Market integration among electricity markets and their major fuel source markets James W. Mjelde ⁎, David A. Bessler Department of Agricultural Economics, Texas A&M University, College Station, TX 77845-2124, United States abstract article info Article history: Received 7 August 2008 Received in revised form 7 January 2009 Accepted 9 February 2009 Available online 20 February 2009 Keywords: VECM Market dynamics Electricity prices Fuel prices Dynamic price information flows among U.S. electricity wholesale spot prices and the prices of the major electricity generation fuel sources, natural gas, uranium, coal, and crude oil, are studied. Multivariate time series methods applied to weekly price data show that in contemporaneous time peak electricity prices move natural gas prices, which in turn influence crude oil. In the long run, price is discovered in the fuel sources market (except uranium), as these prices are weakly exogenous in a reduced rank regression representation of these energy prices. © 2009 Elsevier B.V. All rights reserved. 1. Introduction One goal in deregulating energy markets is to allow the markets respond to supply and demand conditions. As a result of deregulation, more competitive and interrelated market environments are devel- oping in the electricity and natural gas markets (Park et al., 2006, 2008). These changes imply that price determination is more likely to be in the hands of the market participants than in the regulators' hands. Having market participants determining price may allow participants to more quickly respond to changes in major input prices. Specifically, for the electricity market, electricity spot markets may respond to price changes in its major fuel source markets. Further, fuel source prices may in turn respond to changes in electricity prices (Asche et al., 2006). Given the nature and use of the different fuel sources, one would further expect fuel source prices to be at least weakly integrated (Bachmeier and Griffin, 2006). Economic theory and intuition suggest a relationship should exist between input and output (product) prices. Consider the simplest case of a single factor of production (input) used to produce a single product. A static supply and demand model suggests increasing the marginal cost of the input leads to an increase in the product price, ceteris paribus. Likewise, an increase in demand leads to an increase in quantity demanded for the product, therefore, a higher price. Associated with these changes are increasing marginal costs caused by the increased use of the input. Economic theory, however, does not state how such relationships will respond in a dynamic framework which includes feedback in a non-ceteris paribus environment. Further complicating the issue are numerous locations using multiple inputs for power generation with different substitutability and complementary relationships. The degree of price transmission from input to output, therefore, may depend on the cost share of the input factor in question. This discussion suggests at least three efficiencies lead to input and output prices being related, time (dynamics), space (location), and form (transformations). Electricity markets provide a unique opportunity to examine the relationships between input and output prices in a non-ceteris paribus environment. The objective of this study is to characterize the dynamic price information flows among U.S. electricity wholesale spot peak and off- peak prices and the prices for their major fuel sources. Providing information on the dynamics of electricity and fuel source prices allows for a better understanding of price information flow among the markets. To accomplish this objective, interrelationships among electricity prices from two diverse markets, Pennsylvania, New Jersey, Maryland Interconnection (PJM) and Mid-Columbia (Mid-C), and four major fuel source prices, natural gas, crude oil, coal, and uranium, are examined. For several reasons, two electricity spot markets are used. First, Park et al. (2006) find electricity markets in the Western U.S. are separated form the Eastern markets in contemporaneous time, but this separation disappears at longer time horizons. The relationships between the markets are a function of physical assets (such as transmission lines) and institutional arrangements. Second, electricity generation by fuel source varies within regions of the U.S. (Fig. 1). Within the two spot markets, both firm peak and non-peak prices are included because these prices may respond differently to input price changes. To our knowledge, no study to date has examined interdependen- cies among four electricity and four fuel source prices using a multivariate time series framework. This formulation allows for the inclusion of all three forms of efficiency, although individual contributions can not be determined. A vector error correction Energy Economics 31 (2009) 482–491 ⁎ Corresponding author. E-mail addresses: j-mjelde@tamu.edu (J.W. Mjelde), dab@ag.tamu.edu (D.A. Bessler). 0140-9883/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.eneco.2009.02.002 Contents lists available at ScienceDirect Energy Economics journal homepage: www.elsevier.com/locate/eneco