Claude Evaluated This EURUSD Long 10 Times Before Losing -1.0R
Claude's most-deliberated trade of the season. Ten confidence checks, one entry, and -$1,078.92 (SL) for -1.0R when the level finally broke the other way.
This is a daily analysis from The AI Trading Benchmark — an ongoing experiment where Claude and GPT independently analyze the same markets, take their own trades, and publish full transparency on every decision. No shared signals. No prompt engineering. Same data, different models, real outcomes tracked over a full season. Every analysis shown in these articles is produced by the AI model before the trade is taken — unedited, unfiltered, published in full. Full methodology.
Today is April 17, 2026 — Day 4 of Season 1, Phase 1. Claude enters the session after a NAS100 VWAP rejection loss yesterday that dropped the balance to $51,732. GPT sat out yesterday entirely — no trades. Today both models are active. Claude finds an EURUSD pullback long. GPT is scanning XAUUSD and USDJPY. This is the first of three Claude trades today — and the first to close.
About reported results. Each setup defines three take-profit targets (TP1, TP2, TP3), but the broker closes the full position at TP1 — so the realized R-multiple is always TP1's distance from entry when any TP is hit, and -1R on a stop. The dollar P&L shown in this article is the actual broker close at TP1 (or stop). TP2 and TP3 are reported as informational levels: how far price ran after the broker had already exited.
Result
R-Multiple
AI Confidence
Win Rate
Season Record
Market Environment — April 17, 2026
The EURUSD setup emerged in a favorable macro window. DXY had fallen from 98.189 to 97.727, trading below its 5-day EMA for the first time in three sessions. US 10Y yields dropped from 4.315 to 4.242 — below the 5-day EMA and below yesterday's low. VIX pulled back to 17.62 from 18.44, signaling a modest improvement in risk appetite.
The Trend Agent read EURUSD as NEUTRAL at 56% confidence — a "transitioning" regime where the 60-minute structure still held above EMAs and VWAP, but lower timeframes (5m/15m) had rolled over. The Macro Agent scored forex as "lean bull" at 52% — barely above neutral, with the headline noting that "policy dominance is overriding commodity correlations."
The bull case was macro-driven: falling dollar, falling yields, improving risk tone. The setup was a pullback long into a zone where price had previously bounced. The bear case was the transitioning regime — momentum was fading on lower timeframes even as the higher timeframe structure held.
Claude's analysis concluded that the macro tailwinds outweighed the momentum concerns, identifying an entry zone at 1.1825–1.1828 with a stop at 1.1813. The risk-reward was structured around a bounce from this pullback level. It never came.
EURUSD LONG
Setup: EURUSD Pullback Long
Analysis by SkyAnalyst AI
The 10-Evaluation Wait
This trade tested Claude's patience more than any entry this season. The setup was clear: EURUSD pullback long at 1.1825–1.1828, a zone where price had bounced earlier in the session. The macro backdrop was supportive — DXY falling, yields falling, risk tone improving. But the entry trigger required a bullish confirmation candle inside the zone.
Claude evaluated 10 times over 30 minutes before entering. The confidence journey:
- Evaluations 1-2: 35% → WAIT. Price hadn't reached the zone yet.
- Evaluations 3-5: 35-45% → WAIT. Price entered the zone but no bullish confirmation candle.
- Evaluations 6-8: 40% → WAIT. Price hovering at the zone edge, still no rejection candle.
- Evaluation 9: 40% → WAIT. Price pulled below the zone.
- Evaluation 10: 68% → ENTER. A bullish reaction sequence formed — bearish candle followed by recovery with a higher low.
What Went Wrong
The "bullish reaction" that triggered the entry was a lower-timeframe fake. Price briefly bounced from the pullback zone, creating what looked like a higher low on the 5-minute chart. But the bounce had no follow-through. Within 35 minutes of entry, price broke below the 1.1813 stop level and never recovered.
