The idea is to program an indicator that would consider the range of error in determining the crossing of prices and a SMA of the close without the need of intraday data. The purpose is to use this indicator value as any other value calculated at close and use it as means to enter/exit in today's bar instead of tomorrow's. This artifice would allow me to search 1-2-3 bar setups with daily historical data.
As an example let us consider a 20 period SMA. The upper maximum value above today's close was (H-C)/20. The lower minimum was (C-L)/20.
If we consider today's range in % as: 100*(H-L)/C= 5%, it is interesting to note that the maximum range of possible crossing points fall within a relatively narrow 5% / 20 = 0,25% band. In numbers it would look like this (example).
H=10 doll
L=9,5 doll
error range: 0,5/20 = 0,025 doll (1/20 of the bar range)
To use this value in backtesting one could use the closing value of the SMA as the crossing point and consider the STDEV from it to have an idea of the size of possible error. Calculation for EMAs are not much more difficult.
Unfortunately I can only consider the idea theoretically. No scripting skills yet.
rgds
-geo-
As an example let us consider a 20 period SMA. The upper maximum value above today's close was (H-C)/20. The lower minimum was (C-L)/20.
If we consider today's range in % as: 100*(H-L)/C= 5%, it is interesting to note that the maximum range of possible crossing points fall within a relatively narrow 5% / 20 = 0,25% band. In numbers it would look like this (example).
H=10 doll
L=9,5 doll
error range: 0,5/20 = 0,025 doll (1/20 of the bar range)
To use this value in backtesting one could use the closing value of the SMA as the crossing point and consider the STDEV from it to have an idea of the size of possible error. Calculation for EMAs are not much more difficult.
Unfortunately I can only consider the idea theoretically. No scripting skills yet.
rgds
-geo-
Comment