THE ROLES OF HEDGERS AND SPECULATORS IN THE NATURAL GAS AND CRUDE OIL MARKETS

Prof. Ronald D. RippleDirector, CREMECurtin University

30th USAEE/IAEE North American ConferenceWashington, D.C11 October 2011

• Motivation• Market and data• What is excessive trading• Open interest versus trading volume• Index investment data• Shares of trader categories• Granger causality• Conclusions

• Crude oil futures market - NYMEX• Crude oil futures prices (NYMEX, CRB, EIA)• Crude oil futures trading volumes (NYMEX, CRB)• Crude oil futures open interest (NYMEX, CFTC)• Time periods

– Prices: January 2000 – March 2011– Trading volume: January 2000 – March 2011– Open interest:

• NYMEX : January 2000 – March 2011• CFTC-Legacy: January 2000 – September 2011• CFTC-Disaggregated: June 2008 – September 2011• Index Investment Data: Dec 2007 – July 2011 (periodic)

• CFTC-Legacy:– Commercial: long and short– Non-commercial: long, short, and spread– Non-reporting: long and short

• CFTC-Disaggregated:– Producers-Merchants: long and short– Swap dealers: long, short, and spread– Managed money: long, short, and spread– Other reporters: long, short, and spread– Non-reporters: long and short

Crude oil trading volume, NYMEX (million bbl)

2007 2008 2009 2010 2011* Average daily near month volume 240.6 275.9 276.2 329.0 341.7 Ave. daily near month volume: 30-day basis

8.0 9.2 9.2 11.0 11.4

Maximum daily near month volume 502.9 508.7 527.0 605.0 638.2

Max. daily near month vol.: 30-day basis

16.8 16.9 17.6 20.2 21.3

Real “multiple”: Near-month average volume 0.42 0.48 0.48 0.58 0.60 Near month maximum volume 0.88 0.89 0.92 1.06 1.12

Estimated daily physical use is assumed to be 19 MM bbl. * - 2011 average through 31 Mar. Source: NYMEX and CRB data. For a full explanation of the approach, see R.D. Ripple, “Futures trading: What is excessive?”, Oil & Gas Journal, Vol. 106.22, June 9, 2008, pp. 24-32.

Natural gas trading volume, NYMEX (Bcf)

2007 2008 2009 2010 2011* Average daily near month volume 564.5 750.4 954.9 1,121.8 1,304.8 Ave. daily near month volume: 30-day basis

18.8 25.0 31.8 37.4 43.9

Maximum daily near month volume 960.3 1,618.9 2,087.9 2,251.1 2,135.1

Max. daily near month vol.: 30-day basis

32.0 54.1 69.6 75.0 71.2

Real “multiple”: Near-month average volume 0.29 0.39 0.50 0.58 0.68 Near month maximum volume 0.50 0.84 1.09 1.17 1.11

Estimated daily physical use is assumed to be 64 Bcf. * - 2011 average through 31 Mar. Source: NYMEX and CRB data. For a full explanation of the approach, see R.D. Ripple, “Futures trading: What is excessive?”, Oil & Gas Journal, Vol. 106.22, June 9, 2008, pp. 24-32.

0

2

4

6

8

10

4/01/2010 4/02/2010 4/03/2010 4/04/2010 4/05/2010 4/06/2010 4/07/2010 4/08/2010 4/09/2010 4/10/2010 4/11/2010 4/12/2010

Total open interest divided by total trading volume: CLDaily observations 2010

Overall 2010 average 2.19

0

2

4

6

8

10

2/01/2008 2/02/2008 2/03/2008 2/04/2008 2/05/2008 2/06/2008 2/07/2008 2/08/2008 2/09/2008 2/10/2008 2/11/2008 2/12/2008

Total open interest divided by total trading volume: CLDaily observations 2008

11.86

Overall 2008 average 2.59

0

5

10

15

20

25

30

4/01/2010 4/02/2010 4/03/2010 4/04/2010 4/05/2010 4/06/2010 4/07/2010 4/08/2010 4/09/2010 4/10/2010 4/11/2010 4/12/2010

Total open interest divided by total trading volume: NGDaily observations 2010

Average 2007 = 3.45

0

5

10

15

20

25

30

2/01/2008 2/02/2008 2/03/2008 2/04/2008 2/05/2008 2/06/2008 2/07/2008 2/08/2008 2/09/2008 2/10/2008 2/11/2008 2/12/2008

Total open interest divided by total trading volume: NGDaily observations 2008

Average 2007 = 6.28

Index InvestmentData

In US dollars and futures equivalent contracts

Includes index funds, swap dealers, hedge funds, pension funds, and mutual funds.

Source: CFTC

Total notional value is $418.6 billion.

U.S. Futures Market1

(Notional Value > 0.5 billion US$)2 Long Short Net L (S) Long Short Net L (S)

Cocoa 1.6 (0.6) 1.0 53 -20 33Coffee 7.3 (2.7) 4.6 67 -25 42Copper 10.2 (3.2) 7.0 97 -30 66

Corn 24.1 (8.2) 15.9 629 -215 414Cotton 5.1 (1.7) 3.4 96 -31 65

Feeder Cattle 0.9 (0.2) 0.7 13 -3 10Gold 29.3 (7.5) 21.8 160 -41 119

Heating Oil 12.3 (3.4) 8.8 95 -27 68Lean Hogs 5.1 (1.6) 3.5 149 -46 103Live Cattle 9.1 (2.9) 6.3 198 -63 135Natural Gas 19.7 (5.3) 14.4 468 -129 338

