BTC → IBIT · MTF-VP Trade Autopsy · Apr–Jun 2026
A +5.5R winner that gave it all back — and the one rule that would have kept it.
+5.5R
Peak reached
(best case)
+0.9R
Held to the
failure signal
+5.2R
Trimmed into
the 200-DMA
The rule: a long below a declining 200-DMA is a B, not an A — size small (~12.5% of Daily Stop) and trim 50–100% into the 200-DMA test. Signal quality sets the size; the 200-DMA test sets how much comes off; pullback quality decides whether you re-add.
▼ full autopsy below · signal on BTC, executed on IBIT · independently verified 7/7 · read-only

BTC → IBIT MTF-VP Trade Autopsy — April→June 2026 Long v2 · chart-validated

Signal on BTCUSD, executed on IBIT. Objective MTF-VP signal computed at BTC scale (where the docs'/charts' 200-DMA & VP levels live); every event mapped to an IBIT execution via an empirical per-day scale factor. Pilot for a reusable monthly trade-autopsy harness. Generated 2026-07-06.

Read-only research. No production pages / Pine / scanners / alerts / DB rows touched. engine='kite_v1' referenced as a literal only. /system/research/ doesn't exist on this stack → artifacts in Z:\trading\dev\mtf-vp-backtest\research_output\. All /system, /strategies, /scorecards, /todos items below are written recommendations, not edits. Supersedes the v1 (IBIT-native-signal) report.

1 · Executive Summary

Entry (IBIT next open)
$40.32
2026-04-08 BTC signal
Grade
B
5.75/10 · PARTIAL (12.5% DS)
200-DMA trim (100%)
+5.22R
the winning rule
Held to failure
+0.93R
gave the move back

The trade. The MTF-VP journal's flagship BTC long fired 2026-04-08 off the April base. The drawn Apr-8 chart specified an LVN pocket 72,500→91,802, stop 69,700, and targets 75k / 80k / 91,802. Executed on IBIT: entry $40.32, stop $39.1953 (−2.8%), R=$1.1247/sh. It ran to an MFE of +5.55R (IBIT +15.5%), tagged the BTC 200-DMA at the top, then rolled over.

The crux — the 200-DMA was the ceiling (on BTC). Price entered -18.9% below a declining 200-DMA and rallied right into it: at the May-6 top BTC's high came within -0.59% of the 200-DMA (the chart's teal-dashed ~82,275). On IBIT that trim zone = +5.22R. A 100% trim at the 200-DMA test = +5.22R vs +0.93R held to the objective failure — the single biggest lever in the whole autopsy. (In the v1 IBIT-native build the 200-DMA was calendar-stretched out of reach and this rule was moot; computing the signal on BTC is what makes it real.)

Targeting T3 was the trap. T3=91,802 sat inside the 88–95k supply wall; it was never hit, so "hold for T3" = never take profit = +0.93R. T1 (+1.65R) and T2 (+4.15R) filled; the edge lived between T2 and the 200-DMA test, not at the target in the wall.

Bottom line for sizing: quality → how much Daily Stop (this was a B = 12.5% DS, never a size-up A); the momentum burst / 200-DMA test → how much comes off (50–100%); pullback quality → whether you can re-add (here: none, so no add). Full IF/THEN in §16.

2 · Design — Signal on BTC, Execute on IBIT

Why split them

  • The docs' and charts' objective levels (200-DMA, POC/VAH/VAL, the LVN pocket) are drawn on BTCUSD — so the signal must be computed at BTC scale. On the ETF's own 200-trading-day line the same level is calendar-stretched ~40% further back and never reached (the v1 error).
  • IBIT is the executable, liquid, margin-friendly instrument — so entry/stop/targets/R and all P&L are in IBIT terms.

