Published in:
Publication year: 2002
Co-author 1: Paul Stefiszyn
In our previous column, we outlined the development of a multi- factor model of the evolution through time of the forward price curve of a commodity using Principal Component Analysis (PCA), and we briefly introduced the idea of Seasonal PCA. This month, we demonstrate a practical application of the technique.
Certain instruments or contracts in commodity markets have a value that depends on more than one forward price. To value these contracts, it is necessary to develop a realistic model of the simultaneous evolution in time of all the forward prices; that is, a model of the forward price curve. Through the use of these models, we seek to accurately capture the joint behavior of forward prices, and to know how they move together, rather than how they move in isolation.
In our previous column, we outlined the development of a multi- factor model of the evolution through time of the forward price curve of a commodity using Principal Component Analysis (PCA), and we briefly introduced the idea of Seasonal PCA. This month, we demonstrate a practical application of the technique.
Certain instruments or contracts in commodity markets have a value that depends on more than one forward price. To value these contracts, it is necessary to develop a realistic model of the simultaneous evolution in time of all the forward prices; that is, a model of the forward price curve. Through the use of these models, we seek to accurately capture the joint behavior of forward prices, and to know how they move together, rather than how they move in isolation.