Demo data

Year-end 2026 outlook

USD/JPY forecast for 31 Dec 2026

Fresh

Link to this forecast: https://futurenotset.com/forex/usd-jpy/forecast/2026-12-31/. This address always means 31 Dec 2026.

Reference rate
152.40 Japanese yen per US dollar · ECB daily reference rate · frozenJPY per US dollar · frozen
Dates and timesHide dates and times
Rate observed
Friday, 30 Oct 2026 · ECB fixing, about 16:00 CET
Forecast published
31 Oct 2026, 06:12 UTC · rate 15 h old
Target
31 Dec 2026 · settles on the ECB rate for Thursday, 31 Dec

Reference rate from 30 Oct. The ECB publishes no reference rates at weekends or on TARGET holidays. The outcome uses the ECB fixing of Thursday, 31 Dec 2026.

Settles on Thursday, 31 Dec 2026

Movement threshold
1% — see every threshold 5% — see every threshold 10% — see every threshold 50% — see every threshold

Default for currency pairs: 1%

Research estimate AI average · uncalibrated

AI average · uncalibrated Two research runs, by OpenAI and Claude. Each one searches news, central-bank and market data, cites what it used, then estimates. Weighted equally. ECB reference rate for 31 Dec 2026.

±1% from 152.40 ±5% from 152.40 ±10% from 152.40 ±50% from 152.40

Down 1% or more

37%

150.87 or lower

Between −1% and +1%

16%

150.87 to 153.92

Up 1% or more

47%

153.92 or higher

Down 1% or more

150.87 or lower

37%

Between −1% and +1%

150.87 to 153.92

16%

Up 1% or more

153.92 or higher

47%

The most likely of the three outcomes is up 1% or more, at 47%. Down 37% · Within 16% · Up 47% at ±1%.

Down 5% or more

13%

144.78 or lower

Between −5% and +5%

67%

144.78 to 160.02

Up 5% or more

20%

160.02 or higher

Down 5% or more

144.78 or lower

13%

Between −5% and +5%

144.78 to 160.02

67%

Up 5% or more

160.02 or higher

20%

The most likely of the three outcomes is between −5% and +5%, at 67%. Down 13% · Within 67% · Up 20% at ±5%.

Down 10% or more

2%

137.16 or lower

Between −10% and +10%

93%

137.16 to 167.64

Up 10% or more

5%

167.64 or higher

Down 10% or more

137.16 or lower

2%

Between −10% and +10%

137.16 to 167.64

93%

Up 10% or more

167.64 or higher

5%

The most likely of the three outcomes is between −10% and +10%, at 93%. Down 2% · Within 93% · Up 5% at ±10%.

Down 50% or more

<0.1%

76.20 or lower

Between −50% and +50%

>99.9%

76.20 to 228.59

Up 50% or more

<0.1%

228.59 or higher

Down 50% or more

76.20 or lower

<0.1%

Between −50% and +50%

76.20 to 228.59

>99.9%

Up 50% or more

228.59 or higher

<0.1%

The most likely of the three outcomes is between −50% and +50%, at >99.9%. Down <0.1% · Within >99.9% · Up <0.1% at ±50%.

How boundaries are counted

Down includes its boundary: a rate of exactly 150.87 counts as Down.

Up includes its boundary: a rate of exactly 153.92 counts as Up.

Within is strictly between 150.87 and 153.92. The three outcomes cannot overlap and always sum to 100%.

Down includes its boundary: a rate of exactly 144.78 counts as Down.

Up includes its boundary: a rate of exactly 160.02 counts as Up.

Within is strictly between 144.78 and 160.02. The three outcomes cannot overlap and always sum to 100%.

Down includes its boundary: a rate of exactly 137.16 counts as Down.

Up includes its boundary: a rate of exactly 167.64 counts as Up.

Within is strictly between 137.16 and 167.64. The three outcomes cannot overlap and always sum to 100%.

