The time horizon of foreign exchange rate exposure management

Transcription

1 The time horizon of foreign exchange rate exposure management An empirical investigation of the factors influencing the hedging horizon in medium-sized and large, non-financial companies in Scandinavia Af Morten Lindholm Uth og Thomas Enevoldsen Vejleder: Tom Aabo Master of Science in Finance & International Business Aarhus University. Business and Social Science 2011

2 1. INTRODUCTION DELIMITATIONS TERMINOLOGY THESIS STRUCTURE CONCEPTUAL FRAMEWORK THE EXPOSURE OF INTERNATIONAL COMPANIES Sources of risk General exposures Industry specific exposures Firm specific exposures International Parity Conditions Interest rate formation Exchange rate formation Parities and the law of one price Foreign exchange rate exposure Exchange rate exposure, parity conditions and the time horizon HEDGING A RISK MANAGEMENT TOOL The value of corporate risk management Rationales for hedging Risk management techniques Financial hedging Operational hedging and real options Risk management and the time horizon of hedges EMPIRICAL LITERATURE ON THE TIME HORIZON OF FOREIGN EXCHANGE RATE EXPOSURE MANAGEMENT General literature about companies risk management of exchange rate exposure and their use of derivatives Literatures emphasizing the importance of the time horizon in managing exchange rate exposure RESEARCH METHODOLOGY TARGET POPULATION QUESTIONNAIRE DESIGN SURVEY EXECUTION SURVEY RESPONSE RESPONSE BIAS DESCRIPTIVE STATISTICS OF EMPIRICAL DATA COMPANY CHARACTERISTICS Employees...39 i

3 4.1.2 Assets DESCRIPTIVE STATISTICS ON SURVEY SAMPLE QUESTIONNAIRE International involvement and foreign exchange rate exposure Concise conclusion on chapter Debt and derivatives usage for managing foreign exchange rate exposure Concise conclusion on chapter Firm flexibility and industry stability Concise conclusion on chapter REGRESSION ANALYSIS DEFINITION OF VARIABLES & HYPOTHESES Dependent variables Independent variables and hypothetical variables Factor analysis Correlation analysis independent variables Regression model and hypotheses AVERAGE TIME HORIZON OF HEDGING USING FINANCIAL MEANS AVERAGE MATURITY OF FOREIGN DEBT AVERAGE MATURITY OF SHORT-SIGHTED DERIVATIVES AVERAGE MATURITY FOR LONG-SIGHTED DERIVATIVES AVERAGE TIME HORIZON THAT OPERATIONAL MEANS EXCEED THAT OF FINANCIAL MEANS CONCLUSION...83 BIBLIOGRAPHY...86 APPENDIX... I APPENDIX 1: EXPORT SHARE FOR THE SCANDINAVIAN COUNTRIES... I APPENDIX 2: NACE CLASSIFICATION SYSTEM... II APPENDIX 3: COMPLETE QUESTIONNAIRE...III APPENDIX 4: SCREENSHOTS FROM THE STUDSURVEY TOOL...VIII APPENDIX 5: CORRELATION TABLES...XI APPENDIX 6: FACTOR ANALYSIS...XIV ii

4 1. Introduction The economies of the Scandinavian countries are characterized as being small and open with high degrees of foreign trade due to the limited sizes of their domestic markets 1. Thus, foreign exchange rate risk represents a major source of concern for non-financial companies in these countries, due to its ability to influence the prices of inputs and outputs as well as the competitive capabilities of the company in the foreign markets in which it operates. In order to manage these risks, financial managers use different strategies according to the policies within the company. However, the overall impacts of exchange rate fluctuations are extremely difficult to estimate, in terms of impact and length, since much of the exposure derives from expected cash flows not yet on the books. As the magnitude and the length of the exposure is difficult measure it becomes extremely difficult to choose the right timing as well as the maturity of the hedges. Even though financial instruments have become quite sophisticated with a range of different maturities, it is still a difficult task for the companies to completely match the exposure that they have quantified in terms of maturity, with the financial hedge that they can buy. However, shortcomings of the financial derivatives for managing e.g. operational exposure can be mitigated through investments in flexibility whereby the company acquires real options it can use in combination with financial derivatives to manage the overall foreign exchange rate exposure of the company. However, by using a combination of financial and operational means for managing exchange rate exposure, the time horizon for each type of management needs to be carefully considered. Studies on the use of financial derivatives and foreign debt in Scandinavian countries have found that both types of instruments are widely used by firms to hedge foreign exchange rate exposure. 2 Additional research shows, that even though the use of currency derivatives is often the preferred choice when managing foreign exchange rate exposure, firms are becoming increasingly aware of the benefits of strategic approaches such as the use of real options for managing long-term exchange rate exposures. 3 Moreover, the interrelationship between financial hedges and operational hedges known as integrated risk management has drawn the attention of an increasing number of researchers. 4 1 See appendix 1 2 See for example Aabo (2006), Hagelin (2003), Pramborg (2005) 3 Aabo, Simkins (2003), Triantis (2002). 4 See for example Meulbroek (2002) 1

5 While much has been written on the subject of risk management and the extend to which firms use derivatives to manage exchange rate exposure, very little research has focused on the time perspective of these hedges. However, Srinivasulu (1981) has discussed the time horizon of hedging and defines the time horizon of exchange rate operating exposure as follows: the time required for the restructuring of operations through such means as changing products, markets, sources and technology What Srinivasulu argues is that the time horizon for a company s exchange rate exposure is the time it would take the company to exercise its real options. This view allows a company to quantify its exposure, and hedge it using financial means, but it assumes that the company has made investments in flexibility. This serves to highlight an interesting connection between flexibility and the time horizon for which firms choose to hedge their exchange rate exposure which will be discussed in this thesis. Based on an empirical study of medium-sized and large, non-financial firms in Scandinavia, the objective of the present thesis is to investigate factors influencing the time horizon of foreign exchange rate exposure hedges in small open economies. The primary focus is on the role of operational flexibility, but the influence of other factors will also be examined. This research contributes to the existing literature on corporate risk management in two ways. Firstly, it provides empirical evidence on the actual hedging practices, in terms of the actual use of risk management instruments as well as the horizon for these instruments in Scandinavian, non-financial firms. Secondly, this research analyses factors that lie behind the determination of the time horizon for which firms decide to hedge exchange rate exposures as well as the relationship between financial hedges and operational hedges in terms of managing foreign exchange rate exposure. Thus, the central research questions of the present thesis are as follows: 1. To provide actual empirical evidence about the length of the hedging horizon in Scandinavian, medium-sized and large, non-financial companies when hedging their foreign exchange rate exposure using financial means; the extend to which financial means are used as well as the relationship between financial and operational means for exchange rate exposure management. 2. By means of regression analysis to investigate the factors that influence the length of the hedging horizon of financial hedges as well as on the horizon at which the 2

6 benefits of operational hedges exceeds that of financial hedges for managing exchange rate exposure. 1.1 Delimitations In order to create a focused, concise and goal-orientated thesis, choices have to be made in terms of the scope of the thesis as well as the theories used, which inevitably involves tradeoffs. The following section clarifies the delimitations of the thesis and identifies which assumptions the authors have made in a process of composing this thesis. Theoretical delimitations This thesis is placed within the overall frame of corporate risk management with the scope being hedging of exchange rate exposure in non-financial companies. This means that other risks such as interest rate- and commodity price risk will not be considered. Furthermore, the thesis focuses on the time perspective of companies management of their exchange rate exposure and how the interrelationship between the use of financial derivatives and operational flexibility influences this horizon. As earlier empirical literature regarding the time horizon of exchange rate exposure management is rather limited, the conceptual framework is considered in a broader scope than merely the time horizon of companies hedging activities. In this way, the reader will be able gain an overall understanding of hedging and the underlying rationale for engaging in such activities. To explore the research questions, an extensive empirical study was undertaken. Here, the main delimitations were geography, company type, size etc. More specifically, the study was delimited to focus on non-financial, Scandinavian companies. The rationale underlying these delimitations, as well as a series of other delimitations specifically related to the research design, will be accounted for in chapter 3 Research methodology. Additionally, the focus of this thesis is the investigation of the hedging horizon of Scandinavian, medium-sized and large, non-financial firms and not on the scope of these hedges. 1.2 Terminology Time horizon The focus of this thesis is on the time horizon for companies foreign exchange rate exposure management. It is important to distinguish between the time horizon from 3

7 identification to maturity of an exposure and the time horizon (maturity) of a financial instrument and debt. The first is defined as being the time horizon for which companies generally plan and manage their exchange rate exposure. The second time horizon is the maturity of the financial instruments and debt used to cover foreign exchange rate exposure. The difference between these periods occurs as a company does not necessarily hedge an identified 1-year exposure by the use of e.g. 1-year contracts but instead it may use other strategies such as layered hedges. This thesis main focus is on the maturity of financial instruments and debt. Earlier literature does not define when the time horizon defined to be long or short and likewise it is not defined in this thesis. However, it has to be mentioned that foreign debt and currency swaps is seen as being used in long term currency exposure management whereas options and forward contracts is seen as being used to hedge short term currency exposure. This assumption is also evaluated in the descriptive statistics, chapter Thesis structure The thesis is organized as follows. The next chapter covers the conceptual framework which includes the theoretical background and previous research. Chapter three explains the research methodology while chapter four presents detailed descriptive statistics for the population and for the survey responses. Chapter five presents the statistical regression results while chapter six concludes. The structure of the thesis is further illustrated in the figure below. Figure 1: Thesis structure This figure presents the structure of the thesis and illustrates the topics and contents of the following five chapters 4

8 2. Conceptual framework This chapter presents the conceptual framework used in this thesis. This is done in order to create a joint frame of reference within the discipline of foreign exchange rate risk management. In figure 2, each of the three sections within the conceptual framework is presented. The first section contains an overview of international companies exposures with focus on the foreign exchange rate exposure and the time horizon. Section 2 examines how hedging creates value and which instruments companies may use to cover exchange rate exposure. This section concludes by a examining the time horizon within hedging exchange rate exposure. Section 3 presents earlier literature within risk management of foreign exchange rate exposure, as well as the scarce literature emphasizing the time horizon of foreign exchange rate exposure management. Figure 2: Outline of the conceptual framework This figure outlines the three sections within the conceptual framework. The figure shows the main themes of each of the three sections. 2.1 The exposure of international companies This chapter introduces the reader to how the authors perceive which exposures international operating companies generally face, when moving their business from a national to an international level. 5

9 The chapter begins by outlining and describing some the most important risks that an international company faces Sources of risk There may be several reasons as to why corporations decide to expand their operations internationally. Firstly, by expanding internationally, companies get the opportunity to increase their market size and consequently their revenue/profit, by selling their products in a new market. Additionally, doing business abroad increases a company s possibilities of attracting better and cheaper sources of financial and employees etc, increasing productive capacity, and of using foreign channels of distribution (Hollesen, 2010). However, all of these possibilities may have a negative side effect by contributing with a range of risks, which would not have existed to the same degree in the home market. The main focus in this thesis is non-financial companies, for which reason it is specifically the risk of non-financial companies that will be illustrated. The internationalization of the firm leads to a range of risks, which must be analyzed and managed in order to improve the company s overall performance. Within an internationalized company, there may be a range of different exposures in each division and subdivision and it is essential that these exposures are not sub-optimized in each department but handled in combination with the other departments (Brouthers, 1995). The rationale, on which the above is based, is that even if sub-optimization looks successful at business unit level, international companies are exposed to a range of different macroeconomic, firm-specific and political exposures which are interrelated and hereby affect managers basis for decision (Oxelheim & Wihlborg, 1987). At the same time, the basis for decision may be changed due to indirect risks such as a change of strategy by competitors. It is presumed that only top managers have the overview of the integrated effects of this indirect exposure, for which reason the exposure should be seen from the top manager s point of view and not the individual departments or strategic business unit (Miller, 1998). To present an overview of the combined risk of non-financial companies, the authors of this thesis have made the following figure: 6

10 Figure 2: Identification of the main exposures of international companies This figure presents MNE s total risk. The figure maps all possible risk aspects that MNE s may face operating in an international environment. Source: Own contribution inspired from Oxelheim & Wihlborg (1987), Culp (2001), Meulbroek (2002) and Triantis (2005) Internationalized companies are exposed in different ways and, as Miller (1998) argues, many of these exposures are interrelated and have to be managed combined. The total firm exposure may be divided in many ways, but as illustrated in figure 2, this thesis deals with three overall exposures, namely; general, firm specific and industry specific. It should be mentioned that the list is not exhaustive as more exposures may be identified. However, the figure serves to create an overview of the main exposures that non-financials companies face. All these factors may, to various degrees, affect the company s possibilities of sourcing, its flexibility and its export strategies and hence it may affect the time horizon of their risk management when the competitive situation is changing. The following sections will elaborate on each of these different exposures in turn General exposures International as well as national companies are exposed to a range of risk, which are not related specifically to the company or the industry it operates within. These exposures are gathered under the heading general exposures. Natural disasters, war and terror are examples of environmental risk, which will affect the aggregated level of demand and thus possibly that of the individual company. Macroeconomic risk is the 7

11 risk that a country or a region experience decreasing levels of demand due to changes in inflation rates, interest rates and exchange rates. The risk of exchange rate changes may be general as well as firm specific depending on the market(s) the individual company operates within. The business environment may also be influenced in case of political uncertainties, such as government legislation that change tax conditions, legal minimum wages etc Industry specific exposures This category includes a range of risks, which are not only related to the specific firm but to the industry as whole. It includes the risk of problems with the supply of scarce resources, which are needed for the production in all the companies within the given industry, such as materials as well as employees with specific education/knowledge. It may also be the risk that a substitute product is introduced to the market or that demand for the industry s products is decreasing. The industry may also be affected if the market is over floated with new entrants who are forcing prices down. Furthermore a range of companies may be affected if there is general labour strike, which is related to the whole industry Firm specific exposures The category firm specific exposure is subdivided into two categories, financial and operational risks, and relate to risks that are specific for the individual company. The financial risks category contains the risks that the company experiences financial distress problems due to problems with its liquid holdings, thus making it difficult to pay off its short-term liabilities or making attractive investments. At the same time, the company may experience financial distress if it is close to being insolvent, in case of consecutive periods with recession. If the interest rate level changes it may cause a firmspecific exposure, in case there is a mismatch between the sensitivity of the company s assets (and growth options) and its liabilities (Triantis, 2005). The company may also be exposed if the price level of their commodities changes, consequently affecting the competitive situation and hereby its cash flow. Whether the company operates internationally or domestically the company may also be influenced by exchange rate changes, as it may affect the prices on the company s input as well as its output. As this exposure is part of the main topic of the thesis further elaboration follows in chapters and

12 Operational exposure relates to the exposure that a company s operational costs or revenues will change as the company s internal processes or system fails as well as the risk of the personal that operates them makes mistakes (Meulbroek, 2002). It may also relate to changes in legislation that is specific for the company and its products. Furthermore, operational exposure also includes exposures relating to behavioural finance e.g. the risk that optimal rational decisions are not always made as employees and managers act opportunistically. In accordance with the delimitation of the present thesis, emphasis is placed on financial exposure more specifically exchange rate exposure, for which reason the formation of the exchange rates, price levels and interest rate are linked in the following chapter International Parity Conditions Exchange rate changes affects companies expected domestic cash flows and hence the value of a company. The formation of exchange rates is based on interest rates and price levels and the links between these are called international parity conditions. The time horizon for these market mechanisms to create equilibrium in the market is affected by many factors and, for this reason, it may affect the time horizon for which companies manage their exchange rate exposure. This chapter investigates the basic rationale of the link between these market mechanisms, often referred to as the international parity conditions (Eiteman, et al., 2007) A company s exchange rate exposure to already established sales agreements and liabilities is defined as nominal exposure. Opposing this nominal exposure is the economic exposure, which is defined as real, because exchange rate changes affects prices as well as quantity on the company s input and output side for which reason it affects the company s international competitiveness. The interest rate and the exchange rate formation stems from the supply and demand for liquidity. The following chapters investigate the determinants that affect supply and demand for liquidity in different currencies Interest rate formation The interest rate is the cost of liquidity and there is a range of different interest rates levels, depending on the amount, time horizon, risk profile of the debtor, supply and demand of liquidity etc. 9

13 Supply and demand for liquidity is dependent on all private and financial institutions demand for liquidity versus having these liquid assets invested in other assets. It is the Central Banks, ECB and FED, who control the supply of liquidity. This control is done by selling and buying long-term obligations. The demand for liquidity is highly influenced by the private investors demand for liquid assets to make investments, as well as the price level in the society. If a country has a high inflation rate, the demand for liquidity will increase due to increasing prices. When the inflation is increasing the price of being liquid, which is to say the interest rate, will also increase. This is as a matter of supply and demand and the fact that the persons who hold the liquid assets wants to be compensated for the loss of value they experience when they can not get the same set of goods as before because of inflation Exchange rate formation The background for exchange rate formation may be covered by several approaches. However, the authors of this thesis have chosen to use the parity conditions for thus purpose, as the formation is integrated with inflation and interest rates to create a theoretical framework for the global market conditions 5. As was the case with the formation of interest levels, a floating exchange rate is determined by supply and demand. The demand for a country s currency may have a commercial as well as a financial perspective. A commercial perspective could be if a large part of a country s GDP comes from tourism and/or from exporting goods, the demand for its currency will be high. In a financial perspective, the demand for a currency is related to the demand for transactions. At the macro level the exchange rate increases, when more investments into a country or if the country s national bank wants to hold more of the home country s currency in reserve. The supply of a country s currency is similar to the demand but just the other way around. If a country is importing many goods or if foreign investors are selling off assets or shares in the country, the supply of the currency will increase and the exchange rate will decrease. The total supply and demand for a country s currency is closely linked to the central bank s official interest rates and expectation hereof as well as expectations of political and economic corporation between the countries. 5 Besides this approach we could have used the Asset market approach and the Balance of Payments (Eiteman, et al., p 141f, 2007) 10

14 Parities and the law of one price The connection between exchanges rate and interest rates are, as mentioned, often explored in different parity conditions (Eiteman et al., 2007) and the following chapter consist of a discussion of the law of one price (LOP), the Purchasing Power Parity (PPP), the International Fisher Effect (IFE) as well as the Interest Rate Parity (IRP). The LOP is a theory, postulating that prices of identical products should be equal when translated from one currency to another, in the absence of sales restrictions, transport cost and other market imperfections. The rationale behind this theory is that goods that are cheaper abroad will be imported, which will then force the domestic prices down until it is equal to the foreign price (Dufey & Srinivasulu, 1983, p. 55). The spot exchange rate, S t, is then given by the relationship between two individual goods, i, at time, t. For the purpose of illustration, this is could be between a US good measured in dollars, $, and a Danish good measured in Kroner, DKK (Eiteman et al., 2007) The spot exchange rate is given by the formula: S t = P $ i,t / P DKK i,t The real exchange rate is a measure of a company s competitiveness, as this shows the differences in the purchasing power between two countries. This relationship is often outlined in the PPP (Oxelheim & Wihlborg, 1987). PPP ease on the assumptions of LOP and states that the real exchange rates between countries could be found by looking at two baskets of identical goods. If that is not the case, the exchange rates are not in equilibrium and one currency is over/under valuated. According to this absolute version of PPP, the spot exchange rate could simply be calculated by looking at two identical baskets of goods in two countries. (Eiteman et al., 2007) The spot exchange rate is then seen in the following formula: S t = P $ t / P DKK t If the PPP is to hold and the inflation is given by π, the future spot rate in time t+1 will be: S t+1 = P $ t (1 + π $ )/ P DKK t (1 + π DKK ) The PPP implies that even if a country has more inflation than another, the price differences will be offset by an opposite change in the nominal exchange rate, leaving the real exchange rate unchanged If the assumptions of the absolute version are eased, the countries exchange rates are said to be in equilibrium when the general cost of living is, over time, the same in any two countries. This is called the relative purchasing power parity, RPPP. This implies 11

15 that even if a country experiences inflation it will not affect the long-term standard of living of the citizens as the inflation will be offset by a change in the future spot exchange rate between the two countries (Brealey et al., 2007) and (Eiteman et al., 2007). The RPPP place some interest in hedging exchange rate exposure, because if the market effects are not realized in the short-term, the time horizon becomes interesting and companies may use exchange rate exposure management. The PPP has been tested many times and holds poorly over a short period of time, but is more efficient over a long period of time. In this sense it is a poor way to determine the expected short-term spot exchange rates between countries (Rogoff, 1996). This market imperfection is presumed to exist due to the fact that market reacts slowly to real price changes, for which reason companies operating cash flow may be very volatile and hereby making the company exposed to changes in the real exchange rate. On the other hand, the RPPP seems to be quite precise in determining the long run exchange rate equilibrium where different inflations rates are offset by opposite changes in spot exchange rates. The link between the international money markets and the exchange rates may be clarified in IFE and IRP. IFE states that the exchange rate between two countries should change by the same percentage, as the change in the interest rates, but with opposite sign. According to this parity, it does not matter in which country an investor places his investments, as differences in nominal interest rates will be offset by changes in the nominal exchange rate leaving the real return the same. This parity is under the condition that the investment is placed in positions with the same maturity and risk profile. IRP links the national interest rates and the forward rate premium/discount. According to IRP the one-year forward exchange rate and the unbiased future spot rate may, under perfect market conditions, be seen in the subsequent parity: F $/DKK 1 = ((1+i $ ) 1 / (1+i DKK ) 1 ) * S $/DKK 0 Where F $/DKK 1 is the one-year forward rate between US and DK, i $ is the interest rate in US, i DKK is the interest rate in DK and S $/DKK 0 is the spot rate between $ and DKK (Eitman et al., 2007). As seen in the parity conditions exchange rates may affect a firm s domestic cash flow, the value of e.g. obligations or financial positions thus affecting its accounting earnings and/or the value of a company. The time horizon for the parity conditions to fully materialize is empirically uncertain, which supports our interest in the exchange rate exposure management horizon, for companies doing business in an international 12

16 environment. International operating companies may be exposed to exchange rate changes directly or indirectly, depending on the competitive situation in the market for which they operate. Therefore further investigation of the foreign exchange rate exposure is vital in the answering of our research questions Foreign exchange rate exposure Foreign exchange rate exposure is often divided into three overall exposures - transaction, translation and operational/economic exposure. This division is made due to the different influences that the exchange rate fluctuations have the company (Eiteman et al., 2007, s. 251). Transaction exposure is the amount (domestic currency) that a company s current financial obligations/contracts will change due to an unexpected change in exchange rates. Furthermore, as the value of the company s future cash flow changes as a result of the exchange rates, the so does the value of the company according to the DCF model 6. As the obligations have been made and hence the prices have been settled, the exposure is defined as being nominal (Muller & Verschoor, 2005). Translation exposure differs from transaction exposure, as it is the amount that the value of a company s foreign financial statements will change when they are translated from the foreign country s currency to the domestic country s currency. In that sense, it is simply a translation and thereby a change in the company s book value, for which reason it does not affect the company s future earnings or market value (Grant & Soenen, 2004). For the same reason, authors often cover this as the accounting exposure 7. The exact effect of the translation may be difficult to measure, all depending on how the company is disclosing the cash flow translation according to IAS 21 (IASCF, 2008). Operating exposure measure the in present value of the firm in case of an unexpected change in the future foreign exchange rate. The exposure may be divided in to two categories. Firstly, the market value of the firm may change due to the fact that unexpected variation in the exchange rate will affect the relative prices of the goods sold and bought in anticipated transactions (not yet made contract for) and hereby changing 6 See for example Berk, J., & DeMarzo, P. (2007) 7 See for example Grant, R., & Soenen, L. (2004) or Lessard, D. R., & Lightstone, J. B. (July-August 1986) 13

