Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. Executive Summary This thesis is motivated by the challenges of contemporary finance in appraising companies according to their underlying, intrinsic value. Scholars and practitioners within the field of financial analysis have searched in futility for an efficient and accurate valuation technique; a single, magic wand that could estimate companies’ intrinsic value accurately. Instead, several competing contemporary valuation methods occupy the literature today. The purpose of the thesis is to find the intrinsic value of Vestas. To do so, this thesis will evaluate and discuss contemporary valuation models in order to choose the most appropriate models. Vestas is of particular interest being a Danish market leader within a major growth industry. Under the current momentum that wind power enjoys, it is relevant to examine if the intrinsic value of the company deviates from the current market price. In order to answer the thesis statement, the environmental, competitive, and internal situation of Vestas will be addressed. The strategic analyses reveal the profitable prospects for the industry, but also disclose an enormous dependence on political support. Vestas also faces an increasingly tougher competition. A historical financial analysis reveals that Vestas’ has increased profitability, mainly due to low working capital and an improved gross margin. By synthesising the strategic and financial analyses, the intrinsic value of Vestas is then derived under three different scenarios. The estimates are then examined via a sensitivity analysis, which is supplemented by a simulation and a peer group analysis. The thesis finds an intrinsic value estimate of 53.91 EUR but also note significant, theoretical issues within the employed financial theory. This thesis therefore stresses that estimates, share price targets, buy/sell recommendations, and so forth should be treated extremely critically and one must not be seduced by precise formulas with imprecise assumptions (Graham 1974). Bachelor Thesis © Emad Sharghbin 2009 ii Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. The concept of future prospects and particularly of continued growth in the future invites the application of formulas out of higher mathematics to establish the present value of the favored issue. But the combination of precise formulas with highly imprecise assumptions can be used to establish, or rather justify, practically any value one wishes. Benjamin Graham 1973, “The Intelligent Investor”, 4th ed., Harper and Row, New York, pp. 315-316. Bachelor Thesis © Emad Sharghbin 2009 iii Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. Brief Table of Contents EXECUTIVE SUMMARY I PART I – THESIS INTRODUCTION 1 PART II - CRITICAL REVIEW OF VALUATION MODELS 5 PART III – STRATEGIC ANALYSIS OF VESTAS 21 3. COMPANY PROFILE 21 4. EXTERNAL STRATEGIC ANALYSIS 22 5. INTERNAL STRATEGIC ANALYSIS 36 PART IV – FINANCIAL STATEMENT ANALYSIS OF VESTAS 42 6. 42 HISTORICAL FINANCIAL ANALYSIS PART V – VALUATION OF VESTAS 48 7. ESTIMATING THE COST OF CAPITAL 48 8. VALUATION 54 9. CONCLUSION 69 10. BIBLIOGRAPHY 71 11. APPENDICES 77 Bachelor Thesis © Emad Sharghbin 2009 iv Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. COMPREHENSIVE TABLE OF CONTENTS EXECUTIVE SUMMARY I PART I – THESIS INTRODUCTION 1 1.1. INTRODUCTION 1 1.2. THESIS STATEMENT 1 1.3. SOURCES AND LITERATURE 2 1.4. DELIMITATION 3 PART II - CRITICAL REVIEW OF VALUATION MODELS 5 2.1. INTRODUCING THE VALUATION CONUNDRUM 5 2.2. THE TECHNICAL VERSUS FUNDAMENTAL CONUNDRUM 5 2.3. FOUR CATEGORIES OF MODELS WITHIN FUNDAMENTAL VALUATION 7 2.4. THEORETICAL FOUNDATION 7 2.4.1. Asset-based Valuation Models 7 2.4.2. Multiple Valuation Models 8 2.4.3. Accounting-based Valuation Models 10 2.4.3.1. Dividend Discount Models 11 2.4.3.2. Cash-Flow Valuation Models 12 2.4.3.3. Economic Profit Valuation Models 12 2.4.3.4. Residual Income Valuation Models 13 2.4.3.5. Abnormal Earnings Growth Valuation Models 14 2.4.4. Real Options Valuation 15 2.5. CONCLUDING DISCUSSION OF THE FOUR GROUPS OF VALUATION MODELS 16 2.6. THE AUTHOR’S SYNTHESISING CONCLUSION 18 2.7. SUMMARISING THE CRITICAL REVIEW OF VALUATION MODELS 19 2.8. APPLICABILITY OF VALUATION MODELS TO A VALUATION OF VESTAS 19 PART III – STRATEGIC ANALYSIS OF VESTAS 21 3. COMPANY PROFILE 21 4. EXTERNAL STRATEGIC ANALYSIS 22 4.1. ENVIRONMENTAL ANALYSIS – USING THE PESTED FRAMEWORK Bachelor Thesis © Emad Sharghbin 2009 22 v Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. INDUSTRY ANALYSIS - USING PORTER’S FIVE FORCES FRAMEWORK 4.2. 4.2.1. Foundation - Static Analysis 26 26 4.2.1.1. Threat of substitutes 27 4.2.1.2. Rivalry among competing firms 29 4.2.1.3. Threat of new entrants 30 4.2.1.4. Bargaining power of customers 30 4.2.1.5. Bargaining power of suppliers 30 4.2.2. COMPETITOR AND DEMAND ANALYSIS – A WORLD MARKET UPDATE FROM VESTAS’ PERSPECTIVE. 4.3. 5. Putting Porter into Perspective – Dynamic Analysis INTERNAL STRATEGIC ANALYSIS VALUE CHAIN ANALYSIS – USING BARNEY’S VRIO FRAMEWORK 5.1. 31 32 36 36 5.1.1. Primary Activities 37 5.1.2. Supportive Activities 39 5.2. SUMMARISING SWOT TABLE 40 PART IV – FINANCIAL STATEMENT ANALYSIS OF VESTAS 42 6. 42 HISTORICAL FINANCIAL ANALYSIS 6.1. INTRODUCTION 42 6.2. GROWTH 43 6.3. ROIC 44 6.4. CREDIT RISK 46 PART V – VALUATION OF VESTAS 48 7. 48 ESTIMATING THE COST OF CAPITAL 7.1. INTRODUCTION 48 7.2. ESTIMATING THE COST OF EQUITY 48 7.3. ESTIMATING BETA 50 7.4. WACC 52 8. 8.1. VALUATION INTRODUCTION Bachelor Thesis © Emad Sharghbin 2009 54 54 vi Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. 8.2. FORECAST-DRIVERS AND CONSTANT VARIABLES 54 8.3. BASE, BEARISH, AND BULLISH SCENARIO-ANALYSES 56 8.3.1. Base case scenario 56 8.3.2. Bearish scenario 58 8.3.3. Bullish scenario 60 8.3.4. Conclusion on scenario analyses 61 8.4. SENSITIVITY ANALYSIS 61 8.5. SIMULATION OF FORECAST-DRIVERS 64 8.6. PEER GROUP ANALYSIS 65 8.7. COMPARING PRICE TARGETS AND ESTIMATES. 67 9. CONCLUSION 10. BIBLIOGRAPHY 69 71 10.1. ACADEMIC BOOKS AND ARTICLES 71 10.2. REPORTS AND OTHER SOURCES 75 NEWS PAPER ARTICLES 76 11. 77 APPENDICES 11.1. MARKET PREMIUM CALCULATIONS 77 11.2. MARKET WEIGHT REPORT CSE 15/04-09 78 11.3. LUNDHOLM’S CREDIT RISK FRAMEWORK 79 11.4. GUIDE TO THE VALUATION MODEL 80 11.5. BASE SCENARIO – VALUATION SUMMARY 81 11.6. BEARISH SCENARIO – VALUATION SUMMARY 82 11.7. BULLISH SCENARIO – VALUATION SUMMARY 83 11.8. CONSTANTLY HELD VARIABLES IN THE FORECAST 84 11.9. PRICE TARGETS FROM LEADING INVESTMENT BANKS 86 11.10. THREE PEER GROUP ANALYSES: WIND, SOLAR, INDUSTRIALS Bachelor Thesis © Emad Sharghbin 2009 87 vii Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. LIST OF FIGURES Figure 1: Displaying the two main groups of valuation approaches. ............................................... 5 Figure 2: Displaying the four main categories of valuation models. ................................................ 7 Figure 3: PESTED-analysis - Overview....................................................................................................... 23 Figure 4: Impact of PTC expiration on annual installation of wind capacity on US market. . 24 Figure 5: Self-reinforcing relation: Government support, Demand, and Cost reductions. .... 24 Figure 6: Industry Analysis using the Five Forces Framework – Overview ................................ 27 Figure 7 In-depth presentation of Threat of Substitutes analysis................................................... 28 Figure 8: Relative energy prices for different energy sources. ......................................................... 29 Figure 9: Five Forces Framework in a time-oriented perspective .................................................. 31 Figure 10: BTM Consult’s figure showing the top ten suppliers’ market share and presence in 2008. ................................................................................................................................................................... 33 Figure 11: The leading suppliers in the top 10 markets in 2008. .................................................... 34 Figure 12: Value chain activities as well as their competitive and financial implications ... 37 Figure 13 Summarising SWOT-analysis .................................................................................................... 41 Figure 14: Historical growth rates – Geographical segments ........................................................... 43 Figure 15: Decomposing ROIC-tree for Vestas based on 2005-2008 annual reports.............. 44 Figure 16 Expense groups in Vestas in percent of total revenue 2005-2008 ............................. 45 Figure 17: Vestas’ Credit Risk in the Lundholm Framework ............................................................ 47 Figure 18: Estimated market premiums in recent research - S&P 500......................................... 49 Figure 19 Rolling 2-year beta regressions against the MSCI World index. .................................. 51 Figure 20 Historical and future decomposition of Vestas’ WACC. .................................................. 53 Figure 21 Overview of forecast drivers in the different valuation periods ................................. 55 Bachelor Thesis © Emad Sharghbin 2009 viii Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. Figure 22 Development in growth, ROIC, EBITA and gross margin in the base case scenario ................................................................................................................................................................................... 57 Figure 23: Development of EP and FCF in the detailed forecast period. ...................................... 58 Figure 24 Development in growth, ROIC, EBITA and gross margin in the bearish scenario 59 Figure 25 Development in growth, ROIC, EBITA and gross margin in the bullish scenario . 60 Figure 26: The three scenario share price estimates and the weighted share price estimate. ................................................................................................................................................................................... 61 Figure 27: Sensitivity Analysis – Effect on share price, table and graph ...................................... 63 Figure 28: Histogram of simulated share price estimates.................................................................. 64 Figure 29 Peer Group analysis of Vestas against Wind Energy Peers ........................................... 66 Figure 30 Overview of the price targets in EUR from analysts and this thesis. ......................... 67 Bachelor Thesis © Emad Sharghbin 2009 ix Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. Part I – Thesis Introduction 1.1. Introduction Central within renewable energy is wind power. Despite Denmark’s small size and lack of heavy industrial tradition, the global market leader within the field is Vestas Wind Systems A/S (Vestas) Danish (BTM 2008: 25). The company is the third largest Danish company measured on their 10.2 % market weight (Bloomberg: 14/04-09). The main motivation behind this thesis the challenges of contemporary finance in appraising companies according to their underlying, intrinsic value (Penman 2007: vi) Vestas’ position is within a growth industry entailed by huge potentials but also major risks (BTM 2008: vi). Under the current momentum that wind energy enjoys, it is relevant to examine if the intrinsic value of the company deviates from the current market price. This paper is also motivated by the abundance of competing, contemporary valuation methods and their different, theoretical foundations. Although residual income, abnormal earnings, and real options models are prominent within academia, practitioners often revert back to multiples and discounted cash flow models (Barker 1999). Therefore, contemporary valuation models will be evaluated and discussed in order to provide a thorough overview of the valuation models. 1.2. Thesis statement The purpose of this paper is to find the intrinsic value of Vestas, which is of particular interest being a Danish world market leader within a major growth industry. This paper will answer the thesis statement: What is the estimated, intrinsic value of a Vestas’ share in a stand-alone case? In order to answer the thesis statement, the following issues will be addressed. Contemporary valuation models and their theoretical backgrounds will be examined. Their differences, advantages and disadvantages, as well as their usage by practitioners will also be critically reviewed. The review will result in a selection of appropriate models for the valuation of Vestas. Bachelor Thesis © Emad Sharghbin 2009 1 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. In an external strategic analysis, the external factors that can affect the profitability of wind energy will be discussed. The competitive situation of Vestas will also be addressed. An internal strategic analysis shall discuss and clarify Vestas’ competitive advantages. A historical financial analysis will reveal both past and current drivers of Vestas’ profitability. Prior to the valuation, the problematic, theoretical issues concerning cost of capital will be addressed in order to estimate Vestas’ cost of capital. Synthesising the strategic and financial analyses will then enable a forecast of the pro-forma budgets. Using the cost of capital, the intrinsic value of Vestas will be derived under base, bearish, and bullish scenarios. The value estimates’ sensitivity to changes in key variables will be examined via a sensitivity analysis, which will be supplemented by a simulation and a peer group analysis. 1.3. Sources and literature The paper is based on both qualitative and quantitative data. The strategic analyses are primarily based on qualitative, secondary data from market reports, journals, news papers, but also Vestas’ own publications and has therefore been treated very critically. The same applies to the financial analysis, which is mainly based on quantitative, secondary data from Vestas’ annual reports. Even though Vestas’ reports have been verified by publicly chartered accountants, all the material from Vestas is expected to be of a subjective character and displaying Vestas in the best possible light within the limitations of the law (Clatworthy 2005: 63-77). One of the most significant providers of market information within wind energy is Danish BTM Consult. BTM Consult, however, was founded in 1986 by two former Vestas employees and maintains a strong cooperation with Vestas (Jensen 2003: 237). For instance, Vestas also recognises BTM as their market information provider (AR 2008:17). Therefore, special attention has been paid to any bias in the provided information from BTM. A profound literature research was conducted in order to form the theoretical basis. Most of the relevant literature has been incorporated in the making of this thesis, although additional texts could always have been included. Bachelor Thesis © Emad Sharghbin 2009 2 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. 1.4. Delimitation It is necessary to make certain delimitations in order to keep the thesis focused on finding the intrinsic value of Vestas on a stand-alone basis. The thesis could have been based on other scenarios. For instance, Vestas could be acquired, for instance, by its’ competitor, American GE Wind. This would require profound knowledge of the acquirer, possible synergy effects, and a new required rate of return (Koller 2007: 447-52). Doing so would ultimately change the focus of this thesis. The thesis has been performed from an external, long term investor’s perspective without any access to information from internal compartments of Vestas. The different valuation models will be reviewed and discussed in order to have an overview of the different theoretical models, as opposed to using any arbitrary model. However, to maintain focus, there will be no empirical studies of the models’ accuracy or practitioners’ usage of these models. The discussion will be based on previous studies. The financial markets are also assumed to be efficient according to Fama’s Efficient Market Hypothesis, as there will be no in-depth discussion of this topic (Fama 1970). The thesis’ focus within the finance area also means that the accounting standards and techniques of Vestas and shall remain largely uncommented. The terms “intrinsic value of Vestas” and “intrinsic value of Vestas’ share” can obviously be used interchangeably, as it is just a question of dividing by number of shares. Still, focus will be on the latter. The strategic analyses could easily have been extended in size, but the limited space has been spent in other areas. Therefore, the strategic analyses will be concise and summarised in figures. Nevertheless, all major findings have been included and the analyses still provide a comprehensive picture. This thesis has been written using British English. Also, inclusive first person pronouns, “we”, will also be employed to include the reader in the thesis for the purpose of linguistic variation even though there is a single author. Bachelor Thesis © Emad Sharghbin 2009 3 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. The annual report for 2008 was published February 11, 2009. The next quarterly report is publicised April 28, just 3 days before the hand-in of this thesis. Therefore, April 14 has been chosen as the cut-off date. Bachelor Thesis © Emad Sharghbin 2009 4 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. Part II - Critical review of valuation models 2.1. Introducing the valuation conundrum People working within the field of financial analysis have for decades been searching for an efficient and accurate valuation technique, the magic wand that can estimate companies’ intrinsic values accurately (Gentry 2003: 3). Academics have sought after the wand in scientific endeavour, whereas professionals and investors have sought after competitive advantages and comparatively superior returns. Both groups have been unable to develop a universally applicable and accurate technique although various different models have been put forward (Gentry 2003: 3-4). Ideally, the balance sheet of a company would precisely depict the current value of the firm and thus be traded at a P/B ratio of 1.0. As Penman notes, no such perfect balance sheet exists, hence the need for valuation techniques in the first place (Penman 2007: 199). Being able to valuate a firm accurately can make the difference between financial success or failure, both for investors, acquiring firms, and so forth. The purpose of this part of the thesis is to compare and evaluate all major valuation techniques in order to choose the most appropriate for a valuation of Vestas. 2.2. The technical versus fundamental conundrum Generally, we can divide valuation into two supra-groups, technical and fundamental analysis techniques. Figure 1: Displaying the two main groups of valuation approaches. Based on Clatworthy 2005: 47. By far the most dominating analysis technique is fundamental analysis (Clatworthy 2005: 47). Two of the original main proponents behind the fundamentalist’s approach are Graham and Dodd (Gentry 2003: 3). The definition given by Penman is that “fundamental Bachelor Thesis © Emad Sharghbin 2009 5 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. analysis is the method of analyzing information, forecasting payoffs from that information, and arriving at a valuation based on those forecasts… instead of market prices” (Penman 2007: 84 & 73). This is supported by Bauman (Bauman 1996: 1). Within the fundamentalist approach, buying a share is not just buying an arbitrary piece of paper; it means that the investor is buying a piece of a business. Therefore, Penman states “If you are going to buy a business, know the business” (Penman 2007:14). Consequently, the starting point in the fundamentalist’s valuation process is a thorough strategic analysis of the firm. A comprehensive understanding of the firm’s economic factors is necessary before using valuation models (Penman 2007: 16-17). Contrary to the fundamentalist’s approach, technical analysts believe that share prices are determined entirely by supply and demand in the stock market (Clatworthy 1996: 50). Technical analysis involves studying the historical actions of the market instead of studying the firm, its goods or its surrounding market (Edwards 2001). Several scholars have confirmed technical analysis’ predictive powers (Canegrati 2008, Bettman 2009). Irwin also conducted a meta-study of 96 studies regarding technical analysis and concluded that 56 studies yielded positive results in favour of the technique. However, Irwin also stresses that there are serious, problematic issues with data snooping within these studies (Irwin 2007). Bettman suggests a triangulation of fundamental and technical analysis to synthesise the strengths of the two methods into one superior approach (Bettman 2009). Bettman’s suggestion is interesting, but further studies are needed within the area to validate the findings. In conclusion, for the matter of this thesis we regard technical analysis as a tool most beneficial to day traders and the like, with no or very little benefits to long-term investors. Fundamentalist investors are often very dismissive of technical analysis. As the American investment guru Warren Buffett expresses “If past history was all there was to the game, the richest people would be librarians”.1 The focus in this thesis will be entirely on fundamentalist techniques. Fundamental investors focus on long-term profitability and ignore short-term volatility in stock prices (Penman 2007: 19). 