Calculating Expected High and Low

[echelon4x]

Hi all,

I'm interested to know if anyone trading over 1-2 days makes an attempt to calculate the expected high and low prices of the next day's trading session.

The most obvious way to do this is to use conventional pivots and treat say R1 as the expected high and S1 as the expected low.

Another method is to use Tom DeMarks enhanced pivot calculations which also take into account the prevailing direction of recent price action.

Also, Elliot waves could be used, though I find those things so subjective to how you view the chart and over a short timeframe I don't think the technique will work well.

I've looked briefy into Gann's methods and although techniques such as the Gann wheel can be used to predict both price and timing on a longer term scale, I can't see how these techniques can be used for shorter timeframes - anyone who knows otherwise please let me know as I find Gann's work intriguing.

I'm sure that there are loads of other techniques which can be used, however you never seem to be able to find much information how to forecast future prices.

Given that the FX market is probably the most predictable market out there - it's got to be the best one to attempt any sort of forecasting.

Has anyone used any techniques other than the ones I've mentioned above ?

[7thSignalTrader]

Echelon/Nightcat:

Have either of you noticed that you cannot edit your own post in some threads on this board? I created a digital journal and the Motec server threw up two jpg files when I only used the IMG tags to post one (two are not needed). Went back to delete two of the three images and was not able to do so. So, three of the same exact image files are now showing in the journal.

Not very "sharp".

[7thSignalTrader]

BTW - does the Panel show up too large on your screen in the Journal section? It does on mine and I'll play with the pixel sizing on my end to see if I can't make it smaller.

But, if it fits most screens out there, then I will leave it alone.

[flatfish]

Demark HiLo Projections
R= ((H+L+C+L)/2)-L
S= ((H+L+C+L)/2)-H

[cbpetro]

Hello,

Part of my strategy utilizes Neural Nets as a guide for direction, entries and targets. I post majors daily at 00:00 GMT.

CB Petro's Intraday High & Low Price Forecast

You may find them to be useful and if not . There is a small discussion regarding the use of NN and some background info that you may find useful.

cb

[echelon4x]

Hi Guys,

Thanks for all your responses - I've got much more to say on this - back later as my time is limited right now.

7th - I would guess it's just the journals and due dilligence sections which are uneditable, just to prevent anyone from going back and editing posts for unscrupulous means. If you can make the screen smaller that would be handy, though scrolling's no great hardship if you're viewing in mode 800x600 - it's great to see the system in action from the cockpit even if the message is driven home in triplicate !

[DeltaGuy]

Originally Posted by flatfish

Demark HiLo Projections
R= ((H+L+C+L)/2)-L
S= ((H+L+C+L)/2)-H

I have slightly different forumula for DeMark Projections:

Doesn't paste well so here's the link to the foruma:
De Mark

from the Pivot Points page.

[JR97]

I've had good success picking tops and bottoms by using both DeMark projections, fibs, modified bollingers, and other stastical measures of price/time. But in the end, it's all subject figuring out which levels are likely to stick. I just look for confluence among the numbers. I'm trying to get my output to be smart enough to highligh the confluence.