The Airbus-Boeing Launch Aid Disputea Game-theoretic Analysis. Richard John Fairchild

Transcription

1 The irbus-oeing Launch id Disputea Game-theoretic nalysis Richard John Fairchild University of ath School of Management Working Paper Series This working paper is produced for discussion purposes only The papers are expected to be published in due course, in revised form and should not be quoted without the author s permission

2 University of ath School of Management Working Paper Series School of Management Claverton Down ath 7Y United Kingdom Tel: Fax: na Lozano-Vivas & Fotios Pasiouras 0080 GC Parry, M James-Moore, P Graves and O ltinok Glenn Parry, ndrew Graves & Mike James-Moore G C Parry, J Towler & Graves 008 The impact of non-traditional activities on the estimation of bank efficiency: international evidence Legal aspects of electronic signatures Lean New Product Introduction: a UK erospace Perspective Travel Policy: a case study to deliver cost savings to MOD business air travel Glenn Parry theoretical approach to construction project progress drawing on complexity and metaphysics Manthos D Delis, Philip Molyneux & Fotios Pasiouras Richard John Fairchild Regulations and productivity growth in banking The irbus-oeing Launch id Disputea Game-theoretic nalysis

3 The irbus-oeing Launch id Dispute- Game-theoretic nalysis Richard John Fairchild, School of Management, University of ath, UK July nd 008 bstract We present a game-theoretic approach that addresses the longstanding and acrimonious dispute between oeing and irbus over launch aid Our model considers the US and EU governments simultaneous decision to proved launch-aid Further, we demonstrate that subsidies encourage manufacturers vale-creating efforts, and that governments may have an incentive to provide subsidised finance, since governments are keener on aircraft industry success compared with ordinary investors Finally, we demonstrate that launch-aid has both a strategic competitiondistorting effect, and an innovative growth (infant industry effect

4 1 Introduction Over the last couple of decades, irbus and oeing have been engaged in an acrimonious trade dispute over governmental provision of subsidised funding ( launch aid Specifically, oeing has claimed that the EU governments have provided subsidies to assist irbus in the production of the 380, unfairly distorting competition in favour of the European consortium The US company were so upset with this situation that they filed a complaint with the WTO irbus responded by filing its own complaint, that oeing had also received massive subsidies, particularly in producing the oeing 787 (the main rival to the Industrial economists are becoming increasingly interested in this case, since it allows a practical, real-world analysis of dynamic product market competition and innovation under duopoly This provides a basis for theoretical modelling and empirical testing For instance, Esty and Ghemawat (00 employ a Hotelling model of duopoly competition in order to consider irbus and oeing s entry and product differentiation decisions relating to production of their superjumbos Irwin and Pavcnik (004 perform an empirical analysis of the rivals behaviour, informed by game-theoretic considerations Galy and Germain (007 consider the effect of the rivals product market announcements on their relative stock prices However, none of these authors consider the specific debate over launch-aid subsidies Kaivanto (1997, 005 and Kaivanto and Stoneman (007 have conducted extensive research into the launch-aid dispute In his 005 paper, he provides three useful 1 The irbus-oeing dispute is long-standing and complex For excellent and comprehensive discussions see Carbaugh and Olienyk (001; 004 3

5 characterisations of launch aid, whereby it may be considered as an instrument of a infant industry policy, b technology policy, or c strategic trade policy Motivated by his claim that surprisingly little independent economic analysis has informed this debate, we present a game-theoretic model of the launch-aid dispute Kaivanto (005 notes the complex nature of the financial contract involved in the provision of launch-aid to irbus to produce the 380 This contract combines features of debt, equity and state-contingent financial instruments Specifically, irbus is expected to repay the funding to the government on an increasing scale as it makes sales of the 380 (an equity-like, sales-contingent claim Furthermore, given the sales forecasts, the government will not recoup their money for 17 years If irbus does not make sufficient sales, the government will not recoup their investment (therefore, risk has effectively been shifted to the government, who have provided risky debt; irbus has a risk-free position Finally, the launch-aid is subsidised, as the government have not taken into account inflation or the time value of money when calculating the break-even position (cheap loans at below market rates Kaivanto and Stoneman (007 make two contributions to the debate First, they provide a working definition of subsidised launch aid ( in essence, they argue that if the sum paid by the government for the contract exceeds the market value, then there is a subsidy; if not there is no subsidy Secondly, they consider the complex form of launch aid as addressing market incompleteness, so that launch aid provides an efficient risk-sharing form of finance for irbus, not available through the financial markets This quote comes from the online discussion of Kaivanto and Stoneman s (007 paper (see Launch id to the erospace Industry: New Criterion for ssessing Whether it is a Subsidy 4

