School of Business
STOCKHOLM UNIVERSITY
Master thesis 10 credits Spring semester 2006
How Skanska can handle risks
due to price fluctuations in
commodity markets
-Is it Economic Effective to Use a Commodity
Hedge?
Authors: Andrea Arppe Supervisor: Jens Lindberg
Abstract
This thesis is written in commission for Skanska Financial Services (SFS), which is a support unit that services the multinational construction company Skanska AB and Skanska’s Business Units, and coordinates the Group’s relations with financial markets and institutions. Construction companies that undertake construction projects face a number of unique problems. Of particular concern, is the fact that in many instances the final costs may be uncertain subject to substantial change, particularly changes in raw material costs. Increasing crude oil prices could turn projects expected to be profit generating into being unprofitable. The purpose of this thesis is to investigate how to protect a company, which knows it will have to buy a specific energy commodity in the future, against risks due to price fluctuations in the energy commodity market. Bitumen and diesel commodity hedges will be initiated by using the financial derivative instruments futures contracts and swaps. The effectiveness of the hedges will thereafter be evaluated from an economic perspective.
The conclusions from the investigation are that futures contracts are not a good alternative for trying to create an effective bitumen hedge. To avoid the risk of entering a hedge that is neither effective from the perspective of locking in a price nor assumed to result in a financial gain, it is also not recommended to create a diesel hedge by using futures contracts.
When it comes to swap hedges, there is no clear evidence that any of the specific type of swap constructions consistently will result in a total gain or loss. The timing has a major impact on the economic gain of a swap hedge, especially when hedging bitumen exposures. Moreover, it can be said that swap hedging offers great flexibility and possibility to fully lock in diesel and bitumen exposures. For Skanska it is preferable to hedge the diesel (QUSDL50-C-NWE) exposure with a bulk swap hedge for which the currency rate is fixed throughout the lifetime of the hedge. The bitumen (QHFO-ARA) exposure should be hedged with a 1 year swaps or a combined 1 year and 2 year swap hedge construction for which exchange rates are fixed according to the length of the swaps.
Key Words: Skanska, crude oil, bitumen, diesel, financial derivative instruments, futures contracts, swaps, hedging, hedge effectiveness, hedge ratio, prospective test, retrospective test, hypothesis test, IAS 39.
Content
1 Introduction ...5 1.1 Background...5 1.2 Presentation of Problem...6 1.3 Purpose ...8 1.4 Limitations ...8 1.5 Hypotheses ...8 1.6 Contribution...9 1.7 Outline ...9 2 Theoretical Background...102.1 The Oil Market...10
2.2 Hedging...11
2.3 The Futures Market ...12
2.3.1 The Futures Market in Practice...12
2.3.2 Basis Risk ...13
2.4 The Swap Market ...14
2.4.1 The Swap Market in Practice...14
2.4.2 Benefits Versus Risks...15
2.5 IAS 39 ...15
2.6 Earlier Research ...16
3 Method ...18
3.1 Futures Contracts Hedge ...18
3.1.1 Cross Hedge...18
3.1.2 Optimal Number of Contracts ...18
3.1.3 Retrospective Test...19
3.1.4 Prospective Test ...19
3.1.5 Hypothesis Test...20
3.1.6 Futures Hedge Data...20
3.1.7 Currency Exchange Rate Data...20
3.2 Swap Hedge ...21
3.2.1 Swap Hedge Price Data...21
3.2.2 Swap Cases ...21
3.2.3 Swap Evaluation ...22
3.2.4 Currency Exchange Rate Data...22
3.3 Models (explicit for futures)...23
3.3.1 Hedge Ratio ...23
3.3.2 Optimal Number of Contracts ...23
3.3.3 Basis Risk ...24
3.4 Generalization, Reliability, Validity and Criticism of Sources ...25
4 Results and Analysis ...26
4.1 Bitumen Futures Hedge...26
4.2 Diesel Futures Hedge...27
4.3 Diesel Swap Hedge...29
4.4 Bitumen Swap Hedge...31
4.5 Retrospective Tests for Bitumen and Diesel Swap Hedges...33
5 Conclusions and Recommendations...34
5.1 Futures Contracts Hedge ...34
6 Suggestions for Further Research...35 7 References ...36 7.1 Literature ...36 7.2 Working Papers ...36 7.3 Websites ...37 7.4 Verbal Sources ...37 Appendix 1...38 Appendix 2...39 Appendix 3...41 Appendix 4...43 Appendix 5...44 Appendix 6...45 Appendix 7...46 Appendix 8...47 Appendix 9...49 Appendix 10 ...50 Appendix 11 ...51 Appendix 12 ...56 Appendix 13 ...61 Appendix 14 ...67
1 Introduction
1.1 Background
This thesis is written in commission for Skanska Financial Services (SFS), which is a support unit that services Skanska AB and Skanska’s Business Units, and coordinates the Group’s relations with financial markets and institutions.
SFS is responsible for the Group’s debt and for ensuring that the Group has adequate funding. It coordinates and carries out operative financial activities for the Business Units. For projects, SFS provides or procures contract guarantees, insurance and financial solutions. In addition, it manages risks that stem from the Group’s operations, including risks associated with interest rates, foreign exchange, credit and counterparty relationships, funding and liquidity1.
Skanska is a multinational construction company with the mission to develop, build and service the physical environment for living, working and travelling. The vision is to become a world leader in construction-related services and project development. It was founded in Sweden 1887 and today it operates also in the US, UK, Denmark, Finland, Norway, Poland, the Czech Republic and South America. It has been listed on the Stockholm Stock Exchange since 1965. CEO of the company is Stuart Graham and Chairman is Sverker Martin-Löf. Skanska had a revenue of SEK 125 billions and 54 000 employees during 20052.
The thesis will investigate how to handle risk due to price fluctuations in oil based products, by looking at a specific major road project in Poland. The new stretch of motorway is approximately 90 kilometres long and runs from north to south between Gdansk and Nowe Marzy in northern Poland. The entire construction project, which is the largest project to date in Poland, is valued at approximately EUR 500 million (SEK 4.6 billion) and will be led by Skanska Poland (80 percent) and carried out in collaboration with the Polish company NDI (20 percent). Skanska’s share of the contract amounts to EUR 400 million (SEK 3,7 billion) and is included in order bookings for the third quarter of 2005. Skanska Infrastructure Development is part of the ownership and investor consortium Gdansk Transport Company (GTC). Skanska’s share in the company amounts to 30 percent and Laing Roads from the UK, Intertoll of South Africa and NDI of Poland have the remaining shares. Skanska’s investment amounts to approximately EUR 10 million (approximately SEK 94 million).
The project is being conducted as a public-private partnership. The Polish infrastructure ministry and the ownership consortium in which Skanska is a member have signed a concession agreement entailing a total undertaking to design, finance, construct and operate the road. The ownership consortium will be responsible for operation and maintenance during the concession period, which extends until 2039. Payment from the Polish Road Authority will comprise a guaranteed basic payment for access to the road with supplements for traffic volumes through “shadow tolls”3.
1 www.skanska.com 2 Ibid.