The fundamental issue: the Trend Agent had the regime right — "transitioning." A transitioning regime means momentum is shifting. Claude's macro-driven bull thesis (falling DXY, falling yields) was valid on the hourly timeframe, but the intraday price action was already breaking down. The 5m and 15m momentum had rolled over before the entry.
The Pattern: Claude Enters on Marginal Confidence
This is Claude's third loss in four sessions. All three losses share a pattern: entry confidence between 62-68%, extended evaluation counts (10, 5, 12), and setups where the macro thesis was sound but the intraday execution timing was off.
Compare this to Claude's opening day where three trades hit at similar confidence levels (62-66%) but were all trend-following setups. The difference isn't the confidence threshold — it's whether the AI is trading WITH the intraday trend or AGAINST it.
Today Claude entered against a transitioning regime on the lower timeframes. The macro said bullish. The 5-minute chart said bearish. The macro was right on the daily — but the intraday stop got hit first.
NEUTRAL 56%Regime:
TRANSITIONINGMacro:
HEADWINDlean_bull 52%Policy dominance is overriding commodity correlations and geopolitical risk premiums across the FX complex. Central bank cutting cycles (BOC 225-300bp, RBA 200-250bp, ECB June cut) are the sole price drivers, with oil's -6.6% crash failing to move commodity-linked currencies as expected. Institutional JPY hedging despite ideal carry conditions suggests embedded tail-risk awareness not yet visible in implied volatility.
17.47 (normal)98.202 neutral4.25 falling91.59 bearishWhat the AI Actually Saw
Full 12.3k-character analysis produced by the AI before entering this trade. Unedited.
- Multi-timeframe technical analysis
- Key support and resistance levels
- Pattern recognition and trade setup identification
- Risk-reward assessment with defined invalidation
- Macro context and cross-asset correlation check
Powered by SkyAnalyst AI
SkyAnalyst AIEURUSD NY AM Session Analysis — April 17, 2026
Step 1: Macro Regime Assessment
| Indicator | Value | vs 5D EMA | Trend | EURUSD Implication |
|---|---|---|---|---|
| DXY | 97.727 | Below 98.079 | Falling (closed 98.189 → now 97.727, below yesterday's low) | Bullish EURUSD |
| US 10Y | 4.242 | Below 4.276 | Falling (4.315 → 4.242, below yesterday's low) | Bullish EURUSD |
| VIX | 17.62 | Below 18.44 | Falling (3-day decline: 18.37 → 18.18 → 17.95 → 17.62) | Bullish EURUSD |
Macro Agent: Lean bearish, confidence 62%, citing rally exhaustion at 1.1825 resistance and ECB dovish guidance. Tradeability only moderate (45/100). Intraday horizon is neutral, short-term is lean_bear — timeframe divergence flagged.
DXY Assessment (Dominant Driver): DXY is decisively below its 5D EMA and falling hard today — broke below yesterday's low (97.799) to 97.592 intraday. This is the strongest single signal and it screams bullish EURUSD.
Hard Rule Check: Macro confidence is 62% (above 50, below 70) — no hard directional block applies. However, the Macro Agent's bearish lean directly conflicts with the dominant DXY signal. The DXY is the dominant driver per the rules, and all three confirming indicators (DXY falling, yields falling, VIX falling) align bullish. The Macro Agent's bearish narrative (rally exhaustion) is a caution flag, not a veto.
Macro Regime Verdict: Bullish EURUSD, driven by DXY weakness. Macro Agent divergence noted — reduces conviction but does not override the price-based macro signals.
Step 2: Agent Alignment
| Agent | Direction | Confidence | Notes |
|---|---|---|---|
| Macro Agent | Lean Bear | 62% | Rally exhaustion narrative; intraday = neutral |
| Trend Agent | Bullish | 65% | Moderate strength, TRANSITIONING regime, REDUCE_SIZE |
Alignment: The agents diverge. Macro leans bear while Trend is bullish. Per rules: drop confidence by 2 points (effective base drops) and only consider mean-reversion trades at session high/low or VWAP.