1 http://www.guardian.co.uk/world/2008/apr/20/usa.subprimecrisis Bachelor Thesis © Emad Sharghbin 2009 6 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. 2.3. Four categories of models within fundamental valuation Within the fundamental valuation approach, scholars put forward the following four categories of models. Figure 2: Displaying the four main categories of valuation models. Based on Plenborg 2000: 4 and Damodaran 2006: 9. With such a wide range of contemporary valuation models, it is essential to have an overview and critically scrutinise the models prior to any usage (Penman 2007: 18-20). Penman believes a good valuation model has three main ingredients: good thinking, good application and a good balance between costs and benefits (Penman 2007: 21). Plenborg proposes four criteria: the model’s precision, realistic assumptions, usability, and a fair value that is easy to understand (Plenborg 2000: 2-3). Their tenets will form the basis for the assessment of the models, which will follow in the next section, starting with assetbased valuation models. 2.4. Theoretical foundation 2.4.1. Asset-based Valuation Models Asset-based valuation models values all assets of the firm at an estimated fair value, which can easily be expressed algebraically. Total firm value V firm equals market value of assets MV ( A) . V firm MV ( A) Based on Penman 2007: 82. Although the model may seem simple, it can be time-consuming to assess all assets of especially larger firms. Penman also points out several theoretical weaknesses. Market Bachelor Thesis © Emad Sharghbin 2009 7 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. values may be unavailable for illiquid assets. Even if market values are available, they may not be correct measures for the intrinsic value of the specific assets in a particular going concern. Finally, the value of brand assets and firm synergies are notoriously difficult to estimate (Penman 2007: 82). The models are, however, applicable to smaller firms with a high value of inherent goodwill such as restaurants and so forth. The models also apply to valuation of asset-heavy companies, often within primary industries, as most of the assets in these firms are tangible goods with relatively precise market values. Asset-based valuation is also of relevance for firms that are soon expected to be liquidated (Penman 2007: 82-83). Still, the models have a limited scope for usage and are therefore only discussed briefly. 2.4.2. Multiple Valuation Models In the 1930’s, Graham and Dodd argued that analysing firms’ key ratios could reveal investment opportunities (Gentry 2003). Multiple-based valuation today forms a part of most text-books within valuation (Koller 2005, Damodaran 2006, Penman 2007). The reason for the multiples’ popularity is their apparent simplicity (Koller 2005: 362). Multiples are simply a market price variable, often stock price, divided by one of the firm’s value driver; earnings, book value or sales to name a few. Algebraically, this is expressed as follows. Vx is the price for the firm that is being valuated. V y is the value driver for the firm that is being compared, is the multiple on the value driver, and is the pricing error (Plenborg 2000: 18). Vx * V y Based on Liu 2002, and Plenborg 2000. Investors can quickly compare firms’ multiples based on the market values for comparable firms within the same industry, the so-called peer group. Multiples are mainly used to compare publicly traded firms across industries. The models can also be applied to privately held firms. In acquisitions or IPOs of private firms, investors can use multiples from existing publicly-traded firms to estimate the firm’s potential market value. As a complement to more complex valuation techniques, practitioners regularly present Bachelor Thesis © Emad Sharghbin 2009 8 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. valuations partially based on multiples, such as the price to earnings multiple (Koller 2005: 61). The necessary steps in the model appear simple. First, comparable firms, the peer group, are identified. Next, the relevant value multiples are chosen. Finally, the multiples are then aggregated into single numbers by choosing a weighing of the peer group figures and applied to the corresponding value driver of the firm being valued (Penman 2007: 76.) Multiple valuation models do not require detailed multi-year forecasts concerning complex numbers such as profitability, growth, and risk. The belief is that the market is efficient in pricing these. The models’ simplicity is, however, deceiving. The selection of “truly” relevant value drivers and identifying the “truly” comparable peer group is highly problematic. Deciding how to calculate the firm multiples and how to weigh the combined, peer group multiples also cause much problems (Schreiner 2007: 2-4). During the research for this thesis, the amount of literature dealing with multiple-based models is immensely less than literature dealing with accounting-based models. Paradoxically, the most frequently used models in real life receive the least attention from scholars (Schreiner 2007: V). Most importantly, by basing its foundation on market prices, multiple-based models assume that the markets are efficient in their pricing (Fama 1970). This is a highly debatable topic according to fundamentalist investors, and one could discuss if the method is within the fundamental approach in the first place, although “pure” fundamentalist valuation without referring to market prices at some point is difficult (Penman 2007: 7682). Researchers differ in their view on multiples. Penman dismisses multiples referring to the tenet that a valuation of the intrinsic value can under no circumstance be based on existing market prices (Penman 2007: 77). However, both Koller and especially Damodaran approve of the method as a valid and almost necessary sanity check to any accountingbased valuation and also as a source of insight into the value drivers of the firm (Koller 2005: 378 and Damodaran 2005). Koller and Schreiner agree on the superiority of forwarded multiples, that is, estimation of for instance forwarded earnings, instead of using Bachelor Thesis © Emad Sharghbin 2009 9 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. historical data. However, they also disagree, as Koller argues heavily in favour for using adjusted enterprise value multiples to avoid biases from capital structure and nonoperating profit and losses (Koller 2005:366) This makes sense logically, but Schreiner’s empirical data indicate that equity value multiples are superior (Schreiner 2007: 100). There is obviously no consensus within area. To conclude this discussion, we note that stand-alone multiple evaluation can rarely be used except in special cases such as in acquisitions or IPOs of private firms. Some scholars also approve it as a quick complement to other models. Koller is the strongest proponent, arguing that multiple valuation should receive as much attention, as the model it supports (Koller 2005:380) Penman and Plenborg, however, oppose multiple valuation and remain very critical of the model (Penman 2007: 77-78, Plenborg 2000: 24-26). 2.4.3. Accounting-based Valuation Models Prior to dealing with the different forms of accounting-based valuation, a few issues that apply to all models within this category will be addressed. All models assume some form of discount factor based on the cost of capital. To estimate these unobservable costs of capital or equity2, asset pricing models can translate firm risk into expected returns (Penman 2007: 471-76). The most commonly used model is Sharpe’s one-factor CAPM model (Sharpe 1964). Two common alternatives, Fama-French’s three-factor model and the arbitrage pricing theory models shall not be used, as Penman points out, even the one-factor CAPM is demanding…we really don’t know what the cost of capital for most firms is” (Penman 2007: 114-16). The focus in this thesis will therefore be on Sharpe’s CAPM model. The CAPM model claims that the expected rate of return on a security E(ri) is equal to the risk-free rate rf plus the security’s beta i times the expected return of the market E(rm) subtracted by the risk-free rate rf. Beta is a measure of the firm’s correlation with the market and thus the only firm specific measure in the model. E(ri) = rf + i [E(rm) – rf] 2 The choice naturally depends on whether the model presents firm or equity value, respectively. Bachelor Thesis © Emad Sharghbin 2009 10 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. Source: Koller 2005:294. Especially beta and the market risk premium are highly difficult to estimate (Koller 2005: 298). The risk-free rate is relatively easy to assess, as ten-year government bonds can be used as a proxy. For more on this topic, see section 7.2 Estimating the Cost of Equity, page 48. The brief introduction to cost of capital shows the huge uncertainty in any estimate of the cost of capital or equity. Unfortunately, the estimates play an integral part of all the models as the discount factor in the denominator, as we will see briefly. The issue constitutes one of the major weaknesses in all the accounting-based models (Penman 2007: 476). All the models expect for the Discount dividend model also contain problem in the estimation of their terminal value. The idea is that growth tends to mean-revert in the long period (Plenborg 2009: 7). In order to maintain focus on the differences of the models and for the sake of simplicity, infinite valuation periods are assumed for all the following accounting-based models. 2.4.3.1. Dividend Discount Models The accounting-based valuation models all build on Williams’ Dividend discount model from 1938 which was extended with the classic Gordon Growth Model (Gentry 2003: 3). The discounted dividend model (DDM) is as follows. Veq t 1 div (1 re ) t Based on Plenborg 2000: 6. The model provides the equity value of the firm by discounting all future dividends by the required return on equity of the investors. This is the basic idea in all the following accounting-based models. The model seems simple and intuitively correct, as shareholders focus on dividends, at least in the short term (Penman 2007:122). The weakness of the dividend model is precisely its focus on dividends; it ignores value creation in the firm (Plenborg 2000: 6). Its main applicability is to firms in stable industries with stable earnings and fixed payout ratio (Penman 2007:122). In practical terms, such conditions Bachelor Thesis © Emad Sharghbin 2009 11 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. apply to few firms, which partly explain why the model is uncommonly used by practitioners. 2.4.3.2. Cash-Flow Valuation Models Various types of cash flow models exist. Koller presents enterprise, capital and equity cash flow models, but the focus will here be on the enterprise discounted cash flow model (DCF) (Koller 2005:102). DCF provides the total value of a firm and its operations from an enterprise or entity perspective, whereas DDM focuses on the equity perspective. In the DCF, free cash flows to both equity and debt holders are discounted by a weighted average cost of capital (WACC). V firm t 1 FCFF (1 WACC ) t Based on Plenborg 2000: 7 The model is popular and appears logical as cash flows as physical entities that are easy to think about (Penman 2007: 127). Another strong point is the model’s incorporation of tax shield effects by using WACC (Koller 2005:111). Despite its popularity amongst practitioners, the model has serious theoretical flaws. It assumes that all liquidity after payments of dividends is invested in projects returning a rate of return equal to the cost of capital. Historically, some firms tend to “hog” excess liquidity for unprofitable projects instead of paying it out or buying back shares (Plenborg 2000:9). Penman also points out the highly problematic issue that necessary, sound investments are treated as a loss of value, at least in the short term. In real life, investments are necessary for firms’ survival, even if the returns may lie years ahead. Long forecast horizons are therefore needed which is another weakness of the model (Penman 2007: 127). Penman finds the model applicable primarily to firms with stable cash flows or constantly growing cash flows, which are rather unrealistic assumptions (Penman 2007:127). Koller disagrees with Penman and maintains his primary foundation in the DCF framework (Koller 2005: 101-02) 2.4.3.3. Economic Profit Valuation Models This category includes various valuation models that originate as far back as the 19th century (Penman 2007: 186). The models were revived in the 1990’s as several Bachelor Thesis © Emad Sharghbin 2009 12 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. consultancies and accounting firms started using the framework. Examples include McKinsey’s Economic Profit model, Stern Stewart’s Economic Value Added model, or KPMG’s Value Added model (Penman 2007: 89). Here, focus will be on McKinsey’s EP model. Via algebra, the DCF model can be reformulated to the following formula. V firm ICt * ROIC t WACC * ICt 1 (1 WACC ) t t 1 Based on Plenborg 2000: 9 The model shows that a firm is only able to provide an economic profit if its return on invested capital (ROIC) exceeds its costs of capital, WACC. Otherwise, firm value equals invested capital (ICt). Like DCF, EP is also made from the firm perspective. Whereas DCF treats even sound investments as declining cash flow, EP provides a better insight into the actual economic performance of the firm. Thus, both DCF and EP models provide relevant and valuable information, but in different ways (Koller 2007: 116). The advantages and disadvantages of EP models will be discussed in the following section concerning RI models, which are the equity counterpart of EP models. 2.4.3.4. Residual Income Valuation Models The recent popularity of the residual income model (RI) is attributed to the analytical work of Ohlson (Ohlson 1995). In the same way DCF is the enterprise equivalent of the equitybased DDM, EP is the enterprise equivalent of the equity-based RI model. (Plenborg 2000: 10). The formula reflects the difference. REq t re * Eqt 1 t 1 (1 re ) t Veq Eqt * Based on Plenborg 2000: 10 The model shows, that a firm is only able to provide residual income, if the rate of return on its equity (REq) exceeds its required return on equity (re). Otherwise equity value equals equity (Eqt). The more accounting-based EP and RI models assume clean surplus accounting (CSA) in the accounts, which are not required in the more flow-based CF models. CSA requires that Bachelor Thesis © Emad Sharghbin 2009 13 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. all income and expenses are placed on the income statement and not on the statement of equity (Plenborg 2000: 11). The most common examples of breaches of CSA include unrealised gains and losses on securities, derivatives, and foreign currency translations (Penman 2007: 269). If CSA is fulfilled RI and EP can be used across different GAAPs. If not, it can have dramatic consequences. As an example, Daimler-Benz’ 1993 annual reports showed a profit of 168 mDEM in Germany but a loss of 1 bnUSD on NYSE (Plenborg 2009: 5). If the assumption of CSA can be maintained, however, EP and RI models can like DCF models be applied across different types of GAAP and accounting methods (Plenborg 2000:11). EP and RI models possess several advantages according to Penman. Highlights include the focus on the value drivers; profitability and growth in investments. It forecasts a more graspable term, earnings, instead of cash flow which may fluctuate depending on the level of investments (Penman 2007: 175). Analysts and businesses also typically forecast earnings not cash flows. This also makes it easier to validate RI and EP forecasts with the following years’ income statement. It is harder to validate CF forecasts, as investments may alter the picture dramatically. In RI and EP framework more value is also recognised in the short term, which enables shorter forecast horizons. RI and EP models are more robust to value-destroying investments that may grow earnings but at rates less than the required rate of return (Penman 2007: 176). The RI and EP framework is also more robust against earnings creating by accounting manipulation. If earnings are increased via accounting manipulation, it will be offset exactly by the lower book value (Penman 2007: 178). The models, however, requires a profound understanding of accrual accounting and relies on accounting numbers, which necessitates an accounting analysis (Penman 2007:175, Plenborg 2000:15). 2.4.3.5. Abnormal Earnings Growth Valuation Models The Abnormal Earnings Growth model (AEG) is the latest newcomer among the accounting-based models and originates from the 1995 Ohlson framework (Ohlson 1995, Bachelor Thesis © Emad Sharghbin 2009 14 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. Penman 2007:206). The model provides the value of the equity by discounting accountingbased earnings with the cost of equity. Veq t 1 Earnt (1 re ) t Based on Plenborg 2000: 12. The model is intuitively simple, as it also focuses on earnings, which is what many investors focus on and it is also indifferent to different GAAPs (Penman 2007:213). Contrary to EP and RI models though, AEG models do not provide any insight into the drivers of growth, particularly the balance sheet. It also requires an understanding of cumdividend earnings which equals the sum of earnings of the firm and earnings of paid-out dividends (Penman 2007: 213, Plenborg 2000: 13). 2.4.4. Real Options Valuation Finally, real options have received much attention in recent years’ academic literature within valuation. A real option is the business’ option to make or abandon a capital investment in a business project. Real options are options based on a “real”, tangible asset and have many similarities within financial options. For instance, the exercise and share prices from financial options are equal to, respectively, present value of the fixed costs over the lifetime of the investment and present value of cash flows expected from the investment (Leslie 1998). Significant theorists within the field are Trigeorgis (1993, 2005) and Copeland et al. (2001). They put forward that real options can play an important role in the net present value capital budgeting processes that many firms use to choose among projects. The main claim is that traditional net present value budget analyses (NPV) are unable to incorporate the future value of options in a capital budgeting analysis (Trigeorgis 2005 & Block 2007: 255). On future projects, firms may ignore their option to cancel a project. A budget analysis may then inaccurately provide a negative NPV, which then causes the project to be discarded. However, by including the option value inherent in the project, the project may have a positive NPV. Bachelor Thesis © Emad Sharghbin 2009 15 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. Still, practitioners within equity valuation continue to use accounting-based models. The effects of real options are therefore often ignored in contemporary equity valuation (Atauallah 2009: 57). This can lead to seriously flawed investment decisions, as highly profitable investments can be overlooked. For valuation of projects on a business level, Trigeoris formulates the “new” NPV equation as follows. S ( NPV ) P( NPV ) ROV ( FV SV ) The strategic S(NPV) equals the findings of a traditional P(NPV) plus the option value ROV(FV+SV), which is the sum of the flexibility value and strategic value inherent in any project (Trigeoris 2005: 32). As an alternative, McKinsey & Company proposes finding the strategic NPV using a fourstage process. The idea is that the value of the real options should be added to the value of the firm provided by a DCF model (Trigoris 2005, Ataullah 2005: 58, and Ashton 2003: 424). Therefore, the DCF analysis is first applied. Then uncertainty in the project is modelled by identifying cash flows in different scenarios. Next, the inherent flexibility in the management’s different options is then modelled into a decision tree, which can also be estimated using real option valuation (Leslie 2000). As a technique, it is mostly relevant for oil companies in for natural-resource discovery or for IT-companies in technology-related investments and so forth. As several scholars point out, real options are also ideal for medical and biotech companies for example (Block 2007: 255). In these cases, the existing real options framework justifies the relatively high complexity and uncertainty inherent in real options models. For other types of firms, the model is rarely applicable. 