London, 10 January 2010

In an updated paper on "Fundamentals, Trader Activity and Derivative Pricing" published in November, Commodity Futures Trading Commission (CFTC) Chief Economist Jeffrey Harris and his coauthors insist gyrations in crude oil prices can be explained by fundamental factors rather than speculation and investment flows.

Research by Harris and other staff economists at the CFTC linking the rise in oil prices to physical supply and demand, rather than the influence of hedge funds and pension funds, has formed the basis for official reports issued by the Commission, as well as the UK Financial Services Authority (FSA) and IOSCO.

In some form or another, all claim there is no evidence for a speculative effect and downplay the need for position limits or other measures to curb the influence of new financial investors on raw materials pricing.

The CFTC's position has shifted somewhat recently. Chairman Gary Gensler admitted in a pre-confirmation letter last year that the influx of money into commodity indices had probably contributed to the run up in prices and increased volatility ("I believe that rapid growth in commodity index funds was a contributing factor to a bubble in commodities prices that peaked in mid-2008").

As a result, the Commission appears set to propose position limits on crude oil and other energy markets at a public meeting this week. The shift has created some tension between the Commission's political appointees, led by Gensler and Commissioner Bart Chilton, who insist position limits are necessary to avoid excessive concentration and potential price impacts, and the chief economist's department, which has never formally withdrawn its findings that investment flows have no effect.

It will be interesting to see whether the staff line changes when the Commission appoints a new chief economist to replace Harris, who is returning to teach at the University of Delaware. The commissioners themselves are split. While Gensler might favour a new economist more in tune with his own views, other members of the Commission seem more committed to the "no evidence" thesis.

In contrast, the FSA shows no such irresolution. Rejecting calls to impose position limits last month, the Authority continued to insist "the majority of academic studies do not support the proposition that prices have been systematically driven by this increased inflow of financial interest ...

We agree with these conclusions."

CONFIDENTIAL DATA

In fact, the body of academic research is less impressive and decisive than the FSA and other regulators believe. Studies of investors' behaviour and impact on commodity markets are invariably based on U.S. futures and options markets, since the CFTC's weekly commitments of traders report provides the only attempt to classify positions as "commercial" (hedge-related) and "noncommercial" (speculative).

But as papers published by CFTC economists themselves acknowledge, findings based on these highly aggregated data "should be interpreted with caution" since "researchers utilizing broadly aggregated data have been forced to make assumptions about the composition of each classification."

The CFTC has much more detailed breakdowns of positions by maturity and trader type. Its Large-Trader Reporting System (LTRS) from which the published data is compiled classifies traders into no fewer 30 categories (including dealer/ merchant, manufacturer, producer, swap dealer, pool operator, floor trader, etc).

CFTC staff use this for monitoring and enforcement (though one might reasonably ask how they missed the build up of large positions by Amaranth in the natural gas market). But the data remain confidential. Segments have been made available to only a few researchers.

Without access to the detailed data, the vast majority of academic studies have little value. But the few studies which have benefited from access are impossible to verify or critically review because the data on which they are based remain confidential. The evidentiary basis for the "no impact" argument is therefore much thinner than many people suppose.

WHO IS A SPECULATOR?

One of the papers which benefited from access to the confidential data, published by the chief economist's office, found there was no evidence of a link between price changes and the positions of "money market traders" (MMTs) in the natural gas market, and a negative correlation in crude oil (in other words money market traders reduced positions when prices were rising and vice-versa).

It also found MMTs changed their positions less often than other types of traders, including merchants and manufacturers. There was a negative correlation between MMT position changes and other categories of participants. MMT traders responded to position changes initiated by other participants rather than vice-versa. In other words, MMTs were providing liquidity to the rest of the market rather than absorbing it.

The paper identified MMTs with "speculation," but its definition was very narrow. It did include the LTRS categories for commodity pool operators (CPO), commodity trading advisers (CTA), associated persons (AP), and managed money (MM). But it did not include more strategic investors (pension funds, mutual funds) or intermediaries (such as merchants and swap dealers) who may take substantial positions on their own account.

Many large players in commodity markets run mixed physical- and-financial books and are classified as commercial "hedgers" in the merchant, swap dealer and other categories (though it is impossible to know for certain because the CFTC does not even release details of who is classified as what).

More importantly, much of the money that has flowed into commodity markets has come via indices (usually operated by swap dealers) and institutions such as pension funds (who are outside the narrow MMT category).

Finally, the period covered in a detailed paper on price formation published by the CFTC covers the period from August 2003 to August 2004, so predates most of the investment inflow and the concurrent rise in commodity prices.

IMPROBABLE FINDINGS

The published research does not really provide a sound basis for reaching a conclusion one way or another about whether investment flows impact on commodity prices. In fact, the econometrics throws up some very counterintuitive findings.

While hedge funds and other money market traders do not apparently have any impact on the absolute level of prices, or volatility, they do affect the market in benign ways, adding liquidity both nearby and further along the curve, as well as helping to improve the integration of formerly segmented markets at different maturities (nearby and further forward).

Crucially, MMTs appear to be entirely passive, reacting dumbly to changes price and position changes initiated by underlying demand from physical producers and consumers. If they really were this passive, it is hard to see how they earn their fees or returns. While this may describe the trading behaviour of technical funds it is not an accurate description of the big strategic macro funds and index investors.

Moreover, the commodity markets, unlike housing, equities or fixed income, would appear to be the only ones in the world where prices are determined entirely by a series of "exogenous" shocks appearing from nowhere, and where a sudden inrush of funds does not disturb prices for an extended period while the market searches for a new equilibrium.

In a world where policymakers, including Fed Chairman Ben Bernanke, now accept bubbles can and do occur, the characterisation of investors as passive providers of liquidity with no impact on outright price is implausible. More recent research confirms that commodity indexation has indeed changed the way raw materials prices behave.

Rather than endlessly echoing the "no evidence" thesis, the challenge now for academics and policymakers is to develop a more rounded and realistic view of commodity price formation. One that incorporates both fundamental and forward-looking/speculative influences and considers what, if anything, should be done to prevent damaging herd behaviour and bubbles forming.

Ends --