|
PureBytes Links
Trading Reference Links
|
Hi b,
you may see the relative performance of StochRSI with and without the
magic OUT3 condition [Y<=2].
I have posted both formulas.
The OUT3 condition, in general, improves various systems for this
period of 30 last months.
I have not searched this condition for other periods and I have not
any reason to do it.
The way it acts is to cancel some wrong trades, if you examine
carefully trade by trade.
For example, it helped me to avoid a wrong entry 8 days ago, when
some indicators gave a weak "buy".
Note also that it is a composite ticker for the whole N100 market.
The synthesis and the behavior of this set was quite different the
90s. It is not so safe to run composites with the hypothetical recent
synthesis 5 or 10 years ago [this is really ..."artificial"]
If the +240% is interesting to go to +450% or +650% you may use it.
Else, you may trade with +240%. It is excellent, in one way or
another.
The StochRSI is not a universal system [try the whole N100
backtesting and you will see the results].
The magic OUT3 is a universal condition and will help you to avoid
some wrong [and not profitable] entries, due to the [quick] StochRSI
whipsaws.
DT
--- In amibroker@xxxx, "b519b" <b519b@xxxx> wrote:
> Dimitris,
>
> Thank you for posting the code for your OUT123 composite (based on
> the N100 stocks). I have played around with it with the stoRSI on
> the QQQ and the ^NDX. The ^NDX was used to see how the QQQ likely
> would have behaved prior to its creation in March 1999. Here are my
> observations:
>
> 1. It really does well since mid 2000 for QQQ and ^NDX data.
>
> 2. It did very poorly from 1998 to mid-2000. This was tested on the
> ^NDX data from Yahoo. The drawdown was about 50%.
>
> 3. From 1990-1997, it's equity curve moves in a range of -20% to
> +20% with some minor trending.
>
> Unless I have missed something (which is possible), it appears that
> the stoRSI system with the MagicOut123 addition is GREAT for
trading
> in a BURST BURST period (ie, the current bear market), but it is
> equally POOR for trading during a bubble building phase (1998-
1999).
> For more "normal" periods of 1990-1998 (not bubble building or
> bursting), it is more or less random. Although it does display some
> trending results within that period, it ended with about the same
> equity it started with.
>
> b
|