KeltnerChannel conditional order did not trigger

i finally came up with the right code, thanks largely to you, Pete, for which i again thank you.  the solution is in the last line.  bueno!

# conditional orders solution for Keltners

input displace = 0;
input factor = 1.5;
input length = 20;
input price = close;
input averageType = AverageType.exponential;
input trueRangeAverageType = AverageType.exponential;

def shift = factor * MovingAverage(trueRangeAverageType, TrueRange(high, close, low), length);
def average = MovingAverage(averageType, price, length);
def Upper_Band = average[-displace] + shift[-displace];

plot buytrigger = close[1] > upper_band and close < upper_band;

I moved your question over to the "Strategy Guide" topic because this is where we find solutions for questions related to chart strategies and conditional order scripts.

Check the attached screenshot. I have replicated one of your three conditional orders and I used the "Study Alert" tool on Thinkorswim to display the true/false signals being generated by your script. I used the same ticker symbol and time frame as in your example screenshot.

Note the cyan colored line on the lower sub-graph of the chart. It displays a value of 1 when the condition is true and a value of 0 when it is false. Note this is a daily time frame and it is only showing the state of the entry signal as of the close of the bar on each day.

From this view, there is no way to tell how that signal evolved during the course of the day. Did the close of today's daily candle ever trade below the value of the keltner line? There is no way to know for certain because your script is designed to chase a moving target. That keltner line is changing throughout the day. All we see at this point is where the value of that line was at the close of the day. We don't get to see where it was during the course of the day.

I spot-checked a few time stamps from 7/30/25 using the OnDemand feature of Thinkorswim. It looks to me like the close was always above the kletner band for the entire session.

Mystery solved. Hopefully I have provided you enough training to be able to replicate these troubleshooting steps on your own.

If you would like my opinion on this matter. The method you have employed here is not suitable for back-testing. And historical signals you examine in the past do not reflect how those signals would have been executed on a live chart.

Why? Because you have constructed a signal which is based on the current bar's close and the current bar's Keltner channel.This will cause signals which repaint. They change during the course of the trading session. And you only get to see the final state of the signal on each bar and not the condition of that signal during the course of each daily candle.

If you want to fix this, to make it a proper fit for back-testing, you will update the script as follows:

close is less than keltorders ("factor" = 1.5). "buywhencloseisbelow" from 1 bars ago

But you will find the back-test results are very different than what you are using now. Be aware of that and complete your research before implementing this.