Down includes its boundary: a rate of exactly 76.20 counts as Down.

Up includes its boundary: a rate of exactly 228.59 counts as Up.

Within is strictly between 76.20 and 228.59. The three outcomes cannot overlap and always sum to 100%.

The rate on the target date, not whether a level is touched before it. Probability estimates, not guarantees. Methodology and limitations.

USD/JPY · 31 Dec · ±1% Down 37%, within 16%, up 47%
USD/JPY · 31 Dec · ±5% Down 13%, within 67%, up 20%
USD/JPY · 31 Dec · ±10% Down 2%, within 93%, up 5%
USD/JPY · 31 Dec · ±50% Down <0.1%, within >99.9%, up <0.1%

Rate so far and forecast range

USD/JPY since 31 Jul, then where the research estimate puts the rate on 31 Dec 2026.

Middle estimate
153.29
Even odds above or below
50% chance between
147.93 – 158.78
25th to 75th percentile
80% chance between
142.39 – 164.99
10th to 90th percentile
170160150140 31 Jul31 Aug30 Sep31 Oct30 Nov31 Dec RATE HISTORYFORECAST RANGETARGET +1% · 153.92 −1% · 150.87 +5% · 160.02 −5% · 144.78 +10% · 167.64 −10% · 137.16 +50% · 228.59 ↑ −50% · 76.20 ↓ 90th · 164.99 75th · 158.78 Median · 153.29 25th · 147.93 10th · 142.39 152.40 170160150140 31 Jul31 Aug30 Sep31 Oct30 Nov31 Dec +1% · 153.92 −1% · 150.87 +5% · 160.02 −5% · 144.78 +10% · 167.64 −10% · 137.16 +50% · 228.59 ↑ −50% · 76.20 ↓ 152.40
The forecast is only for 31 Dec 2026, 23:59 UTC; the shading between 30 Oct and 31 Dec is interpolated for readability, not a forecast of the path. Percentiles come from the AI average, spread evenly within each threshold band. The models put 84% of the probability outside ±1%.
Chart data as a table
Forecast percentiles for 31 Dec 2026, 23:59 UTC (AI average)
PercentileRate
10th142.39
25th147.93
Median (50th)153.29
75th158.78
90th164.99
USD/JPY rate history, month-ends and the reference date
DateRate
31 Jul 2026153.09
31 Aug 2026150.12
30 Sep 2026154.26
30 Oct 2026152.40

All movement thresholds

AI average for the ECB reference rate for Thursday, 31 Dec 2026, from the 152.40 reference. Rates are Japanese yen per US dollar. All four thresholds are on this page; choosing one updates the panels above.

AI average probabilities for each movement threshold
ThresholdMove Down Within Up Selection
±1% 37% Down 1% or more: 150.87 or lower 16% Between −1% and +1%: 150.87 to 153.92 47% Up 1% or more: 153.92 or higher Selected Show ±1%
±5% 13% Down 5% or more: 144.78 or lower 67% Between −5% and +5%: 144.78 to 160.02 20% Up 5% or more: 160.02 or higher Selected Show ±5%
±10% 2% Down 10% or more: 137.16 or lower 93% Between −10% and +10%: 137.16 to 167.64 5% Up 10% or more: 167.64 or higher Selected Show ±10%
±50% <0.1% Down 50% or more: 76.20 or lower >99.9% Between −50% and +50%: 76.20 to 228.59 <0.1% Up 50% or more: 228.59 or higher Selected Show ±50%

Within ranges exclude both boundaries. Wider thresholds always have a larger Within share: the estimates nest consistently. A 50% move in a major currency pair is extremely unlikely; the estimate is shown rather than hidden so all four thresholds stay comparable. Values shown as “<0.1%” or “>99.9%” are bounds; the underlying values sum to 100%.

