Is LNG Arbitrage Possible in Natural Gas Market?

Is LNG Arbitrage Possible in Natural Gas Market? Jalal Dehnavi, 1 Ferdowsi University of Mashhad, Iran and University of Vienna, Austria Yuri Yegoro...
Author: Solomon Simon
2 downloads 0 Views 167KB Size
Is LNG Arbitrage Possible in Natural Gas Market?

Jalal Dehnavi, 1 Ferdowsi University of Mashhad, Iran and University of Vienna, Austria

Yuri Yegorov, 2 University of Vienna, Austria 14 June 2012

Abstract The dynamics of markets for natural gas during the last years included the higher role of LNG in gas trade, the growing role of spot markets, deregulation, liberalization and competition in national gas markets. Rising flexibility in contracts’ destination clauses created new international arbitrage opportunities. However, technical, contractual and market restrictions, difference in LNG qualities, shipping capacity limitation and high transportation costs are the most important barriers for an arbitrageur in LNG market. This paper studies the main barriers for LNG arbitrage and its cost. Using the time series for natural gas prices in different regional markets, we show to what extent arbitrage between three main hubs was possible, and discuss why the markets do not converge to unique world price for natural gas. Key Words: arbitrage, LNG trade movement, price dynamics. JEL Classification: N7, P28

Acknowledgment: Jalal Dehnavi is grateful to Prof. Dr. Franz Wirl who has invited him for research visit to University of Vienna, where this paper has been written.

                                                             1

E-mail: [email protected] and [email protected], Ferdowsi University of Mashhad (FUM) campus, Azadi Sq., Mashhad, Khorasan Razavi, Iran. P.O. Box: 9177948974. Tel: +98 (0)511-8802000. 2 Corresponding Author; e-mail: [email protected], Tel.: +43-1-4277-38108; Fax: +43-1-4277-38104, University of Vienna, BWZ, Brünner Strasse, 72, A-1210, Vienna, Austria

1    Electronic copy available at: http://ssrn.com/abstract=2110234

1. Introduction Arbitrage is one of fundamental concepts in finance. Arbitrage is defined as “the simultaneous purchase and sale of the same, or essentially similar, security in two different markets for advantageously different prices” (Sharpe and Alexander, 1990). “Such arbitrage requires no capital and entails no risk”, (Shleifer and Vishny, 1997). Indeed, in arbitrage case arbitrageurs are benefit from market in-efficiency and supply-demand imbalances. In natural gas market story is quite different. This paper deals with arbitrage opportunities in the markets for liquefied natural gas (LNG). LNG arbitrage is a physical cargo diversion from one market to another or re-exports of cargoes to spot gas markets in which gas prices are as much as high that covers the costs and profit remains. LNG liquefaction costs, transportation costs and regasification costs are considerable. So one cannot say that arbitrage in gas market is cost less, such as financial markets (paper markets).3 Besides, the traditional destination clause, technical and market restriction, limitation in shipping capacity and relatively high transportation costs are the most important barriers that the arbitrageur faces in natural gas market. However it seems that these limitations will gradually remove. The typical contracts for pipeline gas are long term, while for LNG we often have spot market. Due to substantial transport costs (their share in total costs is much higher than for oil) there is no unique world price for natural gas (see Yegorov and Wirl, 2010). Instead, we have quite different series of spot prices for natural gas in different hubs. With the development of LNG technology those costs have been reduced, but they are still important. In other words, costs throughout the LNG value chain have been declining in recent years.4 However, advocates of unique world market for gas have to take into account that arbitrage is costly, and thus can lead only to some convergence of prices between different regional markets, but not to the full price equalization. In addition, LNG contracts changed remarkably during last decades. Short-term contracts and spot deals have been a usual practice since early 1990s.5 Second, importers have got a right to resale excess volumes or return them to a seller at some fee. This flexibility option allows market players to benefit from arbitrage opportunities playing on price differential across regional markets (Ikonnikova, 2009). For example, in 2007, the Equatorial Guinea LNG project sold its entire LNG output on an FOB basis to BG for 15 years, without a destination clause in the contract. Other significant sources of flexible cargos are Qatar, Trinidad and Tobago, Algeria and Egypt (Zhuravleva, 2009). But even if these difficulties can be overcome, the significant fixed costs, high volatility of the arbitrage and its sensitivity to market conditions, make arbitrage a relatively high risk activity (Holleaux, 2007).                                                              3 . “The high costs of LNG transportation still make it difficult to move the commodity physically over long distances. Only when there is surplus capacity in liquefaction plants and tankers can LNG compete in distant markets”. James. T. James, 2004. 4 . The Global Liquefied Natural Gas Market: Status and Outlook, 2003. 5 . The spot LNG market has been growing from early 1990s. Spot LNG trading accounted for 17 percent of the total gas trade in 2009 (Kanai, 2011).