2.5. Concluding discussion of the four groups of valuation models The first section will briefly discuss the usage of the models among practitioners. Some researchers find that multiples valuation is the most frequently used method by practitioners and accounting-based valuation only second. Barker, for instance, states that the P/E ratio the most important driver and FCF only secondary (Barker 1999). This is Bachelor Thesis © Emad Sharghbin 2009 16 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. backed up by Demirakos, who finds that the multiple P/E is present in 90% of analysts’ reports, whereas DCF only is employed in 40% of the reports (Demirakos 2004: 230). However, more recent research contradicts these claims. For example, Imam finds that DCF has grown in importance among practitioners and is now, although the study focuses on UK analysts (Imam 2008). Block finds that only 14 % of American firms use real options within in their budget analysis, indicating the low dispersion of the method among practitioners (Block 2007: 10). The empirical findings are somewhat in line with the discussions in the previous, theoretical sections. Here, the main conclusion was that accounting-based models stood strongest, while they could benefit from multiples providing a sanity check. Given that all the accounting-based models are based on the same underlying DDM, the choice of model should not matter, as all models should provide the same result. In real life, however, the RI and EP (hereafter just RI) models are less sensitive to changes in the budget variables. The DCF model first dissolves the accounting figures and then rebuilds them into cash flows. The DCF model therefore has to estimate both “normal” and residual income, which increases the sensitivity to changes in variables. Residual income is the sole focus of RI models. Several studies around the millennium proposed that RI outperforms DCF in term of predictability and estimate accuracy, and RI were therefore recommended. (Penman 1998, Olsson 2000, Plenborg 2000: 15). The claim was contested by Lundholm in 2001, as he argued that the whole valuation literature comparing the accuracy of DCF and RI was misguided. He stresses that the theory prescribes that both models provide identical valuations, if implemented correctly (Lundholm 2001). Penman acknowledges Lundholm’s findings, but claims that practice is the final arbiter of competing models (Penman 2001). In other words, the theorists within the field are very dispersed in their views. Generally, there is a somewhat established consensus that the merits of asset-based and real options valuation are only applicable to special cases (Koller 2005:131, Penman 2007:83, Plenborg 2009:16). Little focus is currently given to asset-based and cash flow based models in academia. Koller focuses on using both DCF and EP, stressing that DCF is less sensitive to accounting principles (Koller 2005: 101). Penman is in favour of RI models and criticises AEG models intensely (Plenborg 2000, Penman 2007). Ohlson, one of the Bachelor Thesis © Emad Sharghbin 2009 17 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. main proponents of AEG, still maintains its merits and is supported by other scholars (Jorgensen 2005, Ohlson 2005). 2.6.The author’s synthesising conclusion Summarising the different views provides the following conclusions. Asset-based and real options models can be useful but only under special circumstances. Multiples valuation can provide a quick and dirty sanity check, but also has severe methodological and practical issues, which are only exacerbated by the model’s heavy reliance on market’s pricing of the peers. The accounting-based models stand strongest, and although DCF is most commonly used by practitioners, it receives little attention within academia. Focus is definitely on the RI versus AEG model discussion. For general purposes, the RI and EP models are the most preferred model by the author. They are easy to explain to readers outside the finance area, as it avoids the rather complex term of cum-dividend earnings (Penman 2007: 213). Whereas the RI model is strong in telling the story of a firm’s economic performance, DCF reveals the development of cash flows. Therefore, for general purpose valuations, a combination of the RI or EP model with a DCF model is a theoretically strong. It includes their respective advantages, although it also adds complexity. Bachelor Thesis © Emad Sharghbin 2009 18 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. 2.7.Summarising the Critical review of Valuation Models The following figure summarises the previous, theoretical discussion. Enterprise-based EP and equity-based RI have been put together, as they share most of their advantages and disadvantages, although the models are not completely similar. Figure 3: Summarising the critical review of valuation models. Category Model name Asset-based Replacement Advantages Focus on assets Disadvantages Cannot capture brand value Cannot capture synergies etc. Pricing all assets takes time Useful for Asset-heavy companies Small businesses with goodwill Liquidation Focus on assets No "market" liquidation values Pricing all assets takes time Firms soon to be liquidated Multiples Various Relatively quick No proforma budgets Assumes market pricing Difficult to choose peers Difficult to choose variables Challenging to weigh variables Private firms going public Private firms as target firms Sanity check to other models Accounting -based DDM Simple, only dividends Reliance on cost of equity No focus on value creation Affected by pay-out ratio Stable firms with fixed pay-out ratios DCF CF is easy to think about Incorporates tax shield effects Indifferent to different GAAPs Reliance on WACC Terminal value calculation Assumes reinvestment at WACC Investments = value-destruction Firms with stable CF Firms with constantly growing CF EP & RI Focus on economic perf. Earnings -> easier to forecast Forecasts can be validated Shorter time horizons If CSA, indifferent to GAAPs Robust against earnings manipulation by accounting Robust against earnings manipulation by investments Reliance on WACC / cost of eq. Terminal value calculation Requires CSA Understanding of accrual acc. Requires accounting analysis All other firms AEG Shorter time horizons Investors "buy" earnings Indifferent to different GAAPs Reliance on cost of equity Terminal value calculation Complex cum-div. earnings Requires earnings analysis No insight in value drivers Unrecommendable Ignores earnings drivers and balance sheet RO Captures value of options Complex Requires replicable portfolio Rarely applicable Oil companies in resource discovery Medical and biotech companies Other special cases Real Options 2.8. Applicability of valuation models to a valuation of Vestas The choice of valuation models for Vestas will now be discussed based on the previous, theoretical discussions. Bachelor Thesis © Emad Sharghbin 2009 19 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. Vestas is neither an asset-heavy company nor a very likely candidate to go bankrupt, as will be demonstrated in section 6.4 Credit Risk page 46. Therefore, the asset-based models are irrelevant for the remainder of the thesis. Applying a real options model to Vestas would include replicating the portfolio of Vestas and its projects (Koller 2005: 131). Modelling the portfolio is possible, but would pose severe, time-consuming difficulties for a large, industrial firm like Vestas. Penman points out that a model must have a good balance between cost and benefit (Penman 2007: 21). Subsequently, real options models are not suitable for Vestas, and will not be included in the valuation. As mentioned in the discussion of the models, accounting-based models stand the strongest theoretically. The EP models appear to have the most advantages and their main strength lies in their indication of the firm’s economic performance, whereas a DCF model focuses on the cash flows. The two models shall be used in combination, partly for validation, but mainly because they provide different but valuable information. (Koller 2005: 116) The models’ results will of course be thoroughly scrutinised using scenario and sensitivity analyses, and a simulation of the value estimates. Still, multiples will also be applied as a useful sanity check. Especially forwarded multiples models are highlighted as providing a useful sanity checks to the value found via other models. In conclusion, the valuation will employ DCF, EP and Multiple valuation models. Bachelor Thesis © Emad Sharghbin 2009 20 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. Part III – Strategic Analysis of Vestas The purpose of the strategic analysis part is to identify and discuss both external and internal factors which will then enable an evaluation of Vestas’ future growth and profit potentials (Penman 2007: 14). The analysis will have a top-down approach. The external factors will be analysed first and followed by an analysis of Vestas’ internal factors. The strategic analyses will be summarised in a SWOT analysis, which will provide an overview of the analyses. 3. Company profile Vestas, the largest supplier of wind turbines in the world3, focuses its principal activities on the product development, manufacturing, turnkey delivery and maintenance of wind turbine installations. The company’s sale of wind turbines is by far its main activity and constituted 92 % of its 2008 revenue as sale of services constituted the remaining 8 % (AR 2008: 62). Vestas was originally a hydraulic crane manufacturer which entered the wind turbine market in the oil crisis of the late 1970s. The wind energy market expanded and Vestas entered the US market, in which they invested heavily. When tax advantages on wind turbine investments were removed in the US in 1986, Vestas was forced to sell off all nonwind -related assets and re-emerge as Vestas Wind Systems A/S. In the 1990s, Vestas bought Danish Wind Technology A/S. Germany, Denmark and Spain emerged as the major European markets. Vestas then went public on the Copenhagen Stock Exchange in 1998. In 2001, Vestas sold its 40 % stake in Gamesa Eólica, a Spanish joint-venture from 1994. In 2004, the Danish rival, NEG Micon, was then acquired. 3 Still the largest, one might add, as Vestas has been losing market share in recent years. The runner-up, American GE Wind, is only 1.2 pct. points from Vestas’ 19.8 % market share. (BTM 2008: 24) For more on this topic, refer to section 4.3 Competitor and Demand Analysis – A World Market Update from Vestas’ perspective., p. 32 Bachelor Thesis © Emad Sharghbin 2009 21 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. In 2008, Vestas supplied 5581 MW of wind turbine capacity, a 19.8 % market share (BTM 2008: 24). Vestas reports sales in more than 40 countries and maintains production facilities in Denmark, Germany, Italy, India, the UK, Spain, Norway, Sweden and China. Vestas’ core competence is within the larger >1.5 MW segment of wind turbines, which constituted 92 % of Vestas’ total supplied MW in 2008 (BTM 2008: 32). 80.4 % of global supply in MW in 2008 was within the segment. The company, via its ambitious “10 by 2010” plan, targets to increase its production capacity to 10,000MW by the end of 2010 (AR 2008: 16). 4. External Strategic Analysis 4.1. Environmental Analysis – Using the PESTED Framework Vestas’ profitability, growth potential and value are highly dependent on external factors, which will be analysed using a PESTED model adapted to Vestas based on Grant and Ireland (Grant 2008: 66 & Ireland 2007: 35). The model differs from the more traditional PESTEL, as the political and legislative segments have been combined into one segment. Oppositely, the demographic and socio-cultural segments are split into two separate segments. The summary will precede the actual discussion to provide an initial overview. All the segments, their underlying elements, and the conclusions are summarised in the figure below. In the figure, the sign of the underlying element as well as its importance have been indicated by the coloured cells to the right. Bachelor Thesis © Emad Sharghbin 2009 22 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. Figure 3: PESTED-analysis - Overview Generel Segment Political/Leg. Underlying element Kyoto-protocol Change in EU/US politcal attitude towards the environment US goal to reduce dependence on foreign oil REFIT in EU and PTC in the US Dependence on politicians' subsidies. Without, WE is expensive Effect Magnitude +2 +3 +1 +3 -3 Economic Current financial crisis --> Fewer developers can find finance Dependence on the state of the market WE Short time-span --> Short-term interest --> Higher risks Dependence on prices of raw materials - notably steel Global business --> Currency exposures -3 -2 -1 -1 -1 Socio-cultural Changed attitude towards the enviroment Increased focus on ethical issues Neighbors to wind mills and animal activists +2 +1 -1 Technological Development of IT Potential in +5 MW offshore wind energy Wind research is expected to lower $/kw cost Development of alternative, cost-efficient renewable energy Economies of scale expected to lower $/kw cost +1 +1 +2 -1 +1 Demographic Population growth in countries with high energy consumption (China/US)increasing need for energy globally Constantly +1 +2 Environmental Climate debate dominated by enviromentalists (No "Lomborg") Recent studies indicate melting poles Environmental legislation enforced - Renewable quotas +2 +1 +2 Adapted by the author to fit Vestas. (Model synthesises Grant 2008: 66 & Ireland 2007: 35) As the analysis indicates, wind energy enjoys a massive political support at the moment and especially the European REFIT and the American PTC, economic incentive schemes have boosted the industry. However, the political support is also the industry’s Achilles’ heel. If the political winds suddenly change, Vestas and its competitors may have to reconfigure their businesses like in 1986, when the US tax rebates were suddenly changed, as mentioned in section 3. Company profile, page 21. To illustrate just how sensitive wind energy still is to changes in the political support, the following graph depicts the development on the American market in recent years where the PTC scheme expired three times. Bachelor Thesis © Emad Sharghbin 2009 23 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. Figure 4: Impact of PTC expiration on annual installation of wind capacity on US market. Source: American Wind Energy Association. The existing political support has a positive, self-reinforcing effect, which is crucial to the wind industry for the time being. The major threat is obviously if political support disappears before wind energy is competitive on an unsubsidised basis. Figure 5: Self-reinforcing relation: Government support, Demand, and Cost reductions. Source: Hoogwijk 2008, REN 21. In the economic segment, there are many negative elements and the current crisis is obviously the most urgent issue. Developers have a hard time finding the finance for their projects, as Vestas’ CEO Ditlev Engel also acknowledges (AR 2008:05). The consequences of the crisis are exacerbated by the 1-2 years of typical lead time depending on the project’s size. That is, the full consequences of the crisis will not be revealed before the end of 2009, possibly 2010.4 4 Refer to http://www.vestas.com/en/investor/announcements/company-announcements.aspx Bachelor Thesis © Emad Sharghbin 2009 24 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. The change in the socio-cultural attitude towards the environment can be regarded as the reason for the political change. Still, the socio-cultural change is given a lower importance in the summary. The reason is the obvious fact that no policies are implemented before a political majority in favour for “wind-energy” friendly policies. The socio-cultural change is still important and is assigned a magnitude of +2. The technological segment also brings positive prospects for Vestas. The future potential of off-shore wind farms can solve both the lack of available land and also improve efficiency, as wind speeds are typically higher offshore than onshore. The wind is also more stable offshore, as hills and other land obstacles can create turbulent winds for onshore wind turbines. Still, the current off-shore price €/kw is much higher due to the increased installation and maintenance costs. Ditlev Engel also acknowledges that the off-shore will not play a major part in the short-term (AR 2008:04). Developments in the technological segment could also paralyse the whole industry, if for instance the solar industry was to achieve an enormous, technological breakthrough which would lower the solar power’s €/kw price dramatically. The element is only given a magnitude of -1, as the probability of such a rapid development in the short and medium term is relatively low. These issues will be discussed much further in section 4.2.1.1 Threat of substitutes, page 27. The demographic segment is also interesting. The energy consumption and population growth in the two key markets US and China will impact the wind energy industry in years to come. The countries’ growing need for energy has positive prospects for Vestas. (BTM 2007: 71) Finally, the environmental segment includes many positive elements for Vestas. Recent research is more often in favour of an environmentalist view against global warming, which is believed to be caused by fossil fuels among other things. The issue still remains highly debatable. In Denmark, the debate was temporarily bi-polar due to the “sceptical environmentalist” Bjørn Lomborg. His school of thought occupies very little space in the for practical examples on lead times within wind energy. Bachelor Thesis © Emad Sharghbin 2009 25 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. public debate now. 5 Vestas benefits from a public debate dominated by environmentalists, as the movement obviously is in favour of renewable energy. In conclusion, the external analysis of Vestas’ environment shows a somewhat optimistic prospect for Vestas. There are, however, also some potentially very damaging elements. In the short term, the paradox is that while the political climate has not been much better, the financial climate could not be much worse, as customers struggle to find raise finance (AR 2008: 04-05) In the longer run, the biggest threat is if political support disappears before wind energy is competitive on an unsubsidised basis. A major, technological breakthrough within other energy sources, especially solar energy, could pose a serious threat. 4.2.Industry Analysis - Using Porter’s Five Forces Framework 4.2.1. Foundation - Static Analysis For the later valuation, it is essential to evaluate the industry’s attractiveness, intensity of competition and level of future profitability. The chosen theoretical framework is Porter’s Five Forces Framework (Porter 1985). The following figure provides a concise overview of my analysis of the wind turbine industry using the Five Forces Framework. The first column displays the type of competitive force dealt with. The second column provides the overall conclusion on the level of the threat. The third and fourth columns deal with the underlying elements in depth. The final column indicates the effect for Vestas in which a short, green bar is positive and a long, red bar indicates a negative outlook. The five forces will now be dealt with separately. 5 http://www.lomborg-errors.dk/ Biologist Kåre Fog scientifically disproves Lomborg’s claims. Bachelor Thesis © Emad Sharghbin 2009 26 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. Figure 6: Industry Analysis using the Five Forces Framework – Overview Competitive Force Substitutes Overall Medium Underlying element Nuclear power Low price on fossil fuels Other renewable energy sources Customers' propensity to substitute Wind energy price relative to "broad" substitutes Level of threat Low-Medium Medium-high Medium Low-Medium High Rivals Medium Relatively few global competitors High growth industry Excess capacity and high exit-barriers Different target segments - Low/High MW Dominant national competitor in major markets Ability to copy patents and products Low-Medium Low High Low-Medium Medium-high Low Entrants Low-Medium Economies of Scale High capital intensity High R&D costs High product differentiation Low-Medium Medium Low-Medium Low-Medium Suppliers Medium-High Few suppliers Suppliers' ability to downstream integrate Suppliers' business outside wind energy Medium-high Low Medium-high Buyers Medium Medium-high Low-Medium Low-Medium Fewer, bigger suppliers Operational reliability is imperative Buyers' ability to upstream integrate Effect for Vestas Adapted by the author to fit Vestas. (Theory is based on Grant 2008: 72 and Porter:1980) 4.2.1.1. Threat of substitutes Threat of substitutes in the broad sense includes all sources of electricity including the cheaper, more traditional substitutes, such as nuclear power and fossil fuels. In the narrow sense, however, only sources of renewable electricity are true substitutes. The following figure provides an overview. Bachelor Thesis © Emad Sharghbin 2009 27 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. Figure 7 In-depth presentation of Threat of Substitutes analysis Energy Type Fossil Pros + Cheap - if no environmental taxes + Cheap to install - Cons Carbondioxide Other pollutants Dependence on foreign states - Middle East Potentially facing heavy taxes Nuclear + No carbondioxide + Long time span of 40 years + Cheap running costs - Massive installation costs Dependence on uranium import and prices Waste management problems Security problems - Terrorist threat Expensive to shut down plant Wind + + + + No carbondioxide Quick Installation "Unlimited supply" No dependence of supplying countries - Requires available, windy area Volatility requires advanced electricity net Dependent on political/financial support Solar (thermal) + + + + + + No carbondioxide Decreasing costs Less volatility than wind production "Unlimited supply" No dependence of supplying countries Largest renewable technical potential - Requires lot of space Less developed than wind Dependent on political/financial support Requires available, sunny area Seasonal fluctuations in production Hydro + No carbondioxide + Relatively cheap within renewables - New locations are very limited Large installation costs Often disrupts wild life and agriculture Requires massive political support Biomass + No carbondioxide + Cheap setup costs - Large production requires huge area Drives food prices up Ethical issue - Starvation in poor countries Fossils emit CO2 and other pollutants, whereas nuclear power has a waste management problem in addition to the ever-present post 9/11 “terrorist threat”. That is, fear of terrorists either bombing nuclear plants or using “dirty” nuclear bombs (Congress Report 2005) Biomass’ impact on global food prices poses an ethical issue in the Western world, especially its consequences for the world’s poorest. Solar power has a huge technical potential, but is currently very expensive (REN 21:40). As the following figure shows, other types of renewable energy are currently less price-competitive than wind power.6 Hydro 6 The figure is based on Australian data from 2006, but it is the author’s judgment that it is somewhat representative of the global situation today. Bachelor Thesis © Emad Sharghbin 2009 28 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. power is cheap, but poses no severe threat to wind energy due to the very limited locations available for new sites (USGS 2009).7 Figure 8: Relative energy prices for different energy sources. Levelised costs for energy sources Solar Photovoltaics Geothermal Biomass Solar thermal Wind power (Optimal) Small Hydro power Nuclear Gas: combined cycle Coal 120 89 88 85 55 55 40 37 28 0 20 40 60 80 100 120 140 Cost (AUD/MWh) Source: Graham 2006. On one hand, wind energy is the cheapest and most feasible renewable energy source at the present. On the other hand, the wind industry is very dependent on subsidies and feed-in tariffs, as mentioned in the PESTED analysis. Should the politicians on a global scale suddenly change their mind and erase all supportive schemes overnight, the industry could face serious difficulties. In terms of unsubsidised prices, traditional energy sources such as fossils and nuclear power are still much cheaper. Therefore, the threat is overall of a medium level. 