The Politics and Economics of International Energy (Spring 2009- E657)

Lecture 4

The functioning of the international oil market

Prof. Giacomo Luciani

Outline

 Price formation: past and current regimes  Prices, Supply and Demand  The Market for Brent  Details on the Futures Market  The Crude Oil Market: a Broader Picture  Causes of volatility  Fundamentals or speculation?

 Better benchmarks?

 Conclusion

Price Formation: Past and Current Regimes

Chart of crude oil prices since 1981 © BP 2008

Successive Price Regimes

 Before 1880: “Disorder” in the US  1880-1910: the Standard Oil Trust Regime  1910-1930: Transition  1930-1970: the 7 Sisters Pricing Regime:  Posted prices controlled by the companies  1970-1985: the OPEC Pricing Regime:  Posted prices controlled by the producing countries  1985-87: the Netback Pricing Regime:  Transition period  1987 to date: the Reference Pricing Regime

1970-85: The OPEC Price Regime

 In 1970-85, prices were unilaterally determined by producers.

 In the occasion of political crises, the market (illiquid and opaque) pushed prices higher, and producers consolidated the increases.

 Saudi Arabia dissented from all others: the quest for a long-run equilibrium price.

Development of the Market in the 1980s

 The Brent and WTI markets developed in the early 1980s  Non-OPEC supply responded vigorously to higher prices  The OPEC share was eroded, and Saudi production declined to little more than 3 million b/d.

 In 1985, Saudi Arabia begins a price war.

1985-87: The Netback Pricing Regime

 Netback pricing means that the price of crude is tied to the price of refined products.

 Refined product prices are determined by product markets (e.g. Rotterdam).

 The system lacks transparency and encourages products oversupply.

 Weak prices – soon abandoned

After 1987: The Reference Pricing Regime

 “Reference pricing” means that the price of a crude which is not freely traded is tied by some formula to the price of another crude which is freely traded.

 The two main reference crudes are Brent and WTI (West Texas Intermediate)

Prices, Supply & Demand

The roots cause of oil price instability

   Why are oil prices so unstable? (At times, because at other times they were “kept” very stable) The key reason is that both supply AND demand are rigid to price changes in the short term.

(Demand appears to be more responsive to income than to price shifts)

Non-OPEC Supply

 Non-OPEC supply is competitive and price responsive.

 However:  Production requires huge up front investment   Direct costs are a small component of total costs As long as direct costs are paid, producers are best off maximizing production  In actual practice, non-OPEC production cannot satisfy global demand and producers always maximize production.  OPEC must provide the difference – this is the “call on OPEC”.

OPEC Supply

 OPEC supply is limited by quotas, first established in the 1980s.

 OPEC supply does not respond to prices (except for “cheating”, which occurs anyway but is more tolerated if prices are high).

 In fact, OPEC quotas are (unpredictably) modified to pursue a revenue or price objective.

 If prices are high, production may be reduced because revenue targets are met anyhow – and vice versa.

Global Demand

 Demand too is price inelastic:  Oil satisfies essential needs  Higher per capita income  Concentration in transportation  Taxes isolate consumers from the impact of price changes.

 Demand is influenced primarily by macro trends (income effect) and the weather.

Demand and Supply are Rigid in the Short Term

A’ A

Non-OPEC OPEC

P P’ Price

However, in the longer run…

 Demand and (less so) supply are price elastic.

 Consumers will make the investment required to reduce fuel consumption or switch to a different fuel.

 On the supply side, companies’ cash flow improves, investment increases, new fields are developed.

 However, higher prices may also encourage resource nationalism and slow down investment/production

The Market for Brent

The Market for Brent

 Brent is a field in the UK North Sea.

 The market consists of:  A “spot” market;  A “physical forward” market;  A “futures” market – based at the International Petroleum Exchange (IPE) in London

The spot market, or “dated Brent”

 “Dated Brent” refers to the sale of a specific cargo that is either available in a specific loading slot or that is already loaded and in transit to some destination.

 Main characteristics:  Transactions are bilateral,  Over the counter (OTC),  For variable quantities.

The forward market, or “15-day Brent”

 The 15-day cargo is a standard parcel (500,000 barrels) that will be made available by the seller to the buyer on an unspecified day of the relevant month.

 The clearing involves book-outs or seller’s nominations, which can take place on any day in the period starting fifteen days before the beginning of the relevant month and closing eighteen days before its end.

