Continuous Futures and Back-Adjustment

Each futures contract lives for a limited time, so there is no single price history for "crude oil futures" or "E-mini S&P 500 futures" the way there is for a stock. To chart long term trends or backtest strategies, traders build continuous futures series by joining contracts together. How the contracts are stitched, when to switch from one to the next and whether to adjust for price gaps at each roll, makes a big difference to what the chart shows and to the results of any backtest.

The roll gap problem#

Different contract months trade at different prices. When a series switches from the expiring contract to the next one, there is a jump that reflects the curve, not a real market move.

Methods#

Back adjustment in detail#

Back adjustment keeps the most recent contract's prices as they are and shifts all earlier prices by the cumulative roll gaps.

adjusted past price = actual past price + sum of later roll gaps (new minus old)

Result: day to day changes in the series match what a trader rolling the position would have earned or lost, so profit and loss in backtests is realistic. The level of prices long ago is artificial: a series that has been back adjusted for decades of contango may show early prices far below what was actually traded, sometimes below zero.

Ratio adjustment#

Ratio adjustment multiplies earlier prices by the ratio of the new contract price to the old on each roll date. It preserves percentage returns, which suits indicators based on percentages and long histories. It does not preserve dollar profit and loss per contract.

Roll rules#

The roll date affects the series:

• On expiry: may include illiquid final days.
• Fixed days before expiry: for example, 5 days before last trading day, or before first notice day for physical contracts. See First Notice Day and Last Trading Day.
• Volume or open interest based: roll when the next contract becomes more active. This matches where traders actually trade.

Whatever rule is used for the data should match the rule used in trading. See Rolling Futures Contracts.

Why it matters for backtesting#

• Unadjusted data creates fake profits or losses at rolls.
• Back adjusted data gives correct P&L but wrong price levels, which breaks rules based on absolute price levels or percentage moves from old prices.
• Ratio adjusted data gives correct percentage changes but wrong point values.
• Indicators such as moving averages can give different signals on different series.

Good practice is to compute signals on an appropriate continuous series and compute P&L from the actual contracts traded. See Backtesting Methodology and Historical Data for Backtesting.

Comparison with adjusted stock prices#

Stocks have a similar issue with dividends and splits, handled by adjusted close prices. The logic is the same: adjust history so that returns are correct, accepting that old price levels no longer match what was printed at the time. See Splits and Dividends in Price Data.

Common mistakes#

• Backtesting on unadjusted continuous data.
• Using back adjusted levels for rules like "buy below $50".
• Ignoring the roll rule in data and in trading.

Frequently asked questions#

What is a continuous futures contract?#

A price series created by joining successive futures contracts, used for long term charts and backtests.

What is back adjustment?#

A method that shifts all past prices by the price gaps at each roll, so the series has no artificial jumps and daily changes match real profit and loss.

Why do back adjusted futures prices sometimes go negative?#

Because cumulative adjustments for many contango roll gaps can push old prices below zero, even though actual traded prices were positive.

Next, learn how commodities change hands in Physical Delivery vs Cash Settlement.