Hedging Behavior in Small and Medium-sized Enterprises The Role of Unobserved Heterogeneity.docx

《Hedging Behavior in Small and Medium-sized Enterprises The Role of Unobserved Heterogeneity.docx》由会员分享，可在线阅读，更多相关《Hedging Behavior in Small and Medium-sized Enterprises The Role of Unobserved Heterogeneity.docx（53页珍藏版）》请在淘文阁 - 分享文档赚钱的网站上搜索。

1、* Joost M.E. Pennings is an associate professor in the Department of Agricultural & Consumer Economics, Office for Futures & Options Research at the University of Illinois at Urbana-Champaign and the AST Distinguished professor in Commodity Futures Markets in the Department of Marketing & Consumer B

2、ehavior at the Wageningen University in the Netherlands. Philip Garcia is the Thomas A. Hieronymus professor in Futures Markets with the Office for Futures & Options Research at the University of Illinois at Urbana-Champaign. Correspondence should be addressed to: Joost M.E. Pennings, Office for Fut

3、ures & Options Research, University of Illinois at Urbana-Champaign, 326 Mumford Hall, MC-710, 1301 W. Gregory Drive, Urbana, IL 61801. 0 Hedging Behavior in Small and Medium-sized Enterprises: The Role of Unobserved Heterogeneity Joost M.E. Pennings and Philip Garcia* Forthcoming Journal of Banking

4、 & Finance Copyright Elsevier Science * Joost M.E. Pennings is an associate professor in the Department of Agricultural & Consumer Economics, Office for Futures & Options Research at the University of Illinois at Urbana-Champaign and the AST Distinguished professor in Commodity Futures Markets in th

5、e Department of Marketing & Consumer Behavior at the Wageningen University in the Netherlands. Philip Garcia is the Thomas A. Hieronymus professor in Futures Markets with the Office for Futures & Options Research at the University of Illinois at Urbana-Champaign. Correspondence should be addressed t

6、o: Joost M.E. Pennings, Office for Futures & Options Research, University of Illinois at Urbana-Champaign, 326 Mumford Hall, MC-710, 1301 W. Gregory Drive, Urbana, IL 61801. 1 Hedging Behavior in Small and Medium-sized Enterprises: The Role of Unobserved Heterogeneity Joost M.E. Pennings and Philip

7、Garcia* Abstract We investigate factors that drive derivative usage in small and medium-sized enterprises (SMEs). The influence of these factors on hedging behavior cannot a priori be assumed equal for all SMEs. To address this heterogeneity, a generalized mixture regression model is used which clas

8、sifies firms into segments, so that the hedging response to the determinants of derivative usage is the same within each segment. Using a unique data set of 415 SMEs, containing both accounting and experimental data, we find that factors like risk exposure, risk perception, risk attitude, and the de

9、cision-making unit, among others are useful in explaining hedging behavior. However, the effects of these factors are not homogeneous across all managers, and the roots of the heterogeneity can partially be traced to differences in attitudes, perceptions, and to differences in ownership structure. J

10、EL Classification: C1; G0 Keywords: Derivatives Usage; Hedging Behavior; Unobserved Heterogeneity 2 1. Introduction Financial derivatives, such as futures and options, provide managers with tools to manage price risk. Derivative exchanges and financial institutions facilitate the exchange of these i

11、nstruments through over-the-counter trading. Recently, the competition among financial institutions that provide these services has increased, leading to customized financial products which fulfill user needs better. Accordingly, the interest of financial institutions in identifying the motivation b

12、ehind the derivative usage of different groups of (potential) users has increased (e.g., Fridson, 1992; Angel, Gastineau and Weber, 1997; Nesbitt and Reynolds, 1997). Froot, Scharfstein and Stein (1993), Nance, Smith and Smithson (1993), Mian (1996), Tufano (1996), Gczy, Minton and Schrand (1997), L

13、ee and Hoyt (1997), and Koski and Pontiff (1999) among others have conducted research on the determinants of derivative use. In these studies, large, often publicly-traded companies have been examined. Here, we expand the literature by studying the derivative usage of managers of small and medium- s

14、ized enterprises (SMEs). SMEs do not have different departments, nor do they have separate organizational structures to administer functions such as research & development, quality control, sales, and accounting. The management of these functions rests on one single manager. Moreover, the ownership