Compare sources

Selected threshold ±1%. Every source uses the same event and the same 152.40 reference, and is listed in Down / Within / Up order. OpenAI and Claude differ by up to 1 percentage point at ±1%.

Selected threshold ±5%. Every source uses the same event and the same 152.40 reference, and is listed in Down / Within / Up order. OpenAI and Claude differ by up to 3 percentage points at ±5%.

Selected threshold ±10%. Every source uses the same event and the same 152.40 reference, and is listed in Down / Within / Up order. OpenAI and Claude differ by up to 2 percentage points at ±10%.

Selected threshold ±50%. Every source uses the same event and the same 152.40 reference, and is listed in Down / Within / Up order. OpenAI and Claude give the same estimates at ±50%.

Down, within and up probabilities by source at the selected threshold
Source Down Within Up Updated Coverage Details
AI average Average · uncalibrated 37%16%47% 31 Oct 2026, 06:12 UTC Fresh2 of 2 models Details: AI average — method
OpenAI Research run 38%15%47% 31 Oct 2026, 06:09 UTC Fresh4 cited sources Details: OpenAI — run details
Claude Research run 37%16%47% 31 Oct 2026, 06:10 UTC Fresh4 cited sources Details: Claude — run details
Statistical reference Synthetic fixture 40%19%41% 30 Oct 2026 fixing FreshECB rate history Details: Statistical reference — method
AI average Average · uncalibrated 13%67%20% 31 Oct 2026, 06:12 UTC Fresh2 of 2 models Details: AI average — method
OpenAI Research run 14%65%21% 31 Oct 2026, 06:09 UTC Fresh4 cited sources Details: OpenAI — run details
Claude Research run 13%68%19% 31 Oct 2026, 06:10 UTC Fresh4 cited sources Details: Claude — run details
Statistical reference Synthetic fixture 11%77%12% 30 Oct 2026 fixing FreshECB rate history Details: Statistical reference — method
AI average Average · uncalibrated 2%93%5% 31 Oct 2026, 06:12 UTC Fresh2 of 2 models Details: AI average — method
OpenAI Research run 3%92%5% 31 Oct 2026, 06:09 UTC Fresh4 cited sources Details: OpenAI — run details
Claude Research run 2%94%4% 31 Oct 2026, 06:10 UTC Fresh4 cited sources Details: Claude — run details
Statistical reference Synthetic fixture 1%98%1% 30 Oct 2026 fixing FreshECB rate history Details: Statistical reference — method
AI average Average · uncalibrated <0.1%>99.9%<0.1% 31 Oct 2026, 06:12 UTC Fresh2 of 2 models Details: AI average — method
OpenAI Research run <0.1%>99.9%<0.1% 31 Oct 2026, 06:09 UTC Fresh4 cited sources Details: OpenAI — run details
Claude Research run <0.1%>99.9%<0.1% 31 Oct 2026, 06:10 UTC Fresh4 cited sources Details: Claude — run details
Statistical reference Synthetic fixture <0.1%100.0%0.0% 30 Oct 2026 fixing FreshECB rate history Details: Statistical reference — method
Prediction markets No matching market No contract on USD/JPY settles near December 2026, and no related contract was found. Checked 31 Oct 2026, 06:05 UTC.
Source details
AI average — method

Simple mean of the OpenAI and Claude probabilities for each outcome, rounded so the three sum to 100. Uncalibrated: it has not been adjusted using settled outcomes.

Both models receive the same evidence pack and reference price, so their estimates are not statistically independent. Averaging two runs does not double the information.

OpenAI — run details

Model: [OpenAI model ID from run metadata] · Run: 31 Oct 2026, 06:09 UTC · Evidence pack: [pack ID]

Returns a probability distribution over 26 return bins; all four thresholds are computed from it. Checked to sum to 100 and to nest across thresholds before publishing.

