The timeframe of the chart should not matter, because the market price will be the same on all charts.
As the contracts meet during the roll, the prices equalize during the roll week. They kiss, basically. They are magical friends. *smooch* There is no gap in a roll period that freaks out. (Sorry, I try to analogize with video and music. Sometimes it works, sometimes it does not. Hehe.)
Yes I know that but there is still the gaps on the chart. So the question is, do you take your levels off the actual price which that contract traded at (unadjusted chart) or do you take your levels off an adjusted chart which minimises the roll gaps but shows historical prices which that contract actually didn't trade at. I believe that is correct? If there is no roll gap then why is there one on a chart?
There's no one correct answer for this - It's different for different participants and markets. Typically there's a spread been different contract months - this is your "roll gap". If the spread is positive, the market is in contango, if it's negative, the market is in backwardation. Generally this spread reduces to its minimum level during the few/several days period that most traders roll their positions. CME group have a tool to view this - called Pace of the Roll. Some ag futures traders with long positions prefer to roll from one year to the next each time - the roll gap can be quite significant here of course. In years gone by these were known as Gann Contracts. For commodity futures, with the "cost of carry" (storage) reduces through to delivery. Cost of carry varies across markets (consider risk-free interest rates for financial futures) and is also based upon other supply/demand/storage constraints (consider the Crude Oil fisaco where the May 2020 futures contract went negative in late April 2020). There's also seasonality for agriculture futures too. For a trader looking using the front month of a futures market as a proxy for spot pricing, for analyzing support/resistance levels they might want to use an unadjusted (spliced) chart. For a technical trader with a medium-term position using typical indicators (e.g. moving averages), they would probably want this "roll gap" removed (i.e. backadjusted) since their position is unaffected by the roll gap. For a systematic/quant trader trading off a continuous data series, having a medium-term position, likewise they'd want it removed (i.e. backadjusted) as that "roll gap" is detrimental to their analysis.
Thanks a lot Richard, clear and concise response as always. This is similar to what I was thinking. For traders who are using indicators and mathematical equations a gap is going to throw this out of whack. I think an unadjusted chart is going to be best for what I am after. Most of the time levels from the front month will be all that is needed, unless we are breaking out to new highs or lows. Thanks