Okay, so both IB and SC can do this - but they both have various annoyances. IB in that the way the charts are actually rendered I never know how much I can even trust their multi-leg chart calculations vs a simple calendar spread. SC in the pure tediousness needed to do anything outside the box. IMO, the dev team thinks it's perfectly reasonable to require any chart using a difference, ratio, or overlay study to require an entirely new chart just to provide input to said study (I think that's overly tedious and only valuable when doing really complex stuff).
Using -CLM5+2CLN5-1CLQ5 as an example:
To do it with IB you just use standard combos and for a butterfly I've always used the "pair or leg-by-leg" tab. I also made a "reverse calendar" chart using the leg-by-leg tab just to see it as a component chart here but technically neither of the front and back month spreads are needed just to chart a butterfly:
To do it with SC you use these steps:
- Create a new chartbook just to keep it all in one place.
- Create a new front-month chart (or dupe it from another chartbook into this new chartbook) - and then (unless it's an inter-commodity spread) disable continuous contracts on it (#1 aka CLM5).
- Dupe that first chart and change the contract month to the middle month (#2 aka CLN5)
- Dupe that first chart again and change the contract month to the far month (#3 aka CLQ5).
This is where it gets more complicated:
- Take the middle month chart (#2 aka CLN5) and dupe it 2 more times until it's the 4th and 5th charts in the chartbook (which are now also CLN5 and CLN5).
- On chart #4 (CLN5), use the difference (bar) study and use chart #3 (CLQ5) as the input to the study. Important: set the study to "display as main price graph". This results in one chart that is just CLN5-CLQ5:
- On chart #5 (CLN5), use the difference (bar) study and use chart #1 (CLM5) as the input to the study. Important: set the study to "display as main price graph". This results in one chart that is just CLN5-CLM5.
- On chart #5 (which is now CLN5-CLM5), add a overlay (bar) study and use chart #4 (CLN5-CLQ5) as the input and set region to 2. You now have the front and back spreads as region 1 and region 2. Note: Technically you don't need this overlay region as it's just a nice to have.
- On chart #5, use another difference (bar) study and use chart #4 (CLN5-CLQ5) as the input to the study and make sure that the study setting "chart 2 multiplier" is set to -1. Set region to 3. This results in a region that is made up of this math (remember the study is using the main price graph [CLN5-CLM5] as the "chart 1" side):
(1 * (CLN5-CLM5)) - (-1 * (CLN5-CLQ5)) aka
(CLN5-CLM5) - (CLQ5-CLN5) aka
-CLM5+2CLN5-CLQ5
You can now add bollinger bands or whatever else you want to each region in the chart.
BTW: if you screw up and set chart 2 multiplier to 1 (the default), you'll get this:
(1 * (CLN5-CLM5)) - (1 * (CLN5-CLQ5)) aka
(CLN5-CLM5) - (CLN5-CLQ5) aka
CLQ5-CLM5
Which is Jun/Aug and not what you want. If you screw up and set chart 1 multiplier to -1, you'll get this:
(-1 * (CLN5-CLM5)) - (1 * (CLN5-CLQ5)) aka
(CLM5-CLN5) - (CLN5-CLQ5) aka
+CLM5-2CLN5+CLQ5
Which is the inverse of what you want (or maybe it's what you want).
The spreads can be made further accurate by following these directions such that high/low differences are computed correctly intra-bar:
http://www.sierrachart.com/index.php?page=doc/helpdetails61.html
I haven't setup a butterfly spread chart with those directions just due to the pure complexity but I might try just to see if it can be done.
Now you see why most people use X_Trader or CQG. The really annoying/lame part is that this is *not hard* stuff. There's literally no reason both SC and IB cannot graph spread charts at high accuracy so people can get their work done without so much hair pulling.