Presented analysis of gas price formation mechanism in Germany was prompted by changes brought about by technological advancements and the liberalization and harmonization of natural gas markets in the European Union after the year 2000. Because the data used in the study is generated by nonstationary stochastic processes, the cointegrated vector autoregressive model was applied as the most appropriate. The analysis pointed out that the price of natural gas, oil and the USD/EUR exchange rate influence each other in the long run and thus should be modelled together. Gas price in Germany is driven by both fundamental and financial factors, and so it rises with economic expansion, oil price increases, and the depreciation of the USD. It also reacts to changes in short-term interest rates and the volume of gas production in the US, which confirms that the shale revolution in this country has been consequential for gas prices in Europe, like any other supply shock would have been.
Funding source: Narodowe Centrum Nauki
Award Identifier / Grant number: OPUS 21: DEC—2021/41/B/HS4/04317
The second author kindly acknowledges financial support from National Science Centre under OPUS 21: DEC—2021/41/B/HS4/04317.
A.1. Tables and figures
Table 2 presents the definitions of variables used in the empirical analysis. Figure 3 depicts the residuals of CVAR equations explaining prices of gas, oil and exchange rate. Table 3 presents the results of unit root tests.
|P g||Natural gas price||Import price index (excluding taxes, duties etc.)||Federal statistical office – www.destatis.de|
|P o||Crude oil price||To calculate real price of oil, nominal price of Europe brent spot price FOB (dollars per barrel) was deflated by the EU’s CPI||Energy information administration www.eia.gov (data on brent spot price); OECD https://stats.oecd.org (data on CPI)|
|ex t de||USD/EUR real exchange rate||The real USD/EUR exchange rate was calculated as: ex t de = p t − pe t − s t , where P t and PEt are the US and German CPIs, respectively. S t stands for the nominal USD/EUR exchange rate. Small letters denote natural logarithms||https://www.investing.com (data on nominal interest rate), ECD – https://stats.oecd.org (data on CPIs)|
|Y t act||Economic activity||Y t act is a OECD’s industrial production index. It accounts for the output of the mining, manufacturing, electricity, gas and steam and air-conditioning sectors of the OECD’s countries||OECD – https://stats.oecd.org|
|SI t||The relative stock exchange index||The stock exchange rate index in Germany was measured against the US′ stock exchange rate index SI t = Rn t DJ/Rn t DAX (Rn t DAX, Rn t DJ stand for DAX and dow Jones indexes, respectively). The monthly average value of the indexes on closing was used||https://finance.yahoo.com/|
|R t US||US′ short term interest rate||The real short-term interest rate in the US is calculated as nominal three-month treasury bill rate adjusted for inflation according to the equation: R t US = Rn t US − (P t − P t−s )/P t−1(Rn t US stands for nominal interest rate, P t is US′ CPI)||OECD – https://stats.oecd.org|
|Gas production in the US||U.S. Natural gas gross withdrawals (MMcf). The data were deseasonalized with TRAMO-SEATS procedure||Energy information agency – www.eia.gov|
|vol t||Volume of futures and options contracts on energy products||Total monthly volume of futures and options contract on the ICE futures europe market||www.theice.com|
Year 2015 is the base period for indexes.
|H 0: y ∼ I(1)||H 0: y ∼ I(0)|
|No intercept no trend||With intercept||With intercept and trend||With intercept||With intercept and trend|
A.2. Additional Empirical Results
The CVAR model presented in section 4 of the paper has 8 variables comprising the vector: . As before, the optimal lag length of 2 months was selected and two dummy variables were included (first with a value of 1 in September 2008, and the second one taking a value of 1 in November and December 2015). Assuming that the model has only two long-run relationships (see the cointegration test results, Table 1), in the next step, the existence of weakly exogenous variables was tested. Both economic activity and the volume of futures and options contracts on the ICE exchange were found to be weakly exogenous (LR = 4.82; p = 0.31). The repeated trace test and the maximum eigenvalue test confirmed the previous results (pointing to the existence of 4 and 2 long-term relationships, respectively). After imposing the necessary restrictions, the following results were obtained (the figures in the parentheses are t-Student statistics):
The restrictions enabled all cointegrating vectors (LR = 5.09; p = 0.40) with full economic interpretation to be identified. The system adjusts to the equilibrium path as α 11 = −0.02, α 32 = −0.13 and the tests confirm that the residuals are stationary.
The key conclusions concerning gas price formation following from the above results and those presented in Section 4 are virtually the same. The price of gas in Germany goes up with a rising price of oil, expanding economic activity, and a depreciating US dollar. Increasing gas production in the US reduces the price of gas in Germany. A falling interest rate boosts investments in the commodity market, raising the price of gas. According to the second cointegrating vector, a rising price of oil increases the USD/EUR exchange rate and so does gas production in the US. Increases in the stock index ratio and in the volume of futures contracts and options on the ICE reduce the USD/EUR exchange rate.
