The Japanese Market Calendar: Earnings Season, Ex-Dividend Dates and Golden Week

Some weeks in the Japanese sector table are strange for structural reasons rather than economic ones. If you know the calendar, you can discount those weeks instead of building a theory around them. Educational only, not investment advice.

The fiscal year ends in March

Most Japanese companies close their books on 31 March, not 31 December. This is the single most important calendar fact for reading Japanese equities, and it moves three things.

Guidance season is May, not January. Full-year results and the following year's guidance land in early-to-mid May. That is when the largest single-name dispersion of the year occurs, and it is why May weekly tables often show enormous spread between sectors.

Q3 results land in early February, covering October to December. Second-largest cluster.

The last week of March is window-dressing season. Institutional positioning around the fiscal close can push sectors around for reasons unrelated to fundamentals. Discount that week.

Ex-dividend dates cluster twice a year

Japanese companies overwhelmingly pay semiannual dividends with record dates at the end of March and the end of September. Because so many names go ex-dividend on the same two days, the index and the higher-yielding sectors take a visible mechanical drop.

Banks, Electric Power & Gas and Transportation & Logistics carry higher yields and show the effect most clearly. A red week for Banks that includes the late-September ex-date is not a signal. Check the date before you interpret the move.

Golden Week compresses May

Late April into early May contains a run of national holidays, and the exchange closes for several sessions. Two consequences:

• Weekly return windows in that stretch cover fewer trading days, so the numbers are not comparable to a normal week.
• Global news that breaks while Tokyo is shut gets priced in a single catch-up session on reopening, which produces one outsized day that dominates the week.

There is a smaller version of this in mid-August around Obon, when volume thins and moves exaggerate.

What this means for cross-market comparison

The US and Japan do not share a calendar. In February and May, Japan has an earnings-driven dispersion spike that the US does not; in late January and April, the US does and Japan does not. A "divergence" between the two markets in those windows is often just the two reporting seasons being out of phase.

A short checklist before interpreting a Japanese week

1. Did it contain a March or September ex-dividend date?
2. Was it inside the February or May earnings cluster?
3. Did it include Golden Week, Obon, or the fiscal-close week in late March?
4. How many trading sessions did it actually have?

If any of those is yes, treat the sector table as structurally distorted and wait for a clean week.

The short version

March fiscal close, February and May earnings, semiannual ex-dividend dates, and a week-long May holiday. Four calendar facts that explain most of the Japanese weeks that otherwise look inexplicable.