15、of SMEs is often concentrated. In such a structure, a managers risk aversion can provide an important motivation to manage risk (Mayers and Smith, 1982; Smith, 1995). The wealth of the manager often is directly affected by the variance of the SMEs expected profit, constituting an (extra) motivation

16、to consider hedging 3 (Smith and Stulz, 1985).1 SMEs also differ from large corporations in their capital structure, as bondholders are relatively scarce. Avery and Bostic (1998) and Berger and Udell (1998) show that private equity, bank loans, and personal commitments dominate the capital structure

17、 of SMEs. These differences motivate the importance of considering the managers, along with the firms, characteristics in the investigation of the determinants of derivative usage. We also build on previous work, by incorporating the notion that the motivations of enterprises to use derivatives as a

18、 hedging tool may not be homogenous. Firms from different regions or of different organizational structures may face dissimilar economic constraints and conditions that might lead to a different choice of derivatives. Similarly, managers may possess dissimilar objectives and motivations that can als

19、o result in different derivative decisions. This could be particularly relevant for SMEs, as they show a wide variety of organizational structures. Furthermore, managers of SMEs may have different risk attitudes and risk perceptions, suggesting that these firms may behave differently (e.g., Pennings

20、 and Smidts, 2000). Consequently, we may expect the factors that influence a firms choice of financial instrument to vary across segments of an industry, and common factors to influence firms differently. Clearly, this heterogeneity impacts the efforts of financial institutions in developing appropr

21、iate derivatives, particularly for customized products. Here, we model the effects of “unobserved heterogeneity” on the determinants of firms derivative usage. The term “unobserved heterogeneity” posits two interrelated ideas that are central to the empirical procedure we employ. The first notion is

22、 that not all managers respond similarly to a given change in the determinants of derivative use, but instead that 1 This is consistent with the notion that firms whose portfolios are poorly diversified have a stronger incentive to hedge (Smith, 1995). 4 segments of managers who behave in a similar

23、manner may exist. The second notion is that these segments are not directly observable prior to the analysis. Rather, they are determined by grouping together managers who reveal a similar relationship between the determinants of derivative use and their hedging behavior. In this context, we present

24、 a generalized linear mixture model that simultaneously investigates the relationship between managers derivative usage and a set of explanatory variables for each identified segment in the sample. We demonstrate how managers behave differently regarding derivative usage, and show the importance for

25、 financial institutions to develop an understanding of their customers. We focus exclusively on derivative usage for reducing price risk when buying inputs and selling outputs, and thus do not consider hedging motivated by risky investment projects. Our empirical investigation is in the raw food ind

26、ustry, and, as such, derivative use refers to commodity derivative usage. We pay special attention to the fundamental motivation behind derivative usage as a hedging tool: risk attitude, risk perception and their interaction. Although these factors are recognized in theory as being crucial in motiva

27、ting and explaining derivative usage, to date risk attitudes and risk perceptions rarely appear in empirical studies of derivative usage. Their absence can primarily be explained by two reasons. First, most studies focus on large corporations, rather than on their managers, thus concentrating on fir

28、m characteristics. Second, risk attitudes and risk perceptions are not directly observable and cannot be obtained from accounting data. Measuring these concepts in a realistic and accurate manner is a difficult task. Here, we measure risk attitudes of 415 SMEs in a relevant economic business setting

29、 using computer-guided interviews and an experimental design based on the expected utility model. 5 This paper is organized as follows: Section 2 provides an overview of the determinants that are hypothesized to influence the derivative usage of SMEs. Section 3 discusses the heterogeneity in the rel

30、ationship between the determinants of derivative usage and hedging behavior, followed in Section 4 by the statistical specification of the generalized-mixture model that is able to empirically identify this type of heterogeneity. Section 5 describes the sample and the measurements, in particular the

31、 elicitation of the managers global risk attitude. Section 6 presents the empirical results. A discussion of the findings follows in Section 7. 2. The Determinants of SMEs Derivative Hedging Usage First, we provide background information on the decision context of SMEs operating in a commodity marke

32、ting channel. In particular, we elaborate on the Dutch pork industry, our empirical domain. Subsequently, we discuss the factors that drive derivative usage in this decision context. 2.1. Decision Context Previous studies have focused on the derivative usage of large corporations (e.g Tufano, 1996;

33、Gczy, Minton and Schrand, 1997; Koski and Pontiff, 1999; Rogers, 2002). In these studies the risk-management behavior of CEOs is explained by various factors, among others firm size, CEO risk-taking incentives, tax schedules, and financial distress costs. Here we study hedging behavior of SMEs opera