 As dated Brent, 15-day Brent is bilateral and OTC – but it is standardized.

The futures market

 The futures market was launched by the International Petroleum Exchange (IPE – today International Commodity Exchange - ICE) in 1988  1 contract = 1000 barrels  Contract based on cash settlement and not on physical delivery  If a contract is allowed to expire the settlement price is the Brent index  Central exchange and clearing house rather than bilateral trades  Several months traded

Details on the Futures Market

Options

 Launched by the IPE in 1989  A call option gives the holder the right to buy the underlying futures contract, and a put option the right to sell.

 Call and put options may be combined to design complex risk management strategies.

What are options for?

 Any buyer or seller on the petroleum market faces a price risk.

 Options and futures allow parties facing a structural risk to limit that risk, “selling” it to speculators (or “insurers”).

Hedging

 A producer can sell futures or buy put options to ensure a minimum level of prices.

 A large consumer can buy futures or a call option to ensure against very high prices  Major companies are on both sides of the market and may be doing both things at the same time.

Complex Strategies

 A producer can at the same time buy a put option and sell a call option: in the first step he acquires a guarantee of future revenue, in the second he gives up the extra profit from potential price peaks.  The purchase of a put is financed by the sale of a call. The combination of a put and a call is called a “collar”.

Why so many Paper Barrels?

 Most participants in the futures market are there to manage their risk, not to acquire Brent crude.

 Buying and selling Brent futures and options is an effective strategy because other crude prices follow Brent movements.

The Structure of the Oil Market

 At the center, we find the market for Brent and WTI, which influence each other  Brent and WTI trade few physical and lots of paper barrels  Paper and wet barrels influence each other, but paper barrels are more important  Smaller markets, such as ANS and Dubai, are influenced by Brent and WTI

Other Crudes

 All other major crudes are priced on the basis of formulas which tie them to Brent or WTI  The producing countries oppose the free trading of their crudes, and restrict destination and secondary trading  Formulas are modified from time to time, but the essence remains

Example: Saudi Arabia’s Prices for 04/09

How is the Market Cleared?

 Brent/WTI are not the marginal crudes that balance demand and supply.

 Yet, it is Brent/WTI that make the price.

 The implication is that demand and supply are not necessarily balanced: OPEC and other operators manage supply and/or stocks, given the price.

The Feedback Issue

 Given this market structure, operators that have no interest in Brent crude trade in Brent futures and options, to manage their risk  In this way, a certain feedback is obtained between the global physical oil market and Brent  However, such feedback is limited and partial  The feedback would be greater if all operators hedged systematically; in that case, an excess of demand would raise the price of calls, financial intermediaries would lower the price of puts, producers would be incentivated…

OPEC and Brent/WTI prices

 OPEC has no direct control on Brent/WTI prices. It attempts to exert an influence through signaling and decisions on quotas.

 But quotas are never fully implemented, and in any case are but one of the elements at play: Brent’s price response to quota shifts is unpredictable.

Other influences

 Refined products markets also have a feedback on Brent and WTI, but the prevailing influence is in the opposite direction.

 Except in the US and UK, the price of natural gas is also indexed to crude or products prices  Only the price of coal is truly independent of Brent and WTI

Rotterdam product prices and Gulf Coast product prices © BP 2008

The Oil Market – A broader Picture

Why oil producing countries like reference pricing?

 Accepting pricing out of a marker implies that producers are giving up on an important role  They should naturally be price makers, instead are price takers  Why? First and foremost because in the past they failed in the management of posted prices.

Why there is no Arabian Light market?

 Setting up a market for a major crude, such as e.g. Arabian Light, is not easy because there is just one seller  The seller does not want the responsibility of making prices, because he is afraid of international political pressure or domestic dissension.

Oil Prices - a increasingly complex market constellation Forward Markets Physical Markets Commercial Operators & Independent Traders Over the Counter (OTC) Derivatives Markets Futures Markets Financial Markets Oil Products Markets

DTS/T, Saturday, 25 April 2020

Non-traded Crude oils

40 Modified by GL

Natural Gas Markets

Causes of Volatility

How can we explain this?

ICE Brent Crude Oil Front Month

Main questions about prices

 Why prices are where they are?