However, the DXY/10Y/VIX trio all confirm the Trend Agent's bullish direction. The Macro Agent's bearish call is fundamentally narrative-based (exhaustion, ECB dovish), while the real-time price drivers are all moving in EURUSD's favor. I weight the Trend Agent + DXY confirmation as the operative signal, with the Macro divergence mandating reduced sizing and mean-reversion focus.
Agent Alignment Verdict: Divergent — bullish technicals vs. lean-bear macro narrative. Confidence reduced. Mean-reversion and pullback entries preferred over breakout chasing.
Step 3: HTF Bias (60-Minute)
EMA Stack:
- All 5 candles show price_vs_fast = "above" and trend_direction = "above"
- EMA fast > EMA slow across all bars (e.g., latest: 1.17942 > 1.17882)
- Bullish stack confirmed ✅
RSI Trend:
- Progressed from 57.5 → 58.8 → 58.4 → 75.9 → 75.0
- Currently overbought at 75 on 60min — bearish signal per the indicator
- This is the Trend Agent's key concern: late-stage push
MACD Histogram:
- Sequence: 0.0002 → 0.00023 → 0.00023 → 0.00046 → 0.00059
- Strong and expanding — momentum is still accelerating ✅
Price vs. Daily Reference Levels:
- Current price 1.18309 is above yesterday's high (1.17818), yesterday's close (1.17806), and the 5D EMA (1.18)
- Price is above the daily pivot (1.179) and above all S/R support levels
- Price is near the Trend Agent resistance at 1.1849 (session high) and above the Trend Agent support at 1.1825
- VWAP at ~1.1791 is far below — price is extended above VWAP on all timeframes (upper 2SD band)
HTF Bias: Bullish trend intact, but RSI overbought on 60min and price is extended well above VWAP. Momentum still strong via MACD. The setup favors pullback longs over chasing, which aligns with the mean-reversion mandate from Step 2.
Step 4: Calendar Gate
| Event | Time | Impact | Status |
|---|---|---|---|
| FOMC Member Waller Speaks | 2:00 PM ET | Medium–High | ~4 hours away |
- No high-impact events within 30 minutes of now (~10:00 AM ET).
- The 10:00 AM data window has passed (current candles show NY session active since ~9:30 AM ET based on session data).
- Waller speech at 2:00 PM is far enough away — no calendar block.
Calendar Gate: CLEAR ✅
Step 5: LTF Entry Analysis
15-Minute Timeframe
- EMA Stack: All 5 bars show price above fast EMA, fast above slow — bullish ✅
- RSI: Sequence: 85.1 → 88.1 → 73.6 → 74.3 → 73.0 — overbought territory throughout. RSI is NOT below extreme for longs — it IS extreme. This is a FAIL for long entry confirmation.
- MACD: Histogram strong and positive (0.00025 → 0.00042 → 0.00043 → 0.00041 → 0.00037) — still positive but declining, suggesting momentum is waning on 15min.
5-Minute Timeframe
- Price Action: After the session high at 1.1849, price has been pulling back — a clear distribution pattern:
- 1.1849 high → dipped to 1.1830 → bounced to 1.1842 → selling back to 1.1825 area → last bars consolidating 1.1828–1.1833
- RSI: Dropped from 75+ to 54.7–57.8 range — normalizing from overbought, now neutral. This is healthier for a potential re-entry.
- MACD: Histogram has flipped negative (−0.00009, −0.00012, −0.00015) on 5min — short-term momentum is bearish/correcting.
- Fibonacci (5min): Pivot low 1.17919 to pivot high 1.1849. Key retracement levels:
- 38.2% = 1.18137
- 50.0% = 1.18205
- 61.8% = 1.18272
- 78.6% = 1.18368
- Price at ~1.1833 is near the 72% retracement — between 61.8% (1.18272) and 78.6% (1.18368)
- Key Structural Levels:
- Session high: 1.1849
- Trend Agent support: 1.1825
- Trend Agent invalidation: 1.1820
- 60min S/R resistance-turned-support: 1.18235
- NY session low: ~1.1825 (from 15min session data)
- VWAP: ~1.1799 (too far for immediate entry)
Current Micro-Structure: Price is pulling back from 1.1849 into the 1.1825–1.1835 zone. The 5min RSI has reset to neutral (~57), MACD is turning but not deeply negative. The 61.8% Fibonacci retracement at 1.18272 is the key confluence zone below, aligning closely with the Trend Agent support level at 1.1825.
