| I propose an integrated model of the behavior of a profit-maximizing crude oil producer. I show that the producer's optimization program can be decomposed into an extraction program, which is subject to physical constraints, an investment level program, a development timing program, and an exploration program for new prospects. In the extraction program, I find that it is optimal to extract oil at the maximal feasible rate, unless prices are expected to rise steeply. I also show that such a price process can arise in equilibrium and that prices in such an equilibrium grow at a rate determined by the decrease rate of the maximal feasible extraction rate. In the investment level program, I find that optimal investment is always increasing in the size of the field and, if the current price of oil net of extraction costs is positively related to future net prices, the investment level is also increasing in the net price. In addition, investment increases in the current technology level, unless technological improvements carry very bad news about future prices. Optimal development timing generally has an ambiguous relation to current shock values, and fields are not necessarily developed in a monotonic order by size. However, I establish conditions on the shock process under which development is more likely to occur in higher-price, lower-cost, and higher-technology states and show that the same conditions also ensure that larger fields are developed first. Under the same conditions that ensure monotonic development time dynamics, I show that it is optimal to explore fields in stochastically decreasing order of size and that the total number of prospects explored in a play in a given period responds positively to prices and technology, but is generally declining over the lifetime of the play. The model replicates a number of industry facts. First, crude oil producers respond to price and cost shocks primarily at the extensive margin, by adjusting exploration and development levels. Second, production paths at all levels of aggregation have an inverted-U shape (Hubbert's peak). Third, within each petroleum play the larger deposits tend to be discovered first. Finally, both exploration effort and new discoveries decrease over the lifetime of each exploration region. |