1.2 Presentation of Problem
Construction companies that undertake construction projects face a number of unique problems. The contracts they enter into, and the work they do, can be very complex. The values of individual contracts can be extremely high and there is always the possibility of completion dates being delayed. Of particular concern, is the fact that in many instances the final costs may be uncertain and subject to substantial change. The measurement of contract revenue and costs arising from the fact that many constructions extend over long periods of time and are carried out under very uncertain conditions is a major difficulty when evaluating whether to enter a particular contract or not and when accounting for construction contracts. Such uncertainties could include unexpected construction problems, particularly changes in raw material costs.
To a large extent, the future spends for Skanska’s project in Poland will be influenced if the oil market price in general changes. The most obvious examples are;
Material Total cost Influence from crude oil
Bitumen 55 mil. zl 80%
Diesel 65 mil. zl 90%
Heating oil 15 mil. zl 90%
Plastic pipes and geotextiles 25 mil. zl 50%
Transport 140 mil. zl 20%
Total value based on crude oil price 155 mil. zl
Table 1.1 Estimated future crude oil price exposure for the Polish A1 project4
To investigate how to protect Skanska from price fluctuations in the crude oil market, a distinction has to be made between the physical contract, i.e. the specific order received by Skanska, and financial contracts, i.e. contracts entered into to protect the order value against increasing costs. Future oil prices might be significantly different compared to those used as input in order to estimate the value of physical contracts received. Increasing crude oil prices could therefore turn projects expected to be profit generating into being unprofitable.
In order to balance this crude oil risk to a certain extent, Skanska could consider using
financial derivatives, which are assets whose values derive from that of some other assets5. They are used with the purpose to eliminate the risk by creating an effective hedge, i.e. an investment that is taken out specifically to reduce or cancel out the risk in another investment6. By exploiting financial derivatives, a risk averse buyer can use a hedge strategy
to “lock in” the oil price for a specific asset which it knows it will have to buy in the future. This is desirable from an economic perspective, as it enables a company to estimate the value
of a project and to decide whether it should be accepted or not.
One alternative is to use the financial derivative instrument futures on crude oil, which is a contract to buy/sell a commodity or security on a future date at a price that is fixed today7.
4 Thomas Wieland, Skanska, 2006 5 Hull, 2003, p.704
6 Ibid., p. 707
Another alternative is to use the financial derivative instrument swap, which is an agreement between two companies to exchange cash flows in the future8.
In the case of Skanska’s project in Poland, additionally one risk arises due to the fact that the commodity purchases are settled and paid in the Polish currency zloty, while the financial commodity derivatives prices are settled in US dollars. This problem has to be regarded, when evaluating the effectiveness of using commodity derivatives from an economic perspective. In addition to the economic perspective, another reason for the importance of creating an effective hedge is that the European Union (EU) requires all companies listed on a stock exchange in an EU country to comply with International Accounting Standards (IAS) and to prepare consolidated accounts since 1 January 20059. When using financial derivatives, IAS 39 should be applied by all enterprises. If a financial asset instrument has been taken out to act as a hedge, hedge accounting rules should be followed if certain criteria are fulfilled10. IAS 39 requires a prospective effectiveness test to be met for hedge accounting to be available11. The hedging relationship is considered as effective when actual results are within the range of 80% to 125%12. If the prospective test fails to keep within the stated interval, hedge accounting can not be applied. If a company decides to use a hedging strategy anyway, the hedge results must be realized in the income statement of the accounting period. This might cause fluctuations of major significance in current income13, which is undesirable from the perspective of stock exchange listed companies, like Skanska, as the share price would probably fluctuate due to the result changes. This thesis will not look at the implications of IAS 39 values but use the 80%-125% range for evaluation.
Even if hedges are not considered effective from an accounting viewpoint, they should be effective from an economic perspective14, as the reason for entering a hedging strategy otherwise is no longer motivated. Therefore, a good hedge from an economic perspective does not always go hand in hand with the hedge accounting framework. A hedge that fully locks in a targeted commodity price at the same time as being profitable might generate cash flows that fall outside of the interval necessary for maintaining hedge accounting. In these situations, a company must value the economic benefits associated with a hedge more than the possible benefits from using hedge accounting. There are also high administration costs associated with handling and implementing hedge accounting and this factor must also be considered when evaluating a hedge.
8 Hull, 2003, p.125 9 Webpage deloitte.com 10 Ibid., p.343 11 PriceWaterhouseCoopers, 2005, p.48 12 Ibid. 13 Ibid. 14 www.jpmorgan.com
1.3 Purpose
The purpose of this thesis is to investigate how to protect a company, which knows it will have to buy a specific energy commodity in the future, against risks due to price fluctuations in the energy commodity market. The purpose can be separated into two components;
1. To create a commodity hedge by using a financial derivative instrument.
2. To evaluate whether the hedge is effective or not from an economic perspective.
1.4 Limitations
The thesis will only investigate how to protect a company against risks due to price fluctuations by analysing the oil related products bitumen, (bitumen is an asphalt component) i.e. fuel oil 3,5% (QHFO-ARA15), and diesel, (diesel is used for driving machinery etc.) i.e. fuel oil 50 ppm (QUSDL50-C-NWE16). The only financial derivative instruments that will be used are futures contracts and swaps. IAS 39 is a factor of importance when using financial derivative instruments. However, it will not be within the scope of this thesis except from the 80%-125% criteria of effectiveness when evaluating the hedges. Tender phase issues will be disregarded since it is a thesis on its own.
1.5 Hypotheses
To test whether the created commodity hedges are effective or not according to the criteria of IAS 39, the following hypothesises are formulated17;
Hypothesis 1: 8 , 0 : 0 ≤ ∧ β H 8 , 0 : 1 > ∧ β H Hypothesis 2: 25 , 1 : 0 ≥ ∧ β H 25 , 1 : 1 < ∧ β H
The slope parameterβˆ describes the linear relationship between the value of the bitumen/diesel exposure and the value of the financial derivative instrument. IAS 39’s requirement of effectiveness within a range of 80-125% is translated into the requirement of a regression β between 0,8 and 1,2518.
15 ARA = Amsterdam, Rotterdam and Antwerp 16 NWE = North West Europe
17 For an illustration see Appendix 1 18 Reznek, 2005, p.5
INTRODUCTION
THEORETICAL BACKGROUND
METHOD
RESULTS AND ANALYSIS
CONCLUSIONS AND RECOMMENDATIONS
In order to be confident about the effectiveness, the null hypothesis needs to be rejected at a low error probability level. Therefore a 2 % significance level is chosen (α =0,02)19.
1.6 Contribution
This thesis adds to the rather limited amount of earlier research within this field. It is of interest for companies with large exposures towards future raw material costs and therefore are in need to protect themselves against associated risks. This is also of importance for stock exchange listed companies with a desire to smooth the income20 between different accounting periods, as they are concerned about a stable development of the share price21.
1.7 Outline
The following figure illustrates the outline of this thesis;
Figure 1.1 Outline
19 Reznek, 2005, p.12
20 Income smoothing may be viewed as the deliberate normalization of income in order to reach a desired trend
or level.
2 Theoretical Background
2.1 The Oil Market
The market participants can chose between different qualities of crude oil and refined products, of which sweet light crude oil is the most common. The currently most active exchanges are the New York Mercantile Exchange (NYMEX), on which oil is traded as Western Texas Intermediate (WTI) and the International Petrol Exchange (IPE), where oil is traded as Brent22.
