The Importance of Simplicity

[Snosur4]

Quote from Girlpower:



Actually, funny you should mention that, but - Yes I have - only problem was that the system defaulted them to start the 3rd bar at 10.00 which didn't help much, and started the first bar at 9.20 which was even less help. It doesn't seem to matter what size of bar, they all deafult to 10 a.m.

I'd love to be able to work on say 17 min bars or something like that with 34 as the biggie instead of 30 but I just can't do it. I drew them by hand once for an experiment and they worked very well indeed...

Maybe I will be able to when I get some new equipment (when I finally stop changin my mind about what to get...)

Natalie

Good thought.

FWIW I use Esignal and am very pleased with it.

Of course, it does much more than I reguire, but the data feed is very reliable and that is what is important to me.

You can configure your time templets to start/end any time you wish and the bars will start and progress from there.

I was thinking of 12 min bars....I will watch one next week and we'll compare notes.

Mike

 

[dbphoenix]

Quote from Girlpower:Maybe I will be able to when I get some new equipment (when I finally stop changin my mind about what to get...)

You may want to look at Sierra charts. You can set the bar interval to anything you like, the program is likely to do just about anything you require, and it's only $40/six months. If you use IB, that can be used as a free datafeed. I don't know of a lower-cost option around.

--Db

 

[dbphoenix]

Quote from Snosur4:

I was thinking of 12 min bars....

There's an advantage in using something that not everyone is looking at, which is generally the case with 5, 15, 30, and 60m bars. There is also an advantage in entering in the middle of a bar rather than at the open or close or a tick one side or the other, simply because no one knows what you're doing.

--Db

 

[Girlpower]

Quote from dbphoenix:



There's an advantage in using something that not everyone is looking at, which is generally the case with 5, 15, 30, and 60m bars. There is also an advantage in entering in the middle of a bar rather than at the open or close or a tick one side or the other, simply because no one knows what you're doing.

--Db

hi Db

Good point.

Seeing as how I don't know what I'm doing 3/4 of the time how could anyone else have any clue... :eek:

That's why I need a simple approach that even I can't get wrong

Natalie

 

[wally_]

Quote from dbphoenix:



There's an advantage in using something that not everyone is looking at, which is generally the case with 5, 15, 30, and 60m bars. There is also an advantage in entering in the middle of a bar rather than at the open or close or a tick one side or the other, simply because no one knows what you're doing.

--Db

I have to quote it again, because these are exactly my thoughts.
In other words, I would like to anticipate the breakout and I believe in some cases it can be done, the last 3 days for instance, and as far as non-standard bars are concerned, I use 13 min ones or 7 min ones (13 min ones are more popular than 12 min ones because 13 is a Fibonacci number). Sierra Chart is very good in this respect, I believe Wealth-Lab would do the trick too.

Now, about anticipation on a diferrent occasion, I am hoping to ilustrate this point with pictures of charts, but I need more time to prepare it. I am now busy working on some promising mechanical systems, so this may take a while.

One of the systems produced 126 pts this past October with only 21 trades, 17 of which were winners. This month the system is up 19.25 pts with 6 trades, only one of which was a loser. One of the system signals was today for a cool 9 pts. So I am kinda excited about it, but need to do more backtesting.

Natalie, thanks for your data file. As I understand, the data is for a big contract and not a mini one. Anyway, I hope, I will get the E-mini daily data somewhere or will read them off the 60 min bars. For consistency, I prefer the E-mini data, but thanks anyway.

wally_

 

[Girlpower]

Quote from wally_:



Natalie, thanks for your data file. As I understand, the data is for a big contract and not a mini one. Anyway, I hope, I will get the E-mini daily data somewhere or will read them off the 60 min bars. For consistency, I prefer the E-mini data, but thanks anyway.

wally_

Wally,

I can send you the 24 hours continuous file for ES if that would help. The one I sent was the closest match to normal trading hours that I have.

This file would be OHLC midnight to midnight and not just 8-30 till 4. let me know if you'd like it. Sorry, I can't extrapolate the 8.30 to 4 data...

Natalie

 

[Girlpower]

I've manage to find a solution to what I was saying before about it all defaulting to 10.00 I can now make 16 min and 17 min bars default to midnight, which changes the aspect of the charts completely. A cursory glance shows an improvement to the overall approach I've been using and removes some of the false breaks which have been where my losses have occured.

I'm going to run some tests on it in the next few days and see if the improvement is real or an illusion

1 immediate observation from scanning back through and comparing to the 15 mins chart is that the intraday turnarounds are much more easily visible. Could be a new addition to the general method I'm using in the making - maybe!!! Don't know yet - but looks promising

Natalie

 

[dbphoenix]

Db's Law of Trading #1

No matter how much money you make backtesting a system and no matter how much money you make paper-trading that system, the first day you trade it for real, you will lose truckloads of money.

--Db

 

[Girlpower]

Quote from dbphoenix:

Db's Law of Trading #1

No matter how much money you make backtesting a system and no matter how much money you make paper-trading that system, the first day you trade it for real, you will lose truckloads of money.

--Db

Oh boy - have I ever got that T-shirt! :eek:

Not so much recently though. It's been more about making little improvements for the past couple of years...


Natalie

 

[Lobster]

Quote from mktman:

All indicators are "lagging" no matter what one you use.

mktsurf

I made an indicator that gives me a spike usually about one minute before the ES start a large move.