2    Electronic copy available at: http://ssrn.com/abstract=2110234

The paper presents both empirical data and model of LNG arbitrage. The remainder of this paper is organized as follows. Section 2 deals with world trade movement of LNG. Section 3 investigates natural gas pricing mechanisms and the value of LNG arbitrage costs, including LNG transportation costs. In section 4, a simple model for LNG arbitrage is presented. Section 5 includes some theory about LNG arbitrage considering particularities of LNG market. In section 6, dynamics of price gaps among different markets are investigated. Finally, concluding remarks and policy implications are given in Section 7.

2. World LNG Trade Movement In this part we investigate world LNG trade movement country by country. According to the data presented in Table 1, in 2010, Qatar, Indonesia and Malaysia are the 3 biggest LNG exporters, respectively. Besides, the biggest importers are Japan, South Korea and Spain. Total LNG trade in 2010 experienced 22% increase than 2009. Based on below table in 2010, 19 countries export LNG and 23 countries imported.     Table 1: World LNG Trade Movement 6 TO

US

T&T

Pe

Be

No

Rus

Al

Eg

US Canada Mexico Argentina Brazil Chile Dominica Puerto Rico Belgium France Greece Italy Portugal Spain Turkey UK Kuwait UAE China India Japan S. Korea Taiwan LNG. EXP

--

5.38

0.45

--

0.76

--

1.59

0.08

--

0.08

--

--

0.26

--

--

--

Eq.G

Lib

Nig

--

--

2.07

--

--

--

--

0.16

--

Om

Q

--

--

1.18

--

--

--

--

--

--

2.23

--

1.02

Au

Br

Mal

LNG Imp

--

--

12.33

--

--

2

--

5.72

UAE

Yem

Ind

1.29

--

1.10

--

--

0.25

--

--

--

--

--

0.18

--

--

1.87

--

1.63

--

--

--

--

--

--

--

--

--

--

0.15

--

--

--

--

--

--

1.78

0.09

0.85

0.16

0.08

--

--

--

--

0.08

--

0.89

--

0.59

0.04

--

--

--

--

--

2.78

--

0.52

--

--

--

--

0.17

0.55

1.50

--

--

--

0.25

--

0.08

--

--

--

--

3.07

--

0.82

--

--

--

--

--

--

--

--

--

--

--

--

--

--

--

--

--

0.82

--

0.77

--

--

--

--

--

--

--

--

--

--

--

--

--

--

--

--

--

0.77

0.05

0.08

0.08

--

0.09

--

--

0.17

--

--

0.16

--

5.80

--

--

--

--

--

--

6.43

--

0.35

--

--

0.51

--

6.27

0.73

--

--

3.57

--

2.43

--

0.08

--

--

--

--

13.94

--

0.08

--

--

--

--

0.98

0.08

0.03

--

--

--

--

--

--

--

--

--

--

1.17

--

0.32

--

--

5.75

--

27.56

0.72

0.09

--

--

--

6.18

--

--

--

--

--

--

9.08

--

0.18

--

--

0.05

--

--

--

--

--

2.70

--

0.08

--

--

--

--

--

--

3.01

0.12

3.32

0.63

0.08

3.3

--

12.05

2.62

--

0.34

7.82

0.17

5.54

--

0.18

--

--

--

--

27.54

--

0.26

--

0.08

0.26

--

3.87

0.27

--

--

1.26

--

1.92

--

--

--

--

--

--

7.92

0.18

1.63

--

--

0.94

--

1.25

0.12

--

--

0.40

--

13.89

--

0.26

--

--

--

--

18.67

--

0.33

--

0.09

--

0,09

--

0.33

0.29

--

0.08

0.91

--

0.25

0.09

0.09

--

--

0.24

2.78

--

--

--

--

--

--

--

--

--

--

--

--

0.16

--

--

--

--

--

--

0.16

--

0.07

0.08

0.08

--

0.51

--

0.08

0.08

--

0.17

--

1.61

0.08

0.70

5.21

--

2.45

1.68

12.80

--

0.66

--

--

--

--

--

0.09

0.17

--

0.33

--

10.53

--

0.37

--

--

--

--

12.15

0.85

0.15

--

0.08

--

8.23

0.08

0.57

0.72

--

0.84

3.80

10.15

6.86

0.16

17.66

7.78

17

18.55

93.48

0.35

0.88

0.08

0.08

0.16

3.90

--

0.98

1.85

--

1.18

6.11

10.16

0.25

2.27

1.33

1.05

7.42

6.39

44.44

--

0.51

--

--

0.07

0.67

--

0.17

0.35

--

1.09

0.50

3.75

0.42

--

1.06

--

2.62

3.68

14.90

1.64

20.38

1.82

0.57

4.71

13.40

19.31

9.71

5.16

0.34

23.90

11.49

75.75

7.90

5.48

25.36

8.83

31.36

30.54

297.63

Notes: there are no exact numbers about the LNG export from Algeria and Norway to Italy and Spain, Numbers that are presented in above table are for both LNG and pipeline exports. But LNG exports from Algeria to these two countries are 6.69 bcm and from Norway to so-called countries is 1.79 bcm. Source: Based on BP, Statistical Review, 2011. 

