The rally this week

[bwolinsky]

Quote from Wide Tailz:

"Look inside bar backtesting"

It doesn't work in multicharts, and I'm sure doesn't work in tradestation.

Whether it's ticks, minutes, the same problem arises.

If your simulations don't change significantly when you look inside that should clue you in that it doesn't work, because I see the hourly bars on your chart will experience the same problem, even if you are looking inside the bar, it doesn't work or function properly.

How have your results changed when using lookinsidebarbacktesting than when you're not?

Are you sure it's reading the chronological order, because seeing the hourly chart I know it's not wrt what you posted.

 

[Wide Tailz]

Quote from bwolinsky:

It doesn't work in multicharts, and I'm sure doesn't work in tradestation.

Whether it's ticks, minutes, the same problem arises.

I've sat next to the laptop and watched it execute to the cent when the trailing stop fires. It matches the back test.

The only unknown at this point is slippage, and it appears to be a few cents as well.

You could be right and we learn the most by trying to disprove our results.

I appreciate your input.

 

[Wide Tailz]

Quote from bwolinsky:



If your simulations don't change significantly when you look inside

The simulation returns about 4x more with "look inside bar" turned off. You are seeing my detuned system with very conservative settingz........

 

[bwolinsky]

Quote from Wide Tailz:

I've sat next to the laptop and watched it execute to the cent when the trailing stop fires. It matches the back test.

The only unknown at this point is slippage, and it appears to be a few cents as well.

You could be right and we learn the most by trying to disprove our results.

I appreciate your input.

No, really, the easiest way to tell is if your results change significantly between that setting and regular backtesting.

If that works, let me know, but I've found that it's not functional the way it ought to be.

 

[bwolinsky]

Quote from Wide Tailz:

The simulation returns about 4x more with "look inside bar" turned off. You are seeing my detuned system with very conservative settingz........

Hmm, but I take it you haven't optimized with that setting on?

I know it's a pain, but there are ways to do it, and I suggest genetic algorithms as a solution.

 

[Wide Tailz]

Quote from bwolinsky:

No, really, the easiest way to tell is if your results change significantly between that setting and regular backtesting.

If that works, let me know, but I've found that it's not functional the way it ought to be.

I don't "look inside bar" to the tick level, just to the nearest minute so there is some error in the results.

Let's take this to PM.

 

[bwolinsky]

Quote from Wide Tailz:

I don't "look inside bar" to the tick level, just to the nearest minute so there is some error in the results.

Let's take this to PM.

Sure.

 

[tradingjournals]

Hey Bwol: if you are searching for exposure, I suggest you go somewhere else other than this thread.

Moderators: clean his shit as you see fit.

 

[Tsing Tao]

Quote from tradingjournals:

Hey Bwol: if you are searching for exposure, I suggest you go somewhere else other than this thread.

His text is infinitely more intelligent than anything you write.

Exhibit A:

Quote from tradingjournals:

The post reminds me of poets teaching advanced maths.

 

[tradingjournals]

Tsing Tao just posted, and E/U is at 1.3030 area and was going down. So is it time to go long?