TAUNUSSTAR

Backtest Results

Historical performance analysis of our trading strategies. All results are based on out-of-sample backtesting with no lookahead bias.

Performance Metrics Overview

Model Performance

Comprehensive performance comparison across all strategies

Metric
Position-Sized
Consensus
Buy & Hold
Annual Return
10.61%
10.66%
8.58%
Sharpe Ratio
0.82
0.80
0.46
Sortino Ratio
0.99
0.91
0.59
Calmar Ratio
0.38
0.39
0.15
Max Drawdown
-27.84%
-27.62%
-56.78%
Annual Volatility
12.26%
12.60%
18.07%
Win Rate
85.71%
82.61%
0.00%
Total Trades
14
23
N/A
Backtest Period
36.0 years
36.0 years
36.0 years

Cumulative Returns

Consensus, Position-Sized, and Buy & Hold strategies

Consensus:3734.87%
Position-Sized:3671.89%
Buy & Hold:1825.71%
Consensus vs B&H:+1909.16%
Position-Sized vs B&H:+1846.18%
Consensus Strategy
Position-Sized Strategy
Buy & Hold

Drawdown Comparison

Peak-to-trough declines (lower is worse)

Consensus Max DD:-27.62%
Position-Sized Max DD:-27.84%
Buy & Hold Max DD:-56.78%
Consensus Strategy
Position-Sized Strategy
Buy & Hold

Rolling Returns

252-day rolling annualized returns

Consensus Avg:11.79%
Position-Sized Avg:11.57%
Buy & Hold Avg:9.99%
Consensus Strategy
Position-Sized Strategy
Buy & Hold

Signal Distribution

Distribution of consensus signals across all time periods

Model Analysis

Model Prediction Probabilities

Confidence levels for each time horizon (50% threshold for LONG signal)

When probability > 50%, model signals LONG. When < 50%, model signals CASH.

Consensus Strategy

Historical backtest performance vs Buy & Hold

Annual Return

10.66%

vs 8.58% B&H

Sharpe Ratio

0.80

vs 0.46 B&H

Max Drawdown

-27.62%

vs -56.78% B&H

Outperformance:+2.08%/year

Win Rate

82.61%

Total Trades

23

Time in Market

0.00%

Position-Sized Strategy

Historical backtest performance vs Buy & Hold

Annual Return

10.61%

vs 8.58% B&H

Sharpe Ratio

0.82

vs 0.46 B&H

Max Drawdown

-27.84%

vs -56.78% B&H

Outperformance:+2.03%/year

Win Rate

85.71%

Total Trades

14

Time in Market

0.00%

Position Size History (Position-Sized Strategy)

Strategy allocation over time (0-100%)

Average Position: 66.6%

Position Size History (Consensus Strategy)

Strategy allocation over time (0-100%)

Average Position: 65.4%

Model Voting Pattern & Consensus Strength

Dot color reflects number of models LONG

0/3 LONG (0%)
1/3 LONG (30%)
2/3 LONG (90%)
3/3 LONG (100%)

Strategy Descriptions

Consensus Strategy

Binary approach that invests 100% when the consensus threshold is met.

Rules:

  • • ≥2 models signal LONG → 100% invested
  • • <2 models signal LONG → 100% cash
  • • Simple, easy to implement

Best for: Traders who prefer binary decisions and can tolerate higher volatility.

Position-Sized Strategy

Gradual approach that varies position size based on model agreement.

Rules:

  • • 3 models LONG → 100% position
  • • 2 models LONG → 90% position
  • • 1 model LONG → 30% position
  • • 0 models LONG → 0% (cash)

Best for: Traders who prefer gradual adjustments and smoother equity curves.

Model Training Insights

Model Training Metrics

Classification performance on test set and cross-validation results

HorizonAUCAccuracyPrecisionRecallF1 ScoreCV AUCFeatures
3M
0.776
0.691
0.859
0.664
0.749
0.656 ± 0.213
10
6M
0.869
0.776
0.879
0.804
0.840
0.667 ± 0.174
5
12M
0.927
0.829
0.965
0.811
0.881
0.856 ± 0.110
10
Good
Score ≥ target threshold
Fair
Score slightly below threshold
Needs improvement
Score significantly below threshold

AUC: Target ≥ 0.60 (higher = better predictive power)

Precision: Accuracy of LONG predictions (higher = fewer false signals)

Recall: Ability to identify opportunities (higher = catches more good periods)

CV AUC: Cross-validation performance (lower std = more stable)

Confusion Matrices

How well each model classifies LONG vs CASH signals

3M Horizon

Accuracy: 67.3%

True Negative
1678
18.5%
Predicted: CASH
Actual: CASH ✓
False Positive
1079
11.9%
Predicted: LONG
Actual: CASH ✗
False Negative
1875
20.7%
Predicted: CASH
Actual: LONG ✗
True Positive
4414
48.8%
Predicted: LONG
Actual: LONG ✓
Total predictions: 9,046

6M Horizon

Accuracy: 72.4%

True Negative
1420
15.7%
Predicted: CASH
Actual: CASH ✓
False Positive
1020
11.3%
Predicted: LONG
Actual: CASH ✗
False Negative
1475
16.3%
Predicted: CASH
Actual: LONG ✗
True Positive
5131
56.7%
Predicted: LONG
Actual: LONG ✓
Total predictions: 9,046

12M Horizon

Accuracy: 90.7%

True Negative
1581
17.5%
Predicted: CASH
Actual: CASH ✓
False Positive
388
4.3%
Predicted: LONG
Actual: CASH ✗
False Negative
450
5.0%
Predicted: CASH
Actual: LONG ✗
True Positive
6627
73.3%
Predicted: LONG
Actual: LONG ✓
Total predictions: 9,046

Cross-Validation Results

Train vs Validation AUC per fold - detecting overfitting

3M Horizon CV Results

5 folds completed

Mean Val AUC: 0.761
Gap: 0.0300 (Good)

6M Horizon CV Results

5 folds completed

Mean Val AUC: 0.840
Gap: 0.0366 (Good)

12M Horizon CV Results

5 folds completed

Mean Val AUC: 0.790
Gap: 0.1544 (High Overfitting)

Feature Importance

Top 10 most important features for each prediction horizon

Top 10 Features - 3M Horizon

Most important features for prediction

Top 10 Features - 6M Horizon

Most important features for prediction

Top 10 Features - 12M Horizon

Most important features for prediction

Backtest Methodology

Data & Approach

  • ✓Out-of-sample testing (no lookahead bias)
  • ✓Walk-forward validation
  • ✓Models trained on historical data only
  • ✓Signals lagged by 1 day (realistic execution)

Key Metrics

  • Annual Return: Annualized total return
  • Sharpe Ratio: Risk-adjusted returns (higher is better)
  • Max Drawdown: Largest peak-to-trough decline
  • Win Rate: Percentage of profitable trades

Want to learn more about our models and methodology? View detailed methodology →