Optimal Timing for Introducing a New High-Tech Product: A Real Options Approach

Optimal timing for introducing a new high-tech product: A Real Options approach Hugo Manuel Couto Fernandes Master in Finance Dissertation Supervisor Prof. Paulo Jorge Marques de Oliveira Ribeiro Pereira 2013 Nota Biogr´afica Hugo Manuel Couto Fernandes nasceu em Vila Nova de Famalic˜ao em 1988. Licenciouse em Economia em 2010, na Faculdade de Economia do Porto. Ap´os o final da licenciatura ingressou na empresa TLCI enquanto coordenador do departamento de gest˜ao de receitas. Em 2011 iniciou o Master in Finance na Faculdade de Economia do Porto tendo terminado a parte curricular do Mestrado em Finan¸cas em 2013, cuja disserta¸c˜ao se encontra actualmente sob proposta. Durante o mestrado co-fundou o FEP Finance Club, tendo sido director do mesmo durante os seus dois primeiros anos de actividade. Ainda em 2013 ingressou na SONAE SGPS na ´area de Planning and Management Control, onde se encontra desde ent˜ao. 3 Agradecimentos A elabora¸c˜ao desta disserta¸c˜ao n˜ao teria sido poss´ıvel sem a colabora¸c˜ao e apoio de algumas pessoas, a quem gostaria de agradecer. Em primeiro lugar, ao Professor Doutor Paulo Jorge Marques de Oliveira Ribeiro Pereira, pela sua mais que orienta¸c˜ao na elabora¸c˜ao deste trabalho e pela disponibilidade sempre demonstrada. Agrade¸co-lhe o apoio, dedica¸c˜ao e o facto de ter excedido as suas fun¸c˜oes de orientador e ter sido um verdadeiro parceiro ao longo deste trabalho. Sem ele qual n˜ao teria conseguido concluir este projecto. Segundo, `a Vˆania Campos, por ter sido uma mentora ao longo de toda a disserta¸c˜ao, por me ter feito acreditar que conseguiria sempre alcan¸car os meus objectivos e por saber, de forma misteriosa para mim, sempre o que fazer e o que dizer no momento certo. Agrade¸co-lhe ser o motor que me auxilia a conduzir a minha vida. Gostaria de deixar uma men¸c˜ao especial `a minha fam`ılia, principalmente aos meus pais, pelo seu amor incondicional e por lhes dever tudo, sem nada nunca me ser cobrado. Para al´em de me terem dado a vida, deram-me tamb´em forma¸c˜ao nos meus primeiros anos de vida que foi essencial ao longo do meu crescimento, sendo os meus conselheiros ainda hoje. Agrade¸co ainda terem-me sempre fornecido todas as ferramentas de que preciso, e todas aquelas que nunca precisei, em detrimento do seu pr´orio bem estar. Estou profundamente convicto de que n˜ao poderia ter tido mais sorte ao ter nascido em outro lugar. Tenho, naturalmente, de agradecer profundamente aos meus companheiros de viagem, Iolanda Filipa Gomes e Vasco Ferreira, por me terem acompanhado e guiado durante estes ´ultimos dois anos. Foram fulcrais no meu crescimento e fizeram-me acreditar que tudo ´e poss´ıvel. Ficarei eternamente grato por todo o seu apoio e por estarem sempre dispon´ıveis para mim. N˜ao podia deixar de agradecer `a B´arbara Barros, `a Catarina Fernandes, ao David Ferreira, ao Nuno Peixoto, ao Pedro Oliveira e ao Rui Freitas por sempre me terem feito perceber que os problemas s˜ao relativos e me terem ajudado a descontrair e a clarificar as ideias, e `a Ana Couto e ao Andr´e Santos por sempre terem mostrado interesse neste trabalho. Uma nota de apre¸co `a Faculdade de Economia do Porto, onde decorreu esta parte da minha forma¸c˜ao acad´emica, por todos os meios colocados ao dispor dos alunos para que consigam alcan¸car os seus objectivos, e particularmente a todos os professores bem como a todos os meus colegas. Por ´ultimo resta-me agradecer ao Ricardo Carvalho, por ter sido sempre flex´ıvel e compreensivo comigo e por se revelar um verdadeiro l´ıder. 4 Contents Nota Biogr´afica 3 Agradecimentos 4 Sum´ario 7 Abstract 8 1 Introduction 10 2 Literature Review 16 3 The decision to launch a new generation model under a monopolistic setting 23 4 The decision to launch a new generation product under competition 30 5 The case of the Sony Playstation video-games business 37 5.1 PS1toPS2 ................................ 40 5.2 PS2 to PS3 after the Xbox 360 . . . . . . . . . . . . . . . . . . . . . 43 5.3 PS2 to PS3 before the Xbox 360 . . . . . . . . . . . . . . . . . . . . . 45 5.4 Comparative statics and sensitive analysis . . . . . . . . . . . . . . . 47 6 Conclusions 49 5 List of Figures 1.1 Playstation Shipments (millions units) . . . . . . . . . . . . . . . . . 13 1.2 Estimated iPhone Sales History . . . . . . . . . . . . . . . . . . . . . 14 5.1 Sony Playstation sales per fiscal year . . . . . . . . . . . . . . . . . . 38 5.2 Sony Playstation total sales per unit . . . . . . . . . . . . . . . . . . 39 5.3 PS1 to PS2 case parameters . . . . . . . . . . . . . . . . . . . . . . . 41 5.4 The output values for the PS1 to PS2 case . . . . . . . . . . . . . . . 42 5.5 Old product and new one relationship (no competition) . . . . . . . . 43 5.6 PS2 to PS3 with competitor case parameters . . . . . . . . . . . . . . 44 5.7 The output values for the PS1 to PS2 case with competition . . . . . 44 5.8 Old product and new one relationship (with competition) . . . . . . . 45 5.9 Old product and new one relationship (hidden competition) . . . . . 47 5.10 The value of the triggers with different values on the parameters . . . 47 6 Sum´ario A ind´ustria tecnol´ogica ´e um mercado de grande crescimento com algumas particularidades, uma vez que cont´em alta incerteza, ´e composto por uma competi¸c˜ao agressiva com a possibilidade de entrada de competidores desconhecidos e, por vezes, duas tecnologias diferentes podem estar a competir por um mercado sendo uma substituta da outra. O objectivo desta disserta¸c˜ao ´e determinar o tempo ´optimo para o lan¸camento de uma nova gera¸c˜ao de produtos. Apesar dos modelos existentes, em Op¸c˜oes Reais sobre o tempo ´optimo para a entrada num mercado, ser aplic´avel nalguns mercados, existem outros que s˜ao deixados de fora por estes modelos, j´a que a maioria destes trabalhos assumem que o valor dos seus projetos aumenta sem se considerar o ciclo de vida do produto (o que n˜ao acontece em todas as situa¸c˜oes). Nalguns casos, como a ind´ustria tecnol´ogica, h´a sempre alguns produtos a serem lan¸cados e o valor dos produtos mais velhos come¸ca a verificar uma invers˜ao na sua tendˆencia para crescerem. Dado que a venda dos produtos mais velhos come¸ca a verificar uma descida depois dos novos estarem no mercado, acreditamos que os cash-flows ir˜ao crescer at´e um ponto, a partir do qual come¸cm a decrescer. A inova¸c˜ao trazida por esta disserta¸c˜ao relaciona-se com a introdu¸c˜ao de um modelo em que os cash-flows crescem de forma diferente do ciclo de vida dos produtos, at´e que a empresa lance um novo produto. Quando isto acontece, os cash-flows do produto antigo crescem a uma taxa diferente da anterior (eventualmente uma negativa) e os cash-flows do novo produto podem crescer a uma taxa diferente da do primeiro produto, uma vez que o mercado ´e agora mais maduro e tem uma procura maior. 