Zusammenfassung:
The U.S. natural gas market is crucial for domestic energy and increasingly important in global trade. Structural analyses of this market often adapt oil market models but overlook key features, such as external trade flows, potentially limiting their ability to fully capture its dynamics. This paper extends these analyses by developing a Structural Vector Autoregression model that incorporates external gas flows and distinguishes between domestic and export-driven demand shocks, contributing to policy discussions on price fluctuations, particularly after the surge in U.S. gas exports following the Russia-Ukraine war. The model uses monthly data to reduce information loss and better capture market dynamics compared to models using quarterly data. The results indicate that supply and domestic demand shocks cause price overshoots, followed by a steady decline, with limited effects on economic activity. Export demand shocks cause short- and medium-term price increases and gradually expand supply, while inventory demand shocks trigger brief price spikes with minimal long-term impact. The analysis reveals that failing to control for extreme values in COVID-period data yields counterintuitive results, such as reduced gas supply boosting economic activity. A decomposition of 2022-2023 price fluctuations shows domestic demand and inventory demand shocks were the main drivers, while export demand shocks-though important-played a smaller role, influencing prices through alternating effects from increased LNG exports and maintenance disruptions. Finally, the estimated elasticities suggest that natural gas supply is unresponsive to short-term price changes, while demand exhibits limited responsiveness.