Importance of WalkForward Analysis???

phanz · 2021-09-01T06:00:56+00:00

Hi all, While I have read the danger of over optimization, my understanding is that if your system is not optimized then it is likely to be more robust, and perhaps also more objective as you use the same criteria to study different market. So let say I have decided not to do optimization and try to test the system as original. Then again I read the importance of WalkForward analysis. My understanding is WalkForward analysis is something like optimization also. So if I have decided not to do any optimization, does that mean WalkForward analysis is not important in this case? What are your thoughts and advice? Thank you for your time and have a splendid day ahead.

Viewing 3 posts - 1 through 3 (of 3 total)

Author

Posts
09/01/2021 at 7:00 AM #176557 quote

phanz
Participant

Average

Hi all,

While I have read the danger of over optimization, my understanding is that if your system is not optimized then it is likely to be more robust, and perhaps also more objective as you use the same criteria to study different market. So let say I have decided not to do optimization and try to test the system as original.

Then again I read the importance of WalkForward analysis. My understanding is WalkForward analysis is something like optimization also. So if I have decided not to do any optimization, does that mean WalkForward analysis is not important in this case?

What are your thoughts and advice?

Thank you for your time and have a splendid day ahead.

09/01/2021 at 7:18 AM #176558 quote

GraHal
Participant

Master

In the ideal world (doesn’t exist) and you have the holy grail system (doesn’t exist) that has not been optimised but is acceptably profitable on many timeframes and many markets.

Walk Forward on this holy grail system, would / should show that variable values would give you acceptably profitable results for each of the 5 (or more) WF periods.

Try it and let us know, better still, show us what you get?

phanz and thanked this post

09/01/2021 at 8:29 AM #176565 quote

Nicolas
Keymaster

Legend

The Walk Forward analysis IS an optimization, but divided into 2 periods: In-Sample (the optimization period = IS) and Out-Of-Sample (the test period = OOS).

Example:

I decided to optimize my strategy from 1st January to 30th April (4 months of IS period) and test if it was profitable if it was traded in live between 1st May to 30th June (2 months of OOS period).

It simply let you see if the optimization was relevant for the test period: is the strategy over optimized or not? Of course, there is more to analyze and test with the Walk Forward tool available in ProRealTime.

phanz and thanked this post
Author

Posts

Viewing 3 posts - 1 through 3 (of 3 total)

You must be logged in to reply to this topic.

New

Stuck on this ProBuilder code?

Describe what this topic is trying to build, in plain English, and ProRealAI writes the ProRealTime™ indicator, screener or system for you.

Available in 7 languages

Try ProRealAI

Importance of WalkForward Analysis???

ProOrder: Automated Strategies & Backtesting

New Reply

Author

phanz @phanz Participant

Summary

This topic contains 2 replies,
has 3 voices, and was last updated by Nicolas
5 years ago.

Topic Details

Forum:	ProOrder: Automated Strategies & Backtesting
Language:	English
Started:	09/01/2021
Status:	Active
Attachments:	No files