I would  like a pcf for double top double bottom to use in a scan for v12.

How would I scan for double tops or double bottoms on Forex 15 min charts? Would it be any different on 1 hr, 4 hr or daily charts?

To be able to make the PCF, at least 3 things must be defined.  The first is what is the difference in price that should be allowed between the 2 bottoms/tops to define an equivalent price because usually prices don't match identically.  UP TO 10 PIPS

The second is how much time are you going to search (ie the number of bars between the bottom/tops).  I AM INTERESTED IN INTRADAY ON A 15 Min. CHART AS WELL AS MULTI DAY

And a third is how far back the first bottom/top occurred.  This assumes that the bottom/top is not today since, by definition, if it is today, you can't say this is the bottom or top since the trend has not reversed yet. I AM INTERESTED IN INTRADAY ON A 15 Min. CHART AS WELL AS MULTI DAY

TomTerrific may be able to create double top or double bottom formulas based on his questions and your answers, but I can't.

As stated in my original response, I personally do not think complex pattern recognition is practical in TC2000 using Personal Criteria Formulas and EasyScans.

Im also interested in finding double tops and bottoms in the 15 minute chart or 1 h chart.

i habe look at the links but the problem is that they use percentages which is not practical and cannot be applied to the 15m or 1h charts. 

Im looking for "M" and "W" patterns that have a minimum of 4-5 bars between them. the pattern does not have to be perfect, but it has to allow me to find stocks for further study.

please bruce can you help

thanks

kind regards

I really can't help. I will point out that there is only kind of a problem with using percentages though, because it depends on the answer to the question, "A percentage of what?"

You mentioned, "they use percentages which is not practical and cannot be applied to the 15m or 1h charts." This statement is generally going to be true only when applied to the type of percentage that doesn't translate well between time frames.

There are generally two types of percentages used in the formulas which attempt to identify these types of patterns. One of these translates just fine when changing time frames and the other one does not.

The Scanning for Stocks Printing a Double or Triple Bottom topic mentions both types of percentages at various points. Note that mentioning this topic as an example doesn't mean I think it is necessarily very good. I generally think most of the techniques are pretty bad at identifying these types of patterns (and don't know of a good way).

Let's take a look at the type of percentage that does translate well between time frames.

Note that it uses MAXH10.20 - MINL10.20 to determine the range over the 10 bars ending 20 bars ago. It then looks for where the highest and lowest prices over the most recent 10 bars are not as a percentage of a fixed price, but as a percentage of the range.

You mentioned, "they use percentages which is not practical and cannot be applied to the 15m or 1h charts." This statement is generally going to be true only when applied to the type of percentage that doesn't translate well between time frames. Let's take a look at an example that doesn't translate well.

This doesn't translate well between different time frames because the percentage used is of a fixed price (the highest high of the 10 bars ending 10 bars ago) instead of a range.

To "fix" it so it worked between time frames, you would need to base the percentage on a range of some sort instead of basing it on a percentage of a fixed value. I have no idea if 5% of the range of the 10 bars ending 10 bars ago will work (I suspect not and that it will be too narrow), but the conversion would resurt in the following.

C < MAXC10.10 AND C > MAXH10.10 - .05 * (MAXH10.10 - MINL10.10)