Indicator for Comparing Which of Two Correlated Instruments is Leading/Lagging?
- Thread starter BOC
- Start date
Which platform?
Leading and lagging, versus co-related, are two very different things.
https://tylervigen.com/spurious-correlations
Posit a theoretical connection, first, and then seek to disprove it. It won't *guarantee* that you're posited connection is sound, but it'll sure cut down on noise.
FWIW, there is no *direct* "T/A" tool for testing such a hypothesis, but one this you might do is create a single inter-market ("spread") product, and see how it trades over time. Do that logarithmically, and you'll have percent changes that don't depend so much on units.
https://tylervigen.com/spurious-correlations
Posit a theoretical connection, first, and then seek to disprove it. It won't *guarantee* that you're posited connection is sound, but it'll sure cut down on noise.
FWIW, there is no *direct* "T/A" tool for testing such a hypothesis, but one this you might do is create a single inter-market ("spread") product, and see how it trades over time. Do that logarithmically, and you'll have percent changes that don't depend so much on units.
I do this last one on Koyfin - it’s pretty neat.
If you have ThinkorSwim you can use this custom study that will normalize and compare 10 different instruments. It's called the "Multiple Instruments Comparison" study. It's free.
https://github.com/jshingler/TOS-and-Thinkscript-Snippet-Collection/blob/master/TOS & Thinkscript Collection.pdf
It's good for comparing highly correlated instruments. You can check your read on the market against a basket. Here's a screen of me using a modified version of it.
https://github.com/jshingler/TOS-and-Thinkscript-Snippet-Collection/blob/master/TOS & Thinkscript Collection.pdf
It's good for comparing highly correlated instruments. You can check your read on the market against a basket. Here's a screen of me using a modified version of it.
Posts like these make me feel fortunate to have found the ET community.If you have ThinkorSwim you can use this custom study that will normalize and compare 10 different instruments. It's called the "Multiple Instruments Comparison" study. It's free.
https://github.com/jshingler/TOS-and-Thinkscript-Snippet-Collection/blob/master/TOS & Thinkscript Collection.pdf
It's good for comparing highly correlated instruments. You can check your read on the market against a basket. Here's a screen of me using a modified version of it.