7 Abstract The technological industry is a high growth market of great value with some particularities as it is a market with high uncertainty, has an aggressive competition with hidden competitors probability of entrance, and sometimes two different technologies may be competing for one market, as one is a substitute of the other. The main purpose of this dissertation is to determine the optimal time to launch a new generation product. Although the existing work about optimal timing to enter a market, in Real Options, may be applicable to some markets, there are others that are left outside of these models, as most of these researches assume that the value of their projects will increase without considering the product life-cycle, which does not happen in every situation. In some cases, such as the high-tech industry, there are always new products emerging, and the older products’ value starts to see an inversion on its tendency to grow. Given that the product being sold by the company will see its sales starting to decrease after a new product enters the market, we believe that the cash-flows will increase only until a certain point, after which they will start to decrease. The innovation brought by this dissertation is met by the introduction of a model where the cash-flows grow without considering the product life-cycle until the company releases a new product. When this happens the cash-flows of the old product will grow at a different rate from before (eventually a negative one) and the cash-flows of the new product may evolve at a different rate than those of the first product as the market is now more mature and has a higher demand. 8 Chapter 2 Literature Review Having come a long way since they were firstly coined by Myers (1977), the Real Options theory is now seeing a sustainable growth as an area of academic research and as a tool for business and investment analysis. Nowadays the real options theory has become one of the most important fields in modern finances and its usage is expanding between managers and analysts. Its usefulness also surpasses the companies and project’s evaluation and it is being used as an auxiliary in determining investment strategies, appearing often related with the Games Theory, such as Fudenberg and Tirole (1985) study which analyzes the effects of preemption in games of timing, studying the adoption of a new technology. Some research carried in this area is based on the work of Dixit and Pindyck (1994). In this book the authors develop a model based on the stochastic value of the project evolving randomly, and derive a model where there is a leaderfollower competition. Huisman and Kort (1999) study how the investment under uncertainty combines with the investment under competition. When there is uncertainty the option value of waiting has to be considered by the company’s investment timing decision, because when the company is investing it is also giving up the possibility of doing it later, with more information available. On the other hand, if the company does not invest quickly its competitors may do it, and the company may loose all of the project’s value, or a part of it. Nielsen (2002) expanded the knowledge in this area to situations where there is imperfect competition, with investments having positive externalities and situations in which a monopolist has multiple investment opportunities. Tsekrekos (2003) explains the importance of investment strategies in Real op16 CHAPTER 2. - 17 tions. Real Options, as opposed to financial options, may be held by a few investors at the same time and the exercising of one of the options affects the value of the rest of them. This feature justifies an approach that contemplates all of the option holders. The author builds a model where there is a big threat of preemption and being first in the market may have advantages. These advantages are suitable of being, in part or as a whole, permanent. Thus this author tries to approach the investment strategies combined with Real Options, in an attempt to find an equilibrium between the trade-off of exercising early and gaining the advantages of being a first-mover in the market or holding the option and waiting to see how the market evolves, thus decreasing the uncertainty. Tsekrekos (2003) introduces a concept that may be applicable in a serious number of different markets. In the high-tech industry it is common to see the first mover gaining an advantage that is held throughout the whole lifetime of a generation and, sometimes, even on the transition to the next generation. Paxson and Pinto (2003) approach the theme in a similar way to Tsekrekos (2003) but introduces an innovation: situations where the market share of a company may change due to the increase or decrease of consumers, just as happens in a birth/death process with a Poisson distribution. The conclusion the authors have is that the value function of the Follower is less sensitive to market share changes and to the ratio arrivals/departures than the Leader’s, until the cash-flows exceed the Follower’s trigger. This trigger’s value grows with the market share owned by the Leader, the ratio arrivals/departures and the cash-flows volatility. The Leader’s value function is, still, unstable as the cash-flows get closer to the Follower’s trigger. Armada et al. (2011) continue the work of the previous authors, by considering the hypothesis of temporary or permanent advantages, but adds an innovation: in this case the model has the possibility of hidden competitors, which has a great impact on the decision to invest. In the high-tech industry there is a lot of different areas, and although some of the companies may sell certain types of goods but not others, it is common to see a company entering a new marketing and becoming a previously-hidden competitor (just as happened when Google entered the market of operating systems by introducing Android, or Microsoft entered the tablet hardware production with the Surface). All of this models have in common the usage of a value of the project (or the company) evolving accordingly to a Geometric Brownian Motion. Bollen (1999) CHAPTER 2. - 18 studies an option valuation framework where demand is governed by a product life cycle as opposed to using an expected growth of the value of the project. This is a different approach and very useful when valuing the introduction of goods, instead of entering a certain project. Rodrigues (2009) develops a model where demand can only grow up to a limit, which is determined by segment size and it can behave differently before and after the investment in each segment. So, the company may find it optimal to expand to new segments right after entering a previous segment or postpone it until the demand reaches the trigger. Another work worth mentioning in this field of the Real Options is Ruiz-Aliseda and Wu (2012). These authors develop a model with the entry and exit decisions of companies that act in markets whose demand goes from growth to decline phases at uncertain times, introducing a process that captures this random evolution. Daming et al. (2013) develop a real option game model to describe the impact of technological innovation on the market. These authors distinguish between incremental innovation (“sustainable or evolutionary innovation”) and radical innovation (“sustainable or evolutionary innovation”), basing on Altshuller (1999) definition, and focus on the later, using a Poisson to describe the impact of this technological innovation on the market. The results of this is two Nash equilibriums whose critical points depends on the market response (the radical technological innovation capacity is greater when companies have a moderate market response to an enterprise product). Several authors contributed to the definition of innovation, not only the meaning of the word, but also the concept behind it. One of the most cited is Schumpeter (1934), for