6 Our game-theoretic model contributes to the launch-aid dispute in the following important ways First, we formalise the rival governments simultaneous decisions to offer launch aid Second, we consider why governments may choose to offer subsidies Third, in a major departure from existing analysis, we demonstrate that the government needs to consider the effect of the financial contract on the firm s valuecreating incentives (Kaivanto, for example, only considers the risk-sharing properties of launch-aid Fourth, we contribute to the debate on launch aid subsidies by considering the value-reducing and value-enhancing aspects Indeed, motivated by Kaivanto s (005 classifications of launch-aid (infant industry, technology, and strategic trade policy, we consider two effects of launch-aid; a a competitiondistorting effect (Kaivanto s strategic trade policy effect, and b an innovative expansion effect (Kaivanto s infant industry effect Finally, we consider (as future research behavioural game-theoretic extensions to our model, not considered by previous researchers Overall, our analysis demonstrates that policy-makers should not be quick to criticise launch-aid without considering the advantages, as well as the disadvantages oeing may have a valid claim that launch-aid distorts competition However, launch aid may also add value by providing funding for innovation Furthermore, government funding may be necessary, if such funding is not available privately Even if such funding is available privately, the fact that governments may offer it more cheaply may enhance producers innovative, value-adding efforts Our analysis also demonstrates that, in a free-market, oeing could choose to compete with irbus by also offering launch-aid, rather than filing a complaint at the WTO 5

7 The rest of the paper is organised as follows In section, we present the model In section 3, we compare launch aid with finance from the stock market We discuss policy implications in section 4 Section 5 concludes with suggestions for future research The Model We consider a launch aid/project investment game consisting of two firms i {, } (where represents irbus and represents oeing, and two governments j {, } ll agents are risk-neutral, and the risk-free rate is zero The timeline of the game is as follows Each firm has existing operations-in-place with present value V > 0 t date 0, each firm now has a potential new project available (the irbus 380 and the oeing Jumbo, requiring investment funds I > 0 The firms have insufficient internal funds to meet this requirement Furthermore, they cannot obtain finance from the capital market (we analyse this in more detail in section 3 Each firm can, however, approach its own government for finance (we term government funding launch aid The governments simultaneously decide whether to provide launch-aid If firm i receives launch aid, it invests in date 1 project development (R and D With probability q [0,1], firm i' s R and D succeeds, in which case full scale-production i can go ahead at date With complementary probability, firm i' s R and D fails, in 6

8 which case, full scale production must be abandoned We consider the case where q > q t date 15, if R and D has succeeded, such that full scale-production can commence, the government offers to the firm an equity stake 3 α [0,1] in the project, with the government holding the balance of project equity 1 α [0,1] Note that we assume that the government can only obtain an equity stake in the project, and not in the cashflows V from existing operations t the date production stage, each firm s management exerts value-creating effort e This managerial effort affects success probability, as follows With probability p = γe, the project provides date 3 income g > 0, while with probability 1 p, the project provides date 3 zero income Note thatγ represents managerial ability Therefore, the expected value of the project is pg = γeg If only one firm invests in the new project, g = G If both firms invest in the new project, g = G δ Each firm faces a cost of effort C( e = βe t date 3, payoffs occur and the game ends We consider two effects resulting from the governments simultaneous provision of launch aid Firstly, we consider a competition-distorting effect, whereby the provision 3 Kaivanto (005, 007 identifies the complex nature of launch-aid financial contracts, whereby the government holds a sales-contingent stake in the project (with the government s share of revenues increasing with sales of the 380 In this paper, we do not consider this complexity, but focus on an equity contract However, we argue that this sales-contingent contract is equity-like, and therefore faces similar incentive problems to the one that we consider in our model We leave full analysis of the real-world contract for future research 7

9 of launch-aid affects each firm s present value of existing operations as follows If only one firm (say firm receives launch-aid, N = V Δ, N = V Δ If both firms receive, or if neither firm receives, launch aid, N = N V = The second effect is that launch-aid allows innovation and growth (that is, investment in the new projects that is not possible under market finance We solve the game by backward induction 1: Date : Firms Effort in the Growth Option We take as given that a firm has obtained finance I > 0 for the new project, that R and D has been successful, and that the government has offered to the firm an equity stake α [0,1] in the project We solve for the firm s optimal effort level in the project as follows 4 The firm s expected payoff from the project is = αγeg βe (1 i i Solving = 0, e we obtain the firm s optimal effort level as follows 4 t this stage, we do not consider whether both firms or only one firm, has received launch-aid, and been successful in R and D We keep the situation general, by working with the expected value g, which is affected by whether only one firm, or both firms have successful R and D 8

