The most useful thing to know about the Bank of Japan’s quarterly Outlook Report is this: markets rarely trade the rate decision, which is usually priced in advance — they trade the revisions to the median CPI forecast, the tilt of the risk-balance language, and the names attached to any dissenting votes. A hold accompanied by an upgraded inflation forecast and “risks tilted to the upside” is a hawkish event; a hike paired with a downgraded forecast can trade as a dovish one. This guide is being read against a live backdrop — Japanese bank shares were among the market leaders on August 12, 2026, with the Banks sector ETF proxy up 1.56% and Sumitomo Mitsui Financial Group gaining 2.91% among tracked large caps, while USD/JPY traded near 159 — but everything below is structural and will read the same at any point in the cycle.
What the Outlook Report is and when it appears
The Policy Board (seisaku iinkai — the BOJ’s nine-member decision-making body: the Governor, two Deputy Governors and six other members) holds eight scheduled Monetary Policy Meetings a year. At four of them — typically January, April, July and October — it publishes the Outlook for Economic Activity and Prices, known in Japanese as the Tenbo Repōto (“outlook report”). The summary section, called “The Bank’s View,” is released alongside the policy statement on decision day; the full text, including the risk-balance charts, follows shortly after.
The report contains three things worth a practitioner’s time: a table of board members’ forecasts for real GDP and consumer prices over roughly a three-fiscal-year horizon; a written assessment of upside and downside risks; and charts showing how each individual member sees the risks around their own forecast. The four non-Outlook meetings produce only a statement, which is why the January/April/July/October meetings carry disproportionate event risk.
Reading the median forecast table
The forecast table is the report’s centerpiece. Three conventions matter:
- Two inflation rows. Japan’s headline policy-relevant measure is CPI excluding fresh food — what Tokyo desks call “core CPI” (note this differs from the US definition). The report also shows CPI excluding fresh food and energy, the “core-core” measure, which strips out imported-energy noise and is the better read on underlying, domestically generated inflation.
- Medians, not a house view. The headline figures are the medians of the nine members’ individual forecasts. Brackets typically show the range of the majority of forecasts, excluding the highest and lowest — a wide bracket signals genuine disagreement on the board even when the median looks unremarkable.
- Fiscal years, not calendar years. Japan’s fiscal year runs April to March. The terminal-year forecast — the fiscal year two ahead — is the one to compare against the BOJ’s 2% price-stability target, because it approximates where the board thinks inflation settles once transitory effects wash out.
The single most common retail mistake is reading the level of the forecast rather than the revision. A terminal-year core CPI forecast of around 2% is not news if it stood there last quarter. A revision of even 0.1–0.2 percentage points, in either direction, is the tradeable information — it tells you the board’s conviction about hitting or overshooting the target has shifted. Always read the table side by side with the previous quarter’s table.
Decoding the risk-balance language
Each forecast carries a risk assessment, and the phrasing follows a small, stable vocabulary: risks are “generally balanced,” “tilted to the upside” or “tilted to the downside,” assessed separately for growth and prices and often separately by fiscal year. This language is deliberate and migrates slowly, which is exactly why changes in it are informative. The typical sequence ahead of a policy shift is: balanced risks → risks tilted (say, to the upside) → forecast revised in that direction → policy action. Language that moves one notch is a forward signal even when the numbers do not move at all.
The full report’s risk-balance charts add resolution: each of the nine members indicates whether they see the risks around their own forecast as skewed up, down or balanced. Counting those markers quarter over quarter tells you whether hawkish (or dovish) conviction is broadening across the board before it ever shows up in a vote. A practical rule: a median forecast at 2% with upside-tilted risks is a more hawkish configuration than a forecast slightly above 2% with balanced risks, because it implies the board thinks its own number is more likely too low than too high.
Vote splits and dissents as forward signals
Policy decisions are made by majority vote of the nine members, and the vote count — with dissenting members named and their preferred alternative described — is published in the statement on decision day. Three layers of disclosure follow, each with a lag: the Summary of Opinions (unattributed viewpoints, roughly ten days after the meeting), the full Minutes (released after the subsequent meeting), and verbatim transcripts a decade later. For trading purposes, the same-day vote count and the Summary of Opinions are what matter.
Dissents are the board’s pressure gauge. A lone member voting for an immediate hike while the majority holds is historically one of the more reliable forward signals in BOJ-watching: it establishes that the internal debate has moved from “whether” to “when,” and majority shifts have often followed within a few meetings. Track the trajectory, not the snapshot — a vote moving from 9–0 to 8–1 to 7–2 across consecutive meetings is a countdown, whereas a stable lone dissenter with no sympathizers appearing in the Summary of Opinions is closer to noise. Note also the direction: a dissent against a hike that passes tells you the board’s dovish wing is alive, which caps how fast the majority can move.
