This project explores the efficacy of a "Quality-at-a-Reasonable-Price" (QARP) ranking engine. By combining Value (Earnings Yield) and Quality (Return on Capital), the goal was to determine if a systematic selection framework could generate Alpha (excess return) over a broad equity universe.
To ensure institutional-grade integrity, this research implements:
- Point-in-Time Realism: A mandatory 90-day publication lag to simulate real-world data availability.
- Sector-Neutrality: Stocks are ranked only against their industry peers to avoid comparing disparate business models (e.g., Banks vs. Software).
- Annualized Strategy Spread (Alpha): +1.80%
- Information Coefficient (IC): -0.0129
The engine successfully identified a 1.80% Alpha Spread, proving that the "Top Quintile" of stocks fundamentally outperformed the "Bottom Quintile."
While the negative IC indicates that the factors were not perfectly linear across the entire universe during this period, the positive spread confirms that the model is effective at the "extremes"—it successfully filters out low-quality "junk" stocks while highlighting top-tier opportunities like AAPL, APD, and COR.
The following chart illustrates the "monotonicity" of our factor. By dividing the universe into five groups (Quintiles) based on their Factor Score, we see a clear performance advantage in the top-ranked group.
- Market Regime Sensitivity: The negative IC (-0.0129) suggests that during the specific periods in the dataset, the "middle" of the market was volatile. However, the 1.80% Spread confirms the strategy is highly effective at identifying extreme winners and avoiding extreme losers.
- Factor Decay: The results indicate that "Quality" and "Value" are persistent but not linear; they act as powerful filters for stock selection rather than precise price-target predictors.