USDJPY Long on March 20: Five Evaluations, One Enter, +2.90R to TP3
SkyAnalyst AI journal entry: USDJPY Long on Mar 20, 2026 closed +2.9R on TP3. Full workspace view, decision log, and AI reasoning, unedited.
Before the trade, meet the system.
SkyAnalyst is not one AI trader. It is four specialist agents — each with its own data pipeline, each maintaining state between evaluations, and each required to agree before a position is sized. They don’t chat in prose. They write structured messages to a shared state object that each reads on every evaluation cycle. That’s what makes the system auditable — and it’s what this case study will show, step by step, on a specific setup the trend agent almost passed on.
The morning the rate differential carried the trade
The macro print into NY-AM was a tightening-bullish setup for USDJPY without being a maximum-conviction one. The Macro Agent had written regime as lean_bull at 52 percent confidence to shared state, low enough that the rate-differential argument could not unilaterally drive the trade but high enough that it did not block. US 10-year yields sat at 4.382 against a 5-day EMA of 4.285, above the average and above yesterday's high. DXY held at 99.765 against a 5-day EMA of 99.745, also above. The yield-USD backdrop was the cleanest part of the read.
The risk-tone read was the complication. VIX printed 26.13 against a 5-day EMA of 24.96, which signaled risk-off, and that matters for USDJPY because the yen can strengthen during stress even when yields rise. The Dow sat below its 5-day EMA. Gold was not rallying alongside yields, which removed the cleanest divergence warning, but the broader risk tone was defensive enough that the Risk Agent flagged carry-unwind risk as elevated and reduced the size envelope.
USDJPY itself was offering a textbook pullback long. Price had reclaimed the Tokyo high at 159.021 and was pulling back into the 159.02 to 159.06 band that overlapped the 5-minute EMA cluster and the intraday support shelf. The 60-minute trend was bullish but transitional rather than fully stacked. The 15-minute and 5-minute were clearly up but stretched. The Risk Agent put the stop at 158.94, slightly below the Trend Agent's invalidation at 158.959, with 11.8 pips of room. Multi-frame confluence on the bull side, with VIX as the regime asterisk. Setup grade printed C+: structural read clean, macro lean correct but low confidence, every floor cleared and nothing more.
The setup at 15:18 UTC was a USDJPY pullback long into reclaimed intraday support. Walking through the structural requirement explains why the system held four waits before firing.
What the pattern is
Price has been pushing higher on the 60-minute chart and just reclaimed a prior-session high. Somewhere inside the advance, it pulls back into a short-term support zone, typically the reclaimed level itself or a 5-minute EMA confluence above VWAP. A professional does not buy the touch of resistance. They wait for the retest of support to confirm: a 5-minute candle that pulls into the zone, holds above the reclaimed level, and closes back above the entry trigger with RSI staying above 40.
How pros actually use it on a yen pair
The pullback long on USDJPY trades the rate differential mechanically. When 10-year yields rise faster than 10-year JGB yields, the carry trade refunds, and pullbacks into intraday support get bought by macro flows even when the equity tape is shaky. The tell is whether VIX is moving with yields or against them. Moving with yields, the pullback typically holds. Moving sharply against yields, carry unwind dominates and the pullback fails.
Why this works on a tightening-bullish backdrop
USDJPY is the cleanest expression of US-Japan rate differential at the front end of the curve. When the morning print pushes US yields above their 5-day EMA and DXY confirms, the structural bid for the pair thickens through the session. A counter-trend pullback into reclaimed support is mechanically a re-positioning move, not a regime change. The fade catches the next leg of the directional move after the pullback confirms it lacks downside follow-through.
Why this graded C+ rather than B
Three things kept the grade modest. The Macro Agent's confidence was 52 percent, lean correct but not authoritative under the system's 60 percent confidence threshold. VIX was elevated above its 5-day EMA, which meant carry-unwind risk was real even with yields supportive. The 60-minute EMA stack was not fully stacked bullish, only transitionally so. C+ is the system's notation for tradeable, every floor clears, conviction not high enough for B.
How the system reads this, dynamically not dogmatically
The pullback long on USDJPY is one playbook of many. The same Friday morning the Trend Agent was watching a fresh-breakout long on USDJPY above 159.18 to 159.20 that the system gated as lower quality than the pullback retest, and a no-trade short on USDJPY that did not pass the confluence gate because 10-year yields and DXY both worked against it.