 Political instability  “Paradigm shift”   Speculation – flow of funds Unreliable benchmarks -volatility

Political instability

    Routine explanation It is normally quite easy to find a political disturbance to which higher prices may be attributed: Iraq, Venezuela, Nigeria, today Turkey/Kurdistan… But political instability is the norm, not the exception Therefore, this interpretation is pretty useless ex ante

Paradigm shift

   Prices are high and increasing because oil supply is tight Beginning in 2004, markets signaled a paradigm shift by way of a protracted anomalous contango Markets went back to backwardation in the summer of 2007

The Forward Curve

 At any point in time, we have several prices for the same marker for different maturity future contracts.

 If prices for subsequent months are lower than the closest-month price the market is said to be in backwardation  If prices for subsequent months are higher than the closest-month price the market is said to be in contango

Meaning of contango

   A contango occurs when the market expects future prices to be higher than today’s Normally, a contango occurs when prompt prices are low, backwardation when they are high Backwardation is the “normal state” of a market because holding stocks has a physical and financial cost (interest rate)

What is the impact of backwardation/contango?

  Contango encourages the buildup of physical stocks (you earn money by buying spot and selling futures, while holding the commodity) Backwardation encourages financial commodity investors (you make money by buying futures and waiting for futures to converge to spot prices)

Aissaoui on backwardation



The risks to oil market stability nowadays are likely to be higher in backwardation than in contango. Even assuming a constant appreciation of spot prices (spot yield), the negative roll yields generated by contango tend to inhibit commodity investors. By contrast, the positive roll yields achieved in a backwardated market offer greater prospect of realizing higher returns. In such a case, a surge in commodity investments will only lead to more speculative activities and increase, as a result, the likelihood of renewed oil price spikes and greater volatility. Obviously, OPEC has no interest in such economically damaging oil price behaviour.

Speculation – flow of funds

   Investors move from one asset class to another depending on the conditions of each market As they shift from one asset class to another, they drive prices on one market down and on another market up This is the “flow of funds” effect – it will last as long as funds are transferred between markets

The International Investment Playing Field: Oil is a small piece The Global Trillion Dollar Trading System Equities Bonds Currencies Real Estate Commodities Other Interest Rate Instruments Metals Agricultural Fuels Precious Metals Heating Oil Gasoline Crude Oil Natural Gas Electricity

Source: ESAI HESS ENERGY TRADING COMPANY, LLC January 28-29, 2002

Support for flow of funds explanation

     Mortgage crisis in US Weak dollar Unstable equities Prices have increased for all commodities, not just oil Oil market is backwardated

Volatility

  Volatility is a challenge for all:  Producing countries face unstable revenues   Companies face uncertainty in investment decisions Consumers pay higher prices than would be necessary to stimulate supply Volatility is caused by:  Price rigidity of oil supply and demand in short term   Prevailing influence of futures vs. physical trading Benchmarks shortcomings

Instability and resonance

   The oil market is structurally unstable because of the rigidity of demand and supply Hedging and financial investors thrive on this underlying instability. Through their action, they tend to increase instability This may be viewed as a kind of resonance

Momentum (finance)

  

From Wikipedia, the free encyclopedia

In finance, momentum is the empirically observed tendency for rising asset prices to raise further. For instance, it was shown that stocks with strong past performance continue to outperform stocks with poor past performance in the next period with an average excess return of about 1% per month (Jegadeesh and Titman, 1993, 1999).

The existence of momentum is a according to market anomaly , which economic theory has been struggling to explain. The difficulty is that an increase in asset prices, in and of itself, should not warrant further increase. Such increase, to efficient market theory , is warranted only by changes in demand and supply or new information (cf. cognitive biases fundamental analysis , which belong in the realm of ). Students of financial economics have largely attributed the appearance of momentum behavioral economics . The explanation is that investors are irrational (Daniel, Hirschleifer, and Subrahmanyam, 1998 and Barberis, Shleifer, and Vishny, 1998), in that they underreact to new information by failing to incorporate news in their transaction prices. However, much as in the case of rational traders (Crombez, 2001).

price bubbles , recent research has argued that momentum can be observed even with perfectly

Trading with moving averages

60 58 56 54 68 66 64 62 52 50 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61 WTI prices, January to March 2007: 3 and 6-day moving averages

Where is the damper?

  Resonance is contained if there is a damper; if there is no damper, oscillations may widen indefinitely, until the system collapses.