it is believed that he was the first to propose an explicit definition for the word. Reinganum (1989) studies the timing of innovation and examines combination of the expected benefits, the cost of research and development (R&D) and interactions among competing firms. This author analysis the incentives for individual companies to invest in R&D extent based on the rivalry and appropriability interactions between the companies acting in the market, as well as the effects of licensing (primarily fixed-fee contracts) the technology discovered and the diffusion of innovation over time. On the research about innovation acceptance, Chaney et al. (1991) study how the introduction of a new product affects the market value of the companies, using a market-based analysis, concluding that there is a small increase over a short period of time, but the impact of an introduction varies negatively with the magnitude of CHAPTER 2. - 19 the firm’s systematic risk and negatively with the number of announcements made over the decade. Also the market reaction to the announcement seems to be related to the number of products in the announcement (multiple products seem to induce a larger effect) and the degree of innovation brought by them. Chou and Yang (2011) examine the interaction between the strategic orientation of high-tech companies, namely their innovation orientation, two types of market orientations (responsive and proactive), and new product performance. They conclude that new product performance, when derived from the interaction between innovation orientation and responsive market orientation show that the interaction effect is contributory to firm performance until an optimal level is reached, and then the effect becomes adverse thereafter. On the other hand, when new product performance is derived from innovation orientation and proactive market orientation the interaction effect is detrimental to firm performance until a threshold level is reached and then the effect becomes contributory thereafter. The introduction of a new product must follow some strategies because, as said before: in this business the market has high uncertainty and growth, aggressive competition with the probability of entrance of hidden competitors, sometimes two high-tech industries have two different technologies that are competing for one market, and one is a substitute of the other. As Zahra and Bogner said: “Technology, the sum of a firm’s knowledge and skills, determine the ability of new ventures to offer the products (services), gain market acceptance, survive, and achieve financial success”(Zahra and Bogner 2000, p.136). If the company releases its product too soon it will affect the sales of the older product and will allow the competition to either replicate it at a lower price or have a more technologically-advanced product. If the company releases the product too late it will allow the competition to enter the market and affect the company’s reputation as well as secure the advantages of being a first mover. Takeuchi and Nonaka (1986) address the importance of strategies when releasing a new product. Due to the fast-pacing changes in consumers habits in a competitive world of commercial new product development, companies are emphasizing the new products as a source of new sales and profits, as the old products represent less and less of the market share. According to this author more and more consumers tend to choose the new products, instead of the old ones, in contrast to what happened in the past. The degree of adoption of a new product will depend on the consumers’ perceived value of it as well as its cost. If the consumers find it to have a big improvement of CHAPTER 2. - 20 value over its predecessor they will buy it, otherwise they will buy the old product, taking advantage of the possible price drop, or not purchase any. There is also to consider the complementarity between that product and other products that the consumer may have, as the tendency to adopt the new product increases with its compatibility with other products already held by the consumer. The case of complementary technology in a world of uncertainty is studied by Azevedo and Paxson (2009). These authors find analytical or quasi-analytical solutions for the leader and the follower value functions and their respective investment thresholds, in a world of uncertainty, competition and technological complementarity. In their model, the investment game firms have two technologies available, whose functions complement each other, and the option to adopt both technologies at the same time or at different times, in a context where the evolution of their gains is uncertain. Their results show that given the uncertainty about revenues and the price of the two technologies, it might be optimal for the leader and the follower to adopt the two technologies asynchronously (first, the technology whose price is decreasing at a lower rate and then the technology whose price is decreasing more rapidly). It is common to see some complementarity between different products in this industry. It is the case of the media holders and media readers, the computer and modems, the game consoles and joysticks. It is even common to see complementarity between generations of products: Nintendo’s Wii games and accessories are compatible with its new console (Nintendo Wii U) and Apple iPhone 4 cases and cables were compatible with the Apple iPhone 4S. Rubera et al. (2012) identify the key dynamics that contribute to successful new product rollouts (significant product release, often accompanied by a strong marketing campaign to generate a large amount of consumer hype1) within a region: the type of innovation and the cultural and economic factors. A longer regional rollout strategy is more desirable for technological innovations because new technology may have a steep learning curve and may need some infrastructures (such as an increase of the speed of the internet). The risk avoidance culture of a country is also important as a longer regional rollout strategy for technological innovations achieve greater market share in countries with high uncertainty avoidance. On the other hand a short rollout may benefit the company when it is thought that the product will provide, or at least maintain, high levels of status within an individual’s peer group and in markets with higher economic openness. As economic openness will make it easier for market entrance it will also increase competition as easier market 1http://www.investopedia.com/terms/r/rollout.asp CHAPTER 2. - 21 entry applies to all other new entrants, so the company will benefit from gaining the pioneer advantages. The perceived economic value of high-tech consumer products is studied by Munnukka and J¨arvi (2011). More precisely these authors study the influence of customer-value hierarchy factors on consumers value perceptions of high-tech products. This study is important to understand the degree of adoption of a new product over the old one, which may determine how the market is split between the two products. They identify five categories of factors that could influence the perceived value of a high-tech product, and test each one of them as hypothesis: 1. It’s ability to fulfill the costumer’s latent subjective goals (such as ease of mind, increase in self-esteem, enjoyable and problem-free use experience, and overall effectiveness); 2. The