In late September 2026, the Nikkei 225 closed at 66,364.20 on September 25 with a gain of 1.30% on the day, and the advance was broad, with 13 of 17 TOPIX-17 sectors rising. The session fell in the fiscal half-year ex-dividend cluster week: with the semi-annual dividend record date at the end of September only days away, positioning around interim dividend rights was plausibly part of the picture. When a dividend record date can shape trading across an entire national market, it tells you something structural: Japan’s corporate calendar is unusually synchronized. This guide explains that machinery — the March fiscal-year convention, the quarterly kessan reporting waves, the mechanics of TDnet release timing, and the outsized role of company guidance. The single most useful takeaway: in Japan, the reported quarter is usually the sideshow, and the company’s own full-year forecast is the event.
One date rules the market: the March fiscal year
Roughly two-thirds of companies listed in Tokyo end their fiscal year on March 31, mirroring the Japanese government’s April-to-March fiscal year. This is a convention, not a law, but it is sticky, and it synchronizes almost everything downstream: full-year results cluster in a single spring window, interim results cluster in autumn, dividend record dates pile up on March 31 and September 30, and annual shareholder meetings crowd into late June.
The synchronization has two practical consequences. First, earnings season in Japan is a genuine season — a compressed flood of releases rather than a rolling drip. Second, the semi-annual dividend record dates create mechanical index effects: on the ex-dividend day (one business day before the record date), hundreds of stocks tend to open lower by roughly their dividend amount simultaneously, which can shave a visible chunk off the index for reasons that have nothing to do with sentiment. Exceptions to the March convention exist and are worth memorizing for the names you follow: many retailers close their books in February, Fast Retailing (9983) ends its fiscal year in August, and a minority of globally oriented firms use December.
The kessan calendar: four reporting waves a year
Kessan (決算) means the settling of accounts — in market usage, “earnings results.” The document that matters for traders is the kessan tanshin (決算短信), a standardized earnings flash report filed with the exchange. It is the first public look at the numbers, arriving weeks before the audited statutory filings, and it is what algorithms and headlines key off.
For a March fiscal-year company, the four waves look like this:
| Period | Books close | Typical announcement window |
|---|---|---|
| Q4 / full year | March 31 | Late April to mid-May (the biggest wave) |
| Q1 | June 30 | Late July to early August |
| Q2 / interim (chūkan kessan) | September 30 | Late October to mid-November |
| Q3 | December 31 | Late January to mid-February |
The Tokyo Stock Exchange asks companies to report within 45 days of the period end, and encourages 30 days. Because most firms wait until near the deadline, each wave has a sharp peak: in the May and November waves, well over a thousand companies can report within the peak fortnight, with several hundred landing on the single busiest day. On those days, even professional desks triage. One structural note: since 2024, the statutory quarterly securities reports for Q1 and Q3 have been abolished, so for those quarters the exchange-mandated tanshin is essentially the only disclosure — thinner in detail, and even more dependent on the guidance table inside it.
TDnet timing: why results land at 15:00, 15:30, or mid-session
All timely disclosures flow through TDnet (Timely Disclosure network), the electronic system operated by the exchange group. There is no single mandated release hour, but convention matters. Historically, most companies released results at 15:00, the old cash-market close. When the TSE extended the cash session to 15:30 in late 2024, release practice shifted with it: many issuers now publish at 15:30 or later, though conventions vary from company to company. Some still publish at 15:00, which now lands 30 minutes before the close, producing sudden intraday repricings in otherwise quiet tape. Others deliberately release during the lunch break so the afternoon session can absorb the news.
Three practical rules follow. First, on peak kessan days, expect a burst of headlines in the minutes around 15:30 and be careful interpreting closing prices — the market has not yet reacted to most of that day’s releases, so the real repricing happens at the next morning’s open. Second, treat 15:00 as a live risk window during earnings season, not dead time. Third, remember that scheduled results are only part of the flow: standalone disclosures — especially guidance revisions, covered below — can hit TDnet at any hour of the trading day, which is why some of the largest single-stock moves in Japan occur mid-session with no result on the calendar.
Guidance culture: the forecast is the event
The most distinctive feature of Japanese reporting is the gyōseki yosō (業績予想) — the company’s own full-year earnings forecast, published as specific point estimates (revenue, operating profit, net profit, usually the dividend) inside the tanshin. It is an exchange convention rather than a legal requirement, and a small minority of firms decline to give one, but for the vast majority the forecast table is the first thing the market reads. When full-year results land in May, the new fiscal year’s guidance routinely matters more for the stock than the year just reported.
