Three traders, three completely different approaches, and one shared question: what does it take to set yourself up for a better trading year? This end-of-year roundtable brought Kevin Davey (three-time World Cup Trading Championship finalist), Laurent Bernut (algorithmic short selling specialist), and Scott Welsh (trend following and mechanical systems) together to reflect on what 2021 taught them and what they were changing heading into 2022.
The conversation runs over 90 minutes and covers everything from the challenge of building short index strategies in a decade-long bull market, to using Python for algorithmic short selling, to identifying when a bear market transition is actually underway. Each guest brings a distinct perspective, and the overlap in their concerns reveals something consistent about what serious systematic traders were thinking at that inflection point.
Watch the full episode below, then read on for the complete breakdown.
Kevin Davey: getting better as a strategy builder
Kevin’s word for the year was “variable.” Markets that spent the first half grinding higher before the short side started showing opportunities in September 2021. His biggest focus heading into 2022 was not a specific strategy, but the process of building strategies faster and evaluating them with tougher criteria.
His logic: if you can test 50 strategies in the same time it used to take to test 5, you can afford to raise your pass/fail threshold without reducing the number of viable strategies you find. Tougher filtering for the same throughput of ideas means better strategies in the portfolio.
Kevin’s World Cup Trading Championship entry for 2021 illustrated the risk side of competition trading directly. He reached fourth or fifth place midway through the year with a primarily long NASDAQ strategy. The Omicron variant announcement the day after US Thanksgiving hit both his stock index positions and a crude oil long he had added to break into the top five. His observation: shooting for 150% return to finish top five almost by definition requires taking concentrated risk. When that risk fires against you, the outcome is predictable. The plan was sound. The execution caught bad timing.
On short index strategies: Kevin acknowledged the structural problem. Ten-plus years of bull market data makes it difficult to build and validate strategies for the short side. His approach was to develop short-only strategies that sit on the sidelines 90% of the time and activate only when specific conditions are met. Low frequency with specific triggers, rather than trying to find a short signal that competes with sustained bullish momentum.
Laurent Bernut: algorithmic short selling and the transition to bear markets
Laurent’s focus in this episode was the practical mechanics of algorithmic short selling and how to think about bear market transitions. He runs Alpha Secure Capital and has written extensively on the subject.
On scaling in volatile sideways markets: Laurent’s approach involves reducing position size as volatility increases rather than maintaining fixed sizing. The logic is that volatility-adjusted sizing keeps risk constant even as price swings become larger, preventing the account from getting disproportionately hit during choppy, high-volatility conditions.
On bear market timing: Laurent was direct that predicting the exact transition from bull to bear is not a reliable approach. What is more useful is monitoring specific structural signals: are shorts working when they should not be? Is the market rewarding selling overbought conditions in ways it was not before? Those behavioral changes in the market’s response to strategy signals often precede the formal confirmation of a bear market by enough time to act.
He also discussed using Python for algorithmic short selling, noting that the tooling around short-side research has improved significantly. The ability to systematically screen for and test short setups at scale, rather than relying on discretionary pattern recognition, changes the quality of the analysis available to systematic traders.
Scott Welsh: trend following, “V” moves, and the legend of purple valley
Scott’s section covered what 2021 looked like from a trend following perspective, with a focus on two specific challenges: the nature of “V” price movements and the question of small stops in fast markets.
On V moves: 2021 produced a number of fast, sharp reversals that filled and reversed within days or even hours. For trend following strategies built on slower moving averages or breakout confirmations, these moves are particularly damaging. The strategy triggers on the breakout, the price reverses immediately, the stop gets hit. The stop-and-reverse approach, entering in the opposite direction when the stop triggers, addresses this for some traders. Scott discussed the tradeoffs: you reduce whipsaw losses but add complexity and can end up with losing positions on both sides of the same move.
On small stops in volatile markets: the intuitive response to fast, large moves is to use wider stops. Scott’s experience points in the opposite direction for certain strategy types. Small, tight stops in volatile conditions can actually improve overall system performance when the alternative is watching a position run against you for a long time before exiting.
On doing more of what works: one of Scott’s clearest takeaways was the value of reviewing which strategies performed well in the past year and allocating more capital there, rather than chasing new ideas. Old strategies that have been working quietly in the background are sometimes more reliable than newly built systems with limited live track records. This mirrors Marsten Parker’s observation in an earlier episode about revisiting strategies that were mediocre for years and then suddenly performed strongly.
Shared themes across all three guests
Three patterns ran across all three segments of this roundtable:
- Short side is harder than it looks. Kevin flagged the data problem. Laurent flagged the timing problem. Scott flagged the behavioral problem. Everyone agreed the short side requires more thought than most traders give it.
- Process improvements compound over time. Kevin explicitly linked testing speed to strategy quality. Laurent linked systematic tooling to research quality. Scott linked honest review of what is working to allocation decisions. Each was describing a version of the same thing: the infrastructure of your strategy-building process matters as much as any single strategy.
- Diversification is not just about markets. All three guests maintained portfolios of strategies rather than betting on a single approach. Kevin across futures markets, Laurent across short and long approaches, Scott across trend following timeframes. The diversity of strategy types, not just instrument types, is what reduces the volatility of the overall trading account.
Related episodes
- How to detect a failing trading strategy with Kevin Davey
- Breaking trading rules for higher returns with Scott Welsh
- Overcoming broken strategies with Marsten Parker
Want to think more systematically about how to build and review your trading strategies?
Subscribe to the Better System Trader podcast for weekly interviews with the world’s top systematic traders and quantitative researchers.