benefits perceived by the costumer (no hassle, reliability, ease of use, time saving, desired visual view, independence of time, availability, high quality/cost ratio, and so on); 3. The product’s direct attributes (such as layout of instruments, size, smooth shift, design, frequency of repair, low price, product quality, display, keyboard); 4. Buyers needs and the opinion of others (how the product is perceived by others in the buyers social network affects the perceived economic value of the purchased high-tech product); 5. The extent to which the purchased high-tech product is in use in the buyers social network. These authors conclude that the first hypothesis obtained support but the result suggests a weak model fit. The second and third category have important factors (the direct attributes is the most important one). The forth category is only partly supported and taking into consideration others opinions was found to reduce the product’s perceived value. The fifth category seems to be fully supported with the extent to which the product had diffused into the respondents social network having a positive effect on the consumer’s perception of economic value. This discrepancy (between category four and five) might be explained by consumers tendency to rationalize their buying decisions through personal needs and wants, while social factors operate in the background. CHAPTER 2. - 22 Lee et al. (2012) also conduct a research on the factors that influence the consumer’s perception of the product’s attributes to be an object of interest, which will influence the diffusion of new high-tech products in the market. These authors examine the influence of hierarchical product attributes on actualized innovativeness, and base their study on the relationship between product attributes and actualized innovativeness and in identifying hierarchical patterns in the relationship. The conclusion of the study is that the unexpected and non-essential (tertiary) attributes have a stronger impact than the expected but non-essential (secondary) attributes and the expected and essential (primary) ones. In our model we consider that the product’s cash-flows evolve stochastically according to a Geometric Brownian Motion mixed with a life-cycle (the cash-flows increase until a certain value and then start to decrease). Also, there is hidden competition, and when a competitor enters the market the company may not loose half of its cash-flows (sharing the market in equal parts). After exercising the option the company will have two products sharing the market in different parts, whose value will be dependent on the perceived value of the new product as well as its capability to have a successful rollout. After seeing how the existing work influenced our model we will proceed to our model. Chapter 3 The decision to launch a new generation model under a monopolistic setting In this section we will introduce a model where the company has the control of the technology leap, being able to determine when the next generation starts and, therefore, the competition (or the threat of it) does not exist. This is the case when the product is a new-born good, and the company launching it has the know-how to do it and has the lead on research and development. This often occurs when a new good is released, which creates a new market, and it takes some time before competitors arise. Assume that the company has one product in a market that meets the assumptions mentioned above, and receives a certain amount of cash-flows which we will denominate as x. This xevolves following a Geometric Brownian Motion: dx =αxdt +σxdz (3.1) where αrepresents the risk-neutral expected growth rate of the cash-flows. σis the volatility and, for convenience, we assume that the volatility of the cash-flows is the same independently from the generation available in the market. Finally dz is the increment of a standard Wiener process. As the equation shows, one may know the value of the current cash-flows but its future values are lognormally distributed with a linear growing over the time variance. The model we developed is a two period model: •The first period where the company has one product to sale. This is the 23 CHAPTER 3. - 24 continuation region, as it is not yet optimal for the company to launch the product at that moment. We will later refer to it as the Previous Generation (P) moment; •The second moment where the company releases the new product, as it is optimal to invest (stopping region). We will later refer to this moment as the New Generation (N) moment. This market is assumed to be complete, and it is possible to replicate this company’s options with financial assets, thereby the required rate of return is the riskfree rate of return, denominated as r(and r > α). This means that the company’s value, in the Previous Generation period, according to the Net Present Value (NPV) will be: ∞ Z 0 xe−(r−α)tdt =x r−α(3.2) At some point in time the company will introduce a new and upgraded version of its product. This upgraded version suits more costumers and may expand the market. The cost of producing the first product will now be lower, and the company may choose to sell it with a lower price, reaching to more consumers. Sometimes companies choose the introduction of a new product to introduce the old product into new markets with lower purchase power, trying to increase (or at least maintain) its sales. With two products now being sold it is likely that the market will expand, because of the complementary products. An example of this is the DVD players’ market, as mentioned before, or the smartphone’s market. With the first generation of smartphones being sold at a lower price, the number of devices sold will, most likely, increase because new consumer segments will be reached. With the increase on the number of devices sold the number of available applications will also increase. This will make the smartphones more desirable, so the market will expand. The company’s cash-flows will now have two sources: its first product and its new product. Since the new is an improvement over the old product it is likely to see a cut-down on the old product’s sales. The immediate decrease in sales will be represented by φ∈[0,1], depending on the perception of the increased value that the new product has over its predecessor. It is expected that the new product will have some new features when compared to the old one, and will be released in a more mature market (since the old one had already made its way into the market), CHAPTER 3. - 25 thus it is expected that sales will have a boost. The cash-flows of the new product at the starting point of its sales, as expected by the company are y. This value may also be re-estimated by the company if it feels like the market has somehow changed and the estimations may have to be re-thought. Since there is a new and improved product on the market, the old one will have a new rate of growth αP(that can be positive or negative). The new products growth rate will be αN. The cash-flows from the old product, after the release of the new one turns to be: ∞ Z 0 x(1 −φ)e−(r−αP)tdt =x(1 −φ) r−αP(3.3) Notice that where before was xis now (1 −φ)x, which captures of the losses in sales induced by the existence of a new generation product. One may assume that the two new product will not be good enough to surpass the old one and assign 0 to φ. The model also