Numerical Computing specialist and author Timothy Masters joins us (in his only ever interview) to discuss trading strategy development and validation techniques, including:
• Why trading strategies fall apart in live trading,
• How luck impacts trading strategy results and the impact of selecting the 'best' trading system,
• Why trades in an out of sample dataset are NOT representative of future trades,
• 3 ways to use Monte Carlo Permutation tests in trading strategy development to avoid overfitting, evaluate strategy robustness and assess the quality of your strategy development process,
• The danger of using out of sample trades to compute the probability of drawdowns,
• Entropy and information of indicators and why it's important to strategy development,
• How stationarity really impacts trading strategies and what strategy developers can do about it,
• Using Walk Forward to determine how robust a strategy is in the market,
• Incomplete Beta Distribution to track strategy deterioration,
• Bootstrapping, granular Profit Factor, the predictive ability of technical indicators, counter-trend trading and much more.
► Discover more from Tim at http://timothymasters.info
❓Want weekly trading insights, expert interviews, Sunday Summaries and more❓
Join the free Better System Trader newsletter📰: https://bettersystemtrader.com/newsletter
⭐⭐⭐ Support The Channel ⭐⭐⭐
☕ Buy Me a Coffee: https://buymeacoffee.com/bettersystemtrader
*** Timestamps ***
00:00 Welcome
00:35 Tim Masters background
02:15 Introducing bias when selecting strategies
06:20 How luck impacts trading strategy results
08:35 Impact of selecting the 'best' trading system
10:48 Reason for strategies that fall apart in live trading
11:50 How to use out of sample data
12:40 Traders not following their trading systems
13:55 Using Monte Carlo permutation tests to avoid the "luckiest" trading systems
17:50 3 ways to use permutation tests in trading strategy development
18:14 Overfitting trading strategies to noise
21:00 How to avoid overfitting with Monte Carlo Permutation tests
23:28 Using Monte Carlo Permutation Tests to evaluate performance of an existing model
26:20 Evaluating your trading system development process using Monte Carlo Permutation tests
32:14 How many times to permute data and what to look for
33:30 Bootstrapping in strategy development and validation
34:20 Sensitivity testing of optimized parameters
35:31 Incomplete Beta Distribution to track strategy deterioration
37:33 'Unbiased' measures in strategy development
39:54 Why trades in out of sample dataset are NOT representative of future trades
41:29 Danger of using out of sample trades to compute probability of drawdowns
47:06 Profit Factor and calculate it at a fine granularity
53:10 Entropy and information of indicators and why it's important to strategy development
57:10 How to increase the entropy of indicators
58:38 Are technical indicators predictive and how can we tell?
01:01:12 How stationarity really impacts trading strategies
01:04:45 Using Walk Forward to determine how robust a strategy is in the market
01:07:30 Counter-trend strategies for markets with strong trends
01:12:54 The number 1 takeaway for strategy developers
🎓🎓🎓 MORE TRADING EDUCATION & CONTENT FROM BST 🎓🎓🎓
► Better System Trader website: https://bettersystemtrader.com
► Interviews With Pro Traders: https://interviewswithprotraders.com
► Smashing False Breakouts: https://smashingfalsebreakouts.com
► Trading Market Internals: https://tradingmarketinternals.com
► DPS Trading Technique: https://dpstradingtechnique.com
Disclaimer:
Trading in the financial markets involves a substantial risk of loss and is not suitable for everyone. All content produced by Better System Trader is for informational or educational purposes only and does not constitute trading or investment advice. Past performance is not necessarily indicative of future results.