                                                             6 LNG Exporter countries are USA (US), Trinidad and Tobago (T&T), Peru, Belgium (Be), Norway (No), Russia Federation (Rus), Algeria (Al), Egypt (Eg), Libya (Lib), Nigeria (Nig), Oman (Om), Qatar (Q), United Arab Emirates (UAE), Yemen (Yem), Australia (Au), Brunei (Br), Indonesia (Ind) and Malaysia (Mal).

3    Electronic copy available at: http://ssrn.com/abstract=2110234

LNG trade movements from different countries to different regions shows in table 2, as it is presented approximately 60% of LNG import belong to countries in Asia-Pacific region. On the other hand, till now, Africa is the only region in the world that has never imported LNG. Table 2: World LNG Trade Movements to Different Regions  TO North America S.& Central America Europe & Eurasia Africa Middle East Asia Pacific LNG. EXP

US

T&T

Pe

Be

No

Rus

Al

Eg

Eq.G

Lib

Nig

Om

Q

UAE

Yem

Au

Br

Ind

Mal

LNG Imp

--

6.97

0.79

--

0.84

--

--

2.23

--

--

3.41

--

2.56

--

1.28

--

--

1.87

--

19.95

0.09

4.59

0.16

0.08

--

--

0.17

0.55

1.58

--

0.89

--

0.99

0.04

0.08

--

--

--

--

9.22

0.35

6.22

0.71

0.16

56.98

--

51.98

4.71

0.12

0.34

15.91

0.17

35.84

--

0.52

--

--

--

--

87.75

--

--

--

--

--

--

--

--

--

--

--

--

--

--

--

--

--

--

--

--

--

0.33

--

0.09

--

0.09

--

0.33

0.29

--

0.08

0.91

17.41

0.25

0.09

0.09

--

--

0.24

2.94

1.20

2.27

0.16

0.24

0.23

13.31

0.08

1.89

3.17

--

3.61

10.41

36.20

7.61

3.50

25.27

8.83

29.49

30.31

177.77

1.64

20.38

1.82

0.57

4.71

13.40

19.31

9.71

5.16

0.34

23.90

11.49

75.75

7.90

5.48

25.36

8.83

31.36

30.54

297.63

Source: Based on BP, Statistical Review, 2011. 

3. Natural Gas Pricing Mechanisms and LNG Transportation Costs 3.1. Natural Gas Pricing Mechanisms

During the last two decades, two opposing natural gas (NG) pricing systems have coexisted in international gas exporting contracts. The first mechanism is oil indexed gas pricing system which indexes the gas price to oil or basket of oil products in long-term oil contracts. This mechanism is used mainly in long term contracts in the Middle East, Asia and Continental Europe. For example,  in most of the East Asian LNG contracts, gas price is indexed to a basket of crude imported to Japan called the Japan Crude Cocktail (JCC).7 In contrast, there is another mechanism, gas to gas competition pricing system or spot pricing mechanism, which decouples the prices of gas and oil. This gas pricing system is widely used in competitive markets especially in the US, Henry Hub, and the UK hub, the NBP. Hence, because of different gas pricing mechanisms and due to variety of variables considered in gas pricing mechanisms in different regions, gas prices are quite different in different regions. In recent years, energy and non-energy events have begun to exert substantial pressure on the oil-indexed pricing mechanism. The economic recession starting in late 2008, the substantial growth of LNG capacity and world LNG trade volumes, and rapid increase in natural gas production from shale gas formations lead to oversupply of natural gas. The result of gas oversupply is more flexibility                                                              7

In Indonesian LNG contracts, price formula is linked to Indonesian Crude Price (ICP).

4   

in gas pricing mechanism and more development of spot pricing mechanism. Besides that, in recent years markets became more opened to competition through deregulation and liberalization. Competitive markets for gas are emerging and new gas services are being developed. Fig.1. shows gas prices evolution in different markets. According to Fig.1 gas prices are decoupled, market is not efficient and there is a room for arbitrageurs to make money.

Fig.1. Natural Gas Prices (US $ per mBtu) 14

12

10 Japan cif

8

Eropean Union  cif UK, NBP

6

US, Henry Hub Canada

4

2

0 1994

1996

1998

2000

2002

2004

2006

2008

2010

2012

Source: Based on BP Statistical review, 2011. Note: cif = cost + insurance + freight (average prices).

