It took me 7 years to realize, and I will tell you in 2 minutes.
1. Most things that make money in markets are already discovered
2. Given that, if you think you discovered something new, I'd bet on two things:
a) Either you don't understand what really is you're extracting from the market
b) You made a mistake somewhere
a) Either you don't understand what really is you're extracting from the market
b) You made a mistake somewhere
3. Risk is never eliminated, always shifted somewhere else
4. The naive pursuit of risk elimination, will most likely make it morph into something, that is not as apparent as it could be
5. Keep risk as apparent as you can, that way you have a better shot at controlling it
6. Abnormal returns are usually derived from taking abnormal amounts of risk (extreme example: drug dealers)
7. Work on signal research from an ideas first approach
8. Backtesting is a mere sanity test, not a signal research tool
9. The data will tell any story that you want, if you beat it long enough
10. If you don't have a working model, focus on tasks with maximum impact rather than marginal improvements
11. Focus on what makes money today, not on future potential problems
12. Do useful things in uncompetitive places
13. Compete in places where your skillset is at least slightly above the average competition
14. Often complex theory leads nowhere or to marginal improvements.
15. Complex optimization requires a deep understanding of the underlying assumptions
16. Optimization comes at the expense of something. Usually of robustness.
17. Too many assumptions in your model increase the exposure to risk
18. There will always be a tradeoff between
Generated by Thread Navigator
Press ⌘ + S to quick-export