How the BoJ Rate Decision Moves Forex, on Real Days

The BoJ rate decision in forex moved USD/JPY 17 to 73 pips in the 15-minute candle it landed on, across the four decisions on our calendar from April to September 2026, against about 4 pips for an ordinary candle at those hours. Unlike the other central banks it has no fixed minute: the four landed at 22:54, 23:00, 23:11 and 23:19 New York.

All times here are New York and a candle is named by the time it opens. The Bank's decisions fall in New York's evening and Tokyo's lunchtime. The wider subject is in news trading, and the neighbouring decisions are in the RBA page and in what FOMC means in forex.

BoJ rate decision forex: what is it, and when does it come?

The Bank of Japan's Policy Board sets the policy rate, and the Bank publishes its meeting schedule in advance. It publishes dates, and our calendar's minutes differ from meeting to meeting. The decision on our calendar came at 23:00 on 27 April, 23:19 on 15 June, 23:11 on 30 July and 22:54 on 17 September 2026. The monetary policy statement sat at 23:00 on the last three, and on 17 September the press conference was at 02:30.

So a trader cannot set a clock for the minute the way they can for the jobs report. The decision can come at any of several minutes, and the chart has no warning. That is part of why the candle is large against a quiet evening, and why a 15-minute candle is the unit here: the decision falls inside it, not on its first minute.

BoJ rate decision forex: how far did USD/JPY move on four decisions?

The 15-minute candle the decision landed on, in pips.

The median range is 40.0 pips. The median for the same late-evening candles across all days of our sample is 4.5, so a decision candle ran four to sixteen times an ordinary one. That is a larger multiple than the jobs report's, because the evening is so quiet. In pips it is about half: a median 73.8 on the five jobs reports and 40.0 here, on USD/JPY. The count is bankDecisions.BoJ in scripts/seo-counts.mjs, and it says nothing of what the Bank decided or how it was expected.

What did one real day look like? USD/JPY, 17 September 2026

The evening of Thursday 17 September 2026 was calm until the decision. The low of the session, 155.870, had come at 19:00. Japan's consumer prices at 19:30 were medium impact on our calendar, and price climbed slowly, 30 pips in three and a half hours. Then it pressed under the previous day's high, 156.314: the 22:15 candle reached 156.311, 0.3 pips short.

• 22:30. A 5.6-pip candle, 156.134 to 156.190.
• 22:45, the decision (22:54). The candle opened at 156.175 and ran to 156.867, 55 pips through that high, and closed at 156.653. Range 73.3 pips.
• 23:30. A high of 157.145, and the evening went on up to 157.325 at 01:30.
• 02:30, the press conference. The candle opened at 157.116, fell 61 pips to 156.502, rose 20 over its open to 157.313, and closed at 156.792. Range 81.1 pips.
• 05:45. The high of the day, 158.052, 218 pips off the low of 19:00.
• 13:45. The low of the New York morning, 156.659, 139 pips off the high.

The pair ran through the previous high in the decision minute and kept going, then the press conference made an 81-pip two-way candle in the quiet of the London pre-open and price rose again. It was a trend day with a violent candle in the middle of it, and the decision candle did not hold the high. A hunt of the previous day's high would have closed back under it, and nothing did. A day like this is not a reversal, and a trader who read the press conference candle's 61-pip drop as the turn was wrong within three hours.

What did the other three decisions look like?

We have the candle's numbers and not the candles' stories. On 27 April the 23:00 candle ranged 38.9 pips and closed 30.2 down, nearly its whole range, and held neither end of the day. On 15 June the decision came at 23:19 and its 23:15 candle ranged 17.0 pips, the smallest of the four and still nearly four times an ordinary candle. On 30 July the 23:00 candle ranged 41.0 and closed 24.1 up, and held the day's high.

Read as a set, the four candles closed in both directions: three up and one down. Two closed close to their furthest point and two gave back about a third of their run. None of that is a rule, and four candles are too few to say whether the yen's decision candles tend to hold their move or hand some of it back. A decision that arrives at an unannounced minute in the middle of a quiet evening is a size, and the direction is no easier to guess than on a fixed-time release.

Does the BoJ press conference move USD/JPY more than the decision?

On 17 September it did: 81.1 pips against 73.3. That is the only press conference we read candle by candle, so it is one day. The guides say the press conference can matter more than the statement, and that fits this night, but a statement and a press conference are different texts read at different hours. Whether the second is usually the larger is a question for more decisions than four.

What do the three seats see on a BoJ evening?

The Three Seats read 17 September three ways.

• Retail sees a rate decision it cannot time. The candle came with no warning and ran 55 pips through a level that had held all day.
• The market maker, as we read it, sees a quiet book and a level with stops above it. The previous day's high is where the day's buy stops rest, and a minute nobody can set a clock for reaches them in one candle. That is News-Induced Liquidity, without the clock. Nobody outside can see those orders.
• The observer sees two big candles in a trend. One through a level at 22:45, one two-way at 02:30, and a high at 05:45. All can be read from prices, and none says who was behind them.

The time of day matters for who is awake. At 22:54 New York most of the Western market is asleep and Tokyo is at lunch, so the first candle prints into the thinnest part of the day, which is one reading of why an ordinary evening candle is 4.5 pips and the decision candle at the median is about nine times that. It is a reading, since our data has no volume.

What does this not tell you?

It does not tell you what the Bank decided or how it was expected. It is four decisions on one feed, and the 17 September candle was larger than the other three. A break seen afterwards is not a setup, and our own fixed-rule test of fading hunts on a release lost money after costs. Why price spikes, then reverses, after news looks at days when it did turn.

Course 2 of ForexiaPRO, The Chaos Pattern, teaches this read step by step in lesson 2.3.3, NFP, CPI, FOMC: One Read, Three Calendars.