From Table 3 we can see that LNG prices in different spot markets became decoupled during recent years. This is good news for arbitrageurs. Meanwhile, LNG transportation costs are decreasing gradually.

Table 3: Natural Gas Prices in 3 different regions, 2000 to present (Current US $ per mBtu) Year 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012,M01 2012,M02

US 4.31 3.96 3.36 5.49 5.89 8.92 6.72 6.98 8.86 3.95 4.39 4.00 2.68 2.52

Europe 3.86 4.06 3.05 3.91 4.28 6.33 8.47 8.56 13.41 8.71 8.29 10.52 11.45 11.12

Japan 4.71 4.63 4.28 4.73 5.13 5.99 7.08 7.68 12.55 8.94 10.85 14.66 16.50 16.25

Source: World Bank Commodity Price Data.

5   

3.2. Transportation Distances and Costs between LNG Spot Markets In natural gas market, there are three important spot markets: the Henry Hub, the UK hub (NBP) and Japanese spot market. Table 4 presents distance and transportation costs between 3 largest LNG markets are presented. The UK-Japan route is one with the largest transport cost.

Table 4: Approximate Distance and Transportation Costs between 3 Largest LNG markets Countries US- Japan

Exporting Terminal New Orleans

US-UK

New Orleans

JapanUK

Isle of Grain Dragon Dragon

Importing Terminal Himeji Sodegaura Isle of Grain Dragon Himeji Himeji Chita

Distance (miles) 9424 9141 4777 4468 10917 10742 10826

LNG Transportation Cost (US$/mbtu) 1.13- 1.88 1.09- 1.82 0.573- 0.954 0.53- 0.89 1.308-2.185 1.284-2.148 1.296-2.164

Note: Costs based on a 138000 cubic meter tanker at a charter rate of 65000 $ per day. Source: Research Calculation based on: http://sea-distances.com/, GIIGNL, 2011, EIA, 2003.

3.3. The Value of LNG Arbitrage Costs LNG value chain includes 5 important costs: production (extraction) cost, liquefaction cost, transportation costs, regasification costs and selling costs. In the case of LNG arbitrage production and selling costs doesn’t matter. According to an IEA report that released in 2003 LNG cost decrease throughout its value chain. “Costs throughout the value chain have been declining in the LNG industry in recent years. According to the Gas Technology Institute (GTI), liquefaction costs have decreased 35 to 50 percent over the past ten years, with plant capital costs decreasing from more than US$500 per ton of annual liquefaction capacity to less than US$200 for trains at existing plants (in nominal dollars). Building costs for LNG tankers have decreased from about US$280 million (nominal) in the mid-1980s to about US$155 million in late 2003. Regasification terminal costs have also fallen, though costs tend to be site-specific and can range from US$100 million to more than US$2 billion”. Decreasing the LNG costs is good news for arbitrageurs. However, in our case because there is excess supply in LNG markets we can easily neglect the fixed costs of building liquefaction and regasification terminals. In this part, based on the availability of data (see table 5) and considering the distance between LNG routes (see Appendix 1), LNG transportation cost estimated. Results show that approximately the transportation cost of LNG is between 12 to 20 cents per million Btu per 1000 miles.8                                                              8

Based on table 3 and table 1 in appendix, distance between T&T to Everett is 2032 mile; T&T to Elba Island is 1690 miles, Nigeria to Cove point is 5256 miles, Norway to Cove point is 3975 miles, Qatar to Elba Island is 8716

6   

Table 5: Representative LNG Shipping Rates (Dollars per Million Btu) Exporter Algeria Nigeria Norway Venezuela T&T Qatar Australia

Everett 0.52 0.80 0.56 0.34 0.35 1.37 1.76

Cove Point 0.57 0.83 0.61 0.33 0.35 1.43 1.82

Elba Island 0.60 0.84 0.64 0.30 0.32 1.46 1.84

Lake Charles 0.72 0.93 0.77 0.35 0.38 1.58 1.84

Note: Prices based on a 138000 cubic meter tanker at a charter rate of 65000 $ per day. Source: The Global Liquefied Natural Gas Market: Status and Outlook, Report #:DOE/EIA-0637, Release Date: December 2003, EIA, US energy Information Administration, Based on: LNG shipping Solutions Data.

