The efficiency of natural gas futures markets

AbstractRecent experience with the emergence of futures markets for natural gas has led to many questions about the drivers and functioning of these markets. Most often, however, studies lack strong statistical support.The objective of this article is to use some classical statistical tests to check whether futures markets for natural gas (NG) are efficient or not. The problem of NG market efficiency is closely linked to the debate on the value of NG. More precisely, if futures markets were really efficient, then: 1) spot prices would reflect the existence of a market assessment, which is proof that speculation and the manipulation of prices are absent; 2) as a consequence, spot prices could give clear signals about the value of NG; and 3) historical series on spot prices could serve as “clean” benchmarks in the pricing of NG in long‐term contracts. On the whole, since the major share of NG is sold to power producers, the efficiency of futures markets implies that spot prices for NG are driven increasingly by power prices.On the other hand, if futures markets for natural gas fail the efficiency tests, this will reflect: 1) a lack of liquidity in futures markets and/or possibilities of an excess return in the short term; 2) a pass‐through of the seasonality of power demand in the gas market; 3) the existence of a transitory process, before spot markets become efficient and give clear signals about the value of NG.Using monthly data on three segments of the futures markets, our findings show that efficiency is almost completely rejected on both the International Petroleum Exchange in London (UK market) and the New York Mercantile Exchange (US market). On the NYMEX, the principle of “co‐movement” between spot and forward prices seems to be respected. However, the autocorrelation functions of the first differences in the price changes show no randomness of price fluctuations for three segments out of four. Further, both the NYMEX and the IPE fail, with regard to the hypothesis that the forward price is an optimal predictor of the spot price. Consequently, unless we have an increase in the liquidity of spot markets and an increase in the relative share of NG spot trading, futures markets cannot be considered as efficient.