17 the cash flow from its future operations. This exposure is said to be a direct exposure (Grant & Soenen, 2004, s. 53). The second reason for the market value of the firm to change is indirect, called indirect or competitive exposure, and occurs when the company s competitive situation is considered as being at risk, when the different competitors situation changes due to the unexpected exchange rate variations. This change affects the company s economic environment, giving them a competitive advantage or disadvantage, and hereby its growth opportunities (Marston, 1996). Operating exposure is characterized by having a longer time horizon than translation and transaction exposure, and for this reason it may also be much more difficult to measure and manage/hedge (Lessard & Lightstone, 1986). As opposed to translation and transaction exposure, operating exposure is measured in real terms, because the fluctuations in the exchange rate will affect the relative prices/quantity of the input as well as the output side. Grant & Soenen (2004) finds that the impacts of operating exposure depends on whether the competition is national fragmented (direct exposure has deepest impact) or global fragmented (indirect exposure has deepest impact). Adler & Dumas (1984 p.41) also find that even if a company is not directly involved in international business, its profits may be indirectly affected, due to the fact that its competitors have cost denominated in other currencies. In that sense the company is not directly exposed in terms of their accounts, but in an operational perspective they are definitely indirectly exposed. Hakkarainen et al (1998) finds large differences in companies management of translation, transaction and operating exposure even if recognized. They find that companies with foreign exchange rate policies tend to hedge transaction exposure, whereas translation and operating exposure where not hedged to the same degree. The reason behind this may be the companies difficulties in quantifying the companies operating exposure. Other authors finds that the transaction exposure is often hedged in accordance with the company s hedging policy, whereas the real operating exposure is hedged more irregularly (Bodnar, et al., 1998). This may be explained by the difficulties in the quantification of the exposures and the distinction between indirect and direct exposure (Flood & Lessard, 1986). 14

18 2.1.4 Exchange rate exposure, parity conditions and the time horizon As mentioned, a change in the home country s price of a foreign currency is defined as being nominal (Flood & Lessard, 1986, p. 27). If the hypothesis of the PPP holds true, the change in nominal exchange rate is not a real exposure, but rather a short-term state, which will be counterbalanced by a change in the price level. For this reason the company should not hedge the nominal exchange rates, due to the fact that a company s revenues and expenses will be equalized leaving the profit unchanged (Grant & Soenen, 2004, p. 54). If the change in exchange rate across countries does not trigger an exact opposing change in the price level (inflation/deflation), the change in exchange rate is defined as being real. In other words the prices and exchange rates have not moved in parity with each other. (Flood & Lessard, 1986, p. 27) For this reason, the real exchange rate is often referred to as the inflation-adjusted exchange rate. If the PPP is to hold true, any change in exchange rates between two countries will cause an equal change in the price level and if so the exchange rate change is termed as being 100% passed through (Eiteman et al., 2007, p. 110). In other words, the degree to which an exchange rate change is passed through is termed by what degree of the nominal exchange rate change that has led to a change in the price level. The extent to which an exchange rate change is passed through is according with the LOP and PPP, is,in the long run 100% at both product level, company level as well as at aggregated level. In practice, even if the pass through is 100% at aggregated level in the long run, the extent of the pass through might differ for specific industries, companies and products as well as from country to country (Dufey & Srinivasulu, 1983). Grant & Soenen (2004) finds that elasticity of demand has a positive influence on the operating exposure and hence impact profits as the exchange rate changes. Furthermore, ceteris paribus, pass-through degrees are higher for inelastic products than for elastic product. Hence, industries selling products with very high elasticity of demand may be more exposed to exchange rate fluctuations if the exchange rate change is not passed through compared to industries selling products with inelasticity of demand. This assumption proposes an inclusion of a variable that could account for the price elasticity on the companies input and output in the time horizon determination. Bodnar, et al. (2002) and Yang (1997) find that the degree of pass-through is also linked to industry characteristics such as the degree of competition, product substitution and relative 15

19 marginal cost for domestic and foreign producers. At the same time, price adjustment may be imposed faster in some industries/countries than others, as some markets are more liquid than others. Thus, the degree of pass-through and the time horizon of financial hedges become interesting as the pass-through level depend on the time horizon you have defined. Similarly, previous studies have shown that the extent of pass through is high in countries with high inflation rates and low in countries having low inflation rates (Calvo & Reinhart, 2002). At the same time, the economic policies may have significant influence on the degree of pass through (Engel, et al., 2004). As there is no discernible difference in the inflation rates in the countries within our study 8 differences in inflation rates is not tested and accounted for. This chapter suggests that industry characteristics might affect the time horizon for which companies use financial derivates to hedge exchange rate exposure. Furthermore, the degree of pass-through and the time horizon for which exchange rates are at equilibrium might affect the companies operating flexibility. If a company operates in an industry where input and output prices are volatile due to e.g. the degree of pass through, it may decide to invest more in operational flexibility than if it operated in markets where market mechanisms were faster to create equilibrium. 2.2 Hedging a risk management tool The following section will show how managing foreign exchange rate operating exposure may add value to the firm, illustrate instruments financial managers have at their disposal and relate the different risk management tools to the aspect of time horizon determination. In order for companies to justify undertaking risk management programmes that take up valuable time and resources, such programmes must increase firm value and benefit the shareholders of the company. The finance literature on corporate risk management and its value enhancing nature have opponents as well as advocates. The following section will briefly present some of the arguments for and against risk management. 8 g=labels&plugin=1 en&pcode=tsieb060&tableselection=1&footnotes=yes&labeling=labels&plugin=1 16

20 2.2.1 The value of corporate risk management Miller and Modigliani (1958) presented the idea that in perfect capital markets with no market imperfections such as taxes, costs of financial distress and asymmetric information, the value of a company is independent of its capital structure due to the fact that investors are able to replicate the financial decisions of the company, on their own. This line of reasoning was expanded by Stulz (2003) in what he called the risk management irrelevance proposition and by Smith and Stulz (1985) in the hedging irrelevance proposition. The risk management irrelevance proposition states that hedging a risk does not increase firm value when the costs of bearing the risk are the same whether the risk is borne within the firm or outside the firm by the capital markets. (Stulz 2003 p. 45f) whereas the hedging irrelevance proposition states that if markets are complete and perfect, the value of the firm is independent of its hedging policy (Smith & Stulz 1985 p. 392) Following this logic, companies should not commit to corporate risk management and hedging programmes, as investors can control the risk profile of their portfolio by trading a range of risky assets and thus obtain the amount of risk that suits the risk preferences of the individual investor. Corporate risk management would thus be worthless or possibly a value reducing activity, as the time and resources spent could have been used more effectively elsewhere in the company. The above was based on an assumption that financial markets work perfectly and that no market imperfections exist. However, financial markets are not perfect and market imperfections do exist, and the potential value of risk management lies in a company s ability to successfully address these market imperfections. The next section will describe in more detail the value adding nature of risk management in relation to market imperfections Rationales for hedging Much empirical end theoretical research has been conducted on the determinants of risk management. However, one of the central research papers often cited in later research is the theoretical paper by Smith and Stulz (1985). In this paper, the authors present three explanations for the value increasing nature of corporate risk management. The first of these is related to corporate taxation. What Smith and Stulz explain here is basically that corporate risk management can reduce the tax liability of a company by moving income from higher tax brackets to lower tax brackets if a firm faces a progressive tax structure, 17

21 thus smoothing earnings. This reduction in the company s tax liabilities is what creates value for the company s shareholders. Secondly Smith and Stulz also present the argument that corporate risk management can increase firm value by reducing the expected costs of financial distress. This reduction in the probability of encountering financial distress is achieved by using risk management to reduce the variability of the firm s future value. By reducing the variance of firm value, risk management also reduces the probability of the firm incurring bankruptcy costs, and the reduction of expected bankruptcy costs is valuable to the company and thus its shareholders. The final argument presented in the article by Smith and Stulz ties together risk management and managerial risk aversion. Another rationale for corporate risk management is presented by Froot et al. (1994) and relates to the underinvestment problem. An important element in the value creation process of companies is to engage in positive NPV projects, and in order to do this, companies need to have sufficient internal cash to fund these projects. Following this logic, Froot et al. argues that the goal of any risk manager should be to ensure that the company has enough cash available to make value enhancing investments. The arguments of the authors rest on financial pecking order theory, which states that companies prefer to use internally generated funds to finance new projects. If no such funds are available within the firm, the firms then rely in the debt markets to provide the funds needed, and if no other option is available to them, firms may turn to the equity markets in order to fund projects. The costs associated with each of the three costs increases with internally generated funds as the cheapest source of finance. The authors conclude that the value enhancing nature of risk management in relation to the underinvestment problem comes from its ability align the internal supply of funds with the external demand for funds which, if invested in positive NPV projects adds real value to the company. This section briefly presented some of the theoretical arguments for firms to pursue risk management strategies. The following section relates to the various techniques managers have at their disposal for managing foreign exchange rate exposure Risk management techniques When managing foreign exchange rate exposure, companies can either turn to the financial markets in order to offset the exposure using financial derivatives, or they can use strategic approaches within the firm in order to manage risk. However, most 18

22 companies often use a combination of the two approaches to manage their exchange rate exposure. Generally speaking, financial hedges are best used in order to minimize the risk associated with near-term transactions that are often known and fixed in foreign currency terms, such as transaction and translation risks. Economic exposure on the other hand, where the value of the cash flows are uncertain, are best managed using strategic approaches such as sourcing and production location as well as long term debt (Srinivasulu, 1981) Financial hedging Adler & Dumas (1984) define a hedge as follows: The amounts of the foreigncurrency financial transactions (i.e., forward contracting or its equivalent) required to render the future, real, domestic-currency market value of an exposed position statistically independent of unanticipated, random variations in the future domestic purchasing powers of these foreign currencies. In order to make future cash flows independent of foreign exchange rate changes, companies can use financial hedges. Financial hedging involves the use of financial instruments such as forward contracts, swaps or currency options to manage foreign currency risk. The basic goal of financial hedging is to increase the value of the firm by reducing the variability of the firm s net cash flows. Financial hedging generally involves transactions between a company and a financial intermediary (e.g. banks), that facilitate the transfer of unwanted risks from the company to another company that has the capacity or the will to take on additional risk at a given price. In the following sections, two types of financial hedges will be described, namely short-sighted derivatives (forward contracts and currency options) and long-sighted derivatives (swaps). Short-sighted derivatives The most common way for a company to hedge their foreign exchange rate exposure by means of financial hedging is by using a currency forward contract. A currency forward contract is a contract that fixes the exchange rate of a transaction of a future transaction, and it is usually a contract that is written between the firm and a bank (Berk & DeMarzo, 2007). The benefit of using a forward contract is that the firm is able to lock in a future exchange rate which will enable the firm to completely eliminate the risk or reduce it depending on the amount of their future cash flow the firm chooses to hedge. 19

23 However, a potential drawback of using a forward contract is that the firm will not be able to realize potential gains if exchange rates should move in a favourable direction as the forward contract is a binding agreement to buy or sell currency at a future date. A financial instrument similar to the forward contract is a foreign currency futures contract. However, futures contracts are marked to market on a daily basis which means that the firm is likely to have to make margin calls more frequently than most businesses usually want (Eiteman, et.all, 2007). Thus, futures contracts are not ideally suited for business and are most commonly used by speculators as they do not have access to forward contracts. As the present thesis focuses on hedging, not speculation and on businesses rather that individuals, futures contracts will not be included when referring to short-term hedging instruments. Another method for a company to hedge using financial means is by use of currency options. Where the forward contract described above locks the exchange rate for the future transaction, a currency option allow companies to protect themselves on the downside while retaining the potential for upside gains should the exchange rate move in a favourable direction (Berk & DeMarzo, 2007). The important feature of the currency option is that it gives the buyer of the option the right, but not the obligation, to buy or sell a given amount of foreign exchange at a fixed price per unit for a specified time period (Eiteman, et.all, 2007). Thus, when buying a currency option, you get the benefits of the forward contract as well as the potential upside gains if the exchange rate moves in a favourable direction. However, the cost of obtaining a currency option is also correspondingly higher than that of the forward contract. Currency options can be used for short-term hedges as well as in some cases for longerterm hedges. Most options are traded for time periods of up to one year, but in some cases, two or three year options are also written (Eiteman et al, 2007). Long-sighted derivatives Foreign currency swaps is another type of derivatives that is often used by firms to hedge foreign exchange rate exposure. While the forward contracts and financial options described above are often used to hedge near-term exposures, foreign currency swaps are often used for hedging over longer periods. The motivation when for using currency swaps is generally to replace cash flows in an unwanted currency with cash flows in a desired currency. 20

24 Whether to consider the use of foreign denominated debt as a financial or an operational hedge is difficult as it involves dealings in the financial markets but is often used for strategic purposes. For the purpose of this thesis, foreign denominated debt will be considered a financial hedge in line with currency swaps. Financial hedges are not well suited for hedging operating exposure, as they are both costly and possibly ineffective in doing so. For example, if changes in exchange rates alter a company s competitive situation in a foreign market or if markets for production become unattractive due to these changes, financial hedging is of limited use to the companies. As alternative to financial hedges operational/strategic hedging may be better suited for managing such risks that arise from changes in the firm s operations Operational hedging and real options The following quote from Capel (1997) illustrates that the objective of the operational hedging of operating exposure is the same as the objective of managing transaction and translation exposure using financial means that is to anticipate and manage the effect of unexpected changes in exchange rates on the future cash flows of the company. the objective of economic exposure management should be to exploit the variability of exchange rate rates in order to increase the firm s market value, while ensuring that the downside risk of exchange rates remains sufficiently small (Capel 1997) However, the quote also illustrates an important feature of operational hedging, which is its ability to exploit changes in exchange rates to increase firm value. In order for a firm to successfully address operating exposure using operational hedges, it needs to be able to recognize when the parity conditions, explained in section 2.1, are in disequilibrium and be in a position to act accordingly. In order to exploit the disequilibrium, the firm needs to have the flexibility to shift its operations such as sales, production and sourcing between different countries or to be able to shift its financing sources between different currencies (Eiteman et al. 2007). In order to obtain and exploit this operational flexibility, the firm is required to make initial investments in production plants, sales subsidiaries etc. However, once such investments in flexibility are made, the company is not only able to more efficiently protect itself from changes in exchange rates; the flexibility also gives the firm the ability to actively pursue profit opportunities by shifting operations between countries as exchange rates move. However, such operational shifts are neither instantaneous nor 21

25 free but as long as adjustment time and costs are not prohibitively high, firms can still benefit from its ability to change its operations (Capel, 1997) and (Triantis, 2005). Operational hedging a real options perspective The real options perspective can serve as a good framework in which to view the different operational approaches to managing foreign exchange rate exposure. In line with the article by Aabo & Simkins (2003), this thesis distinguishes between four types of real options that a firm may exploit. These are as follows: I. The option to expand: entering a new market or expanding operations in an existing market II. The option to defer: delay entering a foreign market III. The option to abandon: abandon or temporally reduce/close operations in a foreign market IV. The option to shift: shift production outlets or change sourcing between suppliers in different countries. I) The option to expand. The option for a firm to expand involves the firm s ability to enter a new market or expanding operations in existing markets due to exchange rate changes. A firm with sales subsidiaries in another country is well positioned to exploit such a real option as it is easier for a firm already present on a foreign market to introduce new products or to promote existing products than it is for a firm to enter a completely new market. Thus entering a new market automatically entails an option to later expand production. This type of option can be obtained by acquiring a foreign company of through greenfield investments in a new market. II) The option to defer. The option to defer relates to a firms ability to delay entry into a foreign market as a reaction to adverse market changes as well as the firms ability to wait and see, that is gather more information so as to make an increased informed decision as to enter a new market or not. For the purposes of the present thesis, the option to defer relates to whether changes in exchange rates influence a firm to delay entry into a foreign market that they had already existing plans to enter. III) The option to abandon. Another real option is the option to abandon a foreign market. If exchange rate changes decreases the competitiveness of a firm, increases the price of inputs or similar, the option to abandon is effectively the option to close down operations in the market and sell the assets involved. However, the firm may wish to 22

26 just temporarily reduce or close down its operations. By closing down production or rescaling operations, the firm is able to reduce its foreign exchange rate exposure. IV) The option to switch. The last real option strategy discussed here is the option for a company to shift production between production subsidiaries in different countries or change their supplier due to changes in exchange rates. Investments in production subsidiaries in different countries allow firms the flexibility to change production location based on where the costs of production are cheapest. However, shifting production between countries is not free, but as long as the benefits exceed the costs of switching, the flexibility to change sourcing/production location is valuable to the firm, and can serve to reduce exchange rate exposure. Currency matching If a firm has an anticipated continuous long-term exposure to a particular currency, it can offset the risk associated with the exposure by matching the revenues or costs with corresponding costs/revenues. Matching cash flows can be done in a number of different ways. One way of offsetting long term exposure in a foreign currency is to acquire debt in the same currency. In this way, the firm can use earnings in the foreign currency to service the principal and interest payments in the same currency, thus matching inflow and outflow and eliminating the risk. Alternatively, a firm with cash inflows in say USD could reduce the risk associated with these inflows by seeking out a supplier in the US as use the cash inflows to pay for the raw materials whereby the firm will reduce the risk (Berk & DeMarzo, 2007) Risk management and the time horizon of hedges In the discipline of risk management financial hedges and operational hedges may be seen as substitutes but more companies probably see them as being complementary. As mentioned financial hedges are not well suited for managing operating exposure as the time horizon for the financial hedge contracts are fixed, whereas the time horizon of the operational exposure is unfixed while at the same time difficult to quantify. This emphasizes the importance of choosing a proper time horizon for the financial hedge. Choosing a short horizon may prevent the underlying profitable objectives of the hedge from being met, while choosing a long horizon on the hedge may actually contribute with extra risk as estimation of the expected cash flows may be biased (Lidbark, 2010). 23

27 Whether operational hedges are used as substituting or complementary protections, Capel (1997) stress that it is of much importance to lower the adjustment cost to exploit these real options. If the companies have invested in lowering the adjustment cost the time horizon becomes of less importance. In an ideal world where the adjustment costs are zero and adjustments instantaneous, the time horizon is unimportant, but it is a trade-off between the cost of the real option and the value that operational flexibility creates. If operational and financial hedges are used as complementary the time horizon is of much importance. The time horizon, ceteris paribus, of the financial hedge should match the time horizon that it takes to exercise a real option without prohibitively large adjustment cost to follow. 2.3 Empirical literature on the time horizon of foreign exchange rate exposure management Existing empirical literature treating the time horizon of non-financial companies risk management and specifically the time horizon for their exchange rate exposure management by means of financial and operational hedges is very limited. The following chapter is divided into two parts and deals with the rather limited literature within this field. In accordance with the theoretical delimitations outlined in chapter 1.2, the subject of the first part is the most important literature about the foreign exchange rate exposure of non-financial companies and their management hereof. The second part is a review of the scarce literature, dealing with the aspect of a time horizon when managing foreign exchange rate exposure. Additionally, the last part includes literature on company flexibility and the impact on the time aspect when managing exchange rate exposure General literature about companies risk management of exchange rate exposure and their use of derivatives Many earlier studies have focused on risk management by non-financial companies with a specific focus on their use of derivatives. Allayannis & Ofek, (2001), Bodnar, et al., (1998), Géczy, et al, (1996) and Nance, et al, (1993) all focus on American nonfinancial companies whereas research conducted by Hagelin, (2003), Bodnar & Gebhardt, (1999), Alkebäck & Hagelin, (1999) has focused on European non-financial companies. Both studies in which Bodnar is a co-author have been made with close-end 24

28 questionnaires and have been of much inspiration for similar studies such as Hagelin & Alkebäck (1999) and Hagelin (2003)which were based on Swedish non-financial companies. Nance et al (1993) and Géczy et al (1996) also contributes to the literature about non-financial companies use of derivatives to hedge exchange rate exposure, but contradictory to Hagelin and Bodnar they use information from the stock market and financial databases to conduct their survey. In the 90 s and 00 s increased focus has been on the usage of real options (move production, delay investments, change suppliers etc.) in relation to financial hedging of exchange rate exposure management. On this subject, Capel (1997), Triantis (2000) as well as Aabo & Simkins (2005) have made interesting theories and provided empirical evidence which has also emphasized importance of the time horizon of hedging. The usage of real options in relation to financial hedging is, among others, a matter of the company s flexibility. The relationship between the time horizon of financial and operational hedges is thus interesting, as one could argue that the time horizon that firms should hedge using financial means is equivalent to the time horizon at which they are able to exercise their real options. One of the most quoted and used studies on financial risk management in non-financial companies is the 1998 Survey of Financial Risk Management by U.S Non-Financial Firms (Bodnar et al, 1998). This study was made with a population of 1928 nonfinancial U.S. companies, who were asked to answer 21 closed-end questions about their financial risk management and use of derivatives. Bodnar et al found that there was a large correlation between firm size and the use of financial derivatives, and that especially manufacturing firms were using derivatives. At the same time, they found only a small fraction of the exposure was hedged as the average firm hedged less than 50% of the apparent exposure. They also found that manager s focus was on short-term exposure, transactions exposure and anticipated transaction below one year, rather than anticipated transactions longer than one year and long-term operating exposure. Bodnar also made a similar study with Bodnar and Gebhardt (1997) on German manufacturing companies. The survey showed that German companies used more financial derivatives to hedge exchange exposure than their American counterparts. This difference can be based on the fact that of a more open economy in Germany in Alkebäck and Hagelin (1999) use the same method as Bodnar et al to investigate Swedish non-financial usage of derivatives and compare them to similar international research. They find that the Swedish companies use more financial derivatives than 25

29 their American and New Zealand counterparts and that the reason for hedging is focused on their balance sheet rather than specific operations. Like Bodnar et al (1998) the authors also found a positive correlation between size and use of derivatives. Allayannis and Ofek (2001) used information from a financial database to investigate 500, U.S. Non-financial companies use of derivatives to hedge their exchange rate exposure. They found that the companies, which used derivatives, lowered their exposure significantly. They also found that the decision to hedge was positively correlated with company size, R&D expenses, and foreign sales ratios. Hagelin (2005) made a survey with 101 Swedish non-financial companies who answered a questionnaire regarding the reasons for their use of hedging with derivatives. Hagelin found that companies hedge transactions exposure with currency derivatives to increase firm value by reducing indirect cost of financial distress and reduce the problem of underinvestment. Aabo & Simkins (2005) analyzed Danish non-financial companies exchange rate exposure management based on a closed-end questionnaire. They found that the choice of whether or not to use financial derivatives for hedging foreign exchange rate exposure was significantly affected of the manager s ability/willingness to undertake/exercise real options. Furthermore, they made a model to explain the usage of real options and found that a combination of company size, export share and foreign subsidiaries had a positive effect of the real option usage. These findings help to underline our interest in an analysis of the time aspect and influence of flexibility on exchange rate exposure management by financial means Literatures emphasizing the importance of the time horizon in managing exchange rate exposure Srinivasulu (1981) was one of the first to discuss the aspect of a time horizon when companies are managing their exchange rate exposure. He argues that the time horizon for a company s operating exposure may be measured by the time required for the company to exercise a real option such as changing markets, inputs etc. According to this argument, a company that has flexibility built into its operations is not exposed for longer than the time it takes for the company to exercise its real options. Froot et al. (1993) elaborates on the time horizon seen from an investor s point of view, where each company is part of the investors portfolio, for which reason it is the investor s time horizon that is of interest when hedging the exchange rate exposure. 26

