π―
Best Composite Correlation
T+20
Ο = -0.0374
(p = 0.0001)
π
Cross-Ticker Spread (T+20)
1.3%
Win Rate: 60% Β· Long top / Short bottom
β‘
Strongest Component
Fundamental Trend
Ο = 0.1983 at T+20
π°
Best Score Level
Strong Bull (70+)
+$86.37 per $1K at T+20
π Quick Translation (ELI5 Glossary)
Because data science terms can be confusing, here is what this report actually means:
- Spearman / Pearson (Correlation): A score from -1 to 1. 1 means the score perfectly predicts the price going up. 0 means it's a random guess. -1 means the score predicts the exact opposite.
- p-value (p=): The chance this result is just pure luck. We want this to be below 0.05.
- Observations (n): The number of times we saw this matching pattern in our dataset.
- T+1, T+5, etc.: Trading days. "T+5 Return" is what the stock did 5 trading days (about 1 week) later.
- Score up crosses MA: Our bullish signal where the score breaks above its own 7-day average.
- Low score + 10pt rise: "Recovery Buy", the score was in the dumps (<50), but suddenly jumped 10 points.
- N/A Values: You will see N/A if there wasn't enough historical data recorded to test a specific scenario (common for newer metrics like Fundamental Trend).
1. Composite Score β Forward Returns
Does a higher PulseScore today predict higher returns?
| Horizon | Spearman Ο | p-value | Pearson r | Observations | Per-Ticker Mean Ο |
| T+1 β |
-0.0042 |
0.6254 |
-0.0042 |
13502 |
-0.0156 |
| T+5 β
|
-0.0254 |
0.0039 |
-0.0159 |
12902 |
-0.0832 |
| T+10 β
|
-0.0205 |
0.0235 |
-0.0099 |
12183 |
-0.1042 |
| T+20 βοΈ β
|
-0.0374 |
0.0001 |
-0.0081 |
10753 |
-0.1713 |
2. Component Predictive Power
Which sub-scores carry the most alpha?
| Component | T+1 | T+5 | T+10 | T+20 |
| Short Squeeze | -0.0045 | -0.0083 | 0.0090 | 0.0192 |
| Technical Momentum | 0.0005 | -0.0422 | -0.0309 | -0.0623 |
| Fundamental Trend | -0.0228 | -0.0513 | -0.1140 | -0.1983 |
| Broad Sentiment | 0.0401 | 0.0858 | 0.1126 | 0.0847 |
| Event Risk | -0.0057 | -0.0288 | -0.0844 | -0.0982 |
| Macro | 0.0268 | 0.0084 | 0.0090 | -0.0863 |
3. Score Level β ROI
$1K invested at each score level β what do you make?
| Score Level | Count | T+1 Avg | T+5 Avg | T+10 Avg | T+20 Avg | Best Win Rate | Best P&L ($1K) |
| Bearish (<30) |
4503 |
+0.19% |
+0.94% |
+1.84% |
+3.76% |
54.6% (T+20) |
+$37.60 (T+20) |
| Cautious (30-50) |
7048 |
+0.12% |
+0.63% |
+1.49% |
+3.19% |
51.2% (T+10) |
+$31.91 (T+20) |
| Bullish (50-70) |
2016 |
+0.17% |
+0.42% |
+0.94% |
+1.82% |
51.2% (T+1) |
+$18.17 (T+20) |
| Strong Bull (70+) |
85 |
+0.57% |
+0.51% |
+2.71% |
+8.64% |
60.0% (T+20) |
+$86.37 (T+20) |
4. Score Change Momentum
Do big score jumps predict price moves?
