Dr Clinton Levitt, Copenhagen Business School
I develop a structural model of exploration and extraction that can be used to estimate the effectiveness of different royalty schemes. I focus on four targets: (1)government revenue; (2)where firms explore; (3)volume extracted from existing wells; and, (4)in-situ recovery rates (investment in enhanced recovery methods). Using data from Alberta, Canada, I constructed an extensive geographic information system. These data contain information on over 400 thousand individual wells. For each well, the data include the latitude and longitude, the owner, monthly extracted volumes over the lifetime of the well, and production dates. With these data, I can construct individual production histories of over 400 thousand wells. In addition, data on over 40 thousand hydrocarbon pools are used to construct accurate reserve estimates for each well. Finally, the pool data can be used to accurately estimate extraction costs at the pool level. These data represent a significant improvement over data that have been previously used to study oil and gas investment decisions. These novel data allow me to estimate a sophisticated structural model of firms’ investment decisions.
This study has the potential fill an important gap in the literature and be an important reference for policy makers. Indeed, there has been a renewed interest in understanding the effects of regulatory policy on investment decisions for a number of reasons. First, recent volatility in hydrocarbon prices combined with near-record high prices and record profits have motivated policy makers to reevaluate existing royalty schemes. Second, current debates about peak oil has sparked interest in understanding firms’ exploration decisions over land with heterogenous characteristics. Third, the peak oil debate has also created the need to study how policy can encourage firms to extract more resources from existing pools and invest in enhanced recovery methods. Fourth, land-use policy has gained prominence because of concerns over environmental degradation of production sites.
Investigating policy in the oil and gas industry requires modeling three interrelated decisions: (1)whether to explore; (2)where to explore; and (3) the volume hydrocarbons to extract. Moreover, because exploration and extraction occur over long periods, to capture the intertemporal tradeoffs firms consider, the model must be dynamic. Studying firms’ incentives to explore different areas and their incentives to increase extraction rates from existing pools, requires incorporating the heterogeneity that exists, in both production and cost characteristics, across large regions. Modeling this heterogeneity is extremely important because it essentially represents the firms’ choice set. For instance, if each location were treated as identical, in terms of drilling and extractions costs, then there is no choice concerning where to explore or from which well to extract oil because there is no opportunity cost for choosing one site over another. In fact, the choice is simply, to drill or not to drill, and how much to extract. But, the interesting questions involve the choices firms’ make between different exploration sites and between different extraction possibilities and how policy influences these choices—these questions have not been addressed.
The parameters of the structural model are estimated using general methods of moments (GMM). The theoretical model will produce, for each location, two Euler equations: one illustrating the intertemporal tradeoffs between exploring a field today versus waiting till next period, the other describing the intertemporal tradeoffs between extracting hydrocarbons today versus waiting till next period. From these Euler equations, I will be able to derive moment conditions that are used to construct the GMM estimator.