30 This argument leaves the time horizon less important as investors may hedge on their own hereby deciding the time frame within which they want to minimize exposure. This argument relates to the general discussion about the value creation of risk management, as some may argue that exchange rate exposure in general should be left unhedged, as the shareholders may hedge all exposure themselves. Similar to Srinivasulu, Stulz and Williamson (1996) also discuss the time horizon for the company s operating exposure and makes the same arguments that the horizon for operating exposure is no longer than the time it takes for the company to exercise its real options. Furthermore, they emphasizes that the time horizon is firm specific and depend on the intention of managers. If the purpose of the exchange rate exposure management is to ensure that investments (with positive NPV) are made, the time horizon for the operating exposure is the same as the company s planning horizon. On the other hand, if the purpose is to reduce taxes or shareholders risk, the horizon the company is exposed, is the time horizon for their budgeting. Capel (1997) makes an investigation of real options in relation to economic exposure management and emphasize that companies may benefit by lowering the time it takes to shift production, storage, sales from one currency to another etc., which once more emphasizes the importance of choosing the proper time horizon for risk management in relation to the use of real options. Firms that invest in flexibility (buying real put or call option) by lowering adjustment cost may minimize downside risk. Such an investment makes the time horizon less important as highly flexible companies may exercise these real options leaving profit stable. Capel argues that financial instruments are not ideal for managing operating exposure, for three different reasons. One of the reasons is that the maturities of a standard financial contract is predetermined, often resulting in a mismatch between the time horizon of the exposure and the maturity of the instruments used. The argument emphasizes the need for an investigation of the time horizon of the financial instruments used for foreign exchange rate exposure management. Aabo (1999) made a cross case study of eight large Danish companies exchange rate exposure management, and found that the companies where focusing on their short-term cash flow exposure from obligated transactions, and not so much on the long-term operating exposure. The most important reason for the short term focus is found in the companies possibility of exercising real options and the dynamics of the business environment, which lead to a focus on short-term exposure and not so much on longterm operating exposure. Operating exposure is in theory indefinite, however, in the 27

31 survey only one of the eight companies where analyzing exposure more than 5 years into the future. Only two of the companies hedged 25-75% of their competitive exposure and the rest did not hedge this exposure at all and only one in eight companies used derivatives to hedge competitive exposure. Finally the author states that the different structural settings and management objectives leads to different time horizons when companies are hedging their exchange rate exposure. Aabo also find that none of the companies take their competitors exposure setup into account when making hedging related decisions. The last finding is investigated further in the present thesis. Grant & Soenen (2004) emphasize the role of the time horizon, as financial instruments are specific in regards to amount and time, whereas the amount of operating cash flow is volatile and the time horizon is continuous. They state that the use of short-term financial derivatives only reduces short-term transactions exposure but does not reduce long term operating exposure, as undefined transactions are extending indefinitely. In that way, hedging with short-term financial derivatives does not add value, as, according to the parity conditions, the contracted forward rate is an unbiased predictor of the future spot rate. Aabo (2006) examines Danish non-financial firm s management of exchange rate exposure, and specifically the differences and similarities between the use of debt and derivatives. This investigation is interesting in the field of the time horizon of exchange rate exposure as foreign debt is often used for long term hedging, whereas derivatives (options, forwards) are often used for short term hedging. Aabo finds that foreign debt is an important alternative to currency derivatives, and that the importance of foreign debt is positively related to numbers of countries in which the firms has subsidiaries and the extent to which firm value is based on assets already in place (measured by Tobin s Q). Latest Lidbark (2010) tries to explain a range of factors that managers should consider when they choose the horizon for risk management including exchange rate exposure. He emphasizes the time horizon and argues that the choice of a proper time horizon is a trade-off between choosing too short a time horizon which may mean that the economic benefits from hedging transactions may not be met and that choosing too long a time horizon may cause additional exposure due to forecast errors etc. He states that a proper time horizon should be seen in the perspective of how the company wants to create value by hedging, as short term hedging on a rolling basis does not create value but only minimizing short term variance and hereby helping in a planning objective. As 28

32 Srinivasulu and Stulz & Williamson, he also argues that companies should only hedge for the horizon for which it could become very costly to exercise real options in case of an unfavourable development in the exchange rates. It should be mentioned that Lidbark focuses on the time horizon for which companies identify their exposure rather than the time horizon of the financial instruments used to hedge exchange rate exposure 29

33 3. Research Methodology This section of the thesis begins by presenting the company selection process in terms of the different selection and restriction criteria applied to the initial population. Subsequently, the structure of the questionnaire designed for the survey is presented followed by the survey execution process as well as the survey responses. Concluding this chapter is a test for response bias between our sample of companies and the population as a whole. This chapter is included in the thesis to present the considerations of the authors in terms of the overall execution of the survey. 3.1 Target population In general, much of the research conducted on risk management and hedging practices in firms have used a combination of survey data (primary data) and publicly available data (secondary data) as their foundation. This is because much of this research requires information on the exposures and hedging practices in firms which is often not publicly available. Had this data been publicly available, financial statements could have provided much of the same data while at the same time, mitigating potential problems such as inaccurate or dishonest responses as well as language and interpretation problems, which make up some of the potential shortcomings the questionnaire approach. (Pramborg 2005) The thesis at hand make use of a quantitative questionnaire together with publicly available data rather than financial statements as this allowed us to obtain detailed data about company specific characteristics and approaches in relation to the hedging decisions and general risk management practices that would not have otherwise been available. The data has been collected amongst Scandinavian, non-financial firms of a size larger than the enterprise category small as defined by the European Commission 9. Additional selection and restriction criteria were imposed, all of which are shown in below. Restriction and selection criteria: 1) Geographic region: Scandinavia (Norway, Sweden, Denmark, Finland) 2) Company size: Larger than small companies (EU-classification) See the next section for an explanation of size categories 30

34 3) Industry classification: Non-financial firms (NACE Rev. 2) 11 4) Further restrictions: addresses, web-pages, last available year (2008, 2009) consolidated level (only parent companies) The following subparagraphs will elaborate on each of the restriction/selection criteria in turn. In order to conduct the company selection and to obtain selected financial numbers, this study made use of the Orbis database which is a global database with financial information on more than 60 million companies 12. The Orbis database was used as it contained financial information for all the target countries whereas other databases considered focused on one particular country. Ad 1) Geographic Region The area of interest in the present thesis is Scandinavia. Therefore, a search parameter that excludes all non-scandinavian firms was included in the search leaving us with firms based in Sweden, Norway, Finland or Denmark (including Greenland and the Faroe Islands). The rationale behind the selection of Scandinavian countries was that these countries are all small, open economies that to a large extend are dependent on foreign trade and thus similar in many respects. Additionally, due to the high degree of foreign trade, currency derivatives must be expected to be of great importance to firms in these countries which is ideal for our purposes. (Hagelin, 2003) Ad 2) Company Size In terms of the size of the companies in the sample, all companies smaller than the enterprise category medium-sized as defined by the European Commission were to be excluded. This definition states, that in order to be considered as medium-sized, a company has to fulfil two criteria, namely; to have a headcount of at least 50 employees and to have either a turnover or a balance sheet total in excess of 10 million. 11 See the next section for an explanation of industry classification 12 Orbis is a web based database that contains data about 45 million companies around the world, including: descriptive information, company financials, market research etc. 31

35 Table 1: Definition of small and medium-sized enterprises This table presents definitions of SME s in accordance with the European Commissions definitions as of May For a company to be classified as medium-sized according to these definitions, the company has to fulfil two criteria, namely; 1) to have a headcount of 50 or more employees; and 2) to have either a turnover or a balance sheet total in excess of 10 million Enterprise category Headcount Turnover or Balance sheet total Medium-sized < million 43 million Small < million 10 million Micro < 10 2 million 2 million Source: Thus, in order to exclude small- and micro-sized companies from the initial sample, search criteria for number of employees of at least 50 was included as well as criteria of either a turnover or a total balance of 50 million. For the purposes of this thesis, the total balance variable is used as the measure for firm size as some companies have missing values for turnover, whereas total balance is stated for all the companies in the sample. Ad 3) Industry classification As mentioned in the introduction to the present section, the focus of this thesis is on non-financial firms. In terms of the break down of different industries, the Orbis database allows users to choose between numerous different classifications. For the purpose of the present thesis the NACE 13 classification code was used. All categories of the NACE Rev. 2 with the exception of category K Financial and Insurance activities were included in the sample. Category K being characterized as follows: This section includes financial service activities, including insurance, reinsurance and pension funding activities and activities to support financial services. This section also includes the activities of holding assets, such as activities of holding companies and the activities of trusts, funds and similar financial entities. The exclusion of financial firms rests upon the fact that the focus of the present thesis is on end-users of financial services, not producers. (Hagelin 2003) This leaves us with a sample consisting only of non-financial firms. 13 See appendix 2. NACE is the acronym for Nomenclature statistique des activités économiques dans la Communauté européenne which translates into the statistical classification of economic activities in the European Community 32

36 Ad 4) Further restrictions Further restrictions that were applied to the initial population include the elimination of companies whose last available financial data date back to before This elimination was made to ensure that the financial data used in the thesis was comparable across firms and to account for the financial crisis and credit crunch in the data. Furthermore, a restriction concerning the company level was included to ensure that only companies on a consolidated level were included in the population, as subsidiaries and daughter companies might exhibit different risk management characteristics due to risk management decisions being taken care of by the mother company. Lastly, as the data for the research in the present thesis was collected through the use of online questionnaires, two additional restrictions relating to the contact details of the companies were applied. These included the selection of companies with either an address stated in the Orbis database or if this was not the case, a web-page listed in the database. The inclusion of companies with a web-page was based on a desire to boost the number of companies in the final population in the hope of achieving a subsequent higher response rate. By applying these selection and restriction criteria to the initial sample provided a population of 1399 non-financial, medium+, Scandinavian firms with addresses and 811 companies with web-pages. The 811 company web-pages were searched for either company addresses or the address of the CFO/financial manager. This search yielded 245 company s as well as 68 CFO s. The final population is shown in table 2 below. Table 2: Total population This table shows the total population of companies divided by the source from which they were obtained. Total population s from Orbis database s from company web-page 313 Total population Questionnaire design Detailed data on hedging practices is not readily available in the financial reports or in the Orbis database. Thus, in order to obtain the data necessary to perform the analysis, an electronic questionnaire was distributed to the target population. The questionnaire 14 used was designed based on inspiration from previously conducted research studies 14 The complete questionnaire is presented in appendix 3 33

37 such as Aabo & Simkins (2003), Bodnar et al (1998) as well as previous thesis from the Aarhus School of Business (ASB) library. However, the questions were adjusted were needed to match the purposes of the investigation conducted in this thesis. As the survey was conducted across four different countries with different languages, the survey was conducted in English. The questionnaire was designed and distributed using StudSurvey 15 - a program developed by the IT-department at ASB specifically for the purpose of designing electronic questionnaires. The StudSurvey tool allows users to build and customize questionnaires and to tailor the questions and corresponding answers to satisfy specific requirements. Additionally, the StudSurvey tool has a built in mailing system which allows users to forward the questionnaire directly from the StudSurvey tool and to track the respondents/non-respondents using a unique code given to each firm in the population ( code protection). Finally, StudSurvey allows users to export the answers obtained in the survey into different file-formats which can be used to conduct the subsequent statistical analysis. 16 The electronic questionnaire was favoured over the standard paper-based questionnaire due to certain advantages. Advantages of the electronic questionnaire include lower overall costs; convenience for the respondents; easier management of responses in terms of data processing and follow-ups as well as the option to use intelligent answers that allow respondents to skip certain questions based on previous answers. Additionally, the respondents were not allowed to skip questions, which ensured that no questions were left unanswered. The following figure shows the structure of the questionnaire used for data collection and highlights the overall themes of each section Screenshots of the StudSurvey tool is presented in appendix 4 34

38 Figure 4: Questionnaire structure and contents This figure presents the structure of the survey conducted among medium-sized and large, non-financial firms in Scandinavia. The survey consists of three parts. The first part contains 8 questions while part two contain 4-6 questions. The final part contains 6 questions and the overall themes are described for each section. Part I (8 questions) Part II (4-6 questions) Part III (6 questions) End of survey Part I: Part II: Part III: International involvement and foreign exchange rate exposure Debt and derivatives use for managing foreign exchange rate exposures Firm flexibility and industry stability The questionnaire contains closed-ended questions and is divided into three parts. The first part of the survey consists of 8 questions which all relate to the international involvement of the company and its foreign exchange rate exposure. The second part of the questionnaire contains 4-6 questions as the respondents are able to skip two questions contingent on certain answers to previous questions. The answers in part two concern the degree to which the company uses debt and derivatives when managing their foreign exchange rate exposure. The final part of the questionnaire contains 6 questions which relate to stability of the industry in which the company operates as well as the flexibility of the firm in terms of its operations. The answers to part I, II and III is presented in chapter 4 which contains detailed descriptive statistics for the survey responses. 3.3 Survey execution The companies in this research were contacted by exclusively. Company s were obtained through the Orbis database or company web-pages as described above. The companies were initially contacted in order to obtain the contact details of the company CFO or equivalent. These s were written in English and sent to the general company (e.g. company@company.com or info@company.com) a total of two times. The responses to these s gave us; either the information requested, a response declining to participate due to lack of time/interest or no the s were not answered at all. 35

39 Figure 5: Structure of survey execution This figure presents the structure of the survey execution. After obtaining the company s, companies were contacted two times in order to obtain contact information of the company CFO or equivalent. Next, an invitation to participate in the survey, including an individual link, was sent directly to the financial managers where possible. Otherwise, the was addressed to the general company , asking for it to be forwarded to the CFO. Step 1: Obtaining contact information Step 2: Sending the invitation to participate for the company CFO in the survey Next, an invitation to participate in the survey was sent to the CFO or equivalent where possible. The invitation contained a brief presentation of the study and a survey link. Additionally, the invitation was sent to the companies that had failed to answer the two initial s with the request that the be forwarded the to the company CFO or equivalent. This procedure was repeated two times. However, companies that completed the survey were not contacted further. The structure of the survey execution is presented in figure 5 above. 3.4 Survey response The response rate for the population as a whole as well as by country is presented in table 3 below. 174 companies choose to participate in the survey which corresponds to a response rate of 11,3%. The response rate across countries range from a maximum of 19,6% for Danish companies while Finnish firms had the lowest response rate with just 3,1%. Sweden and Norway firms achieve response rates of 11,7% and 11,6% respectively. In terms of the sources of information regarding the contact information for the company CFOs, it is interesting to note, that while the contact information obtained for Danish firms came exclusively from answers to the s requesting the information, half the contact information for Swedish firms (57) came from information found on the company web-page. The total number of CFO addresses obtained amount to 235 of which 68 come from company web-pages. 36

40 Table 3: Survey response This table presents detailed information about the sources from which the contact information was obtained by dividing it into replies from companies or from the webpage. Furthermore, the table presents the response rate for the survey as a whole and by country. Denmark Finland Norway Sweden Total Step 1: Obtaining contact information for the company CFO replies from companies from web-page Companies without contact information Total population Step 2: Sending the invitation to participate in the survey Non-respondents Total survey response Total population Response rate 19,6% 3,1% 11,6% 11,7% 11,3% Table 4 presents the participating companies divided into categories dependent on the contact information used in the process of sending the invitation to participate in the survey. The table shows that of the 167 companies that answered the initial requesting the contact information of the company CFO, only 70 (41,9%) of these eventually proceeded to complete the survey. 10 companies for which the information had been found on the company web-page participated in the survey which corresponds to 14,7%. Lastly, it is interesting to note that even though they did not answer the two initial regarding the contact information of the company CFO, a total of 94 answers more than half our sample - came from companies to which the invitation had been sent to the general company with a request to forward it to the company CFO. Table 4: Survey response by contact information This table presents the companies that participated in the survey according to contact information obtained before sending the invitation to participate in the survey. Survey response from companies from which contact information was obtained 70 40,2% Survey response from companies for which contact information was found on web-page 10 5,7% Survey response from companies with only the general company address 94 54,0% Total % 37

41 3.5 Response bias This section contains a test for bias between the group of responding and nonresponding companies. The test is conducted in order to test whether the sample can be considered as representative for the overall population. The test is conducted by calculating test statistics for a number of company characteristics, namely; the number of employees as well as three financial indicators (total assets, solvency ratio and leverage). The factors included were obtained directly from the Orbis database. The procedure followed in this section is inspired by the approach used by Pramborg (2005). The test calculates the mean and median for the factors of each group and uses a t-test to test for differences in the means while a Wilcoxon/Mann-Whitney test is used to test for differences in the median values. The tests were conducted using the test for equality function in Eviews 17 and the results from the tests are presented in the table below. Table 5: Results from the response bias test This table presents the results from the response bias test conducted to test for differences between respondents and non-respondents. The test was conducted by comparing the mean and median by means of a t-test and Wilcoxon/Mann-Whitney test respectively. The test is conducted on the following company characteristics: Total assets, number of employees, solvency ratio and leverage ratio. Population (1712) Total Assets (in th. Euro) Number of Employees Solvency Ratio Leverage Ratio Respondents (174) Mean ,16 1,45 Median ,46 1,43 n Non-respondents (1538) Mean ,41 4,26 Median ,48 1,70 n Test for differences (p-values) Mean 0,5120 0,7129 0,3056 0,4609 Median 0,0004 0,0146 0,3017 0,2397 The results of the tests conducted indicate that the survey responses are biased towards large companies, as the test for medians are significant at the 1%-level and 5%-level for total assets and number of employees respectively. As for the rest of the tests conducted, we find no significant differences between the sample and the population as a whole. 17 Eviews 6, a windows-based econometrics software package was used to perform the tests. Eviews will also be used to perform the regression analyses in chapter 5. 38

42 4 Descriptive statistics of empirical data As described in the research design, primary as well as secondary data has been collected as basis for the analysis of the companies time horizon of foreign exchange rate exposure management. As mentioned earlier, to answer the research questions most comprehensively, the quantitative information will be analyzed in two separate parts. In this chapter a detailed descriptive analysis will be performed to create an overview and form basis of estimating any significant differences in the company characteristics as well as the answers they made in the questionnaire. This is done by the use of frequency and cross tabulations and makes an ideal overview to perform a regression analysis performed in chapter Company characteristics The final sample includes 174 Scandinavian companies all fulfilling the selection and restriction criteria outlined in the previous chapter. The objective of this section is to identify any differences between the companies regarding number of employees and total balance (assets) across the countries Employees As seen in table 6, the 174 Scandinavian companies have an average number of employees of 1057 with a standard deviation of 3281, whereas the median company only has 203 employees. The relative high variance is a result of infrequent, high deviations, which is illustrated by the kurtosis of Furthermore, the distribution is positive skewed with a skewness of There is a large difference between the mean numbers of employees divided on countries, where the Danish companies on average have 1795 employees and the Norwegian companies only have 470 employees. Additionally, the number of employees at the Danish companies is positive skewed with a skewness of The number of employees at the Finnish companies is most normally distributed with a kurtosis value on 3.5 and a positive skewness of 1.94, but again it should be mentioned that Finland is only represented of 12 companies 39

43 Table 6: Descriptive statistics on employees and assets This table presents the descriptive statistics for number of employees and the amount of assets for the 174 Scandinavian companies according to country. The tables show the number of employees, constructed from the figures of the last available year (2009) and the assets is in thousand Euros last available year (2009). Employees Std. N Mean Median Deviation Skewness Kurtosis Minimum Maximum Denmark ,50 299, ,01 4,76 24,48 53, ,00 Sweden ,63 162, ,00 3,21 10,27 56, ,00 Norway ,63 188,00 738,02 2,92 9,42 50, ,00 Finland ,08 273,00 954,62 1,94 3,50 63, ,00 Total ,07 203, ,14 7,14 58,32 50, ,00 Total Assets Denmark , , ,53 3,66 14, , ,39 Sweden , , ,70 3,80 15, , ,45 Norway , , ,57 1,56, , ,36 Finland , , ,20 3,26 10, , ,00 Total , , ,19 4,44 22, , , Assets The average amount of assets in our sample on 174 companies is 249,247 thousand Euros, whereas the median company only has assets worth 40,769 thousand Euros. There distribution has a large standard deviation of 616,553 thousand Euros, which again is a result of infrequent deviations which is also seen in the kurtosis value on At the same time the distribution is positive skewed with a skewness of There is a large difference on the average amount of assets in the companies between countries, where the Finnish companies seems to be largest measured on assets with a mean on thousand Euros and the Norwegian companies is smallest measured on assets with a mean on thousand Euros. The group of Norwegian companies is most normally distributed with a positive skewness on only 1.56 and the group is not affected by high deviations with a kurtosis value on only Descriptive statistics on survey sample questionnaire This section of the thesis presents detailed descriptive data for the survey responses. For each question the answers presented are divided into country specific responses in order to highlight potential differences between the four countries in our sample. The number of respondents as well as the equivalent percentage is shown for each country and for the sample as a whole. Furthermore, where possible and relevant, the mean for each country and for the sample is calculated and compared by means of a t-test in order to test for statistical differences between the countries. Most of the survey results 40

44 presented includes answers from all 174 companies in the sample. However, due to the nature of the intelligent questionnaire used for the collection of the data, certain questions will exhibit a lower number of respondents due to the fact that some companies were allowed to skip the question as a result of answers given previously in the survey. Furthermore, comparisons will be made with surveys conducted by Hagelin (2003) and Pramborg (2005) on Swedish firms as well as with Aabo, Simkins (2003). The first section presets the survey results relating to I: international involvement and foreign exchange rate exposure of the companies, while the second and third sections present the results relating to II: debt and derivatives use for managing exchange rate exposures and III: firm flexibility and industry stability. To ease and create an overview of the answers, a concise conclusion is found in the end of each section International involvement and foreign exchange rate exposure This section Table 7 presents the answers to questions 1a and 1b of the survey and relate to the degree of trade that the firms have with foreign firms. With a total of 83,3% of the survey respondents indicating that at least 1% of their consolidated revenues are in foreign currency and a mean of 39,95%, it seems that the firms in general are quite involved internationally, which was to be expected as they operate in small, open economies with limited domestic demands. In terms of the distribution across the intervals, 39,7% of the respondents have between 1-40 % of their consolidated revenues in foreign currency while another 23,6% have a share of foreign revenues which exceed 81%. With an average of 48,57%, Danish firms have the highest share of revenues in foreign costs while Finnish firms have lowest at 25,79%. A contributory cause to the low Finnish value may be that Finland, as the only country, has adopted the Euro, for which reason it is not exposed to 55% of its export trading partners 18. Also, it is interesting to note that 30% of the Danish companies have between 81-99% of consolidated revenues in foreign costs (21,3% total), which suggest a high degree of foreign involvement. In terms of consolidated costs in foreign currency, the picture is quite similar. 90,8% of the companies have foreign denominated cost and only 9,2% of the companies have no foreign costs. The average firm has 32,66% of their costs in foreign currency with 56,9% of the companies having between 1-40%. The distribution across countries is