The current crude oil and oil products can be classified into three general categories; long-term contracts, spot market trade and derivative trade23.
Trade on the international oil market
Trade with derivatives Trade with the physical asset
Options Swaps Futures Forwards Long-term contract Spot market
Trade on the international oil market
Trade with derivatives Trade with the physical asset
Options Swaps Futures Forwards Long-term contract Spot market
Figure 2.1 Trade on the international oil market24
1. A long-term contract refers to trading crude oil or oil products that for a predetermined price are directly delivered from the producers to the oil customers. The prices are often based on the daily spot market quotations25.
2. The spot market trade for a commodity is the market for immediate delivery and payment26.
3. Examples of derivatives that can be traded on the international oil market are options, swaps, futures and forwards.
- An option is the right, but not an obligation, to buy or sell an asset. A call
option is an option to buy an asset at a specified exercise price on or before a specified exercise date. A put option is an option to sell an asset at a specified 22 Geman, 2005, p.19-20, 201 23 Yazdanfar, 2003, p.45 24 Ibid., p.44 25 Ibid., p.45 26 Grinblatt, 2002, p.779
exercise price on or before a specified exercise date27. A distinction is also made between American options, which are options that can be exercised at any time during its life28, and European options, which are options that can be exercised only at the end of its life29.
- A swap is an arrangement between two companies to exchange cash flows in
the future, e.g. in different currencies, or one at a fixed rate and the other at a floating rate. The arrangement defines the dates when the cash flows are to be paid and the way in which they are to be calculated30.
- A futures contract31 is a contract that obligates the holder to buy or sell an
asset at a predetermined delivery price during a specified future time period. Unlike forward contracts, future contracts are normally traded on an exchange. To make trading possible, the exchange specifies certain standardized features of the contract. The future contract is marked-to-market daily and the contract is usually closed out prior to maturity32.
- A forward contract is an agreement to buy or sell an asset at a certain future
time for a certain price33. The main differences between forward and futures contracts are summarized in Table 1.
Forward Futures
Traded on over-the-counter market Traded on an exchange
Not standardized Standardized contract
Usually one specified delivery date Range of delivery dates
Settled end of contract Settled daily
Delivery or final cash settlement usually takes place Contract is usually closed out prior to maturity
Table 2.1 Comparison of forward and futures contracts34
2.2 Hedging
Market participants with the aim to reduce a specific risk, for example oil price fluctuations, might use futures contracts or swaps trying to create a hedge, which, as earlier described, is a trade designed to reduce risk and the uncertainty of future cash flows35.
An example is a construction company who wants to buy a given quantity of oil in the future but does not know what the actual spot price will be upon delivery. The buyer therefore faces the risk of significant price increases. Through buying futures contracts at the same time as the buyer contract for oil, while making an offsetting sale around the time the oil is delivered, a risk averse buyer can use a hedge strategy to “lock in” the oil price. In case of an oil spot price increase the buyer will take a loss on the physical transaction, but in theory the price of
27 Grinblatt, 2002, p.702, 710-711 28 Hull, 2003, p.700
29 Ibid., 2003, p.705 30 Ibid., 2003, p.125
31 For an illustration of futures price formation see Appendix 2 32 Brealey, 2003, p.1044
33 Hull, 2003, p.706 34 Ibid., p.36 35 Hull, 2003, p.70
the futures contracts should rise in line with the physical commodity. A sale of the future contracts will lead to a compensating gain. A perfect hedge is when the gain on selling the futures contracts completely offsets the loss on the physical transaction and the risk as a result is entirely eliminated36. It is more or less impossible to create a perfect hedge in reality, due to basis risk, credit risk, commissions and other transaction costs. Basis risk37 will almost always be present when using futures contracts, perhaps resulting in undesirable results. The aim for the hedger is therefore to create hedges that perform as close to perfect as possible. It is usually much better to be hedged than being at risk for the entire quantity of oil38.
A long hedge is one that involves taking a long position (i.e. buying) in a financial derivative, for instance a futures contract. It is suitable for a company that wants to lock in a price now for a specific asset which it knows it will have to buy in the future. It can also be used to partly offset an earlier entered short position. On the contrary, a short hedge refers to a short position (i.e. selling) in futures contracts. It is suitable when a company expects to sell an asset which it already owns at a specific point in time in the future. It can also be used when an asset is not owned right now but will be owned at some time in the future.
2.3 The Futures Market
2.3.1 The Futures Market in Practice
The pricing opportunities for oil related products widen with the existence of future markets. The futures market is suggested to have the following functions;
• Makes it possible to reallocate risk.
• To collect and distribute information about the market’s judgement of future prices.
• To reallocate venture capital.
• Makes delivery of the underlying asset possible.
The success of a futures market is depending on how well the defined criteria can be satisfied. The commodity has to be traded in large quantities. It should also be homogenous, i.e. no quality differences should exist. Production and consumption should be widely distributed. Trade should take place at an organized exchange with the function similar to an auction market, and the physical commodity should be purchased and sold in a way that causes its price to be volatile in a random or non-systematic manner39.
Brokers and clearing-houses are necessary to make the futures market work. The clearing-house acts as a seller to buyers, and buyer to sellers, i.e. its primary function is to perform as an intermediary in the transfer between traders. Among other things, the clearing-house contributes to the depersonalization of transactions. The individual traders do not have to gain insight about the financial strength of other traders since they are actually doing businesses with the clearing-house and not individuals. It should be emphasized that only brokers belong to the clearing-house and it is their accounts that are settled by this institution. Likewise, individual brokerage firms act as a clearing function for their clients4041.
36 Hull, p.10-14; Geman, 2005, p.6-8 37 Ibid., 2003, p. 75
38 Ibid., 2003, p.70
39 Yazdanfar, 2003, p.54-56
40 For an illustration see Appendix 3
2.3.2 Basis Risk
Basis is the differential that exists between the spot price of a given commodity and the price of the nearest futures contract for the same, or a related commodity. It can also be described as the price difference between the physical and financial market. The forecasting ability and the size of the basis can involve four price relationships;
• The difference between the futures contract and the spot price of the underlying commodity.
• The difference between the price at the futures contract delivery point and the price at a different location.
• The price at the futures contract delivery point and the price of a similar, but not identical, quality commodity at the same location.
• The basis can be minimized if delivery is made and accepted at the same time as the futures contract nears expiration.
Different companies are exposed to different basis risks. Firms that want to hedge but do not make/take deliveries at the futures contract location faces the risk of locational basis. In the theoretical framework, the price differences between two markets will be based upon the transportation costs between them. Rapid changes in supply and demand on the local market can affect and distort this price relationship. The predictability of these changes in market conditions affect the ability to what extent a firm can hedge its exposure towards locational basis risk.
Another kind of risk is the product basis. Firms that want to hedge a purchase or sale of a particular commodity not offered as a liquid futures contract are exposed to this risk. They have to base their hedge on historical values of the relationship between the commodity underlying the contract to the commodity to be hedged.
