Five publishers give dollar-yen’s 200-day moving average as 157.47, 157.59, 158.43, 158.47 and 158.50. That is a 103-pip band on the single most-quoted level in the pair. For most of today spot sat inside it, which means the question every one of those pages was implicitly answering — are we above or below the 200-day — had no answer at all. At 08:21 UTC it got one: dollar-yen took out 156.62 on a Ministry of Finance remark, extended to 156.51, and was back at 157.04 by 09:38. The band is 103 pips wide. The London range was 119. You can size that comparison yourself.
The five figures, and who owns each one
Minkabu’s technical note at 03:15 UTC today gives the 200-day at 157.47, alongside a 10-day at 157.11, a 21-day at 156.48 and a 100-day at 159.26. Gaitame.com’s analysis at 03:17 UTC names 158.50 and calls it resistance. FXStreet’s dollar-yen forecast of 24 September calls it, verbatim, “the key 200-day Simple Moving Average (SMA), at 158.47.” Two further figures — 157.59, which that same house labelled a 200-period rather than a 200-day, and 158.43 — come from this desk’s own reporting on 24 and 25 September and were not re-read at their sources for this article.
Take the two FXStreet numbers together, because they are the sharpest thing here. One house has published 157.59 and 158.47 for the same instrument four days apart, 88 pips apart, and the smaller figure carries a label that quietly says it is not the same object: a 200-period average on an hourly or four-hour chart is not a 200-day average, and a reader skimming for “the 200” will not notice. This desk has published vendor disagreements on consensus figures, on prints, on option notionals and on a session high. A vendor disagreeing with itself because two of its pages mean different things by the same shorthand is a different fault, and it is the one that survives any amount of cross-checking between publishers.
One figure we looked at and cut: a US analysis page putting the 200-day “near 158.5” with spot at 158.826. It is dated 2 September. It would have been a sixth reader and it would have sat neatly on top of Gaitame’s number, which is exactly why we checked the date. Twenty-six days of price action separate it from anything useful today.
A band only matters while price is inside it — and for three hours today it was
At 07:21 UTC Minkabu had spot at 157.54 in a London range of 157.30 to 157.70. That is seven pips above the lowest published 200-day and 96 pips below the highest. On Minkabu’s number you were above the 200-day by a rounding error. On Gaitame’s you were nearly a big figure below it, testing nothing, with resistance a hundred pips away. Both statements were published this morning, about the same instrument, at the same time, by two houses this desk rates highly and reads every day.
Then the Ministry of Finance spoke. Minkabu reports Vice Minister for International Affairs Mimura saying the government’s message on yen weakness is “very clear” and that he hopes the market takes it straightforwardly. Dollar-yen printed 156.62 at 17:21 JST and 156.78 immediately after; by 18:38 JST the low stood at 156.51 and spot had recovered to 157.04, with euro-yen at 178.68 against a session low of 178.20. That remark and that move are one publisher’s account, read in Japanese, and we are labelling it as such — Mimura’s name and title are independently established, today’s quote is not.
At 156.51 the question resolves: you are below all five 200-days, by between 96 and 199 pips. At 157.04 you are below all five again, by 43 to 146. At 157.54 you are inside the band and the indicator is silent. Three states in three hours, and only one of them is a level the five publishers agree on the sign of.
What “short below the 200-day” committed you to today
Suppose your rule is the commonest one in retail dollar-yen: sell the break of the 200-day, buy the reclaim. The five published values give you five entries spanning 103 pips. The London session’s entire high-to-low travel was 119 pips, and against the Tokyo high of 157.84 from our own earlier reporting the day’s range is 133. So the disagreement between vendors about where your trigger is amounts to 87 percent of the range the London session actually produced, and 77 percent of the day’s.
That is not a rounding problem. On the Minkabu number your short triggered at 157.47 and you rode 96 pips. On the Gaitame number your short triggered 103 pips higher and you were stopped, reversed, or never filled depending on which side of the morning you were on. Same rule, same instrument, same day, opposite outcomes, and the only input that differed is which page you had open.
The uncomfortable conclusion is not that one vendor is wrong. It is that a 200-day moving average is a computed object with free parameters — which close, which data source, which session boundary, whether the average is simple or exponential, and whether a gap-heavy weekend counts as one bar or none — and no publisher we read states any of them. If a level depends on parameters nobody discloses, then the level is a vendor opinion presented in the typographic register of a fact, and a system that keys on it has an undeclared dependency on a third party’s spreadsheet. The honest use is as a zone, and today the zone is 157.47 to 158.50, which is wide enough that saying so out loud tells you the indicator should not be carrying much weight this week.
The euro leg, and the only mechanical claim on price left today
There is 4.5 billion euros expiring at 1.1400 at the 14:00 UTC New York cut, with further strikes listed at 1.1350 and 1.1420 and no notionals given for either. The publisher calls the 1.1400 line “exceptionally large and almost sitting on top of current price” and puts the 100-hour moving average at 1.1404 as a near-term resistance layer. This is single-sourced, and we read it on the SwingFish syndication mirror because the originating page returned a 404 on the exact URL search supplied — the fifth consecutive slot that has happened. Treat the notional accordingly. Note also that the mirror did carry the notional this time, which contradicts our own standing note that expiry tables do not come through it.
Around it: the dollar index at 101.12, up 0.15 percent, having touched 101.199 and then slipped under 101 before recovering, so the advance has paused rather than reversed. Sterling is the strongest leg on the board, through the 1.3250 that capped it in Tokyo and printing 1.3262. Euro-yen made 178.20 on the yen spike. On a day with no tier-one release, a 4.5 billion euro strike five pips from a 100-hour average is the only thing on the board with a mechanical claim on price, and a break through 1.1400 at 14:01 tells you something a break at 12:30 does not. That claim was made in print this morning by another desk here and it is still the right frame; what we would add is that it is a claim about euro dollar, and the yen leg has already demonstrated today that the thing which actually moves the board is an official with a microphone and no calendar entry.
What this does not tell you
We did not compute a 200-day moving average ourselves, and we are not claiming any of the five figures is the right one. Three of the five were read at their sources today and two are carried from this desk’s own earlier work; that is a weaker footing than a same-day five-way read would be, and the band could be narrower or wider than 103 pips if all five were refreshed at once. We do not know any publisher’s close convention or data feed, so we cannot say whether the spread reflects a genuine methodological difference or simply a stale page. The 156.51 low, the 156.62 print and the 157.04 recovery are one publisher’s log, and this desk has previously found three publishers four dollars apart on a session high, so treat any single-publisher extreme as an estimate. Friday’s low is itself disputed at two pips — 156.93 on one house, 156.95 on another — which is a useful reminder of the precision the pair is actually quoted to in practice. And nothing here forecasts direction: a pair that travelled 133 pips and closed the move roughly where it started is a pair with no settled view, and we do not have one either.
Related
- FX events calendar — all times UTC.
- Signals
- EA presets