10 Lemma 1: If a government provides launch aid, and the firm s R and D is successful, the firm s optimal effort level during full-scale production is αγg e * = ( β Therefore, the expected value of the growth option is αγ g V = γe * g = (3 β : Date 15: The government s equity offer Given that R and D has been successful, the government makes its equity offer α to maximise its expected payoff; αγ g G = (1 α j γeg = (1 α j, (4 β where we have substituted for the firm s optimal effort level from ( 9

11 The parameter j represents government j' s private benefit from the success of the growth option We assume that > Parameter j provides an interesting angle in this model It differentiates the government from a standard equity investor, who is only interested in the cashflows from an investment The idea behind is that the government is interested in the success of the project because it provides prestige to the country, making the government look good Furthermore, the government s tenure may be tied up in the success of the project (due to the public voting at elections G The government chooses α to maximise equation (3 Solving = 0, α we obtain the government s optimal equity offer as follows: Lemma : If R and D is successful, the government s optimal equity offer at the full scale-production phase is 1 j α = (5 Note that the government s equity offer to the firm is increasing in the government s private benefits This provides an insight into the policy debate over subsidised launch-aid Our analysis suggests that the government is prepared to provide subsidised launch-aid (that is, a higher equity stake than private investors for whom = 0, due to the private benefits it receives from the success of the project That is, 10

12 since the government is interested in the prestige of a successful project, it is prepared to provide more incentives to the firm s management in the form of higher equity Under this argument, subsidised finance is beneficial Substituting the optimal equity offer into equation (4, we obtain the government s indirect payoff from full scale-production; (1 j 4 j γ g G = (6 8β Furthermore, substituting the optimal equity offer into ( and then into (1, we obtain the firm s indirect payoff from full scale-production; α γ g (1 j γ g i = = (7 4β 16β 3 Date 1 Government s Launch aid decision Finally, we solve for the governments simultaneous launch aid decision We solve the following normal for game: Govt \ No Launch aid Launch aid No Launch id 8, 9 10, 11 Launch id 1, 13 14, 15 11

13 These payoffs are as follows: G V = (8 G V = (9 ( Δ = V G ( (1 ( I G q V G Δ = β γ ( (1 ( I G q V G Δ = β γ (1 ( Δ = V G (13 8 ( 4 (1 I G q V G = β δ γ (14 8 ( 4 (1 I G q V Gb = β δ γ (15 1

14 We solve for the governments equilibrium decisions by considering the best responses First, consider firm s best responses If government chooses no launch aid, government will choose launch aid if (1 (8; that is, if (1 I G q Δ β γ If government chooses launch id, government will also choose launch aid if (14 (10; that is, if 8 ( 4 (1 I G q Δ β δ γ Next, consider government s best responses If government chooses no launch aid, government will choose launch aid if (11 (9; that is, if (1 I G q Δ β γ If government chooses launch id, government will also choose launch aid if (15 (13; that is, if 8 ( 4 (1 I G q Δ β δ γ 13

15 In order to focus our analysis, we consider the case where (1 4 γ ( G δ Δ q I Therefore, government will provide 8β launch aid, regardless of government s choice That is, government s dominant strategy is to provide launch aid Next, we consider the effect of the governments launch-aid decisions on combined firm values If neither firm receives launch-aid, each firm s expected profit is V If both firm s receive launch-aid, each firm s profit is i = V q i α γ ( G δ 4β = V q i (1 i γ ( G δ 16β If only firm receives launch-aid, the expected profits are (1 γ 4G = V Δ q, = V Δ 16β Therefore; (1 4 γ ( G δ Proposition 1: Given that Δ q I, such that 8β government s dominant strategy is to provide launch aid, the equilibrium of the launch aid game is affected by and q as follows; 14