contemplates the possibility that the old product will disappear and, in that case, φwould be equal to one. Also notice that αis now αPbecause it is now a different generation, and the old product’s sales will evolve differently from the past. Again, there is the possibility that αPequals α. The model was developed to be suitable to a large variety of markets and conditions. The cash-flows of the new product will be: ∞ Z 0 ye−(r−αN)tdt =y r−αN(3.4) Again, yrepresents the cash-flows of the new product at the starting point of its sales, as expected by the company. Remember that those are estimated by the company, and the estimations may have to change overtime in order to accommodate possible changes in the market. αNis the expected perpetual growth rate of the cash-flows produced by the new product (y). So, when the company has only one product, its entirely cash-flows will come from that product and, since there is no competition, the company could stay there forever. But the knowledge that the release of a second product will boost the market, probably increasing its cash-flows, makes the company wonder about the option to launch a new product. To exercise this option (launching of a new product) the company will have to pay K(here Kmay be seen as all of the costs of commercial- CHAPTER 4. - 32 •h1=h2, if the new product does not impress the consumers, and their negative view of the company being a follower did not change. If this happens, the loss of reputation will affect both products by the same amount; •h1<h2, if the new product disappoints the consumers. The company had already lost reputation due to being late in the market, and the new product failed to meet the consumers expectations, lowering the company’s reputation concerning the new product. So, in the beginning there is a company, acting in the market and receiving xcash-flows, but facing the possibility that a competitor releases a new product, which will lower the company’s cash-flows, and leaving it with the option to enter that generation. As usual the solution must be solved backwards, starting with the scenario where the rival has already released its product, and it is up to the company to choose whether or not to invest, launching a new product as well, and when to do it. After the entrance of the rival firm, the old product is worth: ∞ Z 0 x(1 −h1)e−(r−α)tdt =x(1 −h1) r−α(4.1) Just as happened in the previous model, the company may introduce a new upgraded version of its product, which may expand the market. The company’s cash-flows will have two sources: its first product and its new product. The first part of the cash-flows will be x(1 −φ), while the second part will be (y). Since there is a new and improved product on the market, the old one will have a new rate of growth αP(it will be a positive or negative one, depending on the company’s strategy of releasing it or not in another countries with less consumption power). The new products rate of growth will be αN. The company’s value will be the sum of its assets in place with the option to release a new product. The latteris given by H(x). The new differential equation comes: 1 2σ2x2H00(x) + αH0(x)−rH(x) + x(1 −h1) = 0 (4.2) Following the development of the equation showed in the section before, the solution to this equation will be: CHAPTER 4. - 33 H(x) = E1xβ1+E2xβ2+x(1 −h1) r−α(4.3) In addition H(x) will have to satisfy the following boundary conditions: lim x→+∞H(x)−x(1 −h1) r−α= 0 (4.4) lim x→x∗ H H(x) = (1 −φ)x∗ H(1 −h1) r−αP+y(1 −h2) r−αN−K(4.5) lim x→x? H H0(x(1 −φ)(1 −h1) r−αP(4.6) Once again, the first condition (4.4) ensures that the option component of the value function tends to zero as xtends to infinity. This means that option to launch the new generation product will be worthless when the cash flows of the old product becomes infinity large. The second condition (4.5) is the value-matching condition. If the company decides to invest, it will receive the present value of the cash-flows originated by the old product and the present value of the cash-flows that come from selling the new product (remember that this time the value of the product may be lower because of being a second mover). The parameter Kis the cost to exercise the option, which is, in this case, all of the costs required to release the new product (we assume that K does not depend upon being a first or second mover). The company should launch its new generation when xhits x∗ Hmeans the values of xfor which it is optimal to invest and release a new product, because this values are in the stopping region (New Generation moment). The third condition (4.6) is the smooth-pasting condition, which assures that the function is continuously differentiable along x. After considering the boundary conditions the solution comes: H(x) =                    (1 −φ)x∗ H(1 −h1) r−αP+y(1 −h2) r−αN−x∗ H(1 −h1) r−α−Kx x∗ Hβ2 + +x(1 −h1) r−αfor x≥x∗ H (1 −φ)x(1 −h1) r−αP+y(1 −h2) r−αN−Kfor x<x∗ H (4.7) CHAPTER 4. - 34 The first branch of the equation (the upper one) represents the value that the company receives for being in the market with its first product, after the competitor has entered the market, added with the value of the option to release its new product. The company will likely decrease the older product’s sales in exchange for having bigger sales in the new product. After the launch of the new generation, both products may be in the market, each one contributing with its cash-flows to the company. Nevertheless this model contemplates the scenario where the sales of the old product go directly to zero and any scenario between the new product having part of the market or the total of it. The second branch (lower one) represents the value that the company will receive after the release of the new generation, net of the exercise price. Notice that when there is no reputation costs (h1=h2=0) the value function of H(x) and the trigger x∗ Hbecome, respectivelly, F(x) and x∗ F. This means that, without reputation damage, the company acts as a monopolist. The trigger (the value from which the company will exercise the option to release a new product) will be: x∗ H=β2 β2−1 K−y(1−h2) r−αN (1−h1)(1−φ) r−αP−1−h1 r−α (4.8) Having the solution for the moment when the rival has already entered the market, we will now study the moment where there is the possibility of a competitor entering the market. The company is acting in the market, receiving a certain amount of cash-flows x, and having the option to release a new product. At the same time, there is the possibility that a competitor moves firstly into the market starting that generation. If the company is the first mover it will receive its cash-flows, (1−φ)x r−αP+y r−αN. On the other hand if the rival moves first the company will stop receiving xand will be receiving x(1−h1) for the reasons mentioned before. The company will have lost its option to be a first mover, receiving (1−φ)x r−αP+y r−αNand will have the option to be a second mover and receive (1−φ)x(1−h1) r−αP+y(1−h2) r−αN. Notice how the company goes from having the option F(x) to having the option H(x). Hence the company, at the starting point, will have the option G(x), which may turn out to be an option F(x) with some probability of becoming an option H(x). The possibility of this happening is given by λ, which will capture the probability of the competitor releasing a new product and becoming a first mover. So, the differential