What is the Flagium AI CoinTree Momentum Model?

Methodology Overview

The CoinTree Momentum Model is Flagium AI’s proprietary quantitative engine for measuring Institutional Capital Velocity and Structural Risk Divergence across the NIFTY 500 universe.

Rather than relying on lagging technical indicators, CoinTree normalizes price and volume returns into standardized cross-sectional Z-Scores. This allows risk managers, institutional analysts, and long-term investors to separate organic smart-money accumulation from low-volume retail bull traps.

---

1. Multi-Horizon Signal Construction

Rate of Change (ROC) Horizons

To capture both immediate momentum ignition and multi-quarter structural persistence, CoinTree evaluates Rate of Change (ROC) across 6 independent lookback periods:

---

Cross-Sectional Z-Score Normalization

Raw percentage returns vary wildly depending on broader market volatility. To enable direct comparison across sectors and time horizons, each ROC series is standardized relative to the entire NIFTY 500 universe:

$$Z = \frac{ROC - \mu}{\sigma}$$

Where:

• $ROC$: Individual stock return over the lookback horizon.
• $mu$: Cross-sectional mean return of all NIFTY 500 constituents.
• $sigma$: Cross-sectional standard deviation of returns.

A Z-Score of $+2.0sigma$ indicates that a stock's momentum sits in the top 2.5% of the entire Indian equity universe.

---

Composite Z-Score Weighting Scheme

The standardized signals are aggregated into a unified Composite Z-Score weighted intentionally to prioritize intermediate structural persistence while remaining sensitive to early momentum ignition:

---

2. Fundamental Quality Overlay (Forensic Safety)

High momentum on low-quality businesses is fragile. To ensure momentum is backed by real cash flows and balance sheet strength, Flagium filters CoinTree signals through an institutional quality matrix:

1. ROCE > 20%: Validates capital efficiency and structural pricing power.
2. Revenue & Profit Trajectory: Confirms price momentum is supported by fundamental earnings expansion.
3. Cash Conversion & Balance Sheet Health: Rejects companies exhibiting working capital stress or cash flow divergence.

---

3. Identifying Structural Risk Divergence ("The Silent Top")

The primary power of the CoinTree model lies in its ability to detect Bearish Momentum Divergence during Stage 3 market tops:

• Normal Institutional Uptrend: Stock price reaches a new 52-week high while CoinTree Z-Score expands ($Z > +1.5sigma$). Institutional buying is driving the rally.
• Bearish Risk Divergence: Stock price reaches a marginal new high, but the CoinTree Z-Score drops below $0.0sigma$ and A/D Flow turns negative.
• Forensic Significance: Smart money is actively distributing shares into retail buying. CoinTree flags this structural deterioration 3 to 6 weeks before the stock experiences a Stage 4 price collapse.

---

4. CoinTree Flow & Accumulation/Distribution Regimes

CoinTree integrates Close Location Value (CLV) and Relative Turnover Ratios (RTR) over a 21-day window to classify institutional flow regimes:

$$CLV = \frac{(Close - Low) - (High - Close)}{High - Low}$$

• Strong Accumulation ($Z > +1.5sigma, CLV > 0.5$): Institutions are buying near daily highs on expanding volume (Stage 2 Breakout).
• Silent Accumulation: Price remains rangebound while CoinTree Flow Z-Score climbs steadily (Stage 1 Base).
• Institutional Liquidation ($Z < -1.5sigma, CLV < -0.5$): Aggressive institutional selling (Stage 4 Downtrend).

---

5. Macro Market Breadth & Regime Analysis

In addition to single-stock scoring, aggregate CoinTree outputs reveal universe-wide market health:

• Positive Breadth (% Z > 0): Measures broad market participation.
• Leadership Depth (% Z > +1.0σ): Tracks the expansion or contraction of true market leaders.
• Regime Transitions: A contraction in Leadership Depth while benchmark indices make new highs signals an impending market correction.