Strategy Analysis

What is a VWAP pullback buy on a yield-capped rally?

A VWAP pullback long is a continuation entry inside a contained intraday structure — a setup specifically designed for sessions where the structural read favors longs but the macro environment limits the upside expectation. The thesis is straightforward: the trend is technically up, an opening drive has just printed, and the volume-weighted average price band is the highest-asymmetry place to engage the dominant direction at a contained risk. The model waits for price to retrace into the band, watches for a 5-minute reaction signal, then enters long with the stop placed beneath the next structural level below. The targets sit at the next intraday extensions but with the upper bound capped by the macro headwind.

The setup resolves on three signals. A test of the VWAP zone — for this trade, the 27,760-to-27,780 band defined by the session VWAP at 27,775 and the 60-minute Fibonacci 78.6% retracement at 27,770. A 5-minute reaction signal: a candle close above 27,775 after touching the zone, with RSI above 50, MACD histogram positive, and EMA9 holding as dynamic support. And a higher-timeframe context that has not flipped — yields below the 5-day high (the framework's hard kill-switch on NAS100 longs), no high-impact USD event inside the next thirty minutes, and the Trend Agent invalidation at 27,714.8 untouched. When all three align, the trade is structurally valid at a moderated confidence.

The honest weakness of this setup — the same weakness embedded in the rate-sensitivity gate itself — is that the macro headwind sets a hard ceiling on the structural expectation. NAS100 trades in a yield-capped band when the ten-year sits above its 5-day EMA and approaches the 5-day high; the framework has documented this regime across the season, and the structural targets in this state are tighter by design. TP1 at 27,810 sits only thirty-five to forty points from the entry midpoint, which is roughly 0.5R rather than the standard 1R-to-1.25R structural-target placement. The framework's R:R rejection clause — "structurally inverted setups where TP1 delivers less than 1R and reaching further targets requires breaking through major levels" — is not triggered here because TP2 at 27,862 sits at 1.15R against today's session high and TP3 at 27,930 sits at 2.0R against the daily R1 area. The profile is acceptable, but only with a partial-profit management approach: scale at TP1, trail to breakeven, run the residual.

How Claude structured the trade

The pre-trade analysis ran 14,331 characters — the framework's full structural read with one qualified setup offered, one alternative scenario explicitly rejected on confluence count, and a comprehensive risk-management section that distinguished the structural stop from the hard kill-switch on yields. Five of the framework's seven confluence factors fired bullish for the long side. The Trend Agent ran bullish at 63% confidence under TRANSITIONING regime. The 60-minute EMA stack confirmed the technical uptrend with price above both EMAs. The 5-minute price reaction at the VWAP zone — bounce off the session VWAP at 27,775 with RSI back above 50 — fired the entry trigger. The 15-minute RSI at 55.1 with MACD histogram expanding positive at +8.74 confirmed the continuation. The macro calendar was clear of high-impact events.

Two confluences sat opposed. The 10-year yield direction was the structural headwind — yields above the 5-day EMA, breaking yesterday's high, approaching the 5-day high. And the Macro Analysis Agent ran lean-bull at only 56% confidence without citing rate factors in its bullish reasoning. The framework's process on a 5-of-7 setup where both missing confluences are rate-related is to take the trade at the lower band of the Medium-High range — roughly 6.5 on the 10-point conviction scale — with reduced position sizing and tighter structural targets. The yield kill-switch sat 0.8 basis points away. The framework wrote it explicitly into the analysis: "If US10Y breaks above 4.430% (5-day high), this long setup is immediately invalidated. No longs under any circumstances."

Setup parameters: entry zone 27,760-to-27,780, sitting at the 5-minute VWAP / 60-minute Fibonacci 78.6% confluence. Stop loss 27,705 — placed below the double-bottom at 27,716.5/27,721.3 with a 10-point overshoot buffer, providing approximately 75 points of risk from the entry midpoint, exceeding the framework's 1x 60-minute ATR floor of 60.6 points, and sitting just below the Trend Agent invalidation at 27,714.8. Take-profits at 27,810 (yesterday's high / 60-minute pivot high, ~0.5R), 27,855-to-27,865 (today's session high / 60-minute Fibonacci 100%, ~1.15R), and 27,920-to-27,940 (next structural zone / daily R1 area, ~2.0R). The risk-to-reward profile was 0.5R / 1.15R / 2.0R.

Risk on the trade: $509.49 against the $50,949.19 starting balance, exactly 1.0% — at the upper edge of the framework's REDUCE_SIZE band rather than the 0.75% the analysis suggested as a more conservative read. Position size 12.90 lots, reflecting the index-instrument calibration the framework applies on tick-value differences. The structural stop at the 60-minute ATR floor contained the risk envelope to the 1.0% target.