4. Simple LNG Arbitrage Model Consider that we have three players in LNG market; LNG exporter country, LNG importer country and Speculators (arbitrageur). The possible opportunities for LNG arbitrage using a simple model are presented in Fig.2. Fig.2. Model of Physical LNG arbitrage

Cargo Diversion

2

LNG Exporter

LNG Spot Markets, (HP)

7

3 Long-Term Contract

LNG Spot Markets, (LP) 1

Re- Exporting 5 6

LNG Importer

4

Arbitraguer

Source: Our Model Notes: LNG spot markets (HP) are the LNG markets that LNG price in them is relatively higher than other spot markets (e.g. Japan LNG spot market). LNG spot markets (LP) are the LNG markets that LNG price in them is relatively lower (cheaper) than other spot markets (e.g. Henry Hub).

In principle, rational producer of LNG can also be an arbitrageur, but we do not focus on this case. Before 2006, long-term contracts for LNGs with destination clause and pricing formula were quite common (Zhuravleva, 2009). It is possible that LNG price (based on long term contract pricing                                                                                                                                                                                                  miles. So, we can suggest that transportation cost approximately would be between 12 – 20 cents, per 1 mBtu per 1000 miles.

7   

mechanism) will be higher than LNG price in spot market today. In this case LNG exporter and LNG importer can cooperate and earn money by cargo diversion or re-exporting. Recently, it became acceptable that LNG cargos even in long-term contracts can be diverted to another market with the mutual agreement of both the seller and the buyer. In this case both sellers and buyers have financial incentives; if buyer can buy LNG in regional market in a price equal or less than the contract price, there is a room for earning profit from cargo diversion by seller. However, the profit shares depend on bargaining power of players (buyer and seller). The importer country can re-export all or a part of the LNGs that are imported. In long-term contracts it may happen that importer’s forecast about their future demand is overestimated, so they can re-export the additional gas. For example, USA has a long term contract with Qatar on LNG imports. However, after discovery of economic technology for producing shale gas, US domestic gas price dropped and its demand for LNG imports went down. Now it diverts LNG imports (due to long term agreement with Qatar) to other destinations. A third party (arbitrageur) can buy LNG either from exporter or importer countries and sell it in a spot market. If markets are inefficient, then arbitrageur can buy LNG in markets with lower price and resell it in other markets. In all of the arbitrage cases mentioned above arbitrage is not costless and easy. Arbitrageurs have to overcome LNG transportation limitations and cover the transportation costs. Besides that, they are facing risk of price change during the time of transportation, as well as the negative impact of any arriving LNG vessel on spot price.

5. Some Theory. Particularities of LNG Market In economic theory we name a commodity a good in particular location and in particular moment of time. Suppose that some physically identical good is sold in two markets, A and B, and has prices p(A,t) and p (B,t) there. Equilibrium in markets A and B is determined by the local balance between demand D(i, t) and supply S(i, t), (i=A,B), in each time moment t. This local equilibrium determines price path p(i, t). Besides that, there should be no arbitrage opportunity between markets A and B. If transport cost for unit of this good between markets A and B equals to  and if transportation takes virtually no time, then in equilibrium we do not have a unique price p(A,t)=p(B,t). This would be the case only if arbtrage is costless: =0. Instead, we have the continuum of possible equilibrium prices in each time t, that satisfy the following constraint: |p(A,t)-p(B,t)| .

(1)

8   

Contrary to some other goods with substantial share of transport cost in total price (see Grohall and Yegorov, 2010), LNG has one more property, making arbitrage more difficult. It is related to the huge size of vessels, making arrivals of LNG to different ports a discrete process. Discreteness of vessels makes it unprofitable to send only a fraction of vessel between markets A and B, if there is suspicion that sending the whole vessel would drive price to the extent, that no gain from such trade could be obtained. Since any physical arbitrage between different hubs, A and B, includes time of vessel arrival T, the price difference out of the constraint (1) is still not sufficient for successful arbitrage.The reason is that price in time t+T is not known. This puts additional difficulty for arbitrage seeker. There is one more reason to make arbitrage ambigous, and it is related to absence of unique value of parameter  between each pair of hubs. It is known that LNG transportation includes both fixed and variable costs, and a smart arbiteger should take into account both. Just to give a simple example. Suppose, an arbitrageur would like to buy cheap gas in the USA and send it to Europe. Let the price difference between hubs be substantially above the variable cost of sending one vessel. But is the gas in US is in gas state, there is additional cost to make it liquid, and it comes from the operation on liquefaction plant, that hassubstanbtial cost to be built. However, when we have competition between large firms, owning both plants and vessels, they sometimes can neglect this fixed cost. Putting all together, we can expect that price difference across two hubs is not a process with fixed, but rather with a fuzzy parameter .

