Previous research on extreme values has been used in commodity pricing, more specifically in trying to capture the dynamic behavior of the random occurrences of extreme events, such as price spikes/drops. This research uses the records theory to study the effect of extreme gas prices and the probability of future records. To our knowledge, records theory was not previously applied to gas markets. The aim of this study is to test the stability of three different regional gas markets, each having its own supply and demand characteristics. Records theory, studies observations that are higher than all previous ones, which is equivalent to say the maximum/minimum observation up to present time, and those records, are concentrated in the tail of a given distribution. In this study, several models are developed to test and analyze the stability of three main regional gas markets (U.S, Europe and Asia). The classical model is used for the case where gas prices are independent and identically distributed (i. i. d case). Alternative models, such as Yang model and the discrete-time random walk, are used, where the number of records grows faster than in the i. i. d case and where records are not only concentrated among the first observations. In spite of the non-independent and non-identically distributed properties of the models, the results are distribution free. Consequently, the applicant will not be concerned by identifying the distribution type and the complexity of the models is reduced.