Claude — run details

Model: [Anthropic model ID from run metadata] · Run: 31 Oct 2026, 06:10 UTC · Evidence pack: [pack ID]

Returns a probability distribution over 26 return bins; all four thresholds are computed from it. Checked to sum to 100 and to nest across thresholds before publishing.

Statistical reference — method

Synthetic fixture: lognormal rate at the target with zero median change and 4.1% volatility over the horizon (about 10% a year).

A simple baseline from price history. It is not a market-implied (risk-neutral) estimate, and no claim is made that it is more or less accurate than the models.

AI average

Average · uncalibrated
Fresh
Down
37%
Within
16%
Up
47%

Updated 31 Oct 2026, 06:12 UTC · 2 of 2 models

Details: AI average — method

Simple mean of the OpenAI and Claude probabilities for each outcome, rounded so the three sum to 100. Uncalibrated: it has not been adjusted using settled outcomes.

Both models receive the same evidence pack and reference price, so their estimates are not statistically independent. Averaging two runs does not double the information.

OpenAI

Research run
Fresh
Down
38%
Within
15%
Up
47%

Updated 31 Oct 2026, 06:09 UTC · 4 cited sources

Details: OpenAI — run details

Model: [OpenAI model ID from run metadata] · Run: 31 Oct 2026, 06:09 UTC · Evidence pack: [pack ID]

Returns a probability distribution over 26 return bins; all four thresholds are computed from it. Checked to sum to 100 and to nest across thresholds before publishing.

Claude

Research run
Fresh
Down
37%
Within
16%
Up
47%

Updated 31 Oct 2026, 06:10 UTC · 4 cited sources

Details: Claude — run details

Model: [Anthropic model ID from run metadata] · Run: 31 Oct 2026, 06:10 UTC · Evidence pack: [pack ID]

Returns a probability distribution over 26 return bins; all four thresholds are computed from it. Checked to sum to 100 and to nest across thresholds before publishing.

Statistical reference

Synthetic fixture
Fresh
Down
40%
Within
19%
Up
41%

Updated 30 Oct 2026 fixing · ECB rate history

Details: Statistical reference — method

Synthetic fixture: lognormal rate at the target with zero median change and 4.1% volatility over the horizon (about 10% a year).

A simple baseline from price history. It is not a market-implied (risk-neutral) estimate, and no claim is made that it is more or less accurate than the models.

Prediction markets

No matching market

No contract on USD/JPY settles near December 2026, and no related contract was found.

AI average

Average · uncalibrated
Fresh
Down
13%
Within
67%
Up
20%

Updated 31 Oct 2026, 06:12 UTC · 2 of 2 models

Details: AI average — method

Simple mean of the OpenAI and Claude probabilities for each outcome, rounded so the three sum to 100. Uncalibrated: it has not been adjusted using settled outcomes.

Both models receive the same evidence pack and reference price, so their estimates are not statistically independent. Averaging two runs does not double the information.

OpenAI

Research run
Fresh
Down
14%
Within
65%
Up
21%

Updated 31 Oct 2026, 06:09 UTC · 4 cited sources

Details: OpenAI — run details

Model: [OpenAI model ID from run metadata] · Run: 31 Oct 2026, 06:09 UTC · Evidence pack: [pack ID]

Returns a probability distribution over 26 return bins; all four thresholds are computed from it. Checked to sum to 100 and to nest across thresholds before publishing.

Claude

Research run
Fresh
Down
13%
Within
68%
Up
19%

Updated 31 Oct 2026, 06:10 UTC · 4 cited sources

Details: Claude — run details

Model: [Anthropic model ID from run metadata] · Run: 31 Oct 2026, 06:10 UTC · Evidence pack: [pack ID]

Returns a probability distribution over 26 return bins; all four thresholds are computed from it. Checked to sum to 100 and to nest across thresholds before publishing.