How markets trade the report: forecasts versus the decision
Because the rate decision itself is usually well telegraphed through media channels and market pricing, the reaction function on Outlook days runs mainly through three instruments:
- The JGB curve. The 2-year yield is the cleanest proxy for the expected policy path; the super-long end responds more to term premium and supply. A hawkish forecast revision shows up as front-end-led flattening pressure; a long-end-led move is usually about something else.
- Bank equities. Higher policy rates widen deposit-lending margins, so banks are the equity market’s policy-expectations barometer. Exporters cut the other way via the yen.
- The yen. Upgraded inflation forecasts and upside-risk language are yen-supportive; downgrades weigh on it. With USD/JPY trading near 159 (159.16 at the 2026-08-11 close), the FX channel feeds straight back into the CPI forecast via import prices — one reason risk language and the currency can become self-referential.
Worked example: reading the curve for the priced-in policy path
Here is an actual figure and how to read it step by step. As of August 10, 2026, the Ministry of Finance’s official constant-maturity JGB yields stood at: 2-year 1.618%, 5-year 2.101%, 10-year 2.815%, 30-year 3.957%.
- Start with the 2-year. It approximates the average overnight policy rate the market expects over the next two years. Its gap above the prevailing policy rate is, roughly, the amount of further tightening priced in — compare it against the policy rate on any given day to gauge that cushion.
- Check which end is moving. On that day the 2-year rose 0.7bp while the 30-year rose 3.2bp and the 40-year 4.0bp. A long-end-led move like this reads as term premium and supply dynamics, not a repricing of the BOJ’s path. Had the 2-year led, the correct read would be shifting policy expectations — the kind of move an Outlook forecast revision produces.
- Cross-check with bank stocks. In the TOPIX-17 ETF-proxy data for August 12, 2026 (Nomura NEXT FUNDS closes, which may deviate slightly from the official sector indices), Banks were among the leaders at +1.56%, alongside Energy Resources at +1.57%, with Banks up about 3.6% over five sessions. Banks outperforming while front-end yields grind higher is the equity market confirming the rates market’s hawkish read; banks lagging into rising long yields would flag a term-premium story instead.
Run this same three-step check on any Outlook Report afternoon and you can separate “the market repriced the BOJ” from “the bond market had its own problem” within minutes of the release.
Common mistakes to avoid
- Reading forecast levels instead of quarter-over-quarter revisions.
- Ignoring the core versus core-core split — energy swings can push the headline core forecast around while the underlying trend is unchanged, and the board says as much in the risk text.
- Treating an unchanged forecast table plus a one-notch shift in risk language as “no news.” The language shift is the news.
- Confusing the same-day vote disclosure with the Summary of Opinions — only the former names names; the latter tells you whether a dissenter has company.
- Trading the decision headline before the forecast table is out. On Outlook days, the second headline is usually the one that moves the yen.
FAQ
Where and when is the Outlook Report published?
On the BOJ’s website (English section included) on the decision day of the January, April, July and October Monetary Policy Meetings, with “The Bank’s View” summary accompanying the policy statement and the full text — including risk-balance charts — following shortly after.
Are the forecasts a policy commitment?
No. They are the aggregated individual forecasts of nine members, and policy is decided meeting by meeting. But a terminal-year CPI forecast at or above 2% combined with upside-tilted risk language is the standard justification framework the board uses when it tightens, so the table functions as a conditional roadmap even without commitment.
Do dissenting votes reliably predict the next move?
Not mechanically, but the base rate is good enough to respect. A lone dissent for an immediate hike has historically preceded majority moves by a few meetings — provided the subsequent Summary of Opinions shows other members drifting toward the dissenter’s logic. A persistent lone dissent with no spreading support is a weaker signal.
Why do bank stocks react more to the forecasts than to the decision?
The decision is usually priced before the meeting; the forecast revisions and risk language reset expectations for the path of rates over the next one to two years, which is what drives bank margin assumptions. That is why bank-sector performance on Outlook days is a useful real-time check on how hawkishly the market read the report.
Sources
- Bank of Japan — Outlook for Economic Activity and Prices, statements, Summary of Opinions and Minutes: www.boj.or.jp/en
- Ministry of Finance Japan — official JGB yield data: www.mof.go.jp/english
- JPX statistics — sector and trading data: www.jpx.co.jp/english
- Nikkei Indexes — index and volatility data: indexes.nikkei.co.jp/en
Disclaimer: This is an information and analysis publication, not investment advice. See our Methodology for data sources, standards, and our corrections policy.