In this paper Swedish listed companies' use of capital budgeting and cost of capital estimation methods in 2005 and 2008 are examined. The relation between company characteristics and choice of methods is investigated and both within-country longitudinal and cross-country comparisons are made. Larger companies seem to have used capital budgeting methods more frequently than smaller companies. When compared to U.S. and continental European companies, Swedish listed companies employed capital budgeting methods less frequently. In 2005 the most common method for establishing the cost of equity was by asking the investors what return they required. By 2008 CAPM was instead the most utilised method, which could indicate greater sophistication. The use of project risk when evaluating investments also seems to have gained in popularity, while the use of company risk declined. Overall, the use of sophisticated capital budgeting and cost of capital estimation methods seem to be rising and the use of less sophisticated methods declining.Keywords: Capital Budgeting Method; Cost of Equity; Project Risk; Swedish-listed Companies
INTRODUCTIONophisticated capital budgeting methods are often "highly recommended" by financial management textbooks, e.g. net present value (NPV), whereas others that are simpler are not, e.g. undiscounted payback (e.g. Brealey and Myers, 2003;Lumby and Jones, 2003;Ross et al., 2005;Smart et al., 2007). Theoretical advice is also given on how to estimate the cost of capital. Estimation models like the capital asset pricing model (CAPM) are frequently referred to and instructions how to calculate project discounts rates, contingent on the underlying risk of the project, instead of using a single company discount rate, are presented (e.g. Brealey and Myers, 2003;Lumby and Jones, 2003;Ross et al., 2005;Smart et al., 2007). Furthermore, multinationals are advised to consider risk factors such as political and foreign exchange risk (Shapiro, 2002;Eiteman et al., 2006). Still, the practical use of economic models can, as earlier research has shown (Graham and Harvey, 2001;Sandahl and Sjögren, 2003;Brounen et al., 2004;Danielson and Scott, 2006), deviate from what is prescribed by normative theory. Management, possibly with other goals than the principal (Jensen and Meckling, 1976), decides whether the recommended methods should be employed or not. The goal incongruence together with the information asymmetry between management and the shareholders, might lead to different practical behaviour than what is recommended by textbooks (Narayanan, 1985). And moreover, also in an ideal world where the principalagent problem does not exist or is limited, there might be a discrepancy between theory and practice. Methods that from a theoretical perspective at first sight have flaws can in practice namely be efficient and lead to shareholder wealth maximisation. Guthrie (1997, 2006) showed, for example, that the employment of the payback method could be rational since the payback time approximates the optio...