4.2.1.2. Rivalry among competing firms The threat from rivals is deemed to be of medium threat and consists of opposing forces. Few truly global competitors exist, especially in the growing >1,5 MW segment of wind turbines (BTM 2007: 27 & 32). On the other hand, the exit-barriers are high, due to the huge investments necessary for wind turbine production. Participants could be forced to continue despite neutral or negative economic performance, potentially making competition fiercer. Especially Siemens Wind Power and GE Wind could stay in business for an elongated time since the major industrial conglomerates behind them could choose 7 The US government-run U.S. Geological Survey points out “…most of the good spots to locate hydro plants have already been taken.” http://ga.water.usgs.gov/edu/wuhy.html Bachelor Thesis © Emad Sharghbin 2009 29 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. to supply them with capital injections. The competition is also anticipated to be fierce in the near future with the entrance of several Chinese players. 4.2.1.3. Threat of new entrants Next, the threat from new entrants is overall estimated to be of a low to medium level. There are large economies of scale in the production of wind turbines. Being a successful wind turbine manufacturer requires investments in factories, and R&D departments, often in several countries (BTM 08:V). Finally, there are several different types of wind turbines depending on the speed and type of wind, creating a diversified product portfolio which is harder for new entrants to copy. These observations, however, mainly apply to the case of “truly” new entrants, that is, companies inexperienced within the energy sector. Traditional, large oil companies have a completely different starting point, as their cash flows from fossil fuels could provide the necessary financial funds. Even if a renewable energy subsidiary should provide negative cash flow initially, it may be regarded as a necessary long term investment for many of the large oil companies. However, some traditional oil companies claim they will no longer invest in renewable energy (Guardian 17/04-09). It is difficult to validate such press releases, as they could be made for strategic reasons. Still, this circumstance raises the level of threat to the overall low to medium level. 4.2.1.4. Bargaining power of customers In recent years, buyers of wind turbines tend to be less and less private and professional developers and more often large utilities (BTM 2008: 33). These experienced, professional customers buy much larger quantities and possess much more bargaining power. Still, at least so far, the market has been a seller’s market, and therefore, the force is still within the low to medium range. 4.2.1.5. Bargaining power of suppliers The bargaining power of suppliers is medium-high, There are relatively few suppliers in the industry, for instance one out of three wind turbines currently in operation is fitted with wings from Danish LM Glasfiber, making the supplier a key player in the industry (BTM 2008:114). Hansen Transmissions, now acquired by Suzlon, accounts for 30 % of the Bachelor Thesis © Emad Sharghbin 2009 30 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. world supply of gearboxes, which further adds to the monopolistic situation within some of the components (BTM 2008:115) Also, these suppliers often have major business areas outside wind turbines granting them more bargaining power. However, the suppliers are often unable to integrate downstream, that is, acquire wind turbines manufacturers or manufacture turbines themselves, which does lower their bargaining power slightly. 4.2.2. Putting Porter into Perspective – Dynamic Analysis In order to make Porter’s framework more dynamic, a time-orientation analysis will supplement the previous analysis. The analysis will deal with both short term profitability and long term profitability. Figure 9: Five Forces Framework in a time-oriented perspective Competitive Force Rivals Substitutes Entrants Suppliers Buyers Overall profitability Short-term 2009-2013 Medium Medium Low-Medium Medium-High Low-Medium Medium Long-Term 2014-2023 Change Medium-high Negative Low-Medium Positive Low-Medium Neutral Low-Medium Positive Medium Negative Medium-High Slightly positive As the figure shows, Vestas is already within an attractive industry. Wind energy will maintain its potential if the technological developments within the different renewable energy sources are assumed to be evenly distributed. As wind becomes more price competitive, the threat of substitutes should decrease in particular from the currently cheap fossil energy sources. The bargaining power of suppliers should also decrease. More upstream integration is anticipated. This could be either via acquisitions or for certain components via organic growth. Also, lack of key components should be a thing of the past, as suppliers have expanded their capacities.8 8 A reference to the crisis of 2006, in which shortage of crucial components hit the industry (Business.dk 20/03/07) Bachelor Thesis © Emad Sharghbin 2009 31 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. On the other hand, the bargaining power of buyers is expected to increase, as professional customers from large utilities buy large quantities, which naturally gives them more bargaining power. The intensity of rivalry will also increase, as Vestas has previously enjoyed some sort of first mover advantage, which will decrease due to the entrance of professional conglomerates such as GE Wind and Siemens. Still, Vestas is within a fairly profitable industry with a positive outlook. Vestas’ CEO Ditlev Engels’ is, as expected, far more optimistic, as he states “Honestly it is difficult to identify an industry with a better outlook. Modern energy is the future.” (Vestas SH 2009/1: 2). 4.3. Competitor and Demand Analysis – A World Market Update from Vestas’ perspective. As a supplement to the previous analyses, a competitor and market analysis will follow. The following figure from 2008 shows the top ten suppliers’ market share on the vertical axis and their market presence in number of countries above 50 MW on the horizontal axis. Bachelor Thesis © Emad Sharghbin 2009 32 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. Figure 10: BTM Consult’s figure showing the top ten suppliers’ market share and presence in 2008. Source: BTM 2008: 28 As mentioned in 1.3 Sources and literature, page 2, information from BTM needs to be paid special attention. Besides the different ways of measuring market share, one could also question BTM’s >50MW criteria. Other criteria could display different pictures of the competitive situation. Regardless, the figure gives us a good idea of Vestas’ unique position with its presence in several, major markets. The uniqueness is also illustrated in the following figure, which shows the leading suppliers in the top 10 key markets in 2008. Bachelor Thesis © Emad Sharghbin 2009 33 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. Figure 11: The leading suppliers in the top 10 markets in 2008. Source: Based on BTM 2008:29 The figure shows that in all of the five biggest markets major, national players dominate, which works against Danish-based Vestas. However, Vestas is among the top three suppliers in all of the 10 major markets expect Portugal and China. None of its competitors can boast such a global presence. Globally, there only three large-scale manufacturers of off-shore wind parks, Vestas being one of them. The figure also eminently shows the relative magnitude in MW of the two major markets, US and China. Based on installed capacity9, the two markets represented an astounding 52 % of global demand (BTM 2008:19). China alone experienced a growth rate of 92 % in 2008 (BTM 2008:13). As Vestas “only” grew by 30 %, Vestas lost market share in this market as well. Vestas has relatively impressive growth rates, but the market is simply growing so fast, that they cannot maintain their territory. The Sino-American markets will in all likeliness continue their dominance. The valuation will partly be based on 3 geographical drivers, Europe, Americas, and Asia. Obviously, US and China will constitute major, underlying components of their respective growth drivers. Fortunately for Vestas, the firm is already constructing production plants in both markets, 9 BTM provides both numbers for supplied and installed capacity. The supplied capacity was 111 % of recorded installation. They differ as some turbines are in transit or under construction. For more on these methodological issues, refer to BTM 2008: 28. Bachelor Thesis © Emad Sharghbin 2009 34 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. as well as major R&D centres in Houston, Boston and Colorado (AR 2008: 20). In other words, the company has increased its presence and capacity in these strategic markets in anticipation of the growth (AR 2008: 10). Vestas risks excess capacity and the consequent costs in case of a major drop in demand, which may be the case currently. In anticipation of growth in demand, Vestas strives to maintain all employees, at least in the present (AR 2008: 26). It is yet too difficult how much the firm will be impacted by the global recession, but the quarterly report for Q1 and CEO Engel’s guidance for 2009 will be strong indications of Vestas’ situation. To some extent, Vestas’ falling market share is also a result of the prioritising of Vestas’ executive management. Vestas has since 2005 prioritised EBIT-margin, and net working capital as the two most important financial goals (AR 2008:16). The third most important goal, market share, was replaced by revenue in 2008, as the original goal of a 35 % market share in 2008 was conveniently dropped (AR 2008:24) The actual figure was much lower, 19.8 % (BTM 2008: 24). To some extent, Vestas’ focus on profitability has sacrificed its superior market leadership. A mere 2 years ago, Vestas enjoyed a 28.2 % market share (BTM 2008:29). On the other hand, Vestas has become profitable since its 2005 annual report. In conclusion, this brief walkthrough of Vestas and its competitive situation has displayed its uniqueness as a global player and the difficulties in that national players dominate the five major markets. It also showed the Sino-American dominance on the demand side of global wind energy. Bachelor Thesis © Emad Sharghbin 2009 35 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. 5. Internal Strategic Analysis 5.1. Value Chain Analysis – Using Barney’s VRIO framework The purpose of this section is to identify the competitive advantages of Vestas. The analysis will be conducted using Barney’s synthesis of his own VRIO analysis with Porter’s classic Value Chain framework (Barney 2002, Porter 1985: 33-61). The synthesised model is efficient in valuating organisational strengths and weaknesses. The level of competitive advantages within each value-creating activity will be judged according to Barney’s four criteria behind the VRIO abbreviation. Activities can be valuable, rare, costly to imitate and exploited by the organisation (Barney 2002: 150-62). In order to maintain competitive parity, an activity has to be valuable, if the company is to exploit it. If the activity is also rare, exploitation may grant the firm a temporary advantage over competitors. If the activity is also costly to imitate, competitors are deterred from pursuing it. If the company’s strategies and policies enable the firm to exploit the activity, the company has a sustained competitive advantage (Barney 2002: 150-51). It is worth noting that Barney strongly recommends building the resource-based analysis on intra-organisational information. However, obtaining and analysing such data is outside the scope of this report, so the analysis will be conducted using external data as an approximation for intra-organisational information (Barney 2002: 161-62). In the analysis, all information from Vestas has been regarded as particularly biased and treated very critically. The following figure presents an overview of the findings, and subsequently, the activities will be discussed further in depth. Bachelor Thesis © Emad Sharghbin 2009 36 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. Figure 12: Value chain activities as well as their competitive and financial implications Costly to Value Chain Activity Valuable Rare Imitate Primary Logistics Production Marketing and Sales Support Supporting Infrastructure HRM R&D Exploited by Competitive Organisation Implications Financial Implications Parity Temporary advantage Temporary advantage Parity Neutral Positive Positive Neutral Temporary advantage Sustained advantage Sustained advantage Positive Superior Superior The theoretical foundation is based on Barney (Barney 2002: 150-51). Value Chain as proposed by Barney , procurement has been ignored.( Barney 2002: 137) 5.1.1. Primary Activities Within logistics, Vestas has implemented a Six Sigma System with its suppliers as one of the most important tools among Vestas’ quality improvement initiatives (AR 2008: 018). Six Sigma is fundamentally about identifying and eliminating causes of defects and errors in business processes, notably manufacturing (Zu 2008) While some authors believe Six Sigma to be just traditional quality management in a new packaging (Clifford 2001), others have found empirical data proving the benefits of the Six Sigma approach. (Pande 2000). A discussion of the theory is beyond the scope of this thesis, but Vestas’ initiative is considered positive. Secondly, Vestas’ “sourcing excellence”, one of Vestas 12 must-win battles, recommends suppliers to set up business in the US and China. Vestas and its suppliers will in this way be ready for the expected growth within these two key markets. (AR 2008: 018), Finally, Vestas’ policy of having two suppliers of all items should also decrease potential bottlenecks (AR 2008:23). Outbound logistics is fairly complicated in the industry. Vestas has established several long-term relationships with transportation providers, which is valuable, but not rare. All the above-mentioned initiatives are certainly valuable, but also fairly common within the industry. Within logistics Vestas has presently no comparative advantage, only competitive parity. Bachelor Thesis © Emad Sharghbin 2009 37 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. The production activity is different. Vestas is in the process of building factories in Spain, China and US (AR 2009: 21). Having production facilities in these major markets should give Vestas an advantage, although the advantage of in-sourcing production is debated within the industry (Finansnyheder: 20/03-09).10 Furthermore, Vestas has experienced fewer failures recently (AR 2008: 59). Finally, another one of the 12 must-win-battles, “Production Excellence”, has increased the already strong output by 4,5 % from 2007 to 2008 underlining Vestas’ production capability (AR 2008: 18). The relatively high productivity, the recent initiatives in Vestas’ pipeline, and “global” production facilities are very valuable. The latter especially is relatively rare in an industry where some of Vestas’ competitors are mainly national players. They can, however, be imitated and especially the Chinese manufacturers Goldwind and Sinotel have a cost advantage due to the low Chinese wages. Thus, Vestas currently enjoys a temporary competitive advantage. The sales and marketing activity has recently participated in the design of the wind turbines. The intention is to create turbines that satisfy customer needs based on the input from the sales people (Inside 01/09: 6). Also a New Group Marketing and Customer insight Business unit has been established recently in order to maintain continuous dialogues with customers (Inside 01/09: 33). Being present throughout the world with established sales offices and business units is also certainly valuable and fairly rare, but can be imitated. Vestas benefits from a temporary, competitive advantage within this field. The support activity has recently been strengthened tremendously by the creation of the business unit “Vestas Spare parts and Repair” (AR 2008:29). The unit will among other things be occupied with Vestas’ 24 hour monitoring of more than 11,500 wind turbines globally (AR 2008: 19). These steps are more than necessary, after Vestas has experienced several negative incidents with turbines malfunctioning and wings falling off (Ingeniøren 29/01-08). Vestas’ support function is valuable, but has underperformed in the past. Currently, the firm barely obtains a comparative parity within the activity, although this may change in the long term. 10 Some professionals say Vestas should outsource production of towers & gearboxes. (Finansnyheder: 20/03-09) Bachelor Thesis © Emad Sharghbin 2009 38 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. 5.1.2. Supportive Activities The infrastructure of Vestas is relatively strong. Its infrastructural strength lies in the organisation of several different business units within all relevant geographical areas, (AR 2008: 23-24). This strong foundation is only supplemented with new business units such as the repair business unit and Vestas China (Shareholder information 01/09: 5). The company has a temporary competitive advantage, which may however be replicated. Within human resource management, the firm has a relatively strong employer brand and a strong graduate programme, which may attract above-average skilled employees in the near future (AR 2008: 5)11. The initiative “People before Megawatt” entails education of existing employees (AR 2008: 19). The huge e-learning programmes within all levels and units of Vestas should both upgrade the skills of employees and improve the employee retention rate further (Inside 01/09: 13). The programmes are supplemented by 11 worldwide training centres (Inside 01/09: 36). Finally, in a time where absence is increasing in some industries, Vestas enjoys a low total absence due to illness of 2.5 % compared to the average of 3.9 % in Danish industries (AR 2008: 100). On the negative side, the financial crisis may force Vestas to lay off employees, even though they intend not to (AR 2008: 10, 18). A massive lay-off will of not be beneficial to the HRM activity. Also, 74 % of Vestas’ employees work within the Euro zone, so Vestas probably also has relatively high wage expenses (AR 2008: 24) In conclusion, Vestas’ history as one of the earliest wind turbine manufacturers and sole Danish survivor has created a unique organisation. Vestas possesses a sustained competitive advantage. In terms of research & development, Vestas has recently opened the industry’s largest R&D centre in Aarhus with another R&D centre in Singapore. This is in addition to its already existing R&D centre in Colorado, US (AR 2008: 21). These expansions are equivalent to an increase by 58 % or 1345 employees from 2007 to 2008. These R&D centres will be accompanied by coming R&D centres in the UK and in the energy capital of US, Houston (AR 2008: 21). Vestas also formed a strategic alliance with Boeing to collaborate on 11 Some observers find the wind energy industry to be inept in recruitment (Epn 07/04-09) Bachelor Thesis © Emad Sharghbin 2009 39 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. technology research, although Vestas R&D director points out that it will take years for specific results to amount (Finansnyheder 17/03-09). There is no doubt that Vestas invests heavily in R&D and the capabilities Vestas has built up within this activity are valuable, rare and costly to imitate. Thus, Vestas clearly has a sustained competitive advantage. In conclusion, the VRIO-based value chain analysis has visibly identified Vestas’ areas of competitive advantage. Its core competencies lies within its talent-attracting HRM and global, extensive R&D. Clearly, both are sustained competitive advantages. 