How Financialization Changed Commodity Prices and Markets
Book: Commodities: Markets, Performance, and Strategies
Editors: H. Kent Baker, Greg Filbeck, Jeffrey H. Harris
Publisher: Oxford University Press, 2018
ISBN: 9780190656010
The Evidence: Did Financialization Actually Change Anything?
This is Part 2 of Chapter 25 by Kyle Putnam and Ramesh Adhikari. Part 1 covered the who and how of financialization. This part digs into the empirical evidence. Did all that financial money actually distort commodity prices, returns, and volatility? The short answer: it depends on who you ask and how they measured it.
Haase, Zimmerman, and Zimmerman (2016) reviewed 100 empirical research papers on the effects of financialization. They sorted the studies by what was being measured: prices, returns, risk premiums, spreads, volatilities, and spillover effects. The scorecard is messy. Evidence points to financialization having a positive impact on prices, a negative impact on returns and risk premiums, and mixed results for everything else.
Here is an interesting detail. When researchers used indirect or proxy measures for speculation, most found positive, reinforcing effects from financialization. When they measured speculation directly, many found negative or weakening effects. The measurement choice seems to influence the conclusion.
What the Data Shows
Looking at summary statistics across five commodity sectors (foods and fibers, grains and oilseeds, livestock, energy, and precious metals), the financialization era (2001-2016) showed higher average returns for all sectors except livestock compared to the pre-financialization period (1986-2000). Standard deviations were also higher across the board. Risk-adjusted measures generally improved.
Average commodity prices rose to four times their pre-2000 levels by 2011 before declining. Singleton (2013) found that investor flows had a positive and significant effect on crude oil futures prices, especially from money-managed accounts around the 2008 boom and bust. Henderson, Pearson, and Wang (2014) found that financial institutions hedging commodity-linked notes had a measurably positive impact on futures prices around pricing dates and a negative impact when the trades were unwound.
Volatility also spiked. A clear volatility jump happened around 2007, persisting through 2009 before falling back. Silvennoinen and Thorp (2010) found that an increase in noncommercial short open interest increases futures price volatility. Adams and Gluck (2015) attributed most volatility spillovers to a “style effect,” where index traders buy and sell commodities and equities together. The spillovers were one-way: stocks affected commodities, but not the reverse.
Tang and Xiong (2012) found that non-energy commodities included in major indexes (S&P GSCI and DJ-UBSCI) showed much greater volatility increases than non-indexed commodities. Oil price volatility was spilling over to non-energy commodities through index investment, particularly between 2007 and 2008.
The Correlation Problem
One of the clearest effects of financialization has been the rise in correlations. Average commodity-equity correlations went from roughly zero before 2000 to about 0.50 by the late 2000s. Commodity-to-commodity correlations also surged, especially between oil and non-energy commodities.
Tang and Xiong (2012) found that non-energy commodity futures became much more correlated with oil after 2004, and the increase was more pronounced for indexed commodities. Since oil carries the biggest weight in most commodity indexes, and since there is no fundamental economic reason for corn or cotton to move in lockstep with crude oil, they argued that the correlation increase reflects the financialization process.
This matters because rising equity-commodity correlations undermine one of the main reasons investors bought commodities in the first place: diversification. If commodities and stocks move together, the portfolio benefit shrinks.
But Adhikari, Putnam, and Maroney (2016) found that diversification benefits still persist, especially when adding commodity futures to a U.S.-only portfolio of stocks and bonds. The benefit is smaller when the reference portfolio is already globally diversified. So commodities still help, just not as much as they used to.
The Other Side: Evidence Against Financialization Effects
Not everyone agrees that financial investors changed the fundamental character of commodity markets. A significant body of research pushes back.
Irwin, Sanders, and Merrin (2009) argued that little systematic evidence links financialization to changes in risk and return properties. They noted that attacks on speculation are a recurring pattern during periods of extreme market turbulence, and the arguments favoring a direct link are “conceptually flawed.”
Sanders and Irwin (2011a) used CFTC data on swap dealers and found little evidence that their net positions predict weekly returns for agricultural and energy commodities. Sanders and Irwin (2011b) found the same for CIT position changes in corn and wheat. Irwin and Sanders (2012) used quarterly index investment data across 19 commodities and found little evidence that changes in index positions affect returns or volatility.
Buyuksahin and Harris (2011) ran causality tests and found the opposite of what financialization critics claimed. Price changes caused position changes, not the other way around. In other words, speculators were following prices, not pushing them.
Irwin et al. (2009) pointed out a logical inconsistency in the theory: if long-only index traders were responsible for price increases, the increases should have been uniform across all commodity markets. They were not. Cross-sectional tests showed significant positive correlations in only one of 12 models.
Bhardwaj, Gorton, and Rouwenhorst (2016) took a longer view, benchmarking recent experience against 45 years of history. They concluded that variation in business cycle conditions, not financialization, explains the changes observed since 2005. The rise in equity-commodity correlations was merely a business cycle artifact that peaks during severe economic turmoil and has since returned to long-run averages.
So Who Is Right?
The chapter’s conclusion is honest: the debate is far from settled. Mixed results come from different focus variables, sample periods, and statistical methods. But there is one encouraging sign. Visual inspection of prices, volatility, and correlation data shows a behavior change beginning around 2012, with metrics starting to return to long-term averages despite the continued presence of index investors. If financialization were permanently distorting markets, you would expect those effects to persist.
Regardless of the debate, the chapter notes that commodity futures still appear to be a valuable diversification tool and can provide substantial returns. The practical implication is that investors should not avoid commodities just because the financialization debate is unresolved.
My Take
This is one of those debates where both sides have legitimate points. The correlation evidence is hard to dismiss. Something clearly changed when $300 billion of index money showed up in a market that was not used to it. But the causality evidence is weak. Price changes leading to position changes, rather than the reverse, is a pretty significant finding for the anti-financialization camp.
What I find most interesting is the 2012 reversal. Correlations and volatility metrics started normalizing even though financial investors stayed in the market. That suggests the initial shock of financialization may have been temporary, and markets adapted. If that interpretation is correct, then the regulatory response (Dodd-Frank, position limits) may have been an overreaction to what was essentially a transition period.
Previous: Financialization of Commodity Markets: When Wall Street Discovered Raw Materials