Why guidance runs conservative
Japanese guidance is habitually cautious, for institutional rather than cultural-cliché reasons. Missing your own published forecast forces a formal downward revision — a reputational event management works hard to avoid — while beating it costs nothing and permits a mid-year upgrade. Exporters reinforce this by setting foreign-exchange assumptions several yen more conservative than the prevailing spot rate, embedding a cushion that converts into “upside surprises” if the currency simply stays put. The result is a familiar cadence: guide low in May, revise up alongside the November interim results, and finish the year ahead of plan. Practitioners adjust with the shinchoku-ritsu (進捗率), the progress rate — first-half profit as a percentage of full-year guidance. A company at 60% progress after six months with unchanged guidance is quietly signaling a beat; one at 35% is a downward-revision candidate even if no revision has been announced. Independent benchmarks such as the Toyo Keizai Kaisha Shikihō quarterly handbook, whose analysts publish their own forecasts, exist precisely because company guidance cannot be taken at face value.
Guidance revisions: where the biggest single-day moves live
A gyōseki shūsei (業績修正) is a formal earnings-forecast revision. Exchange timely-disclosure rules require one promptly whenever a company’s internal expectations diverge from published guidance by roughly 10% on revenue or roughly 30% on the profit lines. Two consequences make revisions the most explosive items on TDnet. First, revisions that trip the mandatory threshold are large by construction — a change of around 30% or more on the profit lines — though companies can and do voluntarily disclose smaller revisions as well. Second, they are unscheduled — the rule says “promptly,” so revisions drop on random days and at random hours, often mid-session, when no one is braced for them. This is why seasoned observers of the Japanese market say that mid-year revisions, not the headline quarterly results, drive the largest single-day moves. Revisions do cluster somewhat ahead of the Q2 and Q3 waves, as companies true up before reporting, so the weeks just before each peak deserve as much attention as the peak itself.
Worked example: placing Toyota on the calendar
Take one verified real-world data point: Toyota Motor (7203) closed at ¥2,989.50 on September 25, 2026. Here is how a calendar-aware reader interprets that date and price together, step by step:
- Identify the fiscal position. Toyota is a March-year company, so in late September 2026 it is days from closing its interim (first-half) period on September 30 of the fiscal year ending March 2027.
- Flag the dividend mechanics. September 30 is also the interim dividend record date. The stock trades ex-dividend one business day earlier; all else equal, it would be expected to open lower by roughly the interim dividend on that day, though the actual move also reflects whatever else the market is pricing at the time — not a sell signal, just arithmetic to keep in mind. Any dividend-rights positioning supporting the broad market into that week is the mirror image of the same mechanics.
- Mark the next kessan. Interim results will arrive in the late-October-to-mid-November wave, via TDnet.
- Know what to read first. When that tanshin lands, the priority order is: any change to full-year guidance, then the forex assumption behind it, then the progress rate against guidance — and only then the quarter’s headline profit.
Common mistakes
- Reading a “profit down 20%” headline as bearish without checking guidance — if full-year guidance was maintained or raised, the market often looks straight through a weak quarter.
- Ignoring the progress rate, which frequently telegraphs revisions weeks before they are announced.
- Forgetting the ex-dividend clusters in late September and late March, and misreading a mechanical index dip as risk-off.
- Treating a 15:30 release as “after hours, no rush” — the repricing concentrates in the next morning’s opening auction, and gaps there can be large.
- Assuming the earnings calendar is the full risk calendar — standalone revisions can hit any stock on any trading day.
FAQ
When is earnings season in Japan?
Four times a year, keyed to the dominant March fiscal year: late April to mid-May for full-year results, late July to early August for Q1, late October to mid-November for the interim, and late January to mid-February for Q3. The May and November waves are the largest.
What time of day do Japanese companies report?
There is no mandated hour. Since the cash close moved to 15:30 in late 2024, many releases hit TDnet at or just after 15:30, but a meaningful number still publish at 15:00 (now mid-session) or during the lunch break, and unscheduled guidance revisions can appear at any time of day.
Why do Japanese companies guide so conservatively?
Because the disclosure regime penalizes misses asymmetrically: falling short of published guidance forces a formal downward revision, while beating it allows a flattering mid-year upgrade. Conservative currency assumptions by exporters amplify the pattern. The habitual result is guide-low-in-May, raise-in-November.
Where can I find the announcement schedule and the disclosures themselves?
The exchange publishes forthcoming earnings announcement dates, and all tanshin and revision filings appear on TDnet the moment they are released; company IR pages mirror them, usually with English versions following the Japanese original.
Sources
- Japan Exchange Group — listing rules, timely disclosure and market statistics: www.jpx.co.jp/english
- Nikkei Indexes — official index data and methodology: 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.