You need to make sure you are not subject to confirmation bias:

Quoting 

Confirmation bias, also called confirmatory bias or myside bias, is the tendency to search for, interpret, favor, and recall information in a way that confirms one's preexisting beliefs or hypotheses.

In other words, you need to complement your hypothesis with statistic. You need to define what resistance/support you consider in your observation. are these mathematical levels, or objective levels such as prior day high,/low, opening range etc and then try to correlate these levels to the ES and post your stats. Failing to do this will lead to misunderstanding.

I see your point except I didn't reach this conclusion based on the conclusion that the SandP was dominant and then backtested to make it fit.
If anything it was the opposite as I trade the Dax and the Dow so didn't really want it to be the case.
I could do a statistical compilation but then I'd have to devote time to compiling data rather than trading it.

Anyway, perhaps it would be a good idea for others to do their own research, especially for the US markets, and see what they think.
I have added a section in my original post as I realised that I was focussing on the European/Dax session when that was not really the main thrust of my argument.

Happy trading!

I have no experience doing what you are doing, so forgive the intrusion. However, it seems to me that the correlation might better apply to momentum and trend, rather than actual points of declension where price trend will bottom or top "to the tick," as it were. Also...a different issue: wouldn't it be better to look at supply and demand than S&R? Yes, price often changes at support or resistance, but typically it does so or a reason. That reason may be external or fundamental factors, but isn't it usually where there is already pent up supply or demand. It won't always change at S&R; won't it almost always stop or change at supply or demand?

The correlation seems dubious at best . The correlation coefficient goes under 0 a lot and clearly a visual tells me it is a bit hit and miss but hey whatever works for you . Myself i trade every chart on its OWN merits . I like to measure things

Ozquant 

The correlation seems dubious at best . The correlation coefficient goes under 0 a lot and clearly a visual tells me it is a bit hit and miss but hey whatever works for you . Myself i trade every chart on its OWN merits . I like to measure things

Hi there,

I'm not sure that the correlation measurements provide an accurate picture.
I'm not actually saying they are perfectly correlated for intraday trading, that's kind of my point.
If the Dow is at a reversal point before the SandP has reached it's own reversal point (i.e the open/close or name your area) where it is likely to turn, then the Dow will just continue until the SandP does it's thing (if it does it at all).
It's an issue of timing-basically don't expect the Dow to reverse until the SandP is ready and willing!

@Ozquant,
if you watch daily-charts, you are absolute right, the daily %-change in the FDAX is "normally" much larger as the ES, but if you watch intraday-swings, it really works fine, but you always have to watch if this 2 instruments work on each day a similar way, because there are sometimes days where the FDAX moves completely different in relation to the ES, has to do with German/European related news.

As a short-term trader, I use the ES reactions to S/R levels to take quick reversal trades on YM during the US session. I don't know much about the actual statistics, but I know that I trade at an 80% success level, and most of my trades are based on ES hesitation -- which often means a quick reversal on YM -- at known levels such as Y-high, D-low, vwap, etc.

We've been in New Zealand the past couple of months, so I have also become familiar with FDAX and the London session. That is a seriously wild instrument, but for scalping, trading levels on FESX works very well. FESX is slow, but when it hits major levels as described above, FDAX will react, often violently. I don't have an opinion on which way the correlation between FDAX and the ES works, but I don't much care, as FESX correlation works just fine.

Just my two cents. I think arguing over which instrument leads is not really worth the bother, though the statistics might be interesting. I found what works for me, and that's what is important. I think that's what everyone here might be trying to say, despite the disagreement on correlations.

Michael