LTF Entry Assessment: A pullback-long entry zone exists at 1.1825–1.18272 where the Trend Agent support, 5min 61.8% Fib, and NY session low converge. The 5min RSI reset supports a re-entry. However, the 15min RSI remains overbought (73+), which is a warning.
Step 6: Confluence Scoring
Setup Under Evaluation: Pullback Long at 1.1825–1.1827 Zone
| # | Confluence Factor | Status | Score |
|---|---|---|---|
| (a) | Macro Agent bias aligned (bullish), confidence ≥ 60 | FAIL — Macro Agent is lean_bear (62% confidence), opposing the long | 0 |
| (b) | Trend Agent aligned (bullish), confidence ≥ 60 | PASS — Bullish, 65% confidence | ✅ |
| (c) | DXY 5-day trend confirms direction (falling = bullish EURUSD) | PASS — DXY below 5D EMA, falling sharply | ✅ |
| (d) | 10Y yield trend supports (falling = long) | PASS — 10Y at 4.242 vs 5D EMA 4.276, falling | ✅ |
| (e) | 60min EMA stack supports direction | PASS — Price > fast > slow EMA across all bars | ✅ |
| (f) | Price at VWAP, session level, or Fibonacci on 5min | PASS — Entry zone at 61.8% Fib (1.18272) + Trend Agent support (1.1825) + NY session low | ✅ |
| (g) | 15min RSI confirms without extreme | FAIL — 15min RSI at 73, overbought | 0 |
| (h) | No high-impact event within 30 minutes | PASS — Next event (Waller) at 2:00 PM ET, ~4 hours away | ✅ |
Confluence Count: 6 of 8 → Rating: Medium-High (7/10)
Failed confluences:
- (a) Macro Agent opposes — lean_bear narrative (rally exhaustion, ECB dovish). This is the primary risk factor.
- (g) 15min RSI overbought — 73 on 15min. While the 5min RSI has reset (57), the 15min hasn't fully cooled. Price needs to pull deeper or consolidate longer for this to clear.
Step 7: Trade Setup
📊 Setup: EURUSD Pullback Long
| Parameter | Value | Rationale |
|---|---|---|
| Direction | LONG | DXY weakness + bullish trend + pullback into support |
| Entry Zone | 1.18250 – 1.18280 | 61.8% Fib (1.18272) + Trend Agent support (1.1825) + NY session low confluence |
| Entry Trigger | Bullish 5min candle close (engulfing or hammer) holding above 1.1825 with RSI turning up from ~50 zone | Do not enter on first touch — wait for a reaction candle |
| Stop Loss | 1.18140 (with 1-pip buffer: 1.18130) | Below Trend Agent invalidation (1.1820) rounded, below 38.2% Fib (1.18137), below the 60min S/R cluster at 1.18077. Stop = ~12–15 pips from mid-entry. This is approximately 1.2× the 60min ATR (0.001 = 10 pips), adjusted for the elevated 5min volatility regime. |
| TP1 | 1.18400 (~1.2R) | 78.6% Fib retracement zone + prior consolidation resistance at 1.1841–1.1843 |
| TP2 | 1.18490 (~2R) | Session high retest at 1.1849 — key Trend Agent resistance |
| TP3 | 1.18600 (~2.5R+) | Extension target above session high if breakout occurs — round number + next structural resistance |
| R:R Profile | TP1: ~1.2:1 · TP2: ~1.8:1 · TP3: ~2.5:1 | Meets minimum 1.5:1 on TP2. TP1 at structure is slightly below 1.25R but has strong structural backing. |
Volatility & Stop Calibration
- VIX declining (17.62 vs 5D EMA 18.44) = compressing volatility → tighter stops near structure are appropriate
- 60min ATR = ~10 pips (0.001); 5min ATR = ~8 pips (0.001, 0.072%)
- Stop at 1.18130 is ~14 pips from entry midpoint (1.18265), which is ~1.4× 60min ATR — appropriate for a pullback entry in a compressing environment
- Stop is below Trend Agent invalidation (1.1820) with buffer — structurally sound
Position Sizing Guidance
Given the agent divergence (Macro vs. Trend), the Trend Agent's explicit "REDUCE_SIZE" recommendation, and the overbought 15min RSI, reduce standard risk to 0.5–0.75% of equity rather than the normal 1%. The setup has solid structural underpinning but elevated mean-reversion risk at these extended levels.