6. Dynamics of Price Gaps among different markets In this part, we investigate the possibility of LNG arbitrage among gas hubs considering the dynamics of price gaps among 3 different markets (Henry Hub, NBP, and Japan spot market). For simplification we assume that LNG transportation costs as a fixed number. However, as we mentioned earlier there are tree estimation of LNG transportation costs (minimum cost, average cost and maximum cost) that we use the boundary costs (minimum and maximum) in our study. The results are presented on Fig.3-5. All the graphs show the evolution of price differential between two corresponding hubs (red). The upper (lower) curve adds (subtracts) the cost of arbitrage. The arbitrage is profitable if all three curves are in the upper area (then Japanese price exceeds US price by more than arbitrage cost, and profitable arbitrage is from US to Japan) or in the lower area (then the arbitrage from Japan to US is profitable). Here we do not take into account, that arrival of even one vessel can change the price in destination point, and that the relevant price is one of later

9   

date, since transportation needs time). Thus, positivity or negativity of all curves is only a necessary (but not sufficient condition!) for arbitrage. Fig.3. Dynamics of price gap between Japan and US Part A: Minimum Transportation cost 14 12 10 8 6

(PJ-Pus)-Cmin

4

PJ-PUS

2

(PJ-PUS)+Cmin

0 -21996 1998 2000 2002 2004 2006 2008 2010 2012 -4 -6

Part B: Maximum Transportation Cost 14 12 10 8 6 (PJ-PUS)-Cma x 4

PJ-PUS (PJ-PUS)+Cma x

2 0 1996 -2

1998

2000

2002

2004

2006

2008

2010

2012

-4 -6

Source: BP, Statistical Review, 2011 and World Bank Commodity Price Data, March, 2012. Notes: PJ is gas price in Japan; PUS is gas price in US hub (Henry Hub); Cmin and Cmax are minimum and maximum transportation costs, respectively.

As it is shown on Fig.3, the gap between LNG prices in Japan and USA before 2008 was not high enough to allow for arbitrage. Since 2009 it became well above transport cost, and thus allowed for arbitrage. Since the price of making and transporting LNG has a downward trend, arbitrage cost has been 10   

a bit larger than we use before 2003 and a bit lower after. However, variable transport cost also depends on the dynamics of oil price, and due to its positive trend in the last 10 years. Although our graph suggests a possibility of profitable arbitrage after 2008, account for fixed cost is also needed. Fig.4. Dynamics of price gap between Japan and UK Part A: Minimum Transportation cost 7 6 5 4 3

(PJ-PUK)-Cmin

2

PJ-PUK

1

(PJ-PUK)+Cmin

0 -1 1996 1998 2000 2002 2004 2006 2008 2010 -2 -3 -4

Part B: Maximum Transportation Cost 7 6 5 4 3

(PJ-PUK)-Cmax

2

PJ-PUK

1

(PJ-PUK)+Cmax

0 -1 1996 1998 2000 2002 2004 2006 2008 2010 -2 -3 -4

Source: BP, Statistical Review, 2011 and World Bank Commodity Price Data, March, 2012. Notes: PJ is gas price in Japan; PUS is gas price in UK hub (NBP); Cmin and Cmax are minimum and maximum transportation costs, respectively.

11   

Looking on Fig.4, one can see that the price gap between Japan and UK did not allow for arbitrage for all years, except for the years 2009-11. But the gain is small, and it is not clear if it covers fixed costs and associated risks of price change.

Fig.5. Dynamics of price gap between US and UK Part A: Minimum Transportation cost 8 7 6 5 4 3

(PUK-PUS)-Cmin

2

(PUK-PUS)

1

(PUK-PUS)+Cmin

0 -11996 1998 2000 2002 2004 2006 2008 2010 2012 -2 -3 -4

Part B: Maximum Transportation Cost 8 7 6 5 4 3

(PUK-PUS)-Cmax

2

(PUK-PUS)

1

(PUK-PUS)+Cma x

0 -11996 1998 2000 2002 2004 2006 2008 2010 2012 -2 -3 -4

Source: BP, Statistical Review, 2011 and World Bank Commodity Price Data, March, 2012. Notes: PUK is gas price in UK; PUS is gas price in US hub (Henry Hub); Cmin and Cmax are minimum and maximum transportation costs, respectively.

12   

Dynamics of price gap between UK and US (red) accounting for maximal transport cost Cmax We see that arbitrage opportunities did not exist only in 1996-2005 (not all lines are below or above zero) After 2006 arbitrage can take place even for high cost. We need to think about some explanation for existence of arbitrage opportunity between gas market in the USA (Henry Hub) and other markets in the recent years. It is definitely not market equilibrium, or we have missed something. We know that USA re-export some LNGs coming from Qatar. But why don’t they export domestically produced gas, if the price difference favors that? First, we need to account for high fixed cost of liquefaction plants in the USA (that probably are not yet built). It might also happen that there are some national regulations (like high export tax). It might be better for USA to keep additional gas reserves (mostly non-conventional) for future generations and to stay in a kind of autarchy in the middle term, just producing and consuming their own gas, maintaining regional price for it below one in other regions. Or it might happen that low current natural gas price

in US is

because of implicit subsidies for shale gas and some expectations of costs to grow. If this is the case, then rational investor might not build liquefaction plant, if the expected price gap between US and other markets in future will not cover transport cost plus this high fixed cost.

