Chapter 29 · Residual Honesty, a Girsanov Laboratory | Rulers Before Models

Rulers Before Models, Chapter 29: Residual honesty, a Girsanov laboratory. Covers: Subtracting the sample mean removes 15% of annual return on the strength of a 0.75-sigma estimate; hedging first turns the same drift into an 11.9-sigma object; Five per-lag residual correlations all clear 0.25 individually; the joint Ljung-Box statistic Q(5)=12.32 fails at p=0.031 Chapters: 0:00 The quietest class of failure 1:50 The density that has to integrate to one 3:30 Four procedures that claim to remove drift 5:00 Hedge first: 64 years of data becomes 64 days 6:40 Five clean correlations that are not clean 8:10 Zero autocorrelation, double the variance 9:40 The false-alarm rate that turns off the gate 10:50 A spread alone turns a price into a band 12:30 Beating a formula is not an economic claim 14:10 $1.80 crosses a CUSUM threshold on day two 15:30 A claim earns a card field by surviving the lab Book: Rulers Before Models, Chapter 29: Residual honesty, a Girsanov laboratory Series: https://blog.stokeonstocks.com/series... Informational and educational purposes only. Not investment advice. #RulersBeforeModels #StokeOnStocks #CreatorHarness