- 1Our composite score (0-100) combines six independent factor scores
- 2Each factor is z-scored, winsorized, and weighted according to its academic evidence
- 3The weights are: Profitability 30%, Momentum 25%, Value 15%, Low Volatility 10%, Investment 10%, Short Interest 10%
- 4Higher scores mean the stock ranks well across multiple dimensions
- 5The system is fully transparent — no black boxes
#From Raw Data to Composite Score
Our ranking process follows five steps for each of the approximately 3,000 stocks we cover:
Step 1: Calculate Raw Factor Values
For each stock, we calculate six raw factor measures:
| Factor | What We Measure |
|---|---|
| Profitability | Cash-based operating profitability (Ball et al., 2016) |
| Momentum | 12-month return, skipping the most recent month |
| Value | Intangible-adjusted price-to-book (Arnott et al., 2021) |
| Low Volatility | 60-day annualized realized volatility (inverted) |
| Investment | Year-over-year asset growth (inverted — lower is better) |
| Short Interest | Days-to-cover ratio (inverted — lower is better) |
Step 2: Z-Score Normalization
Raw factor values are on wildly different scales. A profitability ratio might range from -0.5 to 0.8, while momentum returns range from -80% to +300%.
We normalize each factor to a z-score:
Z = (Value - Mean) / Standard Deviation
After z-scoring, every factor has a mean of 0 and standard deviation of 1. This makes them directly comparable.
Step 3: Winsorization
Extreme outliers can distort rankings. We winsorize z-scores at ±3 standard deviations, following Tukey (1962). This means any z-score beyond ±3 is capped at ±3.
This prevents a single extreme data point from dominating a stock's overall score.
Step 4: Weighted Combination
We combine the six z-scores using our fixed weights:
Composite Z = 0.30 × Profitability + 0.25 × Momentum + 0.15 × Value + 0.10 × Low Vol + 0.10 × Investment + 0.10 × Short Interest
Step 5: Percentile Ranking
Finally, we convert the composite z-score to a percentile (0-100): - Score of 95 = better than 95% of stocks - Score of 50 = exactly average - Score of 5 = worse than 95% of stocks
#Why These Weights?
Our weights reflect the strength and consistency of academic evidence:
| Factor | Weight | Rationale |
|---|---|---|
| Profitability | 30% | Most robust factor; works in all market regimes; largest risk-adjusted premium |
| Momentum | 25% | Largest raw premium; confirmed across 200+ years and 40+ countries |
| Value | 15% | Long history but weaker recent performance; improves with intangible adjustment |
| Low Volatility | 10% | Primarily a risk reduction factor; smaller but very consistent premium |
| Investment | 10% | Moderate premium; overlaps somewhat with profitability |
| Short Interest | 10% | Useful negative screen; limited by data availability |
These weights are static — they don't change based on market conditions. Research shows that static weights outperform tactical timing.
#Star Ratings
We assign star ratings using fixed score thresholds inspired by AQR-style factor research. Each tier is defined by the stock's composite score:
| Stars | Score Threshold | Rating | Meaning |
|---|---|---|---|
| ★★★★★ | Score 75+ | Strong Buy | Strong Buy tier |
| ★★★★☆ | Score 65–74 | Buy | Above-average |
| ★★★☆☆ | Score 50–64 | Hold | Average |
| ★★☆☆☆ | Score 40–49 | Reduce | Below-average |
| ★☆☆☆☆ | Score below 40 | Avoid | Avoid tier |
Unlike Wall Street analyst ratings (where 90%+ are "Buy" or "Hold"), our score-threshold distribution is transparent and informative. A 3-star rating means average — not bad, but not exceptional.
#What Makes a 5-Star Stock?
To earn 5 stars, a stock must rank well across multiple factors simultaneously:
- Strong cash profitability (high margins, efficient operations)
- Positive price momentum (recent uptrend)
- Reasonable valuation (not overpriced relative to fundamentals)
- Low volatility (stable, not erratic)
- Conservative investment (disciplined capital allocation)
- Low short interest (smart money isn't betting against it)
It's very rare for a stock to score poorly on one factor and still reach 5 stars. The composite score inherently favors multi-dimensional quality.
#The Bottom Line
Our composite score distills the wisdom of 46+ academic papers into a single, actionable number. It's not perfect — no model is — but it systematically identifies stocks with characteristics that have historically predicted outperformance.
#Further Reading
Last updated: February 1, 2026