Interpolating commodity futures prices with Kriging

TL;DR

This paper explores using Kriging, a Bayesian interpolation method, to construct commodity futures price curves by incorporating trends, seasonalities, and bid-ask spreads, aiding market participants in forecasting future prices.

Contribution

It introduces a novel application of Kriging for interpolating commodity futures prices, accounting for market-specific features like bid-ask spreads and seasonal patterns.

Findings

Kriging effectively models futures price term structures.

Incorporating bid-ask spreads improves interpolation accuracy.

The method captures seasonal and trend components in futures prices.

Abstract

The shape of the futures term structure is essential to commodity hedgers and speculators as futures prices serve as a forecast of future spot prices. Commodity markets quotes futures prices on a selection of maturities and delivery periods. In this note, we investigate a Bayesian technique known as Kriging to build a term structure of futures prices by embedding trends and seasonalities and by taking into account bid-ask spreads of market quotations on different delivery periods.