Where is the damper in the oil market? Where is the limit to oscillations? What is the “tipping point”? Unless we have an answer, there will be no reversal of expectations

Attempts to launch new benchmarks

   In 2007, two parallel attempts started to launch a new ME sour futures contract:  NYMEX/DME Oman futures  ICE ME sour futures Oman supports the DME and will price its oil out of that contract DME contract is physically deliverable

Fundamentals or Speculation?

Section III: Speculators were definitely active Our estimate of investment flows to commodities (in US$ bn) 160 120 80 40 0 Est'd cumulative funds flows into commodity indices (US$bn)

At these volumes, allocations matter immensely, when GSCI increase gasoline allocation by 3% this is 10-15% of all contracts

Source: BME, Goldman Sachs, DJ-AIG, Industry Documents, Nymex and Bloomberg

A confirmation from the futures exchange Outstanding futures contracts exploded with prices “Open Interest” WTI Crude 3,500 3,000 2,500 2,000 1,500 1,000 500 120 110 100 90 80 70 60 50 40 30 20 Futures and Options WTI price (RHS)

One contract is 1000 barrels. Over the past four years WTI alone has seen 3bn barrels of new futures contracts (new long or short)

At times prices can overshoot physical fundamentals  The past 6 months are critical example of how intimately linked physical fundamentals and prices are $150 Brent $/b $130 35 28 $110 21 $90 $70 $50 $30 Jan-07 Apr-07 Jul-07 Oct-07 Brent 1-6 differential Jan-08 Apr-08 Brent spot Jul-08 0 -7 14 7 Source: Bloomberg, UBS

Part V: The second fundamental driver – costs Both oil companies… 82 72 62 52 42 32 22 12 2 1994 1997 2000 2003 2006 2009E 2012E 3 2 5 4 7 6 9 8 Oil price Operating costs 18.00% 16.00% 14.00% 12.00% 10.00% 8.00% 6.00% 4.00% 2.00% 0.00% 0.00

Higher prices for the same rates of return 1999 1998 20.00

2000 2002 2004 40.00

2005 60.00

2006 2009E 2011E 2012E 2007E 80.00

100.00

Oil price (US$/b)

The price rise is rooted in fundamentals: Prices rose exponentially, as a result of lost OPEC output and rising demand, triggering: a demand response, capex for new supply, resource nationalism, and even higher prices In Soft Markets, Inventories Count; in Tight Markets, Capacity Counts Annual % Change in Real Oil Price 200% Spare Capacity Drives Price 150% 100% Forward Days Consumption in Inventories Drive Price

2003-2008

1973 / 1974 50% 1987 0% 1991 1986 1998 (50%) 50% 60% 70% 80% 90%

________________

Annual OPEC Capacity Utilisation 1972–2008

Source: U.S. DOE/EIA; LCMC Estimates.

100% 72

But flows played a major role in exacerbating the situation : Ccost inflation cannot explain price rise of 2007-08

PPI data appear to explain oil prices well until the divergence in Oct-2007. Recently, there has been a convergence, with September 2008 data suggesting $81 for deferred prices.

Average monthly 5-yr out WTI prices regressed on cost indicators

$/bbl 140 130 120 110 100 90 80 70 60 50 40 30 20 10 - R-squared through Oct-07 is 95% A recent convergence of prices to levels implied by cost indicators.

5-year out Dec WTI Fitted WTI w PPI costs Fitted WTI w PPI DW costs Fitted WTI w PPI DW DXY

Aug-08

$114.90

$80.92

$79.80

$88.21

73

HIGH OIL PRICE STIMULATES: BIOFUELS, UNCONVENTIONAL OIL, EOR AND NEW FRONTIERS

Even so, the marginal cost of supply was changing

Base marginal cost used to be $10 to 20, unconventional oil sands push to $90, now what?

Producing basins by marginal cost

130 120 110 100 90 80 70 60 50 40 30 20 10 0 0 Other OPEC 40 Other Conventional Oil Renewable Power Conventional Power Biofuels US (Corn Based) Oil Shale Oil Sands (mining) CtL GtL Deep Water Artic EOR FSU 80 Venezuelan Heavy Oil Oil Sands (In-Situ) 100 Biofuels ( Sugar Cane Based) 120 Conventional liquid sources Non-conventional liquid sources Emerging sources of transportation energy Million Barrels of Oil Equivalent per Day

Source: Booz Allen/IEA - Assumed average vs. marginal costs; 10% return for conventional and 13% return for unconventional technologies; no subsides for biofuels; no carbon offset costs; after severance and production taxes

74

Financial Activity: Trading Crude Oil Boomed Until this Year Daily trading volume on NYMEX crudes is over 6 times total global consumption; Trading volume on ICE crudes is over 5 times total global consumption.