equation will be: CHAPTER 4. - 35 1 2σ2x2G00(x) + αG0(x)−rG(x) + x+λ(H(x)−G(x)) = 0 (4.9) As usual G(x) must satisfy a set of boundary conditions1: lim x→+∞G(x)−x r−α= 0 (4.10) lim x→x∗ G g(x) = x∗ G(1 −φ) r−αP+y r−αN−K(4.11) lim x→x? G G0(x) = 1−φ r−αP(4.12) After considering the boundary condition (4.10), the solution comes: G(x) = C2xγ2+h1x r−α−λ+H(x) (4.13) Solving equation (4.13) according to the boundary condition of equation (4.11) we obtain the value of C2: C2=(1 −φ)x∗ G r−αP+y r−αN−h1x∗ G r−α−λ−H(x∗ G)−K1 x∗ Gγ2 (4.14) By replacing equation (4.14) in equation (4.13) we obtain the complete solution: G(x) =                    (1 −φ)x∗ G r−αP+y r−αN−h1x∗ G r−α−λ−H(x∗ G)−Kx x∗ Gγ2 + +h1x r−α−λ+H(x) for x≥x∗ G (1 −φ)x r−αP+y r−αN−Kfor x < x∗ H (4.15) This time the trigger value does not have an analytical solution. In order to find it one most solve numerically the following equation: γC2x∗ G γ2+h1x∗ G r−α+λ+H0(x∗ G)x∗ G=(1 −σ)x∗ G r−αP(4.16) 1The economic interpretation of the boundary conditions is similar to the ones represented previously. CHAPTER 4. - 36 where γ2is the following equation: γ2=1 2−µ σ2−sµ σ2−1 22 + 2r+λ σ2(4.17) If the probability of a rival entering the market (λ) or the effects of this entrance in the reputation of the company (h1and h2) were null G(x) would become F(x) (see chapter 3). After developing the models we will next apply them to real-world situations. Chapter 5 The case of the Sony Playstation video-games business In this chapter we will apply the model to a real-world situation. For this purpose we collected some data about the Sony Playstation sales1. We will present three different situations: •One where the company introduces its product earlier than its rivals. This is the case of the Playstation 1 transition to the Playstation 2 (PS1 to PS2); •One where the company introduces its product after the competition, loosing reputation. This is the case of the Playstation 2 transition to the Playstation 3 (PS2 to PS3); •One where the company there is the possibility of a rival entering the market. this is the case of the Playstation 2 transition to the Playstation 3 before the Microsoft Xbox 360 was released. In the final section of this chapter we will present a comparative statics and sensitive analysis. Notice, however that this exercise is just for demonstration purposes. We do not have enough data, and do not intent, to make a final assessment of the optimal timing for this company to introduce a next generation product. We will have some assumptions in order to keep the application simple. The reason we chose the video game industry is because we feel it illustrates well enough the high-tech market. 1From vgsales.wikia.com and Sony’s financial reports 37 CHAPTER 5. - 38 It is easy to distinguish between generations because they seem to last for five to seven years and usually there is something in common between the products available for that generation (for example, the forth generation is known as the 16bit era because of the hardware used). It is important to know the perfect timing to release a product, as the Sony Playstation benefited from being launched before the Sega Saturn, but the Sony Playstation 3 suffered from being sold after the Xbox 360. It is a market where the old product continues to be sold (Sony continues to sell its old generation products five years after the release of the new generation), which is good to see how the model fits reality. Being a high-tech market there is always high uncertainty as new technologies are constantly being discovered and new ways of competition (nowadays it seems that smartphones and social network mini-games are stealing consumers from home video games), with hidden competitors (Microsoft only entered the market in 2002 with the first Xbox). Also, there seems to be an effect on the sales of the old generation when the new generation starts. We start by taking a look at Sony Playstation’s sales per fiscal year (ending in March) in the Figure below: Figure 5.1: Sony Playstation sales per fiscal year CHAPTER 5. - 39 In the first column we can see the fiscal year to which the sales refer. PS1, PS2 and PS3 refer to the first, second and third generations of Sony Playstation’s consoles. On the fifth column we can see the name of Sony’s competitors, the year when they release their video game system and in which generation it belonged. In order to know the cash-flows provided by each console we must assume that they only come from the sales of the consoles, ignoring the software and accessories sales (as we do not have enough information for that). The prices of the consoles are not static, and tend to lower throughout time. Given that we do not have the sales per month, we assume that when there is a price drop it affects the prices of all the consoles sold that year. So, imagining that Sony lowered the price of one console in the middle of a year, we will assume that price to all the consoles sold that year. Since we do not have the production costs we will assume the cash-flows to be equal to the operating incomes. We will assume the operating incomes to be a fixed percentage (the 2010 values) of the sales. This means that we assume the net cashflows of a year to be “sales (in dollars)” times the 2010 “Operating Income/Sales 2010” ratio for the gaming division and assumed it to be constant. Figure 5.2: Sony Playstation total sales per unit As the risk-free we used the 10 years Japanese Government bonds yields of August 1994 to August 1995 (4%), as the PS1 was released in 1995, Sony is a Japanese company and Japan is the world’s biggest video-game markets. In order to know the value of the cash-flows growth, or drift, (α) we used the CHAPTER 5. - 40 cash-flows shown before. But since we are working in a risk-free environment, we have to find the risk-neutral drift (α∗)2. The risk premium was obtained using the CAPM model with a βand a premium for the risk given of the entertainment tech industry by Damodaran (7%) 3, a risk-free rate as mentioned before, and Sony’s financial reports to know the debt-to-equity ratio and the tax rate. According to Sony’s financial reports, the value of the investment, Kis 24.57 million euros. The value of φwas considered to be: •0.614 in the first example (because when Sony released the PS2, the sales of the PS1 lowered 61.4%); •0.3 in the second and third examples (the consumers do not see an increase in quality from the PS2 to the PS3 as high as it was from the PS1 to the PS2 - as may be seen from year 2006 to 2007) As a simplification for the project’s volatility (σ) we used Sony’s share’s daily returns volatility: •in the first example, from 1995 until 2000 (year of release of the PS2). The value of σis, then, 0.3; •in the second and third example, since the PS3 came after the PS2 (which was released in 2000), and was released in 2006 we considered only the period between 2000 and 2006, as Sony was evaluating when to release the new generation during that period. So, σis 0.369. 5.1 PS1 to PS2 Sony Playstation was a huge success for the company, as the console was one of the most remarkable consoles of all time, being the most sold one during that generation. Sony first introduced Playstation (or PS1) during the year 1995, almost at the same time as one big competitor (Sega Saturn), and a few moments before other major competitor (Nintendo 64). With these products being released with few 2Since µ−α=δand r−α∗=δand considering that µ=r+RP where RP stands for the Risk Premium (CAPM), it comes: r+RP −α=r−α∗, and so: α∗=α−RP where α∗is the risk-neutral drift and αis the risk-adjusted drift. 