Bruce I have been struggling trying to write a PCF that represents a possible MACD histogram turning point.  That is the histogram 2 days ago is higher than the histogram 1 day ago and the histogram today is higher than the histogram yesterday.  Seems like it should be simple but I am getting a little confused on the formulas for the MACD and that of signal line.  Can you help?

The exact formula would depend on your particular MACD Histogram settings. If you are using an Exponential MACD Histogram 12,26,9 then the formula would be:

XAVGC12 - XAVGC26 - XAVG(XAVGC12,9) + XAVG(XAVGC26,9) > XAVGC12.1 - XAVGC26.1 - XAVG(XAVGC12.1,9) + XAVG(XAVGC26.1,9) AND XAVGC12.1 - XAVGC26.1 - XAVG(XAVGC12.1,9) + XAVG(XAVGC26.1,9) < XAVGC12.2 - XAVGC26.2 - XAVG(XAVGC12.2,9) + XAVG(XAVGC26.2,9)

Understanding MACD

Thank you.

Hi Bruce,

I have struggling with how to write a formula for 10 EMA, 20 EMA and 50 SMA all three crosing one another. Your reply is highly appreciated. Thanks. 

If all three moving averages have to cross each other during the same bar, then the Condition Formula is relatively straightforward:

SGN(XAVGC10 - XAVGC20) <> SGN(XAVGC10.1 - XAVGC20.1) AND SGN(XAVGC10 - AVGC50) <> SGN(XAVGC10.1 - AVGC50.1) AND SGN(XAVGC20 - AVGC50) <> SGN(XAVGC20.1 - AVGC50.1)

If the crossovers can happen over more than one bar, you would need to determine the maximum number of bars over which the crossover can occur. The simplest version of such a formula would just replace all of the .1s in the formulas with this number of bars instead.