Continuous Futures Series and Roll-Return Construction
Summary
The document explains why futures backtests need a method for joining prices from contracts with different expiration dates. Contango and backwardation can create price gaps at the splice, so the article compares three approaches: additive Panama adjustments, proportional adjustments, and a weighted rollover series. It describes how additive shifts can distort historical levels and returns, while proportional adjustments preserve percentage continuity but complicate signals based on absolute prices.
The article focuses on a perpetual series that gradually shifts weight from a farther contract to the nearer one over several business days. A Python example builds a date-indexed weight matrix from settlement dates and applies it to two WTI crude oil contracts; the displayed output illustrates a continuous transition. This series can support backtest research, but gradual rolling entails holding both contracts and may raise transaction costs. The example is limited to two contracts and assumes suitable settlement data and roll dates; a continuous series is a research construction, not a record of a single tradable contract.
Key ideas
- Futures expirations create overlapping contract histories that must be joined for continuous backtests.
- Panama adjustment can introduce drift and distort historical price levels and return calculations.
- Proportional adjustment preserves percentage continuity but requires adjusting signals based on absolute prices.
- A weighted rollover transitions exposure from far to near contracts across several days.
- The rollover example builds weights from expiry dates and applies them to settlement prices.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.