@Baron, I gave it another try. Giving up because I do not need people like Kevin policing my posts. You can delete my account. Kevin wants the forum for himself. If you do not block him, he deserves it.
@ronblack I'm not sure who you're responding to? I don't see any posts by Kevin in this thread. Also the reference you made to a previous thread has a post by Kevin but all the text has been deleted??
Let’s get back to the topic. This is more important than you think. Most researchers, even the big guys fail to consider it. This is bad. Really bad.
On a related note, Norgate's historical index constituents are supported in Wealth-Lab for anyone who wishes to test on data which is free from survivorship bias. -Eugene
Great news! By the way, another article in same blog with example where "CAGR drops to less than 8% from more than 20%" Also I found another interesting one using volume to rank sp500 stocks. Bias is high if not considering delisted with artificial gains in the order of 100% for some year that disappear after proper backtest.
Volume is a poor ranking technique and reeks of selection bias. Dollar volume/turnover is far more interesting. But I agree on survivorship bias - this is critical to any trading system. Anybody who attempts to analyze the market/create a trading system without it is a fool.