3http://pages.stern.nyu.edu/ adamodar/ on 4th of May, 2013 CHAPTER 5. - 41 months apart, Sony and its competitors were able to put all the cards in the table at once, and fight for the consumers preference. In this first example we will consider that Sony was the Leader of the market, and that the PS1 was so successful that the competition could not danger Sony reputation on the market when the company decided to release the PS2, so the model fits the case where the company controls the technology leap, acting as a monopolist. In Figure 5.1 one may see that in 1999 Playstation 2 (PS2) sold 1.41 millions of units. PS2 was released one month before the ending of the fiscal year, and so we decided to add this values to the 2010 fiscal year, and ignore that month, for simplicity purposes. Following the rationale mentioned before, the values we considered were: Figure 5.3: PS1 to PS2 case parameters When Sony is in the market, in 1995, with the Playstation 1 it is receiving 72 million dollars (x), with a risk-neutral rate of growth of -0.088. Sony has the option to release the PS2 by paying the exercise price (K=24.57). If Sony releases the PS2, the cash-flows of the PS1 will be x(1 −φ). Also, by releasing the PS2 Sony will receive its corresponding cash-flows: 178 million dollars (y) growing at a neutral-risk rate of −0.123. The sales of the PS1 will only increase to a point, after which they will start to decrease as the product is not new and lost some of its appeal. In which point should Sony release the PS2 and start a new generation? Imagining that Sony has the control of the market, and the entrance of a challenger does not affect the sales of the company. The project’s trigger (the moment CHAPTER 5. - 48 As the effect of the loss of reputation in the old product increase so does the trigger. This is because the loss in the sales of the old product will rush the company to release a new generation to compensate this loss. The increase of h2leads to an increase in the trigger of the company when the effect is irreversible, but leads to a decrease in the trigger of the company when the competitor has not yet entered the market. If the competitor has already entered the market the company should see the loss in the new generation as a sunk cost, and this should decrease the appeal to release a new product. If the competitor has not yet entered the market, but there is the possibility of doing so, the company should release the new generation sooner, trying to avoid the effect of the reputation loss. At least, as it would be expected, as the chances of rivals entering the market increases so will the trigger of the company, leading to an early increase to avoid the reputation losses. Chapter 6 Conclusions Following the research done in Real Options we developed two new models whose purpose was to determine the optimal time for a technological company to jump from one generations of product into another. These two models contributed to the field of investigation in Real Options by introducing some innovations, combining the stochastic evolution of the cash-flows with the product’s life cycle, and seeing the option to enter the market as a put option, by which the company has the option to abandon the current generation, receiving in return the value of the new one. We analyzed the value of the high-tech market, perceiving that this market is largely based upon research and development, with current high rates of growth on the development of new technologies and discoveries being made. We also described how the consumers tend to have a fast-paced adoption of new products, leaving the old ones with a small share of the market. This combined with the costs of releasing a new product make it decisive to correctly estimate the perfect moment to start a new generation, as there are major variables considerations to be aware of, such as the level and speed of acceptance of the new product, or the effect of the complementary products and even marketing issues, among other variables. In order to gain the needed knowledge to study this subject we reviewed some research conducted before, such as the beginning of the Real Options theory, some models used to evaluate options when the cash-flows are stochastic, others were the focus is on the product life-cycle, we studied researches about innovation (what it is, what is the impact and how it is accepted by consumers), and ended our research with studies about the strategy used in high-tech industries when releasing new products. In our first model we introduced a model where the company had the control of 49 CHAPTER 6. - 50 the technological leap, being capable of determining when to start a new generation. In this case the company is acting in the market with the current generation but has the option of giving up a part, or the total amount, of the cash-flows being received by that product (because the old product may share the market with the new one) as well as an exercise cost (the total amount of costs demanded to introduce and commercialize the new product), in exchange of the value of a new generation. When the company exercises the option it will receive the new project’s value, loosing the exercise price and the old product will loose immediately part of its cash-flows, and will evolve with a different growth rate. We found out that, in order for the model to make sense when the company exercises its option, one of two things must happen: either the company is investing because of the value of the new product’s cash-flows, since its older product will be affected by this new one (when the value of the new generation exceeds the exercise price), or the company is investing because of the synergies the new generation will have with the old one when both share the market (when the exercise price is higher than the value of the new generation). In the second model we introduced the effect of competition. This time, the company does not fully control the technological leap, and does not determine when to start a generation. In this model, the company is active in the maket with a product, but faces the risk of competition, with others being pioneer to enter the new generation market (by releasing a new product first), and leaving the company with only the option to settle for being the follower. In this reality the company has an option to enter the market as being the firstmover, but that option is only valid until a competitor decides to move first. In that case the company looses that option, and is left with the option of entering the market, where there is already competition. If the company looses the opportunity to be the first mover, it will have consequences in the firm’s reputation, which will influence its cash-flows, reducing them by an amount between 0% and 100%. We saw how the models contemplates different possibilities of loss for each product (the decrease in reputation may affect the new product and the old one by a different dimension). The model was obtained by determining what happens when the competitor has entered the market, and then including that in the model where there is only the possibility of that happening. We also pointed that if there is not the probability