The assumption that complexity produces better trading results is one of the most persistent and damaging beliefs in systematic trading. After a three-year project comparing simple systems against AI-generated multi-indicator strategies, David Stendahl arrived at a counterintuitive conclusion: the results were essentially the same, and the simple systems had one important advantage the complex ones did not.
David is a longtime systematic trader and developer with deep experience in TradeStation and multi-market portfolio trading. He trades across dozens of markets using a combination of intraday, daily, and weekly timeframes, running momentum systems, pattern-based systems, and trend systems. This episode focuses on what he learned from that comparative research, what it takes to build strategies you can actually trade with confidence, and why understanding the premise of a system matters as much as the performance metrics.
Watch the full episode below, then read on for the complete breakdown.
The Three-Year AI vs. Simple Systems Project
During COVID, with extra time available, David and a developer spent three years running a systematic comparison. On one side: simple systems built the traditional way, from the ground up with clear logic – what he calls the “write it down on a piece of paper” approach. On the other: systems generated by AI that ingested large numbers of indicators and combined them to produce trading signals automatically.
The AI approach was fast, thorough, and produced genuinely good systems. The simple approach was slower but produced systems with clear internal logic. The finding: “The results were fairly similar. They’re both good quality systems. They were both stable.”
The net difference was comprehension. With simple systems, you understand what is causing trades to be generated. You can anticipate setups, see them developing, and feel confident when the signal fires. With AI systems, “you never know when it’s going to trigger. You’re just waiting.” That lack of comprehension creates subtle but important problems when the system hits a rough patch and you need to decide whether to keep trading it or investigate further.
The Three Types of Simple Systems
David categorises his simple strategies into three styles, each with its own logic:
- Momentum/mean reversion: Price has moved too far from a baseline measure. David uses value charts – a tool he co-created that measures how stretched price is from a normalised average – to identify when the market has gone too far and is likely to revert.
- Pattern-based: Looking for divergence conditions and other price patterns that historically precede reversals or continuations. He uses these extensively, noting that divergence-based signals are some of the most reliable he has found across multiple markets.
- Trend/breakout: Volatility-oriented breakout systems that identify when price is establishing a new directional move. Not moving average crossovers, but volatility-based breakout confirmation.
The key principle across all three: minimise variables and inputs. Fewer variables means a smaller universe of alternative equity curves. Fewer alternative curves means less risk of inadvertently drifting into a dangerous one.
Adaptive Systems: The One Concession to Complexity
David’s simple systems have one feature that adds complexity: they adapt to market conditions. Rather than using static parameter values, his systems adjust their variables dynamically based on whether the current environment looks like a bull market or a bear market.
“The underlying general concept is tried and true and pretty simple – nothing too dramatic. But they have to adjust themselves over time so that the variables are constantly adapting to whatever the environment is. That’s where the complexity may come from.”
This is a deliberate design choice. A system that was calibrated for 2019 conditions may not be well-suited to 2022 conditions. Rather than rebuilding the system, an adaptive mechanism handles the adjustment automatically.
Understanding the Premise Before Trading
The practical advantage of simple systems emerged most clearly when David described actually sitting down to trade. With a simple system, you can watch the setup developing. You can see the conditions approaching the trigger point. That anticipation creates a comfort level that helps you pull the trigger consistently when the signal fires.
He tells the story of a lunar-based strategy – buy on the full moon, exit on the new moon – that has outstanding backtested performance on the S&P 500. It is perfectly simple and perfectly comprehensible. But he would not trade it because he does not believe in the premise. “Knowing the logic and having a strategy you actually believe in – that makes a difference to whether you follow through consistently.”
A system with strong performance metrics that you do not understand or believe in will be abandoned at the first extended drawdown. A system with perhaps slightly lower metrics but clear, logical premise will be followed through.
Diversification Across Markets, Timeframes and Styles
David trades actively across a wide range of markets: stock indices, currencies, commodities, metals, energies, softs. He also runs across multiple timeframes – 45-minute, 60-minute, 90-minute, 120-minute, daily, and some weekly bars. The goal is to find where the market is making moves and have a system ready to capture them regardless of asset class or timeframe.
“I’ll go wherever the market is making moves. If I have to trade intraday or end of day, that makes no difference to me.”
This breadth of deployment means individual system drawdowns have limited impact on overall portfolio performance. Diversification across non-correlated markets and timeframes does more for equity curve stability than any amount of optimisation applied to a single strategy.
Simple Doesn’t Mean Outdated
When the question arises of whether simple systems can still work in modern markets – with better data access, more computing power, and more sophisticated participants – David’s answer is consistent with his research: yes, with some nuance.
“A lot of my systems have done well in equity markets, but the market has been bullish. Anybody could be making money. But if you look at cotton, coffee, some of the more volatile markets – my simplistic systems have fared well 20 years ago, 10 years ago, and they still work pretty well now.”
Related episodes
- How to build Mean Reversion trading strategies – Stefan Friedrichowski
- How to avoid trading strategies that degrade quickly – Timothy Masters
Want more on building stable systematic trading strategies? Subscribe to the Better System Trader podcast for weekly interviews with the world’s top systematic traders and quantitative researchers.
