expect readings of +6.0 and +4.6 over those same quarters. Forecasts are not outcomes, but the sign difference shows that corporate Japan itself does not yet expect the recovery to become uniform.^[1]

Labor scarcity is universal. The cost of solving it is not.

One problem spans every size category: not enough people. The survey’s employment BSI—calculated as the share reporting an insufficient workforce minus the share reporting an excessive one—was 26.7 for large corporations, 34.3 for mid-sized companies and 27.8 for the smaller-company group at the end of September. For large companies, the shortage balance has remained positive for 61 consecutive quarters since September 2011.^[1]

But the management consequences differ. Among smaller companies, the most frequently cited effect of labor shortages was higher personnel costs from wage increases, selected by 47.1% of respondents in that size group. Increased workload and working hours followed at 42.9%, then stalled skills transfer and employee development at 30.8%. Large companies put heavier workloads and longer hours first at 62.8%, followed by wage-driven personnel costs at 41.7% and higher recruiting costs at 41.5%.^[1]

Smaller firms are not simply sitting out the investment cycle

The survey also complicates the idea that smaller companies are too weak to invest. All three size groups reported a net shortage of production and sales equipment: the equipment BSI was 4.4 for large companies, 5.8 for mid-sized companies and 9.7 for the smaller-company group. Across all company sizes, planned FY2026 capital spending—including software and excluding land purchases—was revised to 11.0% above the previous year.^[1]

What companies want to buy differs. Software was the most frequently selected investment target among large corporations, at 55.7%. Smaller corporations put tools, instruments and fixtures first at 44.9%, followed by software at 42.1% and information equipment at 40.2%. The contrast is useful: “investment” can mean a large digital platform in one boardroom and a machine, instrument or labor-saving device on a smaller company’s shop floor.^[1]

Profit forecasts improved sharply—but that is not the same as equal relief

At the aggregate level, the survey became more optimistic about fiscal 2026. Sales are now expected to rise 4.1% from the previous year, while ordinary profits are projected to increase 2.3%. In the June survey, the comparable projections were +3.3% for sales and -2.4% for ordinary profits. Information and communications equipment manufacturing was one of the largest positive contributors to the new profit estimate, with a 29.8% full-year increase in the government’s industry table.^{[1] [3]}

Those sales, profit and capital-spending figures are published on an all-size basis. They should not be read as proof that profitability improved equally for large and smaller companies. In fact, the size gap in the BSI is a reminder that an aggregate improvement can coexist with very different operating margins, bargaining power and staffing costs.

A 2004 survey with older roots—and a very modern problem

The Business Outlook Survey has been conducted jointly by the Cabinet Office and Ministry of Finance since April–June 2004. Its predecessors include the Ministry of Finance’s business outlook work dating to fiscal 1983 and a corporate trends survey dating to fiscal 1984. The current survey covers corporations capitalized at ¥10 million or more, with special thresholds for utilities and finance. For the latest round, 14,448 corporations were sampled and 11,244 responded, a 77.8% response rate. The survey reference date was August 15.^{[1] [2]}

The size labels are the survey’s own statistical categories. “Large corporations” means capital of ¥1 billion or more; “mid-sized” means ¥100 million to less than ¥1 billion; and the survey’s “small and medium-sized corporations” category means ¥10 million to less than ¥100 million. Readers should not automatically substitute other legal or policy definitions of an SME. In the survey’s estimated population, 843,001 of 877,075 corporations fall into that smallest capital category—a reminder of why the experience of smaller firms cannot be treated as a side note.^[1]

The long-running divide has become harder to ignore in the inflation-and-wage era. Japan’s 2025 SME White Paper found that smaller firms generally had weaker price pass-through power than large companies and linked inadequate pass-through to weaker productivity and profit dynamics. In his 2026 New Year statement, the Small and Medium Enterprise Agency commissioner said roughly 60% of SMEs were confronting labor shortages and put price pass-through, growth investment and productivity improvement at the center of policy. That policy agenda lines up closely with the pressures visible in the September survey.^{[5] [6]}

What would make this a broader recovery?

The next Business Outlook Survey is scheduled for December 10. The easiest headline will be whether the smaller-company BSI finally crosses zero. The more important questions are harder: Can smaller firms raise selling prices enough to fund wage increases? Can labor-saving investment relieve staff shortages rather than merely add another financing burden? Does the deeply negative view of domestic conditions begin to improve? And do stronger sales and profit expectations reach beyond the industries currently benefiting most from machinery, electronics and capital spending?^[1]

Japan’s corporate recovery is not a story of large firms thriving while every smaller business declines. The smaller-company index improved substantially this quarter. But the ability to convert demand into margins, margins into wages, and wages into a stable workforce is not evenly distributed. The latest survey’s +5.3 and -10.9 are therefore best read not as two verdicts on Japan, but as two speeds inside the same economy.