5.2. Summarising SWOT table A typical SWOT-analysis is used as an independent, analytical tool, and enjoys much popularity among practitioners but less in the academic world. Several authors point out the weaknesses of SWOT as an analytical tool (Barney 2007: 11). Especially Grant is critical in saying that: “…[A]rbitrary classification of external factors into opportunities and threats and internal factors into strength and weaknesses is less important than a careful identification of these external and internal factors followed by an appraisal of their implications.” (Grant 2008: 13) While their observations are duly noted, the factors were identified and discussed rigorously in the strategic analyses. In the following the SWOT framework shall be used to summarise the previous strategic analyses. Bachelor Thesis © Emad Sharghbin 2009 40 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. Figure 13 Summarising SWOT-analysis Strengths Weaknesses Global production and sales Relatively high wage expenses. Strong presence in key growth markets Weak support - turbines crashing New SBU’s: China and Repair Centre Decreasing market share Strong HRM Industry’s strongest R&D investments Still market leader One of 3 firms with off-shore capabilities Opportunities Threats Strong political support Dependence on political support Kyoto Protocol, REFIT & PTC Short-term: Current financial crisis' Changed social attitude Dependence on state of the market Wind research can lower €/kw Threat from other renewables Ever increasing need for energy New entrants: major industrial/energy firms Environmental legislation National dominance in key growth markets High growth industry “Broad” substitutes cheaper than wind energy This will conclude the strategic analysis for now, as we move on to the historical financial analysis. Bachelor Thesis © Emad Sharghbin 2009 41 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. Part IV – Financial Statement Analysis of Vestas 6. Historical Financial Analysis 6.1.Introduction In this section, the historical performance of Vestas will be analysed. This critical section will together with the strategic section create the basis for the drivers of the valuation model. The areas of focus will be on growth and return on invested capital, ROIC. Growth and ROIC are the most important parameters to examine in relation to forecasting future performance (Penman 2007:371, Koller 2005: 61). The analysis shall cover all annual reports from 2004, which is after the NEC Micon acquisition and thus more representative of today’s Vestas. No comparison between Vestas and its competitors has been made, as Vestas has few competitors that are truly similar, as previously mentioned in section 4.3 Competitor and Demand Analysis, page 32. The focus will be on analysing the trends within Vestas’ key ratios. In line with the prescriptions provided by Penman and Koller, all the statements of Vestas have been reformulated. They are available on the enclosed Excel Sheet – see Vestas Model Base case scenario.xls. The reformulation separates operating from financing activities. Investors put money into a business based on the profitability of the operating activities of the firm. The financing activities only act as a supportive activity and should therefore be excluded in any valuation of a firm’s operational performance (Penman 2007: 244-249). A good example is the much renowned story of how German auto-maker Porsche in 2008 made 80% of its 9 bnEUR profits on financial derivatives linked to rival Volkswagen (Wall Street Journal 08/10-08). Obviously, Porsche’s ability to make profits from its operations, producing cars, did not increase. The story exemplifies precisely why operating and financing activities need to be separated. As mentioned in section 1.4 Delimitation, page 3, the accounting standards and techniques of Vestas and so forth shall remain uncommented. However, note that minor corrections have been made. As an example, some financial leases are included under property, but Bachelor Thesis © Emad Sharghbin 2009 42 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. they are so insignificant that their effect is ignored (AR 2008: 68). Also, it is necessary to comment on currency effects that could change the growth figures dramatically. The functional currency of Danish Vestas is still DKK. The presentation currency in their annual reports and consequently also in this thesis is EUR (AR 08:52). The Danish Krone’s strong tie to the Euro does not terminate currency risks completely, as 40 % of Vestas’ income still originates from outside the euro-zone, while 80 % of the costs are within the euro-zone (AR 2008: 24). However, Vestas hedges contracts by means of forwarding contracts and also hedges any additional, exposed net cash flows (AR 2008: 24). Therefore, currency effects are ignored assuming that Vestas’ currency risk exposure is somewhat insignificant (AR 2008: 87-88). 6.2. Growth Growth is as mentioned one of the most important drivers for value-creation. As profit margins are mean-reverting to some extent, it is evident that long-term growth in revenue is essential for value-creation (Koller 2005: 189). The next section shall therefore deal with the revenue growth in Vestas. As the following figures displays, Vestas has experienced impressive growth rates in the 2004-2008 period with a compounded annual growth rate of 21 % (CAGR). Decomposing the growth in geographical areas reveals that Americas, hereunder the US, provided the highest growth. Asia however delivered a somewhat more sluggish growth while Europe supplied the biggest increase in terms of absolute figures. Figure 14: Historical growth rates – Geographical segments Growth in Vestas' operating revenues 2005 2006 2007 2008 CAGR Europe 45,7% 0,2% Americas 221,9% -5,1% Asia/Oceania -13,8% 25,8% 20,5% 67,3% -1,6% 37,0% 10,2% 6,8% 19,3% 41,3% 7,9% Total 26,1% 24,2% 20,6% mEUR 4.000 Geographic segments Vestas' revenue Geographic segments 3.000 4.000 2.000 3.500 Growth rates in (EUR) Bachelor Thesis © Emad Sharghbin 2009 7,6% 1.000 3.000 0 2.500 2.0002004 2005 2006 2007 2008 mEUR 51,6% Europe North and South America Asia and Oceania 1.500 1.000 500 0 43 2004 Europe 2005 2006 North and South America 2 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. The growth rates are remarkable, but are also expected within a growth industry such as wind energy. Even more so during a time of massive political and economical support, as discussed in the strategic analysis. Also note, that Vestas’ relative growth is lower than the market, since Vestas has lost market share in recent years. However, Vestas has focused on profitability rather than market share, as the next section will display. The three geographical areas, Europe, Americas, and Asia, are also selected as the main forecast drivers. The decomposition of the historical growth rates of these areas enables me to estimate more precise, future growth rates for these drivers in the valuation. Although Penman suggests a very in-depth analysis of growth and sustainable earnings, the above-mentioned analysis shall suffice (Penman 2007: 409-437). 6.3. ROIC ROIC is one of the best value drivers in an evaluation of a firm’s ability to create value for its shareholders (Koller 2005: 183). As discussed in the review of valuation models, a firm’s ROIC has to exceed WACC for it to make any economic profit; otherwise the firm is breaking even or even destroying value. The strength of ROIC is also its focus on the profitability of operating activities, rather than the more traditional ROE and ROA that mix up financing and operating activities (Penman 2007: 378). ROIC measures the operating profitability of capital employed as it is net operating profit less adjusted taxes (NOPLAT) divided by invested capital (IC) (Koller 2005: 182). ROIC NOPLAT IC The invested capital can be calculated exclusive and inclusive of goodwill. Obviously, excluding goodwill measures profitability on the physical assets held. Including goodwill is superior in measuring the historical performance for the company’s shareholders (Koller 2005: 183). Management must create value for the firm with any acquisitions and consequent goodwill. If ROIC does not exceed WACC, the firm is destroying value for the shareholders. Focus will therefore be on ROIC inclusive goodwill. The sources of the recent improvements in Vestas’ ROIC can be found by decomposing the ROIC in a ROIC-tree. Figure 15: Decomposing ROIC-tree for Vestas based on 2005-2008 annual reports Bachelor Thesis © Emad Sharghbin 2009 44 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. Vestas has improved its ROIC significantly within recent years. Although the cash rate has some impact, the development is mainly caused by a significant increase in Vestas’ operating margin, here defined as the EBITA-margin. S,G,A,R&D12 expenses have increased somewhat, but have been offset by the lowered depreciation rate. Sales expenses have increased due to the increase in the global sales forces, which reflects Vestas global effort. The main driver behind the change in operating margin is the change in gross margin. Vestas has simply been better at reducing the cost of sales, as illustrated in the following figure. Figure 16 Expense groups in Vestas in percent of total revenue 2005-2008 12 The cryptic abbreviation represents sales, general, administrative, and research & development costs. Bachelor Thesis © Emad Sharghbin 2009 45 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. Expenses in % of revenue Revenue in each year = 100% 1,8% 100% 2,3% 3,9% 96% 4,9% 2,4% 5,5% 2,0% 5,7% 86% 82% 79% 2006 2007 2008 R&D Sales, Gen. & Adm. Exp. Cost of sales 0% 2005 Vestas has successfully reduced the cost of materials used. Obviously, the drop in raw material price in 2008 has been beneficial, as a major part of the costs of a wind turbine is raw materials. Steel alone can amount to 15 % of the total costs. Still, CEO Ditlev Engel has had an increased focus on the EBIT margin since his takeover from former CEO Svend Siggård in 2005, and the efforts have had positive results. (AR 08:16). Returning to the ROIC-tree, it is obvious that the strengthened ROIC-margin is also carried forward by an increase in average capital turnover. While fixed assets have not been employed more effectively, net working capital has decreased rapidly, which is a positive trend. The main reason is prepayments from customers, which partially outweighs the effect of capital employed in assets and trade payables (AR 2008:10). Most factors are believed to maintain their positive trend, except for the working capital ratio which cannot be expected to maintain its current low level (AR 2008: 26). Although Koller has documented a somewhat mean-reverting process in the ROIC for high performers, Vestas will probably be able to maintain its high ROIC, at least in the coming years (Koller 2005:148). This assumes that the financial crisis will only have short, temporary effects and that the strong, political support is maintained. Overall, the historical financial analysis indicates a strong performance. Both high growth rates and higher efficiency in form of increasing EBIT-margins have been accomplished. 6.4.Credit Risk As the last step in the historical financial analysis of Vestas, the inherent credit risk in Vestas will be touched upon. Even CEO Engel acknowledges its impact on Vestas (AR 2008: Bachelor Thesis © Emad Sharghbin 2009 46 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. 5). The theoretical foundation will be the framework provided by Falkenstein in Lundholm, in which the credit risk is based on 5 key ratios13 (Lundholm 2007: 115). Within the different ratios, industrial firms are grouped into 10 deciles. The firms are then assigned historical default probabilities according to their individual deciles. Refer to appendix 11.3 Lundholm’s Credit Risk Framework, page 79. The following table shows that Vestas is an unlikely candidate to default within the near future. Its default probability is remarkably below the historical average of 5 percent for industrial firms. This is both including and excluding the outlying quick ratio. (Lundholm 2007: 112). Figure 17: Vestas’ Credit Risk in the Lundholm Framework 2004 2005 2006 2007 2008 Decile (1st is best) Default prob. NI/Assets -4% -8% 1% 5% 10% 2 2% Liabilities/Assets 60% 69% 65% 65% 63% 6 5% EBIT/Int. exp. -128% -370% 489% 2953% 7422% 1 1% Sales growth 52% 8% 26% 24% 4 3% Quick Ratio 60% 47% 54% 56% 34% 10 9% Avg. excl. Quick ratio 2,8% Avg. incl. Quick ratio 4% The decreasing quick ratio is caused by an increase in its denominator, current liabilities.. The increase in current liabilities is mainly due to increases in prepayments and trade payables (AR 2008:10). Vestas may be unable to pay back these trade payables that are all due within a year if sufficient cash in-flows are not attained (AR 2008:87). On the other hand, Vestas enjoys an enormous EBIT/Int. exp. ratio, so they seem improbable to default on their loans. Overall, Vestas’ inherent credit risk is not significant at the present time. With a solid foundation in the financial and strategic analyses, it is now possible to move on the valuation part of this thesis. 13 A sixth ratio, inventory holding time, has been ignored, as this is not applicable to Vestas. Bachelor Thesis © Emad Sharghbin 2009 47 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. Part V – Valuation of Vestas 7. Estimating the Cost of Capital 7.1. Introduction The purpose of this section is to find the weighted average capital cost (WACC) of Vestas, which serves as the discount factor in the later DCF and RI valuations. Both are made from the enterprise perspective, hence the need for the full cost of capital and not just cost of equity (Plenborg, 2000: 6-11). The WACC is the approximation for the average opportunity cost of the all investors. The WACC treats the different types of investors according to the source of capital – debt, equity, convertible debt and various other sources (Penman 2007: 473-75). As the WACC is a weighted model which takes the tax shield effect of debt into consideration, it can be used as a discount factor while ignoring debt. D/V and (E/V) equal the target level of debt(equity) to market-based enterprise values. rd and (re) equal cost of debt (equity), while Tm equals the firm’s marginal tax rate. (Koller 2005: 292) D E WACC * rd * (1 Tm ) * * re V V 7.2.Estimating the Cost of Equity The WACC requires an estimate for cost of equity re. Estimating the cost of equity poses a difficult issue within financial theory. Penman points out, "we really don’t know what the cost of capital for most firms is” (Penman 2007: 114-16). The CAPM model claims re is equal to the risk-free rate rf plus the security’s beta i times the expected return of the market E(rm) subtracted by the risk-free rate rf. Beta is a measure of the firm’s correlation with the market and thus the only firm specific measure in the model. E(ri) = rf + i [E(rm) – rf] Source: Koller 2005:294. Some scholars are dismissive of CAPM models. Penman points out the circularity in traditional CAPM models as they use market-weights instead of intrinsic value-weights. Bachelor Thesis © Emad Sharghbin 2009 48 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. That is, market-prices are used in the estimation of equity cost of capital, which are then use to find underlying value (Penman 2007:693). Instead, Penman suggests assessing risk by conducting analysis of the determinants of fundamental risks. Such an analysis is outside the scope of this thesis. Therefore, the school of thought behind Koller will be adopted and a traditional CAPM shall be employed. As previously mentioned in the discussion of the valuation models, the thesis is based on the CAPM model as the best available option in lack of an empirically valid asset pricing model. In the CAPM, especially beta and the market risk premium are highly difficult to estimate (Koller 2005: 298). The risk-free rate is relatively easy to assess, as ten-year government bonds can be used as a proxy. Vestas has a majority of estimated 60 % of international shareholders in 2008 (AR 2008: 119). Therefore, US bonds are used, but alternatively Eurobonds could also be used. Final option is Danish government bonds, although their relatively strong ties to Eurobonds may call for simply using these bonds instead. To find the market risk premium, the American S&P 500 index has been employed. The main reasons are that total return index data is easily available back until 198814 and Vestas’ majority of international shareholders. However, choosing among major indices has little importance, as they are strongly correlated (Koller 2005: 310). The following table presents the author’s results and compares them to the results of the main researchers in the field. For the calculations refer to appendix 11.1 Market Premium calculations, page 77. Figure 18: Estimated market premiums in recent research - S&P 500 Aritmetic average Geometrical average My estimate Arnott Dimson 4,8% 2,4% 5,0% 3,0% 3,0% - Welch Shiller 6,0% 6,0% 4,3% - Wright 7,0% 5.5% Ibbotson 9,4% 7,1% Based on Arnott (2002), Dimson (2003), Welch (2000), Shiller (2000), Smithers (2003), and Ibbotson (2006). As the table shows, there are significant differences in the estimates of the market risk premium. Wright proposes that arithmetic averages are conceptually superior, but less 14 Strictly speaking, one can find the total return from 1963 by adding dividends to the price index return, but for the purpose of this assignment, 20 year-old data from 1988 is deemed sufficient. Bachelor Thesis © Emad Sharghbin 2009 49 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. stable than geometrical averages (Wright 2004:3). Still, geometrical averages, which are compounded annualised returns, appear more logical. Wright suggests using geometrical averages as the basis and then adjusting to the arithmetic average to some extent (Wright 2004:4). In Wright’s study, the confidence intervals of the estimations spans over a range as much as four percentage points (Wright 2004:4) Obviously, this imprecision can have dramatic consequences in the valuation processes, as different premiums lead to different discount rates and ultimately different intrinsic values. Market premium’s importance shall be illustrated later on in the valuation by the means of a sensitivity analysis. Paradoxically, the recent drop in the stock markets leads to the of lower historical risk premiums. Koller suggests that the problem can be solved by including long time series in the regressions (Koller 2005: 303). But as Arnott points out: “It is dangerous to shape future expectations based on extrapolating these lofty historical returns....” “the observed excess return and the prospective risk premium, two fundamentally different concepts that, unfortunately, carry the same label—risk premium.” (Arnott 2002:80&64). We are using historical market data as input in a fundamental analysis trying to estimate future risk premiums. These are not always strongly correlated and we are also violating Penman’s fundamentalist tenet about avoiding using market prices in a valuation of underlying value (Penman 2007: 19) In lack of stronger theoretical foundation, we have to continue regardless of the inherent imprecision. The estimated risk premium of in the author’s study in all likeliness is affected by the recent drops in the market. As current investors are assumed to be increasingly risk-averse, the underlying prospective risk premium must have risen as well. The long-term value of the risk premium is therefore estimated to be 5.5 %. 7.3. Estimating beta In estimations of betas, scholars frequently debate two common issues with regard to datachoice, the frequency of observations and the time necessary for good approximations. Koller suggests rolling five-year monthly data, but also notes that 2 year weekly data is used by the information service provider Bloomberg (Koller 2005: 308). Using daily data Bachelor Thesis © Emad Sharghbin 2009 50 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. entails more noise and would also cause a challenge with regards to holidays.15 The author’s studies are based on 2-year weekly data. The figure below displays the equity betas for Vestas and its two closest competitors Gamesa and REpower on rolling two-year basis. The equity betas have been estimated by using total return on the stocks regressed on the MSCI World Index. It is beneficial to use a large market index such as MSCI World, and not the local Danish OMX20 index. Vestas is one of the most dominant stocks in the OMX with a massive 10,2 % market weight as of April 15. Using this index would create a huge bias, as Vestas’ own stock price is too highly correlated with the OMX20 index (Koller 2005: 310). See appendix 11.2 Market Weight Report CSE 15/04-09, page 78. The Dresdner Kleinwort Wasserstein’s index of renewable energy firms has been included, as an indication of the renewable sector’s risks. The weighted-average of a group of firms is of course less volatile than a single firm (Koller 2005:311). As the index’ beta is above 1, the whole sector is evidently more sensitive to the state of the market - as previously discussed in the strategic analyses. Figure 19 Rolling 2-year beta regressions against the MSCI World index. Evidently, beta fluctuates remarkably across time, as betas can be highly volatile (Penman 2007: 113-16) Like market premiums, different betas can alter valuations significantly. 