A short hedger (a seller of futures) may face a loss if the basis widen during the time the hedge is held. This is because the spot prices have fallen more or risen less then futures prices. In a decreasing market the short hedger’s spot loss would exceed a gain on the short futures transaction. On the contrary the spot gain in an increasing market would be exceeded by the loss on the futures transaction. In contrast, for a long hedger (a buyer of futures) a widening basis would experience a gain because of the futures price in relation to the spot price. In summary, a narrowing basis results in gains for the short hedger and losses for the long hedger42. This is summarized in the following table.
Table 2.2 Potential basis changes43
2.4 The Swap Market
2.4.1 The Swap Market in Practice
As earlier mentioned a swap is an agreement to exchange cash flows in the future according to prearranged terms. Commodity swaps are cash settled and do not involve any physical delivery of the underlying commodity44. In comparison to futures contracts, swaps are traded
Over The Counter (OTC), which is a telephone- and computer-linked network of dealers who do not physically meet. Trades are done over the phone and usually between two financial institutions or between financial institution and one of its corporate clients45. Therefore it must be at least two parties in a swap and they are sometimes labelled as party and counterparty. They may arrange the swap directly with each other or indirectly. In the latter case, a bank or financial institution acts as an intermediary46. Swaps are customized transactions and perfectly suited for hedging activities47.
Swaps are a generalization of forward contracts. The buyer of a swap makes periodic payments to the seller of the swap at a fixed price per unit for a given notional quantity of a commodity. The seller pays the buyer an agreed upon floating price (reference price from Platts) for a given notional quantity of the commodity underlying the swap, where the underlying commodities are usually the same48. The fixed price of a swap is set at the beginning of each period and paid at the end of the period. It is derived from the yield curve of the underlying commodity49. Therefore, the most significant factor in determining the price of a swap is the term structure of forward prices in the market (i.e. yield curves of oil) that is being used by swap participants. For the derivation of each payment of the floating price, the existence of a reliable index is critical. The floating price is usually defined as the market price or an average market price, the average being calculated using spot commodity prices over some predefined period50.
43 Webpage nymex.com 44 Geman, 2005, p. 283-284 45 Hull, 2003, p. 2 46 Ibid., p. 129 47 Geman, 2005, p.284 48 Ibid., p.210 49 PriceWaterouseCoopers, 2005 p. 52 50 Geman, 2005, p.284
Rising Market Falling Market
Spot/Futures Position Spot Rises Less Than Futures Spot Rises More Than Futures Spot Falls Less Than Market Spot Falls More Than Futures Bought the spot/sold the
futures Loss Gain Gain Loss
Sold the spot/bought
The main difference between futures and swaps is that a swap contract can offer a single fixed price for an entire period while a portfolio of futures contracts offers a sequence of different prices for each delivery month51. Although, it can be shown theoretically, that it is possible for all futures prices to be the same, it is more likely that futures will be in an upward (contango52) or downward (backwardation53) trend. The relative value of a swap contract compared to the portfolio of futures contracts will therefore depend on the slope (i.e. trend) of futures prices and whether the swap participant in question is hedging a short or long position on the physical oil market.
2.4.2 Benefits Versus Risks
The attraction of swaps is four-fold. Firstly, there are no costs for buying/selling swaps. Secondly, they are straight forward financial transactions and can thereby be traded without incurring specific quality risks and other related delivery problems of the underlying commodity, which is normally associated with physical oil contracts. Thirdly, since swaps can be tailored exactly to meet the demands and requirements of each participant, they can really offer the possibility a perfect hedge. Fourthly, and finally, swaps are not constrained by the more or less limited term-structures of the prevailing forward and futures market and can therefore be traded far into the future. A conclusion from the above stated is that swaps with its unique tailoring and flexible abilities are well suited for filling the gaps left out by other traded derivatives in the oil market. As a result, business involving oil swaps has increased rapidly over the past few years54.
The buyer of the swap reduces its exposure to the volatile commodity prices in the markets. It still, however, bears some risk. This is because there may be a difference between the spot price and the average spot price. The buyer is still purchasing the commodity and paying the spot price, and from the seller it receives the last month’s average spot price. Another risk associated with private arrangements between two parties is the credit risk. There is a possibility that one party will get into financial difficulties and default. That would affect the intermediary negatively as it still has to honour the contract it has with the other party. Moreover, liquidity issues such as getting out of the agreement or selling one side of the contract are frequently encountered problems55.
2.5 IAS 39
When using financial derivatives, IAS 39 should be applied by all enterprises. If a financial asset instrument has been taken out to act as a hedge, hedge accounting rules could be followed if certain criteria are fulfilled56. IAS 39 requires two kinds of effectiveness tests57;
• A prospective effectiveness test – a forward-looking test of whether a hedging relationship is expected to be highly effective in future periods. It is required, at a minimum, at the inception of the hedge and at the time an entity prepares its interim or
51 Long, 2000, Supplement 3 52 For an explanation see Appendix 2 53 Ibid.
54 Long, 2000, Supplement 3 55 Hull, 2003, p. 145
56 Ibid., p.343
annual financial statements58. If this test is not passed, hedge accounting must be discontinued prospectively59.
• A retrospective effectiveness test – a backward-looking test of whether a hedging relationship has actually been highly effective60, where actual results are within a range of 80% to 125%61, in a past period. It is required, at a minimum, at the time an entity prepares its interim or annual financial statements62.
Any ineffectiveness, i.e. gain or loss arising including ineffectiveness within the 80% to 125% interval, on re-measuring the hedging instrument and the hedged item should be recognised in the income statement in the period63. This might cause fluctuations of major significance on the income statement.
Hedge accounting seeks to reflect the results of hedging activities by reporting the effects of the derivative and the risk being hedged in the same period as offsetting losses and gains. It seeks to match the timing of gain and loss recognition by changing the timing of recognition of gains and losses on either the hedged item or the hedging instrument. This avoids much of the volatility that would arise if the derivative gains and losses were recognised in the income statement, as required by normal accounting principles64.
2.6 Earlier Research
Earlier research in Sweden within the field to investigate the effectiveness in using commodity futures contracts to hedge against commodity price exposures from an economic perspective seem to be of a rather limited amount.
A master thesis arguing about whether an active hedging strategy could improve the profitability for a company with a high commodity exposure or not, has been written at Lund University. It has chosen the aluminium company Rexam as a case study to analyse whether an improvement of its aluminium costs by using forwards or options could be statistically and economically motivated.
At Jönköping University a Bachelor thesis has been written about the volatility associated with commodities and the optimal hedge ratio for copper, gold, oil and cotton. The study aims to conduct an analysis of the variance in different commodities contracts and provide evidence of the optimal hedge ratio in the respective commodities.
When it comes to IAS 39, there are several qualitative Bachelor and Master Thesis discussing its impact on companies65.
When taking a look outside of Sweden, the range of earlier hedge effectiveness research is much wider, especially from the US. For example, Sparks Companies 2001 made an analysis of hedge effectiveness for the new cash-settled corn and soybean futures contracts
58 PriceWaterhouseCoopers, 2005, p.14 59 Ibid. p.48 60 Ibid., p.14 61 Webpage deloitte.com 62 PriceWaterhouseCoopers, 2005, p.14 63 Webpage deloitte.com 64 PriceWaterhouseCoopers, 2005, p.7-9 65 Göthlin, 2005; Hallén, 2005; Hansson, 2004
that were to begin to be traded at the Minneapolis Grain Exchange (MGEX) in February 200266.
