Best Futures SignalsContract / Cash / Execution

Futures / Original case research

Continuous Futures: Back-Adjusted Charts Versus Real Contracts

Trace a continuous chart through its component contracts without treating a revised historical price as a broker fill.

An observation, its retained evidence and a bounded conclusion
Illustrative research workflow; the worked record below is fictional.

Which price series answered the question?

A continuous chart helps examine behaviour across expiring contracts. An order needs the actual exchange product and month. Those identifiers can coexist in a research record, but neither replaces the other. Ask whether the chart is unadjusted, difference-adjusted or constructed another way; save the provider's method and version. A screenshot's familiar root symbol does not tell you which prior prices were transformed. This is especially important when somebody later compares a fixed alert level with a rebuilt chart.

Sierra Chart documents separate contract files, rollover settings and back adjustment. That construction reference supports a question about series identity, not a universal claim about every data vendor. Freeze the data export used for a study and preserve the constituent-month map. If later data correct a bad settlement or missing month, keep the correction version rather than rewriting what the original observer could see. A research-reference history is useful without being an execution history.

Reconstruct the boundary, not a synthetic fill

Separate the old contract's last included observation, the new contract's first included observation and the adjustment applied to earlier values. Compare an alert's stored level against the series actually used when it was issued. Do not silently translate a synthetic chart level into the new contract by guessing the basis. A defensible mapping states its rule before outcomes are known. Actual order receipts must still identify the exchange contract and executable prices. Missing constituent data limit that reconstruction; they do not make the platform fraudulent.

A versioned record to retain

Record chart symbol, trading symbol, roll rule, adjustment type, data snapshot, session template and the time each month became active. Store the original and revised exports separately. If the provider supports only a continuous research reference, label its results that way. For an execution claim, request the specific contract, order state and fill ledger. The useful conclusion may be that a historical hypothesis can be reproduced while its broker implementation remains untested; there is no need to invent a subscriber return to complete the analysis.

When comparisons are comparable

Two records with the same root ticker may still use different cutover dates and transformed price levels. Align construction choices before comparing signal frequency or threshold behaviour. Then disclose any differences that remain. This review treats a back-adjusted reference as an analytical object rather than rejecting it wholesale. Its proper role is the question: a strategy diagnostic can be informative, while a claim that a customer traded at every synthetic price requires evidence the diagnostic does not contain.

Completed fictional record

Fictional contracts A and B trade at 100 and 103 at the selected roll boundary. A difference-adjusted display adds three to earlier A prices, so an earlier stored A observation of 98 displays as 101. No customer acquired A at 101 merely because the chart changed. An original alert at 98 remains an alert at the original series level; its actual fill is unknown unless a separate A receipt exists. The ledger keeps A=98, transformation=+3, display=101 and the B reference=103 as four different fields. A later screenshot cannot merge them into one entry. The example reconciles a display transformation, not a profitable trade.

Recorded objectValue or stateInterpretation
Original A observation98Research observation
Adjustment+3Display transformation
Rebuilt display101Not a fill
New B reference103Separate contract

Primary references and scope

Public documents inspected 13 October 2026. Cases here are original fiction; sources support bounded mechanisms or product functions, not an account test.

Continue this investigation