GOVERNANCE WATCH
By InsidEntity Editorial Desk · Sep 18, 2026 · 11 min read
A rating that holds still while a central bank acts is not a rating that has missed something. It is a rating telling you where the pressure is not. Japan is the clearest case this instrument has produced.
Yesterday this publication ran four economies through the Economic Sustainability Rating and showed that multiplying inflation by unemployment reorders them against the inflation headlines. Canada had the calmest consumer prices of the four and the second-heaviest rating, because 6.4% of its workforce was out of work.
Japan sharpens that argument further. Japan’s Statistics Bureau released the August national consumer price index this morning, 18 September. Headline inflation held at 1.9%. Hours later the Bank of Japan raised its policy rate to a level the country has not seen in 31 years. The Economic Sustainability Rating, calculated from an input published the same day, still reads 5, near the best end of the scale.
The rating is not stale. The CPI input was published this morning. The rating simply does not measure the thing the central bank just acted on.
The reading
Economic Sustainability Rating: 5
| Input | Value | Period |
|---|---|---|
| Consumer price inflation | 1.9% | Year to August 2026 |
| Unemployment, seasonally adjusted | 2.4% | July 2026 |
| Product | 4.56 | 1.9 x 2.4 |
| ESR | 5 | Rounded |
The ESR multiplies a country’s inflation rate by its unemployment rate. The scale runs from 1 to 150 and runs the opposite way to this platform’s company instruments: a low ESR is the Benchmark end, a high ESR is the Unsustainable end, where on the Company Risk Rating and the Financial Stability Rating the five-point scale runs the other way. A reader who assumes the scales run in the same direction will read this article backwards.
Japan’s rating has held at 5 across the last three readings. Today’s August CPI release kept headline inflation at 1.9%, while the latest available unemployment rate remains 2.4% for July. The resulting product is 4.56, which rounds to 5.
Set against yesterday’s four:
| Economy | Inflation | Unemployment | Product | ESR |
|---|---|---|---|---|
| Japan | 1.9% | 2.4% | 4.56 | 5 |
| United States | 3.4% | 4.1% | 13.94 | 14 |
| United Kingdom | 3.1% | 4.9% | 15.19 | 15 |
| Canada | 3.0% | 6.4% | 19.20 | 19 |
| India | 4.82% | 5.0% | 24.10 | 24 |
Japan’s 5 is just over one third of the United States’ 14, which was the lightest of the four. It is close to the best reading the instrument can produce, and it has been sitting there while everything around it moved.
The decision the rating did not register
On 18 September 2026 the Bank of Japan raised its policy rate by 25 basis points to 1.25%, effective 24 September, from 1.00%. It is the highest level since 1995. The Board split 7-2, with Toichiro Asada and Ayano Sato dissenting in favour of holding.
The pace matters as much as the level. This hike came three months after the previous one, against roughly six-month intervals earlier in the normalisation cycle the Bank began in March 2024. Whatever else the Board decided, it decided to go faster.
The Bank’s reasoning was forward-looking. Its statement records that underlying CPI inflation “has been approaching 2 percent” and that “there is a risk that it will deviate upward to a level above the price stability target of 2 percent”, with inflation expected to reach levels generally consistent with that target between the second half of fiscal 2026 and fiscal 2027.
On where it sees the pressure, the statement is specific, and it is worth quoting rather than paraphrasing. The year-on-year rate of increase in the producer price index “has continued to be high”, reflecting “the impact of the expansion in AI-related demand, in addition to high crude oil prices and the depreciation of the yen”.
None of those three is an ESR input, and neither is the index they are cited against. The ESR reads consumer prices. The Bank was reading producer prices, a currency, and a forecast.
Asada and Sato both preferred to maintain the existing guideline. Asada’s recorded position was that “it could not necessarily be said that the economic situation was strong”. Sato’s was that “current economic and price developments did not appear to have substantially accelerated”. Headline inflation, the figure the ESR uses, held at 1.9% in the same release. Core inflation, which excludes fresh food, eased to 1.7% in August from 1.8% in July and has now run below the Bank’s 2% target for a seventh consecutive month.
So the central bank tightened on the strength of pressures the rating does not count, against a core price series that was falling and a headline series that held still. The rating stayed at 5, correctly, because the two things it measures did not change.
The same figure, two opposite meanings
Japan’s policy rate is now 1.25% and its inflation rate is 1.9%. Subtract one from the other and the real policy rate is -0.65%. The Bank has reached a 31-year nominal high and money is still cheaper than free in real terms, which is a large part of why the cycle is not finished.
The ESR treats that 1.9% as one of its two inputs. On the methodology’s scale, combined with 2.4% unemployment, it produces a Benchmark reading. The same 1.9%, set against a 1.25% policy rate, is the reason the Bank is still moving.
One figure, two opposite meanings, and the instrument cannot tell them apart because it never sees the rate.
The yen sits upstream of the rating’s own input
The yen weakened on the day its own central bank raised rates, trading around 157 to the dollar, while the 10-year government bond yield fell 4.9 basis points to 2.947%. The 7-2 split signalled a slower path than the headline implied, and the rate differential with the United States has kept the carry trade profitable in spite of the move.
The level is easy to overstate. At around 157 the yen is weak but not at a fresh low: it weakened past 160 in March 2026, its weakest since July 2024, when authorities intervened at roughly 161.95. The accurate framing is sustained multi-year weakness with the 2026 low already behind it, not a new record.