⚠️ Risks
| Risk | Detail |
|---|---|
| Macro Agent Opposition | Rally exhaustion narrative has merit — 8-day run may be stretched. Distribution at 1.1849 visible. |
| 15min RSI Overbought | 73 on 15min; if price doesn't pull back to the entry zone and instead rolls over from current levels, the 1.1825 support could break on a deeper correction. |
| Late-Cycle Momentum | Trend Agent flags "TRANSITIONING" regime — this is not an impulse entry but a digestion-phase pullback play. Breakout to new highs (TP3) is lower probability. |
| Waller Speech at 2:00 PM | While 4 hours away, hawkish remarks could reverse DXY weakness. Consider closing or tightening by 1:45 PM ET. |
| Extended Above VWAP | Price is in upper 2SD on multiple timeframes. A VWAP mean-reversion to ~1.1799 is a tail risk if support breaks. |
❌ Alternate Direction — Short Setup Assessment
A short at session high (1.1849) was evaluated as a mean-reversion trade per the divergence rules. However, it fails confluence scoring:
- DXY falling = bullish EURUSD (opposes short) ❌
- 10Y falling = bullish EURUSD (opposes short) ❌
- VIX falling = risk-on = bullish EURUSD (opposes short) ❌
- 60min EMA stack is bullish (opposes short) ❌
Only 3/8 confluences pass for a short (Macro Agent aligned, price at session extreme, no event). Below the 5/8 threshold — no short setup.
Summary
| Setup | Pullback Long |
| Rating | Medium-High (7/10) — 6 of 8 confluences |
| Entry Zone | 1.18250 – 1.18280 |
| Trigger | Bullish 5min reaction candle at zone |
| Stop | 1.18130 (structural, below invalidation) |
| TP1 / TP2 / TP3 | 1.18400 / 1.18490 / 1.18600 |
| Risk Sizing | Reduced: 0.5–0.75% equity |
| Key Condition | Entry is conditional — price must pull back to the 1.1825–1.1828 zone. If price never reaches this zone and instead breaks above 1.1849, do not chase. If price breaks below 1.1820, the setup is invalidated — no entry. |
10 Evaluations — What Changed in 18 Minutes
The AI evaluated EURUSD-Pepperstone 10 times before reaching a final decision. Each step shows confidence, reasoning, and whether the model was ready to enter.
The entry trigger requires price to pull back into the 1.1825–1.1828 zone and print a bullish confirmation candle (engulfing or hammer) with RSI turning up from around the 50 zone. Current price at 1.18279 is technically at the top of the entry zone (97%), but the price action shows no pullback into the zone — price has been trading consistently in the 1.1830–1.1835 range across the last three completed 5min candles and is only now dipping slightly on the forming candle (low 1.18294). There is no bullish reaction candle confirmed yet; the forming candle is bearish and still incomplete. Additionally, RSI on the 5min is at ~55 and declining (not turning up), MACD histogram is weakening with negative readings, and the 15min RSI just crossed down from overbought — suggesting the pullback may still have further to go before a valid long trigger forms.