45 similar to that of revenues with Danish firms averaging the highest share (39,96%) and Finnish firms the lowest (19,13%). Table 7: Survey results for questions 1a & 1b This table presents the survey results for question 1a: Approximately what percentage of your company s consolidated revenues is in foreign currency? and question 1b: Approximately what percentage of your company s consolidated costs is in foreign currency? The table presents the answers for the total population of 174 companies according to country, and calculates the mean using the midpoints of the intervals from 0% - 100%. Q1a Consolidated revenues Q1b Consolidated Costs Denmark Sweden Norway Finland Total Denmark Sweden Norway Finland Total 0% 9 (12,9%) 12 (18,5%) 5 (18,5%) 3 (25%) 29 (16,7%) 5 (7,1%) 6 (9,2%) 4 (14,8%) 1 (8,3%) 16 (9,2%) 1-20% 16 (22,9%) 23 (35,4%) 6 (22,2%) 3 (25,3%) 48 (27,6%) 21 (30%) 25 (38,5%) 12 (44,4%) 8 (66,7%) 66 (37,9%) 21-40% 6 (8,6%) 8 (12,3%) 3 (11,1%) 4 (33,3%) 21 (12,1%) (14,8%) 2 33 (19%) (14,3%) (26,2%) (16,7%) 41-60% 6 (8,6%) 6 (9,2%) 3 (11,1%) 0 (0%) 15 (8,6%) 12 5 (7,7%) 3 (11,1%) 0 (0%) 20 (11,5%) (17,1%) 61-80% 10 5 (7,7%) 4 (14,8%) 1 (8,3%) 20 (11,5%) 15 7 (10,8%) 2 (7,4%) 0 (0%) 24 (13,8%) (14,3%) (21,4%) 81-99% 21 (30%) 9 (13,8%) 6 (22,2%) 1 (8,3%) 37 (21,3%) 6 (8,6%) 5 (7,7%) 2 (7,4%) 1 (8,3%) 14 (8%) 100% 2 (2,9%) 2 (3,1%) 0 (0%) 0 (0%) 4 (2,3%) 1 (1,4%) 0 (0%) 0 (0%) 0 (0%) 1 (0,6%) Total 70 (100%) 65 (100%) 27 (100%) 12 (100%) 174 (100%) 70 (100%) 65 (100%) 27 (100%) 12 (100%) 174 (100%) Mean 48,57% 32,70% 41,37% 25,79% 39,95% 39,96% 29,96% 26,26% 19,13% 32,66% Even though the largest gaps in mean values are between Danish and Finnish firms, the only statistical significant 19 difference in the mean values is between Danish and Swedish firms in terms of their share of consolidated revenues in a foreign currency. Aside from this, no statistical differences exist between the different countries in respect to both consolidated revenues as well as costs. A possible explanation for the lack of a significant statistical difference between Denmark and Finland even though Finland exhibits a lower mean than Sweden is that the response rate for Finnish firms is low compared to that of the Swedish firms, which reduces the statistical power. Table 8: Survey results for question 2 The table presents the survey results for question 2: Approximately what percentage of your company s consolidated assets is in foreign currency? The table presents the answers for the total population of 174 companies according to country, and calculates the mean using the midpoints of the intervals from 0%- 100%. Q2 Consolidated assets Denmark Sweden Norway Finland Total Consolidated 0% 17 (14,3%) 18 (27,7%) 10 (37%) 1 (8,3%) 46 (26,4%) Assets 1-20% 22 (31,4%) 31 (47,7%) 7 (25,9%) 9 (75%) 69 (39,7%) 21-40% 9 (12,9%) 4 (6,2%) 3 (11,1%) 1 (8,3%) 17 (9,8%) 41-60% 8 (11,4%) 5 (7,7%) 2 (7,4%) 1 (8,3%) 16 (9,2%) 61-80% 7 (10%) 5 (7,7%) 4 (14,8%) 0 (0%) 16 (9,2%) 81-99% 7 (10%) 2 (3,1%) 1 (3,7%) 0 (0%) 10 (5,7%) 100% 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) Total 70 (100%) 65 (100%) 27 (100%) 12 (100%) 174 (100%) 19 Significant at 10% level 42

46 Mean 28,66% 18,60% 23,31% 14,17% 23,07% Table 8 shows that the average company has an average of 23,07% of their total assets in foreign currencies. As was the case with the two previous questions, Danish firms have the highest share (28,66%) of assets in foreign currency while Finnish firms have the lowest (14,17%). In general, firms tend to have a low share of assets in foreign currency with 39,7% having between 1-20% of total assets in foreign currency while 26,4% have no foreign assets at all. These results suggest that it is more important for Scandinavian firms to have an international presence through business transactions than it is to have an actual physical presence abroad. In terms of differences across countries, we find that no statistical differences exist between in respect to their amount of assets in foreign currency. The survey responses concerning the average time horizon for which the companies have covered their foreign exchange rate exposure using financial means is presented in table 9. The table includes the answers from 128 firms (74%) that confirmed the use of either foreign debt or currency derivatives. Table 9: Survey results for question 3 The table presents the survey results for question 3: At the present time, what is the average time horizon that your company has covered its foreign exchange rate exposures by use of financial means? (i.e. forward contracts, options, swaps and debt in foreign currency) The table presents the answers for the 128 companies that use either short- or long-sighted derivatives or foreign debt. Firms are divided according to country and mean values are calculated using the midpoints of the intervals.* Q3 Average time horizon of financial hedges Denmark Sweden Norway Finland Total Average 0-1 months 15 (28,3%) 27 (54,0%) 5 (29,4%) 1 (12,5%) 48 (37,5%) time horizon 1-3 months 10 (18,9%) 9 (18,0%) 2 (11,8%) 2 (25%) 23 (18,0%) in general 3-6 months 6 (11,3%) 4 (8,0%) 4 (23,5%) 1 (12,5%) 15 (11,7%) 6-12 months 17 (32,1%) 7 (14,0%) 2 (11,8%) 2 (25,0%) 28 (21,9%) 1-2 years 3 (5,7%) 1 (2,0%) 3 (17,6%) 2 (25,0%) 9 (7,0%) 2-4 years 1 (1,9%) 1 (2,0%) 1 (5,9%) 0 (0%) 3 (2,3%) >4 years 1 (1,9%) 1 (2,0%) 0 (0%) 0 (0%) 2 (1,6%) Total 53 (100%) 50 (100%) 17 (100%) 8 (100%) 128 (100%) Mean (months) 6,8 4,5 7,8 7,9 6,1 *the mean is calculated by midpoints in the interval, so 0-1 months equals 0,5, 1-3 months equals 2, 3-6 months equals 4,5, 6-12 months equals 9, 1-2 years equals 18 months, 2-4 years equals 36 months and 4> years is set to equal 48 months As seen in the table, the average firm in our survey has covered its foreign exchange rate exposure by use of financial means for 6,8 months. Norwegian and Finnish firms cover their exposure the longest with averages on exposure at 7,8 months and 7,9 months respectively, while Swedish firms have the shortest average horizon with 4,5 months. Most of the firms indicate that they cover their exchange rate exposure by 43

47 financial means for 0-1 months (37,5%). The majority of the companies (89.9%) cover its exchange rate exposures for less than a year while 3,9% of the companies answered that they cover their exposure by financial means for more than 2 years. Survey questions 4 and 5 were designed to investigate the degree of international involvement of the companies in terms of their presence in foreign countries, represented by the number of sales and production subsidiaries as well as the number of currencies the companies were exposed to. In terms of production subsidiaries abroad (table 10), approximately 55% of the responding companies stated that they do not have production subsidiaries abroad while another 35,6% have between 1-5 production subsidiaries located in foreign countries. On average, the companies have 1,98 production subsidiaries abroad. Danish firms have almost three times as many production subsidiaries as Finnish firms with 2,44 compared to 0,83. However, no significant statistical differences are observed between the countries. The large difference between Finnish and Danish firms is consistent with our findings from question 2 that showed that Danish firms have a much larger share of their assets in foreign currency. Table 10: Survey results for question 4a and 4b The table presents the survey results for question 4a: How many production subsidiaries does your company have abroad and question 4b: How many sales subsidiaries does your country have abroad The table presents the answers for the total population of 174 companies according to country, and calculates the mean using the midpoints of the intervals. Q4a Production subsidiaries Q4b Sales subsidiaries Denmark Sweden Norway Finland Total Denmark Sweden Norway Finland Total 0 36 (51,4%) 38 (58,5%) 15 (55,6%) 7 (58,3%) 96 (55,2%) 18 (25,7%) 28 (43,1%) 13 (48,1%) 6 (50%) 65 (37,4%) (14,8%) (14,8%) 3 (25%) 36 (20,7% (22,9%) (30,8%) (33,3%) (25,3%) (21,4%) (21,5%) (12,9%) 2 (3,1%) 6 (22,2%) 1 (8,3%) 18 (10,3%) 12 (17,1%) 10 (15,4%) 7 (25,9%) 1 (8,3%) 30 (17,2%) (5,7%) 1 (1,5%) 1 (3,7%) 0 (0%) 6 (3,4%) 11 (15,7%) 8 (12,3%) 1 (3,7%) 0 (0%) 20 (11,5%) (2,9%) 2 (3,1%) 1 (3,7%) 0 (0%) 5 (2,9%) 5 (7,1%) 3 (4,6%) 0 (0%) 0 (0%) 8 (4,6%) (1,4%) 1 (1,5%) 0 (0%) 0 (0%) 2 (1,1%) 3 (4,3%) 1 (1,5%) 1 (3,7%) 2 7 (4%) (16,7%) >24 2 (2,9%) 1 (1,5%) 0 (0%) 0 (0%) 3 (1,7%) 6 (8,6%) 1 (1,5%) 1 (3,7%) 0 (0%) 8 (4,6%) Total 70 (100%) 65 (100%) 27 (100%) 12 (100%) 174 (100%) 70 (100%) 65 (100%) 27 (100%) 12 (100%) 174 (100%) Mean 2,44 1,76 1,85 0,83 1,98 5,99 3,11 3,17 4,04 4,34 In terms of sales subsidiaries, a larger proportion of the companies have sales subsidiaries abroad compared to production subsidiaries. 62,6% of the responding companies have sales subsidiaries abroad with the majority having between 1 and 9. The average company has 4,34 sales subsidiaries located abroad, which is more than twice the average for production subsidiaries. The reason for this is likely to be found in 44

48 the significantly higher setup and construction costs associated with setting up production subsidiaries as opposed to sales subsidiaries. A significant difference 20 in the means exist between Danish firms which have the most sales subsidiaries abroad with 5,99 on average and Swedish firms which have the fewest with an average of 3,11. Regarding the number of significant currencies that a company is significantly exposed to (table 11), our findings show that most of the companies (81,1%) are significantly exposed to 1-5 currencies while only 8% have no exposure at all. On average, firms are exposed to 3,22 currencies with Danish firms being the most exposed (3,72) and Finnish firms the least (2,5). Table 11: Survey results for question 5 This table presents the survey responses for question 5: How many foreign currencies is your company significantly exposed to? The table presents the answers for the total population of 174 companies according to country, and calculates the mean using the midpoints for the intervals with 9 as the highest value possible. Q 5 Number of currencies Denmark Sweden Norway Finland Total Number of 0 4 (5,7%) 6 (9,2%) 3 (11,1%) 1 (8,3%) 14 (8%) significant (31,4%) 24 (36,9%) 8 (29,6%) 7 (58,3%) 61 (35,1%) currencies (44,3%) 31 (47,7%) 15 (55,6%) 3 (25%) 80 (46%) (12,9%) 3 (4,6%) 1 (3,7%) 1 (8,3%) 14 (8%) >9 4 (5,7%) 1 (1,5%) 0 (0%) 0 (0%) 5 (2,9%) Total 70 (100%) 65 (100%) 27 (100%) 12 (100%) 174 (100%) Mean 3,72 2,95 2,94 2,5 3,22 As can be seen from table 12 below, the most important currency for the sample of responding companies is the Euro. 51% of the responding firms stated the Euro as the most important currency in terms of exposure, while the Euro was the second most important currency for 29% of the companies. Table 12: Survey results for question 6 This table presents the survey responses for question 6: Which foreign currencies are the most important ones? The table shows the answers for the four most important currencies. The number of respondents varies for each answer. Most important currency Q 6 Importance of currencies 1st 2nd 3rd 4th Euro 85 (51%) 45 (29%) 16 (12%) 4 (4%) DKK 16 (10%) 16 (10%) 11 (8%) 12 (11%) NOK 14 (8%) 7 (5%) 26 (19%) 11 (10%) SEK 10 (6%) 13 (8%) 16 (12%) 11 (10%) Pound 5 (3%) 14 (9%) 22 (16%) 21 (19%) $Dollar 31 (18%) 50 (33%) 20 (15%) 20 (19%) Yen 0 (0%) 2 (1%) 8 (6%) 6 (6%) Rouble 2 (1%) 0 (0%) 3 (2%) 3 (3%) Other 5 (3%) 6 (6%) 13 (10%) 20 (19%) 20 Significant at 10% level 45

49 Total 168 (100%) 153 (100%) 135 (100%) 108 (100%) The dollar is the most important and second most important for 18% and 33% respectively, while the Danish Krone is the third most important currency followed by the Swedish Krona and the British Pound. The answers to the different categories does not completely correspond to the answers given to question 5 where the companies were asked about the number of currencies they were significantly exposed to. While 75 of the responding companies stated that they were exposed to two or less currencies (leaving = 99), the answers with regards to the 3 rd most important currency contains 135 answers. In order to assess the real exposure (net exposure) that companies face and to investigate the use of currency matching as a strategic hedge, questions 7 and 8 investigate the relationship between foreign currency revenues and costs. The question on whether a company s foreign currency revenues had been larger or smaller than corresponding costs within the last year (table 13) show that the majority of companies (70,7%) had foreign currency revenues larger than costs and the answers were fairly consistent across countries. Table 13: Survey results for question 7 This table presents the survey responses for question 7: Within the last year, has your company s foreign operating revenues been larger or smaller than your company s foreign operating costs? The table presents the answers for the total population of 174 companies according to country. Q7 Foreign operating revenues vs. costs Denmark Sweden Norway Finland Total Foreign operating revenues larger than foreign operating costs 48 (68,6%) 44(67,7%) 21 (77,8%) 10 (83,3%) 123 (70,7%) Foreign operating costs larger than foreign operating revenues 22 (31,4%) 21 (32,3%) 6 (22,2%) 2 (16,7%) 51 (29,3%) Total 70 (100%) 65 (100%) 27 (100%) 12 (100%) 174 (100%) Following the answers given to question 7, the companies were asked about the extend to which they match foreign currency revenues/costs with corresponding costs/revenues the answers. These answers are presented in table 14. From table 14 below, we see that companies with foreign revenues larger than costs, match 39,03% of these on average while companies with costs exceeding revenues match 30,12% on average. The answers given are evenly distributed across both country and interval levels with the exception of Finnish firms in Q8b. However, with only two respondents, not much meaning can be derived from the mean of 5%. Overall, we find 46

50 that 77,6% of the respondents use matching at least to some extend. No significant statistical differences in either of the two questions. Table 14: Survey results for question 8a and 8b This table presents the survey responses for question 8a: How much of your company s foreign operating revenues are offset by matching foreign operating costs? and question 8b: How much of your company s foreign operating costs are offset by matching foreign operating revenues? The table presents the answers for the 123 and 51 responding companies according to country, and calculates the mean using the midpoints of the intervals. Q8a Foreign operating revenues matched by costs Q8b Foreign operating costs matched by revenues Denmark Sweden Norway Finland Total Denmark Sweden Norway Finland Total 0% 5 (10,4%) 8 (18,2%) 4 (19%) 1 (10%) 18 (14,6%) 8 (36,4%) 9 (42,9%) 3 (50%) 1 (50%) 21 (41,2%) 1-20% 11 8 (18,2%) 5 3 (30%) 27 (22%) 6 (27,3%) 2 (9,5%) 0 (0%) 1 (50%) 9 (17,6%) (22,9%) (23,8%) 21-40% 11 8 (18,2%) 3 2 (20%) 24 (19,5%) 3 (13,6%) 1 (4,8%) 0 (0%) 0 (0%) 4 (7,8%) (22,9%) (14,3%) 41-60% 10 4 (9,1%) 2 (9,5%) 2 (20%) 18 (14,6%) 1 (4,5%) 3 (14,3%) 0 (0%) 0 (0%) 4 (7,8%) (20,8%) 61-80% 6 (12,5%) 4 (9,1%) 4 (19%) 1 (10%) 15 (12,2%) 0 (0%) 2 (9,5%) 1 (16,7%) 0 (0%) 3 (5,9%) 81-99% 4 (8,3%) 11 (25%) 2 (9,5%) 1 (10%) 18 (14,6%) 4 (18,2%) 2 (9,5%) 2 (33,3%) 0 (0%) 8 (15,7%) 100% 1 (2,1%) 1 (2,3%) 1 (4,8%) 0 (0%) 3 (2,4%) 0 (0%) 2 (9,5%) 0 (0%) 0 (0%) 2 3,9%) Total 48 (100%) 44 (100%) 21 (100%) 10 (100%) 123 (100%) 22 (100%) 21 (100%) 6 (100%) 2 (100%) 51 (100%) Mean 37,88% 41,74% 38,05% 34,95% 39,03% 25,36% 34,24% 41,50% 5% 30,12% Concise conclusion on chapter The purpose of the above section was to examine the international involvement and foreign exchange rate exposure of the companies in our sample by looking at their international operations. We find that 83,3% of the companies have revenues in foreign currencies and that the average firm has 40% of the total revenues in foreign currencies. In terms of costs in foreign currency, 90,8% of the companies have such cost with an average of 32,7% per firm. These findings are similar to the results found by Pramborg (2005) who find that the average Swedish firm in his sample has 43,5% of its revenues in foreign currency while foreign costs amount to 34,3% of total costs. Other indicators of the international involvement of companies show that the average company has 23% of total asset in foreign currencies, that 45% have production subsidiaries abroad (avg. 1,98) while 62,6% of the companies have foreign sales subsidiaries abroad (avg. 4,34). The average company in our sample was exposed to 3,22 currencies with the Euro and the U.S dollar as the most important ones. In terms of offsetting currency risks by matching foreign currency inflows with corresponding outflows, 77,6% of the companies answered that they match at least to some degree. On average, companies with foreign revenues exceeding foreign costs 47

51 match 39,03% of these revenues while companies with costs exceeding revenues in foreign currencies match 30,12% on average. The study by Pramborg (2005) also examined the extend to which Swedish firms match inflows and outflows in foreign currencies. Pramborg found that roughly 72% of the Swedish firms had used matching at some point while roughly 60% used currency matching frequently. These results are quite similar to the ones found here. Lastly, this section presents the answers to the question regarding the average time horizon that the companies in our survey have covered their foreign exchange rate exposure using financial means. We find that the average company has a time horizon of 6,1 months Debt and derivatives usage for managing foreign exchange rate exposure This section of the chapter presents the survey findings to part two of the questionnaire relating to the debt and derivatives use of the companies. The first question in section two relates to the foreign debt ratios of our sample of companies. (table 15) The results show that of the 174 companies, 37,4% do not have any debt denominated in foreign currency. An additional 45,4% have between 1-40% of there consolidated debt in foreign currency. The average firm has a foreign debt ration of 18,43%. In terms of differences between countries, Danish firms have the highest share of total debt in foreign currency with an average of 25,55% which is statistically different 21 from that of Swedish firms which average 12,62%. Otherwise, there are no significant statistical differences between the means of the four countries. Table 15: Survey results for question 9 This table presents the survey responses for question 9: Approximately what percentage of your company s consolidated debt is in foreign currency? The table presents the answers for the total population of 174 companies according to country, and calculates the mean using the midpoints of the intervals from 0% - 100% Q9 Consolidated foreign debt Denmark Sweden Norway Finland Total Consolidated 0% 19 (27,1%) 25 (38,5%) 13 (48,1%) 8 (66,7%) 65 (37,4%) foreign 1-20% 18 (25,7%) 30 (46,2%) 5 (18,5%) 1 (8,3%) 54 (31%) debt 21-40% 14 (20%) 4 (6,2%) 5 (18,5%) 2 (16,7%) 25 (14,4%) 41-60% 10 (14,3%) 1 (1,5%) 3 (11,1%) 0 (0%) 14 (8%) 61-80% 6 (8,6%) 5 (7,7%) 0 (0%) 0 (0%) 11 (6,3%) 81-99% 3 (4,3%) 0 (0%) 1 (3,7%) 1 (8,3%) 5 (2,9%) 100% 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) Total 70 (100%) 65 (100%) 27 (100%) 12 (100%) 174 (100%) Mean 25,55% 12,62% 16,28% 13,29% 18,43% 21 Significant at 1% level 48

52 The question regarding the average loan period of the company s foreign debt arrangements is presented in table 16 and illustrates that of the 174 responding, 109 companies (62.6%) use foreign debt. The average time horizon for these foreign debt arrangements is 16,6 months which should be seen in contrast to the very large share (47,7%) of the companies with debt maturing within three months. It is noteworthy that only one company had foreign debt with a loan period of 3-6 months and only 9 (8,3%) had a loan period of 6-12 months. It seems that the Scandinavian companies are either borrowing in foreign denominated for currencies for a quite long period or for a very short periods of time. Table 16: Survey results for question 10 This table presents the survey results for question 10: What is the average loan period for your company s foreign debt arrangements. The table presents the answers for the population of 109 companies that used foreign debt. The companies are divided according to country and the mean is calculated using the midpoints of the intervals*. Average time horizon of foreign debt Q10 Average time horizon of foreign debt Denmark Sweden Norway Finland Total 0-1 months 7 (13,7%) 16 (40%) 1 (7,1%) 0 (0%) 24 (22%) 1-3 months 13 (25,5%) 9 (22,5%) 4 (28,6%) 2 (50%) 28 (25,7%) 3-6 months 0 (0%) 0 (0%) 1 (7,1%) 0 (0%) 1 (0,9%) 6-12 months 5 (9,8%) 4 (10%) 0 (0%) 0 (0%) 9 (8,3%) 1-2 years 7 (13,7%) 3 (7,5%) 3 (21,3%) 1 (25%) 14 (12,8%) 2-4 years 9 (17,6%) 3 (7,5%) 3 (21,3%) 0 (0%) 15 (13,8%) >4 years 10 (19,6%) 5 (12,5%) 2 (14,2%) 1 (25%) 18 (16,5%) Total 51 (100%) 40 (100%) 14 (100%) 4 (100%) 109 (100%) Mean 19,7 11,6 19,36 17,5 16,6 *the mean is calculated by midpoints in the interval, so 0-1 months equals 0,5, 1-3 months equals 2, 3-6 months equals 4,5, 6-12 months equals 9, 1-2 years equals 18 months, 2-4 years equals 36 months and 4> years is set to equal 48 months. Swedish firms stand out from the rest with an average loan period of 11,6 months compared to an average of 19,7 months and 19,36 months for Danish and Norwegian firms respectively. However, it is to be mentioned that no significant differences are observed between the countries. The next question in the questionnaire, for which the responses are not presented, examine whether the responding firm had used either short-sighted (forward contracts or options) or long-sighted derivatives (swaps) within the last two years. Out of the total population of 174 firms, 64 firms (37%) stated that they used short-sighted while only 34 firms (20%) confirmed that they used long-sighted derivatives. Thus, the following two questions relate to the average maturity of these two types of derivatives. The average contract period for short-term derivatives used to cover exchange rate exposure is presented in table 17. The average contract period of the short-term derivatives was 6,3 months, with 90,6% of the companies having contracts, with 49