7. Concluding remarks and policy implications It has been argued that LNG will drive the world markets for natural gas from regional to a unique world price, similarly to what we observe for oil market and some other commodities. Indeed, we observe the growing role of LNGs in the world trade in natural gas. Here we study the evolution of gas prices in different hubs (located in US, UK and Japan). Further, we estimate the costs of arbitrage across locations. The paper studies to what extent prices for natural gas in different hubs converge to unique world price and whether arbitrage between any pair of hubs can exist. We find that there are practically no arbitrage opportunities between Japan and UK during almost the whole observation period. This happens because of large distance and high transport costs. At the same time, the price gap between US and UK is often larger than transport cost between these locations, potentially allowing for arbitrage. Interestingly, before 2006 prices in the USA have been higher those in UK, but after 2008 the reverse happened. This, we have either to conclude that those markets are not in equilibrium, or that we have underestimated the possible opportunities and associated risks.

13   

Appendix 1: Sea Transportation Distance for Different LNG Routes Exporter

Algeria

Egypt

Equatorial Guinea

Nigeria

Norway

Trinidad and Tobago

Importer Spain Italy India France Turkey Greece Netherlands Portugal Japan Spain India China UK Netherlands US Portugal turkey France Japan Argentina Korea Greece Italy Kuwait Taiwan Chile Greece China Japan Korea Taiwan Chile Mexico Spain Argentina US India china UK France Japan Korea Thailand Turkey Kuwait UAE Brazil Greece Netherlands Portugal Taiwan Belgium Spain US Japan Korea India UK France Dominican Netherlands Italy Portugal Taiwan Spain US Argentina Canada India

Sea Transportation Distance (miles) min Max Average 113 1675 589 456 684 533 4421 4421 4421 520 1260 770 1500 1500 1500 1270 1270 1270 1714 1714 1714 568 568 568 9491 9491 9491 1554 2580 1873 3142 3153 3147 6556 6556 6556 3041 3041 3041 3346 3346 3346 5495 6640 6255 2182 2182 2182 603 603 603 1430 2771 1880 7607 8111 7913 7490 7490 7490 7764 7764 7764 540 540 540 1299 1299 1299 3414 3414 3414 6824 6824 6824 10493 10493 10493 540 540 540 9516 10602 10052 10591 10955 10765 10578 10651 10620 9657 9657 9657 6752 6752 6752 6214 6214 6214 3359 3914 3683 4662 4995 4828 5256 5256 5256 7053 7136 7094 9328 10328 9985 4206 4469 4337 3980 4091 4035 10600 10790 10695 10354 10657 10466 8708 8708 8708 5059 5059 5059 7588 7588 7588 7209 7209 7209 2811 2811 2811 4899 4899 4899 4493 4493 4493 3417 3417 3417 9440 9440 9440 4424 4424 4424 2045 3155 2632 3975 5455 4715 12154 12344 12226 12211 12140 12188 7665 7665 7665 1423 1599 1511 1889 3349 2619 4613 4613 4613 1401 1401 1401 4196 4196 4196 2398 2398 2398 11238 11238 11238 3976 3417 3643 1690 2247 2095 4628 4920 4774 2150 2150 2150 8463 8463 8463

Number of Routes 12 3 1 3 1 1 1 1 1 5 2 3 1 1 4 1 1 3 12 1 1 1 1 1 1 1 1 2 5 3 1 1 1 6 2 1 2 4 2 2 4 3 1 1 1 1 1 1 1 1 1 1 6 2 3 3 1 2 2 1 1 1 1 1 5 7 2 1 1

Exporter

Oman

Russia

Yemen

Australia

Peru

Qatar

Malaysia

Indonesia

14   

Importer Spain Japan Korea India Taiwan Japan China Korea Thailand Taiwan Japan US China Korea Mexico France India UK Taiwan Belgium Japan Korea China India UAE Taiwan Kuwait Spain US China Korea Thailand Taiwan Mexico Japan Spain Japan China UK Korea US Turkey Mexico Argentina Canada India France Brazil Thailand Kuwait UAE Italy Chile Greece Netherlands Portugal Belgium Taiwan (Lumut)-Korea (Lumut)-Japan (Bintulu)-Japan (Bintulu)- Korea (Bintulu)- China (Bintulu)-India (Bintulu)-UAE (Bintulu)-Kuwait (Bintulu)-Taiwan Japan Korea