Platinum 0.8 (0.1) 0.7 9 -1 7RBOB Unleaded Gas 12.3 (3.0) 9.3 103 -24 78

Silver 8.9 (1.9) 7.0 43 -9 34Soybean Meal 0.9 (0.1) 0.8 23 -2 22Soybean Oil 4.4 (1.2) 3.2 124 -34 90Soybeans 19.3 (5.6) 13.7 264 -76 188

Sugar 11.3 (3.5) 7.8 351 -106 245Wheat (CBOT) 13.2 (5.4) 7.8 333 -137 196Wheat (KCBT) 1.9 (0.5) 1.4 42 -12 30WTI Crude Oil 58.8 (17.1) 41.7 657 -192 465

Subtotal (>0.5 billion US$) 256.3 (75.7) 180.6Subtotal (<0.5 billion US$) 1.0 (0.3) 0.7

Total Notional US Mkts 257.3 (76.0) 181.3Total Not'l Non-US Mkts 65.4 (19.9) 45.5

Total All Markets 322.7 (95.9) 226.8

August 31, 2011Notional Value

(Billions US$)Futures Equivalent Contracts3

(Thousands)

Index Investment shares: Crude oil Percentage of Index Investment

Commodities* Percentage of combined futures positions**

June 30, 2008 25.5 18.4 June 30, 2009 25.9 23.8 June 30, 2010 22.0 24.8 June 30, 2011 20.9 23.9 Sources: CFTC (Index Investment Data and COT) and ICE *- Percentage of WTI Crude Oil net long with respect to Total All Markets net long notional value. **- The combined futures open interest positions include the NYMEX WTI futures contract, WTI crude oil financial, and the ICE WTI Crude Futures.

-0.25

-0.2

-0.15

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

13/0

6/20

06

13/0

8/20

06

13/1

0/20

06

13/1

2/20

06

13/0

2/20

07

13/0

4/20

07

13/0

6/20

07

13/0

8/20

07

13/1

0/20

07

13/1

2/20

07

13/0

2/20

08

13/0

4/20

08

13/0

6/20

08

13/0

8/20

08

13/1

0/20

08

13/1

2/20

08

13/0

2/20

09

13/0

4/20

09

13/0

6/20

09

13/0

8/20

09

13/1

0/20

09

13/1

2/20

09

13/0

2/20

10

13/0

4/20

10

13/0

6/20

10

13/0

8/20

10

13/1

0/20

10

13/1

2/20

10

13/0

2/20

11

13/0

4/20

11

13/0

6/20

11

13/0

8/20

11

Crude oil: net-long shares of open interest by trader category

PMN/OI

SWN/OI

MMN/OI

ORN/OI

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

Crude oil: TotSpread/OI

-0.3

-0.2

-0.1

0

0.1

0.2

0.3

0.4

13/0

6/20

06

13/0

8/20

06

13/1

0/20

06

13/1

2/20

06

13/0

2/20

07

13/0

4/20

07

13/0

6/20

07

13/0

8/20

07

13/1

0/20

07

13/1

2/20

07

13/0

2/20

08

13/0

4/20

08

13/0

6/20

08

13/0

8/20

08

13/1

0/20

08

13/1

2/20

08

13/0

2/20

09

13/0

4/20

09

13/0

6/20

09

13/0

8/20

09

13/1

0/20

09

13/1

2/20

09

13/0

2/20

10

13/0

4/20

10

13/0

6/20

10

13/0

8/20

10

13/1

0/20

10

13/1

2/20

10

13/0

2/20

11

13/0

4/20

11

13/0

6/20

11

13/0

8/20

11

Natural gas: net-long shares of open interest by trader category

PMN/OI

SWN/OI

MMN/OI

ORN/OI

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

Natural gas: TotSpread/OI

tttt SPSPSP 23121

ttitittt NOINOISPSPSP 2,51,423121

Granger causality tests

Following the approach employed by Stoll and Whaley (2009), I first estimate the relations between the change in futures prices and its own lagged values. Then I employ a variable-addition test of the lagged values of changes in net open interest positions; one trader category at a time. [Microfit econometric software is employed.]

The estimation is then reversed whereby the change in net open interest positions is estimated against its own lagged values, and then I perform the variable-addition test of lagged values for changes in the futures price.

I also test for Granger causality between categories of traders (commercial versus non-commercial); testing for the influence of lagged values in both directions.

[None of the variables exhibit unit roots.]

For example: (SP is settlement price; NOI is net open interest; i designates commercial or non-commercial)

(1)

(2)

Grange causality test results

Crude oilThe change in non-commercial net OI is Granger caused by the change in commercial net OI, but it is not Granger caused by the change in price.

The change in commercial net OI is caused by the change in price, but not by the change in non-commercial net OI.

The change in crude oil futures settlement price is not Granger caused by changes in either commercial or non-commercial net OI.

Natural gasThe change in non-commercial net OI is Granger caused by the change in commercial net OI, but it is not Granger caused by the change in price.

The change in commercial net OI is Granger caused by the change in non-commercial net OI (but only at the 7% level), but not by the change in price.

The change in natural gas settlement price is not Granger caused by changes in either commercial or non-commercial net OI.

Conclusions• No evidence of excessive trading• No evidence of changes in net futures

positions influencing futures price changes• Some evidence that non-commercial trader

activity is influenced by commercial trader activity, but not the reverse (but perhaps weakly for natural gas)

H.R. Stoll and R.E. Whaley (2009) “Commodity index investing and commodity futures prices,” http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1478195 orhttp://openmarkets.cmegroup.com/wp-content/uploads/2010/02/Stoll-Whaley-Report.pdf

THE ROLES OF HEDGERS AND SPECULATORS IN THE NATURAL GAS AND CRUDE OIL MARKETS

Thank you!

Questions and/or comments?

Prof. Ronald D. RippleCurtin University