15 Different countries have different holidays, and therefore the MSCI is updated on days where Vestas is not traded, but this could be avoided by matching the days of the stock and index. Bachelor Thesis © Emad Sharghbin 2009 51 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. Vestas has not changed its business model significantly within the last years and is assumed to continue their present business model. Therefore, it can be assumed that historical data can provide an estimate for the future beta. Calculating Vestas’ beta on a 2001-09 period returns a beta of 1.68 with an R2 of 24 %. However, in ’03-04 Vestas’ merger with former competitor NEC Micon evidently changed Vestas’ beta temporarily. On a 2006-09 period, the regression results in beta of 1.86 and a R2 of 49 %. Again, we see that while the formulas may be highly precise, the assumptions and the end result are relatively imprecise. In lack of better theoretical foundation, we continue by estimating a beta of 1.6 for Vestas. 7.4. WACC Vestas has in recent years had a solvency ratio of roughly 30-35 %. The firm pursues to maintain a solvency ratio of at least 40 % on a book value basis (AR 2008:92). It is feasible to use ten-year US government bonds as a proxy for the risk-free rate due to the 60 % majority of international shareholders in Vestas (AR 2008: 119 & Koller 2005:296). Furthermore, Vestas informs that their weighted interest rate is 6.0 % in 2008 (AR 2008: 77). The future interest rate is estimated to be maintained in the 6 % region, but still 50 basis points higher, as the borrowing can be more expensive for corporations in the current market. By using the WACC formula previously presented on page 48, it is possible to calculate the past and future WACC. The underlying data behind the risk-free rate beta, return on debt, and marginal tax rates are found in the annual reports, and thus represent the values on 31/12 of the relevant year. Bachelor Thesis © Emad Sharghbin 2009 52 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. Figure 20 Historical and future decomposition of Vestas’ WACC. 2004 Rf 2005 4,29% 4,37% 2006 4,67% 2007 2008 4,21% 2,18% Forward WACC 3,9% Beta 1,51 1,44 2,01 1,63 1,98 1,60 Risk Premium 5,5% 5,5% 5,5% 5,5% 5,5% 5,5% Re 12,58% 12,29% 15,72% 13,17% 13,04% 12,74% EqV / MV 74,3% 81,8% 95,7% 98,3% 97,7% 97,7% Rd 3,10% 4,50% 4,90% 6,30% 6,00% 6,50% DV / MV 25,7% 18,2% 4,3% 1,7% 2,3% 2,3% 30% 30% 28% 28% 28% 28% 9,90% 10,62% 15,20% 13,03% 12,85% 12,56% Marginal tax rate WACC In order to calculate the future WACC, the previously estimated beta of 1.6 has been employed. Even though Vestas has a target capital structure, one could argue in favour of using changing WACC values over the years, as the market value of equity/debt ratio is unlikely to remain constant. However, one can also regard the found WACC as an average of the different, theoretically correct WACC’s (Lundholm 2007:203). Alternatively, one could calculate different WACCs or in severe cases, use the adjust present value model, as Koller suggests (Koller 2005: 326) However, the added complexity is likely to exceed the potential benefits, as Lundholm points out (Lundholm 2007:204). In conclusion, the estimated WACC of 12.56 % shall be used as a constant in the valuation. Bachelor Thesis © Emad Sharghbin 2009 53 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. 8. Valuation 8.1. Introduction Now that the WACC is also estimated, the two main parts, strategic analysis and historical financial analysis come together in this section and form the basis for the valuation of Vestas. First, the choice of forecast-drivers and variables held constant will discussed. In order to give a more nuanced picture of Vestas’ situation, the valuation will be conducted under three scenarios; base, bearish and bullish scenarios. These will be largely based on the conducted strategic and financial analyses. Naturally, the base scenario is the most probable, but the two more extreme scenarios have also been developed to give investors a detailed picture of how differently things may develop for Vestas. In other words, the scenarios depict the downside and upside of the share. All three scenarios include both Discounted Cash Flow and Economic Profit models. Next, a sensitivity analysis and a simulation on a selection of the variables will follow, which eminently shows the investor the inherent uncertainty in the valuation. In the sensitivity analysis, it will be clearly depicted how sensitive the valuation is to isolated changes in the variables. The simulation of key variables will indicate the overall effect of variable deviations on the share price. Finally, Vestas will be compared to its peers via a peer group analysis, which is based on the market’s pricing of shares Note that the chosen cut-off date for the valuation is April 14, 2009, and that the valuation value naturally differs from a valuation conducted on January 1, 2009. However, the issue has been tried to be solved via a mid-year correction factor, which takes into account how much of the year has already passed. 8.2. Forecast-drivers and constant variables For all three scenarios, the valuation will be conducted in using 3-stage models, including a short-term, long-term and continuing value periods. The most relevant and practically feasible forecast-drivers differ in the three valuation periods. Forecast-drivers are variables that are feasible to forecast. Variables that are insignificant or difficult to forecast Bachelor Thesis © Emad Sharghbin 2009 54 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. shall be held constant. The variables held constant in all three scenarios are commented on in section 11.8 Constantly held variables in the forecast, page 84, where their values are also available. For most of the constantly-held variables, historical figures have formed the basis. Three important variables shall be commented on here. The WACC is unlikely to remain the same in the 3 scenarios, but it is challenging to find sound, supportive arguments for different betas and thus different WACC’s in the strategic scenarios. As mentioned previously, there is much insecurity attached to the WACC in the first place. The WACC has therefore been maintained at the same level in all three scenarios. For all three scenarios, ROIC and growth are expected to decrease as new entrants will eventually be attracted as discussed in the strategic analyses. The following figure shows the different forecasts drivers. Figure 21 Overview of forecast drivers in the different valuation periods Period Year Revenue Growth Europe Revenue Growth Americas Revenue Growth Asia Gross Margin Total Revenue Growth EBITA-Margin Detailed forecast period Continuing Value Short-term Long-term 2009-2013 2014-2023 > 2024 RONIC Growth in NOPLAT In the short-term, the chosen main drivers are revenue growth forecasts that have been divided into 3 geographical zones. This enables detailed forecasts, which are naturally based on the previous strategic analyses. China and the US will constitute a major part in the growth drivers from Asia and Americas, respectively. Obviously, one could have forecasted separately on the on- and off-shore segments, but the off-shore segment only constitutes 5 % of the market forecast for the period 2009-2013 (BTM 2008:61) Alternatively, one could also try to forecast on country level. The increased complexity would go against Penman’s tenet of maintaining “a good balance between costs and benefits” (Penman 2007: 21). Bachelor Thesis © Emad Sharghbin 2009 55 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. In the long-term, revenue growth and EBITA-margin were chosen, although EBITDA or EBIT margin could work as well. For the continuing value, RONIC and growth on NOPLAT are the obvious drivers (Koller 2005: 273-75). RONIC, return on newly invested capital, is equal to ROIC in 2023. The growth rate in the terminal period is an estimated 3.5 %. This may seem high, but will be dealt with in the sensitivity analysis. Internally, Vestas uses a 2.5 % long-term growth rate, but acknowledges that it is well below the expected growth rate (AR 2008:62) Finally, Vestas’ own forecasts have been noted, but the forecasts of this thesis are based on an independent basis. As an example, Vestas’ 7,200 mEUR forecast for 2009 is deemed improbable and such subjective considerations permeate the models. 8.3. Base, Bearish, and Bullish scenario-analyses The following three subsections will present the main findings from the three scenario analyses. The main focus is on the short term. However, the models are printed in hardcopy on page 81 to 83 and also available electronically on the attached CD-rom. A guide to the models is available in the appendices as well, on page 79. 8.3.1. Base case scenario The strategic analysis showed the unique and strong position Vestas enjoys in the current market. Combining this with the strong indications from the financial analysis of Vestas, it is probable that Vestas will be able to maintain its high growth and increasing profitability. In the base case scenario, Vestas is forecasted to experience a moderate growth rate of 8 % in 2009. Especially the Americas will show very little growth due to the difficulties for customers to raise finance. Asia and Europe will still have a modest growth, but Vestas will be 500 mEUR short from reaching their current guidance of 7200 mEUR revenue. The current low, raw material prices will enable a slightly increasing gross margin. In 2010, however, growth will start to increase, mainly picking up in the Americas. In 20112012, Asia and Americas will be the main growth drivers, but also Europe will show some growth as a consequence of the EU’s binding commitments to reduce carbon dioxide emissions. In 2013 the current PTC runs out in the US, which constitutes a major part of Bachelor Thesis © Emad Sharghbin 2009 56 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. Vestas’ business in the Americas (AR 2008: 25).16 It is difficult to forecast the political development in the field. The PTC is expected to be followed by a similar supportive system. Some disruptions in the market are anticipated, however, which is why the growth is comparatively less in this year. China will constitute a major growth driver in the AsianPacific region especially from 2011 and onwards. Figure 22 Development in growth, ROIC, EBITA and gross margin in the base case scenario Drivers in 2009-2013 - Base case scenario 2009 2010 2011 2012 2013 75 mEUR 45% 35% 55 Revenue Growth Europe North and South America Asia and Oceania Total Revenue Growth 15% 0% 15% 11% 20% 20% 20% 20% 25% 30% 30% 27% 25% 30% 30% 27% 25% 35 10% 15 30% 22% (5) Gross margin ROIC 22% 22% 24% 24% 24% (25) 18% 15% 23% 27% 29% 25% 15% 5% 2005 2009 2013 2017 2021 -5% Revenue - Left axis -15% EBITA margin - Right axis ROIC - Right axis WACC - Right axis The current, comparatively high ROIC cannot be maintained in the short term. However, after 2010, the ROIC will yet again improve before reaching its maximum in 2013-14. The forecasts in this scenario provide an estimated share price of 52.77 EUR. For the base case scenario alone, the development of EP and DCF in the detailed forecast period shall be commented on. The development for the bearish and bullish scenarios are available in the appendices and on the enclose CD. 16 The US constituted 78 % of Vestas’ sales in Americas and 24 % of total sales. Bachelor Thesis © Emad Sharghbin 2009 57 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. mEUR Figure 23: Development of EP and FCF in the detailed forecast period. 2.500 2.000 1.500 1.000 500 0 (500) Development in FCF and PV(FCF) 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 Free Cash Flow PV of FCF mEUR Development in EP and PV(EP) 1.400 1.200 1.000 800 600 400 200 0 2009 2010 2011 2012 2013 2014 2015 EP 2016 2017 2018 2019 2020 2021 2022 2023 PV of EP The models both provide the same value estimate, as the theory prescribes. Their graphs, however, differ as they tell different stories. In the DCF model, it is obvious that 2009 provides a negative cash flow. This is at least partially due to the large capital investments. The DCF cash flows decrease more slowly than the EP and needs a longer time horizons as discussed in the theory. The EP models clearly indicates that the next, two years will be less profitable but then a significant amount of the forecasted economic profit is earned in the 2011-2015 period. As discussed under the theory of the valuation models, DCF and EP obviously tell different, but valuable stories. For more on the model, refer to appendix 11.5 Base Scenario – Valuation Summary, page 81. 8.3.2. Bearish scenario In the bearish scenario, Vestas is forecasted to experience a very modest growth in 2009. The Americas will in fact decrease in terms of total annual installed MW, due to especially a drop on the US market as well as intensified competition. Further market share will be lost to GE Wind. Asia and Europe will still have some growth. Vestas will be very far from their current guidance of 7200 mEUR revenue. The estimate is in the 6200 mEUR range, a minor improvement from 2008. The reason why the revenue does not decrease even further is due to the nature of the wind energy business. Revenue is only recognised upon complete Bachelor Thesis © Emad Sharghbin 2009 58 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. delivery for supply-only17 contracts, whereas revenue from turn-key18 projects are recognised as the individual turbines are delivered (AR 2008: 54). This means that Vestas will have significant revenue in 2009 based on contracts from 2007-08, and therefore the full effect will not be seen until 2010. Despite the current low raw material prices, the gross margin will be under pressure and decrease. In 2010, growth will start to increase, mainly picking up in the Americas. In 2011-2012, all three regions will show some growth. In 2013, the end of the current PTC in the US will cause a significant contraction of the American market. Figure 24 Development in growth, ROIC, EBITA and gross margin in the bearish scenario 75 2010 2011 2012 2013 Gross margin ROIC 25% 35 15% 15 Revenue Growth Europe North and South America Asia and Oceania Total Revenue Growth 35% 55 Drivers in 2009-2013 - Bearish scenario 2009 45% mEUR 10% -10% 5% 4% 15% 20% 10% 15% 15% 20% 15% 16% 15% 20% 15% 16% 15% -20% 15% 13% 16% 16% 18% 18% 19% 8% 5% 10% 11% 13% (5) (25) 5% 2005 2009 2013 2017 Revenue - Left axis ROIC - Right axis 2021 -5% -15% EBITA margin - Right axis WACC - Right axis After five unprofitable years due to relatively large capital expenditures and excess capacity, ROIC will again exceed WACC, but only modestly throughout the long-term period. The forecasts in this scenario provide an estimated share price of 17.89 EUR. For more 17 on the model, refer to appendix 11.6 Contracts, where Vestas “just” has to deliver the wind turbine. The customer is then responsible for maintenance and so forth (AR 2008: 54). This type of contract is less risky for Vestas, as operating risk is transferred to the customer. The share of supply-only contracts has been increasing (AR 2008:24). 18 Contracts, in which Vestas has to deliver the wind turbine and maintain service and reparations in typically 5 years (AR 2008: 54). Vestas is responsible until the “key has been turned”, hence the name. Bachelor Thesis © Emad Sharghbin 2009 59 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. Bearish Scenario – Valuation Summary, page 82. 8.3.3. Bullish scenario In the bullish scenario, Vestas is forecasted to experience a significant growth rate of 22 %. From all regions, but especially the Americas, orders will flow in heavily in Q3-Q4 2009. A precondition for such a development is that developers and utilities will find it easier to raise finance. A significant part of the income will be recognised in 2009, as excess capacity will enable quick execution of orders. Vestas’ forecasted revenue will be in the 7400 mEUR range, thereby exceeding the current guidance of 7200 mEUR revenue. The current low raw material prices will enable continuingly increasing gross margins. In 2011-2013, Asia and Americas especially will show impressive growth, carried forward by China and the US as mentioned in the competitor and market analysis. The EU’s binding commitments will also drive the European market forward as discussed in the PESTED analysis. In 2013, the current PTC will be followed by equivalent, or even reinforced legislation, which will enable Vestas to maintain significant growth in the US. Figure 25 Development in growth, ROIC, EBITA and gross margin in the bullish scenario Drivers in 2009-2013 - Bullish scenario 2009 2010 2011 2012 2013 Europe North and South America Asia and Oceania Total Revenue Growth 25% 15% 20% 22% 25% 25% 35% 26% 25% 40% 35% 30% 30% 40% 35% 33% 30% 40% 35% 34% Gross margin ROIC 24% 24% 26% 26% 26% 22% 21% 31% 36% 41% Revenue Growth mEUR 75 45% 35% 55 25% 35 15% 15 (5) 5% 2005 2009 2013 (25) The current, comparatively high ROIC will Revenue - Left axis ROIC - Right axis 2017 2021 -5% -15% EBITA margin - Right axis WACC - Right axis decrease in 2009-2010 due to a low level of orders. Also, Vestas will have significant capital expenditures. However, the benefits of these expenditures will be reaped in the 2012-2016 period, where Vestas will experience a considerable ROIC. This scenario assumes that no competing renewable energy receives a breakthrough and that political support remains strong, as discussed in the PESTED-analysis. The forecasts in this scenario provide an estimated share price of 99.06 EUR. For more on the model, refer to appendix 11.7 Bullish Scenario – Valuation Summary, page 83. Bachelor Thesis © Emad Sharghbin 2009 60 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. 8.3.4. Conclusion on scenario analyses The bearish and bullish scenarios are intentionally extreme and deemed equally probable. Therefore, they will both weight 10 %, whereas the base case weighs 80 % in the final estimate. Using the three scenario estimates, the weighted share price estimate is 53.91 EUR. In conclusion, we see Vestas does enjoy a unique position within a unique growth industry, which is in line with the strategic analyses. The conclusion on the valuation part is that the estimated target price of 53.91 seems reasonable, although it was also evident that apparently precise valuations are in fact very imprecise and very sensitive to slight adjustments, and therefore the conclusion is made with certain reservations. Obviously, the two extreme scenarios exemplify how uncertain intrinsic value estimates are even for the same firm in different scenarios. Figure 26: The three scenario share price estimates and the weighted share price estimate. Share price Weights of the scenarios 10% 125 10% 99,06 € EUR 100 75 50 25 53,91 € 52,77 € 17,89 € 0 Bearish 80% Bearish Base Bullish Base Bullish Different Scenarios Weighted average The high level of uncertainty arises even without adjusting for possible variations in the variables. This will be further examined in the following section. 