Bill, Yes, my eyes fooled me on the .png. I thought the bar two to the right was lower making it the pivot bar.
I did try it on longer timeframes also and did not get anything I could see using at this point. However, you did not ask me if I thought the method would work, just about drawing lines, so I will refrain from further comment on that subject. Best regards, Dennis On Jun 29, 2008, at 11:12 AM, bilbo0211 wrote: > Hi Dennis, > >> First of all the right hand (starting) example you posted did not >> look > to be drawn from the correct lower right hand point. < > > Your assuming there is a correct set of pivots to use, that is not the > case. Every triad of pivots is correct ( i.e., 2 pivot lows with the > intervening high and vice versa) in the sense that they can point to > significant pivot points; however, they don't all lead to significant > reversal points. > There are a lot of them (also remember some CIT points are pointed to > by more than 1 triangle) which is why I need to do a detailed analysis > to see how useful they can be. Duffy says significant pivot points and > trend lines that touch more than 2 pivots yield better results but I > cannot confirm that (yet). > >> I took a one minute ES chart and plotted as many as I could for the > last couple of days. I found just one that hit a CIT bar. I would > not bet a dime on these odds.< > > Not really a good choice. If you recall, I mentioned that some times > the reversal only lasts for as little as 1 bar. A 1 bar (or few) > reversal on a 1 minute chart is not likely to be significant; however, > trading off 10 minute bar CIT points on a 1 minute chart could work > out. Also remember that the precision with which the lines are drawn > is critical. > > Wavemechanic is right on target when he says this is not a trading > method, it is just another trading tool to be used in combination with > others. > > Wavemechanic also referred to Reversal Magic in his post and it does > appear to be partly based on the square and angle construction in > Duffy's book. > > The RM vendor posted some projected Forex currency hourly CIT points > in this forum message thread: > > http://www.forex-tsd.com/suggestions-trading-systems/4252-reversal-magic-trading-system-4.html#post194232 > > In the specific post that link points to, he demonstrates how you > could have traded traded NZD/USD on 2/29/08. The results were 7 > trades, 4 wins, 3 losses, 131 pips profit (over $1000 per lot traded, > I don't know the exact NZD/USD conversion rate, he also is likely to > have chosen this day because of the good results). > > This is an interesting post because ot gives some insight into how you > can interpret the CIT points for trading purposes. > > Bill > > > --- In [email protected], Dennis Brown <[EMAIL PROTECTED]> wrote: >> >> Hi Bill, >> >> You are right I did not understand that you were only talking about >> time axis. >> >> I have no theory that says it is a study with an edge. I can only >> come up with why you might be fooled into thinking it is. >> >> First of all the right hand (starting) example you posted did not >> look >> to be drawn from the correct lower right hand point. These are daily >> bars (I usually work in minute bars), but if you mark every bar that >> could be considered a CIT bar, you will find that there are a lot of >> them. If you plot every possible 3 point setup, you will also find >> that there are a lot of them, of which you only picked 3. Now if you >> look at the probabilities of pointing to a CIT bar, you might find >> that any random selection has about a 25% chance of hitting one >> (eyeball guess). The eye/brain has the wonderful capacity to filter >> out all the choices that don't fit the desired pattern. >> >> I took a one minute ES chart and plotted as many as I could for the >> last couple of days. I found just one that hit a CIT bar. I would >> not bet a dime on these odds. >> >> The theory I previously outlined to you has some logic to it, and it >> does have a statistical bias based on human behavior. >> >> However, I encourage you to test it out for your self. There may be >> something I am missing. >> No need for an AFL, just hand draw 20, then pick what price you would >> propose buying (like the open) and selling (like the close 2 or 3 >> bars >> later) at and see how it works. Fill out every case, without seeing >> into the future where they cross and then see how many are right. If >> you can not get about 75% right, you will have a hard time making >> money with this setup. >> >> Best regards, >> Dennis >> >> On Jun 28, 2008, at 10:21 PM, bilbo0211 wrote: >> >>> Hi Dennis, >>> >>> I think you missed the point of the CIT indicators. This is >>> strictly a >>> time based indicator it is not a S/R point. It points to a specific >>> bar on which the trend should change direction. >>> >>> The point where the 'trend' lines cross is usually not very close to >>> the price. It usually points to the exact bar where price changes >>> trend. >>> >>> That is why I said "it must be tapping into some underlying market >>> symmetry or cycle". >>> >>> Look at the 3 examples drawn on the current NDX daily chart pointing >>> exactly to 3 pivot high bars (black lines) on the file I just posted >>> in the files section (changeInTrend1.png). >>> >>> Bill >>> >>> >>> ------------------------------------ >>> >>> Please note that this group is for discussion between users only. >>> >>> To get support from AmiBroker please send an e-mail directly to >>> SUPPORT {at} amibroker.com >>> >>> For NEW RELEASE ANNOUNCEMENTS and other news always check DEVLOG: >>> http://www.amibroker.com/devlog/ >>> >>> For other support material please check also: >>> http://www.amibroker.com/support.html >>> Yahoo! Groups Links >>> >>> >>> >> > > > > ------------------------------------ > > Please note that this group is for discussion between users only. > > To get support from AmiBroker please send an e-mail directly to > SUPPORT {at} amibroker.com > > For NEW RELEASE ANNOUNCEMENTS and other news always check DEVLOG: > http://www.amibroker.com/devlog/ > > For other support material please check also: > http://www.amibroker.com/support.html > Yahoo! Groups Links > > >