SkyAnalyst doesn't favor any single strategy. The confluence math picks the playbook each evaluation cycle. The fact that USDJPY scored a C+ pullback long while EURUSD had been shorting the bounce two sessions earlier is a property of the tape, not a contradiction in the system. Different pairs, different rate-differential vectors, different playbooks. The math reads the tape first, then fits the pattern dynamically to what is actually there.
USDJPY NY AM Session Overview
USDJPY is trading in a macro-supportive but risk-fragile environment for longs. The US 10Y yield is above its 5-day EMA (4.382 vs 4.285) and has broken above yesterday’s high, while DXY is also above its 5-day EMA (99.765 vs 99.745), so the yield/USD backdrop is tightening-bullish. The Macro Analysis Agent is only lean_bull with 52% confidence, so its view is not authoritative under your 7+ rule, but its structural rate-differential argument still supports USDJPY on dips. Gold is not rallying alongside yields; it is below its 5-day EMA and down over the last several days, so no gold/yield divergence warning there.
Risk sentiment is the main complication. VIX is above its 5-day EMA (26.13 vs 24.96), which signals risk-off, and that matters because JPY can strengthen during stress even when yields rise. We were not given NYAD values, but with VIX elevated and the Dow below its 5-day EMA and below yesterday’s low, the broader tape is defensive enough to treat this as carry-risk elevated. Since USDJPY is also trading near the upper end of today’s range, this creates a carry unwind risk warning: upside continuation is possible, but reversals can be sharp.
No USD high/medium-impact events are scheduled today, so there is no calendar-based entry restriction for this session.
Trend Agent says BULLISH, 61% confidence, weak strength, transitioning regime, with price above VWAP and near-term bull control intact above 158.959 invalidation. That aligns with yields, but the 60m structure is still transitionary, not a clean trend. On the 60m chart, price is above VWAP and momentum has improved, but the broader EMA structure remains not fully bullish. On the 15m/5m, trend is clearly up, but both are overbought, which argues against chasing highs and in favor of a pullback-and-hold long only.
Directional Bias: Bullish
Volatility: High
Setup #1: USDJPY LONG
- Entry: 159.02-159.06
- Stop Loss: 158.94
- Targets: TP1=159.15, TP2=159.28, TP3=159.40
- R-Multiples: TP1=1.4R, TP2=2.8R, TP3=4.3R
- Quality Score: 7.1/10
- Confidence: Medium - 10Y and DXY confirm upside, Trend Agent is bullish with 6+ confidence, price is above VWAP on 15m/5m, and the entry zone is a retest of near-term intraday support/EMA structure. Main risks are elevated VIX, overbought short-term RSI, and 60m trend still being transitional rather than fully stacked bullish.
Execution trigger: Only take if a 5m candle pulls back into 159.02-159.06 and then closes back above 159.06 with 5m RSI holding above 40 and price reclaiming/holding above the 5m fast EMA. Best case is a retest-and-hold above the broken local intraday level around 159.02/159.03.
Why it passes confluence gate:
- 10Y yield direction matches long
- Trend Agent bullish with confidence >= 6
- 5m trigger at a key level with RSI confirmation
- Entry zone is at session structure / short-term fib-retest area
- Macro backdrop is lean bullish on rate differential, though confidence is below 6
- Price is above VWAP on intraday frames
Invalidation condition: Cancel if 5m closes below 158.96, or if price loses 158.959 Trend Agent invalidation decisively. That would break near-term structure and raise probability of a carry-unwind reversal.
No Short Setup
I do not recommend a short for the NY AM session here. Although risk-off conditions can support JPY, the short side does not pass the confluence gate cleanly because:
- 10Y yields are rising, which works against shorts
- DXY is firm, which works against shorts
- Trend Agent is bullish
- Price is above VWAP on 15m and 5m
A short would be fighting the stronger rates signal and would only become valid after a loss of 158.959 with follow-through.
Bottom Line
For NY AM, this is a buy-the-pullback, not chase-the-breakout market. The only acceptable setup is a retest long into 159.02-159.06 with tight confirmation. If price stays stretched near 159.15-159.18 without pulling back, No Trade—the reward compresses too much relative to the structural stop, especially with VIX elevated. Risk should be reduced versus normal because this is a high-volatility, carry-sensitive regime.