53 maturities of less than a year. The most frequently hedged contract period was 3-6 months (32,8%) followed by contracts of 1-3 months (26,6%) and 6-12 months (25%) No companies hedged for more than 4 years, using these derivatives. These findings support our hypothesis that options and forwards are primarily used to manage shortterm exchange rate exposure. Table 17: Survey results for question 12 This table presents the survey results for question 12: What is the average maturity for short-sighted derivatives (forward contracts and options) used to cover your company s foreign exchange rate exposure. The table presents the answers for the 64 responding companies according to country and calculates the mean using the midpoints of the intervals*. Average maturity for short-term derivatives Q 12 Average maturity for short-sighted derivatives Denmark Sweden Norway Finland Total 0-1 months 1 (3,3%) 3 (16,7%) 0 (0%) 0 (0%) 4 (6,3%) 1-3 months 7 (23,3%) 6 (33,3%) 3 (30%) 1 (16,7%) 17 (26,6%) 3-6 months 12 (40%) 3 (16,7%) 4 (40%) 2 (33,3%) 21 (32,8%) 6-12 months 10 (33,3%) 4 (22,2%) 1 (10%) 1 (16,7%) 16 (25%) 1-2 years 0 (0%) 1 (5,6%) 2 (20%) 2 (33,3%) 5 (7,8%) 2-4 years 0 (0%) 1 (5,6%) 0 (0%) 0 (0%) 1 (1,6%) Total 30 (100%) 18 (100%) 10 (100%) 6 (100%) 64 (100%) Mean (months) 5,3 6,5 6,9 9,3 6,3 *the mean is calculated by midpoints in the interval, so 0-1 months equals 0,5, 1-3 months equals 2, 3-6 months equals 4,5, 6-12 months equals 9, 1-2 years equals 18 months, 2-4 years equals 36 months and 4> years is set to equal 48 months. The Danish companies have the shortest average contract lengths (5,3 months) while finish companies have the longest (9,3 months). Although the difference may seem large, the difference is not significant, which is likely to be caused by the small finish sample of only six respondents. Table 18 presents the answers given to the question regarding the average maturity for long-sighted derivatives (swaps) used to manage a company s foreign exchange rate exposure. The average contract period of the contracts are 22,5 months which is considerably higher than the average found for short-term contracts (options/forwards), which had an average length of 6,3 months. Table 18: Survey results for question 13 This table presents the survey results for question 13: What is the average maturity for long-sighted derivatives (swaps) used to cover your company s foreign exchange rate exposure. The table presents the answers for the 34 respondents according to country and calculates the mean using the midpoints of the intervals *. Average maturity for long-term derivatives Q 13 Average maturity for long-sighted derivatives Denmark Sweden Norway Finland Total 1-3 months 0 (0%) 1 (14,3%) 0 (0%) 0 (0%) 1 (2,9%) 3-6 months 1 (5,3%) 1 (14,3%) 1 (16,7%) 0 (0%) 3 (8,8%) 6-12 months 1 (5,3%) 2 (28,6%) 1 (16,7%) 2 (100%) 6 (17,6%) 1-2 years 7 (36,8%) 1 (14,3%) 4 (66,7%) 0 (0%) 12 (35,3%) 2-4 years 6 (31,6%) 2 (28,6%) 0 (0%) 0 (0%) 8 (23,5%) 50

54 >4 years 4 (21,1%) 0 (0%) 0 (0%) 0 (0%) 4 (11,8%) Total 19 (100%) 7 (100%) 6 (100%) 2 (100%) 34 (100%) Mean (months) 28,8 16,4 14,3 9 22,5 *the mean is calculated by midpoints in the interval, so 0-1 months equals 0,5, 1-3 months equals 2, 3-6 months equals 4,5, 6-12 months equals 9, 1-2 years equals 18 months, 2-4 years equals 36 months and 4> years is set to equal 48 months. Of the 34 companies who used swaps, 29,3% had swap contracts with a maturity of less than a year while 70,7% had swap obligations with a period above one year. The most frequently used contract period was 1-2 years (35,3%) followed by contracts with a length of 2-4 years (23,5%). No significant statistical differences exist between the countries, which is likely to be caused by the low number of respondents. The next question in the survey had the objective of examining the relative importance of financial and operational hedges in terms of the time horizon for which they are considered important for managing foreign exchange rate exposure. Table 19 below present the answers to this question. Table 19: Survey results for question 14 This table presents the survey results for question 14: Generally financial means (e.g. forward contracts, options, swaps and foreign debt) are used to hedge the short and medium term, while operating means (e.g. changing suppliers and moving production) are used to hedge the long term. In the case of your firm, at what time horizon would you say that the importance of operational means exceeds the importance of financial means for protecting your company against adverse changes in exchange rates? The table presents the answers for the total population of 174 companies according to country and calculates the mean using the midpoints of the intervals*. Importance of operational means exceed financial means Q 14 Financial vs. operational means Denmark Sweden Norway Finland Total 0-1 months 20 (28,6%) 22 (33,8%) 6 (22,2%) 4 (33,3%) 52 (29,9%) 1-3 months 14 (20%) 7 (10,8%) 1 (3,7%) 1 (8,3%) 23 (13,2%) 3-6 months 10 (14,3%) 6 (9,2%) 4 (14,8%) 2 (16,7%) 22 (12,6%) 6-12 months 8 (11,4%) 10 (15,4%) 2 (7,4%) 2 (16,7%) 22 (12,6%) 1-2 years 8 (11,4%) 9 (13,8%) 5 (18,5%) 1 (8,3%) 23 (13,2%) 2-4 years 4 (5,7%) 4 (6,2%) 6 (22,2%) 2 (16,7%) 16 (9,2%) >4 years 6 (8,6%) 7 (10,8%) 3 (11,1%) 0 (0%) 16 (9,2%) Total 70 (100%) 65 (100%) 27 (100%) 12 (100%) 174 (100%) Mean (months) 10,4 12,1 18,2 10,1 12,2 *the mean is calculated by midpoints in the interval, so 0-1 months equals 0,5, 1-3 months equals 2, 3-6 months equals 4,5, 6-12 months equals 9, 1-2 years equals 18 months, 2-4 years equals 36 months and 4> years is set to equal 48 months. On average, the responding companies considers the importance of operational means to exceed that of financial means for managing foreign exchange rate exposure for exposures above 12,2 months. In other words our sample of companies considers operational means (e.g. changing suppliers) to be more important as tool of hedging exchange rate exposure, when the exposure exceeds one year. 29,9% of the responding companies stated that the importance of operational means exceeded the financial means after 0-1 months while about 50% of the remaining answers were evenly distributed among the intervals 1-3 months, 3-6 months, 6-12 months and 1-2 years. 51

55 9,2% of the participating companies considered financial means more important than financial means for managing exposures of more than 4 years, which indicate very little flexibility. In terms of the distribution across countries, the Finnish companies had the shortest horizon (10,2 months) while Norwegian companies believed that financial means were better suited to manage exposures of up to 18,2 months Concise conclusion on chapter The aim of this section was to highlight the debt and derivatives use of the companies in our sample and their corresponding time horizons. We found that the average firm in our sample had 18,5% of its debt denominated in foreign currency with a total of 63% firms stated that that they used foreign debt. Similarly, in his study on foreign exchange risk management by Swedish and Korean non-financial firms, Pramborg (2005) find that 71% of Swedish firms used debt for hedging purposes. Likewise, Hagelin (2003) finds that 53% of the Swedish firms in his sample used debt denominated in foreign currency to manage their foreign exchange exposure. In our sample, 61,5% of the Swedish firms use foreign denominated debt. In terms of the use short-term derivatives, 36,8% of the companies indicated that they used forward contracts or options to manage their exchange rate exposure with an average maturity of these contracts of 6,3 months. Furthermore, 19,5% of the companies indicated that they used swaps to manage foreign exchange rate exposure. The average maturity of these contracts was 22,5 month. Finding a higher frequency of use, Hagelin (2003) finds that 60% of the Swedish firms in his sample used currency derivatives for hedging while Pramborg (2005) found that roughly 79% of Swedish firms use forward contracts while 75% used swaps which are significantly higher proportions than found in this study. The following figure summarizes the results of the survey in terms of the average time horizon for the different financial instruments used to hedge foreign exchange rate exposure Table 20: Summary of findings Thus table summarizes the answers to questions 10, 12 & 13 for the total sample used in each question. Average maturity of...foreign currency denominated debt 16,6 months short-term derivatives (forward contracts, options) 6,3 months...long-term derivatives (swaps) 22,5 months 52

56 These finding confirm our assumptions and earlier findings that debt and swaps are used to hedge long-term foreign exchange rate exposure whereas short-term derivatives (forward contracts and options) are used to hedge short-term exchange rate exposure. Section also examined the time horizon for which managers consider the benefits of operational means (change suppliers, move production) to exceed that of financial means. The results indicate that for exposures longer than one year (12,2 months), operational means are considered superior to financial means when managing exchange rate exposure. Whether this horizon is consistent with the actual time horizon that company s hedge using financial means, provides an interesting topic for further analysis. The results of this thesis indicate that the two horizons do not coincide. (6,1 vs. 12,2 months) Firm flexibility and industry stability The third section of the survey was included to investigate the flexibility of Scandinavian companies and the stability of the industries in which they operate. In question 15 (table 21) the companies were asked to characterize the general business environment in which they primarily operated. As the data for this question is ordinal (answer on likerts scale), it makes no sense to calculate mean values, as the gap between two answers is not necessarily the same. 48,8% of the responding companies characterize their business environment as stable or fairly stable while 28,2% describe it as volatile or fairly volatile. Table 21: Survey results for question 15 This table presents the survey results for question 15: How would you characterize the general business environment in which your company primarily operates? The table presents the answers for the total population of 174 companies according to country. Stability of business environment Q15 Business environment Denmark Sweden Norway Finland Total Very stable 5 (7,1%) 9 (13,8%) 1 (3,7%) 0 (0%) 15 (8,6%) Fairly stable 23 (32,9%) 35 (53,8%) 8 (29,6%) 4 (33,3%) 70 (40,2%) Not particularly stable or 20 (28,6%) 8 (12,3%) 9 (33,3%) 3 (25%) 40 (23%) volatile Fairly volatile 20 (28,6%) 10 (15,4%) 6 (22,2%) 4 (33,3%) 40 (23%) Very volatile 2 (2,9%) 3 (4,6%) 3 (11,1%) 1 (8,3%) 9 (5,2%) Total 70 (100%) 65 (100%) 27 (100%) 12 (100%) 174 (100%) In terms of national differences, Swedish companies describe their business environment as quit stable with more than 53% of the companies answering that they 53

57 find the business environment fairly stable. Aside from this, the answers are pretty evenly distributed across the countries for each category. The ability of a firm to use real options serves as an indicator of how flexible and well equipped a company is in terms of managing foreign exchange rate exposure using operational means. Table 22 describes the extend to which the responding companies are able to exploit different real options, in the coming 4-year period as a response to exchange rate changes. Table 22: Survey results for question 16 This table presents the survey results for question 16: In the coming 4 years is it likely that your company will undertake any of the following actions partly or fully due to exchange rate changes? The table presents the answers for the total population of 174 companies according to country and calculates the mean depending on how many numbers of yes the respondents have stated. The respondents had 7 different real actions that they should respond to. These 7 are seen in table 16 Numbers of real options likely to be undertaken within the next 12 months Q 16 Number of real options Denmark Sweden Norway Finland Total 0 19 (27,1%) 26 (40,0%) 6 (22,2%) 4 (33,3%) 55 (31,6%) 1 15 (21,4%) 11 (16,9%) 3 (11,1%) 4 (33,3%) 33 (19,0%) 2 6 (8,6%) 7 (10,8%) 6 (22,2%) 3 (25%%) 22 (12,6%) 3 13 (18,6%) 6 (9,2%) 5 (18,5%) 1 (8,3%) 25 (14,4%) 4 10 (14,3%) 10 (15,4%) 4 (14,8%) 0 (0%) 24 (13,8%) 5 4 (5,7%) 3 (4,6%) 2 (7,4%) 0 (0%) 9 (5,2%) 6 1 (1,4%) 2 (3,1%) 1 (3,7%) 0 (0%) 4 (2,3%) 7 2 (2,9%) 0 (0%) 0 (0%) 0 (0%) 2 (1,1%) Total 70 (100%) 65 (100%) 27 (100%) 12 (100%) 174 (100%) Mean 2,1 1,7 2,3 1,1 1,9 The average company in our sample of companies is likely to partly or fully undertake 1,9 of the seven different real options, in order to manage exchange rate changes, with Norwegian companies likely to use the most real options (2,3) and the Finnish companies the least (1,1). Table 23 illustrates the seven different types of real options that a firm may exploit in order to react to changes in exchange rates. 50,6% of the companies would consider changing its sourcing within the next four years due to exchange rate changes, making it the most popular real option strategy. The second most popular strategy of managing foreign exchange rate exposure by operational means is to increase current capacity or to extend present capacity, which 43,1% of the companies would consider. A little more than a third (36,2%) might enter a foreign market in case of a positive or negative exchange rate development while about a fifth would consider shifting production (22,4%) or temporarily reducing /closing down operations (21,8%). The least popular real option is to delay an entry to a foreign market (8,6%) totally abandoning a foreign market (7,5%). 54

58 Table 23: Additional descriptive statistics for question 16 This table presents the survey results for question 16: In the coming 4 years is it likely that your company will undertake any of the following actions partly or fully due to exchange rate changes? The table presents the answers for the total population of 174 companies according to country and identifies what kind of real options that the companies are likely to undertake. Change sourcing Increase capacity/extend production Enter foreign market Q 16 Types of real options Denmark Sweden Norway Finland Total Yes 36 (51,4%) 29 (44,6%) 17 (63%) 6 (50%) 88 (50,6%) No 34 (48,6%) 36 (55,4%) 10 (37%) 6 (50%) 86 (49,4%) Yes 35 (50%) 24 (36,9%) 13 (48,1%) 3 (25%) 75 (43,1%) No 35 (505) 41 (63,1%) 14 (51,9%) 9 (75%) 99 (56,9%) Yes 30 (42,9%) 19 (29,2%) 12 (44,4%) 2 (16,7%) 63 (36,2%) No 40 (57,1%) 46 (70,8%) 15 (55,6%) 10 (83,3%) 111 (63,8%) Shift production Yes 17 (24,3%) 15 (23,1%) 7 (25,9%) 0 (100%) 39 (22,4%) No 53 (75,7%) 50 (76,9%) 20 (74,1%) 12 (100%) 135 (77,6%) Temporally reduce/close operations Delay entry Yes 15 (21,4%) 13 (20%) 8 (29,6%) 2 (16,7%) 38 (21,8%) No 55 (78,6%) 52 (80%) 19 (70,4%) 10 (83,3%) 136 (78,2%) Yes 7 (10%) 5 (7,7%) 3 (11,1%) 0 (0%) 15 (8,6%) No 63 (90%) 60 (92,3%) 24 (89,9%) 12 (100%) 159 (91,4%) Abandon foreign market Yes 8 (11,4%) 5 (7,7%) 0 (0%) 0 (0%) 13 (7,5%) No 62 (88,6%) 60 (92,3%) 27 (100%) 12 (100%) 161 (92,5%) Total 70 (100%) 65 (100%) 27 (100%) 12 (100%) 174 (100%) In the next question (question 17) respondents where asked to consider to what degree their competitors currency cost structure where similar to that of their own. (table 24) Currency cost structure was defined as the composition/combination of currencies in which your company has its primary foreign operating costs. In other words, the degree to which the company s exposure setup is similar to that of their competitors. As in question 15 the data is ordinal for which reason it is not possible to point out any country specific differences by using means. Table 24: Survey results for question 17 This table presents the survey results for question 17: To what degree is your competitors currency cost structure similar to your company s currency cost structure? The table presents the answers for the total population of 174 companies according to country. How would you characterize the cost structure of your main competitors? Q 17 Currency cost structure Denmark Sweden Norway Finland Total Similar 10 (14,3%) 20 (30,8%) 7 (25,9%) 1 (8,3%) 38 (21,8%) Somewhat similar 27 (38,6%) 20 (30.8%) 17 (63%) 9 (75%) 73 (42%) Neither similar or dissimilar 21 (30%) 15 (20,1%) 2 (7,4%) 1 (8,3%) 39 (22,4%) Somewhat dissimilar 5 (7,1%) 4 (6,2%) 1 (3,7%) 0 (0%) 10 (5,7%) Dissimilar 7 (10%) 6 (9,2%) 0 (0%) 1 (8,3%) 14 (8%) Total 70 (100%) 65 (100%) 27 (100%) 12 (100%) 174 (100%) Of the responding companies, 73 companies (42%) describe its competitors cost structure somewhat similar as their own and is the most frequent characterization of all 55

59 the companies. At the same time 38 companies (21,8%) characterizes their competitors cost structure as being similar which is to say that 63,8% view their currency cost structure as similar to that of their competitors. On the other side only 24 companies (13,7%) answered that they would characterize their competitors cost structure as somewhat dissimilar or dissimilar, while the remaining 39 companies (22,4%) answered that they would characterize it as being neither similar nor dissimilar. Norwegian companies see the most similarities between their own and the currency cost structure of their competitors with 88,9% having answered similar or somewhat similar. Finnish companies follow suit with 83,3% of the responding companies characterizing their competitors cost structure as similar or somewhat similar. To investigate what the company s primary motivation for hedging foreign exchange rate exposure was, they where asked to indicate whether it was to minimize the volatility of the accounting earnings or it was to minimize volatility of cash flows. (table 25) The question where asked to test whether the time horizon of the financial hedges was influenced by accounting concerns. The influence of these factors on the time horizon of our dependent variables would have been investigated further in the regression analysis. Table 25: Survey results for question 18 This table presents the survey results for question 18: When your company hedges its foreign exchange rate exposure, what is your company primarily trying to manage with the hedge? The table presents the answers for the total population of 174 companies according to country. Primary motivation for hedging Volatility of accounting earnings and the market value of the company Q 18 Motivation for hedging Denmark Sweden Norway Finland Total 34 (48,6%) 27 (41,5%) 8 (29,6%) 6 (50%) 75 (43,1%) Volatility of cash flows 36 (51,4%) 38 (58,5%) 19 (70,4%) 6 (50%) 99 (56,9%) Total 70 (100%) 65 (100%) 27 (100%) 12(100%) 174 (100%) However, in the questionnaire send to the companies, they were asked if the primary motivation was to hedge the volatility of accounting earnings and the market value of the firm or to hedge volatility of cash flows. As the market value of the firm consists of the market value of equity and debt whereas accounting is the company s earnings reported in the income statement, they can not meaningfully be combined. As this was an unlucky mistake during the distribution of the questionnaires the variable is not further analyzed as we would not be able to conclude anything meaningful from the variable, and due to the potential bias it could cause. 56

60 In order to get a further indication of the market stability in which the companies operate they where asked to consider the likelihood that their foreign revenues would deviate from the budgeted within the next 12 months. The answers are presented in table 26 below, and as in question 15 and 17 we are dealing with ordinal data, for which reason means is not calculated. Table 26: Survey results for question 19 This table presents the survey results for question 19: How likely is it that your actual foreign revenues will deviate by +/- 10% from your company s budgeted foreign revenues within the next 12 months? The table presents the answers for the total population of 174 companies according to country. Likelihood of deviations from budgeted foreign revenues Q 19 Variance of budgets Denmark Sweden Norway Finland Total Very likely 7 (10%) 9 (13,8%) 4 (14,8%) 2 16,7%) 22 (12,6%) Likely 26 (37,1%) 16 (24,6%) 7 (25,9%) 3 (25,0%) 52 (29,9%) Neither likely nor 21 (30%) 18 (27,7%) 8 (29,6%) 4 (33,3%) 51 (29,3%) unlikely Unlikely 10 (14,3%) 14 (21,5%) 5 (18,5%) 3 (25,0%) 32 (18,4%) Very unlikely 6 (8,6%) 8 (12,3%) 3 (11,1%) 0 (0%) 17 (9,8%) Total 70 (100%) 65 (100%) 27 (100%) 12 (100%) 174 (100%) Most of the responding companies found deviations either likely or neither likely nor unlikely with answers corresponding to 52 (29,9%) and 51 (29,3%) respectively. Overall, more companies, 42,5% compared to 28,2, found likely that there foreign revenues would deviate from the budgeted within the next 12 months, No significant differences exist between the countries. The last question of the survey related to the amount of total foreign revenues, that a company would be in a position to change into another currency without prohibitively high adjustment costs within a year. (table 27) Table 27: Survey results for questions 20 This table presents the survey results for question 20: How much of your cost base in foreign currency could your firm change (without prohibitively high adjustment costs) within the next 12 months if exchange rates move against your firm? The table presents the answers for the total population of 174 companies according to country, and calculates the mean using the midpoints of the intervals from 0%-100% What part of your foreign cost may be changed to another currency within the next 12 months? Q 20 Changeable costs Denmark Sweden Norway Finland Total 0% 10 (14,3%) 14 (21,5%) 9 (33,3%) 3 (25%) 36 (20,7%) 1-20% 42 (60%) 36 (55,4%) 12 (44,4%) 7 (58,3%) 97 (55,7%) 21-40% 8 (11,4%) 7 (10,8%) 2 (7,4%) 0 (0%) 17 (9,8%) 41-60% 5 (7,1%) 2 (3,1%) 1 (3,7%) 2 (16,7%) 10 (5,7%) 61-80% 3 (7,1%) 2 (3,1%) 2 (7,4%) 0 (0%) 7 (4,0%) 81-99% 1 (1,4%) 2 (3,1%) 0 (0%) 0 (0%) 3 (1,7%) 100% 1 (1,4%) 2 (3,1%) 1 (3,7%) 2 (0%) 4 (2,3%) Total 70 (100%) 65 (100%) 27 (100%) 12 (100%) 174 (100%) Mean 18,71% 18,29% 17,41% 14,17% 18,04% 57

61 While 20,7% of the companies were in no position to change any of their foreign cost within the next 12 months, making them somewhat inflexible in the short-run, the average firm was estimated to be able to change 18,04% of their foreign cost. The most frequent response (55,7%), was that 1-20% of foreign costs could be changed to other currencies within the next 12 months. The answers to this question were evenly distributed across countries as well as intervals, with no significant differences observed Concise conclusion on chapter The purpose of this section was to investigate the flexibility and stability of the firms in our sample. The section examined the perceived flexibility of the business environment and found that 48,8% of the companies described the business environment they operate within as being stable or fairly stable whereas 28,8% of the companies found their business environment to be volatile or fairly volatile. Furthermore, this section examined the likelihood that a firm would experience in their budgeted revenues and found that 42,5% of the companies found it likely or very likely that their revenues budget would deviate by +/- 10% or more within the next 12 months whereas only 28,2% indicated that it was unlikely or very unlikely. These questions were included to proxy for stability and hence these results are surprising as we would have expected them to be somewhat correlated which is not the case. The company s potential to use real options was also examined in this section and we found that on average, 1,9 of the seven choices was likely to be used by the firms in our sample. The most likely real option to be used was change sourcing (51%) or to increase capacity or extend production (43,1%) whereas the least frequent possible real options to undertake was to abandon a foreign market (7,5%). Aabo, Simkins (2003) found that the average Danish firm was likely to undertake 2,2 real options due to exchange rate changes even though the time horizon used in their study was 2-3 years as compared to four years in this study. Similar results in terms of the most likely and unlikely real option to be utilized was found by Aabo, Simkins. The option to change sourcing was the most likely (54%) while the option to abandon a foreign market was the least likely (19%). Lastly, as a measure of the companies short-sighted flexibility they where asked how much of their currency cost structure it was possible to change within the next 12 58