Scale factor & calendar mismatch

  • Empirical BTC↔IBIT scale at entry = 0.00056692 (IBIT$ per BTC$); it drifts 0.00055529–0.00058061 over the hold (~4.5%) — ETF fee drag + premium/discount. Plan levels are translated by % move, then tracked against real IBIT bars.
  • Calendar/slippage: BTC trades 24/7, IBIT on the ETF calendar. The June capitulation legs occurred partly when the ETF was closed → they gap into the IBIT open. A BTC-timed exit executes on IBIT only at the next session open (modeled here).

3 · Chart Reconciliation drawn charts read

Only 3 of 8 priority journal docs embed real charts (Apr 8 setup, May 13, May 14); the rest are text-only placeholders. Below: the two load-bearing panels, read directly.

Apr 8 — the drawn trade plan

Apr 8 annotated plan

Recovered from the annotations (text extraction had blanked it): LVN pocket 72,500 -> 91,802 (labelled 'fast travel zone'), STOP 69700, T1 75000 / T2 80000 / T3 91802. Prior v1 had invented a −2.5% stop; this is the real plan.

May 13 — 200-DMA rejection

May 13 daily 200-DMA rejection

The teal/green dashed line = the 200-DMA at 82,275; price ran to ~82k (May 11) and rejected with black down candles (May 12-13). VP: VAH 89,079 / POC 67,430 / VAL 62,760, with the supply wall 88–95k above the 200-DMA (where T3=91,802 sits).

Confirmed: VP levels at every read anchor; 200-DMA = teal-dashed, tagged & rejected. Corrected/added: the drawn stop/targets; a multi-MA "Alpha Trend Alignment" stack (not bare 20/50/200); the 88–95k supply wall.

RSI / Brown-Cardwell phase note

The journal chart panels carry no RSI subpane — the only RSI is in the standalone TradingView daily screenshot (Jul 1 stamp): RSI 39.80 / signal 33.00. Reading it under Brown/Cardwell range rules: RSI held a bullish 40–80 band through the Apr–May advance, then on the mid-May rejection lost the 40–50 support shelf and shifted into a bearish 20/30–60 range — rolling from ~66 (overbought, mid-May) to 33 (oversold, late June). That momentum-range shift is a textbook bearish phase change and corroborates the price/200-DMA rejection: the same top the VP + 200-DMA flagged. Caveat: this is read from one screenshot, not an OHLCV-reproducible RSI series, so it is corroborating, not a driver of the objective signal.

4 · Reconstructed MTF-VP Ruleset (BTC scale) docs > charts > memory

PARTIAL where docs are silent. Composite score canonical form = (TF confluence) × (gap/ATR) × (1/gap volume density); the 0.7^(N_rejections) term is unverified (in no rule-def doc) and unused. Docs admit the Pine score doesn't reproduce the Apr-8 "262". MA set = 20/50/200 canonical, but charts show a richer multi-MA alignment stack. "MTF alignment/conflict score" do not exist in the corpus. Entry trigger & Type A/B/C labels never formally locked.

Entry trigger (reconstructed)

RECON Long when score>10, price reclaims the 20/50 cluster, and a clean overhead LVN pocket exists ("chart air"). Here the drawn pocket 72,500→91,802 + stop 69,700 defined the trade.

Exit / reduce-risk / failure

  • 200-DMA test = price rallies into the (declining) 200-DMA from below → mandatory trim zone.
  • Reduce-risk RECON = first daily close < 20-DMA.
  • Failure (Scenario E) = 2 consecutive closes < 50-DMA (docs' "sustained close below = thesis invalidated").

