Market phenomena - page 67

 
tol64:
There seems to be another parameter. The distance in pips to which the order is set. Although it might be the closest possible price.

This is the point. This distance equals an op-open, because if you read the first post, with an op-open less than zero, the next op-open will be greater with a ver. 75%. And vice versa. That is, if the op-open is, say, - 0.00150 pips and inside the current bar the price has once again tested the op-open-0.00150, then the probability that the next bar will open at a price higher than 50%, and even reaches more than 60%.

So there is no adjustable parameter, it is determined by open[0]-open[1].

 
alexeymosc:

This is the point. This distance equals an op-open, because if you read the first post, with an op-open less than zero, the next op-open will be greater with a ver. 75%. And vice versa. That is, if the op-open is, say, - 0.00150 pips and inside the current bar the price has once again tested the op-open-0.00150, then the probability that the next bar will open at a price higher than 50%, and even reaches more than 60%.

So there is no parameter configurable, it is determined by open[0]-open[1].

I've wanted to show the point for a long time. Illustration of the situation for open[0]-open[1]<0.

 
alexeymosc:

I've wanted to show the point for a long time. Illustration of the situation for open[0]-open[1]<0.


Did the result you presented above take into account the spread? At 5 minutes the spread in such a scheme could derail everything.
 
tol64:

Did the result you presented above take into account the spread? On a 5 minute spread with this scheme, the spread could negate everything.
I've been putting in 10 five-digit spreads (on a four-digit spread it's 1). But otherwise, yes, you have to look for a broker with a more or less stable spread of one four-digit point.
 
alexeymosc:
Now I'm downloading the 5 minute history from Ducas for 5 years. This is going to be the bomb, I'm going to test the survivability of the strategy on this huge array.

This is a 5 minute history test from January 21, 2007. The vertical dash marks the beginning of the chart I gave earlier (from March 2011). Strange. For the last 2 years or so it has been mainly growth, before that it was mainly decline. Has the market gone weird again? )))

The test has the same range parameter. Optimization ... did not help.

Very strange.

 
alexeymosc:

This is a 5 minute history test from January 21, 2007. The vertical dash marks the beginning of the chart I gave earlier (from March 2011). Strange. For the last 2 years or so it has been mainly growth, before that it was mainly decline. Has the market gone weird again? )))

The test has the same range parameter. Optimization ... did not help.

Very strange.


Yeah. That's it. And if you also make the spread, for example, 5 points, add slippage, then the picture gets even worse. :)
 
alexeymosc:

Very strange.

There's nothing strange about it.

The strategy works wonderfully on the exact part of the story for which the statistics have been gathered...

 

Plus another year... Yes, I agree that the phenomenon is local. Sorry. That's how volatile the market is.

 
tol64:

Yes. That's it. And if you also make the spread, for example, 5 points, add slippage, then the picture gets even worse. :)
Where do you get the 5 pips spread on the eu? In the kitchens? )) And where does the slippage come from? maybe the requotes, that's right...