Another example is an article about optimal hedging in futures markets with multiple delivery specifications, published in the Journal of Finance. Optimal hedging strategies in futures markets allowing delivery for more than one quality of the underlying asset were derived. The effectiveness of the optimal hedging strategies was then compared with a full hedge and with a no-hedge strategy67.
66 Sparks Companies, 2001 67 Kamara, 1987
3 Method
3.1 Futures Contracts Hedge
3.1.1 Cross Hedge
As earlier mentioned, it is not always possible to buy or sell futures contracts of exactly the same commodity or financial instrument as the entity being hedged. For instance, a company that forecasts a need to purchase diesel in the future might decide to buy gas oil futures to try to lock in the price, as diesel futures contracts are not traded on the commodity exchanges. This make it necessary to carry out with cross hedging, a technique for hedging an asset with a derivative contract which has a different but related underlying asset than the one being hedged. This includes calculating the hedge ratio h*, i.e. the ratio of the size of the position taken in futures contracts to the size of the exposure. If the objective of the hedger is to minimize risk, setting the hedge ratio equal to 1,0 is not necessarily optimal. To derive the optimal hedge ratio, a regression between the changes in historical spot prices against the changes in historical futures prices for the same period needs to be done. A number of historical time points will be selected and the changes in the spot/futures prices will be observed and analysed. Depending on the length of the hedge it is optimal to choose a time interval that is consistent with the time for which the hedge is in effect. This provides the necessary input needed for calculating the optimal number of futures contracts needed for the hedge68.
In order to evaluate whether it is possible to create optimal hedges to protect a construction company against its bitumen, i.e. fuel oil 3,5% (QHFO-ARA), and diesel, i.e. fuel oil 50 ppm (QUSDL50-C-NWE), price exposures; Brent futures (CL 3M69) will be used as a proxy for bitumen and gas oil futures (QS 3M) will be used as a proxy for diesel, as there are no outstanding futures contracts on bitumen or diesel. In order to try to lock in the price fully, it would be necessary to buy futures contracts all at once that expire at the different points in time at which commodity purchases are planned to be carried out, for example in 6 months, 1 year, 1.5 years and so on70. The reason for only selecting futures contracts stretching over 3 months is that bitumen and diesel contracts for longer periods are not liquid enough or not even available on the commodity exchanges. Instead, a new hedge will be entered into every 3 months, with the result of continuously locking in the price 3 months ahead.
3.1.2 Optimal Number of Contracts
To derive the number of futures contracts that would be optimal to enter into, the different quantities of total future bitumen and diesel purchases with hence taken to Skanska’s Polish A1-project need to be estimated. This has been done by Skanska. By dividing each of the estimated figures for total future bitumen and diesel spends by the remaining 3 years of construction until completion71, the annual quantity expected to be consumed can be derived for each commodity.
68 Hull, 2003, p.78-80 69 M = months
70 Webpage aerweb.com
The purchases are expected to be made at the beginning of each quarter. Due to seasonal reasons72, 10% of the costs are expected to be consumed during Q1 and Q4 respectively, and 40% of the costs during Q2 and Q3 respectively.
The size of the futures contracts is of 1000 barrels for Brent73 and 100 metric tons for gas oil74. From this information it is possible to calculate the optimal number of futures contracts needed to be bought in order to carry out with the hedge.
3.1.3 Retrospective Test
The most basic test for hedge effectiveness is the dollar-offset method. Implementation of the method consists of dividing the observed values of the commodity purchase value being hedged, by the corresponding values of the hedging instrument. The ratios are compared to the effectiveness intervals stated by IAS 39 (i.e. 80%-125%). IAS 39 both permits the dollar-offset ratio to be calculated by dividing the absolute period-to-period change in the purchase value by the absolute period-to-period hedge return, and on a cumulative basis. In most cases, the last mentioned is the most favourable for the reason that temporary discrepancies between the hedged item and its hedge may be absorbed. In this thesis, the dollar-offset test on a monthly basis will be used for the retrospective test, both on a period-to-period basis and on a cumulative basis. However, in real life IAS 39 requires the choice between those two alternatives to be specified before initiation of the hedge75. The ratios will both be derived by dividing the commodity purchase value and the value of the hedging instrument stated in US dollars, but also converted into the Polish currency zloty, as this is the currency of interest for Skanska’s A1-project in Poland. In the latter case, the volatility of the Polish currency will also have an impact on the effectiveness of the hedge. Finally, the basis will be calculated for the different testing time points.
3.1.4 Prospective Test
The dollar-offset method can easily disqualify a high quality hedge because of uncharacteristic performance in a single test period. However, IAS 39 also permits alternatives such as statistical tests, like regression analysis. Statistical testing for hedge effectiveness is based on valuation of the likelihood that the hedge will be effective in the future. The underlying assumption is that the statistical relation valid in the past will also hold for the future. Even though the dollar-offset method shows that the hedge is ineffective, the regression analysis might allow hedge accounting for the current period. If a regression indicates that the hedge can be expected to be effective in future periods, a company can continue to apply hedge accounting irrespective of the outcome of the dollar-offset test in the current period76. In this thesis, the prospective test will constitute of this kind of statistical test, represented by a rolling regression, i.e. three new values will continuously be added as input for each new regression at the same time as three historical values are excluded77.
72 Thomas Wieland, Skanska, 2006 73 Webpage theice.com
74 Webpage platts.com 75 Reznek, 2005, p.2 76 Ibid., 2005, p.3
3.1.5 Hypothesis Test
To evaluate whether the hedging relationship is expected to be highly effective or not in future periods, a hypothesis test of the coefficients generated by the prospective tests will be made to see if the average gain or loss achieved is within the 80%-125% interval based upon the criteria of IAS 39. In case they are, a backward-looking test of whether the hedging relationships have actually been highly effective or not, i.e. within the 80%-125% interval, during the past period will also be carried out. The tests will influence the decision about whether a company should implement a hedging strategy to protect themselves towards price fluctuations in the commodity market or not.
3.1.6 Futures Hedge Data
Monthly data of prices for Brent futures (CL 3M) and gas oil futures (QS 3M) is collected from Bloomberg. Spot prices data for bitumen, i.e. fuel oil 3,5% (QHFO-ARA), and diesel, i.e. fuel oil 50 ppm (QULSD50-C-NWE), is collected from Reuters data base on a monthly basis. In order to get enough scope for a valid prospective test, data of more than 12 observations and a data series covering a longer history, for example one year, are recommended for each regression78. To be able to evaluate the effectiveness of futures hedges during a longer time horizon, a bitumen hedge is assumed to be initiated 2001-12-31. The construction at Skanska’s A1-project in Poland is, as earlier mentioned, expected to continue during 3 more years79. Brent futures prices and bitumen spot prices data is therefore collected on a monthly basis from 1999-01-31 to 2006-04-30, making it possible to base each prospective test on 36 observations (to fulfil the earlier mentioned criteria of at least 12 observations) covering three historical years (to correspond to the remaining time of during which the Polish A1-project will be under construction). Gas oil futures prices and diesel spot prices data is collected on a monthly basis between 2002-09-30 to 2006-04-30, for the reason that fuel oil 50 ppm only is available from that date. To be able to base each prospective test on 36 observations, the diesel hedge is assumed to be initiated 2005-09-30.
