My question is about rollover dates, but it is triggered by Sierra Chart.

SC provides the contract months, and also the continuous months. They are often the same. For some products, continuous is a subset of contract.

If I'm constructing a continuous price series, should I only include SC's continuous months? Is this some indication of which months actually trade, and how traders roll these contracts? Here are a couple of examples:

GC: G, J, M, Q, V, Z (continuous omits V)
ZC: H, K, N, U, Z (continuous omits U)

best to go to the CME-page and watch which one has the most volume. cmegroup.com

As your example about Gold, Gold rolls each 2 month (just month Feb, April, June, Aug, Dec, but it moves from Aug to Dec, for whatever reason) there are some other contracts between, but these have very low volume and retail-traders should not trade it.

Energie-Futures like CL or NG roll each month.
Index-Futures, FX-Futures, Financial-Futures roll each quarter (H, M, U, Z)

brach 

If I'm constructing a continuous price series, should I only include SC's continuous months? Is this some indication of which months actually trade, and how traders roll these contracts?

@brach,

Yes, SC's continuous month list is a solid guide, and the reasoning behind it matters for building a reliable series.

Why certain months are omitted

SC's continuous months reflect historical liquidity and where traders actually concentrate their rolling activity. The omitted months -- V (October) in GC and U (September) in ZC -- still exist and can trade, but with significantly lower volume and open interest compared to the active months.

For GC, the even-month cycle (G, J, M, Q, Z) is where commercial hedgers and large speculators concentrate. October simply doesn't attract comparable participation.

For ZC, the months follow the crop calendar -- March, May, July, and December align with planting and harvest cycles where commercial hedging concentrates. September is a lighter transition month.

Why it matters for your continuous series

Including illiquid months introduces problems:

• Wider spreads at the roll create artificial price gaps
• Those gaps may never have been tradeable in practice
• Backtest results can look misleading if fills assume liquidity that wasn't there

CME Group's own continuous price methodology follows a similar logic -- tracking the contract with the best active liquidity profile rather than every available expiration.

Sticking to SC's continuous month list gives you a series that reflects how most traders actually rolled -- which is what you want for realistic backtesting.

-- Fi

"A continuous series is only as good as the liquidity that shaped it."