This long term forecast of the price of natural gas at the citygate in each of the fifty U.S. states differs from others because of the oil price forecast that feeds into it. The oil price forecast combines the empirical restriction that prices to 2060 will be just as volatile as they have been throughout the "OPEC Era" of oil pricing, since 1973, with the theoretical restriction that OPEC maximizes its profits in a forward looking way. There is a theoretical and empirical case that volatility, itself, is profitable to OPEC, delineated in four articles that trace the forecast's main intellectual thread:
Pindyck, R.S. (1978). Gains to producers from the cartelization of exhaustible resources, Review of Economics and Statistics 60(2), 238-251, April. http://www.jstor.org/stable/1924977
Mork, K.A. (1994). Business cycles and the oil market, Energy Journal 15, special issue, 15-38, July. https://doi.org/10.5547/ISSN0195-6574-EJ-Vol15-NoSI-3
Vatter, M. (2017). OPEC’s kinked demand curve, Energy Economics 63, 272-287, March. https://doi.org/10.1016/j.eneco.2017.02.010
Vatter, M. (2019). OPEC’s risk premia and volatility in oil prices, International Advances in Economic Research 25(2), 165-175, May. https://doi.org/10.1007/s11294-019-09734-7
Vatter, M. (2026). A volatile oil price forecast, USAEE Working Paper No. 23-610, SSRN Electronic Journal, 0[10.2139/ssrn.4670880]. http://dx.doi.org/10.2139/ssrn.4670880
The oil price forecast used here is Scenario 1 from Vatter (2026). To make it more practical, using conventional econometric methods, I have extended the forecast downstream to natural gas by state, which affects the value of numerous assets in the energy industries and beyond. In many states, volatility in the price of natural gas is driven largely by weather, both as regular seasonal variation and deviations therefrom. Here, I layer the former, but not the latter, on top of the volatility in the price of crude oil. The model quantifies the risk not forecasted, and could, then, be used for Monte Carlo simulations conditioned on the forecast, a smaller problem than simulating unconditioned price risk. (The forecast also includes benchmark prices for diesel.)
The forecast shows weaker convergence in trend and volatility across states than was observed historically, and the econometric model is designed to reflect both. Earlier pipelines followed the highest value, lowest cost routes, and many, though not all, regional markets were integrated. At the long term margin today, returns have diminished, costs, including administrative costs, have risen, and a monetized social preference for environmental quality that rises more than in proportion to income has contributed to that. The high cost of expanding pipelines in the densely populated Northeast has a similar effect on persistence of price shocks as the long distances that pipelines must cover in the sparsely populated Great Plains and Rockies.
The distinct feature here, though, is OPEC-driven volatility. At any point in time since the mid-1970s, one could look back to 1973 and see volatility in the price of crude oil, but turn forward and not see it in forecasts or futures curves. Therefore, I do not hesitate to recommend this forecast as a base case in asset-evaluation, but, if not, also recommend it as a tool to stress test such evaluations against a smooth base case. I offer this scenario free of charge to encourage analysts to use it. I used SciSpace in the preparation of the quantitative forecast and documentation. Please email inquiries to Marc Vatter at marcATappliedecon.net.
Pindyck, R.S. (1978). Gains to producers from the cartelization of exhaustible resources, Review of Economics and Statistics 60(2), 238-251, April. http://www.jstor.org/stable/1924977
Mork, K.A. (1994). Business cycles and the oil market, Energy Journal 15, special issue, 15-38, July. https://doi.org/10.5547/ISSN0195-6574-EJ-Vol15-NoSI-3
Vatter, M. (2017). OPEC’s kinked demand curve, Energy Economics 63, 272-287, March. https://doi.org/10.1016/j.eneco.2017.02.010
Vatter, M. (2019). OPEC’s risk premia and volatility in oil prices, International Advances in Economic Research 25(2), 165-175, May. https://doi.org/10.1007/s11294-019-09734-7
Vatter, M. (2026). A volatile oil price forecast, USAEE Working Paper No. 23-610, SSRN Electronic Journal, 0[10.2139/ssrn.4670880]. http://dx.doi.org/10.2139/ssrn.4670880
The oil price forecast used here is Scenario 1 from Vatter (2026). To make it more practical, using conventional econometric methods, I have extended the forecast downstream to natural gas by state, which affects the value of numerous assets in the energy industries and beyond. In many states, volatility in the price of natural gas is driven largely by weather, both as regular seasonal variation and deviations therefrom. Here, I layer the former, but not the latter, on top of the volatility in the price of crude oil. The model quantifies the risk not forecasted, and could, then, be used for Monte Carlo simulations conditioned on the forecast, a smaller problem than simulating unconditioned price risk. (The forecast also includes benchmark prices for diesel.)
The forecast shows weaker convergence in trend and volatility across states than was observed historically, and the econometric model is designed to reflect both. Earlier pipelines followed the highest value, lowest cost routes, and many, though not all, regional markets were integrated. At the long term margin today, returns have diminished, costs, including administrative costs, have risen, and a monetized social preference for environmental quality that rises more than in proportion to income has contributed to that. The high cost of expanding pipelines in the densely populated Northeast has a similar effect on persistence of price shocks as the long distances that pipelines must cover in the sparsely populated Great Plains and Rockies.
The distinct feature here, though, is OPEC-driven volatility. At any point in time since the mid-1970s, one could look back to 1973 and see volatility in the price of crude oil, but turn forward and not see it in forecasts or futures curves. Therefore, I do not hesitate to recommend this forecast as a base case in asset-evaluation, but, if not, also recommend it as a tool to stress test such evaluations against a smooth base case. I offer this scenario free of charge to encourage analysts to use it. I used SciSpace in the preparation of the quantitative forecast and documentation. Please email inquiries to Marc Vatter at marcATappliedecon.net.