16 (1 4 γ ( G δ a If Δ q I, both governments provide launch- 8β aid Therefore, = V q (1 γ ( G δ 16β q (1 γ ( G δ 16β (1 4 γ ( G δ b If Δ q < I, only government provides 8β launch-aid Therefore, = V q (1 γ 4G 16β This proposition provides an insight into the irbus-oeing dispute oeing has complained to the WTO that the European governments have provided an unfair competitive advantage to irbus by providing launch-aid However, in a free-market, both firms are entitled to provide launch-aid to their respective industries Proposition 1 suggests that each government s decision to provide launch-aid is driven by i the private benefits it obtains from the development of its industry, ii the government s assessment of the success probability of the project s innovation, and iii the effect of competition In particular, irbus may be more prepared than oeing to provide launch aid due to higher private benefits from the success of irbus; is high (the US government may have other priorities; is low, and higher assessment of the success probability of airbus ( q is high, while q is low The effect is further strengthened by the competition effect ( G δ, whereby the market may only support one firm in equilibrium 15

17 Finally, we note that combined firm values are maximised when both governments provide launch-aid Furthermore, firm values are increasing as i increases, since governments will increase the amount of equity provided to firms, in order to increase effort incentives Hence, rather than launch-aid being bad, it provides growth The competition effect merely shifts profits between the firms 3 Finance from the Stock Market Thus far, we have focussed on launch-aid from the government Now, we consider stock market finance However, we focus on the case where γ 4G q SM < I, 8β (16 Where q SM represents the stock market s assessment of the success probability of R and D Note that, for the stock market, = 0 ccording to (16, neither firm will be able to obtain finance from the stock market The stock market is not prepared to provide finance because a it has a low assessment of the success probability of R and D, and b private investors do not obtain private benefits from the success of the project i 16

18 Therefore in the absence of launch-aid, combined firm values are V Therefore, launch-aid is value-adding This contributes to the dispute over subsidised launchaid 4 Policy implications dispute exists at the WTO that firms should not receive subsidised launch aid, as it distorts competition subsidy has been defined by the WTO as a form of financing that is cheaper than that in the general stock market Our model addresses this dispute as follows; a Launch-aid has a negative aspect (competition distortion, and a positive aspect (allowing growth b Launch-aid may be the only method of financing growth available, since governments are interested in helping the firm grow, and furthermore, governments may be more far-sighted than myopic stock markets c Even in the case where the stock-market provides finance, launch-aid may enhance value-creating incentives, as governments give more equity to firms, due to the governments private benefits from growth d In a free market, one government should not complain that the other is providing launch aid The other government is perfectly entitled to provide launch-aid to its own industry However, it may be the case that the other government chooses not to, for the reasons identified previously 17

19 5 Conclusion and Further Research We have presented a game-theoretic approach to the irbus-oeing launch-aid dispute Our model provides the basis for the following future research Firstly, we should develop the model to incorporate the WTO dispute This may naturally lead to the inclusion of a governmental bargaining stage into our analysis (for example, the governments may bargain with each other over the provision of subsidies, while threatening to take the case to WTO if negotiations break down Secondly, we may develop a behavioural game-theoretic approach, incorporating incomplete contracts, bounded rationality, and regret as follows irbus claim that oeing are also providing subtle, indirect subsidies However, due to incomplete contracts, this may be difficult to prove in law This encourages oeing to continue to provide indirect subsidies However, oeing may misdirect their innovative resources, such that R and D fails This may lead to regret, anger, and spite, escalating the dispute at the WTO Thus far, we have considered possible negative factors resulting from a behavioural approach to the irbus-oeing dispute We finish on a positive note, again using behavioural game-theory We could use Sally s (001 empathy game to demonstrate the world benefits of the US and EU governments putting their differences behind them, and acting together to promote international innovation and economic growth 18

20 References: Carbaugh, R, and J Olienyk (004 oeing-irbus Dispute: Sequel Global Economy Journal Esty,, and P Ghemawat (00 irbus vs oeing in Super Jumbos: Case of Failed Preemption Working Paper Number: Galy, N, and L Germain Market Competition and Share Prices: The Case of the irbus-oeing Duopoly Working Paper Irwin, D, and N Pavcnik irbus Versus oeing Revisited: International Competition in the ircraft Market Journal of International Economics 64, 3-45 Kaivanto, K (1997 UK Launch id and Experience Working paper Kaivanto, K (006 Premise and Practice of UK Launch id The Journal of World Trade, Volume 40, Issue 3 Kaivanto, K, and P Stoneman (007 Risk Shifting, Technology Policy and Sales Contingent Claims: When is Launch id to the erospace Industry a Subsidy? Working Paper 19

21 Sally, D (001 On Sympathy and Games Journal of Economic ehavior and Organisation Vol 44,