A Structured Backtest Using EdgeRater on the S&P 100
When You Can’t Take Every Trade
In backtesting, trading strategies are typically evaluated by taking every signal. This allows us to measure the strategy’s statistical edge—its average performance over many trades. In real trading, however, that’s often not possible.
You may have multiple valid signals at the same time
But capital for only 2–4 positions
So the question is no longer: “Does this strategy work?”
It becomes: “Which trades should I take first?”
The 5DD Strategy
This study focuses on 5DD (Five Days Down), one of the core Swing Systems Template developed by Dr. Ken Long and implemented in EdgeRater.
What is 5DD?
A mean reversion strategy
Triggered after five consecutive lower closes
Designed to capture short-term rebounds after downside exhaustion
In simple terms: 5DD looks for stretched downside conditions where price is likely to revert toward the mean.
Where Logic Chain Comes In
This is where Dr. Ken Long’s Logic Chain framework becomes practical.
Logic Chain organizes decisions into: Market → Sector → Strategy → Symbol
Instead of treating all 5DD signals equally, we ask
Is the market supportive (Risk-On vs Risk-Off)?
Is the stock’s sector supportive?
Study Objective
We tested the following question:
When we can only take a few trades, does choosing 5DD signals that align with the market and sector make a difference in results?
Methodology:
We used EdgeRater’s batch testing framework to simulate multiple conditions:
Universe: S&P 100 (broad, noisy environment)
Strategy: 5DD
Exit: 10-day hold
Capital Constraints
1 MAX → only 1 position
2 MAX
3 MAX
4 MAX
This forces trades to compete for capital, just like in real trading.
Three Approaches Tested
1) Baseline
Take signals as they appear
No prioritization
2) Sector Alignment
Look at trades where the stock’s sector is favorable
3) Logic Chain
Look at trades where both market and sector are aligned
Understanding the Metrics
To compare results, we focus on two measures:
1) FE50 / MD50 - Reward to risk
FE50 (Final Equity at 50th percentile) → Median final equity
MD50 (Max Drawdown at 50th percentile) → Median drawdown
2) Expectancy — $ Per Trade
Average profit/loss per trade
Market & Sector Alignment Using RL30Slope Z
To apply Logic Chain, we use: RL30Slope Z-score to measure tremd strength:
Applied to:
Market (SPY)
Sectors (XLK, XLY, XLP, etc.)
(Explanation of RL30Slope Z is available in an earlier article)
Results: 5DD on the S&P 100
Within this test framework, the following pattern was observed:
Baseline (No Alignment)
FE50/MD50 ≈ 4.13
Expectancy ≈ $55 per trade
Sector Alignment
FE50/MD50 ≈ 9.35
Expectancy ≈ $119 per trade
Logic Chain (Market + Sector)
FE50/MD50 ≈ 11.85
Expectancy ≈ $170 per trade
Results show: Trades in aligned Market–Sector environments showed stronger outcomes within this study.
What Actually Improved
Logic Chain does not change the 5DD signals, It changes: How trades are prioritized when capital is limited
Observed effects
Higher average outcomes
Higher win rates
Higher expectancy
From Research to Practical Use
A practical interpretation of these findings is:
Generate all valid signals
Identify which trades occur in more favorable conditions
Use that information to:
Prioritize trades when capital is limited
Trades are not eliminated—only ranked differently
Extending Beyond 5DD
Inside OBUG, we tested Logic Chain across additional Dr. Ken Long Swing Systems in EdgeRater:
ORL (Over-Reaction Long)
CH (Channeling)
WO (Washout)
TS (Triple Screen)
MPRC (Max Pain Range Compression)
Different strategies showed different sensitivities to market and sector conditions.
Why This Matters for Traders
Most traders focus on: Entries, Indicators, and Signal tweaks.
But in practice: Performance is often influenced by how capital is allocated across multiple opportunities
Inside OBUG
This article represents some of the work being done inside the Owl Bundle User Group (OBUG). If this type of structured, evidence-based research resonates with you, you may find value in joining us HERE.
Disclaimer
This article is provided for educational and informational purposes only.
The concepts, strategies, and results discussed are based on historical backtesting and analysis, and are intended to illustrate differences between approaches within a specific test framework. These results do not guarantee future performance.