I’m working on building systematic strategies and wanted to open a discussion with others doing similar work.
Rather than “signals that print money,” I’m interested in:
• Research workflows that scale
• Signal evaluation beyond backtest Sharpe
• Risk management and portfolio-level thinking
• Where people think retail quants still have an edge
What approaches have you personally moved away from?
What areas are you doubling down on now?
Would love to learn from different perspectives from students to industry folks.
Quant strategy discussion: what’s actually worth researching now?
byu/Dre_dev ininvesting
Posted by Dre_dev
1 Comment
I found a lot of useful stuff in the books of Marcos Lopez de Prado. The explains the scientific approach behind evaluating strategies: how to do backtests, what are common pitfalls, how to do feature importance, ensemble modeling, feature engineering, and other good stuff.
It is less about where to look, but more about how to test. You’ll throw away a lot of bad/noisy signals by using the right framework.
The best advice I have is to look on data that people don’t write about a lot, and data that you need to do a lot of preprocessing for. This way, you’ll end up with features that might not be priced in yet