Connors and Raschke NR4 Historical Volatility System. Compares the 6 day volatility to the 100 day volatility. When this ratio drops below 50%, a buy stop and sell stop bracket the current price, expecting historical volatility to revert to the mean. Four day Narrow-Range (NR4) and Inside Day patterns are used to filter the trades to increase probabilities
Although yesterday was NR7 day Volatality is above 50% could be the reason? May be AW10 can clarify
Code:
Ticker Date/Time Buy Close VolRatio NR4day NR7day insidedayBuyStop ,SellStop,
NIFTY 7/27/2009 1 4572.300 0.53 1.00 1.00 0.00 4595.33 4530.02
Regards
ken
Thanks Ken for bringing in new dimension to NRx setup by using Historical Volatility. The system that u mentioned uses 100day Historical volatility to last 7 days volatilty to
filter our NR4 signals. My views on this are (I might be wrong..but open to discuss them here)
1) NR4 signals occur more frequently hence we need to have some mechanism for filtering them. Historical Volatilty approach is one such method.
NR7 doesn't occur so frequently and days range is also narrow.. so not sure if historical volatilty parameters need to be adjusted further.
2) Linda's backtesting results given in her book were before 2008. All volatility based systems busted last year due to extreamly high volatility (beyond 5 - Standard deviation) so
I would rather question the basic concept of using last years volatility data for filtering. Is it fair enough to use the numbers that worked in < 2 to 3 Std Dev market in market that is beyond this limit.
or vice versa.
3) I have not backtested NR4 + Historical Volatility based approach.. so can't comment on whether it works or not.. but quite agree with the concept. Will keep this as part of
my to-do list to test this later. Only concern is, this complicates the basic setup and brings in statistics into it. That goes against my basic philosophy of keeping things simple as far as possible.
Leave the complexity for others to handle.
Happy Trading.