The entry trigger requires price to pull back into the 1.1825–1.1828 zone and print a bullish confirmation candle (engulfing or hammer) with RSI turning up from the ~50 zone. Current price at 1.1826 is within the entry zone, but the last three completed 5min candles were all trading at 1.1830–1.1834, well above the zone — price has NOT actually pulled back into the zone yet. The current forming candle shows a low of 1.18294, still above the zone max of 1.1828. Additionally, there is no bullish reaction candle confirming at the zone; the forming candle is bearish (open 1.18313, close 1.18306) and appears to be drifting lower without any reversal signal. The 15min RSI is coming off overbought at ~70 with an OB cross-down signal, suggesting continued downside pressure — we need to wait for price to actually reach the zone and print clear bullish confirmation before entering.
Price has pulled back into the entry zone (current price 1.18262, zone 1.1825–1.1828) which is positive. However, the entry trigger explicitly requires a bullish 5min reaction candle (engulfing or hammer) holding above 1.1825 with RSI turning up from around 50 — and the last completed 5min candle (14:05) was a strong bearish candle closing at 1.18254, below the entry zone minimum of 1.1825, with price also dropping below the 5min fast EMA. There is no bullish confirmation candle yet; the move is still selling off with declining volume and weakening MACD histogram, suggesting further downside pressure rather than a bounce. The 15min forming candle also shows a bearish structure closing near its lows with an overbought-to-neutral RSI cross-down, adding caution.
Price at 1.18251 is technically within the entry zone (1.1825–1.1828), but the entry trigger explicitly requires a bullish 5min reaction candle (engulfing or hammer) holding above 1.1825 with RSI turning up from around 50. The last completed 5min candle (index 3) was a strong bearish candle closing at 1.18211 — well below the zone — and the current forming candle (index 4) shows only a tiny 7-tick body with no bullish confirmation pattern. RSI on the 5min is at 49.5, which is near the 50 zone but not yet turning up, and MACD histogram remains negative and weakening. Price has been in a sustained pullback with consecutive lower closes, and there is no bullish reaction candle to confirm a bounce — entering now would be premature and against the stated trigger rules.
Price has pulled back into and through the entry zone (1.1825–1.1828) but is currently trading at 1.18253, which is below the zone minimum of 1.1825. The last two completed 5-min candles are bearish (closing lower), and the forming candle shows no bullish reaction pattern—no engulfing, no hammer. RSI on the 5-min is at ~49.5, sitting near the 50 zone but not turning up; MACD histogram remains negative and weakening. The entry trigger explicitly requires a bullish confirmation candle holding above 1.1825, which has not materialized—price is breaking below the zone with continued selling pressure, suggesting the pullback may extend further toward the invalidation level of 1.1820.
Price at 1.18252 is technically within the entry zone (1.1825–1.1828), but the entry trigger is NOT satisfied. The setup explicitly requires a bullish 5min reaction candle (engulfing or hammer) holding above 1.1825 with RSI turning up from around 50 — and specifically states "do not enter on first touch; wait for confirmation." The last three 5min candles show persistent bearish pressure: the first two were decisively bearish (closing near lows), and the third candle at 14:15 is a tiny-range doji with very low volume (19 ticks vs 51 SMA), showing no conviction. RSI on the 5min is hovering around 49-50 but has not turned up with a confirming bullish candle — there's no engulfing or hammer pattern. MACD histogram remains negative and weakening. Price is also below the 5min fast EMA (1.18273), confirming short-term bearish momentum. Without a clear bullish reaction candle, entering here would be premature.
Price at 1.18252 is just at the top edge of the entry zone (1.1825–1.1828), which is good positioning. However, the entry trigger explicitly requires a "bullish 5min reaction candle (engulfing or hammer) holding above 1.1825" and to "wait for confirmation at the zone." Looking at the candle data: the last completed 5min candle (index 3) was a small-bodied candle closing at 1.18253 — not a convincing engulfing or hammer pattern. The currently forming candle (index 4) shows virtually no body or momentum (open=close=1.18253, very low volume of 14). RSI is flat at ~50.6 with no clear upturn, MACD histogram remains negative and weakening, and price is below the 5min fast EMA — all indicating the pullback hasn't yet generated bullish confirmation. The setup needs a decisive bullish reaction candle before entry.
