I took a screenshot of tradingview with the two spreads NQ/ES (blue) and RTY/YM (orange) and the super spread (NQ/ES)/(RTY/YM) in black.



The super spread seems to track the difference between the two spreads just fine... It always touches 0% as soon as the normal spreads get back together. Please disregard the first cyan circles, as I mixed up the lines a bit...

But I understand, it's probably only good to track the differences from a theoretical point of view, once you want to build a position, you gotta pay close attention to the notional values and how they might impact these percentage calculations...

alphagamma 

Dude, Josh!!! That's a slam dunk! Your explanation makes it easy to understand. Don't know why I didn't think of that appraoch myself... Dividing really yields the results that I want! Now I understand the logic behind the spread constructs with 2 tickers. What if I add 4 tickers to the equation? For example, If I want to track the difference between NQ-ES and RTY-YM, what should I enter then? (NQ1!/ES1!)/(RTY1!/YM1!)? Is that correct? So buying that spread (NQ1!/ES1!)/(RTY1!/YM1!) results in being long NQ-ES and being short RTY-YM? What about 3 tickers: 2*ES-YM-NQ?

you have to be sure the Value of each tick of Spread that you are buying/ selling.
after being market neutral, you can crop the Volatility

You can't/shouldn't use percentage or ratio's with continuous contracts, where the continuous contract is constructed by splicing together absolute price series. Every time there is a rollover all your historical percentages and ratio's will change, making any form of historical back testing obsolete.

to make such calculation on spreads, you will need a CQG (CQG Qtrader, is on a transactional mode

alphagamma 

If I want to track the difference between the NQ and the ES futures on a percentage basis, I can type in NQ and add ES on top of it to see when the two futures are far apart. Is there a synthetic futures formula that would allow me to track the spread between NQ and ES in a systematic fashion? I tried to use NQ-ES but this leads to some confusing, at least to me, results: As you can see, the 2 (ES orange, NQ magenta) futures and the NQ-ES spread (blue line) start out at 0% but, even though the ES and NQ drift apart and meet once again, as indicated by the green circles, the blue line (the spread), doesn't fall back to the 0% line as it should, because the gap between NQ and ES is closed at the green circled spots. Why isn't the NQ-ES spread at 0% once NQ and ES meet and above zero (in case NQ gets stronger than ES) and below zero (in case NQ gets weaker than ES)?

alphagamma 

Why isn't the NQ-ES spread at 0% once NQ and ES meet and above zero (in case NQ gets stronger than ES) and below zero (in case NQ gets weaker than ES)?

@alphagamma,

That's worth asking -- this trips up a lot of people and the fix is straightforward once you see why subtraction breaks down here.

The core issue: NQ and ES have very different point values. NQ trades around 4x the price level of ES, so subtracting NQ - ES gives you an absolute point difference that keeps growing over time even when their percentage moves are identical. That's why your blue spread line drifts away from zero -- it's tracking a point gap, not a relative performance gap.

The fix: use division instead of subtraction.

In TradingView, type NQ1!/ES1! as your symbol. This gives you the ratio between the two contracts. When NQ and ES are moving in lockstep on a percentage basis, the ratio stays flat. When NQ outperforms, the ratio rises. When ES outperforms, it falls.

Set your chart scale to "Percent" (right-click the price scale, select Percent) and you'll see exactly the relative divergence you're looking for. When the two converge at your green circle spots, the ratio line will actually return to baseline -- which is the behavior you expected from subtraction but couldn't get.

One thing to watch: continuous contracts (the 1! symbols) use back-adjusted prices, which can create small distortions around quarterly rollovers. If you need clean data for a specific period, you can use the actual contract months instead -- for example, NQM26/ESM26 for the June 2026 contracts. For general monitoring this usually isn't a big deal, but it's worth knowing about.

As josh pointed out in this thread, the ratio approach works for any pair of related futures -- not just NQ and ES. Same logic applies if you ever want to compare SI against GC, for instance.

-- Fi

"The right formula makes the invisible obvious."