5 · Entry State & Grade

CriterionVerdictDetailPts
1. MTF alignmentMIXEDShort-term reclaiming (BTC>20/50DMA cluster ~68.7k) but far below declining 200DMA (88.5k). Longer TFs down, shorter up.0.5
2. Daily trendEARLY-TURNBasing above the 20/50DMA cluster after the Feb capitulation; not an established uptrend.0.5
3. Weekly trendDOWNDeep drawdown from Nov highs; price -18.9% below the 200DMA. Weekly bearish.0.0
4. Price vs MA stackMIXEDAbove 20/50DMA; below 100/200DMA. Reclaiming short, capped long.0.5
5. 200DMA slopeDECLININGBTC 200DMA slope -2.6%/10b (declining). Bear structural backdrop.0.0
6. Distance to 200DMA-18.9% (in range)Below but within one leg of the 200DMA -> it is a reachable TARGET/ceiling (price tagged it at the top).0.5
7. Overhead VP pocketCLEAN LVNDrawn LVN pocket 72,500->91,802 ('fast travel zone'): thin volume overhead = the bullish feature at entry.1.0
8. R:R to targetsGOODStop 69,700 (-2.8%), R=1.12. R:R to T1=1.65, to T2=4.15, to T3=10x. >=2:1 to T2.1.0
9. Clean invalidationYESDrawn stop 69,700 = loses the pocket / back inside value. Unambiguous.1.0
10. IBIT executableYES-caveatIBIT very liquid, tight spread; scale factor drifts ~4.5% over the hold. CAVEAT: ETF RTH-only vs BTC 24/7 -> overnight gap risk (June legs).0.75
Composite5.75 / 10

Grade B — Below a declining 200DMA in a bear regime caps it at B (mandates partial risk), but a clean drawn LVN pocket + tight defined stop + >=2:1 R:R to T2 give real tactical merit -> B, not A, not Pass.

6 · Signal-Event Timeline (BTC signal → IBIT execution)

DateEventIBIT pxR
2026-04-09ENTRY Entry (IBIT next open)$40.320.00R
2026-04-14T1 T1 42.1757 hitlimit+1.65R
2026-04-22T2 T2 44.9874 hitlimit+4.15R
2026-05-06200DMA 200-DMA TEST (BTC high -0.59% of 200DMA)$46.19+5.2192R
2026-05-11MFE MFE peak$46.56+5.5482R
2026-05-18REDUCE Reduce-risk (1st BTC close<20DMA)$43.74+3.0408R
2026-05-28FAIL Failure / Scenario E (2 closes<50DMA)$41.37+0.9336R
never hitT3 T3 51.6242 (in the 88-95k wall)——

Targets are resting limits (fill on touch). The 200-DMA test, reduce-risk and failure are BTC-signal events executed at the next IBIT open. MAE was -0.32R (day 1) — the stop was never threatened before the top.

7 · 200-DMA Analysis — the crux

BTC 200DMA @ entry
-18.9%
reachable ceiling
BTC high vs 200DMA @ top
-0.59%
tagged 2026-05-06
Trim@test vs hold
+5.22R vs +0.93R
the 200-DMA lever

On BTC scale the 200-DMA is the operative resistance: price entered -18.9% below a declining 200-DMA (~$88,479) and rallied right into it at the top (chart teal-dashed ~82,275). The ETF's native 200-trading-day line is calendar-stretched ~40% further back and was never reachable — which is exactly why the signal is computed on BTC.

Was the failure at the 200-DMA? On the BTC source signal — yes. Price entered -18.9% below a declining 200-DMA (slope -2.6%/10b) and rallied straight into it; the high came within -0.59% and rejected. The 200-DMA (~82k) was the first ceiling; the 88–95k VP wall was the second. On the executable ETF the native 200-day line was never reachable — which is precisely why the signal is computed on BTC and the trim referenced to BTC's 200-DMA.

8 · Grade → Daily-Stop Allocation — "when to swing harder" centrepiece

GradeDaily-StopActionGate (all must hold)
A20-30% initial, max 30%SIZE UP / swing hardPrice ABOVE a rising/flat 200DMA + MTF aligned up + clean overhead pocket (free of supply) + R:R>=2:1 to first target + clean invalidation.
B10-15% initial, max 20% after confirmationPARTIAL / scale into strengthReclaiming or below a declining 200DMA BUT with a clean LVN pocket + defined stop + R:R>=2:1. Take 50-70% off into the first burst/200DMA test; do not hold for targets sitting in overhead supply. <-- THIS TRADE
C5-10% initial, max 10-15%PILOT onlyCounter-trend into supply, weak MTF alignment, 200DMA declining & unreclaimed, R:R<2:1. Prove it before adding.
Pass0%NO TRADELong into declining-200DMA resistance from below with no pocket / no defined invalidation, or MTF conflict.