Eleven evaluations and thirty minutes of patience

Claude's evaluation log on this trade runs eleven rows over thirty minutes — the longest qualifier sequence the framework has produced this season. Six waits, then one enter at evaluation six, then four more waits, then the disciplined enter at evaluation eleven. Reading the sequence in order is the cleanest available illustration of how the framework processes a setup that sits inside a yield-capped regime where the trigger candle keeps drifting in and out of the entry zone.

Evaluations one through four, between 14:06 and 14:10 UTC, recorded waits at 45-to-48 percent confidence as the trigger candles repeatedly closed below the 27,775 threshold. Price held inside the entry zone but the bearish-completion candles at indices 3 (closing 27,767.2) and at 14:13 (closing 27,754) failed the framework's required confirmation. The 5-minute MACD histogram weakened, RSI drifted between 53 and 60, and the volume on the forming candles ran below the 20-bar SMA. The framework's rule on a setup of this grade — entry zone repeatedly tested, completion candle failing, momentum decelerating — is to wait for the trigger to print rather than relax the standard.

Evaluation five at 14:13 recorded the framework's first enter consideration at 48 percent confidence — and rejected it. Price had spiked to 27,801.9, sitting at 210% of the entry zone, twenty points above the band's upper bound. The forming candle showed a V-shaped bounce rather than a clean pullback-and-hold. The R:R math was the structural problem: stop at 27,705 implied 97 points of risk against TP1 at 27,810 only eight points away. The framework's chase rule fired: "entering 20+ points above the zone top offers a poor risk-reward with the stop at 27,705 (nearly 97 points away vs. TP1 only ~8 points away)." The decision held at wait.

Evaluation six at 14:15 recorded the framework's first enter at 66% confidence — premature, by the discipline standard. Price sat at 27,794.9, at 175% of the zone, with the forming candle bouncing back from 27,774.9 to above 27,775, RSI at 60.3, MACD histogram positive at +3.11. The decision logged enter, but the actual broker execution did not fire — the framework's downstream logic appears to have rejected the entry on the same R:R caveat the eval log noted explicitly: "the candle is still forming, price is slightly above the ideal entry zone which reduces R:R, the 15m MACD line remains below zero, and breadth is negative." The trade was not taken at this print. The framework continued to monitor.

Evaluations seven through ten, between 14:20 and 14:34, recorded four further waits at 45-to-48 percent confidence as price repeatedly extended above the entry zone (27,809 at evaluation seven, 27,797-27,807 at evaluation eight, 27,793 at evaluation nine, 27,801 at evaluation ten). The R:R math remained unfavorable across all four prints. The framework declined to chase. The 5-minute MACD histogram weakened from "strong" to "weak" across this window. The pullback structure was forming as the framework's analysis had anticipated, but the entry trigger kept printing at the wrong side of the zone.

Evaluation eleven at 14:36 recorded the disciplined enter at 64% confidence. Price sat at 27,793.4 — above the entry zone but with the forming 5-minute candle holding above EMA9 at 27,771.3 as dynamic support, RSI at 57.8, MACD histogram positive at +3.01. The most recent completed 5-minute candle had closed at 27,801.4 above the 27,775 trigger level. Price had pulled back from the NY AM high of 27,809.7 to 27,771.2 and bounced cleanly, confirming the VWAP pullback structure with price now above VWAP at 27,757. The framework's reasoning recorded the same caveats it had recorded across the prior ten waits: "volume is below the 20-bar SMA on the latest candle, the 15m candle is still forming, the regime is transitioning with only 62% bullish confidence, the R:R is compressed since price is already 13 points above the entry zone max, reducing the reward to TP1 (~17 points) against a ~88 point stop." The entry was not textbook-perfect. The framework took it because the structural reference at the entry midpoint had effectively printed and every macro factor that had been opposed at the start of the qualifier sequence remained in the same configuration.

That is the framework's trigger discipline working at the lower band of the Medium-High conviction range. Eleven evaluations across thirty minutes, ten waits, then one enter when the structural reference had printed and the alternative was to skip the trade entirely. The model declined to chase on the R:R math at four separate prints above the zone. It only entered when the residual upside to TP1 was acceptable against the stop the framework's structural rule required.

How the trade played out

Entry executed at 27,780 at 14:36:23 UTC. The 5-minute candle that printed at 14:40 closed at 27,802 — the trade twenty-two points in profit four minutes after entry. By 14:45, twenty-eight points up. By 14:50, the position had reached the analysis-side TP1 at 27,810 — yesterday's high / 60-minute pivot high, the framework's first structural target. Broker exit at 27,811.1, one tick past the TP1 reference, captured the structural scale the framework's R:R math had defined.

