Korva is a scroll of the internet’s best ideas.
Check every decision against history
The most common mistake among smart investors is having no game plan. A game plan starts with studying each decision: if it were made in these circumstances, how would it have worked in the past? The record also builds an understanding of how things work.
The next step is to turn the decision into a rule and program it into a computer. Then the rule can be run across the world to find everywhere the same circumstances exist, with a track record behind each one. A collection of those rules, made uncorrelated with each other, supplies the diversification.
The rules should be timeless and universal. If one didn't work over a long period in the past, the reason it failed then and would work now has to be understood before it is used.
Why do backtests mislead?
If many rules are tested on the same historical data, some will fit the past by chance and look strong without any real cause behind them. Researchers David Bailey, Jonathan Borwein, Marcos Lopez de Prado, and Qiji Zhu wrote a 2014 paper describing how this overfitting inflates the apparent results of backtested strategies. A rule backed by a clear explanation of why it works is less likely to be one of those flukes.
How many investing 'factors' are real?
Campbell Harvey, Yan Liu, and Heqing Zhu catalogued more than 300 published factors claimed to explain stock returns in a 2016 paper. They argued that many were probably false discoveries, since so many tests had been run on the same data. They suggested that new findings should clear a higher statistical bar than the standard threshold.
Why do active traders lag the market?
Brad Barber and Terrance Odean studied tens of thousands of household brokerage accounts in a 2000 paper titled 'Trading Is Hazardous to Your Wealth.' The households that traded the most earned the lowest net returns, trailing the market after costs. Their analysis pointed to overconfidence and frequent trading costs as the main drivers.
More from My First Million