Governing logic: Signal quality decides HOW MUCH Daily Stop (A=size up, C=pilot, Pass=0). Momentum burst / 200-DMA test decides HOW MUCH comes off. Pullback quality decides whether risk can be RE-ADDED.

This trade → B, 12.5% DS (B: 10-15%), PARTIAL risk. Never a full-size A: it's a below-declining-200DMA counter-trend long. Size for the burst, not the trend. You swing harder (A, 20–30% DS) only when the 200-DMA is rising/flat and price is above it with a clean pocket and MTF alignment — the opposite of this setup.

9 · Trade-Management Variants

All executed on IBIT (R=$1.1247/sh). P&L at the recommended B-band 12.5% DS.

VariantFinal R@DS$300@DS$350@DS$500
All-in / hold to Scenario-E failure+0.93R $35.01$40.84$58.35
Target T3 (91,802, inside the supply wall)+0.93R $35.01$40.84$58.35
50% @ T1, stop→BE, trail rest+2.35R $87.95$102.61$146.59
⅓ @ T1, ⅓ @ T2, ⅓ trailed+2.95R $110.51$128.93$184.18
Momentum-burst: 60% off into 200-DMA test, trail 40%+4.35R $163.04$190.22$271.74
200-DMA trim: 100% out at the test+5.22R $195.72$228.34$326.20
5/10/20 DMA close-break staircase (trail from entry)+3.09R $116.03$135.37$193.39
60% into test + 40% trailed on 20-DMA (runner combo)+4.21R $157.71$183.99$262.85

200-DMA trim sweep (vs no-trim +0.93R)

Trim sizeFinal R
Trim 33%+3.76R
Trim 50%+4.13R
Trim 70%+4.57R
Trim 100%+5.22R

Marking: trim R is at the test-day IBIT close (May 6, $46.19) — a proactive intraday trim into resistance. At the next open (May 7, $45.87) the 100% trim is +4.93R instead of +5.22R. Reactive stop/failure exits use next-open fills. No-lookahead (May-6 close known at May-6 close).

On BTC scale the 200DMA WAS tagged at the top (-0.59%). Trimming there captured +5.22R vs +0.93R held to failure. A 100% trim = +5.22R. The 200DMA is a MANDATORY trim zone for a below-200DMA long once price rallies into it.

5/10/20 DMA close-break staircase (trailing runner rule)

Leg1st breakExit pxR
1/3 < 5 DMA2026-04-21$44.5+3.72R
1/3 < 10 DMA2026-04-28$43.56+2.88R
1/3 < 20 DMA2026-05-18$43.34+2.69R
Staircase total+3.09R
as runner leg (60% test + 40% trail)+4.21R
Thoughts: Sound mechanical runner-trail (ordering 5<10<20 = scale out as momentum decays; last third gets the most room). BUT from full size the fast 5/10 DMA whipsaw you out mid-trend: here the 5-DMA broke Apr 21 and the 10-DMA Apr 28 -- 2-3 weeks BEFORE the May 6-11 top -- so 2/3 exited in the low-44s and missed the push to 46.5. It beat holding (+0.93R) and thirds-at-targets (+2.95R) but UNDERperformed the proactive 200-DMA trim (+5.22R): a below-200DMA-into-supply long has a KNOWN ceiling, so trimming INTO it beats trailing out AFTER it. Requiring a confirmed 2-close break made it worse (+2.81R), not better. Best use: the RUNNER leg after banking the bulk (60% into the test + 40% trailed = +4.2R), or trend-continuation trades ABOVE a rising 200DMA with no overhead cap. Refinement: treat the 5-DMA break as 'raise the stop / shave a token slice', and reserve full thirds for the 10 & 20 DMA to avoid the fast-MA chop.