Decision log
First evaluation, 15:13 UTC, confidence 62 percent. Price has pulled into the 159.02 to 159.06 retest band. The structural premise is intact: 10-year yields above their 5-day EMA and rising, DXY above its 5-day EMA, Trend Agent bullish at 61 percent, 60-minute structure bullish above VWAP. But the most recent 5-minute bar has not closed back above 159.06 with confirmation. I want a closed-bar reclaim with RSI above 40 before the trigger clears. Declining.
Second evaluation, 15:14 UTC, confidence 62 percent. The bar in progress is forming a small bullish body inside the band but has not closed above the 159.06 trigger. Volume sits at the 60-period average rather than meaningfully above it. The premise has not weakened, the trigger simply has not printed. Declining.
Third evaluation, 15:15 UTC, confidence 63 percent. Confidence ticked up as the 5-minute closed marginally above 159.05, but the close was inside the band rather than above the trigger. RSI sits at 41, just above the 40 floor I treat as meaningful. The closed-bar reclaim above 159.06 has not printed. Declining.
Fourth evaluation, 15:16 UTC, confidence 62 percent. The next bar opened lower and probed back toward 159.04, then closed mid-band as an indecision body. Cross-Asset returned supportive, DXY held above its 5-day EMA, but the trigger condition I require has not printed on a closed bar. Below the trigger means no entry, full stop. Declining.
Fifth evaluation, 15:18 UTC, confidence 63 percent. The most recent 5-minute closed back above 159.06 with RSI holding at 44 and price reclaiming the 5-minute fast EMA, the next print opened higher with no lower wick into the band. Volume came in above the 60-period average. This is the trigger I have been watching for across the prior four evaluations. Entering long at 159.058, stop 158.94, TP1 159.15, TP2 159.28, TP3 159.40.
On a $100k account at 2.0% risk per trade.
Each trade risks +$2,000 (1R). The system's actual scale-out behavior may differ, see disclaimer.
| Scenario | R-multiple | Profit on $100k |
|---|---|---|
| Stop hit (invalidated) | -1R | −$2,000 |
| TP1 hit | +0.78R | +$1,560 |
| TP2 hit | +1.88R | +$3,760 |
| TP3 hit (max potential)Actual | +2.9R | +$5,800 |
All six agents combined.
What this trade teaches
The four waits across five minutes are the discipline beat. Confidence floated between 62 and 63 percent on every evaluation, never spiking, never collapsing. A discretionary trader watching a flat 62 percent score across four cycles would have either entered on the first read because "the score is above the floor" or skipped the trade entirely because "the score is not climbing." The system did neither. It waited for the closed-bar reclaim above 159.06 with RSI holding above 40, and entered when both conditions printed on the same evaluation cycle.
The trade closed at +2.90R (TP3) over fifty-four hours and forty-six minutes, with zero drawdown from entry to exit. Most of that hold was the weekend session into Sunday's open, which the Risk Agent had pre-sized for. The position did not require active management between the Friday entry and the Sunday close. That outcome is a property of the rate-differential trade on a tape that was supportive of the carry.
The score never moved more than one percentage point across five evaluations. The trigger was the closed bar above 159.06 with RSI holding the 40 floor, not the score crossing any threshold the score had never crossed. - From the post-trade review
The shape contrasts with the same week's paired EURUSD shorts, documented in the March 18 EURUSD short at +1.07R and the March 18 rally-fade short at +1.81R. Different instrument, different direction, different playbook. The dollar-soft tape that fueled the EURUSD shorts also produced a yen-pair long, because USDJPY tracks the rate differential while EURUSD tracks the dollar's relative strength against the euro specifically. Two correct reads, two opposite directions, both ran to TP3. The MTD book stood at 27 trades and +3.54R net after this trade closed.
From the desk
This trade did not look special on the setup card. A C+ grade. A 63 percent confluence score on the entry evaluation, with the score never moving more than one point across five cycles. The Macro Agent gating regime as lean_bull at 52 percent rather than confirmed bull. None of those numbers, on their own, would have any reader marking this as the +2.90R closer of a mixed week.
What separated it from the breakouts that stopped earlier in the quarter was the tape, and the tape was specific. We do not say "this will run 34 pips over fifty-four hours." We say "this clears every floor, the bias is intact across timeframes, the rate differential supports the direction, the closed-bar reclaim has finally printed with RSI confirmation." The system places the stop above structural invalidation, sets targets at the next three references, and lets the position run.