62 months and on average, firms were able to change 18% of their costs to other currencies. 59

63 5. Regression Analysis While the previous chapter provided detailed descriptive statistics on the results from our survey, the purpose of the present chapter is, by means of regression analysis, to examine the relationship between various company characteristics and the time horizon for which Scandinavian, non-financial firms choose to hedge their foreign exchange rate exposure using financial means. Additionally, the following chapter investigates the influence company characteristics have on the time horizon at which the importance of operational means exceed that of financial means for managing foreign exchange rate exposure. These tests are conducted in order to answer the second part of the problem statement outlined in chapter Definition of variables & hypotheses The following section describes the dependent and independent variables used in the regression analyses, which serve to investigate and test the relationships outlined above. At the same time, predicted relationships between the dependent and independent variables are formulated in various hypotheses to be tested Dependent variables The focus of this thesis is on the investigation of factors that influence the time horizon of foreign exchange rate exposure hedges, in small open economies. This is done by examining the time horizon of Scandinavian companies use of financial hedges as well as the perceived horizon at which benefits of operational hedges exceeds that of financial hedges. In total the following five dependent variables will be tested: 1) Time horizon for use of financial means in general (see section 5.2 for results) 2) Time horizon for foreign debt (see section 5.3 for results) 3) Time horizon for short-sighted derivatives use (see section 5.4 for results) 4) Time horizon for long-sighted derivatives use (see section 5.5 for results) 5) Time horizon at which the importance of operational means exceed financial means (see section 5.6 for results) 60

64 All the dependent variables are assessed on an ordinal scale and are thus, ordered variables. Each of the five variables is explained in more detail in the beginning of each of the sections relating to the particular variable Independent variables and hypothetical variables To test the different hypotheses, which are to be investigated in this regression analyses, a number of independent variables are used as regressors. As some of the independent variables, often called hypothetical or latent variables, are difficult to observe and quantify, proxy variables are used in the model (Møller Jensen & Knudsen, 2006). Some of the proxy variables come from the questionnaire while others come from the financial database, Orbis. These different proxy variables are used to test whether they influence the time horizon in accordance with the hypotheses which will be outlined in chapter This structure is inspired from the work of e.g. Nance et al. (1993), Bodnar et al. (1998) and Hagelin (2003) who also uses a questionnaire combined with public available information to analyze companies exchange rate exposure management. In accordance with the conceptual framework presented in chapter 2, the following independent variables have been identified to influence the time horizon of foreign exchange rate exposure management: 1. The companies foreign involvement 2. The companies possibilities of being flexible 3. The stability of the company s primary market 4. The net exposure 5. The competitors currency cost structure 6. The size of the company 7. The companies financial situation The variables are related to similar variables confirmed by previous studies to have some relationship to the hedging practices of corporations. However, as not much has been written (empirically or theoretically) on the subject of time horizon for foreign exchange rate hedges, the predicted relationships are primarily derived from the proxies influence on derivatives use and the corresponding time horizon of these derivatives. For example, company size might be associated with increased use of foreign debt, and as foreign debt is assumed to be used to hedge exposures with longer 61

65 horizons, we would expect size to have a positive influence on the length of the time horizon. Figure 6: Determinants of the FX hedging horizon to be tested This figure presents an overview of the different independent variables expected to influence the hedging horizon of Scandinavian, medium-sized and large non-financial firms. 1. The degree of a company s foreign involvement is measured by five different proxy variables all inspired from former work of several authors, see e.g. Allayannis & Weston (2001), Nydahl (1999) and Jorion (1990). Firstly, foreign involvement is proxied by the average of a company s foreign cost and revenues out of total revenues and costs (Q1). Furthermore, foreign involvement is proxied by the company s foreign asset ratio (Q2) as well as by the share of the company s total debt that is denominated in foreign currency (Q9). Lastly, the degree of foreign involvement is proxied by the average number of sales and production subsidiaries the company has abroad (Q4) as well as the numbers of currencies the companies are significantly exposed to (Q5). Aabo (2006) finds a positive relationship between a company s foreign involvement (measured by foreign subsidiaries) and the propensity of foreign debt use over derivatives, for exchange rate exposure management. Assuming that foreign debt is most commonly used for long term hedging, an assumption supported by the findings in chapter 4, we test whether this also means that a higher degree of foreign involvement is associated with longer horizons for exchange rate exposure hedges. 62

66 2. Several authors have discussed and analyzed how a company s flexibility is linked to its use of financial derivatives. Aabo & Simkins (2003) find that a company s decision on whether or not to use financial hedges is related to company size, amount of export as well as the manager s ability to/intentions of exercising various real options relating to the flexibility of the company. These findings do not necessarily entail that the degree of flexibility directly affects the time horizon of the financial hedges. However, it does emphasize the interesting connection between real options and time horizon of financial derivatives due to the fact that one might argue that the duration of a company s financial hedges should match the time horizon at which it is possible for the company to exercise its real options, in order for the company to fully protect itself. (Capel, 1997) As much of the discussions relating to flexibility involve the use of real options, the first proxy variable for the flexibility is based on the question regarding the company s possibility of exercising a number of different real options within the next 4 years (16). The question listed seven different real options all relating to the four general types of real options described in chapter 2. Additionally, the company was asked how much of their short-term cost base in foreign currencies that it could change, without prohibitively high adjustment costs, within the next 12 months (Q20). This variable is also used as a proxy for a company s flexibility. The question is inspired by Capel, who state that a company who has production abroad and has the possibility to move this production home, actually own a put option to change its most intensive cost items from the foreign currency to its home currency (See Capel, 1997, page 89). The two variables may be interrelated, but Q16 is viewed as the company s long-term operational flexibility whereas Q20 is viewed as a measure of the company s short-term operational flexibility. Both of the proxy variables are predicted to have a negative relationship to the time horizon of the different depend variables. The more flexible a company is the less incentive to cover long-term foreign exchange rate exposure, as the company may exercise some of its real options, eliminating long-term exposure. 3. To test whether there is a relationship between the industry stability and the company s hedging horizon, two question pertaining to stability where constructed. As discussed in chapter 2.1 there may be a connection between the price elasticity of a company s input and the output and the time horizon of the exchange rate formation. Grant & Soenen (2004) finds a positive relation between the intensity of competition 63

67 (price elasticity, seller concentration etc.) and operating exposure. Assuming that the long term nature of operating exposure described in chapter 2.1 is correct, we wish to investigate whether this has a positive influence on the dependent variables. As the price elasticity of the company s input/output, seller concentration etc. may be difficult to measure, the respondents where asked about the stability of the business and possible budget variations likely to occur within one year. The point is that the more intense the competition within the industry is, the more unstable and volatile is the company s future cash flows likely to be. These variables are both ordered variables ranging from 1-5, where the first question asked about the stability of the general business environment in which the company operates (Q15), the second variable inquired about the likelihood that the company s actual revenues would deviate from the budgeted by +/-10% within the next 12 months (Q19). 4. Pramborg (2005) found that currency matching was the most widely used internal hedging technique used by Swedish firms to manage foreign exchange rate exposure. Thus, even if a company may have a large proportion of its total revenues and/or cost in foreign currencies, this may not necessarily be a measure of a company s actual foreign exchange rate exposure. If some of the revenues (cost) are matched by cost (revenues) in the same currency the exposure is not real. Hence, to measure a companies net exchange rate exposure, the respondents where asked to state how much of the companies operating revenues (costs) that where matched by operating costs (revenues) in the same currency (Q8). Consequently, if the cost/revenues are matched 100% there is not necessarily any actual real exposure creating a net exposure of 0. Whether this variable has a positive or negative influence on our dependent variables is unknown. 5. To further investigate whether a company s business environment, influences the time horizon of its hedging strategy, the company s indirect or competitive exposure is analyzed. The indirect exposure would in an ideal case be measured by the interaction of the company s sensitivity to exchange rate related movements with the markets sensitivity to exchange rate related movements (Muller & Verschoor, 2005), but due to limitation of information another variable was constructed in the questionnaire. The respondents where asked to asses the similarity of the currency cost structure between that of their company and that of their competitors (Q17). This 64

68 variable assumes that the company is aware of its competitor s exchange rate setup, and is thus, assumed to quantify the company s indirect exposure. The more similar their competitor s exchange rate setup is, the less is the indirect competitive exposure. Whether there is a positive or negative relationship between similarities of the competitors currency cost structures and the time horizon of the financial hedges is unknown as no previous research exists. However one may argue that if e.g. all your competitors have covered its foreign exchange exposure for a short period and your company has covered its FX exposure for a long time horizon, your competitive situation is different than that of your competitors. Whether this is a competitive advantage or actually has become a speculative situation is unknown and depends on the outcome. 6. Previous research has indicated a connection between the size of a company and its hedging practices. Warner (1977) has found various factors, which might cause small firms to hedge more than large companies. He argues that due to relatively higher costs of financial distress and progressive taxation systems (small firms has more income in the progressive region), small firms tend to hedge relatively more than larger firms. Block and Gallagher (1986) on the other hand, argue that due to economies of scale in terms of derivatives use, large firms tend to hedge relatively more. In this thesis, the size of the company is included in the regression analysis to see if there may be an association between the size and our dependent variables. The rationale for this relationship is that the larger a company is, ceteris paribus, the less flexible it is. If the company is large, it may be difficult to restructure the whole organization quickly in response to advantageous changes in changing exchange rates. Furthermore advantages of economies of scale, is achieved by reducing input cost pr unit by operating on a large scale. By scaling down production in one place by moving it abroad, the economies of scale may be somewhat diminished. This may indicate that they plan further into the future and hence hedge for longer time periods. At the same time companies that are relative small do not have the same possibilities of exploiting economies of scale for which reason they may choose to invest in flexibility instead, while leaving long-term exposures unhedged. Thus, size is expected to have a negative effect on the flexibility of a company and so, larger companies are expected to hedge for longer periods in terms of their exchange rate exposure management. Allayannis and Ofek (2001) find that that there is a positive relationship between company size and the 65

69 use of debt Assuming that this relationship is true, this thesis test whether size is positive related to hedging horizon (under the assumption that debt is primarily used as long term hedging). In this thesis, size is measured by the natural logarithm of the number of employees and the natural logarithm of its total balance (assets). 7. Although it is normally used as a part of the financial distress hypothesis, (see Smith & Stulz, 1985), solvency, (total equity/total assets) for the purposes of this study, is used as a measure of the company s financial situation. Solvency is a measure of a company s ability to meet its long-term obligations. In other words if a company has a high solvency ratio, the probability of bankruptcy and financial distress in a long run, is lower than for companies with low solvency ratios. Additionally, some credit rating firms use solvency as part of their credit rating, for which reason, a lower solvency means lower credit rating. Kedia and Mozumdar (2003) find a positive relation between credit quality and the use of foreign debt and as debt is often used for long-term hedging there may be a positive relation between solvency and the maturities of the financial derivatives. Leverage (debt/assets) is also used to test whether it influences the time horizon. Clark and Judge (2009) find that debt plays a role in the choice of companies hedging strategies. They find that leverage has positive influence on the use of swaps and a negative influence on the use of foreign debt. As both foreign debt and swaps are used to hedge foreign exchange rate exposure, it is unknown whether leverage has a positive or negative on the time horizon of the hedges. An overview of the different variables proxying for the underlying latent variables is presented in figure 7. This figure also illustrates the source of the variable, and where appropriate, the related question in the survey. Additionally, the figure defines each of the variables. 66

70 Figure 7: Definition of proxy variables This figure presents the proxy variables used in the following regressions as well as the factor it has replaced. In addition, a description/definition of each variable is given and the source of information stated. # Proxy variables Independent factor Definition Source Q Average of the amount of costs and revenues in foreign currency 1) Foreign Revenues and Costs (FRC) Foreign Involvement as part of all cost and revenues Survey 1 2) Foreign Assets Ratio (FAR) Foreign Involvement Foreign Assets/Total Assets Survey 2 3) Number of Subsidiaries (SUB) Foreign Involvement Average number of production and sales subsidiaries abroad Survey 4 4) Number of Currencies (CUR) Foreign Involvement Number of currencies significantly exposed to Survey 5 5) Foreign Debt Ratio (FDR) Foreign Involvement Foreign Debt/Total Debt Survey 9 6) Real options (REALOPT) Flexibility 7) Flexible costs (FLEXCOST) Flexibility 8) Business environment (BUSS) Stability 9) Variance of budgeted foreign revenues (VAR) Stability 10) Net exposure (NETEXP) Net Exposure 11) Competitors currency cost structure (COMP) Competitors Variable ranging from 0 to 7 depending on how many real options the company is considering to use within the next 4 years Survey 16 Ordered variable of whether the company may change the currency of its cost (without prohibitively high adjustment costs) within the next 12 months if exchange rates move against their firm Survey 20 Ordered variable ranging from 1-5 depending on the companies characterization of stability within the general business environment where the company operates Survey 15 Ordered variable ranging from 1-5 depending on how likely it is that the companies actual revenues will deviate more than +/- 10% from the budgeted within the next 12 months Survey 19 Share of company's foreign operating costs(revenues) that are offset by matching foreign operating revenues (cost) Survey 8 Ordered variable ranging from 1-5 depending on how similar the company considers their competitors exchange rate exposure to be Survey 17 13) Number of employees (EMP) Size Logarithm of the number of Employees Orbis - 14) Total balance (BAL) Size Logarithm of the total Balance (assets) Orbis - 16) Solvency ratio (SOLV) Financial situation Total Equity/Total Assets Orbis - 17) Leverage (LEV) Financial situation Total Debt/Total Assets Orbis Factor analysis In order to examine whether it would be relevant to group the different proxies into indices, a factor analysis was conducted to test the relationship between the variables. The results of the factor analysis are presented in appendix 6. From appendix 6 we see that the factors proxying for foreign involvement are all grouped together with high coefficients. Thus as a proxy for foreign involvement, an index containing the five variables was created. The foreign involvement index will be used in the regression models conducted later in this section Correlation analysis independent variables This section examines the relationships between the independent variables used in the regression models. Regression models may suffer from multicollinearity. The problem of multicollinearity arises when two or more independent variables in a regression have a high (but not perfect) correlation. The concept of multicollinearity is not well defined, 67

71 which means that there is no exact number at which we can conclude that multicollinearity is a problem. (Wooldridge, 2006) However, a generalization can be derived from Wooldridge (2006) p.103 in which he states that: everything else being equal, for estimating β j, it is better to have less correlation between x j and the other independent variables. Thus, in order to interpret the degree of correlation between independent variables, this thesis uses a rule of thumb suggested by Hair et al. (2003) 22 which order the correlation coefficients based on their values, and variables with moderate correlations (>0,4) will be addressed. A potential solution to the problem is the partial exclusion of correlated variables in order to reduce multicollinearity. However, this procedure may lead to bias, so a trade-off between the two has to be considered. Due to differences in the response rates for the dependent variables used in the regression models, five different correlation tables are presented in appendix 5. While the questions relating to the time horizon at which operational means exceeds financial means for hedging foreign exchange rate exposure (Q_21) uses the full sample, the remaining dependent variables use smaller sample sizes. Concerning the correlation coefficients for the model containing the complete sample of 174 companies, high correlation exists between foreign involvement and number of employees (0,43), between foreign involvement and real exposure (-0,45) and between total balance and number of employees (0,76). Similar results are found for the variables used in the regression concerning the time horizon that companies have covered their exchange rate exposure using financial means. Here, high correlations exist between foreign involvement and the two size variables number of employees (0,47) and total balance (0,41). Additionally, high correlation exist between the two size variables (0,81) Next, results are found for the variables used in the regression concerning the time horizon for foreign debt which includes 109 companies. Here, high correlation exists between foreign involvement and the number of employees (0,49), foreign involvement and total balance (0,47) as well as between number of employees and total balance (0,83). The variables used in the regression concerning the time horizon for short-term derivatives, which includes 64 companies, show high correlation for a number of variables concerning foreign involvement. High correlation exists between foreign ,2 slight, negligible correlation; 0,21-0,4 small; 0,41-0,7 moderate; 0,71-0,9 high; 0,91-1,0 very strong correlation 68

72 involvement and real options (0,41), real exposure (-0,51), total balance (0,49) and number of employees (0,61). Additionally, as in the previous correlations tables, high correlation exists between number of employees and total balance (0,84) as well as for a new set of variables, namely leverage and solvency (-0,46). Lastly, the correlation table for the variables used in the regression concerning the time horizon of long-term derivatives show high correlation between foreign involvement and real exposure (0,51); between number of employees and the three variables foreign involvement (0,63), real options (0,45) and total balance (0,85) as well as between the similarity of competitors exposure setup and the variability of cash-flows (0,42). Based on these correlation coefficients, the regressions conducted will be using total balance as proxy for firm size while number of employees will serve to test the robustness of the models. This was done due to high correlation between the two variables in all the models. Various other correlations exist between variables in the different models, and these variables will be included selectively in the regression models. Problems relating to correlated variables will be addressed in the regression analyses Regression model and hypotheses The regressions made for the purposes of this research are ordered probit models, which are based on the following general model: y = λ 0 + x i λ i + x n λ n + ω, where y represents the dependent variable, x i to x n represents each of the independent variables used; λ i to λ n is the estimated coefficients for each independent variable while ω is the error term representing the value of factors that are not represented in the model. In our estimations the following ordered probit regression model is used to test the different hypotheses: y = λ 1 FIINDEX + λ 2 REALOPT + λ 3 FLEXCOST + λ 4 BUSS + λ 5 VAR + λ 6 NETEXP + λ 7 COMP + β 8 EMP + β 9 BAL + β 10 SOLV + β 11 LEV + ω Models for each of our five dependent variables will be estimated to test various hypotheses. The dependent variables used in all of the five regression models are ordered variables for which reason ordered probit models are estimated. As the regression models are based a nonlinear function (probit model), we will not be able to say much about the magnitude of the impacts that the independent variables have on the 69

73 five dependent variables. However, we will be able to interpret the direction of the impact based on the sign of the coefficients presented for each variable. That is to say, if a variable is significant with a positive coefficient, we conclude that the variable has a positive effect on the dependent variable (time horizon). Additionally, maximum likelihood is used to estimate the models. (Wooldridge, 2006 p. 383ff.) For the purpose of conducting the regression analyses, Eviews was used as it offers the functionality to generate outputs with the relevant information (p-value and coefficients) for every variable. Additionally, Eviews calculates the LR-index (Pseudo R2) as well as the Probability (LR stat) which can be used to evaluate the fit and quality of the different models. Some of the independent variables have been reversed in order for the variables used as proxies for the same latent variable, to exhibit the same relationships. E.g. the answers to question 15 have been reversed so that the answers are comparable to the variable VAR used as proxy for stability. Based on earlier findings and the conceptual framework, the following hypotheses are made to further investigate the research questions 23 : 1) Based on the findings of Aabo (2006) foreign involvement (FIINDEX) is predicted to have a positive relationship towards the time horizon for which the companies has covered its foreign exchange rate exposure using financial means. Additionally, size is hypothesized to have a positive influence on the dependent variables. This is based on the findings of Allayannis and Ofek (2001) who find a positive relationship between size and use of foreign debt. 2) Based on the theoretical arguments of Srinivasulu (1981) and Capel (1997) as well as the finding of Aabo & Simkins (2003) the short-term flexibility (FLEXCOST) and long-term flexibility (REALOPT) is predicted to have a negative influence on the maturity of the companies financial foreign exchange rate derivatives and foreign debt. 3) Stability, proxied by BUSS and VAR, is hypothesized to have a positive influence on the dependent variables used in the five regression models. 5.2 Average time horizon of hedging using financial means The following section presents an ordered regression model that explains the effects that various independent variables have on the following question: at the present time, 23 Net exposure, financial situation, the competitors cost structure and financial situation are predicted to show lower significant influence on the dependent variables for which reason no hypotheses are stated. 70

74 what is the average time horizon that your company has covered its foreign exchange rate exposure by use of financial means? (i.e. forward contracts, options, swaps and debt in foreign currency). The dependent variable used in the regression is an ordered variable with values ranging from 1-7 as below. Value Average time horizon is months months months months years years 7...>4 years As we saw in chapter 4, the 128 companies who used financial means to hedge their foreign exchange rate exposure, had average maturities of 6,2 months, and 89% of the companies covered its exposure with financial means for less than a year. The following models are specified and the results from the regressions are presented in table 28. Model 1 is a full model that includes all the independent variables, whereas model 2 and 3 substitute foreign involvement and total balance due to correlation (0,41). Model 4 focuses on the significant independent variables from the first three models. The dependent variable is measured in the direction from short to long horizon, which is to say that a positive coefficient is associated with a longer hedging horizon. The variable foreign involvement is significant with a positive coefficient at the 1%- level or 5%-level in all the models in which it is included. The positive coefficient indicates that companies with more foreign involvement also tend to cover its exchange rate exposure for a longer time horizon using financial means. This is line with our hypothesis based on the finding of Aabo (2006) who found a positive relationship between foreign involvement (measured on number of subsidiaries) and the amount of foreign debt issued. This hypothesis was based on the fact that debt is normally used to cover long term operating exposure 71