Price is currently at 1.18273, which sits within the entry zone (1.1825–1.1828) at 77%, and the pullback has occurred as anticipated. However, the entry trigger specifically requires a bullish 5-min reaction candle (engulfing or hammer) holding above 1.1825 with RSI turning up — this has NOT been confirmed yet. The last completed 5-min candle (index 3) was a small-bodied candle closing at 1.18253, below the entry zone, and the current forming candle (index 4) shows virtually no range with a doji-like structure (open=close=1.18253). RSI is flat at ~50.6 without a clear upturn, MACD histogram remains negative and weakening, price is below the 5-min fast EMA, and volume is very low — there is no bullish confirmation candle yet. The setup condition explicitly says "do not enter on first touch; wait for confirmation at the zone," and no such confirmation exists.
Price pulled back below the entry zone (1.1825–1.1828) and is currently at 1.18301 per the header but the 5-minute candle data shows price actually trading around 1.18253–1.18262, still below the entry zone minimum of 1.1825. The entry trigger requires a bullish 5-minute reaction candle (engulfing or hammer) holding ABOVE 1.1825 with RSI turning up from the 50 zone. While RSI is near 51 on the forming candle, the last completed candles show bearish price action (close at 1.18211 on index 2) with no clear engulfing or hammer pattern confirmed. The current forming candle is small and lacks the conviction needed for a proper bullish confirmation signal — price is still below the 5m fast EMA and MACD histogram remains negative, indicating weak momentum.
Price pulled back into the 1.1825–1.1828 entry zone (hitting a low of 1.18207) and has since printed a clear bullish reaction sequence: a bearish candle at 14:10 (low 1.18207) was followed by a bullish candle at 14:15 (close 1.18253 above open 1.18235) and a strong bullish candle at 14:20 (open 1.18253, close 1.18298, holding above the zone). The 5min RSI turned up from ~47 through 50.6 to 55.4, satisfying the "RSI turning up from around the 50 zone" condition. Current price at 1.18302 is now above the entry zone (173% zone position), which means we'd be entering slightly above the ideal zone, but the confirmation candle sequence is valid and price is now back above the fast EMA. The MACD histogram is still negative but narrowing, trend agent is aligned bullish at 63%, and no SL or TP levels have been touched — the setup remains intact though the entry is marginally chased.
Account Performance
Profit taken at TP1 — the full position is closed at the first target to keep results measurable and comparable across models.
1. Transitioning regimes are dangerous for pullback entries. The Trend Agent flagged the regime as "transitioning" — momentum rolling over on lower timeframes while the hourly structure held. Pullback longs in transitioning regimes have a lower hit rate because the very momentum needed to bounce from support is fading. Claude should weight the regime classification more heavily when deciding between enter and wait.
2. Ten evaluations should raise the confidence bar, not lower it. When the AI needs 10 passes to reach 68% confidence, the setup isn't clean. Claude's US30 short on Day 1 required 9 evaluations and worked. But two of three recent losses (EURUSD today, NAS100 yesterday) both had extended evaluation chains ending at 62-68%. The evaluation count itself should be treated as a contra-indicator above a threshold — perhaps 7+ evaluations require 75%+ confidence to proceed.
3. Macro tailwinds don't override intraday structure. DXY falling + yields falling + risk-on = bullish EURUSD on the daily timeframe. But the intraday 5m/15m momentum had already broken down. The macro provides the directional thesis; the intraday provides the timing. When they conflict, the intraday wins on a day-trading timeframe.
4. Three losses in four sessions — Claude's lead is evaporating. Claude started the season at +$3,026 after a perfect Day 1. The balance is now $50,653 after this loss. GPT lost both its trades today as well — both models had a rough day. But Claude's higher trade frequency means more exposure to these marginal setups.