Daily Trading Volume in WTI and Brent Crude Oil on NYMEX and ICE vs. World Oil Consumption m bbl/d 700 600 Trading Volume on NYMEX World Crude Oil Consumption Trading Volume on ICE 500 400 Add OTC trading and the daily trading volume may be >20 time global production 300 200 100 0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 75

– – – – Ultimately, Four Factors Explain Price Changes Long-term structures • Adequacy of capital asset base / spare capacity:

benign for next two-three years

Short-term market forces • Supply, demand, inventory, weather and level of economic growth:

two-three years bearish for next

Politics and geopolitics • OPEC’s ability to defend a price floor, risk of a disruption • Saudi Arabia’s ability to protect it’s preferred ceiling:

bearish for next two to three years

Flow of funds into/out of commodity investments • No doubt, fund flows grew in importance as market tightened • Deleveraging has curtailed the impact of fund flows, but what happens when deleveraging is over? • Will fund flows/higher corporate risk management, with lower open interest have a more pronounced impact?

• With re-emergence of exchange traded oil, products, natural gas, will transparency create more/less volatility? 76

Better Benchmarks?

Confidential

Benchmark peculiarities

Edward L. Morse Chief Energy Economist Lehman Brothers

June 2007

Benchmark peculiarities

Each benchmark diverges from global supply-demand dynamics in different ways, opening up interesting trading opportunities

 WTI landlocked; refinery maintenance and Canadian upstream disproportionately affect world’s most liquid benchmark  Brent waterborne, but limited volumes, field maintenance and quality control affect world’s second most liquid benchmark  Both WTI and Brent are light sweet benchmarks in a world where most crude is medium or heavy sour  Platts Dubai (and ICE Dubai) is based on easily influenced partial contracts, though it can function in concert with Platts Oman  DME Oman so far suffers from lack of liquidity and insufficient interest from Middle Eastern OPEC term sellers and Asian term refinery buyers 1

WTI-Brent: A broken relationship Dec 28, 2005 $64 $63 $62 $61 $60 $59 $58 $57 $56 $55 1M 5M WTI 9M 13M 17M May 24, 2007 21M Brent 25M $74 $72 $70 $68 $66 $64 $62 $60 $58 1M 5M 9M

________________ Source: Reuters

WTI 13M 17M 21M 25M Brent 29M 29M May 1, 2006 $78 $77 $76 $75 $74 $73 $72 $71 $70 $69 1M 5M 9M 13M 17M WTI August 31,2007 21M Brent 25M $75 $74 $73 $72 $71 $70 $69 $68 $67 $66 $65 1M 5M 9M WTI 13M 17M 21M 25M Brent 29M 29M 2

WTI local factors sometimes matter most

The arbitrage dynamic is failing with respect to WTI

 Chicago and Cushing used to be the arbiters of international and mid-continent crude oil – Main continental sources of crude are WTI/WTS and Canadian imports – Waterborne crudes come up Capline or Seaway Seaway

________________ Source: Canadian National Energy Board

3 Capline

What’s changed?

Canadian barrels have displaced international barrels in western PADD 2

 Light sweet waterborne crude prices will likely act as a ceiling to US and Canadian light sweet inland crudes (ignoring freight).  And Canadian flows into the mid-continent continue to increase — they are up 700k b/d over the past decade, and 140k b/d (12.5%) in 2006. Canadian oil production adding 150k b/d per annum with no where to go but PADD 2 m b/d 5.0

4.0

3.0

2.0

1.0

0.0

m b/d 1.3

1.2

1.1

1.0

0.9

0.8

0.7

0.6

Conventional (lhs) Tar sands (lhs) NGLs (lhs) Canada exports to PADD 2 (rhs)

________________ Includes exports through PADD 4 Source: EIA, Canadian Association of Petroleum Producers, Lehman Brothers Estimates

4

Midwest refineries plagued by unplanned outages

Tulsa, OK: power outage

•6-7 Feb, 16-22 Mar, 65k b/d for 9 days •6-7 May, 85k b/d for 2 days

Whiting, IN: fire

•22 Mar-End Aug, 90k b/d for 5+ mos.