Statistical reference

Synthetic fixture
Fresh
Down
11%
Within
77%
Up
12%

Updated 30 Oct 2026 fixing · ECB rate history

Details: Statistical reference — method

Synthetic fixture: lognormal rate at the target with zero median change and 4.1% volatility over the horizon (about 10% a year).

A simple baseline from price history. It is not a market-implied (risk-neutral) estimate, and no claim is made that it is more or less accurate than the models.

Prediction markets

No matching market

No contract on USD/JPY settles near December 2026, and no related contract was found.

AI average

Average · uncalibrated
Fresh
Down
2%
Within
93%
Up
5%

Updated 31 Oct 2026, 06:12 UTC · 2 of 2 models

Details: AI average — method

Simple mean of the OpenAI and Claude probabilities for each outcome, rounded so the three sum to 100. Uncalibrated: it has not been adjusted using settled outcomes.

Both models receive the same evidence pack and reference price, so their estimates are not statistically independent. Averaging two runs does not double the information.

OpenAI

Research run
Fresh
Down
3%
Within
92%
Up
5%

Updated 31 Oct 2026, 06:09 UTC · 4 cited sources

Details: OpenAI — run details

Model: [OpenAI model ID from run metadata] · Run: 31 Oct 2026, 06:09 UTC · Evidence pack: [pack ID]

Returns a probability distribution over 26 return bins; all four thresholds are computed from it. Checked to sum to 100 and to nest across thresholds before publishing.

Claude

Research run
Fresh
Down
2%
Within
94%
Up
4%

Updated 31 Oct 2026, 06:10 UTC · 4 cited sources

Details: Claude — run details

Model: [Anthropic model ID from run metadata] · Run: 31 Oct 2026, 06:10 UTC · Evidence pack: [pack ID]

Returns a probability distribution over 26 return bins; all four thresholds are computed from it. Checked to sum to 100 and to nest across thresholds before publishing.

Statistical reference

Synthetic fixture
Fresh
Down
1%
Within
98%
Up
1%

Updated 30 Oct 2026 fixing · ECB rate history

Details: Statistical reference — method

Synthetic fixture: lognormal rate at the target with zero median change and 4.1% volatility over the horizon (about 10% a year).

A simple baseline from price history. It is not a market-implied (risk-neutral) estimate, and no claim is made that it is more or less accurate than the models.

Prediction markets

No matching market

No contract on USD/JPY settles near December 2026, and no related contract was found.

AI average

Average · uncalibrated
Fresh
Down
<0.1%
Within
>99.9%
Up
<0.1%

Updated 31 Oct 2026, 06:12 UTC · 2 of 2 models

Details: AI average — method

Simple mean of the OpenAI and Claude probabilities for each outcome, rounded so the three sum to 100. Uncalibrated: it has not been adjusted using settled outcomes.

Both models receive the same evidence pack and reference price, so their estimates are not statistically independent. Averaging two runs does not double the information.

OpenAI

Research run
Fresh
Down
<0.1%
Within
>99.9%
Up
<0.1%

Updated 31 Oct 2026, 06:09 UTC · 4 cited sources

Details: OpenAI — run details

Model: [OpenAI model ID from run metadata] · Run: 31 Oct 2026, 06:09 UTC · Evidence pack: [pack ID]

Returns a probability distribution over 26 return bins; all four thresholds are computed from it. Checked to sum to 100 and to nest across thresholds before publishing.

Claude

Research run
Fresh
Down
<0.1%
Within
>99.9%
Up
<0.1%

Updated 31 Oct 2026, 06:10 UTC · 4 cited sources

Details: Claude — run details

Model: [Anthropic model ID from run metadata] · Run: 31 Oct 2026, 06:10 UTC · Evidence pack: [pack ID]

Returns a probability distribution over 26 return bins; all four thresholds are computed from it. Checked to sum to 100 and to nest across thresholds before publishing.

