GWU study finds direct evidence financial traders affect commodity prices
“A recently-published study by George Washington University and the University of Illinois has examined the relationship between commodity-linked notes and the commodity futures markets,” wrote Jim Collura, vice president for government affairs at The New England Fuel Institute, www.nefi.com.
The study concludes that they ’cause significant price changes in the underlying future markets, and therefore provide direct evidence of the impact of ‘financial’ trades on commodity futures prices.”
Here is the summary:
New study finds Commodity Linked Notes have immediate and permanent impact on commodity futures prices
“CLNs are issued by and are obligations of financial institutions and have payoffs linked to the price or change in price of a commodity or commodity futures contract, commodity index, or basket of commodities or commodity futures.”
The recent paper New Evidence on the Financialization of Commodity Markets1 is an important contribution to the contentious debate over the impact of financial speculation on commodity futures markets. Below is the paper’s abstract, followed by a brief summary of its major findings and an explanation of its significance in relation to recent literature.
Abstract
“Following the recent, dramatic increase in commodity investments by financial institutions, academics, practitioners, and regulators have engaged in a heated debate over whether financial institutions’ trades and holdings have affected commodity prices and their return dynamics. Distinct from the prior literature, this paper examines the price impact of commodity investments on the commodities futures markets using a novel dataset of Commodity-Linked Notes (CLNs). CLN issuers hedge their liabilities by taking long positions in the underlying commodity futures subsequent price changes in the underlying futures markets, and therefore provide direct evidence of “financial” trades on commodity futures prices.”
Summary
“The results [of Henderson, Pearson and Wang (2012)] support the view that the trades of financial institutions play an important role in price formation in the commodities future markets” and, as noted in the abstract above, their approach uses a “novel dataset”‘examining CLN issues and the related hedging transactions in the referenced commodity futures markets’that has significant advantages over previous literature attempting to deal with the financialization hypothesis.
Using “nearly the entire universe” of publicly issued CLNs in the U.S., the results show issuers initial hedging trades for issues with proceeds greater than or equal to $2 million, $5 million and $10 million raise the underlying commodity futures prices by an average of 37, 40, and 51 basis points, with similar results across all commodity sectors for CLNs linked to single commodities and commodity baskets consisting of small numbers of commodities. Combined, these make up over 67 percent of new financial inflows linked to publicly issued CLNs. However, the findings show no evidence of price impact on futures prices for CLNs based on diversified commodity indices’where average proceeds were smaller and, thus, smaller hedging transactions spread across many different commodities’as well as for CLN issues with proceeds of less than $2 million. Theoretically, then, CLNs linked to diversified commodity Brennan Baker (brennan@maryknollogc.org) Maryknoll Office for Global Concerns
indices could still have significant price impact if the issue proceeds are great enough. In addition to the price impacts noted above, the analysis shows that as proceeds of CLN issues increase so does the magnitude of the price impact on the referenced commodity futures price, and, most importantly, that these price impacts are permanent, showing no signs of reversal. The report concludes, “Commodity prices are no longer determined only by the supply and demand in the real sector. Instead, they are also affected by the financial institutions’ behavior, which reflects a fundamental process of financialization of commodities.”
The approaches of previous studies, both those confirming the financialization hypothesis and those refuting it, have been criticized for one reason or another. Previous literature has taken several general forms: One branch, which “focuses on the relation between net positions of non-commercial traders and futures price changes,” has been criticized for using Commodity Index Trader (CIT) supplement to the Commitments of Traders (COT) weekly report, because it aggregates traders’ positions by category (The non-public daily index investment data used by Buyuksahin and Robe (2011) refuting financialization suffers the same limitations). Another branch analyzes the increasing return correlations between commodity futures prices and other financial assets or the co-movement among different commodities futures prices. A third branch focuses on existing long-only commodity index investments during the “roll activity.”
In contrast to those just mentioned, many of the studies that fail to find evidence of the price impact of financial traders rely on vector autoregressions that measure lagged changes in financial investors’ positions (e.g., Irwin and Sanders 2010, among others). And while so-called Granger causality tests’another statistical tool used in this approach’suffer many fundamental limitations2 in relation to the particularities of the project, this paper argues, in particular, that “commodity futures prices rapidly incorporate information