formula check

[creer]

hi alex

can you please have a look at this formula to make sure it is correct.

in most cases it seems to backtest properly but on APOL from feb 24th to may 25th the backtest doesn't take the trades where it should.

when applied as a study to the chart you can see where it should take the trade so it is clearly wrong in the formula somewhere.

thanks for your time alex as always.

kind regards

richard stanton

[ACM]

richard
Before moving forward with this formula you may want to read this post that deals with moving averages having a negative offset (ie shifted backwards) and the implications these have in backtesting.
Alex

[creer]

thanks alex for the response.

i do take the results with a grain of salt but i am still unsure why it works most of the time and misses a couple of trades altogether.

richard

[ACM]

richard
The first problem in all the Sets is that one of the conditions looks for the SMA3 to be >/< than the SMA18 on the current bar. That is not possible (other than in backtesting where the efs has knowledge of the future) in as much as on the current bar there is no SMA3 value. So all the conditions need to be shifted back one bar to reflect the -1 offset of the SMA3.
The second problem - which is probably the cause of why certain trades are not triggered - is that the two averages may at times be equal on the prior bar and because you only define >/< then the condition does not verify.
Try the attached revision of the efs and see if it does what you want. Note however that the backtest results will be somewhat different (and very likely not as "good" although more realistic)
Alex

[creer]

sorry for the delay in thanking you alex. i have been away.

i have revised the formula to the attached and i am using an envelope as a stop. i really need it to be a multiple of the ATR as in the builtinMAChannel.

is there any way you could help me on this.

perhaps the formula needs to be converted to editor so that the variables can be imputed on the fly so to speak.

i have no idea how to do that though.

thanks for any help you may be able to give alexis.

regards

richard

[ACM]

richard
As far as I can see the formula is working fine. You just need to reduce the percentage of the envelope from 5% to a value that is more appropriate for the security and interval you are using.
In the chart below for example I modified the percentage to 0.25.
With regards to how to create studies in efs (without the use of the Formula Wizard) a good place to start would be the Guide to Developing EFS Indicators which is available in the eSignal KnowledgeBase.
Alex