3.1.7 Currency Exchange Rate Data
The currency exchange rate spot data for the retrospective tests of the futures contract hedges has been collected on a monthly basis from Reuters database between 2002-09-30 and 2006-04-30.
Forward exchange rates on 1M, 3M, 6M, 9M and 1Y has also been collected from Reuters database between 2002-09-30 to 2006-04-30.
78 Reznek, 2005, p.4
3.2 Swap Hedge
An investigation of bitumen and diesel swap hedges covering 1Y80, 2Y and 3Y will be carried out, based on the fact that projects that construction companies, like Skanska, undertake often stretch over longer time periods. The swap hedge formation can be separated into three different steps;
3.2.1 Swap Hedge Price Data
Bitumen and diesel swap prices for 1M, 3M, 3Q and 4Q will be used in this thesis. Bitumen spot and swap quotes data, i.e. fuel oil 3,5% (QHFO-ARA), for 1M, 3M, 3Q and Q4 will be collected from Reuters database. Diesel swap quotes data, i.e. fuel oil 50 ppm, is traded as a forward premium (QGO-50P) added to the diesel spot price (QULSD50-C-NWE). The spot prices and the forward premiums for 1M, 3M, 3Q and Q4 are collected from Reuters. All data is collected on a monthly basis between 2002-09-30 and 2006-05-31.
Due to the fact that bitumen and diesel swap prices are only available for 1M, 3M, 3Q and 4Q, it is necessary to create swap price quotes for 1Y, 2Y and 3Y bitumen and diesel swaps. From Nordea bank, forward prices stretching over 1M, 6M, 1Y, 2Y and 3Y for diesel and bitumen are received 2006-05-24. The fixed 1Y, 2Y and 3Y swap prices for diesel and bitumen on 2006-05-31 can mathematically be derived from the received forward prices. The derived prices will be compared and divided by the outstanding spot prices on the same dates. Then, the fixed curvature for the future prices will be applied throughout the database. The discount or premium of the outstanding spot prices for different maturities will be multiplied backwards with the spot prices between 2002-09-30 and 2006-05-31 (i.e. by using the fixed curvature for future prices 2006-05-31). The prices received have a very good correlation and R2 towards the actual short-term forward curve (about 99%). The swap quotes derived will therefore be based on what the yield curve looks like 2006-05-31.
3.2.2 Swap Cases
The first hedge construction will consider the effect of applying bulk volume, i.e. the volume that is left to consume at the end of the hedge has been considered. The hedge will consist of a hedge for one year at a time. The spot price used for pricing the first hedge initiated will be used also for the second and third hedges. However, the forward premiums for each of the hedges will correspond to the time point at which each hedge is initiated. The exchange rate (zl/$) is locked in for the entire period. The hedge will be initiated 2002-10-31 and closed out 2005-09-30. In order to analyse possible timing issues, the same swap hedge will be assumed to be entered into at two new time points set to 2003-01-31 and 2003-04-30. Similarly to the futures contract hedges, the volume will be calculated with consideration taken to Skanska’s forecasted consumption for the remaining construction time of the project.
The second hedge construction will consist of a hedge for one year at a time. The first 1Y swap will be initiated 2002-10-31 and closed out 2003-09-30. The second 1Y swap will be started 2003-10-31 and closed out 2004-09-30. The third and final 1Y swap will be initiated 2004-10-31 and closed out 2005-09-30. In order to analyse possible timing issues, the same swap hedge will be assumed to be entered into at two new time points set to 2003-01-31 and 2003-04-30. The exchange rate (zl/$) will be locked in (hedged) on a period-to-period basis (i.e. a new currency hedge for every new swap initiated) over the whole swap period, i.e. 3 years, in order to capture the currency effect of the hedge.
The third hedge construction will consist of one 1Y swap and one 2Y swap. The first 1Y swap will be initiated 2002-10-31 and be closed out 2003-09-30. The final and second 2Y swap will be started 2003-10-31 and closed out 2005-09-30. In order to analyse possible timing issues, the same swap hedge will be assumed to be entered into at two new time points set to 2003-01-31 and 2003-04-30. The exchange rate (zl/$) will be locked in for first one year and then two years ahead.
In total, three different cases will be carried out for each of the bitumen and diesel swap hedges. This is summarized in the following table.
Time Swap Case Specification
2002-10-31 - 2005-09-30 1 Hedging with 1 year swap (Bulk - Exchange rate fixed for 3 years)
2003-01-31 - 2005-12-31 1 ― װ ―
2003-04-30 - 2006-03-31 1 ― װ ―
2002-10-31 - 2005-09-30 2 Hedging with 1 year swaps (Exchange rate fixed for 1 year ahead)
2003-01-31 - 2005-12-31 2 ― װ ―
2003-04-30 - 2006-03-31 2 ― װ ―
2002-10-31 - 2005-09-30 3 Hedging with 1 and 2 year swap (Exchange rate fixed for first 1 year then 2 year ahead)
2003-01-31 - 2005-12-31 3 ― װ ―
2003-04-30 - 2006-03-31 3 ― װ ―
Table 3.1 Illustration of three different swap cases
3.2.3 Swap Evaluation
The swaps will be evaluated by calculating the following differential for each month throughout the hedging period;
fixed swap price * fixed exchange rate * volume - spot price * spot exchange rate * volume
When presenting the results of the evaluation, the part of the equation on the right side of the subtraction sign will be called “No Hedge” while the part of the equation on the left side of the subtraction sign will be named “Hedge”. The results from the different swaps will be presented as time series in diagrams, where the zloty costs from the hedge are accumulated over the entire hedge period. An analysis of the different outcomes illustrated in the diagrams will represent the decision criteria for whether it is preferable or not to hedge bitumen and diesel exposures by using swaps.
A backward-looking retrospective zloty-offset test will also be carried out with the same approach as for the futures contracts81.
3.2.4 Currency Exchange Rate Data
The currency exchange rate spot data for the swap hedges has been collected on a monthly basis from Reuters database between 2002-09-30 and 2006-05-31.
Forward exchange rates on 1M, 3M, 6M, 9M and 1Y has also been collected from Reuters database between 2002-09-30-2006-05-31.
81 Skanska will only consider initiating a swap hedge based on the effectiveness from an economic perspective.
3.3 Models (explicit for futures)
3.3.1 Hedge Ratio
The following notations will be used;
=
∆S Change in the spot price, S, during a period of time equal to the life of the hedge
=
∆F Change in the futures price, F, during a period of time equal to the life of the hedge = S σ Standard deviation of ∆S = F σ Standard deviation of ∆F =
ρ Coefficient of correlation between ∆Sand ∆F =
*
h Hedge ratio that minimizes the variance of the hedger’s position
The optimal hedge ratio is the product of the coefficient of correlation between ∆S and ∆F and the ratio of the standard deviation of ∆S to the standard deviation of ∆F;
The optimal hedge ratio, h*, is the slope of the best fit line when ∆S is regressed82 against
∆F. The hedge effectiveness can be defined as the proportion of the variance that is eliminated by hedging. This isρ2, or 2 2 2 * S F h σ σ =
The parametersρ, σF and σS are usually estimated from historical data on ∆Sand∆F. The implicit assumption is that the future will in some sense be like the history. The number of equal non-overlapping time intervals are selected and the values of ∆Sand ∆Ffor each of the intervals are observed. It is optimal that the length of each time interval is the same as the length of the time interval for which the hedge is in effect. In practice, this sometimes severely limits the number of observations that are available and a shorter time interval is used83.
