As you know it's a common practice, the validity of which I was trying to ascertain recently, for traders to "check" the status of indicators in higher time frames as confirmation of signals generated in the lower time frames.
Using a simple example of a price oscillator crossover as a trade trigger, ie., when the value of the oscillator is above zero on a 5 minute chart (closing basis).
Before taking the trade I want to check that the price oscillator is also above zero on a 15 minute chart at the exact point in time.
Below is an example of one technique of checking the higher interval period using (-1 )and (-2) and while this method eliminates the "forward looking" problem when backtesting (which inflates the profitability) it is based on indicator readings of "older"completed bars.
What technique would you recommend (getSeries?)in a strategy EFS running on a 5min chart/int, to check the value of the same price oscillator value on the longer 15 minute interval?
(similiar to a person bringing up a 15 minute after getting the 5 minute signal to visually confirm at that point in time that the oscillator is above 0, and if so initiating the trade?
If the backtester cannot accurately be used, is there anyway to test the profitability of multiple interevals using tick replay and specifying (0) and (-1) on higher level timeframes?
Thanks.
Using a simple example of a price oscillator crossover as a trade trigger, ie., when the value of the oscillator is above zero on a 5 minute chart (closing basis).
Before taking the trade I want to check that the price oscillator is also above zero on a 15 minute chart at the exact point in time.
Below is an example of one technique of checking the higher interval period using (-1 )and (-2) and while this method eliminates the "forward looking" problem when backtesting (which inflates the profitability) it is based on indicator readings of "older"completed bars.
PHP Code:
var myStudy15 = osc(10,20,false,inv(15));// 15min study
if(!Strategy.isLong()){
if( myStudy15.getValue(-2)<0 && myStudy15.getValue(-1)>0)
Strategy.doLong(etc etc)
}
}
(similiar to a person bringing up a 15 minute after getting the 5 minute signal to visually confirm at that point in time that the oscillator is above 0, and if so initiating the trade?
If the backtester cannot accurately be used, is there anyway to test the profitability of multiple interevals using tick replay and specifying (0) and (-1) on higher level timeframes?
Thanks.
Comment