The recorded broker outcome: net P&L of +$348.30 (TP1), balance after this position $51,297.49, exit price 27,811.1, duration approximately fourteen minutes from entry to broker scale. The blended R-multiple on the analysis-side full potential, with TP1 hit and the residual stopped at breakeven (the framework's standard scale-out plan): +0.4R (TP1).

The honest read on this outcome is that the broker scale at TP1 captured the structural target the framework had defined, but the residual position did not run further. The position then exhausted before reaching TP2 at 27,855-27,865, with the analysis-side TP3 at 27,920-27,940 unreached. This is the expected outcome on a yield-capped rally: the structural bounce engages the contained pullback, the partial-scale at the structural target catches the move the framework had anticipated, and the residual fails to extend because the macro headwind asserts itself across the broader session.

The discipline check is intact. Stop placement was structural — below the double-bottom at 27,716.5 with the slippage buffer satisfying the framework's 1x 60-minute ATR floor, just below the Trend Agent invalidation at 27,714.8. Position size at 12.90 lots reflected the REDUCE_SIZE flag from the TRANSITIONING regime call. The trigger discipline held through ten wait evaluations across the thirty-minute qualifier window before the framework's required confirmation printed in completed form. The trade reached its structural TP1 in line with the pre-defined R:R math.

What this trade says about Claude's process

Three reads, in the order the data suggests.

First, the framework's REDUCE_SIZE rule held under the visible structural headwind. The yield kill-switch sat 0.8 basis points away from killing the setup entirely; the Macro Agent did not cite rate factors in its lean-bull reasoning; the Trend Agent flagged the regime as TRANSITIONING. The framework's process for that combination is to either skip the trade or take it at the lower band of conviction with reduced sizing and tighter targets. Claude took the trade. The 1.0% risk allocation on 12.90 lots — at the upper edge of the REDUCE_SIZE band rather than the 0.75% the analysis suggested — represents the framework's calibration on a setup where five of seven confluences fire and the missing factors are both rate-related. The structural stop at the 60-minute ATR floor contained the risk envelope to the 1.0% target. The trade's downside was bounded; the upside was capped by design.

Second, the eleven-evaluation qualifier sequence is the discipline signal of the day. The framework declined to chase a +20-point overshoot at evaluation five on the R:R math alone. It rejected the premature enter at evaluation six on the same caveats it had been recording for the prior thirty minutes. It waited four more evaluations as price repeatedly extended above the entry zone. It only entered at evaluation eleven when the structural reference had printed and the residual upside to TP1 sat at an acceptable ratio against the stop the framework's structural rule required. The eval log shows a model that did not relax its standards under the visible macro headwind. Process consistency, not stoicism, is what the framework required across the thirty-minute window.

Third, Day 15 is the methodical-strikes register that opens Week 4 after the Council reclaimed the lead at the end of Week 3. The NAS100 scale at TP1 for +$348.30 (TP1), paired with the parallel EURUSD short for +$1,003.11 (TP2), produces a session-net of +$1,351.41 — the second-largest single-session print of the season, behind Week 3's Day 13 / Day 14 cluster. GPT did not trade Monday's tape; the cross-model gap widens to $3,521.82. The framework's posture across Monday's tape is the same posture it has held across the prior fifteen sessions: identify the structurally-valid setup, place the stop at the structurally-correct level, scale at the framework's pre-defined target, run the residual. Day 15's NAS100 scale is the smaller of the two prints and the more disciplined of the two qualifier sequences. The session's larger arithmetic comes from the EURUSD trade. The session's process consistency comes from this one.

The April 23 NAS100 echo

One cross-reference frames Monday's structural scale. On April 23, Claude had taken a NAS100 long that stopped out for -1.0R (SL) — a structurally-clean setup that the tape did not pay. That trade's macro context had supported longs unambiguously: yields below the 5-day EMA, falling VIX, weakening DXY. The framework's structural read had been correct; the entry was at a confluence-rich level; the position stopped out on the timing function. May 4's trade is the inverse — same instrument, same direction, similar VWAP-pullback posture, but the macro context is the structural opposite. Yields are the headwind, not the tailwind. The framework's response is to take a smaller position, tighter target, and partial-scale at the structural reference rather than run the larger arc.

The implication for Monday's tape is that NAS100 trades two different structural distributions depending on the rate environment. Under falling yields, the index runs the larger structural arc and the framework can size to the standard 1.0% with TP3 sitting at 3.0R or beyond — the May 1 setup is the cleanest available illustration. Under rising yields approaching the 5-day high, the index trades a yield-capped band and the framework's structural targets compress to TP1 at 0.5R and TP2 at 1.15R — the May 4 setup is the cleanest available illustration of that regime. Both setups deliver positive R-multiples to the framework's structural target. The difference is the magnitude. The framework's job is to recognize the regime and calibrate the position accordingly. The market's job is to deliver the move within the band the regime defines.

The next NAS100 setup will use the same envelope against a different read.