3.3.2 Optimal Number of Contracts
The following notations will be used;
=
A
N Size of position being hedged (volume)
=
F
Q Size of one futures contract (volume)
=
*
N Optimal number of contracts for hedging
82 Regression analysis is in statistics a technique for finding the line of best fit. 83 Hull, 2003, p.79 F S h σ σ ρ = *
The futures contracts used should have a nominal value ofh *NA. The number of futures contracts required is therefore given by84;
F A Q N h N*= * 3.3.3 Basis Risk
The basis risk, which is fundamental to understanding hedging, is defined as;
) ( , S F t Basis T t T t = − where = t
S the spot price when the hedge is being initiated at time t
=
) (t
FT the futures price at time t, where T is the time to maturity of the futures contract
and the time when the hedge is closed out.
The basis is usually quoted as a discount or premium; the spot price as a discount or premium to the futures price85. As the market participants analyse their risk in a mark-to-market perspective at date t and not only a date T, the basis risk is often defined as the variance86 of the basis; )) ( ( ) ( 2 )) ( ( ) ( )) ( ( 2 2 2 S F t S F t S FT t t T t T t σ σ ρσ σ σ − = + − where
ρ= the correlation of coefficient between the futures and spot price series. The equation shows that basis risk is zero when;
1. variances between the futures and spot prices are identical and
2. the correlation coefficient ρ between spot and futures price is equal to one.
In practice, the second condition is the most stringent one and the magnitude of basis risk depends mainly on the degree of correlation between spot and futures prices87.
84 Hull, 2003, p.80 85 Ibid., p.14
86 The variance is the mean squared deviation from the expected value, i.e. a measure of variability. 87 Geman, 2005, p.14
3.4 Generalization, Reliability, Validity and Criticism of Sources
The conclusions drawn from this investigation can be applied by other companies than construction companies like Skanska, with an exposure towards the same underlying commodities. It is also of relevance for all companies listed on a stock exchange in an EU country that as a result need to comply with International Accounting Standards (IAS) since 1 January 2005. As a consequence, the degree of generalization of the results can be considered as high.
Reliability is the question of whether a method generates the same results irrespective of whom the investigatoris and the time of the survey88. When it comes to the futures contracts hedge evaluation, the measuring method used is well suited for the purpose of the investigation and recommended by well-known accounting companies89. Moreover, it is easy to implement for a practician, which arguments for that the reliability of this thesis is good. Though, the reliability of the swap hedges tests could be questioned due to the fact that the created data series are based on a number of subjective assumptions. What advocates for a high reliability, is that the swap hedge method is developed in cooperation with Skanska and Nordea. It means that results generated by other researchers with a similar approach and the same access to data should not differ significantly in comparison to this study.
An investigation of the validity of the study constitutes of whether the methods used in the study measure what they are supposed to measure90. The thesis can be considered to have a high validity, as the methods used are highly accepted as testing instruments for these kinds of investigations. By choosing input variables and parameters carefully, flaws associated with these kinds of measuring methods can be minimized.
The data sources from which the data has been collected, i.e. Reuters and Bloomberg, are used by many different and well-established market actors, and can therefore be considered as reliable.
88 Gustavsson, 2003 p.55
89 Deloitte, PriceWaterHouseCoopers 90 Gustavsson, 2003, p.62
4 Results and Analysis
4.1 Bitumen Futures Hedge
A graphical look at the bitumen data91 shows that the spot prices and the futures prices follow each other pretty good, but with an initial spread which increases significantly in the beginning of 2004 and continues to widen throughout the rest of the time series92. The graph suggests that the hedge might be ineffective. Clear and substantial variations over significant time periods indicate a volatile hedge performance. The graphical comparison often gives a good hint about the hedge performance, but it does not provide a statistical test of effectiveness. The bitumen data generates 18 different regressions for the prospective test93.
Regression - Prospective Test
β = 1,36 R2 = 99 % 5 10 15 20 25 30 35 40 5 10 15 20 25 30 QHFO-ARA ($/bbl)
Brent Futures 3M ($/bbl) Brent Futures 3M
Linear (Brent Futures 3M)
Graph 4.1 QHFO-ARA – Example of prospective test 2002-09-30
According to the hypothesis tests94, all of the 18 derived coefficients pass the lower test criteria, i.e. the null hypothesis is rejected. However, none of them pass the upper test criteria, i.e. the null hypothesis is accepted. The hypothesis tests are illustrated by the following graph;
91 For an illustration see Appendix 4
92 For an illustration of the historical price development of Brent futures 3M see Appendix 5 93 For an illustration see Appendix 6
Prospective Test - QHFO-ARA 0,6 0,7 0,8 0,9 1,0 1,1 1,2 1,3 1,4 1,5 1,6 12-0 1-20 01 03-0 1-20 02 06-0 1-20 02 09-0 1-20 02 12-0 1-200 2 03-0 1-200 3 06-0 1-20 03 09-0 1-200 3 12-0 1-20 03 03-0 1-20 04 06-0 1-200 4 09-0 1-20 04 12-0 1-200 4 03-0 1-200 5 06-0 1-20 05 09-0 1-20 05 12-0 1-200 5 03-0 1-20 06 Coefficient β Upper Boundary Low er Boundary
Graph 4.2 QHFO-ARA – Prospective test for the entire period
As a result, the hedge can not be expected to be effective in the future for none of the tests. For this reason, there is no meaning in entering a hedge and to continuously evaluate it retrospectively.
4.2 Diesel Futures Hedge
A graphical look at the diesel data95 shows that the spot prices and the futures prices track each other very well96. The graph suggests that the hedge might be effective for some specific time periods. The diesel data generates three different regressions for the prospective test97. According to the hypothesis tests, all of the three derived coefficients pass both the lower and the upper test criteria, i.e. in both cases the null hypothesis is rejected.
95 For an illustration see Appendix 4
96 For an illustration of the historical price development of gas oil futures 3M see Appendix 5 97 For an illustration see Appendix 6
Table 4.1 QULSD50-C-NWE versus gas oil futures 3M
As a result, the hedge can be expected to be effective in the future for all of the tests. For this reason, it makes sense to enter a hedge and to continuously evaluate its performance retrospectively.
The retrospective test98 passes the IAS 39 criteria only two times out of seven on a period-to-period basis, with results ranging between 0,69-8,36. On a cumulative basis all of the ratios fall outside of the interval, with 0,52 as the lowest value and 2,03 as the highest value. The narrower value interval for the cumulative method is in line with the earlier comment that the cumulative method often smoothes out temporary discrepancies, in comparison to the period-to-period test. According to the actual results, the hedges can not be considered as effective within the framework of IAS 39. The explanation for the ineffectiveness is the big variations of the basis, ranging between -0,3 and 28. In order for the hedge to be effective, the basis has to stay fairly constant throughout the hedge99.