| Score Change | Count | T+1 Avg | T+5 Avg | T+10 Avg | T+20 Avg | Best Win Rate | Best P&L ($1K) |
| Big Drop (<= -4.0) |
3015 |
+0.08% |
+0.38% |
+1.65% |
+3.22% |
51.2% (T+20) |
+$32.23 (T+20) |
| Drop (> -4.0 to <= 0.0) |
5519 |
+0.15% |
+0.72% |
+1.39% |
+3.02% |
52.6% (T+10) |
+$30.25 (T+20) |
| Rise (> 0.0 to <= 4.0) |
2457 |
+0.03% |
+0.84% |
+1.48% |
+3.02% |
52.8% (T+20) |
+$30.24 (T+20) |
| Big Rise (> 4.0) |
2511 |
+0.34% |
+0.88% |
+1.66% |
+3.91% |
52.4% (T+10) |
+$39.11 (T+20) |
5. Score Trend Reversal
Lowβrising = buy signal, Highβfalling = sell signal
π’ Buy Signal (Score β Crosses MA)
| Horizon | Count | Avg Return | Win Rate | P&L ($1K) |
| T+1 | 1902 | +0.26% | 51.6% | +$2.55 |
| T+5 | 1820 | +0.72% | 50.8% | +$7.19 |
| T+10 | 1732 | +1.39% | 51.9% | +$13.91 |
| T+20 βοΈ | 1518 | +2.99% | 50.8% | +$29.89 |
π΄ Sell Signal (Score β Crosses MA)
| Horizon | Count | Avg Return | Drop Rate | Avoided Loss |
| T+1 | 2017 | +0.13% | 48.2% | $-1.29 |
| T+5 | 1918 | +0.64% | 49.3% | $-6.36 |
| T+10 | 1789 | +1.57% | 46.9% | $-15.74 |
| T+20 βοΈ | 1595 | +2.95% | 48.3% | $-29.47 |
π΅ Recovery Buy (Low Score + 10pt Rise)
| Horizon | Count | Avg Return | Win Rate | P&L ($1K) |
| T+1 | 1381 | +0.27% | 51.3% | +$2.74 |
| T+5 | 1297 | +0.69% | 50.7% | +$6.94 |
| T+10 | 1245 | +1.92% | 52.7% | +$19.16 |
| T+20 βοΈ | 1058 | +3.78% | 52.3% | +$37.79 |
π’ Red to Higher Tier (3 Days < 30, Break Above)
| Horizon | Count | Avg Return | Win Rate | P&L ($1K) |
| T+1 | 401 | +0.24% | 49.1% | +$2.38 |
| T+5 | 377 | +1.06% | 52.0% | +$10.63 |
| T+10 | 365 | +2.08% | 50.7% | +$20.84 |
| T+20 βοΈ | 323 | +3.53% | 55.4% | +$35.31 |
π΄ Top Tier Drop Sell (3 Days β₯ 70, Drop Below)
| Horizon | Count | Avg Return | Drop Rate | Avoided Loss |
| T+1 | 4 | -0.78% | 50.0% | +$7.79 |
| T+5 | 4 | -0.52% | 50.0% | +$5.17 |
| T+10 | 4 | +1.90% | 25.0% | $-18.98 |
| T+20 βοΈ | 4 | +9.38% | 25.0% | $-93.78 |
6. Cross-Ticker Ranking (Long/Short)
If we bought the top 20% of tickers daily and shorted the bottom 20%. A positive spread means our top picks beat the bottom picks.
| Horizon | Top Quintile | Bottom Quintile | Spread | Win Rate | L/S P&L ($1K) |
| T+1 |
+0.12% |
+0.27% |
-0.15% |
54.0% |
$-1.53 |
| T+5 |
+0.75% |
+1.09% |
-0.34% |
49.6% |
$-3.40 |
| T+10 |
+1.73% |
+1.71% |
+0.02% |
51.6% |
+$0.23 |
| T+20 βοΈ |
+3.80% |
+2.51% |
+1.29% |
60.2% |
+$12.95 |
Score β Price Overlay
Top tickers by data coverage β normalized score vs price
Per-Ticker Correlation Distribution
π Recommendations
Increase weight for Fundamental Trend β strongest predictive correlation at T+20 (Ο=0.1983).
Review Short Squeeze β weakest signal (Ο=0.0192) at T+20. Consider reducing weight or improving data quality.
Score reversal buy signal is actionable β 50.8% win rate at T+20. Consider surfacing in the UI.
Strong Bull (70+) is the optimal entry zone β best risk-adjusted T+20 returns.