34、ting in the same commodity marketing channel. This decision context differs from that of large companies, and factors not considered for large companies, such as education level of the manager(s) of the SME, and the influence of the 6 SMEs decision-making unit, might be relevant for SMEs. In additio

35、n, the psychological concepts of managers risk attitudes and risk perceptions may be particularly relevant for SMEs. In this paper, we study derivative usage for companies in the Dutch pork marketing channel. This marketing channel consists of producers (hog farmers), wholesalers (companies that tra

36、de live hogs) and processors (slaughterhouses and meat packers). This type of marketing channel exists for a wide variety of commodities, such as soybeans, wheat, beef and cotton. The companies in commodity marketing channels, especially the producers and wholesalers, are relatively small. For examp

37、le, the average sales of the Dutch hog producers in our sample was $185,000 in fiscal year 1997 (see Table 1). Most producers and wholesalers are family owned, with the manager often the owner. Processors are more diverse, but are generally larger than producers and wholesalers, and consist of priva

38、te- limited companies, as well as sometimes publicly-traded companies (see Table 1). The production process is relatively straightforward in these commodity channels. In the Dutch pork marketing channel, producers raise piglets to hogs which are sold to the wholesalers, who then sell them to the pro

39、cessor. The majority of these transactions are spot market transactions; cash-forward contracts are rare. This commodity marketing channel reflects the decision context of the early work done on hedging behavior (e.g., Blau, 1944; Johnson, 1960; Working, 1962). Companies in a commodity marketing cha

40、nnel are exposed to the spot price risk of the commodity. With a coefficient of variation (CV) of 0.19, the Dutch hog prices fluctuate widely (based on daily observations over the period 1990-97), even compared to the prices of US soybeans (CV 0.14.), which is generally considered to be a risky comm

41、odity. The ex ante risk exposure of these firms is determined by hog price fluctuations and the number of times 7 that they enter the spot market. Wholesalers and processors enter the hog spot market on a weekly or sometimes daily basis, in contrast to hog producers, who, depending on the production

42、 system employed, may enter the spot market as few as 4 times a year (since piglets are raised to slaughter hogs in three to four months) (see Table 1). At present, there is only one risk management tool available: the hog futures contract traded at Euronext (the result of a merger of the exchanges

43、in Amsterdam, Brussels, London and Paris) and the pork belly futures contract traded at the Chicago Mercantile Exchange. Several factors have been identified to explain why firms use derivatives as a hedging tool. The combined work of Froot, Scharfstein and Stein (1993), Nance, Smith and Smithson (1

44、993), Mian (1996), Tufano (1996, 1998), Gczy, Minton and Schrand (1997), Lee and Hoyt (1997), Koski and Pontiff (1999), and recently Graham and Rogers (2002) and Rogers (2002) among others provide a discussion of these factors. This paper provides a brief overview of the primary determinants of deri

45、vative usage relevant for the empirical decision context (commodity marketing channel of SMEs), concentrating on the use of (commodity) derivatives as a means to reduce SMEs input and output price risk. Particular attention is given to the “fundamental determinants” behind risk management and the us

46、e of derivatives: risk attitude and risk perception. While previous studies have accounted for managerial risk aversion indirectly, by measuring risk aversion through proxies like officers and director share ownership (Tufano, 1996), we focus on managers risk attitudes, explicitly recognizing that r

47、isk attitude is a psychological concept. We hypothesize that the risk-attitude concept is particularly important for managers of SMEs (e.g., Pennings and Smidts, 2000). We first discuss the influence of the managers risk attitude and risk perception on derivative use, as well as the managers educati

48、on level, followed by the characteristics of the firm. 8 2.2. Managers Characteristics Influencing Derivative Usage Risk aversion has been a key element in understanding hedging behavior. Marshall (1919), Keynes (1930), Hicks (1939) and Kaldor (1939) in the first part of the previous century, argued

49、 that hedging was motivated by risk aversion. Using normative models, various authors have shown that the hedge ratio is determined by the decision-makers risk attitude (Ederington, 1979). The well known mean-variance models illustrate this relation between risk attitude and hedging behavior (Levy and Markowitz, 1979). Hence, we expect Risk Attitude (RA) to be an important determinant of an SMEs hedging behavior. Risk aversion refers to a preference for a guaranteed out