8.4. Sensitivity Analysis Just how sensitive the estimates are to even slight changes in the variables will be displayed in the following. The sensitivity analysis will be based on the base case scenario which will be used as a proxy for the weighted average. The analysis shall reveal the effect of an isolated, marginal change in selected variables, that is, while all other variables are held constant. In the section concerning estimation of the cost of capital, it was noted Bachelor Thesis © Emad Sharghbin 2009 61 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. several times, that the calculations were very uncertain. Therefore, the effect of these estimates on the share price estimate is extraordinarily interesting (Plenborg 09:4). Bachelor Thesis © Emad Sharghbin 2009 62 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. Figure 27: Sensitivity Analysis – Effect on share price, table and graph Change in %-points -1 +1 +2 Revenue Growth ('09-'23) -17% -9% 10% 21% EBITA margin ('09-'23) -32% -16% 16% 32% Growth in NOPLAT (≥2024) -1% 0% 3% 6% ROIC (≥2024) -4% -2% 0% 1% Market Risk Premium 70% 27% -19% -33% WACC 37% 16% -13% -23% Change in beta -0,5 -0,25 +0,25 +0,50 56% 23% -17% -29% Sensitivity Analysis 80% Effect in % on share price -2 60% 40% 20% 0% -20% -40% -2 -1 0 +1 +2 Change in %-points Beta Revenue Growth ('09-'23) EBITA margin ('09-'23) Growth in NOPLAT (≥2024) ROIC (≥2024) Market Risk Premium WACC Beta Note that beta varies with ±0,5 in the graph - not in percentage points like the other variables. As the table and graph show, the share price is affected tremendously by changes in all variables, except for growth in NOPLAT and ROIC in the terminal period. The percentage of value arising from the terminal period is in all three scenarios less than 50 %. The reasons are the relatively long detailed forecast period of 15 years and that the forecasts indicate that a comparatively significant amount of Vestas’ future, economic profit will be earned in detailed forecast period. Therefore, the model is fairly insensitive to change in the terminal value drivers, ROIC and NOPLAT. Revenue and EBITA margin, however, have a great effect on the share price. Unsurprisingly, however, the share price is most sensitive to the estimate market risk premium, beta and WACC (Penman 2007: 113-16). The sensitivity analysis has accurately displayed by how much the intrinsic value varies due to even slight changes in variables. This is one of the inherent weaknesses of the accounting-based models, as discussed in the review of the models. The estimated share price is extremely sensitive to changes especially in the cost of capital estimates, which are essential in the accounting-based models. The models’ basic foundation is built on the ability to discount using the cost of capital. Bachelor Thesis © Emad Sharghbin 2009 63 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. 8.5. Simulation of forecast-drivers The previous sensitivity analysis displayed the sensitivity of isolated, marginal changes in the drivers. The following simulation of the forecast-drivers will show the combined effect of variation in the drivers. It is also based of the base case scenario model. For each driver a simulation interval has been defined in each year. The interval defines the upper and lower boundaries of the simulation and has subjectively been defined as the maximum and minimum realistic boundary after the author’s judgment. The share price estimate is then simulated 1000 times. The model is enclosed on the CD as “Vestas Model Simulation model.xlsm”. In order to force some realism into the model and avoid very profitable years to be followed by very unprofitable years, a random number is drawn for revenue growth and the EBITA margin. The number then affects the following year’s development and so forth. Having simulated the value, the model then comes up with an estimate for the share price. The median of the 1000 simulations is 50.79 EUR with the arithmetic mean close by at 50.33 EUR. Although the figures are close to the 53.91 EUR weighted estimate, the following graph reveals how much the simulated estimates vary. Figure 28: Histogram of simulated share price estimates Histogram - Distribution of simulated share price No. of simulations 35 30 25 20 15 10 5 71 68 65 62 59 56 53 50 47 44 41 38 35 32 0 Simulated share price in EUR The histogram shows that the values vary significantly ranging from 34 to 69 EUR. The histogram is of course affected by the width of the chosen simulation interval, but still shows the potential variation in the estimated share price. In their different ways, the Bachelor Thesis © Emad Sharghbin 2009 64 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. scenario analyses, the sensitivity analysis, and the simulation model all display the uncertainty in the intrinsic value estimates. 8.6. Peer Group Analysis Finally, a peer group analysis shall be conducted in line with the conclusion from section 2.4.2 Multiple, page 8. Here, it was found that a peer group provides a useful sanity check of estimates found via other models, although it breaks Penman’s fundamentalist tenet about avoiding market based prices (Penman 2007: 19). Another problem with the approach was the difficulty in finding “true” peers and “true” multiples. In Vestas’ case, there is no other true, global peer to Vestas, as previously mentioned. In terms of market share, GE Wind is the closest peer, but as with Siemens Wind, both firms are only parts of conglomerates making comparisons troublesome. Therefore, the only available peers are Spanish Gamesa, German Nordex and REpower. REpower is owned19 by Indian Suzlon, which is also included. In the analysis, the multiples of the chosen peers shall be examined and compared to Vestas. In the following figure, PE and EV/EBIT(DA) multiples are used for comparison. EVmultiples’ strength lie in that they are unaffected by capital structure making comparison more correct (Koller 2005:366). Furthermore, scholars agree on the superiority of forwarded multiples. (Koller 2005, Liu 2002, Schreiner 2007). The focus is therefore on multiples for 2009-2010, although historical comparisons are also feasible. The multiples are based on the median of analysts’ forecasted multiples. 19 http://www.wwindea.org/home/index.php?option=com_content&task=view&id=175&Itemid=40 Bachelor Thesis © Emad Sharghbin 2009 65 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. Figure 29 Peer Group analysis of Vestas against Wind Energy Peers Peer Group - Wind Energy Author's estimates Company namc Price (USD) MV PE PE EV/EBITDA EV/EBITDA EV/EBIT EV/EBIT EV/EBIT EV/EBIT (mUSD) (2009) (2010) (2009) (2010) (2009) (2010) (2009) VESTAS WINDSYSTEMS 58,4 10.806,3 18 14 9 8 12 10 15 GAMESA CORPN.TEGC. 17,4 4.233,3 19 13 8 6 14 11 NORDEX 16,5 1.299,3 17 22 10 10 12 13 121,2 1.089,9 19 17 8 7 10 7 1,3 1.983,8 8 11 6 6 7 7 Arit. average excl. Vestas 16 16 8 7 11 10 Arit. average excl. Vestas and Suzlon 19 17 9 8 12 10 MV-weighted average 16 14 8 7 12 10 Vestas premium: Arit. average 10% -8% 17% 3% 7% -2% Same excl. Suzlon -6% -17% 8% -3% -4% -9% 7% 1% 19% 9% -1% -3% Fair price: Arit. avg. excl. Suzlon 40 35 46 41 41 39 Fair price: MV-weighted 45 43 51 47 42 41 REPOWER SYSTEMS SUZLON ENERGY Vestas premium: MV-weighted VWS price (2010) 15 43 Source: Thomson Financials Datastream The market assesses Vestas differently depending on how the multiple is calculated. One can also discuss whether or the not the slightly outlying Suzlon should be included. One could argue that outliers that alters the estimates significantly should be excluded from a statistical point of view. Obviously, this issue shows the problem with multiple valuation, as it is easy to manipulate the result by choosing the exact firms and multiples that give the desired result (Penman 2007: 78). The conclusions from the strategic analyses suggested that Vestas might be traded at a premium given its unique position as a global wind turbine manufacturer. The figure, however, gives the ambivalent results that the share is traded at both premiums and discounts, depending on multiple. The estimates from the base case scenario have also been included in the upper right corner, which are somewhat comparable to the consensus, especially in 2009. Multiples have also been conducted with Vestas among solar energy manufacturers and also with industrials such as GE, Siemens, and Caterpillar. Vestas does trade to a significant premium compared to the traditional industrials, indicating the market’s belief in wind energy as a growth industry. Vestas also trades a significant premium in comparison to solar energy manufacturers. This underlines wind energy’s strong position within renewable energy, which is in line with the conclusions from the PESTED analysis. For more on the additional Bachelor Thesis © Emad Sharghbin 2009 66 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. multiple analyses refer to appendix 11.10 Three peer group analyses: Wind, Solar, Industrials, page 87. 8.7. Comparing price targets and estimates. The following figure shows the share price targets from leading investment banks in comparison with the author’s estimated intrinsic value. Figure 30 Overview of the price targets in EUR from analysts and this thesis. Price targets in EUR from leading investment banks 120 100 EUR 80 60 40 20 0 Apr-08 Jun-08 Jul-08 Sep-08 Price target in EUR Oct-08 Dec-08 Feb-09 Mar-09 May-09 Author's estimated intrinsic value Source: www.kurstarget.dk – Vestas. See appendix 11.9, page 86. Coincidentally, the estimated intrinsic value is somewhat comparable with the consensus target prices. But in reality, the figures cannot be compared. The analyst is paid to come up with typically 12-month target prices. In most cases that means that intrinsic value are estimated similarly to this thesis, but are adjusted to reflect market moods, news flow, and other short-term perspectives. Rather than estimating the intrinsic value and ignoring any noise in the market, the analyst should normally be more interested in being able to predict the share price in 12 months. Doing so provides food on the table, and suits on his body. In a perfect world according to fundamentalists, one could imagine that all investors would ignore such short-term fluctuations and conduct solid, fundamentalist valuations as the sole basis for investment decisions. Share price fluctuations, “noise”, could only come from different estimates, not market mood or other short-term perspectives (Penman 2007: 9). In this fundamentalist world, bubbles would (almost) not exist and the difference between market prices and intrinsic values would theoretically be much smaller (Penman 2007: 67). Market price will converge towards underlying value over time under the Bachelor Thesis © Emad Sharghbin 2009 67 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. fundamentalist approach. As Bhojraj states, “current share price is like a noisy proxy for the true but unobservable intrinsic value” (Bhojraj 2002). However, reality is that in contemporary finance, markets do not necessarily act according to theory and the same applies to valuation models and so forth. Therefore, the author stresses the need for treating the theory and models very critically. Bachelor Thesis © Emad Sharghbin 2009 68 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. 9. Conclusion The purpose of this paper was to find the intrinsic value of Vestas in a stand-alone case. To do so, fundamental valuation models and their theoretical backgrounds were first discussed and evaluated in the search for the most appropriate theoretical frameworks. Asset-based and real options models were found to be useful mostly under special circumstances. Multiples valuation could provide quick and dirty sanity checks, but also had severe methodological and practical issues. For general purposes, the RI and EP models were the most preferred models by the author, as they are strong in telling the story of the firm’s economic performance. DCF is good in telling the development of cash flows. Therefore, a combination of the RI or EP model with a DCF model is theoretically strong. For this thesis, a DCF and EP combination was chosen, supplemented by a multiple analysis. Having chosen the models, focus was again on Vestas. Overall, the future looks somewhat bright for Vestas. The current political support is a essential for the industry, but also show the Achilles’ heel of the industry. In the longer run, the biggest threat would be if the support disappears before wind energy is competitive on an unsubsidised basis. The industry analysis indicated an attractive industry. The main negative aspects, however, include the bargaining power of buyers and the intensity of rivalry which are both expected to increase. The historical financial analysis revealed that Vestas has increased profitability mainly due to a drop in working capital and improved gross margin. Prior to the valuation, the problematic, theoretical issues concerning cost of capital were addressed in order to estimate Vestas’ cost of capital. Synthesising the strategic and financial analyses then enabled a forecast of the pro-forma budgets. Using the cost of capital, the intrinsic value of Vestas was derived under three different scenarios. Vestas is a company within a profitable industry with good growth opportunities. Much of the growth, however, is already incorporated in the stock. The thesis’ intrinsic value estimate of 53.91 EUR was an attempt to include all the mentioned observations and discussions into a single figure. As the sensitivity analysis and simulation indicated the Bachelor Thesis © Emad Sharghbin 2009 69 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. estimate may still be far from the true intrinsic value, but represents the author’s best estimate. An interesting part of this thesis was the continuous, general discussions about the severe, theoretical flaws in contemporary finance; in the financial theories, valuation models, and so forth. Estimates, share price targets, buy/sell recommendations and so forth should therefore be treated extremely critically and one must not be seduced by precise formulas with imprecise assumptions (Graham 1974). …[P]recise formulas with highly imprecise assumptions can be used to establish, or rather justify, practically any value one wishes.(Graham 1974) Bachelor Thesis © Emad Sharghbin 2009 70 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. 10. Bibliography 10.1. Academic books and articles Arnott, Robert D. 2002 “What risk premium is "normal"?” Financial Analysts Journal, Vol. 58, No. 2, pp. 64-85. Ashton, D. et al. 2003 “An Aggregation Theorem for the Valuation of Equity Under Linear Information Dynamics “,Journal of Business Finance & Accounting,vol. 30(3&4), pp. 416440. Atauallah, A. et al. 2009, “Non-linear equity valuation”, Accounting and Business Research, Vol. 39. No. 1. pp. 57-73. Barker, R. 1999, “The role of dividends in valuation models use by analysts and fund managers”, The European Accounting Review 8 (2), pp. 195-218. Barney. J.B. 1991, “Firm resources and sustained competitive advantage”, Journal of Management, Vol. 17, pp. 99-120. Barney, J.B. 2007, “Gaining and sustaining competitive advantage”, 3rd ed., Pearson Prentice Hall, United States of America. Bauman, M.P., 1996, “A review of fundamental analysis research in accounting”, Journal of Accounting Literature, Vol. 15, pp. 1-33. Bettman, J. et al. 2009, “Fundamental and technical analysis: substitutes or complements?”, Accounting and Finance, Vol. 49,pp. 21–36. Bhojraj, Sanjeev et al., 2002. “Who is My Peer? A Valuation-Based Approach to the Selection of Comparable Firms.” Journal of Accounting Research, Vol. 40, No. 2., pp. 407-439. Block, S. 2007, “Are “Real Options” actually used in the real world?, The Engineering Economist, vol. 52, pp. 255-267. Canegrati, Emanuele 2008. "A Non-random Walk Down Canary Wharf," MPRA Paper No. 9871. Clatworthy, Mark 2005, “Transnational Equity Analysis”, 1st ed., John Wiley and Sons, Wiley Finance, Chichester. Bachelor Thesis © Emad Sharghbin 2009 71 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. Clifford, L. 2001.,“Why you can safely ignore Six Sigma” Fortune, Vol. 143, No. 2, p. 140 Congress Report, CRS Report 2005 “Nuclear Power Plants: Vulnerability to Terrorist Attack”, The Library of Congress, 4/2-2005. Copeland, T. et al. 2001, “Real Options: A Practioner’s Guide” Texere LLC, New York. Demirakos, Efthimios et al. 2004, “What valuation models do analysts use”, Accounting Horizons, vol. 18, no. 4, pp. 221-240. Dimson, E. et al. 2003, “Global evidence on the Equity Risk Premium”. Journal of Applied Corporate Finance”, Vol. 15, No. 4, Dixit A. et al. 1995 “The Options Approach to Capital Investment” Harvard Business Review, Vol. 73, no. 3., pp. 105-116 Edwards, R.D. et al. 2001, “Technical analysis of stock trends” 8th ed., St. Lucie Press, Washington D.C. Fama, Eugene 1965, "Efficient capital markets: a review of theory and empirical work." Journal of Finance, vol. 25, no. 2, pp. 383-417. Gentry, James A. et al. 2003, ”Learning about intrinsic valuation with the Help of an Integrated Valuation Model”, FMA European Meetings Dublin June 5. Graham, Benjamin 1973, “The Intelligent Investor”, 4th ed., Harper and Row, New York Graham, P. 2006, “The heat is on: the future of energy in Australia” CSIRO. Grant, Robert 2008, “Contemporary Strategy Analysis”, 6th ed. Blackwell Publishing, United Kingdom Hoogwijk, Monica et al. 2008, “Global Potential of Renewable energy sources: A literature assessment.” REN 21 publication Ibbotson Associates 1998 “Stocks, Bonds, Bills and Inflation”. Year‐end summary report. Chicago. Imam, Shahed 2008, “The use of valuation models by UK investment analysts”, European Accounting Review, vol. 17, no. 3, pp. 503-535. Bachelor Thesis © Emad Sharghbin 2009 72 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. Ireland, R. et al. 2007,” The Management of Strategy” 8th ed., South-Western Cengage Learning, Canada. Irwin, S. et al. 2007, “What Do We Know About the Profitability of Technical Analysis?” Journal of Economic Surveys, 21, No. 4, pp. 786-826, Blackwell Publishing Oxford. Jensen, Ib Konrad 2003, “Mænd i modvind – Et dansk industrieventyr”, Børsens Forlag, Denmark. Jorgensen, Bjorn et. al. 2005, “An Empirical Assessment of the Valuation Accuracy of The Abnormal Earnings Growth Valuation Model”, Vol. 6, no. 10, presented by Korea University. Koller, Tim et al. 2005, “Valuation: Measuring and Managing the Value of Companies, 4th ed., John Wiley and Sons. Leslie, K. J. et al. 1998, “The Real Power of Real Options” Corporate Finance, lss. 158, p. 13 Lundholm, Russell 2007, “Equity Valuation and Analysis” McGraw-Hill, New York. Lundholm, Russell 2001, “On comparing residual income and discounted cash flow models of equity valuation: A response to Penman”, Contemporary Accounting Research, Vol. 18, no. 4, pp. 692-712. Ohlson, James A. 1995, “Earnings, Book Values, and Dividends in Equity Valuation”, Contemporary Accounting Research, Vol. 11, No. 2, pp. 661-687. Ohlson, James A. 2005, “On Accounting-Based Valuation Formulae”, Review of Accounting Studies, Vol. 10, pp. 323–347, Pande, P.S. et al. 2002. ”The Six Sigma way, Team Field book: An Implementation Guide for Process Improvement Teams”. McGraw-Hill, New York. Penman, Stephen 2001, “On comparing cash flow and accrual accounting models for use in equity valuation: A response to Lundholm and O’Keefe”, Contemporary Accounting Research, Vol. 18, no. 4, pp. 681-689. Penman, Stephen 2007, “Financial Statement Analysis and Security Valuation”, McGrawHill, Singapore. Plenborg, Thomas, 2009, “Værdiansættelse af virksomheder”, Børsens Håndbøger, no. 2. Porter, Michael 1985,"Competitive Strategy", The Free Press, New York. Bachelor Thesis © Emad Sharghbin 2009 73 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. Schreiner, Andreas 2007, “Equity Valuation Using Multiples: An Empirical Investigation” University of St. Gallen, DBA Dissertation. Sharpe, W.F. 1964, “Capital Asset Prices: A Theory of Market Equilibrium Under Conditions of Risk!, Journal of Finance, Vol. 19, pp. 425-442. Shiller R. 2000, “Measuring bubble expectations and investor confidence”. Journal of Psychology and Markets, Vol. 1, No. 1. Trigeorgis, L. 1993, “Topics in real options and applications” Financial Management, vol. 22(3), pp. 202–223. Trigeorgis, L. 2005, “Making use of real options simple: An overview and applications in flexible/modular decision making” The Engineering Economist, vol. 50(1), pp. 25– 53. Welch I. 2000, “Views of financial economists on the equity premium and on professional controversies” Journal of Business, Vol. 73, No. 4. Wright, Stephen et al. 2003, “A Study into Certain Aspects of the Cost of Capital for Regulated Utilities in the U.K.”, On behalf of Smithers & Co Ltd, commissioned by U.K. regulators. Zu, Xingxing et al. 2008: “The evolving theory of quality management: The role of Six Sigma”, Journal of Operations Management, Vol. 26, pp. 630-650. Bachelor Thesis © Emad Sharghbin 2009 74 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. 10.2. Reports and other sources “Acquisition of REpower by Suzlon is important step in international cooperation”: http://www.wwindea.org/home/index.php?option=com_content&task=view&id=175&Ite mid=40 Biologist Kåre Fog scientifically disproves Lomborg’s claims, page 26. http://www.lomborg-errors.dk/ BTM Consult Aps 2008, “BTM World Market Update 2007”, A. Rasmussens Bogtrykkeri, Ringkoebing, Denmark. Kurstarget: Showing the share price estimates of analysts concerning Danish firms. www.kurstarget.dk USGS 2009. US government-run U.S. Geological Survey points out “…most of the good spots to locate hydro plants have already been taken.” http://ga.water.usgs.gov/edu/wuhy.html Vestas SH, Shareholder information 2009/1, 12 page publication. Available at http://www.vestas.com/en/investor/shareholder/shareholder-information Bachelor Thesis © Emad Sharghbin 2009 75 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. News paper articles Business.dk 20/03/07: “Også i år vil Vestas mangle komponenter”. http://www.business.dk/article/20070320/borsnyt/103201015/ Børsen 11/03-09: “Vestas and Boeing to collaborate on technology research projects” http://borsen.dk/?treeid=2614&id=80302&urlpart_firm=vestas_wind_systems Børsen, 6/11-08: ”Vestas/CEO: Kunder ønsker ikke større møller” http://borsen.dk/?treeid=2613&id=55466&urlpart_firm=vestas_wind_systemsA Børsen, 22/04-09: ”Rival vil slå Engel i 2010” http://borsen.dk/investor/nyhed/155904/ Epn, 07/04-09: ”Vindbranchen er dårlig til rekruttering” http://epn.dk/brancher/energi/alternativ/article1657373.ece Finansnyheder, 20/03-09: ”Vestas: Vindekspert forstår ikke tårnfabrik i USA” http://www.finansnyheder.dk/news/shownewsstory.aspx?storyid=10213859 Finansnyheder, 17/03-09: “Vestas/R&D-chef: År til resultater af Boeing-samarbejde” http://www.finansnyheder.dk/news/shownewsstory.aspx?storyid=10207806 Guardian 17/03-09: “Shell dumps wind, solar and hydro power in favour of biofuels”. http://www.guardian.co.uk/business/2009/mar/17/royaldutchshell-energy Guardian 20/04-08: Warren Buffett quote, page 6: http://www.guardian.co.uk/world/2008/apr/20/usa.subprimecrisis Ingeniøren, 29/01-08: ”Vinge rev sig løs fra Vestas-mølle og fløj 40 meter væk” http://ing.dk/artikel/85123-vinge-rev-sig-loes-fra-vestas-moelle-og-floej-40-meter-vaek Wall Street Journal 08/10-08: “As rivals stall, Porsche engineers a financial windfall”. Bachelor Thesis © Emad Sharghbin 2009 76 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. 