Skip to main content

LUP Student Papers

LUND UNIVERSITY LIBRARIES

Crack spreads in crude oil futures - Hedging and speculation in the oil market

Raihle, William and Moell, Gustaf (2005)
Department of Economics
Abstract
This study was performed using daily spot prices, from January 1997 through July 2004; of the three assets that constitute the crack spread i.e. crude oil, gasoline and heating oil - all listed on the New York Mercantile Exchange (NYMEX). Our objective has been to characterize the dynamics of the spread and subsequently create trading strategies to yield returns that surpass that of a passive approach. We used a simple AR (1)-model that incorporates historical prices and the existence of auto regression for predicting future spread-prices; in addition a GARCH (1,1) model which uses the heteroskedastic features in time series to predict volatility. The AR (1)-model yielded somewhat inconclusive results, while the GARCH model proved to be... (More)
This study was performed using daily spot prices, from January 1997 through July 2004; of the three assets that constitute the crack spread i.e. crude oil, gasoline and heating oil - all listed on the New York Mercantile Exchange (NYMEX). Our objective has been to characterize the dynamics of the spread and subsequently create trading strategies to yield returns that surpass that of a passive approach. We used a simple AR (1)-model that incorporates historical prices and the existence of auto regression for predicting future spread-prices; in addition a GARCH (1,1) model which uses the heteroskedastic features in time series to predict volatility. The AR (1)-model yielded somewhat inconclusive results, while the GARCH model proved to be distinctively more accurate. We tested a number of strategies within the realms of each model, in order to find the one with the highest return. A number of strategies, primarily based on the GARCH (1,1) forecast, generated returns exceeding that of a passive approach . The best prediction of the spread was within the GARCH model when using
different kinds of volatility measures as a screening process for market intervention, taking a long position in the 1:1 crack spread. We can also conclude that it proved difficult to reach a higher percentage of successful trades than about 55% in any of the strategies when using our day-trading approach. (Less)
Please use this url to cite or link to this publication:
author
Raihle, William and Moell, Gustaf
supervisor
organization
year
type
H1 - Master's Degree (One Year)
subject
keywords
crack spread, crude oil, intercommodity spreads, 1), AR(1), GARCH (1, Economics, econometrics, economic theory, economic systems, economic policy, Nationalekonomi, ekonometri, ekonomisk teori, ekonomiska system, ekonomisk politik
language
English
id
1644325
date added to LUP
2005-06-22 00:00:00
date last changed
2010-08-05 14:10:04
@misc{1644325,
  abstract     = {{This study was performed using daily spot prices, from January 1997 through July 2004; of the three assets that constitute the crack spread i.e. crude oil, gasoline and heating oil - all listed on the New York Mercantile Exchange (NYMEX). Our objective has been to characterize the dynamics of the spread and subsequently create trading strategies to yield returns that surpass that of a passive approach. We used a simple AR (1)-model that incorporates historical prices and the existence of auto regression for predicting future spread-prices; in addition a GARCH (1,1) model which uses the heteroskedastic features in time series to predict volatility. The AR (1)-model yielded somewhat inconclusive results, while the GARCH model proved to be distinctively more accurate. We tested a number of strategies within the realms of each model, in order to find the one with the highest return. A number of strategies, primarily based on the GARCH (1,1) forecast, generated returns exceeding that of a passive approach . The best prediction of the spread was within the GARCH model when using
different kinds of volatility measures as a screening process for market intervention, taking a long position in the 1:1 crack spread. We can also conclude that it proved difficult to reach a higher percentage of successful trades than about 55% in any of the strategies when using our day-trading approach.}},
  author       = {{Raihle, William and Moell, Gustaf}},
  language     = {{eng}},
  note         = {{Student Paper}},
  title        = {{Crack spreads in crude oil futures - Hedging and speculation in the oil market}},
  year         = {{2005}},
}