Application of the Schwartz-Smith Model (2000) in Copper Derivatives Pricing

Peer reviewed: 
No, item is not peer reviewed.
Scholarly level: 
Graduate student (Masters)
Date created: 
Copper futures
Analysts' forecasts
Two-factor model
Kalman filter

In this paper, we explore the use of Schwartz and Smith two-factor model incopper pricing. We used both Copper future data from LME and Analyst Forecast datafrom Bloomberg (LME) and World Bank as input to generate futures curve and spotcurve. The Schwartz- Smith model incorporates the long-term equilibrium prices thatcommodity price will approach in the long-term and short-term mean reversioncharacteristic of commodity prices. To estimate the state variables and modelparameters, Kalman filter technique was used to update the state variables throughiteration and Maximum likelihood approximation to compute the term structure, sinceKalman filter is able to estimate model's parameters when the model relies on nonobservabledata. This model is able to explain the copper's term structure in an intuitiveway. We begin by describing the input data in section 2 and explaining the short-term andlong-term model in section 3. In section 4, we discuss the estimation process using theKalman filter and, in section 5 we describe the empirical result by applying the model toCopper futures and forecast data. In section6, we offer the concluding remarks.


MSc in Finance Project-Simon Fraser University

Document type: 
Graduating extended essay / Research project
Copyright remains with the author.
Senior supervisor: 
Eduardo Schwartz
Beedie School of Business-Segal Graduate School