[StephaneFima]

Thank you Craig, this is very clear now!

[StephaneFima]

In fact, I do not understand why we have to use NorgateIndexConstituentTimeSeries to check whether a stock was in an index or not if a “current and past” watchlist already provides this.

[StephaneFima]

Thank you Said. I think I see the point now.
In other words: when you nest a static variable within a loop (for example), you are explicitly asking Amibroker to store the value of the variable whatever happens next, for each stock?

[StephaneFima]

Craig,

There is something I do not understand quite well: why not using adatabase including historical constituent AND using IsIndexConstituent ?

Thanks

[StephaneFima]

September Russell 1000 MOC results = +0.83%

[StephaneFima]

Hi Julian,

One little question for you: in the picture you are showing, I can see a box called “Auto order entry configuration”, what is it for? (I have version 1.2.1) Thanks!

Stephane

[StephaneFima]

no pb on my side

[StephaneFima]

Nick, you are probably right, there may be some emotional stuff here (but I refrained myself from switching off the model).
I am pretty confident about my system (I did several times MC analysis, backtested using different parameters, … and everything was OK).

The only thing which was indeed not in my comfort zone was the fact that I didn’t control my entries and I couldn’t do anything about it.

Now with a little bit less than -3.0%, I am perfectly in my DD acceptable range (and I agree with Scott, that’s not so bad). At the end of the day, I am quite happy that it happened at the beginning of trading this system, I learned something (and I think I will keep learning!!).

[StephaneFima]

The API did its job.

Maybe a way to avoid the situation where the API would have to sell back excess positions would be to place stop limit orders with enough room between the limit and the stop.

Yesterday was tough but could have been worst. -2.71% on my MRV system.

[StephaneFima]

Yeah, no big deal but could have been.
I’ll keep you updated on Levente’s answer

PS: Indeed, a big market gap down today and continuing to go down …

[StephaneFima]

I’ve been using Batch Trader including the MOC option for the last week and until today everything went right.

This morning I placed around 80 orders with a maximum of 10 open positions.

The US opening was very bad and the API reacted in a very annoying way.

It bought every position that touched its stretched price (i.e. around 20) and then, as it was above the maximum position number, it sold the excess ones. Doing so, I had (small) losses between the buy and sell prices, but on top on this I had to pay commissions for nothing.

I asked Levente how to fix this and I am waiting for his answer.

[StephaneFima]

Big day today as I really started trading live my mean reversion system with MOC exits.

– Backtest over the last 10 years over the Russell 1000 and without Index filter gives a CAR of around 40% and Max DD of c. 10%
– Backtext with index filter gives a CAR of c. 25-30% and MaxDD of around 5% to 8%
– Monte Carlo with 1000 iteration is in line with the above results.

As everybody says here: next 1000 trades (or is it 10 000?? )

[StephaneFima]

WOW!!! Excellent!

[StephaneFima]

Hi everyone!

I’ve not been involved lately in this forum, but it didn’t mean that I stopped thinking, testing and trading :whistle: . In fact, this is quite the opposite

1) I started paper trading a mean reversion strategy. So far, so good: 22 orders filled, 15 positive (average gain of +2.2%), and 7 negative (average loss of -1.1%)
2) I am also testing a trend following system (results are not good enough so far). I will try 2 or 3 other ides on it and if I
3) When done with point 2, I will try a rotational system

[StephaneFima]

I am under Windows 10 since its launch some time ago.
Much better than Windows 8.
No major problems on my side, except some issues on number formatiing.

[Author]

Posts