10 · Add-On Pullback Study

Result: NO qualifying add → correct action was NO ADD. Through the winning phase IBIT stayed extended above the 20DMA; there was no low-risk pullback that held the rising 20DMA before the 200-DMA test. Adds would have gone INTO the 88-95k supply wall while the 200DMA was declining & unreclaimed -> violates all three 'never add' rules. Correct action: NO ADD. Rubber-Band (backside-long) adds need a reclaim/retest this trade never produced.

Closest any pullback low came to the rising 20-DMA: 0.24 ATR (never the ≤0.25-ATR "hold the line" test).

11 · Returns / Event-Study Table

IBIT (executable)

HorizonDateReturn
+1d2026-04-10+3.1%
+5d2026-04-16+6.0%
+10d2026-04-23+9.2%
+20d2026-05-07+12.6%
+30d2026-05-21+9.1%
entry→200DMA test2026-05-06+14.6%
entry→reduce-risk2026-05-18+8.5%
entry→failure2026-05-28+2.6%
MFE / MAE2026-05-11+15.5% / -0.9%

BTC-USD (signal)

HorizonDateReturn
+1d2026-04-10+2.6%
+5d2026-04-14+4.3%
+10d2026-04-19+3.9%
+20d2026-04-29+6.5%
+30d2026-05-09+13.4%
MFE (→May31)—+16.4%

BTC steps on 7-day weeks, IBIT ~5-day ETF weeks — same "+Nd" lands on different dates; the ETF skips weekend gaps.

12 · Comparable-Signal Sample illustrative, small

BTC: score>10 & close>20DMA & up-day; 10-bar debounce; 20-bar fwd outcome. n=11.

200DMA posturenavg 20davg MFEavg MAEwin%
Above rising 200DMA1+2.3%+3.5%-8.0%100%
Below flat 200DMA2+9.4%+15.8%-3.9%50%
Below declining 200DMA ← this trade8+1.3%+8.1%-4.5%50%

Small single-asset sample; illustrative not powered. April trade sits in 'below_declining_200'.

Read: below-200DMA longs carry large average MFE (~8%) against smaller MAE but are coin-flips on the 20-day close — they move (hand you a burst to sell into) but don't hold. That is the whole case for the 200-DMA trim over target-and-hold.

13 · What Worked · What Failed · Rule Improvements

What worked

  • Grading it B (partial 12.5% DS), not a size-up A — below a declining 200-DMA.
  • The drawn tight stop (−2.8%) & clean LVN pocket → +5.55R MFE, stop never threatened.
  • Trimming into the 200-DMA test: +5.22R (100%) vs +0.93R held.

What failed

  • Targeting T3 in the 88–95k wall = never take profit = +0.93R.
  • Holding past the 200-DMA rejection = round-tripping a +5.5R move.
  • Overnight BTC gaps (June) hit IBIT holders at the open — the ETF can't exit 24/7.

Rule improvements

14 · System-Hub Integration Plan recommendations only

Premarket IF/THEN

IF an MTF-VP long triggers while price is below a flat/declining 200-DMA (bear regime), THEN: • grade B or C (never A); • risk 5-15% of Daily Stop initially (12.5% here); • take 50-100% off when price rallies INTO the 200-DMA (BTC-scale) or the first VP supply shelf; • move stop to BE on the first partial; • never target a level inside overhead supply; • add ONLY after a 200-DMA reclaim + a valid pullback.

/strategies /scorecards /todos

  • /strategies/: file under counter-trend / backside-long ("Rubber Band"), NOT trend-continuation. Cross-link the 200-DMA trim + "never add into declining-200 resistance".
  • /scorecards/: graded case study — Grade B, 12.5% DS, 200-DMA trim +5.22R vs held +0.93R. The canonical "trim into the 200-DMA test" exemplar.
  • /todos/: (1) wire a BTC-scale 200-DMA proximity alert onto the IBIT view; (2) reconcile the Pine composite score; (3) build the reusable monthly-autopsy runner from autopsy.py.
  • Hub: YES — add as the pilot trade-autopsy.