Sea Transportation (miles) min max 4159 4260 5700 5873 5300 5750 773 773 4719 4719 581 1233 1410 2244 1363 1763 3356 3356 1967 1967 6433 6435 6373 8146 5108 5802 5625 6243 8313 8313 2993 4505 1703 1703 4735 4735 5268 5268 4690 4690 3017 3995 3526 3613 2770 3053 3857 3857 5231 5231 2430 5231 5041 5041 9053 9639 10215 10215 9325 9603 9266 9274 11027 11027 9740 9740 10298 10298 8576 8576 4710 5925 6170 6576 5098 5935 6137 6428 5706 6156 8716 9796 3722 3722 9922 9922 8630 9023 8007 8007 1236 1290 4684 6015 8197 8621 4326 4326 354 354 231 231 4438 4774 10040 10040 3696 3696 6509 6509 5291 5291 6277 6277 5230 5230 2014 2850 2405 2999 1256 2530 2124 2124 1256 2205 3337 3337 4110 4110 4479 4479 1350 1350 2211 2857 2043 3149

Distance average 4210 5844 5600 773 4719 1039 1581 1630 3356 1967 6434 7545 5514 5964 8313 3749 1703 4735 5268 4690 3518 3584 2911 3857 5231 5231 5041 9401 10215 9464 9270 11027 9740 10298 8576 5165 6458 5686 6249 6006 9445 3722 9922 8826 8007 1263 5349 8409 4326 354 231 4606 10040 3696 6509 5291 6277 5230 2571 2434 2382 2152 1836 3337 4110 4479 1350 2470 2742

Number of Routes 2 6 3 1 1 16 4 3 1 1 2 3 3 3 1 2 1 1 1 1 18 3 2 1 1 1 1 6 1 2 2 1 1 1 1 6 10 4 3 3 3 1 1 2 1 2 1 2 1 1 1 2 1 1 1 1 1 2 3 4 14 3 4 1 1 1 1 22 8

China UK France Brazil Spain Korea Chile Kuwait Puerto Rico Dominic Rep Chile Greece Netherlands Italy Taiwan Belgium

12479 3734 1618 1732 9230 9303 7596 10541 560

13007 4064 4147 3245 13721 9685 7596 10541 560

12743 3899 2882 2488 11475 9494 7596 10541 560

2 2 2 2 2 2 1 1 1

679

679

679

1

7051 4965 4102 5180 10174 3985

7051 4965 4102 5180 10174 3985

7051 4965 4102 5180 10174 3985

1 1 1 1 1 1

Libya UAE

China Thailand Taiwan Mexico Chile Spain India Kuwait

1856 1610 1455 6850 9135 1061 1227 390

2227 1610 1972 6850 9135 1496 1244 390

2041 1610 1713 6850 9135 1244 1235 390

2 1 2 1 1 3 2 1

Source: Based on “The LNG Industry”, International Group of Liquefied Natural Gas Importers (GIIGNL, 2011).

References 1. BP Statistical Review, 2011. 2. Holleaux D., “Value of Transatlantic Arbitrage”, LNG 15 Conference, Barcelona, 24-27 April, 2007. 3. Grohall G., Yegorov Y. “A Land Far Away”, FIW-Research Report 2010/11, N° 02, http://www.fiw.ac.at/fileadmin/Documents/Publikationen/Studienpool_II/02.ResearchReport.Grohall.Yeg orov.A Land Far Away.pdf. 4. http://sea-distances.com/, GIIGNL, 2011, 5. Ikonnikova S., 2009, Strategic model of LNG arbitrage: analysis of LNG trade in Atlantic Basin, in 32nd International Association for Energy Economics Conference Proceedings (online). 6. Sharpe, W.F., Alexander, G.J., 1990. Investments. Prentice Hall, Englewood, NJ. 7. Schelifer A., Vishny R. W. “The Limits of Arbitrage” The Journal of Finance, Vol. LII. No.1. March, 1997. 8. Jensen James T. “The Development of a Global LNG Market, Is it Likely? If so, When?” Oxford Institute for Energy Studies, (Registered Charity, No. 286084), 2004. 9. “The Global Liquefied Natural Gas Market: Status and Outlook”, Report #:DOE/EIA-0637, Release Date: December 2003, EIA, US energy Information Administration, Based on: LNG shipping Solutions Data. 10. World Bank Commodity Price Data, March, 2012. 11. Yegorov Y., Wirl F. “Gas Transit, Geopolitics and Emergence of Games with Application to CIS Countries”, USAEE - IAEE WP 10-044, February 2010, 29 p. (http://ssrn.com/abstract=1560563 ). 12. Zhuravleva. P. “The Nature of LNG Arbitrage, and an Analysis of the Main Barriers for the Growth of the Global LNG Arbitrage Market”, Oxford Institute for Energy Studies, NG 31, June 2009.

15   

Suggest Documents