Did Amaranth Advisors LLC engage in interday price manipulation in the natural gas futures market?

Abstract

Price manipulation in US energy markets has become a heated political and economic issue because of events such as the Enron scandal, California's electricity crisis, and volatile movements in natural gas and petroleum prices. This article tests to determine whether Amaranth Advisors LLC manipulated natural gas prices during 2006. Amaranth is of particular interest because, during 2006, it controlled as much as 80 per cent of the natural gas open interest on the New York Mercantile Exchange and InterContinental Exchange. As a result of its dominant market position and suspicious trading activities, the Commodity Futures Trading Commission (CFTC) charged Amaranth in 2007 with attempted intraday price manipulation in the natural gas derivatives market, and the Federal Energy Regulatory Commission (FERC) followed the day after with charges that the fund engaged in perfected intraday price manipulation in the natural gas spot market. This study complements the CFTC and FERC investigations by testing whether Amaranth engaged in interday price manipulation in the natural gas derivatives markets during 2006. Using Granger causality tests, we find no convincing evidence of one-way causality from changes in Amaranth's positions to changes in the prices of natural gas derivatives contracts.

Detecting price artificiality and manipulation in futures markets: An application to Amaranth

“…Pirrong (2004) applies his methodology to the specific case of Ferruzzi soybean manipulation claim and concludes that soybean futures prices in the Chicago area were artificial as a result of Ferruzzi's manipulative activities in 1989. Marthinsen and Gai (2010a) used a Granger causality model to analyze whether Amaranth's trades affected the prices of natural gas futures in 2006. In a follow-up article, the authors examined whether Amaranth's spread trading affected prices of calendar spreads, particularly the winter-summer spreads (Marthinsen and Gai, 2010b).…”

“…We follow the Granger causality specification set forth in Marthinsen and Gai (2010a), who base their approach on the work of Hartzmark (1991), Wang (2001), and Sanders et al (2004)…”