•22 Mar-02 Jul, 90k b/d for 103 days •09 Jul-15 Jul, 260k b/d for 7 days

Toledo, OH: power outage

•25 Apr-18 May, 110k b/d for 24 days

Lima, OH: fire

•19 Jul-06 Aug: 170k b/d for 19 days

Coffeyville, KS: flood McKee, TX: fire Cushing

•01 Jul-Mid Sep: 100k b/d for 2+ •16 Feb-20 Jun, 60k b/d for mos.

Wynnewood, OK: fire

125 days •27 Apr-04 May, 35k b/d for 8 days •16 Feb-12 Apr, 105k b/d for

________________ Note: Unplanned outages for US mid-continent crude distillation units since 01 Jan 07 Source: Lehman Brothers estimates

•27 Apr-09 May, 17k b/d for 13 days 5

WTI seasonal weakness: likely to happen again

US turnarounds, particularly in PADD 2, make WTI more susceptible to weakness during turnaround seasons than Brent or Dubai

  The 2007 US refinery maintenance season begins with significantly more refineries down compared to fall 2006. – As of September 1, 2007 there are nearly five times as many planned outages as in 2006 Current record levels of capacity offline combined with expectations for strong mid-continent refinery maintenance may result in significant inventory builds at Cushing, OK.  making unplanned outages a significant threat.

US refinery turnarounds, 2006-07 kb/d 1,600 1,400 1,200 1,000 800 600 400 200 0 High margin environment and tightening product specs have strained refineries and caused systemic restart delays, Unplanned refinery outages, 2004-07 kb/d 500 450 400 350 300 250 200 150 100 50 0 Fall' 04 Fall '06 Fall '07 allow ance 2006 total outages 2007F total outages 6

Brent, meanwhile, suffers from North Sea field declines

North Sea production is no longer traded internationally in large volume. African grades tied to Brent better reflect global markets.

 North Sea crude production has been declining at over 5% per year or at least 200k b/d annually   Geopolitical risks of disruption in Nigeria and Middle East affect waterborne crudes more than WTI – OPEC cuts have also put strain on Asian markets, which forces African grades to move east instead of west – East-West competition for African grades (due to faster Asian demand growth and OPEC cuts) puts more upward pressure on Brent than WTI, since African crudes trade at a differential to Brent Partially offsetting relatively greater Brent tightness is the continued ramp-up of FSU output – A reversal of OPEC cuts could have a similar effect North Sea total oil production and exports to US, 1999-2006 k b/d 6,500 6,000 5,500 5,000 4,500 4,000 3,500 3,000 1995 1996 1997 1998 1999 UK exports to US (rhs) 2000

________________

UK + Norw ay Production (lhs)

Source: UK Dept of Trade and Industry, Norwegian Petroleum Directorate

2001 k b/d 800 700 600 500 2002 2003 2004 2005 Norw ay exports to US (rhs) 0 2006 400 300 200 100 7

Brent versus other benchmarks (cont.)

Seasonality in Brent prices linked to both upstream maintenance and light product demand

   Any major field maintenance can change the sulfur quality of BFOE – – Platts instituted a sulfur de-escalator but it only kicks in if sulfur levels move above 0.6% Refiners which are unsure about the quality of the crude oil they receive will tend to discount BFOE versus the rest of the market Field maintenance in the North Sea occurs at different times every summer, generally tightening light-sweet balances and increasing the Brent-Dubai spread Summer demand peak for gasoline also can contribute to a wider Brent-Dubai spread Brent-Dubai Spread: Summer peaks caused by field maintenance and demand seasonality 6 4 2 0 16 14 12 10 8 10/27/04, 13.61

8/31/05, 9.12

7/17/07, 7.22

8/8/06, 6.14

________________ Source: Bloomberg

8

Platts Dubai has its own peculiarities

Dubai is the least transparent and most easily influenced benchmark

   Platts Dubai is based on 90k b/d Dubai declining oil production. Means only four cargoes per month available – Two companies responsible for most traded cargoes, giving them asymmetric influence over pricing, mostly to the detriment of producers Dubai currently used with Platts Oman as underlying crude price for most major term contracts between Middle Eastern producers and eastern refiners – Dubai crude (30.4º gravity, 2.13% sulfur) approximates Saudi Light (32.7º, 1.8% sulfur) – Eastern refiner crack hedges also based on Dubai Refiners have incentive to sell Platts Dubai during the Platts window, depressing the price of the benchmark on which their term contracts are based – If a refiner is short, the two firms prominent in the trade can push Dubai prices up to higher quality Platts Oman prices (also deliverable against Platts Dubai) – Producers could also influence the price higher if they desired – If ICE Dubai became liquid, it would just create an additional incentive to influence the Platts Dubai price at ICE Dubai settlement 9