On March 20 at 15:13 UTC the Macro Agent had written regime lean_bull at 52 percent into the shared state. The Trend Agent on each of its five evaluations read that value verbatim and weighted it accordingly: lean correct but below the 60 percent confidence threshold meant the trade had to clear the structural and rate-differential floors on its own merits. If the Macro Agent had been chatting in prose about mixed signals, the Trend Agent would have had to interpret tone. It does not, so it did not. The coordination between the four agents is the product, and the multi-session continuation through the weekend is what the coordination produced.
From the SkyAnalyst Team.
At a Glance
What this teaches about AI-driven trading
How does the system handle a position that holds across the weekend session?
The Risk Agent pre-sizes positions for expected hold duration. On a Friday entry into a multi-session setup, the size envelope accounts for the weekend gap risk and the lower liquidity around the Sunday open. The stop placement is anchored to structural invalidation, not to a fixed pip distance, so the Risk Agent's sizing reflects the structural distance. The position does not require active management between the Friday entry and the next session's open if the underlying premise holds.
Why did the system enter at 63 percent confidence when the threshold is 60 percent?
Confidence above threshold is a permission to act once the trigger prints, not an instruction to act when the score crosses the floor. On March 20 the score sat between 62 and 63 percent for all five evaluations, never spiking and never collapsing. The trade fired on the fifth evaluation because the closed-bar reclaim above 159.06 printed with RSI holding above 40 and confirmation volume, not because the score crossed any new threshold the score had not already cleared.
What does +2.90R translate into dollar terms for a typical account?
On a hypothetical $100,000 account at 2 percent risk per trade, 1R equals $2,000, so +2.90R (TP3) translates to roughly +$5,800 of potential return. That figure assumes the position is held to the highest take-profit reached. In live execution the broker scales out at TP1 for risk management, so the recorded broker P&L is smaller than the full-arc R-multiple shown.
How does the system trade USDJPY long when the broader dollar tape leans soft?
USDJPY tracks the US-Japan rate differential at the front end of the curve, not the dollar's strength against every other currency. When 10-year US yields rise faster than 10-year JGB yields and DXY confirms above its 5-day EMA, the structural bid for the pair thickens even on sessions where the broader dollar tape weakens against the euro or sterling. The system reads the rate differential as the primary input for USDJPY direction and the cross-asset confirmation as the secondary check, rather than reading the dollar index in aggregate as a single directional signal.
Run your markets with SkyAnalyst
Seven-day free trial. No credit card. Full access to the Trend Agent, Macro Agent, and six-factor confluence scoring.
Trading involves substantial risk of loss. Past performance is not indicative of future results. The analysis shown was produced by an AI model operating on SkyAnalyst’s live trading infrastructure; it is shared for educational and research purposes only and is not financial advice. About reported results. Each model outputs three take-profit targets (TP1, TP2, TP3) per trade. In live execution, models typically scale out at TP1 for risk management — the broker position records this as a TP1 exit. The R-multiples and dollar returns shown in this article reflect the full potential of the trade: where the market actually traveled to (the highest take-profit hit, or stop loss) before the setup was invalidated or exhausted. This lets readers see the complete arc of each setup, not just where the position was closed. Simulated returns in this article are calculated against a hypothetical $100,000 account at 2% risk per trade (1R = $2,000). These are educational reference figures and do not reflect any specific account or broker execution. Your actual result depends on your position size, your risk parameters, and live market conditions.
From the SkyAnalyst Journal
March 2026 Monthly Recap: 42 Trades, Variance Both Directions, Break-Even Net
Forty-two trades. Twenty-two winners, twenty losers, 52.4 percent win rate. Net minus 0.13R, essentially flat on a TP1 baseline. The month produced both the deepest published drawdown and the bumper week of the record.
USDJPY Short on March 31 - Closing the Month on a TP3 Winner
A pullback short on USDJPY entered at 159.23 ran to TP3 at 158.75 in 2h 32m, closing at +3.20R. The closing-day winner of a March that finished -0.13R / 22W-20L on the TP1-baseline tally.
EURUSD Long on March 31 - Second of Two Same-Day TP3 Winners
A Bullish Pullback long on EURUSD entered at 1.1520 ran to TP3 at 1.1558 over four hours and seventeen minutes, closing at +1.58R. The second of two TP3 winners on the closing day of a near-flat March.