ed The code is correct, as far as I can see. What you have coded is for the variable to have a value of 1 when the midband (which is the Pds EMA) crosses up through the previous bar's low. In the green circle, this appears to be the case. On the 18th, the midband blue line is higher than the low of the blackbar on the 17th. BTW.. I assume you intend doing more with this code, hence the existence of the Hiband and Loband variables, so I won't comment more on them, but the HH5 variable is more simply written, without changing the values as: HH5:= Cross(midband, Ref(L, -1)); MS will return a Boolean TRUE (1) or FALSE (0) based on the algorithm, so it has the same value. -- As an alternative to Cross, particularly for more complicated scenarios you might like to use a code more like: x:= midband>Ref(L, -1); x AND Alert(x=0,2); which is a lot more flexible and extensible. Hope this helps. If you have any problems, just shout! wabbit :

Thanks for the reply Wabbit but unfortunately the Black bar of the 17th has a low of 9.01 and the Blue line Mov() on the 17th has a value of 8.968 and there is a clear gap between them, so in fact there should not be a boolean "true" on the bar of the 18th as it clearly failed the test. thers shouls be no "true"over the 18th at all. The signals on either side are correct, having said that the problem appears more than once further along. Has ther ever been mention of a "bug"in the cross function? BTW you are correct it is part of a more complex test - I just stripped out the noise to focus on this anomaly. Ed

Thanks Jose & Wabbit I cut & pasted Jose's code into the indicator - swapped { Price bar crosses over EMA and clears it, yesterday } crossOver:=Ref(Cross(midband,L),-1); Gave me what I wanted - thanks Wabbit & Jose. It was my logic at fault Ed

Thanks Wabbit Some times it's hard to stop and get out of thinking in Tradestation mode. Cheers Ed