No registration required! (Why?)

Correlation of Market Movements


In this topic pippa was asking about how markets relate to one another.

I thought that I would kick off a new topic that shows a scientific way to measure the relationship between two markets. I created a spreadsheet that you can download and use as a template for doing this sort of thing. If you can operate a spreadsheet then you can do this yourself.

Here are the results from the spreadsheet.

Market correlation of closing prices from 1 June 2007 to 13 July 2007.
	ES	NQ	YM	ZB
ES 1.00 0.68 0.96 0.01
NQ 0.68 1.00 0.77 -0.11
YM 0.96 0.77 1.00 0.05
ZB 0.01 -0.11 0.05 1.00

The correlation coefficient shown above measures how closely one market moves in relation to another. The correlation is a value from -1 to 1. -1 Indicates a perfect negative correlation, 1 indicates a perfect positive correlation and 0 indicates no correlation.

The ZB and NQ, for example, have a weak negative correlation. This means that when the NQ moves up the ZB will move down. The YM and ES have a strong (96%) correlation. The YM and ES will almost always move in step with each other.

The attached spreadsheets show the daily closing prices of the markets but there's nothing stopping you from doing this on intraday prices. Make sure that you line-up the times of your data series correctly if you're comparing European or Eastern markets with the US markets.

Click link to access uploaded file:
correlation.xlsx

Click link to access uploaded file:
correlation.xls

Currently, I just eyeball it and take in the general information. I'm not sure how I could build a TradeStation indicator to plot it real-time. But, I shall be thinking about it!
The correlation formula is not very difficult and I know that on eSignal you can download one for free from the eSignal studies web site. I'm sure that there's already one available on the TradeStation web site that you can download.
I'll certainly look into further. Thanks for the suggestion.
Comments received by email and published with author's permission:
quote:
Your statistical "correlation coefficient" is correct especially because the basic correlation lies in the fact that each and every on of the YM components also reside in the ES so it stands to reason that when YM starts to move (Up or Down), the ES will shadow that move with reciprocity between the ES moves and the YM then tagging along;

On the other hand there is no commonality between NQ or ZB so they will reflect a move, one against the other, except in the most rare of occasions;

Moreover, and in ZB representing the 30 year Treasury Bonds while NQ represents the Nasdaq 100, it is easy to understand that advancing Markets negatively impact the Bond Markets, with the reverse at play when Bonds advance to the disadvantage of the Equity Markets . . .

quote:
As for correlating ER2 against the YM and/or ES, a quick look at the way ES, YM and NQ have been setting Higher Highs and Higher Closing Prices, it should be clear that NQ (which has a good number of its components in ES and even a couple in YM (eg: MSFT is an NQ component, is a YM component and is also an ES component)) can have a correlated relation to the "more sophisticated Indices";

Those moves, on the other hand, are more of a "push me - pull you" variety;

ER2, on the other hand, virtually stands alone (see how it has been lagging below its early June High) without participating in the current "Big Boys" Market Run;

That's because ER2 is comprised of the "aspiring, promising newer corporation" which have yet to make their mark and may never do so because if and when they start to make that mark, the larger corporation may use Merger and Acquisition tactic to take the smaller caps out;

Another factor at work on the ER2 is that it was "Rebalanced" at the close of Quadruple Witching Expiration with that nature of that "Rebalancing" still being digested by Analysts, Managers, the Institutions, et al.;

Similarly, ES is still in the process of being adjusted to fill in the 8 vacancies caused by the disappearance of those equities taken out by M&A;

Those "disappearances" were complete by the end of June (ie: end of Q2 / end of the 1st half;

So far, WFR, PCP, and 2 others have been added to $SPX (ES), with 4 others still to be identified;

Those additions will assuredly bring about Volatility in the equities to be added and that, in turn, will bring about Volatility in ES;