Backtesting term

Continuous (back-adjusted) futures contract

A continuous futures contract is a chain of expiring contracts glued into one line so a strategy can be tested across years instead of months. How you glue it decides whether the edge you see afterward is real.

A single futures contract trades for a few months, then expires. To test a strategy across 13 years of data you need one unbroken line, so you splice the expiring contract onto the next one, and the next, into a single continuous (back-adjusted) contract. Do the splice carelessly and every roll date leaves a phantom gap or a phantom profit in the chart, a jump no trader alive ever actually collected.

Back-adjusting is the fix: shift every price before each roll by the size of the gap, and the seam disappears, so the chart reads as one smooth line instead of a staircase of jumps. That's necessary, and mostly harmless close to today's date. Go back far enough, though, and the stacked shifts can push old prices toward zero, even negative, and a careless percentage-return calculation on that adjusted series stops meaning what it looks like it means.

The trap is quiet because it never announces itself. The equity curve looks clean and the strategy looks profitable. The real cause is how a dozen separate contracts got glued together, and it has nothing to do with the rules. It's close kin to look-ahead bias: both hand a backtest money that was never collectible, one by leaking future information into the past, the other by leaking a splicing artifact into the price series itself.

We run 13 years of CME futures and handle every roll on purpose instead of letting default settings decide, because a bad stitch is enough to manufacture an edge out of nothing, no rule change required. The check that matters is simple: whether the adjusted series still matches the price a trader could actually have gotten filled at, roll after roll, for the life of the backtest.

Get that wrong and the backtest stops testing your strategy and starts testing an artifact: a contract that never existed. The chart looks real. The fills inside it aren't. That gap between a smooth continuous series and the choppy, expiring reality underneath it is exactly where a backtest can quietly stop being a backtest of anything tradeable at all.

The research behind this

External research, linked for context and further reading. FoxAlgo is independent and not affiliated with these authors or publishers.

These are the terms behind The No List — the full audit, every strategy and indicator named, with its verdict and the exact reason it lived or died.

Get The No List →