of a rival entering the market, or if this entrance does not affect the company, then all the models would behave as the model where the company controls the technological leap. CHAPTER 6. - 51 In the fifth chapter we used a real-world example: the video-game industry, with Sony being the company analyzed. The goal was to create three different realities and evaluate how the model suits this market. In the first scenario we evaluated the transition from the Playstation to the Playstation 2, imagining that Sony does not have competition. We concluded that the company was close to its trigger when it released the new product (the theoretical optimal timing to exercise the option would be when the cash-flows were almost 45 million dollars, but the company exercised when they were 52 million). Also we saw how the value of the project increases with the value of the cash-flows, even though the company holds a real option to abandon the first generation. In the second scenario we re-evaluated the transition from the Playstation to the Playstation 2 (PS2), considering that a rival (Microsoft Xbox 360) has already entered the market, affecting Sony’s reputation and the cash-flows generated by Playstation 2 and Playstation 3 (PS3). In this case Sony had the option to abandon part of the cash-flows of the PS2 in exchange for the cash-flows of the PS3, knowing that both consoles will sell less, as the company is not seen thenceforth as a pioneer. In this case Sony should’ve exercised the option when PS2’s cash-flows were almost 76 million dollars, but the company rushed and released the new product when the old one still had cash-flows of 142 million dollars. In the third scenario we evaluated the Playstation 2 to the Playstation 3, when the competition has already entered the market, with the loss of reputation affecting Sony’s products. In this case Sony had the option to enter the new generation market being a first mover, but with the probability of a competitor entering the market in any time, turning the company’s option to be a first mover into an option to be a follower. This fear of the entrance of competition increases the optimal timing to invest, so the company should exercise earlier. We saw that Sony did the right move by not exercising the option before the release of the Microsoft Xbox 360, but then released the PS3 too soon, when the only option was to be a second mover. We also saw that if the probability of a rival entrance, or its effects on Sony’s cash-flows, was null, the trigger would be the same as in the model where the company controls the technology leap. To conclude we conducted a sensitive analysis and verified that the increase in volatility and increase in the loss of the old product due to the introduction of the new one decrease the trigger of the company, making it invest later. The increase in loss in reputation in the old product due to the entrance of a rival, and the increase in the probability of a rival entering, increase the trigger, making the company CHAPTER 6. - 52 exercise the option earlier to avoid this losses. The increase of the damaged done to the reputation of the new product due to the entrance of a rival before the company has two effects: it increases the trigger of the company if the rival has not yet entered, making the company release the product earlier to avoid being a follower; it decreases the trigger, if the rival has already entered the market, and the losses ar a sunk cost, making the company exercise the option later. For further studies we would suggest using these models in areas different from the video-game industry (or even in other areas besides technology) and trying to develop a model where the innovation that can lead to a new generation is endogenous in the model. Bibliography Altshuller, G.: 1999, The Innovation Algorithm: TRIZ Systematic Innovation and Tecnical Creativity, Technical Innovation Center, Inc. Armada, M. R., Kryzanowski, L. and Pereira, P. J.: 2011, Optimal investment decisions for two positioned firms competing in a duopoly market with hidden competitors, European Financial Management 17(2), 305–330. Azevedo, A. and Paxson, D.: 2009, Uncertainty and competition in the adoption of complementary technologies, Hull University Business School, Hull . Bollen, N. P.: 1999, Real options and product life cycles, Management Science 45(5), 670–684. Chaney, P. K., Devinney, T. M. and Winer, R. S.: 1991, The impact of new product introductions on the market value of firms, Journal of Business pp. 573–610. Chou, C. and Yang, K.-P.: 2011, The interaction effect of strategic orientations on new product performance in the high-tech industry: A nonlinear model, Technological Forecasting and Social Change 78(1), 63–74. Daming, Y., Xiaohui, Y., Wu, D. D. and Guofan, C.: 2013, Option game with poisson jump process in company radical technological innovation, Technological Forecasting and Social Change . Dixit, A. K. and Pindyck, R. S.: 1994, Investment under uncertainty, Princeton university press. Fudenberg, D. and Tirole, J.: 1985, Preemption and rent equalization in the adoption of new technology, The Review of Economic Studies 52(3), 383–401. Huisman, K. J. and Kort, P. M.: 1999, Effects of strategic interactions on the option value of waiting, Tilburg University Netherlands. 53 BIBLIOGRAPHY 54 Lee, K., Khan, S. and Mirchandani, D.: 2012, Hierarchical effects of product attributes on actualized innovativeness in the context of high-tech products, Journal of Business Research . Munnukka, J. and J¨arvi, P.: 2011, The value drivers of high-tech consumer products, Journal of Marketing Management 27(5-6), 582–601. Myers, S. C.: 1977, Determinants of corporate borrowing, Journal of financial economics 5(2), 147–175. Nielsen, M. J.: 2002, Competition and irreversible investments, International Journal of Industrial Organization 20(5), 731–743. Paxson, D. A. and Pinto, H.: 2003, Leader/follower real value functions if the market share follows a birth/death process, Real R&D Options. Oxford: ButterworthHeinemann pp. 208–227. Reinganum, J. F.: 1989, The timing of innovation: Research, development, and diffusion, Handbook of industrial organization 1, 849–908. Rodrigues, A.: 2009, Market segmentation under uncertainty, NEGE, School of Economics and Management, University of Minho . Rubera, G., Griffith, D. A. and Yalcinkaya, G.: 2012, Technological and design innovation effects in regional new product rollouts: A european illustration, Journal of Product Innovation Management 29(6), 1047–1060. Ruiz-Aliseda, F. and Wu, J.: 2012, Irreversible investment in stochastically cyclical markets, Journal of Economics & Management Strategy 21(3), 801–847. Schumpeter, J. A.: 1934, The theory of economic development: An inquiry into profits, capital, credit, interest, and the business cycle, University of Illinois at Urbana-Champaign’s Academy for Entrepreneurial Leadership Historical Research Reference in Entrepreneurship . Takeuchi, H. and Nonaka, I.: 1986, The new new product development game, Harvard business review 64(1), 137–146. Tsekrekos, A.: 2003, The effect of first-movers advantages on the strategic exercise of real options, Real R&D Options pp. 185–207. BIBLIOGRAPHY 55 Zahra, S. A. and Bogner, W. C.: 2000, Technology strategy and software new ventures’ performance: exploring the moderating effect of the competitive environment, Journal of business venturing 15(2), 135–173.