StrategyQuant crash...
 

[Sticky] StrategyQuant crash course 2019  

Page 2 / 2
  RSS

coensio
(@coensio)
Member Admin
Joined: 7 months ago
Posts: 135
21/06/2019 8:34 am  

yeah the whole europe is one hot place right now...but 42 is really really extremely hot 😉


ReplyQuote
jpygbp
(@jpygbp)
New Member
Joined: 2 months ago
Posts: 2
23/06/2019 11:22 pm  

I noticed in the GBPUSD slippage retest, that the slippage is set to zero.  Is this correct?  What is the rationale in this step?


ReplyQuote
coensio
(@coensio)
Member Admin
Joined: 7 months ago
Posts: 135
24/06/2019 8:17 am  
Posted by: jpygbp

I noticed in the GBPUSD slippage retest, that the slippage is set to zero.  Is this correct?  What is the rationale in this step?

The rationale here is to be able to test if our system has not been curve fitted to the 'random noise' on EURUSD pair. The assumption here, is that 'noise' by definition cannot be reproduced and so cannot be exactly the same on the another (but correlated) pair like GBPUSD.

The 'normal' slippage will only add additional 'noise' component, however it should not have large impact on the final result. I decided to go without slippage....

In the end you need to find out what is working for you by yourself 😉

 

 

 

 


ReplyQuote
volpiemanuele
(@volpiemanuele)
New Member
Joined: 3 weeks ago
Posts: 2
05/07/2019 8:43 am  

Hi,

I try your custom project and I wante to know:

- how to change the dates shown in each process as we are in July today;

- how to adapt the project to other markets (indices) where data starts from 2013 instead of from 2003 as in forex;

- if the various steps in the project can be sufficient to test the quality of the strategies and if more specific ones can be added- .

- if there are other configurations available also to conduct specific tests;

thanks

 

 


ReplyQuote
coensio
(@coensio)
Member Admin
Joined: 7 months ago
Posts: 135
05/07/2019 12:51 pm  
Posted by: volpiemanuele

Hi,

I try your custom project and I wante to know:

- how to change the dates shown in each process as we are in July today;

- how to adapt the project to other markets (indices) where data starts from 2013 instead of from 2003 as in forex;

- if the various steps in the project can be sufficient to test the quality of the strategies and if more specific ones can be added- .

- if there are other configurations available also to conduct specific tests;

thanks

 

 

1. The workflow is based on a fixed time window length for generation and testing if you want to use 'newer' data then you need to shift all dates by the same amount of time (e.g.: +1year). An exception is the start date of final test with real spread then you can use all of available data.

2. Just get sufficient amount of data, you can export it from Tradestation or buy online or subscribe to SQ data program.

3. There is an infinite number of possible workflows/configuration, this is just most basic workflow you can get...consider it as a simple example. But yes there are more advanced techniques to test for robustness like WFM and additional testing on portfolio level, but this is not covered in this particular example..

Feel free to change it, adjust parameters add or remove test until you are happy with the final results...

gr

Chris

 


ReplyQuote
volpiemanuele
(@volpiemanuele)
New Member
Joined: 3 weeks ago
Posts: 2
06/07/2019 8:28 am  

Thanks


ReplyQuote
Page 2 / 2
Share:
  
Working

Please Login or Register