DME Oman has yet to take off

Until key Middle Eastern suppliers switch to DME, the exchange may not attract sufficient liquidity to be successful

  DME based on 800k b/d Oman production Oman crude (33.34º gravity, 1.04% sulfur) also approximates Saudi Light  For the contract to be successful, it needs aggressive buyers and sellers, as well as a strong market making program  Obstacles: – Not going to be significant producer hedging (Non-OPEC oil hedged with WTI and Brent, unlikely to change until DME already successful) – Refiners will only be aggressive sellers of DME if their term contracts are based on DME versus Platts Dubai/Oman – Saudi Arabia would likely have to assign some production to price off of the DME before the contract can become successful – Moving in the wrong direction: Qatar is de-linking its crude from the DME – Refiners still have huge remaining crack hedge positions on Platts Dubai 10

Potential revolutionary impact if DME Oman succeeds  Oil market could have a successful sour contract that could replace WTI/Brent as more representative  Asian buyers would find good reasons to use DME Oman for hedging long-haul purchases and for refinery margin hedging  Would require all Middle East producers to understand market mechanisms that their customers prefer, may induce some to manage their market exposure through the DME  Could eliminate “Asian Premium” that has characterized Middle East ‘delivered’ crude sales, reduce sour/sweet spreads  Would bring pricing back to Gulf where most of world’s reserves and incremental production are located 11

Benchmark relationships could radically change with refinery investment

Refinery investment has stagnated for a generation, but 2009-10 could be a turning point, especially East of Suez

In the aftermath of the 2005 hurricanes we ran out of spare refining capacity, but investments made since oil prices began rising early this decade should cause refinery capacity additions to outpace demand growth in 2009 by as much as 2:1 Global CDU Capacity Additions (k b/d) Global Upgrading Capacity Additions (k b/d) (1)

1,599

2006 2007 2008 2009 2010 2011 2012

1,005 1,667 1,678 1,596 2,459 2,832 3,108

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

________________ Source: Lehman Brothers Estimates.

1.

Middle East Rest of World

Includes coking, catalytic cracking, and hydrocracking units and expansions.

4.0

1.6

1.4

1.2

1.0

0.8

0.6

0.4

0.2

0.0

809

2006 China

704 1,152

2007 2008 South Asia

1,284

2009 2010 Middle East

960 1,116

2011 2012 Rest of World 12

Light-heavy spreads to narrow as a result

As the downstream bottleneck is eroded, so should the light heavy premium erode

 Current environment holds upside potential for light sweet benchmarks and light-heavy spreads – As complex refineries break down, more simple crude capacity than what was lost is needed to manufacture the same amount of refined light product • Most likely the additional crude demanded by the simple refineries will be light sweet  But as the downstream bottleneck is eroded, so should the light-heavy crude oil premium erode 13

Conclusions

Conclusions

 Oil markets are not straightforward, not what most people think  The markets perform a fundamental functions but have serious shortcomings  Markets are perfectible but any change will affect some very large interests and is therefore difficult  Looking for feasible changes that will improve the stability and transparency of the markets is an important priority

A new price regime?

      There is widespread dissatisfaction with the current price regime An agreed target price?

Curbing of “speculators”?

Better benchmarks?

Or more physical trading of the main crude oil streams?

I believe the latter is the answer, but implementation is difficult.

Physical vs. Paper

  Increasing the weight of physical trading vs. paper trading may limit oscillations because physical excess supply or demand will be visible and will contribute to reversing expectations in the paper market.

The final outcome depends on the behavior of the physical market

Controlled auctions

   Major producers could organize auctions in such a way that they would influence prices (this is done through establishing a demand curve and then setting quantity to be sold, thus determining the price).

As a border case of the above procedure, major producers could simply fix (post) a price. However, if they did so not price discovery mechanism would be in place.

The combination of a free secondary market with controlled auctions for primary sales probably offers the best compromise between price discovery and an effective damper.