VALUE MANAGER
15.0No clear value dislocation from price path alone.
Mean-revert gap 45.6%.
BIST · 2026-10-28
RALYH — AVOID
Ral Yatırım Holding · operator [email protected]
Bu sayfa ne işe yarıyor?
Tek kararın tez, kanıt, replay ve settlement kaydı.
Hangi soruya cevap veriyor?
Bu karar hangi girdilerle alındı ve sonra ne oldu?
Kararda nasıl kullanılmalı?
Öğrenme için kullanın. Eski karar yeni sinyal değildir.
Hangi motora veri sağlar?
Outcome review · Learning Engine · Signal history
Scores
- prediction: 19.6
- confidence: 92.9
- opportunity: 35.6
- timing: 85.3
- portfolio fit: 5.0
Bull Case
RALYH · (Ral Yatırım Holding) · Financials · Holding: defensive bounce case only — 5d -40.8% / 20d -86.4%. Need reclaim of short-term structure before sizing. Lead model Ensemble (6 models) assigns p(up)=0.20; portfolio fit 5.0 with position not held.
Bear Case
RALYH: extreme vol (96.5%) and max drawdown 95.8% compress edge — false breakouts likely. Risk stack vol=100.0 · dd=100.0 · liq=0.0 · sector=100.0 → total 85.0.
Ensemble & Uncertainty
- consensus p(up): 0.196
- agreement: 87.1
- dispersion: 0.032
- calibrated conf: 0.929
- risk of wrong: 0.071
- data quality: 78.4
- 1d p10–p90: -0.003 → 0.030
- 5d p10–p90: -0.003 → 0.001
source: QBMosaic weighted ensembleproxy / heuristicas of 2026-10-28 09:35 UTC
Same-as-of model votes weighted by learned StockModel weights, then gated by calibration, agreement and data quality.
Point-in-Time Intraday Timing
- availability: INSUFFICIENT
- readiness: —
- session: 2026-10-28
- activity: LOW
- RVOL-T: 0.01
- deviation: -0.63pp
- Current intraday quality 2.1 is below 60
source: Yahoo Finance observed 5-minute barsproxy / heuristicas of 2026-10-28 09:35 UTC
Quality-, freshness- and sample-gated timing readiness. Volume, activity and deviation are non-directional and never change BUY into AVOID.
DQS contribution
daily 35.0 · models 25.0 · intraday 0.5 · settled 18.0
- Intraday average quality is 2.7
- Intraday diagnostics report 1 quality issues
- 3 intraday sessions are incomplete
Why Not?
No hard NO_ACTION gate fired.
Catalysts
Yahoo path features (n=90) · asOf=2026-10-28 · model=Ensemble (6 models) · p(up)=0.20 · regime=strong-down · vol=extreme · fit=5.0 · not held · sector=Financials · evidence=no verified research evidence
Invalidation
Invalidate if total market risk > 95.0, drawdown extends beyond 100.8%, or 5d return worsens by >3pp without 20d repair. Liquidity score > 25.0 also voids size-up.
Key Assumptions
Rule thesis over Yahoo closes (sample 90). Mean-revert proxy 45.6%. Intraday timing INSUFFICIENT; activity is non-directional. Verified evidence count 0; unverified imports are excluded. No live LLM; committee text is numeric templates. XK030/sharia screens are membership-only — not a fundamental audit.
Exit Conditions
Remain flat. Re-open only after total risk < 60 and regime leaves strong-down/mixed with rising p(up). Exit any residual on further DD extension.
Paper Position
Replay Closure
Decision Evidence Pack
No verified point-in-time research evidence was attached to this decision.
AI Committee
GROWTH MANAGER
15.0Growth case needs stronger directional odds.
p(up)=0.20 from Ensemble (6 models) · Holding.
MACRO MANAGER
15.0Intermediate trend pressure argues patience.
20d regime -86.4% with vol 96.5% · Financials. No verified financial/news/disclosure evidence.
TECHNICAL MANAGER
15.0Tape soft (strong-down) — wait for reclaim.
5d -40.8% / 20d -86.4%; mean-revert 45.6%. Intraday timing INSUFFICIENT; no eligible observed 5-minute context.
RISK MANAGER
10.0Risk budget strained — size down or avoid.
Total 85.0 · volR 100.0 · ddR 100.0 · ann.vol 96.5%.
SENTIMENT MANAGER
90.0Signal confidence supportive.
Confidence 92.9; decision AVOID. No verified financial/news/disclosure evidence.
CIO MANAGER
51.0Aggregate call: AVOID.
Opp 35.6 · fit 5.0 · risk 85.0 · not held.