Statistical reference

Synthetic fixture
Fresh
Down
<0.1%
Within
100.0%
Up
0.0%

Updated 30 Oct 2026 fixing · ECB rate history

Details: Statistical reference — method

Synthetic fixture: lognormal rate at the target with zero median change and 4.1% volatility over the horizon (about 10% a year).

A simple baseline from price history. It is not a market-implied (risk-neutral) estimate, and no claim is made that it is more or less accurate than the models.

Prediction markets

No matching market

No contract on USD/JPY settles near December 2026, and no related contract was found.

Distribution of possible moves

AI average, built from the threshold probabilities and spread evenly within each band. Heights show probability per percentage point.

Down band Within band Up band Selected threshold Reference, 0%

Histogram of the bands listed in the table beside: probability per percentage point for each band between −10% and +10%.

Hover a band for its price range and probability. Not drawn: 2% of outcomes below −10% (137.16) and 5% above +10% (167.64). The ±50% boundaries (76.20 and 228.59) are far outside this axis.

The same distribution as a table
Change Rate Prob. Per point
Below −10% Below 137.16 2% —
−10% to −5% 137.16–144.78 11% 2.19
−5% to −1% 144.78–150.87 24% 6.03
−1% to +1% 150.87–153.92 16% 7.86
+1% to +5% 153.92–160.02 27% 6.73
+5% to +10% 160.02–167.64 15% 3.01
Above +10% Above 167.64 5% —
Total 100%

Why these estimates?

Directional factors follow the quote (Japanese yen per US dollar): “higher” means USD/JPY rises. Each item says whether it is a reported fact or the model’s inference.

What could push USD/JPY higher?

  1. [Reported fact] Reported fact

    [Fact quoted from source 2 with its own date, for example a published inflation figure.] 2

  2. Risk sentiment Model inference

    A broad shift in global risk appetite could favour the US dollar against the yen before the target date. 3

What could push it lower?

  1. Rate differential Model inference

    If the Bank of Japan keeps rates higher for longer than the Federal Reserve, the yen tends to gain against the US dollar. 1

  2. [Reported fact] Reported fact

    [Fact quoted from source 4 with its own date.] 4

Events before the target date

Scheduled dates come from official calendars. Scenarios are hypothetical and have no date.

Confirmed dates

  1. 9 Dec 2026

    US Federal Reserve policy decision

    FOMC meeting 8–9 Dec; statement at 2:00 p.m. ET. Federal Reserve calendar

    Confirmed date

  2. 17 Dec 2026

    ECB monetary policy meeting

    Decision followed by a press conference. ECB meeting calendar

    Confirmed date

  3. 31 Dec 2026

    Target observation

    ECB reference rate for Thursday, 31 Dec settles this forecast.

    Forecast target

  4. 26–27 Jan 2027

    US Federal Reserve meeting

    After the target date, so it cannot affect this outcome.

    After target

Hypothetical scenarios

  • A sudden global risk-off move that shifts demand toward safe-haven currencies.

    Scenario · no date

Prediction-market context

No matching market Related market: none Checked 31 Oct 2026, 06:05 UTC

No contract on USD/JPY settles near December 2026, and no related contract was found. Market evidence is left out rather than shown as 0%.

Options-implied (risk-neutral) estimates are a different kind of evidence and are not shown in this version. How sources are matched

What this forecast may miss

  • Moves between fixings. The ECB rate is one daily observation; market rates at other times can differ.
  • Weekend news. Anything after 31 Oct 2026, 06:12 UTC enters only with a new version.
  • Calibration. Very few USD/JPY forecasts have settled; the AI average is uncalibrated.

Forecast versions

Each version keeps its own reference rate and source timestamps. Earlier versions are stored, never overwritten. Shown at ±1%; every version holds all four thresholds.

Stored versions of this forecast
Published Reference rate Down / Within / Up at ±1% Version
31 Oct 2026, 06:12 UTC 152.40 · 30 Oct fixing 37% / 16% / 47% Current version · viewing