75 Table 28: Estimated regression models for the average time horizon in general This table presents the results of four regression models for the dependent variable average time horizon for financial means in general For each independent variable, the coefficient and p-value are presented. Variables significant at the 1%-level, 5%-level and 10%-level, are market with *, ** and *** respectively. Below each model, a probability measure; a measure to evaluate the goodness of fit and the number of respondents included is presented. The independent variables used are: FIINDEX an index, 0-100, containing five variables for the level of foreign involvement, REALOPT variable ranging from 0-7 depending on how many real options the company is considering to use within the next 4 years, FLEXCOST ordered variable ranging from 1-7 depending on the amount of costs the company is able to change, BUSS ordered variable coded 1= very volatile, 2= fairly volatile. 3= not particularly stable or volatile, 4= fairly stable, 5= very stable, VAR - ordered variable coded 1= very likely, 2= likely. 3= neither likely nor unlikely, 4= unlikely, 5= very unlikely, NETEXP - the share of company's foreign operating costs(revenues) that are offset by matching foreign operating revenues(costs) on a scale from 0-100, COMP - ordered variable coded 1= dissimilar, 2= somewhat dissimilar. 3= neither similar nor dissimilar, 4= somewhat similar, 5= similar, EMP the natural logarithm of the number of employees, BAL the natural logarithm of total assets, SOLV the ratio of total equity to total assets, LEV the ratio of total debt to total assets. Proxy variables M-1 M-2 M-3 M-4 Coeff. p-value Coeff. p-value Coeff. p-value Coeff. p-value Foreign involvement (FIINDEX) 0,0112 0,0532*** 0,0148 0,0056* 0,0135 0,0123** Real options (REALOPT) 0,0019 0,5901 0,0038 0,2541 0,0021 0,5479 Flexible costs (FLEXCOST) -0,0001 0,9758-0,0003 0,9340 0,0006 0,8751 Business environment (BUSS) -0,0071 0,0254** -0,0076 0,0164** -0,0069 0,0302** -0,0074 0,0176** Variance of budgeted foreign revenues (VAR) -0,0028 0,3635-0,0042 0,1612-0,0029 0,3405 Net exposure (NETEXP) -0,0049 0,1055-0,0072 0,0104** -0,0049 0,1044 Competitors currency cost structure (COMP) -0,0050 0,1033-0,0060 0,0494-0,0046 0,1319-0,0053 0,0835*** Number of employees (EMP) (See robustness test) Total balance (BAL) 0,1809 0,1805 0,2898 0,0176** 0,1468 0,2712 0,1517 0,2550 Solvency ratio (SOLV) -0,0078 0,0503*** -0,0077 0,0517*** -0,0075 0,0572*** -0,0080 0,0445** Leverage (LEV) -0,0045 0,2648-0,0048 0,2311-0,0040 0,3101 Prob(LR statistic) 0,0000 0,0001 0,0001 0,0000 Pseudo R-squared 0,0589 0,0531 0,0548 0,0553 n The variable proxying for the stability of the business environment is significant at a 10%-level in all four models with negative coefficients. This indicates that companies that operate in more stabile business environments are expected to hedge for shorter time using financial derivatives and debt than companies in more volatile environments. The direction of the coefficient is not positive as hypothesized. We had expected that a stable business environment would make the companies more certain of their expected future cash flows for which reason which would allow them to hedge for longer periods. Net exposure is significant at 5%-level in 3 out of 4 models. The coefficient is positive which means that the less a company matches cash flow in and out of the firm, the longer the time horizon that company cover its currency exposure by financial means. However, it is interesting that net exposure is only significant when the proxy for foreign involvement is included in the regression. 72

76 Variables that are not significant in explaining the time horizon for which Scandinavian non-financial firms hedge their foreign exchange rate exposure using financial means include the firms ability to change its costs, the likelihood of budget deviations as well as the financial indicators leverage and solvency. To test the robustness of the four models, the variable total balance was substituted by number of employees. This procedure is copied in following four regression models. However, when re-running the four models, none of the variables showed any significant changes. Model 1 and 2 was significant at 5%-level whereas models 3 and 4 were significant at a 1%-level. The models have Pseudo R-squared values between 0,426 and 0, Average maturity of foreign debt The following section presents an ordered regression model that explains the effects that various independent variables have on the following question: what is the average loan period for your company s foreign debt arrangements. The dependent variable used in the regression is an ordered variable with values ranging from 1-7 as shown below. Value Average time horizon is months months months months years years 7 >4years Of our sample 174 companies, 109 indicated that they had debt denominated in foreign currency. The average loan period for these foreign debt arrangement was 16.6 months, with a large share of the companies (47,7%), with debt contracts with maturities for less than 3 months whereas 16,5% that had foreign debt with loan periods above 4 years. The following models are specified and the results from the regressions are presented in table 29. Once again, model 1 is a full model including all independent variables. Model 2 excludes foreign involvement as it was moderately correlated (0,47) with the variable total balance. Model 3 excludes total balance while including foreign involvement, and as before, model 4 includes all the variables that at least have been significant at a 10%- level. The dependent variable is measured in the direction from short to long horizon, which is to say that a positive coefficient is associated with a longer hedging horizon. 73

77 Table 29: Estimated regression models for the average time horizon in general This table presents the results of four regression models for the dependent variable average time maturity of foreign debt For each independent variable, the coefficient and p-value are presented. Variables significant at the 1%-level, 5%-level and 10%-level, are market with *, ** and *** respectively. Below each model, a probability measure; a measure to evaluate the goodness of fit and the number of respondents included is presented. The independent variables used are: FIINDEX an index, 0-100, containing five variables for the level of foreign involvement, REALOPT variable ranging from 0-7 depending on how many real options the company is considering to use within the next 4 years, FLEXCOST ordered variable ranging from 1-7 depending on the amount of costs the company is able to change, BUSS ordered variable coded 1= very volatile, 2= fairly volatile. 3= not particularly stable or volatile, 4= fairly stable, 5= very stable, VAR - ordered variable coded 1= very likely, 2= likely. 3= neither likely nor unlikely, 4= unlikely, 5= very unlikely, NETEXP - the share of company's foreign operating costs(revenues) that are offset by matching foreign operating revenues(costs) on a scale from 0-100, COMP - ordered variable coded 1= dissimilar, 2= somewhat dissimilar. 3= neither similar nor dissimilar, 4= somewhat similar, 5= similar, EMP the natural logarithm of the number of employees, BAL the natural logarithm of total assets, SOLV the ratio of total equity to total assets, LEV the ratio of total debt to total assets. Proxy variables M-1 M-2 M-3 M-4 Coeff. p-value Coeff. p-value Coeff. p-value Coeff. p-value Foreign involvement (FIINDEX) 0,0194 0,0095* 0,0254 0,0002* 0,0201 0,0036* Real options (REALOPT) 0,0018 0,6746 0,0048 0,2587 0,0017 0,6917 Flexible costs (FLEXCOST) -0,0064 0,2451-0,0077 0,1590-0,0060 0,2834 Business environment (BUSS) -0,0057 0,1857-0,0063 0,1443-0,0070 0,1014 Variance of budgeted foreign revenues (VAR) -0,0026 0,5422-0,0036 0,3947-0,0013 0,7562 Net exposure (NETEXP) 0,0016 0,6811-0,0006 0,8778 0,0021 0,5853 Competitors currency cost structure (COMP) -0,0072 0,095*** -0,0073 0,0866*** -0,0066 0,1202-0,0084 0,0461** Number of employees (EMP) (See robustness test) Total balance (BAL) 0,3386 0,055*** 0,5302 0,0008* 0,3652 0,0330** Solvency ratio (SOLV) -0,0161 0,0044* -0,0133 0,0160** -0,0162 0,0042* -0,0154 0,0052* Leverage (LEV) 0,0102 0,2958 0,0095 0,3150 0,0109 0,2508 Prob(LR statistic) 0,0001 0,0008 0,0002 0,0000 Pseudo R-squared 0,0918 0,0741 0,0822 0,0750 n As seen in table 29, foreign involvement is significant in all models at a 1%-level, which clearly illustrates that there is a relationship between the degree to which a company is involved internationally and the length of its foreign debt arrangements. The direction of the coefficient is positive as expected, which means that the more foreign involved a company is; the longer is the loan period of their foreign debt contracts. These finding are in line with the findings in chapter 5.2. Foreign involvement was hypothesized to have positive influence on the general foreign exchange rate exposure management by use of financial means, mainly because more foreign involved (measured by subsidiaries) companies issue more foreign debt (Aabo, 2006). This finding extends the conclusion that firms with a high degree of international use more foreign debt by showing that these companies also tend to issues debt with a longer maturity. 74

78 The variable competitors currency cost structure is significant at a 10%-level in model 1 and 2 and at a 5%-level in model 4. As seen in the negative sign of the coefficient, the variable has a negative influence on the foreign debt loan period. This means that the more similar the competitor s currency cost structure is, the shorter is the companies loan period of the foreign debt contracts. Company size, measured by total balance, is significant at respectively the 10%, 1% and a 5%-level in model 1, 2 and 4. However, when size is measured by number of employees, it is only significant in model 2, which indicates that size, measured by total balance, has a positive influence on the loan period of the company s foreign debt. This is in line with our prediction that large companies may be less flexible than small companies for which reason they may cover their exposure further into the future instead of investing in operational flexibility. Lastly, the variable solvency is significant in all models at 1% or 5%-levels. The coefficient for solvency is negative in all the models, which is to say that companies with high solvency ratios tend to have shorter debt contracts than companies with lower solvency ratios. Overall, it may be concluded that foreign involvement and company size measured by total balance both have a positive influence on the contract period of the companies debt arrangements, while the solvency ratio has a negative influence on the time horizon for a company s foreign debt arrangements. It is less reliable whether the variable competitor s currency structure has a negative influence on the contract length. Running a robustness check the variable relating to a company s currency cost structure became insignificant in two out of three models, indicating that the variable is sensitive to changes in the size variable. All models are significant at a 1%-level and they have Pseudo R-squared values from 0,0741 to 0, Average maturity of short-sighted derivatives The following section presents an ordered regression model that explains the effects that various independent variables have on the following question: what is the average maturity for the short-sighted derivatives (forward contracts and options) used to cover your company s foreign exchange rate exposure? The dependent variable used in the regression is an ordered variable with values ranging from 1-7 as shown below. 75

79 Value Average time horizon is months months months months years years 7...>4 years 64 of the 174 respondent indicated that they either used forward contract, options or both and the average maturity of these contracts was 6,3 months. 90,6% of the companies had average contract lengths less than a year and only one company indicated that the average contract period was above 2 years. The following models are specified and the results from the regressions are presented in table 30. Model 1 is a full model including all independent variables. In order to find the model with the highest fit the various models were run, where variables were excluded on the basis of the size of their p-values. In model 2, variables with p-values above 70% were excluded and in Model 3 variables with p-values above 30% were removed. Model 4 includes all the variables that at least have been significant at a 10% significance level. The dependent variable is measured in the direction from short to long horizon, which is to say that a positive coefficient is associated with a longer hedging horizon. The results of these models show that the variables business environment and variance of budgeted foreign revenues are the only variables that significantly explain the time horizon for short-sighted derivatives. The variable business environment is significant in all models at a 10%-level with a negative coefficient indicating that, the more stabile the companies characterize their general business environment the shorter is the maturity of their forward contracts and options used to hedge foreign exchange rate exposure. The variable variance of budgeted foreign revenues is significant at all four models at the 5%-level and has a positive coefficient which means that the less volatile the company s budgeted revenues, the longer is the maturity of the forward contracts and options used to cover their foreign exchange rate exposure 76

80 Table 30: Estimated regression models for short-sighted derivatives This table presents the results of four regression models for the dependent variable average time horizon for short-sighted derivatives For each independent variable, the coefficient and p-value are presented. Variables significant at the 1%-level, 5%-level and 10%-level, are market with *, ** and *** respectively. Below each model, a probability measure; a measure to evaluate the goodness of fit and the number of respondents included is presented. The independent variables used are: FIINDEX an index, 0-100, containing five variables for the level of foreign involvement, REALOPT variable ranging from 0-7 depending on how many real options the company is considering to use within the next 4 years, FLEXCOST ordered variable ranging from 1-7 depending on the amount of costs the company is able to change, BUSS ordered variable coded 1= very volatile, 2= fairly volatile. 3= not particularly stable or volatile, 4= fairly stable, 5= very stable, VAR - ordered variable coded 1= very likely, 2= likely. 3= neither likely nor unlikely, 4= unlikely, 5= very unlikely, NETEXP - the share of company's foreign operating costs(revenues) that are offset by matching foreign operating revenues(costs) on a scale from 0-100, COMP - ordered variable coded 1= dissimilar, 2= somewhat dissimilar. 3= neither similar nor dissimilar, 4= somewhat similar, 5= similar, EMP the natural logarithm of the number of employees, BAL the natural logarithm of total assets, SOLV the ratio of total equity to total assets, LEV the ratio of total debt to total assets. Proxy variables M-1 M-2 M-3 M-4 Coeff. p-value Coeff. p-value Coeff. p-value Coeff. p-value Foreign involvement (FIINDEX) 0,0082 0,4177 0,0098 0,2244 0,0061 0,3778 Real options (REALOPT) 0,0000 0,9954 Flexible costs (FLEXCOST) 0,0069 0,3608 0,0066 0,3593 Business environment (BUSS) -0,0105 0,0699*** -0,0101 0,0576*** -0,0098 0,0616*** -0,0095 0,0694*** Variance of budgeted foreign revenues (VAR) 0,0126 0,0298** 0,0127 0,0227** 0,0122 0,0278** 0,0109 0,041** Net exposure (NETEXP) -0,0023 0,7424 Competitors currency cost structure (COMP) -0,0010 0,8412 Number of employees (EMP) Total balance (BAL) -0,1997 0,4076-0,2029 0,3804 Solvency ratio (SOLV) 0,0004 0,9566 Leverage (LEV) 0,0038 0,9420 (See robustness test) Prob(LR statistic) 0,4243 0,0746 0,0377 0,0216 Pseudo R-squared 0,0530 0,0521 0,0439 0,0399 n Both of the variables were included to test whether there was any connection between stability on the market where the company operated and the horizon of their derivative usage. However, as the sign of the coefficients are not the same, not much can generally be said about the influence of stability on the horizon of short-sighted hedges. Running a robustness check, none of the variables showed any sensitivity to changes in the size variable. Model 1 was not significant, Model 2 was significant at a 10%-level and Model 3 and 4 was significant at a 5%-level. The significant models have Pseudo R-squared values from 0,0399 to 0,

81 5.5 Average maturity for long-sighted derivatives The following section presents an ordered regression model that explains the effects that various independent variables have on the following question: what is the average maturity for the long-sighted derivatives (swaps) used to cover your company s foreign exchange rate exposure? The dependent variable used in the regression is an ordered variable with values ranging from 1-7 as shown below. Value Average time horizon is months months months months years years 7...>4 years Only 34 companies answered that they used currency swaps to manage their foreign exchange rate exposure and the average maturity on the contracts was 22,5 months. The following models are specified and the results from the regressions are presented in table 31. Model 1 is a full model including all independent variables. Due to correlation (0,50), the variables foreign involvement and net exposure were used separately in models 2 and 3. Variance of foreign budgeted revenues was removed in all models due to its high p-value. Model 4 includes all variables that at least have been significant at a 10%-level. The dependent variable is measured in the direction from short to long horizon, which is to say that a positive coefficient is associated with a longer hedging horizon and vice versa. The results of these regression models find that the variable relating to the competitors currency cost structure is significant at a 1%-level in two models and at a 5%-level in the two other models. This indicates that the similarity of the exposure setup between a company and that of their competitors have a strong influence on the length that the company hedges using swaps. The coefficient is negative, which implies that the more similar the competitor s currency cost structure is to that of a company, the shorter the horizon for the company s long-sighted derivatives contracts. Similarly we find that the variable relating to the leverage ratio is significant at a 5%- level in all four models, with a negative coefficient. This means a higher leverage ratio would result in the expectations of a shorter time horizon for long-sighted derivatives. 78

82 Running a robustness check, none of the variables showed any sensitivity to changes in the size variable. Model 1 was significant at 10%-level, model 2 and 3 was significant at a 5%-level and model 4 was significant at a 1%-level. The models have Pseudo R-squared values from 0,1419 to 0,1736. Table 31: Estimated regression models for long-sighted derivatives This table presents the results of four regression models for the dependent variable average time horizon for long-sighted derivatives For each independent variable, the coefficient and p-value are presented. Variables significant at the 1%-level, 5%-level and 10%-level, are market with *, ** and *** respectively. Below each model, a probability measure; a measure to evaluate the goodness of fit and the number of respondents included is presented. The independent variables used are: FIINDEX an index, 0-100, containing five variables for the level of foreign involvement, REALOPT variable ranging from 0-7 depending on how many real options the company is considering to use within the next 4 years, FLEXCOST ordered variable ranging from 1-7 depending on the amount of costs the company is able to change, BUSS ordered variable coded 1= very volatile, 2= fairly volatile. 3= not particularly stable or volatile, 4= fairly stable, 5= very stable, VAR - ordered variable coded 1= very likely, 2= likely. 3= neither likely nor unlikely, 4= unlikely, 5= very unlikely, NETEXP - the share of company's foreign operating costs(revenues) that are offset by matching foreign operating revenues(costs) on a scale from 0-100, COMP - ordered variable coded 1= dissimilar, 2= somewhat dissimilar. 3= neither similar nor dissimilar, 4= somewhat similar, 5= similar, EMP the natural logarithm of the number of employees, BAL the natural logarithm of total assets, SOLV the ratio of total equity to total assets, LEV the ratio of total debt to total assets. Proxy variables M-1 M-2 M-3 M-4 Coeff. p-value Coeff. p-value Coeff. p-value Coeff. p-value Foreign involvement (FIINDEX) 0,0005 0,9634 0,0083 0,4209 Real options (REALOPT) 0,0062 0,5008 0,0045 0,6046 0,0045 0,6026 Flexible costs (FLEXCOST) 0,0115 0,1968 0,0115 0,1799 0,0143 0,0994*** 0,0126 0,1331 Business environment (BUSS) -0,0189 0,0608*** -0,0177 0,0708*** -0,0139 0,1338-0,0103 0,1742 Variance of budgeted foreign revenues (VAR) -0,0044 0,5776 Net exposure (NETEXP) -0,0104 0,2294-0,0112 0,1371 Competitors currency cost structure (COMP) -0,0182 0,0274** -0,0201 0,0076* -0,0196 0,0107** -0,0214 0,0041* Number of employees (EMP) (See robustness test) Total balance (BAL) -0,0224 0,9410-0,0007 0,9980-0,0618 0,8361 Solvency ratio (SOLV) -0,0126 0,3365-0,0107 0,3798-0,0102 0,4237 Leverage (LEV) -0,1217 0,0265** -0,1234 0,0205** -0,1070 0,0446** -0,1050 0,0305** Prob(LR statistic) 0,0537 0,0231 0,0400 0,0052 Pseudo R-squared 0,1736 0,1706 0,1554 0,1419 n Average time horizon that operational means exceed that of financial means The following section presents an ordered regression model that explains the effects that various independent variables have on the following question: generally financial means (e.g. forward contracts, options, swaps and foreign debt) are used to hedge short and medium term, while operational means (e.g. changing suppliers and moving production) are used to hedge the long term. In the case of your firm, at what time 79

83 horizon would you say that the importance of operational means exceeds the importance of financial means for protecting your company against adverse changes in exchange rates? The dependent variable used in the regression is an ordered variable with values ranging from 1-7 as shown below. Value Average time horizon is months months months months years years 7...>4 years On average the respondents to this question answered that they considered the importance of operational means to exceed the importance of financial means at a time horizon of 12,2 months. The following models are specified and the results from the regressions are presented in table 32. Table 32: Estimated regression model for the time horizon at which the importance of operational means exceed financial means This table presents the results of four regression models for each of the two dependent variables average time horizon at which the importance of operational means exceed financial means. For each independent variable, the coefficient and p-value are presented. Variables significant at the 1%-level, 5%- level and 10%-level, are market with *, ** and *** respectively. Below each model, a probability measure; a measure to evaluate the goodness of fit and the number of respondents included is presented. The independent variables used are: FIINDEX an index, 0-100, containing five variables for the level of foreign involvement, REALOPT variable ranging from 0-7 depending on how many real options the company is considering to use within the next 4 years, CHANGECOST ordered variable ranging from 1-7 depending on the amount of costs the company is able to change, STAB ordered variable coded 1= very volatile, 2= fairly volatile. 3= not particularly stable or volatile, 4= fairly stable, 5= very stable, BUSVAR - ordered variable coded 1= very likely, 2= likely. 3= neither likely nor unlikely, 4= unlikely, 5= very unlikely, NETEXP - the share of company's foreign operating costs(revenues) that are offset by matching foreign operating revenues(costs) on a scale from 0-100, COMP - ordered variable coded 1= dissimilar, 2= somewhat dissimilar. 3= neither similar nor dissimilar, 4= somewhat similar, 5= similar, EMP the natural logarithm of the number of employees, BAL the natural logarithm of total assets, SOLV the ratio of total equity to total assets, LEV the ratio of total debt to total assets. Proxy variables M-1 M-2 M-3 M-4 Coeff. p-value Coeff. p-value Coeff. p-value Coeff. p-value Foreign involvement (FIINDEX) 0,0112 0,0532*** 0,0148 0,0056* 0,0135 0,0123** Real options (REALOPT) 0,0019 0,5901 0,0038 0,2541 0,0021 0,5479 Flexible costs (FLEXCOST) -0,0001 0,9758-0,0003 0,9340 0,0006 0,8751 Business environment (BUSS) -0,0071 0,0254** -0,0076 0,0164** -0,0069 0,0302** -0,0074 0,0176** Variance of budgeted foreign revenues (VAR) -0,0028 0,3635-0,0042 0,1612-0,0029 0,3405 Net exposure (NETEXP) -0,0049 0,1055-0,0072 0,0104** -0,0049 0,1044 Competitors currency cost structure (COMP) -0,0050 0,1033-0,0060 0,0494** -0,0046 0,1319-0,0053 0,0835*** 80

84 Number of employees (EMP) (See robustness test) Total balance (BAL) 0,1809 0,1805 0,2898 0,0176** 0,1468 0,2712 0,1517 0,2550 Solvency ratio (SOLV) -0,0078 0,0503*** -0,0077 0,0517*** -0,0075 0,0572*** -0,0080 0,0445** Leverage (LEV) -0,0045 0,2648-0,0048 0,2311-0,0040 0,3101 Prob(LR statistic) 0,0000 0,0001 0,0001 0,0000 Pseudo R-squared 0,0589 0,0531 0,0548 0,0553 n As in all the other regression, model 1 is a full model including all independent variables. In model 2 and 3, the variables for foreign involvement and net exposure were used separately due to correlation. Model 4 is a full model including all independent variables that at least have been significant at 10%-level. As we can see from table 32, the variable foreign involvement is significant at respectively a 10%, a 5% and a 1% level in the three models where it is included. The positive coefficient indicates that foreign involvement has a positive influence on the time horizon where operational means exceed the importance of financial means. In other words, companies that are more internationally involved tend to consider financial means important for a longer time horizon than companies with less foreign involvement. Additionally, the variable relating to the stability of a company s business environment is significant at 5%-levels in all models. The sign of the corresponding coefficients are negative, which means that, the more stabile the companies would characterize the business environment in which they operate to be, the shorter the time horizon where operational means exceeds the importance of financial means. Thus, companies in stabile surroundings consider operational means more important than financial means for managing foreign exchange rate exposures within a shorter horizon. The variable competitors currency cost structure is significant at 5% and 10%-levels in two models whereas it was insignificant in two other models. However it was very sensitive to the changes in the size variable, for which reason it cannot be considered as having a significant impact on the dependent variable. Solvency is significant at 5% level in model 4 whereas it is significant at a 10%-level in the first three models. The coefficient is negative, which means that there more solvent at company is the shorter is the time horizon where operational means exceeds the importance of financial means. Running a robustness check none of the above conclusion was changed and in fact all the variables became more significant. 81

85 All models are significant at 1%-level and having Pseudo R-squares from 0,0539 to 0,0589. As a summery of the present chapter, figure 8 presents an overview of all the independent variables and their impact on the different dependent variables used in the regression models. Where variables have had different significance levels in the different models, the average significance level is used. Variables that have only been significant at a 10% level in one of four models are not included. Figure 8: Regression results This figure shows all the independent variables and their influence on the five different dependent variables tested in the regression analyses. Arrows up indicate that the independent variable has a positive impact on the time horizon of the general FX exposure management and debt as well as the time horizon of the financial derivatives used to hedge FX exposure. Similarly, arrows up mean that operational hedges exceed the importance of financial hedges further in the future. Arrows down mean that the independent variable has a negative impact on the dependent variable. Variables that have been significant at different significance levels are categorized in the mean significance level. Variables that have only been significant at a 10%-level in one out of 4 models are not included. 82