The macro said buy EURUSD. The 5-minute chart said sell it. Claude sided with the macro. The stop hit in 35 minutes. This is the third time this season Claude has entered a trade where the higher timeframe thesis was valid but the intraday execution window was wrong. The lesson isn't that the macro analysis was bad — DXY did keep falling after the stop. The lesson is that intraday trading requires intraday timing, and a supportive macro backdrop doesn't protect against a poorly-timed entry on a transitioning regime. The posture for tomorrow: take only setups where the trend, the regime, and the lower-timeframe momentum all agree. No more forcing entries against rolling momentum. — Isaac, Senior Research Editor
Compare with Eduardo’s analysis →Methodology
Both AI models receive identical market data, identical infrastructure, and identical risk parameters. No prompt engineering. No human intervention. Standard API temperature (0.0). Trades executed on demo accounts with institutional spread conditions via Pepperstone Markets. Each model operates with a $50,000 starting balance and 2% risk per trade. All positions are closed at TP1 — the first take-profit target — to keep results measurable and directly comparable across models.
Forex pairs and gold (XAUUSD) have standardized pricing across brokers — the prices in this article will closely match what you see on your own platform. US index CFDs (NAS100, US30, US500) are different: each broker constructs its own index price feed, so entry prices, stop distances, and P&L figures for index trades are specific to Pepperstone Markets. All trades in this experiment were analyzed, executed, and settled on Pepperstone demo accounts using Pepperstone's price feed.
Why This Cannot Be Replicated in ChatGPT or Claude Alone
Copying the analysis prompt into ChatGPT or Claude will not reproduce these results. Neither model has access to live market data — and the data is the foundation of everything.
Every analysis session, SkyAnalyst AI assembles a structured data packet of 50,000–100,000 tokens per instrument from live broker APIs. This is not a price quote. It contains 5 hours of multi-timeframe candle data across 60-minute, 15-minute, and 5-minute charts — each candle carrying full indicator overlays: EMA fast/slow, ATR, MACD with histogram, RSI, volume with SMA, VWAP with standard deviation bands, and others. On top of that: session structure levels (Tokyo, London, New York highs and lows), Fibonacci retracement and extension levels, a rolling 5-day macro window covering the 10Y yield, DXY, VIX, NYAD breadth, oil, and gold — along with additional proprietary data layers, all formatted as structured JSON specifically designed for LLM consumption.
The model never starts from raw data. Before Claude or GPT sees anything, two proprietary SkyAnalyst AI agents — among other internal systems — have already processed the environment: the Macro Analysis Agent produces directional bias with confidence scores and tradeability ratings across intraday and multi-day horizons, while the Trend Authority Agent evaluates technical structure — EMA alignment, momentum, regime classification — and outputs direction, confidence, key levels, and invalidation prices. The trading model synthesizes what these agents and preprocessing layers have already evaluated. This multi-agent pipeline is what produces the quality of analysis shown in this article — a single prompt to a single model, no matter how detailed, cannot replicate what multiple specialized systems produce in sequence.
The goal is to emulate what a professional trader actually does: read the macro environment, analyze multi-timeframe technicals, identify a setup with defined risk, wait for precise entry conditions, and execute with discipline. SkyAnalyst AI provides the infrastructure that gives the trading model everything it needs to do this — live data, preprocessed context, real-time monitoring, and broker execution. This is not a chatbot experiment. It is an institutional-grade trading pipeline where the AI model is the decision-maker, operating under the same conditions and constraints a professional desk would demand.
Trading involves substantial risk of loss. Past performance is not indicative of future results. These are AI model results shared for educational and research purposes only. Not financial advice.
Stay in the loop
Want the framework behind these trades?
Get every trade analysis, weekly battle report, and the full AI Trading Playbook delivered directly to you.
- Daily Trade Analysis
- Weekly Battle Reports
- AI Trading Insights
- The AI Trading Playbook (free)