Auping, W.L., Pruyt, E., de Jong, S., and Kwakkel, J.H. (2016). The geopolitical impact of the shale revolution: exploring consequences on energy prices and rentier states. Energy Pol. 98: 390–399, https://doi.org/10.1016/j.enpol.2016.08.032.Search in Google Scholar
Bachmeier, L. and Griffin, J.M. (2006). Testing for market integration crude oil, coal and natural gas. Energy J. 27: 55–71, https://doi.org/10.5547/issn0195-6574-ej-vol27-no2-4.Search in Google Scholar
Brown, S.P.A. and Yücel, M.K. (2008). What drives natural gas prices? Energy J. 29: 45–60, https://doi.org/10.5547/issn0195-6574-ej-vol29-no2-3.Search in Google Scholar
CEER (2011). CEER vision for a European gas target model. Conclusions paper. Ref. C11-GWG-82-03 (Accessed 1 December 2011).Search in Google Scholar
Davidson, R. and MacKinnon, J.G. (1993). Estimation and interference in econometrics. Oxford University Press, New York.Search in Google Scholar
Frankel, J.A. (2014). Effects of speculation and interest rates in a “carry trade” model of commodity process. J. Int. Money Finance 42: 88–112, https://doi.org/10.1016/j.jimonfin.2013.08.006.Search in Google Scholar
Fratzscher, M., Schneider, D., and Robays, I. (2014). Oil prices, exchange rates and asset prices. In: Working paper series no. 1689. European Central Bank.10.2139/ssrn.2442276Search in Google Scholar
Gomez, V. and Maravall, A. (1996). Programs TRAMO and SEATS, instructions for the users. In: Working paper no. 9628. Banco de Espana.Search in Google Scholar
Guerra, A., Shen, A., and Zhao, T. (2012). Determinants of natural gas spot prices, FMI 3560-01. Available at: https://silo.tips/download/determinants-of-natural-gas-spot-prices#modals.Search in Google Scholar
Habro, I., Johansen, S., Nielsen, B., and Rahbek, A. (1998). Asymptotic inference on cointegrating rank in partial systems. J. Bus. Econ. Stat. 16: 388–399, https://doi.org/10.2307/1392608.Search in Google Scholar
IGU (2019). Global gas report 2019, Available at: https://www.igu.org/resources/global-gas-report-2019-2/ (Accessed 25 May 2020).Search in Google Scholar
Kębłowski, P., Leszkiewicz-Kędzior, K., and Welfe, A. (2020). Real exchange rates, oil price spillover effects, and tripolarity. E. Eur. Econ. 58: 415–435, https://doi.org/10.1080/00128775.2020.1753212.Search in Google Scholar
Kwiatkowski, D., Phillips, P., Schmidt, P., and Shin, Y. (1992). Testing the null hypothesis of stationarity against the alternative of unit root. J. Econom. 54: 159–178, https://doi.org/10.1016/0304-4076(92)90104-y.Search in Google Scholar
KEMA. (2013). Entry-exit regimes in gas. A project for the European Commission. DG ENER, https://ec.europa.eu/energy/sites/ener/files/documents/201307-entry-exit-regimes-in-gas-parta.pdf.Search in Google Scholar
L’Hegaret, G., Siliverstovs, B., and Hirschhausen von, C. (2005). International market integration for natural gas? A cointegration analysis of gas prices in Europe, North America and Japan. Energy Econ. 27: 603–615.10.1016/j.eneco.2005.03.002Search in Google Scholar
Lütkepohl, H., Saikkonen, P., and Trenkler, C. (2001). Maximum eigenvalue versus trace tests for the cointegrating rank of a VAR process. Econom. J. 4: 287–310, https://doi.org/10.1111/1368-423x.00068.Search in Google Scholar
Sartore, D., Trevisan, L., Trova, M., and Volo, F. (2002). US dollar/euro exchange rate: a monthly econometric model for forecasting. Eur. J. Finance 8: 480–501, https://doi.org/10.1080/13518470210160894.Search in Google Scholar
Villar, J.A. and Joutz, F.L. (2006). The relationship between crude oil and natural gas prices. Energy Information Administration, Office of Oil and Gas. https://www.semanticscholar.org/paper/The-Relationship-Between-Crude-Oil-and-Natural-Gas-Villar-Joutz/bad22c2a9a6227f401f2959e795e9b179192e446.Search in Google Scholar
Wu, Y. (2011). Gas market integration: global trends and implications for the EAS region. In: ERIA discussion paper series, ERIA-DP-2011-07.Search in Google Scholar
© 2022 Walter de Gruyter GmbH, Berlin/Boston