I've been checking out the APE article in TASC Apr 2006. It's pretty interesting. Here's a combination of the first 3 indicators mentioned. I can't figure out how to correct the Alpha reading to truly return a value between 0-1. I tried using a couple of different methods (including an external dll), but no success. So, what I wound up doing was normalizing the original method and plotting that instead. Here's what I came up with. Original [code:1:a0f0183825]{ User Input } i1:=Input("APE periods",2,252,20); i2:=Input("Plot smoothing periods",1,100,1); i3:=Input("Plot (APE=1 Ratio=2 Alpha=3)",1,3,1); { Calculations } APE1:=Max(Abs(H-O),Abs(L-O)); APEn:=Max(Abs(HHV(H,i1)-Ref(O,-(i1-1))),Abs(LLV(L,i1)-Ref(O,-(i1-1)))); APEratio:=APEn/APE1; APEalpha:=Log(APEratio)/Log(i1); { Plot } If(i3=1,Mov(APEn,i2,E), If(i3=2,APEratio, If(i3=3,APEalpha,0)));[/code:1:a0f0183825] Normalized [code:1:a0f0183825]{ User Input } i1:=Input("APE periods",2,252,20); i2:=Input("Plot smoothing periods",1,100,1); i3:=Input("Plot (APE=1 Ratio=2 Alpha=3)",1,3,1); i4:=Input("Normalize Alpha (No=1 Yes=2)",1,2,2); { Calculations } APE1:=Max(Abs(H-O),Abs(L-O)); APEn:=Max(Abs(HHV(H,i1)-Ref(O,-(i1-1))),Abs(LLV(L,i1)-Ref(O,-(i1-1)))); APEratio:=APEn/APE1; APEalpha:=Log(APEratio)/Log(i1); APEalphaN:=(APEalpha-Lowest(APEalpha))/Max(Highest(APEalpha)-Lowest(APEalpha),0.0001); { Plot } If(i3=1,Mov(APEn,i2,E), If(i3=2,APEratio, If(i4=1,APEalpha,APEalphaN)));[/code:1:a0f0183825]

Regarding the Rate of Return Time Frame, Chris Young states, "Although it is interesting to consider the potential returns for a trade, the real interest should be in the return rate for a trade. A trade that returns 5% in five days should be much more interesting tahn a trade that returns 5% in a year. The APE plot can be modified to look at relative return rates... I have taken the APE curves for the three stocks, subtracted 1.5% for commissions and slippage (as an example), and divided the result for each day by the number of days. For example, for the value of the five-day average peak excursion, I subtracted 1.5% and divided the result by five to give the relative return rate for five-day excursions." It sounds deceptively simple, and it may well be very simple, but trying to match anything that resembles Figure 7 on p. 46 eluded me. Sorry, I don't have a way to scan the figure for this thread. I think it would be a pretty easy Excel plot.

A third cousin once removed? Hmm... I know about them guys... Thanks, I'll look it over!

I cannot reproduce this. What security are you loading? What are your user input values?