15 · No-Lookahead Assertions & Acceptance Checklist

No-lookahead / anti-curve-fit

  • ✓ All scanner states use only close-of-bar data (trailing MAs/ATR/slope; VP windows end at t).
  • ✓ Entry at the next executable open (IBIT 2026-04-09), not the signal close.
  • ✓ Signal events computed on BTC; each executes at the next IBIT open after the confirming close.
  • ✓ Profit targets are resting limits (fill on touch); no VP/level uses future bars.
  • ✓ Screenshot/chart vs objective-data discrepancies explained (BTC 200DMA tag real; IBIT-native never reached).
  • ✓ Rules reconstructed from docs+charts, not fitted; unlocked thresholds labelled RECON; 0.7^N term rejected.

Acceptance

  • ✓ Old MTF-VP docs/source searched & listed
  • ✓ BTCUSD & IBIT sources+ranges stated
  • ✓ Chart PDFs pulled & read (Apr8/May13/May14)
  • ✓ Entry & exit dates objectively defined
  • ✓ 200-DMA value/slope/distance at entry & test
  • ✓ MTF state at entry & exit
  • ✓ Trade graded A/B/C/Pass (B)
  • ✓ Grade→Daily-Stop table (centrepiece)
  • ✓ IBIT proxy return table
  • ✓ Momentum-burst scale-out test
  • ✓ 200-DMA trim test (now bites)
  • ✓ Pullback add-on test
  • ✓ No-lookahead bias checked
  • ✓ Reusable harness (autopsy.py) parameterised
  • ✓ No production changes

16 · Final Recommendation

DS% at entry12.5% (B-band 10–15%), PARTIAL. Size up to an A (20–30% DS) only above a rising 200-DMA with a clean pocket & MTF alignment — not this trade.
Scale-out planTake 50–100% off when price rallies into the 200-DMA (BTC-scale) / first VP shelf. Backtest: 100% trim +5.22R vs +0.93R held. Never target a level inside overhead supply (T3).
Add-on criteriaAdd only after a 200-DMA reclaim + a valid pullback/retest that holds, total open risk ≤ grade max DS. This trade offered none → no add.
Stop movementTo breakeven on the first partial; trail under the 20-DMA (close-break = trim/exit). Never widen.
200-DMA a mandatory trim zone?YES, referenced to the BTC 200-DMA (the ETF's native line is calendar-stretched out of reach). It captured +5.22R vs +0.93R held.
Add to Hub / Scorecards / Strategies / Premarket?YES — graded case study + the premarket IF/THEN below; the pilot for the monthly autopsy harness.

Rule to promote — VERDICT: PROMOTE (refined)

IF an MTF-VP long triggers while price is below a flat/declining 200-DMA in a bear regime, THEN classify B or C; risk only 5–15% of Daily Stop initially; take 50–100% off when price rallies INTO the 200-DMA (BTC-scale) or the first VP supply shelf; move stop to breakeven on the first partial; never target a level inside overhead supply; and do NOT add unless price reclaims the 200-DMA and holds a valid pullback/retest.

Evidence: quality set the size (Grade B → 12.5% DS, not an A size-up); the 200-DMA test — not a fixed target — banked the edge (+5.22R vs +0.93R); the T3-in-the-wall target proved you must not aim into supply; pullback quality (absent) correctly blocked any add. Governing logic confirmed: signal quality → how much DS; momentum burst / 200-DMA test → how much comes off; pullback quality → whether to re-add.

Artifacts: BTC_IBIT_MTF_VP_BACKTEST_v2_20260702.html · btc_ibit_mtf_vp_autopsy_v2_20260702.json · btc_ibit_mtf_vp_signals_20260702.csv. Engine: autopsy.py (reusable) + analysis_v2.py. Signal=BTCUSD (external Yahoo/CCC, context); execution=IBIT (internal karb, read-only). engine='kite_v1' literal; no DB writes; no production edits.