- ago
Hi,
I just started with intraday backtesting. First of all, I am super satisfied with the speed that this is executued in comparison to WL6. That's just awesome...

Now it came to my attention, that wl8 is doing some automatic split correction for intraday data. This is, in general, a good thing, but for some special cases, I would prefer to do the test without that aka using the pure original data and take care of the handling of these events in my code. Therefore the question, is it possible to turn this correction off?

Best Regards,
Patrick
0
428
3 Replies

Reply

Bookmark

Sort
Cone8
 ( 6.24% )
- ago
#1
First, we're talking about IQFeed data, right?

The title is "Turn off intraday open correction" but you mentioned the auto split corrections.
Are we really talking about splits or correcting the opening price for the IQFeed Provider's RSO option?

Also, in case it provides the solution you're looking for, you can get the price at the level it was actually traded in the market that day using the SplitRev indicator.
0
- ago
#2
Yes, iqfeed data I'm working with now.

Sorry, maybe I am confused something. I just saw that this message:
"IQFeed" "No Daily bars for intraday open correction"
and thought that was the reason.

So, we are talking about splits regarding intraday data. At least as I know, iqfeed delivers unadjusted intraday data, so with splits or reverse splits you would have large gaps. Through these adjustments, the calculation of indicator values is not guaranteed to be the same in comparison to using just the original data. Therefore my question, is it possible using the unadjusted data for intraday testing?

0
Cone8
 ( 6.24% )
- ago
#3
QUOTE:
Through these adjustments, the calculation of indicator values is not guaranteed to be the same in comparison to using just the original data.
Why wouldn't it be? Is there a specific indicator in question?


QUOTE:
Therefore my question, is it possible using the unadjusted data for intraday testing?
It's possible if you reverse adjust it.. and then deal with the "cliffs".

Considering we're dealing with these values with double precision, I'm still having difficulty understanding where you're coming from.

Again, if you need to know the value actually traded historically, for example, to backtest a price filter, you only need to apply the SplitRev indicator.
0

Reply

Bookmark

Sort