A weak yen makes Japan cheap for inbound visitors and improves exporters’ translated earnings. It also raises the cost of everything Japan imports, which in an economy meeting almost all of its crude requirement from imports means energy, and through energy nearly everything else.
Trace the chain: a weaker yen raises the yen price of imported crude, higher energy costs feed into the consumer price index, and the consumer price index is an ESR input. The exchange rate sits upstream of the rating’s own arithmetic without ever appearing in it. By the time a currency move shows up in Japan’s ESR, the Bank of Japan has already been acting on it for months.
The unemployment input has a structural component
Japan’s unemployment rate fell from 2.5% to 2.4%. Underneath that, the number of unemployed fell by 80,000 to 1.67 million, while the labour force fell by 240,000, three times as fast.
A labour force contracting three times as fast as the number of unemployed puts downward pressure on the unemployment rate independently of the broader cycle. That does not make 2.4% wrong; it is the correctly calculated rate. It means one of the ESR’s two inputs carries a demographic component the instrument cannot separate from cyclical conditions, and a reader working across a table of countries would not know that from the number alone.
Why stability is information
The ESR did what it is defined to do. It measured two variables, and those two variables did not move. What moved was the currency, the price of borrowing, the oil price, and the Bank’s forecast of where inflation will be a year from now. A two-input instrument that started registering those things would no longer be a two-input instrument, and its readings would stop being comparable across the 145 markets the platform covers.
Read alongside a 31-year-high policy rate, a stable Benchmark reading tells an analyst something useful and specific: whatever is driving the central bank has not yet arrived in consumer prices or in the labour market. That is a different message from a rating that is deteriorating, and it points to a different set of questions at the next layer down.
This is the same finding the four-economy piece produced from the other direction. There, the American rating held at 14 while the monthly inflation pace quadrupled, because annual inflation and unemployment were unchanged. Here, Japan’s holds at 5 while the policy rate reaches a generational high. In both cases the instrument is correct and incomplete at once, and knowing which is which is the skill.
The governance question
Every company rating this platform publishes begins from two questions. Who is in the room, and how did they get there. Those questions are asked of audit committees as a matter of routine and almost never asked, in public, of a monetary policy board, even though its decisions reach further than any board’s.
In February 2026 the Takaichi government nominated two academics, Toichiro Asada and Ayano Sato, to the Bank of Japan’s Policy Board. Contemporary reporting characterised both as reflationists and read the appointments as echoing the government’s own policy orientation. Long-term Japanese government bond yields moved on the news at the time. On 18 September the Board divided 7-2, and Asada and Sato were the two votes against.
The appointments therefore form part of the institutional context surrounding the September vote. They do not establish why either member voted as they did. Both recorded economic reasons for opposing the increase, and the dissent itself is the documented fact. The governance observation is narrower: the composition of a monetary-policy board can matter to the interpretation of a policy decision, just as board composition matters when assessing a corporate issuer.
The next layer
The methodology sets out an order. Economic environment first through the ESR, then governance risk through the Company Risk Rating, then company risk through the Health Status Report, then financial risk through the Financial Stability Rating. Macro to micro, because the environment a company operates in shapes what its own numbers mean.
Start at the top and the country reads 5, near the Benchmark end. Stop there and a Japanese subject would appear to need less scrutiny than one domiciled almost anywhere else. But a Japanese exporter and a Japanese domestic retailer have identical ESRs and opposite exposures to a yen at 157: one has just had its competitive position improved and its input costs worsened, the other only the second. A policy rate of 1.25% reprices a leveraged balance sheet and does very little to one carrying net cash.
None of that is visible from the country line. All of it is visible one level down. The ESR tells you what environment you are entering. It does not tell you which company inside that environment warrants further scrutiny, and it was never built to.
This piece does not claim the Bank of Japan was right or wrong to raise; that judgement belongs to people with a mandate. It does not claim Japan’s rating is mismeasured, or that any Board member voted other than according to their own economic judgement. Its claim is narrower: a country-level indicator built on two inputs will sometimes hold perfectly still at the Benchmark end while a central bank acts on pressures those two inputs cannot capture. That is the reason the methodology places the ESR first in a sequence rather than alone, and the reason a stable Benchmark reading is the beginning of the work rather than the end of it.
Independent research. Not financial advice. No allegation of wrongdoing is made against any individual named.
Know more. Risk less. Decide better.
Sources: Bank of Japan, Statement on Monetary Policy, 18 September 2026, read directly; Japan Statistics Bureau, Consumer Price Index for August 2026, released 18 September 2026; Japan Labour Force Survey for July 2026; contemporary reporting on the February 2026 Policy Board nominations; InsidEntity Economic Sustainability Rating; and the companion piece of 18 September 2026 on the United States, United Kingdom, Canada and India.
Note on figures: Inflation and unemployment figures are as published by the relevant national statistical agencies and are the most recent releases as at 18 September 2026. Japan’s inflation and unemployment readings are for different months, August and July respectively, since inflation is reported monthly and unemployment on a lag; the August CPI release used here was published the same day as this piece. The ESR of 5 shown here is computed from those published inputs; the platform’s country page is the authoritative record. Quoted passages from the Bank of Japan are taken from its statement of 18 September 2026. Market levels are same-day and move continuously.