86 6. Conclusion The present thesis contributes to the existing literature on corporate risk management in two ways. 1) by providing empirical evidence on derivatives use and the length of the hedging horizon in Scandinavian, medium-sized and large, non-financial companies when hedging their foreign exchange rate exposure using financial and operational means, and 2) through the statistical analysis of the determinants of the length of the hedging horizon of financial hedges as well as the investigation of the perceived horizon at which the benefits of operational hedges exceeds that of financial hedges for exchange rate exposure management. The empirical findings illustrate that when dealing with foreign currency derivatives and foreign debt for managing foreign exchange rate exposure, 63% of the Scandinavian companies used debt while 37% and 20% of the responding companies had used short- and long-sighted derivatives within the last two years. In total, 74% of the responding companies used either debt or financial derivatives. These findings are fairly consistent with findings found in Sweden (Hagelin 2003, Pramborg 2005) which showed that 53% and 71% of the Swedish firms in their respective samples used foreign debt for hedging. However, in terms of the use of currency derivatives, Swedish firms exhibit a higher frequency of use than the results found in this research as described in section Furthermore, the empirical data collected in the survey, supports the idea that forward contracts and options are most often used to hedge short term exposures while swaps and foreign debt is primarily used to hedge long term exposures. The results of the survey showed that while short-sighted derivatives (forward contracts and options) had an average maturity of 6,3 months, long-sighted derivatives and foreign debt had average maturities of 22,5 months and 16,6 months respectively, thus confirming our expectations. This thesis also examined the relative importance of financial and operational means when hedging foreign exchange rate exposure and found that operational means are considered appropriate for managing exposures above one year (12,2 months) whereas financial means are considered the best alternative for managing short term exposures of up to one year. Thus, incorporating the view of Srinivasulu as described in the introduction, firms should hedge exposures of up to one year using financial means and 83

87 use operational means to manage the exposures exceeding this horizon. Whether firms actually uses this approach would be an interesting topic for further analysis. When analysing the determinants of a company s horizon for which it hedges its foreign exchange rate exposure using financial means, we found mixed results. The investigation of the factors that influence the time horizon of a company s financial hedges in general found that the measure of a company s foreign involvement and the variable for the company s net exposure are significant in explaining the length of these hedges, both contributing to longer hedging horizons. Additionally, the variable for the stability of the business environment showed a moderately significant negative impact on the hedging horizon using financial means in general. The results confirm our hypothesis that foreign involvement has a positive influence on the horizon. The determinants of the horizon for a company s debt arrangements were also investigated. While solvency seem to be negatively related to the horizon of a company s foreign debt arrangements, the measure of foreign involvement as well as company size measured by total balance show a significant, positive influence on the horizon for the foreign debt arrangements of Scandinavian medium-sized and large, non-financial firms. The positive influence of foreign involvement and size confirm our hypotheses stated in section Additionally, the influence of foreign involvement extend the conclusion that firms with a high degree of international involvement use more foreign debt (Aabo, 2006) by showing that these companies also tend to issue debt with longer maturities. In terms of the influence of the independent variables on the horizon of short-sighted derivatives used for exchange rate exposure management, the two factors relating to stability had significant influence on the horizon. While the variable relating to budget variance had a positive impact on the horizon, the variable for the stability of the general business environment had a negative impact on the horizon. The fact that the two variables had opposite signs was surprising, as we would have expected the two variables to show the same relationship due to the fact that they are both used as proxies for the underlying latent variable namely, stability. When analysing the factors influencing the horizon of long-sighted derivatives, we find the variable representing the similarity of competitors currency cost structure to be a clear indicator of the horizon for long-sighted derivatives with a negative coefficient, indicating that the more similar your company s cost structure is to that of your 84

88 competitors, the shorter your horizon. Additionally, leverage is significant at the 5%- level also with a negative influence on the horizon. Finally, the relationship between the importance of financial and operational means for managing foreign exchange rate exposure is investigated. The findings this investigation suggest that companies with high levels of foreign involvement consider financial hedges more important than operational hedges for longer horizons whereas the industry stability and solvency show a negative relationship suggesting that financial means are considered important for a shorter horizon. Surprisingly, the overall results of our analysis suggest that the two proxies for flexibility do not have any significance in explaining the time horizon for any of our dependent variables. This result is surprising as flexibility was hypothesized to have a negative influence on the horizon. Similarly, the variable for the stability of the general business environment which was expected to have a positive influence on the horizon turned out to have a significantly negative influence in three of the five models. On the other hand, the positive influence of foreign involvement on the time horizon is partly confirmed as the variable is significant in explaining the horizon in three of the five models. Potential shortcomings of this include a bias towards large companies as illustrated by the results of our response bias test. This consideration was underlined by the section on descriptive statistics for the responding companies which indicated that the sample contained outliners in terms of firm size. Additionally, the research includes a certain degree of subjectivity in some of the survey questions such as the use of long-sighted derivatives in the question regarding the maturity of swaps. Generally speaking this does not necessarily make our results any less valuable. However, certain care should be taken when interpreting the results. As mentioned previously in this section, further analysis into the possible relationship between the exercise horizon for real options and the use of financial derivatives to hedge the interval between could be an interesting extension of the present research. Another potential extension of the present research would be to include the scope of the hedges. Lastly, the geographical area of the research could be reduced in order to account for the potential differences between countries. 85

89 Bibliography Aabo, T. (1999). Exchange Rate Exposure Management: An emperical Study into the Strategies and Practices of Industrial Companies. Aarhus School of Business, International Studies. Aarhus: Aarhus School of Business. Aabo, T. (2006). The Importance of Corporate Foreign Debt in Managing Exchange Rate Exposures in Non-Financial Companies. European Financial Management, vol.12 no. 4, Aabo, T., & Simkins, B. (2003). Interaction between real options and financial hedging: Fact or fiction in managerial decision-making. Review of Financial Economics December, 2003, p Adler, M., & Dumas, B. (1984). Exposure to Currency Risk: Definition and Measurement. Financial Management, 13 (2) p Alkebäck, P., & Hagelin, N. (1999). Derivativ usage by Non-financial firms in Sweden with an international comparison. Joumal of International Financial Management and Accounting, 10, 2, 1999, p Allayannis, G., & Ofek, E. (2001). Exchange rate exposure, hedging, and the use of foreign currency derivatives. Journal of International Money and Finance, 20, Allayannis, G., & Weston, J. P. (2001). The Use of Foreign Currency Derivatives and Firm Market Value. The Review of Financial Studies, 14, no. 1, 2001, p Berk, J., & DeMarzo, P. (2007). Corporate Finance. California: Pearson Addison Wesley.2e Block, S. B., & Gallagher, T. J.(1986) The use of interest rate futures and options by corporate financial managers, Financial Management (15) 1986, Bodnar, G. M., & Gebhardt, G. (1999). Derivatives Usage in Risk Management by US and German Non-Financial Firms: A Comparative Survey. Journal of International Financial Management and Accounting, 10 (3), p Bodnar, G. M., Hayt, G. S., & Marston, R. C. (1998). Wharton Survey of Financial Risk Management by U.S. Non-financial Firms. Financial Management, 27 (4), Brealey, R. A., Myers, S. C., & Marcus, A. J. (2007). Fundamentals of corporate finance. London Business School: McGraw-Hill. Brouthers, K. D. (1995). The Influence of International Risk on Entry Mode Strategy in the Computer Software Industry. Management international reviews, vol 35, 1995, p Calvo, G. A., & Reinhart, C. M. (2002). Fear of Floating. The Quartely Journal of Finance, vol 117, no. 2, May, 2002 p Capel, J. (1997). A Real Option approach to Economic Exposure Management. Journal of International Financial Management and Accounting, 8, 2 p

90 Clark, E., & Judge, A. (2009). Foreign Currency Derivatives versus Foreign Currency Debt and the Hedging Premium. European Financial Management, 15 (3), Culp, C. L. (2001). The Risk Management Process. New York: John Wiley & Sons, Inc. Dufey, G., & Srinivasulu, S. L. (1983). The case for corporate management of Foreign Exchange Risk. Financial Management, 12, p Eiteman, D. K., Stonehill, A. I., & Moffett, M. H. (2007). Multinational Business Finance (Vol. 11). Los Angeles: Pearson. Addison Wesley. Engel, C., Deveraux, M. B., & Storgaard, P. E. (2004). Endogenous exchange rate passthrough when nominal prices are set in advance. Journal of International Economics (63), Eurostat. (n.d.). Eurostat. Retrieved 01 08, 11, from &tableselection=1&footnotes=yes&labeling=labels&plugin=1 Flood, E. J., & Lessard, D. R. (1986). On the meassurement of operating exposure to foreign exchange rate: A conceptual aproach. Financial Management, 15 (1), Froot, K. A., Scharfstein, D. S. and Stein, J. C., (1993) Risk Management: Coordinating Corporate Investment and Financing Policies. The Journal of Finance 48, 5, pp Froot, K. A., & Klemperer, P. D. (1989). Exchange rate pass-through when market share matters. American Economic Review, vol 79, 4 p Froot, K. A., Scharfstein, D. S., & Stein, J. C. (1994). A framework for risk management. Journal of applied corporate finance, vol 7 (3) p Géczy, C.; Minton, B. A.; Schrand, C. (1997): Why Firms Use Currency Derivatives, The Journal of Finance, Vol. 52, No. 4, September 1997, p Grant, R., & Soenen, L. (2004). Strategic Management of operating exposure. European Managent Journal, vol 27, no.1 p Hagelin, N. (2003). Why firms hedge with currency derivatives, an examination of transaction and translation exposure. Applied Financial Economics 13, 2003, p Hair, J., Babin, B., Money, A. and Samouel, P., Essentials of Business Research Methods. Wiley, US., (2003). Hakkarainen, A., Joseph, N., Kasanen, E. and Puttonen, V., (1998). The Foreign Exchange Exposure Management Practices of Finnish Industrial Firms Journal of International Financial Management and Accounting 9, 1, pp Heij, C., De Boer, P., Franses, P. H., Kloek, T., & Van Dijk, H. K. (2004). Econometric Methods with Applications in Business and Economics. Oxford: Oxford University Press. Hollesen, S. (2010), Global Marketing: A Decision-Oriented Approach, Prentice Hall, 4ed 87

91 IASCF, I. A. (2008). A Guide through International Financial Reporting Standards. Nielsen. Jorion, P. (1990). The Exchange-Rate Exposure of U.S. Multinationals. Journal of Business, 63, Kedia, S., & Mozundar, A. (2003). Foreign currency-denominated debt: an empirical examination. Journal of Business, 76, Lessard, D. R., & Lightstone, J. B. (1986, July-August). Volatile Operatins can put operations at risk. Harvard Business Review, Lidbark, J. (2010, 06 21). Finance-magazine. Retrieved 01 10, 2011, from Møller Jensen, J., & Knudsen, T. (2006). Analyse af spørgeskemadata med SPSS teori, anvendelse og praksis. Odense: Syddansk Universitetsforlag. Magee, S. P. (1973). Currency Contracts, Pass Through and Devaluation. Brookings Paper on Economic Activity, 1, Marston, R. C. (1996). The effects of industry structure on economic exposure. Wharton School, University of Pennsylvania, March, 1996, Meulbroek, L. K. (2002). A Senior Manager s Guide to Integrated Risk Management. Journal of Applied Corporate Finance, vol 14, no. 4, p Miller, K. D. (1998). Economic Exposure and Integrated Risk Management. Strategic Management Journal, vol. 19, 1998, p Modigliani, F., & Miller, M. H. (1958). The cost of capital, corporation finance and the theory of investment. The American economic review, vol 28, June, p Muller, A., & Verschoor, W. F. (2005). Foreign exchange risk exposure: Survey and suggestions. Journal of multinational Financial Management, Nance, D. R., Smith, C. W., & Smithson, C. W. (1993). On the Determinants of corporate hedging. The Journal of Finance, vol 48, no.1, March 1993, p Nydahl, S. (1999). Exchange rate exposure, Foreign Involvement and Currency hedging. European Financial Management, Oxelheim, L., & Wihlborg, C. G. (1987). Macroeconomic Uncertainty: International Risks and Opportunitiesfor the Corporation. New York: John Wiley and Sons. Pramborg, B. (2005): Foreign exchange risk management by Swedish and Korean non-financial firms: A comparative survey, Pacific-Basin Finance Journal, 13, 2005, p Rogoff, K. (1996). The Purchasing Power Parity Puzzle. Journal of Economic Litterature, 2, Smith, C. W., & Stulz, R. M. (1985). The Determinants of Firms' Hedging Policies. Journal of financial and quantitative analysis, 20 (5),

92 Srinivasulu, S. L. (1981). Strategic Response to Foreign Exchange Rate Risks. Columbia Journal of World Business, spring,p Stulz, R. M. (2003). Risk Management and Derivatives (Vol. 1). Thomson South Western. Stulz, R., & Williamson, R. (1996). Identifying and Quantifying exposures. Ohio: The Charles A Dice Center for Research in Financial Economics, August 1996, p Triantis, A. J. (2005). Corporate Risk Management: Real options and financial hedging. In M. Frenkel, U. Hommel, & M. Rudolf, Risk Management (pp ). Berlin: Springer-Verlag. Warner, J. (1977). Bankruptcy costs: Some evidence. Journal of Finance (32), Wooldridge, J. M. (2006), Introductory Econometrics: A Modern Approach. Third edition. Thomson South-Western Jiawen, Yang (1997), Exchange rate pass-through in U.S manufacturing industries, the review of economics and statistics, vol. 79 (1) p

93 Appendix Appendix 1: Export share for the Scandinavian countries Table 33: Export share for the Scandinavian countries This table represents the total export of goods and services in % of GDP. The average for each decade is calculated as by the numbers of e.g It shows that all the Scandinavian countries all have averages above the entire world as well as countries within the European Union. The numbers for the European Union are calculated by the average of all the companies which where within the EU at the end of each year. Country/Region Average 1980's Average 1990's Average 2000's Denmark 35,5% 38,2% 48,8% Finland 27,8% 32,7% 42,2% Norway 39,3% 39,1% 44,3% Sweden 33,3% 36,2% 48,0% European Union 26,9% 29,2% 36,9% World 18,4% 20,5% 25,9% Source: Own contribution and data collected from: I

94 Appendix 2: NACE classification system Figure 9: NACE rev. 2 classification system This table represents the 21 main categories of the NACE classification system. All of the categories in the figure, except category K, are included in the main selection criteria of this thesis. II

95 Appendix 3: Complete questionnaire PART I. INTERNATIONAL INVOLVMENT AND FOREIGN EXCHANGE RATE EXPOSURE Q1. Approximately what percentage of your company's consolidated revenues and costs are in foreign currency? (Please choose the option that is closest to your estimate)* Consolidated revenues Consolidated costs 0% 1-20% 21-40% 41-60% 61-80% 81-99% 100% Q2. Approximately what percentage of your company's consolidated assets is in foreign currency? (Please choose the option that is closest to your estimate) * Consolidated assets 0% 1-20% 21-40% 41-60% 61-80% 81-99% 100% Q3. At the present time, what is the average time horizon that your company has covered its foreign exchange rate exposures by use of financial means?(i.e. forward contracts, options, swaps and debt in foreign currency) (Please choose the most appropriate option according to your estimate) * 0-1 months 1-3 months 3-6 months 6-12 months 1-2 years 2-4 years >4 years Q4. How many subsidiaries does your company have abroad? (Please choose the most appropriate option) III

96 >24 Production subsidiaries abroad Sales subsidiaries abroad Q5. How many foreign currencies is your company significantly exposed to? (Please choose the most appropriate option)* Number of significant currencies >9 Q6. Which foreign currencies are the most important ones? (Please choose one option per row or write the currency in the field provided) * Most important 2nd most important 3rd most important 4th most important Euro DKK NOK SEK Pound $ Dollar Yen Rouble Other If other, please state Q7. Within the last year, has your company's foreign operating revenues been larger or smaller that your company's foreign operating costs? (Please choose the most appropriate option)* Foreign operating revenues have been larger than foreign operating costs Foreign operating costs have been larger than foreign operating revenues Q8a. How much of your company's foreign operating revenues are offset by matching foreign operating costs? (Please choose the most appropriate option according to your estimate)* 0% 1-20% 21-40% 41-60% 61-80% 81-99% 100% Foreign operating revenues matched by foreign operating costs IV

97 Q8b. How much of your company's foreign operating costs are offset by matching foreign operating revenues? (Please choose the most appropriate option according to your estimate)* 0% 1-20% 21-40% 41-60% 61-80% 81-99% 100% Foreign operating costs matched by foreign operating revenues PART II. DEBT AND DERIVATIVES USE FOR MANAGING FOREIGN EXCHANGE RATE EXPOSURES Q9 Approximately what percentage of your company's consolidated debt is in foreign currency? (Please choose the option that is closest to your estimate) Consolidated foreign debt 0% 1-20% 21-40% 41-60% 61-80% 81-99% 100% Q10. What is the average loan period for your company's foreign debt arrangements? (Please choose the appropriate option according to your estimate) 0-1 months 1-3 months 3-6 months 6-12 months 1-2 years 2-4 years >4 years Q11. Within the last two years, has your company used financial derivatives (e.g. forward contracts, options or swaps) to hedge your company's foreign exchange rate exposures? (Please choose the most appropriate option)* Short-sighted derivatives (forward contracts and options) Long-sighted derivatives (swaps) Yes No Q12. What is the average maturity for the short-sighted derivatives (forward contracts and options) used to cover your company's foreign exchange rate exposure? (Please choose the most appropriate option according to your estimate)* 0-1 months 1-3 months 3-6 months 6-12 months 1-2 years 2-4 years >4 years V

98 exposure? (Please choose the most appropriate option according to your estimate)* 0-1 months 1-3 months 3-6 months 6-12 months 1-2 years 2-4 years >4 years Q13. What is the average maturity for the long-sighted derivatives (swaps) used to cover your company's foreign exchange rate exposure? (Please choose the most appropriate option according to your estimate)* 0-1 months 1-3 months 3-6 months 6-12 months 1-2 years 2-4 years >4 years Q14. Generally financial means (e.g. forward contracts, options, swaps and foreign debt) are used to hedge the short and medium term, while operating means (e.g. changing suppliers and moving production) are used to hedge the long term. In the case of your firm, at what time horizon would you say that the importance of operational means exceeds the importance of financial means for protecting your company against adverse changes in exchange rates? (Please choose the most appropriate option according to your estimate) * 0-1 months 1-3 months 3-6 months 6-12 months 1-2 years 2-4 years >4 years PART III. FIRM FLEXIBILITY AND INDUSTRY STABILITY Q15. How would you characterize the general business environment in which your company primarily operates? (Please choose the most appropriate option) * Very stable Fairly stable Not particularly stable or volatile Fairly volatile VI

99 Very volatile Q16. In the coming 4 years, is it likely that your company will undertake any of the following actions partly or fully due to exchange rate changes? (Please choose one option in each row) * Enter foreign market in which your company did not operate before Abandon a foreign market (stop selling in that market) Change sourcing between suppliers in different countries Shift production between production outlets in different countries Delay entry into a foreign market Temporally reduce or close operations in a foreign market Increase capacity and/or extend production in a foreign country Yes No Currency cost structure: the composition/combination of currencies in which your company has its primary foreign operating costs. Q17. To what degree is your competitors' currency cost structure similar to your company's currency cost structure? (Please choose the most appropriate option) * Similar Somewhat similar Neither similar or dissimilar Somewhat dissimilar Dissimilar Q18. When your company hedges its foreign exchange rate exposure, what is your company primarily trying to manage with the hedge? (Please choose the most appropriate option) * Volatility of accounting earnings and the market value of the company Volatility of cash flows Q19. How likely is it that your actual foreign revenues will deviate by +/- 10% from your VII

100 company's budgeted foreign revenues within the next 12 months? (Please choose the most appropriate option)* Very likely Likely Neither likely nor unlikely Unlikely Very unlikely Q20. How much of your cost base in foreign currency could your firm change (without prohibitively high adjustment costs) within the next 12 months if exchange rates move against your firm? (Please choose the most appropriate option according to your estimate)* Changeable foreign currency 0% 1-20% 21-40% 41-60% 61-80% 81-99% 100% Appendix 4: Screenshots from the StudSurvey tool Screenshot 1: Start page and stats for the survey responses Screenshot 2: Form builder, in which questions and answer categories could be constructed VIII

101 Screenshot 3: Sending the invitation to participate Screenshot 4: Tracking respondents & non-respondents IX

102 X

103 Appendix 5: Correlation tables Table 34: Correlation table chapter 5.2 This table represents the correlation coefficients for all the independent variables that are included in the ordered probit regression model in chapter 5.2. The variables are Foreign involvement index (FIINDEX), Real options (REALOPT), Flexible costs (FLEXCOST), Business environment (BUSS), Variance of foreign budgeted revenues (VAR), Net exposure (NETEXP), Competitors currency cost structure (COMP), Numbers of employees (EMP), Total balance (BAL), Solvency ratio (SOLV) and Leverage ratio (LEV). The marked variables have correlations above 0,4. Table 35: Correlation table chapter 5.3 This table represents the correlation coefficients for all the independent variables that are included in the ordered probit regression model in chapter 5.3. The variables are Foreign involvement index (FIINDEX), Real options (REALOPT), Flexible costs (FLEXCOST), Business environment (BUSS), Variance of foreign budgeted revenues (VAR), Net exposure (NETEXP), Competitors currency cost structure (COMP), Numbers of employees (EMP), Total balance (BAL), Solvency ratio (SOLV) and Leverage ratio (LEV). The marked variables have correlations above 0,4. XI

104 Table 36: Correlation table chapter 5.4 This table represents the correlation coefficients for all the independent variables that are included in the ordered probit regression model in chapter 5.4. The variables are Foreign involvement index (FIINDEX), Real options (REALOPT), Flexible costs (FLEXCOST), Business environment (BUSS), Variance of foreign budgeted revenues (VAR), Net exposure (NETEXP), Competitors currency cost structure (COMP), Numbers of employees (EMP), Total balance (BAL), Solvency ratio (SOLV) and Leverage ratio (LEV). The marked variables have correlations above 0,4. Table 37: Correlation table chapter 5.5 This table represents the correlation coefficients for all the independent variables that are included in the ordered probit regression model in chapter 5.5. The variables are Foreign involvement index (FIINDEX), Real options (REALOPT), Flexible costs (FLEXCOST), Business environment (BUSS), Variance of foreign budgeted revenues (VAR), Net exposure (NETEXP), Competitors currency cost structure (COMP), Numbers of employees (EMP), Total balance (BAL), Solvency ratio (SOLV) and Leverage ratio (LEV). The marked variables have correlations above 0,4. XII

105 Table 38: Correlation Q14 This table represents the correlation coefficients for all the independent variables that are included in the ordered probit regression model in chapter 5.6. The variables are Foreign involvement index (FIINDEX), Real options (REALOPT), Flexible costs (FLEXCOST), Business environment (BUSS), Variance of foreign budgeted revenues (VAR), Net exposure (NETEXP), Competitors currency cost structure (COMP), Numbers of employees (EMP), Total balance (BAL), Solvency ratio (SOLV) and Leverage ratio (LEV). The marked variables have correlations above 0,4. XIII