11. Appendices 11.1. Start End Frequency Name Code CURRENCY 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 Market Premium calculations 1989 Start 1989 Aritmethical avg. 4,84% 2008 End 2008 Geometrical avg. 2,96% Y Frequency Y S&P 500 COMPOSITEName - TOT RETURN US TREASURY IND (~U$) , CONSTANT MATURITIES 10 YR - MIDDLE RATE S&PCOMP(RI)~U$ Code Y74758 U$ CURRENCY Return on investment - Payment of interest = Excess 288,12 1988 9,17 379,41 1989 7,93 1,32 1,08 1,24 367,63 1990 8,15 0,97 1,08 0,89 479,63 1991 6,86 1,30 1,07 1,24 516,18 1992 6,7 1,08 1,07 1,01 567,09 1993 5,77 1,10 1,06 1,04 575,7 1994 7,81 1,02 1,08 0,94 792,04 1995 5,64 1,38 1,06 1,32 973,9 1996 6,34 1,23 1,06 1,17 1298,82 1997 5,74 1,33 1,06 1,28 1670,01 1998 4,7 1,29 1,05 1,24 2002,11 1999 6,41 1,20 1,06 1,13 1837,36 2000 5,1 0,92 1,05 0,87 1618,98 2001 5,17 0,88 1,05 0,83 1261,18 2002 3,92 0,78 1,04 0,74 1622,94 2003 4,21 1,29 1,04 1,24 1784,96 2004 4,29 1,10 1,04 1,06 1887,93 2005 4,37 1,06 1,04 1,01 2186,13 2006 4,67 1,16 1,05 1,11 2306,23 2007 4,21 1,05 1,04 1,01 1452,95 2008 2,18 0,63 1,02 0,61 Source: Datastream 13/03-09 Bachelor Thesis © Emad Sharghbin 2009 77 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. 11.2. Ticker KFX MAERSKB DC Equity MAERSKA DC Equity CARLB DC Equity COLOB DC Equity DNORD DC Equity DCO DC Equity DANSKE DC Equity DSV DC Equity FLS DC Equity GEN DC Equity LUN DC Equity NKT DC Equity NDA DC Equity NOVOB DC Equity NZYMB DC Equity SYDB DC Equity TOP DC Equity TRYG DC Equity VWS DC Equity WDH DC Equity Market Weight Report CSE 15/04-09 Name A P Moller - Maersk A/S A P Moller - Maersk A/S Carlsberg A/S Coloplast A/S D/S Norden Danisco A/S Danske Bank A/S DSV A/S FLSmidth & Co A/S Genmab A/S H Lundbeck A/S NKT Holding A/S Nordea Bank AB Novo Nordisk A/S Novozymes A/S Sydbank A/S Topdanmark A/S TrygVesta AS Vestas Wind Systems A/S William Demant Holding % Weight Shares Last Price in the Index in the Index 11,5 2,2 28300,0 11,3 2,2 27800,0 5,6 118,9 252,5 2,8 42,4 353,0 1,4 44,6 166,0 1,5 48,9 165,0 7,3 698,8 56,3 1,7 190,2 47,2 1,7 53,2 169,5 1,9 44,9 233,5 3,7 196,9 100,8 0,5 23,7 123,3 3,1 457,7 36,0 23,1 526,5 237,3 3,9 54,3 386,0 1,1 67,5 85,3 1,8 16,7 586,0 3,5 68,0 276,5 10,2 185,2 297,5 2,6 59,0 238,8 Source: Bloomberg, 15/04-09 13:34 Bachelor Thesis © Emad Sharghbin 2009 78 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. 11.3. Lundholm’s Credit Risk Framework Based on their different ratios, industrial firms are grouped into 10 deciles. The firms are then assigned historical default probabilities according to their deciles. The historical, average sum is 5 %. (Lundholm 2007) Bachelor Thesis © Emad Sharghbin 2009 79 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. 11.4. Guide to the valuation model The following guide will hopefully make it easier to understand the model. After several months of work the model may seems simple, but the author has recognises the difficulties in understanding others’ model. The two dark blue sheets to the left are the input sheet in which the balance sheet and income statement data has been typed, which result in the third sheet, the historical statements. The fourth sheet, the regular blue WACC sheet, depicts both historical and future calculations of the WACC. The turquoise fifth and sixth sheets are also input sheets. In the Driver sheet, the shortterm and long-term drivers are typed in. In the Forecast Driver sheet, various, necessary information is typed in, but the sheet also contains financial statement. The seventh sheet, Results, is, as the name indicates, the results of both the historical and future forecasts. The final four green sheets all contain “end” information. The eight sheet, Valuation summary, summarises the valuation, and calculates the share values. This is one the most important sheet in the model. The ninth sheet, Output, contains, the ROIC tree and various other information, whereas the tenth sheet, Table, only includes a minor table of the development in ROIC and EBITA and so forth in the detailed forecast period. The final sheet, Credit risk, is based on Lundholm’s framework and calculated the credit risk of Vestas based on 5 key accounting ratios. Bachelor Thesis © Emad Sharghbin 2009 80 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. 11.5. Base Scenario – Valuation Summary Vestas - Base Case EUR Value of Operations: DCF approach Free Cash Discount Year Flow Factor 2009 (228) 2010 451 2011 493 2012 642 2013 822 2014 850 2015 1.053 2016 1.097 2017 1.271 2018 1.456 2019 1.422 2020 1.606 2021 1.738 2022 1.797 2023 1.991 Cont. Value 22.369 Cash paid out 1/1-2009 to 15/04-2009 Operating Value 0,888 0,789 0,700 0,622 0,552 0,490 0,435 0,387 0,343 0,305 0,271 0,240 0,213 0,190 0,168 0,168 16 Continuing value-% Detailed forecast period value-% Mid -Year Adjustment Factor Operating Value (Adjusted) PV of FCF (202) 356 345 399 454 417 458 424 436 444 385 386 371 341 335 3.766 9.115 41,3% 58,7% 1,098 10.006 Value of Operations: Economic Profit Economic Discount Year Profit Factor PV of EP Operating Value Excess Mkt Securities 10.006 41 2009 210 2010 108 2011 458 2012 670 2013 865 2014 935 2015 911 2016 861 2017 972 2018 1.082 2019 961 2020 1.046 2021 1.145 2022 1.087 2023 1.153 Cont. Value 10.524 Cash paid out 1/1-2009 to 15/04-2009 Present Value of Economic Profit Invested Capital (incl. goodwill) 187 85 321 416 477 458 397 333 334 330 260 251 245 206 194 1.772 6.265 2.850 Financial Investments Excess Pension Assets 26 0 0,888 0,789 0,700 0,622 0,552 0,490 0,435 0,387 0,343 0,305 0,271 0,240 0,213 0,190 0,168 0,168 Operating Value Mid -Year Adjustment Factor Operating Value (Adjusted) Value of Equity Enterprise Value Equity Value 9.115 1,098 10.006 Comparison of key ratios 10.073 (123) (162) (2) 0 0 0 0 0 (13) 9.773 Debt Capitalized Operating Leases Retirement Related Liability Preferred Stock Minority Interest Long-Term Operating Provision Restructuring Provision Future Stock Options Stock options No. shares (millions) Reduction of share capital from 1/1-2009-15/04-09 No. shares (millions) on 15/04-2009 Value per Share 185 0 185 52,77 Latest year -High Latest year -Low Value Difference - High Value Difference - Low 93,97 24,16 -43,8% 118,4% Evaluation of entry and exit multiples From: To: 2005 2008 Revenue growth (CAG) Adjusted EBITA growth (CAG) NOPLAT growth (CAG) Invested capital growth (CAG) Averages 2009 2013 2014 2018 2019 2023 2008 26,4% -275,3% -241,2% 13,9% 21,3% 22,3% 18,1% 13,7% 12,4% 7,2% 7,5% 10,0% 7,6% 4,2% 4,2% 6,3% Operating Value Adj. EBIT/Revenues 2024 10.006 Excess Mkt Securities Financial Investments 22.369 41 26 4,4% 11,7% 10,6% 8,8% Enterprise Value 10.073 22.369 Revenues/Invested Capital (pre-Goodwill) ROIC (after tax, pre-Goodwill) ROIC (after tax, including Goodwill) Average Economic Profit 7,7 48,2% 12,9% (17) 4,7 40,9% 23,9% 462 4,9 38,5% 26,9% 952 4,6 30,2% 23,4% 1.078 Revenue Adjusted EBITA NOPLAT 6.035 733 637 43.379 3.253 2.407 Cash Tax Rate WACC 6,9% 13,0% 26,0% 12,6% 26,0% 12,6% 26,0% 12,6% 1,7 13,7 15,8 0,5 6,9 9,3 = 22.9 % Growth in NOPLAT = 3.5 % mEUR RONIC in the continuing value 2.500 2.000 1.500 1.000 500 0 (500) Enterprise / Revenue Enterprise / Adjusted EBITA Enterprise / NOPLAT 41,3 % 58,7 % Development in FCF and PV(FCF) Continuing value-% 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 Free Cash Flow Detailed forecast period value-% PV of FCF mEUR Development in EP and PV(EP) 1.400 1.200 1.000 800 600 400 200 0 2009 2010 2011 2012 2013 2014 EP Bachelor Thesis © Emad Sharghbin 2009 2015 2016 2017 2018 2019 2020 2021 2022 2023 PV of EP 81 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. 11.6. Bearish Scenario – Valuation Summary Vestas - Base Case EUR Value of Operations: DCF approach Free Cash Discount Year Flow Factor 2009 (463) 2010 98 2011 60 2012 79 2013 191 2014 377 2015 479 2016 526 2017 578 2018 665 2019 652 2020 670 2021 720 2022 688 2023 787 Cont. Value 9.593 Cash paid out 1/1-2009 to 15/04-2009 Operating Value 0,888 0,789 0,700 0,622 0,552 0,490 0,435 0,387 0,343 0,305 0,271 0,240 0,213 0,190 0,168 0,168 16 Continuing value-% Detailed forecast period value-% Mid -Year Adjustment Factor Operating Value (Adjusted) PV of FCF (411) 77 42 49 106 185 209 203 198 203 176 161 154 130 132 1.615 3.230 50,0% 50,0% 1,098 3.545 Value of Operations: Economic Profit Economic Discount Year Profit Factor PV of EP Operating Value Excess Mkt Securities 2009 (101) 2010 (261) 2011 (77) 2012 (49) 2013 37 2014 143 2015 164 2016 195 2017 229 2018 258 2019 224 2020 172 2021 197 2022 139 2023 152 Cont. Value 1.011 Cash paid out 1/1-2009 to 15/04-2009 Present Value of Economic Profit Invested Capital (incl. goodwill) (90) (206) (54) (30) 20 70 71 75 79 79 61 41 42 26 26 170 380 2.850 Financial Investments Excess Pension Assets 0,888 0,789 0,700 0,622 0,552 0,490 0,435 0,387 0,343 0,305 0,271 0,240 0,213 0,190 0,168 0,168 Operating Value Mid -Year Adjustment Factor Operating Value (Adjusted) Value of Equity 3.230 1,098 3.545 Comparison of key ratios 3.545 41 26 0 Enterprise Value 3.613 (123) (162) (2) 0 0 0 0 0 (13) 3.313 Debt Capitalized Operating Leases Retirement Related Liability Preferred Stock Minority Interest Long-Term Operating Provision Restructuring Provision Future Stock Options Stock options Equity Value No. shares (millions) Reduction of share capital from 1/1-2009-15/04-09 No. shares (millions) on 15/04-2009 185 0 185 17,89 Value per Share Latest year -High Latest year -Low Value Difference - High Value Difference - Low 93,97 24,16 -81,0% -26,0% Evaluation of entry and exit multiples From: To: 2005 2008 Revenue growth (CAG) Adjusted EBITA growth (CAG) NOPLAT growth (CAG) Invested capital growth (CAG) Adj. EBIT/Revenues Averages 2009 2013 2014 2018 2019 2023 26,4% -275,3% -241,2% 13,9% 12,8% 1,4% -1,6% 10,8% 8,8% 11,3% 12,1% 6,7% 7,0% 2,6% 2,6% 5,5% Operating Value 2008 2024 3.545 Excess Mkt Securities Financial Investments 9.593 41 26 4,4% 5,5% 8,0% 6,9% Enterprise Value 3.613 9.593 Revenues/Invested Capital (pre-Goodwill) ROIC (after tax, pre-Goodwill) ROIC (after tax, including Goodwill) Average Economic Profit 7,7 48,2% 12,9% (17) 4,1 17,5% 9,9% (90) 4,1 25,5% 16,2% 198 4,0 21,6% 15,1% 177 Revenue Adjusted EBITA NOPLAT 6.035 733 637 24.837 1.490 1.153 Cash Tax Rate WACC 6,9% 13,0% 22,6% 12,6% 22,6% 12,6% 22,6% 12,6% 0,6 4,9 5,7 0,4 6,4 8,3 = 14.5 % Growth in NOPLAT = 3.5 % mEUR RONIC in the continuing value 1.000 800 600 400 200 0 (200) (400) (600) Enterprise / Revenue Enterprise / Adjusted EBITA Enterprise / NOPLAT 50,0% Development in FCF and PV(FCF) 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 50,0% Continuing value-% Detailed forecast period value-% Free Cash Flow PV of FCF Development in EP and PV(EP) 300 mEUR 200 100 0 (100) 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 (200) (300) EP Bachelor Thesis © Emad Sharghbin 2009 PV of EP 82 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. 11.7. Bullish Scenario – Valuation Summary Vestas - Base Case EUR Value of Operations: DCF approach Free Cash Discount Year Flow Factor PV of FCF 2009 (177) 2010 646 2011 758 2012 1.019 2013 1.382 2014 1.458 2015 1.926 2016 2.179 2017 2.604 2018 3.052 2019 3.016 2020 3.078 2021 2.912 2022 3.267 2023 3.010 Cont. Value 42.296 Cash paid out 1/1-2009 to 15/04-2009 Operating Value (157) 509 531 633 763 715 838 843 894 930 817 740 622 619 507 7.120 16.925 0,888 0,789 0,700 0,622 0,552 0,490 0,435 0,387 0,343 0,305 0,271 0,240 0,213 0,190 0,168 0,168 16 Continuing value-% Detailed forecast period value-% Mid -Year Adjustment Factor Operating Value (Adjusted) Value of Operations: Economic Profit Economic Discount Year Profit Factor 2009 376 2010 330 2011 799 2012 1.185 2013 1.694 2014 2.187 2015 2.090 2016 2.135 2017 2.423 2018 2.692 2019 2.635 2020 2.456 2021 2.249 2022 2.431 2023 2.121 Cont. Value 24.879 Cash paid out 1/1-2009 to 15/04-2009 Present Value of Economic Profit Invested Capital (incl. goodwill) 42,1% 57,9% 1,098 18.579 Operating Value Mid -Year Adjustment Factor Operating Value (Adjusted) Comparison of key ratios From: To: Adj. EBIT/Revenues 334 260 559 737 935 1.072 910 826 832 821 713 590 480 461 357 4.188 14.075 2.850 16.925 1,098 18.579 Operating Value Excess Mkt Securities 18.579 41 Financial Investments Excess Pension Assets 26 0 Enterprise Value 18.646 (123) (162) (2) 0 0 0 0 0 (13) 18.346 Debt Capitalized Operating Leases Retirement Related Liability Preferred Stock Minority Interest Long-Term Operating Provision Restructuring Provision Future Stock Options Stock options Equity Value No. shares (millions) Reduction of share capital from 1/1-2009-15/04-09 No. shares (millions) on 15/04-2009 185 0 185 99,06 Value per Share Latest year -High Latest year -Low Value Difference - High Value Difference - Low 93,97 24,16 5,4% 310,0% Evaluation of entry and exit multiples 2005 2008 Revenue growth (CAG) Adjusted EBITA growth (CAG) NOPLAT growth (CAG) Invested capital growth (CAG) 0,888 0,789 0,700 0,622 0,552 0,490 0,435 0,387 0,343 0,305 0,271 0,240 0,213 0,190 0,168 0,168 Value of Equity PV of EP Averages 2009 2013 2014 2018 2019 2023 2008 26,4% -275,3% -241,2% 13,9% 29,0% 34,5% 29,8% 16,6% 16,5% 11,4% 11,6% 14,2% 9,0% 0,5% 0,5% 7,9% Operating Value 2024 18.579 Excess Mkt Securities Financial Investments 42.296 41 26 4,4% 13,9% 12,8% 9,4% Enterprise Value 18.646 42.296 Revenues/Invested Capital (pre-Goodwill) ROIC (after tax, pre-Goodwill) ROIC (after tax, including Goodwill) Average Economic Profit 7,7 48,2% 12,9% (17) 5,3 54,6% 32,9% 877 5,6 52,8% 39,9% 2.305 5,1 35,5% 29,8% 2.378 Revenue Adjusted EBITA NOPLAT 6.035 733 637 75.994 6.080 4.463 Cash Tax Rate WACC 6,9% 13,0% 26,6% 12,6% 26,6% 12,6% 26,6% 12,6% 3,1 25,4 29,3 0,6 7,0 9,5 RONIC in the continuing value = 25.7 % Growth in NOPLAT = 3.5 % Enterprise / Revenue Enterprise / Adjusted EBITA Enterprise / NOPLAT 42,1% Development in FCF and PV(FCF) 4.000 57,9% mEUR 3.000 2.000 1.000 Continuing value-% 0 (1.000) 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 Free Cash Flow Detailed forecast period value-% PV of FCF Development in EP and PV(EP) 1.400 1.200 mEUR 1.000 800 600 400 200 0 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 EP Bachelor Thesis © Emad Sharghbin 2009 PV of EP 83 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. 11.8. Constantly held variables in the forecast This appendix provides an overview of all variables held constant in all the scenarios and the simulations. It refers to the sheet “Forecast Driver” in the 3 scenario models. The list is made according to how the sheet is divided. For an overview, click on the boxed “1” which is in row with the letters of the columns in Excel. This will collapse all the different sections. Clicking on the plusses will expand the sections. OPERATIONS: S,G&Adm. costs: Average of last 3 years to reflect increasing trend. R&D R &D is more and more important, and one could argue that it would receive increasing funding. The variables is held constant at 3,5 % of revenue, which is above the past funding of 2-2,5 % of revenue. Working Capital: All variables are based on the average of the last 5 years. However, Accounts receivable, Accounts payable, and other current liabilities have all grown significantly within recent years, therefore these are only based on last 3 years’ average. BALANCE SHEET ITEMS: Capex Vestas’ guidance for 2009 includes 1200 mEUR on investments in non-current assets, of which 1000 is in tangibles, both excl. dep. However, the company has said that they will lower capex depending on Q1, so capex is forecasted around 900 inclusive of depreciation. Capex for 2010 and 2013 have both been put at 3 % of revenue, while 2011-2012 is assigned 5 %. A slow down is anticipated in 2010, that will be followed by an expansion in 20112012, which will be stabilised again in 2013. Of course, in the 3 different scenarios, capex will be different, but it is outside the scope of this report to estimate exactly how much additional capex a bullish scenario necessitates and so forth. Bachelor Thesis © Emad Sharghbin 2009 84 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. Goodwill: Goodwill is expected not to be amortised, and also, Vestas is not expected to acquire any firms. Hence goodwill is kept at 0. OFF-BALANCE SHEET ITEMS: Intangibles: Additions and amortisations are both put at last 3 years average. Op. leases: Calculated b y finding CAGR for 2004-2008 then using it each year. Interest rate is same as interest rate on debt. Options: Value of Options Outstanding is put at 12,615 mEUR according to AR 2008: 82. Preferred stock: 0, as there is none in Vestas. PROVISIONS & NON-OP- P&L: Long-term debt rate: As mentioned in the section about WACC, it is estimated to be around 6,5 % forward-going. TAX: Marginal, effective tax: Vestas’ own guidance forecasts 28 % corporate tax (AR 2008: 11). Although it is possible to estimate effective tax rate based on in which countries Vestas makes its profits, the added complexity outweighs the benefits. Deferred tax: Average of last 5 years. LONG TERM & CV DRIVERS: Cash tax rate: Average of last 5 years. Closing Net PPE & Other IC: Average of last 3 years to reflect decreasing trends. Cumulative GW & Intangibles: Written up with average of historical mean in 200408, which was 45 mEUR. Bachelor Thesis © Emad Sharghbin 2009 85 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. 11.9. Price targets from leading investment banks Date Price target in DKK 23-04-2009 435 22-04-2009 200 07-04-2009 260 27-03-2009 240 25-03-2009 238 12-03-2009 223 10-03-2009 270 09-03-2009 240 26-02-2009 260 13-02-2009 525 12-02-2009 345 12-02-2009 380 10-02-2009 272 05-02-2009 425 30-01-2009 499 23-01-2009 330 22-01-2009 414 28-10-2008 250 15-10-2008 380 02-10-2008 590 20-08-2008 810 19-08-2008 750 14-08-2008 750 11-07-2008 850 11-07-2008 545 04-07-2008 850 24-06-2008 650 07-06-2008 540 21-05-2008 675 Price target in EUR 58 27 35 32 32 30 36 32 35 70 46 51 37 57 67 44 56 34 51 79 109 101 101 114 73 114 87 72 91 Bank Goldman Sachs Credit Suisse JPMorgan Standard & Poors´s Norddeutsche L. DZ Bank Deutshe Bank Morgan Stanley Society General Danske Bank ING Jyske Bank Royal Bank of Scotland Nordea Jefferies Citigroup Piper Jaffray Merrill Lynch Cheuvreux UBS Handelsbanken Alm. Brand Henton Dresdner Kleinwort Kaupthing/FIH Dansk Aktie Analyse HSBC Gudme Raaschou Carnegie ABG Sundal Collier Price targets in EUR from leading investment banks 120 100 80 60 40 20 0 Feb-08 Jun-08 Sep-08 Dec-08 Price target in EUR Bachelor Thesis © Emad Sharghbin 2009 Mar-09 Jul-09 Author's estimate 86 Error! Use the Home tab to apply Overskrift 1 to the text that you want to appear here. 11.10. Three peer group analyses: Wind, Solar, Industrials Available on the enclosed CD – Peergroup_Vestas.xls. Peer Group - Wind Energy Author's estimates EV/EBITDA EV/EBIT EV/EBIT EV/EBIT Price MV PE PE EV/EBITDA (USD) (mUSD) (2009) (2010) (2009) (2010) (2009) VESTAS WINDSYSTEMS 58,4 10.806,3 18 14 9 8 12 10 14 GAMESA CORPN.TEGC. 17,4 4.233,3 19 13 8 6 14 11 Company namc (2010) (2009) 16,5 1.299,3 17 22 10 10 12 13 121,2 1.089,9 19 17 8 7 10 7 1,3 1.983,8 8 11 6 6 7 7 Arit. average excl. Vestas 16 16 8 7 11 10 Arit. average excl. Vestas and Suzlon 19 17 9 8 12 10 MV-weighted average 16 14 8 7 12 10 Vestas premium cf. arit. average 10% -8% 17% 3% 7% -2% Vestas premium cf. arit. avg. excl. Suzlon -6% -17% 8% -3% -4% -9% 7% 1% 19% 9% -1% -3% -6% -17% 8% -3% -4% -9% Price according to MV-multiples 45 43 51 47 42 41 Price according to avg. excl. Suzlon 40 35 46 41 41 39 NORDEX REPOWER SYSTEMS SUZLON ENERGY Premium MV-weighted Premium/discount ex. Suzlon VWS price on 27/04-09: EV/EBIT (2010) 13 43 Source: Thomson Financials Datastream EV/EBITDA EV/EBIT EV/EBIT Price MV PE PE EV/EBITDA (USD) (mUSD) (2009) (2010) (2009) (2010) (2009) (2010) VESTAS WINDSYSTEMS 58,4 10.806,3 18 14 9 8 12 10 SOLARWORLD 25,6 2.864,7 14 11 6 5 8 7 3,2 37,2 3 2 2 1 3 1 SOLON 12,3 154,1 7 5 9 7 12 10 PHOENIX SOLAR 41,5 277,7 10 8 6 5 6 5 9 7 6 5 7 6 14 11 6 5 8 7 105% 118% 58% 66% 58% 67% Solar energy SUNWAYS Arithmetic average excl. Vestas MV-weighted average Vestas premium cf. arithmetic average EV/EBITDA EV/EBIT EV/EBIT Price MV PE PE EV/EBITDA (USD) (mUSD) (2009) (2010) (2009) (2010) (2009) (2010) VESTAS WINDSYSTEMS 58,4 10.806,3 18 14 9 8 12 10 GENERAL ELECTRIC 12,1 127.886,6 12 13 22 21 41 42 SIEMENS 65,1 59.531,4 10 10 6 6 8 9 ABB 'R' 15,0 34.778,9 13 16 7 8 8 10 CATERPILLAR 33,6 20.229,3 26 21 17 16 38 30 ALSTOM 62,3 17.919,6 12 11 7 6 9 8 20 Industrials 15 14 12 11 21 Arithmetic average excl. Vestas, GE, and Caterpillar 12 12 7 7 8 9 13 13 15 14 27 27 18% 0% -21% -34% -44% -52% 15% Caterpillar Vestas premium cf. arithmetic average excl. GE and47% 42% 13% 38% 8% Arithmetic average excl. Vestas MV-weighted average Vestas premium cf. arithmetic average Source: Thomson Financials Datastream. Bachelor Thesis © Emad Sharghbin 2009 87