The reasons for the basis fluctuations could be the difference between the price at the futures contract delivery point and the spot price different location. Skanska purchases the diesel in Poland, at the same time as the gas oil futures contracts offer delivery in the Netherlands in Amsterdam, Rotterdam or Antwerp. Transportation costs for delivery between the local areas and the specific futures contracts’ delivery point might result in price differences.
It could also be because of the fact that diesel futures are not offered as a liquid futures contract on the market at present, and gas oil futures contracts instead have been used as a proxy for the hedge. Even though gas oil and fuel oil 50 ppm are closely linked products, their qualities are not exactly the same. This should obviously affect the pricing between the two products.
Moreover, the basis might also be explained by the time difference between the diesel purchases and the expiration of the futures contracts. The timing effect could create sever discrepancies in some months when the future contract is closed out several days before taking delivery on the spot prices.
When looking at the retrospective test when its values are converted into zloty100, it can be seen that the IAS 39 criteria is passed only two times out of seven on a period-to-period basis, with results ranging between -4,30 to 2,16. 2005-12-31 the ratio is negative, which is explained by an appreciation of the zloty/dollar exchange rate. Therefore, it
98 For an illustration see Appendix 8
99 For an illustration of the relationship between the futures prices, the spot prices and the basis; see Appendix 9 100 For an illustration see Appendix 8
QULSD50-C-NWE Versus Gas Oil Futures 3M Conditional On True Beta Equal to 0,8
Coefficient Observed T-Value Critical T-Value Result
09/30/2005 0,89 8,40 2,42 Reject Ho
12/31/2005 0,90 6,92 2,42 Reject Ho
03/31/2006 0,92 11,93 2,42 Reject Ho
Conditional On True Beta Equal to 1,25
Coefficient Observed T-Value Critical T-Value Result
09/30/2005 0,89 -34,76 -2,42 Reject Ho 12/31/2005 0,90 -24,28 -2,42 Reject Ho 03/31/2006 0,92 -34,21 -2,42 Reject Ho ∧ β ∧ β
is of significant importance to consider the risk of currency fluctuations when entering into a commodity hedge, as it might have a major impact on its effectiveness. On a cumulative basis, none of the tests are passed. If the currency is locked in by using forward contracts101, the results are similar to the dollar-denominated test but the different values are far better than for the test with the unhedged currency risk.
From an economic perspective, all of the hedges with a dollar/zloty-offset ratio less than 1 can be regarded as a gain, i.e. the gain on the futures position is higher than the loss on the purchase position, i.e. spot position. This is valid for all of the 3M-hedges initiated 2005-12-31. For all of the hedges entered into 2005-09-30 and 2006-03-31, the gain on the futures position is less than the loss on the purchase position, i.e. the dollar/zloty-offset ratio is above 1. Those hedges can therefore be regarded as a loss from an economic perspective.
Due to the fact that it is only possible to buy futures contracts stretching over 3 months and Skanska’s project stretches over 3 more years, it is questionable whether the effort needed to enter a hedge can be motivated, as only a very small part of the project can be covered from the day it is accepted.
Finally Skanska will be exposed towards the risk of sell/buy spreads every time the futures contract will be rolled over for another 3 month period. The spread might vary from time to time and might cause Skanska extra unnecessary cost.
4.3 Diesel Swap Hedge
In case 1 the results102 show that it would have been preferable from an economic perspective to stay hedged by using a bulk swap stretching over 3 years, as it would have resulted in a lower cost compared to being unhedged.
Case 1
0 5 10 15 20 25 30 10-31 -200 2 12-31 -200 2 02-28 -200 3 04-30 -200 3 06-3 0-20 03 08-3 1-20 03 10-3 1-20 03 12-3 1-20 03 02-2 9-20 04 04-3 0-20 04 06-3 0-20 04 08-3 1-200 4 10-3 1-20 04 12-3 1-20 04 02-2 8-20 05 04-3 0-20 05 06-3 0-20 05 08-3 1-20 05 Time M il li on Zl o tyNo Hedge (Bulk) Hedging with 1 year swaps (Bulk - Exchange rate fixed for 3 years)
Graph 4.3Case 1 - Diesel swap hedge
101 For an illustration see Appendix 8
After the first year the hedge still generates a loss but after approximately 1,5-2 years a gain on the hedge starts to appear. The total gain on the hedge is 3,7 million zloty103, when the hedge is closed out (i.e. 3 years later). Over 3 years the spot price has increased by about 254% and it would certainly be surprising if a 3 year hedge would not profit from this opportunity. The exchange rate is also locked in over the entire period, which dampens the total gain that could have been earned on the hedge. The main reason for this is that the spot exchange rate has appreciated towards the dollar by about 20% during this 3 year period and by locking in the exchange rate from the beginning of the swap it is not possible to capture the gain from positive movements of the exchange rate.
Case 2, i.e. hedging with 1 year swaps, results in a gain as well. The fixed exchange rate for 1 year ahead in addition to the favourable exchange rate movements has a positive impact on the hedge. The total gain on the hedge is approximately 0,7 million zloty104.
The fact that the fixed swap price is updated once a year makes it more adjusted to the spot price movements. This has negative effects on the gain of the hedge in comparison to a swap stretching over 3 years. From an economic perspective both hedges are profitable. Though, the preferred hedge choice would be the 3 year bulk swap with the exchange rate fixed over the entire period. It offers Skanska the opportunity to lock in the price both for the underlying commodity price risk as well as the risk of fluctuating exchange rates over the entire 3 year period. It makes it a good choice when the underlying commodity against which price risk protection is desirable is volatile. What can be better then locking in Skanska’s underlying commodity exposure for a long period and at the same time gain from it? However, it is important to keep in mind that the historical gain from an economic perspective might not be reiterated in the future. Even though a swap initiated in the future will fully lock in the commodity exposure in a similar way, the hedge might result in a loss from an economical perspective due to a different price development in the future. This investigation has been carried out when the forward curve has been an increasing function of maturity, i.e. the situation of contango. If the forward curve would have been a decreasing function of maturity, i.e. the situation of backwardation, the outcome of the hedge evaluation might have been different.
The evaluation of the timing effect on the hedge for case 1 and case 2105 indicates that it is a factor of importance. They clearly illustrate the importance of timing and its economic impact on the hedge. Hedging initiated at 2002-10-31 and 2003-01-31 result in a gain for both of the cases, but hedging initiated 2003-04-30 results in a loss. The question is whether the timing effect motivates or disqualifies the decision to hedge. The future cannot be foreseen and the risk of facing a loss must be weighed against the advantages of knowing the total cost of a large scale project.
In case 3 all of hedges generate positive results106. The timing effect influences the total result but it seems to be of less importance than for the other two cases. The best result occurs when the hedge is initiated 2003-01-31 and closed out 2005-12-31, with a total gain of 4,8 million zloty. This is illustrated in the following graph;
103 For an illustration see Appendix 11 104 Ibid.
105 Ibid. 106 Ibid.
