EURUSD H1 BBWidthRa...
 

EURUSD H1 BBWidthRatio  

Page 2 / 2
  RSS

coensio
(@coensio)
Member Admin
Joined: 6 months ago
Posts: 122
12/01/2019 9:48 am  

Two additional notes:

1. Not many brokers provide reliable M1 candle data.

2. The spread is not included in OHLC. 

So not every strategy can be generated, so when you use M1 OHLC concentrate on higher timeframes >M30, with a wider TP/SL range so spread does not have much impact on the results. 

This post was modified 5 months ago by coensio

ReplyQuote
BladeRunner
(@bladerunner)
Active Member
Joined: 6 months ago
Posts: 17
14/01/2019 6:21 pm  

Another view/question about the 99% tests is also if one of the benefits/functions of Robustness tests isn't also to account for broker/data variability. When I test for example for spread or slippage variations, in my mind I'm thinking also about robustness against differences between brokers.


ReplyQuote
coensio
(@coensio)
Member Admin
Joined: 6 months ago
Posts: 122
24/01/2019 8:39 am  
Posted by: BladeRunner

Another view/question about the 99% tests is also if one of the benefits/functions of Robustness tests isn't also to account for broker/data variability. When I test for example for spread or slippage variations, in my mind I'm thinking also about robustness against differences between brokers.

Yes, this is also true. If data is well time synchronized then the slippage and spread validation (Monte-Carlo) tests are covering for the differences between brokers (to some extent).  There is no golden rule, the goal is to minimize unknown factors.


ReplyQuote
coensio
(@coensio)
Member Admin
Joined: 6 months ago
Posts: 122
04/02/2019 5:12 pm  

I'm marking this SQ 3.8.2 strategy as 'outdated' since it is developed according to a very old workflow and also old SQ version. We will now focus on SQ-X strategies only.


ReplyQuote
-hg
 -hg
(@hg)
New Member
Joined: 6 months ago
Posts: 1
09/02/2019 1:09 pm  
Posted by: coensio

Yes I export OHLC data directly from MT4 terminal from my live chart from live account (using the same datacenter that is used for trading), then I load this OHLC data file to SQ in datamanager (you need to setup properly all symbol settings), and I concentrate only on strategies that are using OHLC information. So in theory it would no matter if you retest it with higher precision since the OHLC already contains most of the required information.

Do you use the OHLC to gain speed during optimization or does it has another reason?


ReplyQuote
coensio
(@coensio)
Member Admin
Joined: 6 months ago
Posts: 122
09/02/2019 2:10 pm  

Do you use the OHLC to gain speed during optimization or does it has another reason?

The main reason is lack of historical tick data from brokers I use. The best possible approach would be to switch to Dukascopy broker and use their accurate tick data or find a broker which is very similar to Dukascopy (some advise Purple-Trading broker because of good correlation?).

Most successful traders I know, use OHLC data (with fixed spread) for strategy generation and most of the validation tests (to save processing time). Some critical tests are done with high resolution M1 OHLC data. Then, Dukascopy tick data is used only as final verification of the resulting strategies, to see how those systems will react to real live (variable) spread.  

 

 

This post was modified 5 months ago by coensio

ReplyQuote
Page 2 / 2
Share:
  
Working

Please Login or Register