From theory to practice - page 666

 
Evgeniy Chumakov:
Tossed the updated robot in the mail.

Yeah. Thanks.

 

I'll disappear again for a while... Gotta run tests on 10 year history, check distributional excesses during trades and a few other things. By the New Year the Grail will be there.

Maybe here, the kitten will show up again (the one who A_K) - I'll talk to him, so the readers won't get bored.

Good luck to everyone!

 

Wait. Or am I doing something wrong. Do you have this in your EURUSD tester?


This buy trade has eaten up a lot.

 
Evgeniy Chumakov:

Wait. Or am I doing something wrong. Do you have this in your EURUSD tester?


This buy trade ate up a lot.

There are such trades (2-3 per year) on every pair. They reduce the total profit for the year by a factor of 2. That's why I take a time-out - there is some extreme asymmetry and excesses observed at the moment of entering the trade. We need to analyse this carefully.

 
Evgeniy Chumakov:


Here's a look, a prime example of AUDCAD 2018:

In the upper left corner:

penultimate line - coefficient of asymmetry=0.089 (we have a left-sided bevel at the time of the buy entry, while it should be a right-sided bevel)

The last line - the excess=136.54. In general, it is beyond understanding.

Here it seems - both the price and the sum of increments are out of the bottom boundary. Buy and be happy...

But as a result..:

The wildest drawdown, defeat and disgrace. Immediately lose 150(!!!) points.

There you go.

 
I'm looking at the distribution function with this variance and the sum of the increments, so the signals are coming in between 0.45 and 0.55 almost in the centre.
 
Evgeniy Chumakov:

Around September, I couldn't bear to wait any longer.


 
The deals are different.
 

Today's deal closed:

+40 pips as if from a bush.

And that's the way it should always be. Amen.

 
Roman Kutemov:
The deals are different.


How did you do the calculations? Write the algorithm point by point

Reason: