COMPETE, COOPERATE, OR BOTH? INTEGRATING THE DEMAND SIDE INTO PATENT DEPLOYMENT STRATEGIES FOR THE COMMERCIALIZATION AND LICENSING OF TECHNOLOGY

竞争、合作或两者兼有？将需求侧整合到技术商业化和许可的专利部署策略中

LENOS TRIGEORGIS University of Cyprus 塞浦路斯大学（LENOS TRIGEORGIS，注：此处可能为专有名词或拼写错误，保留原词）

FRANCESCO BALDI University of Turin LUISS Guido Carli University 弗朗西斯科·巴尔迪 都灵大学 路易吉·卡洛·圭多大学

RICHARD MAKADOK Purdue University 理查德·马卡多克 普渡大学

Profiting from innovation typically involves a choice between commercializing a patented technology in the product market to exploit proprietary advantage (i.e., competition) or licensing the technology to an incumbent in the market for ideas (a form of cooperation). A firm may thus deploy a patented technology in ways that may differ in their aggressiveness toward, or accommodation of, competitors. We analyze the deployment of patented technology employing either competition or collaboration modes, or both together (i.e., coopetition), as well as switching among them across demand states or over time, or delaying these choices until more information is available. We thus view a patent as a bundle of real options that enables a firm to manage not only the classic tension between commitment and flexibility but also the tension between competition and cooperation. We develop theory and propositions to predict which of these patent deployment modes will be chosen by an innovator facing an established firm as a function of the strength of the technology, market or bargaining power, and other market conditions, particularly the level and volatility of market demand. 从创新中获利通常需要在两个方向之间做出选择：一是在产品市场中商业化专利技术以利用专有优势（即竞争），二是将技术许可给市场中的现有企业（一种合作形式）。因此，企业可能会以不同的方式部署专利技术，这些方式在对竞争对手的攻击性或妥协性方面可能存在差异。我们分析专利技术的部署方式，包括采用竞争模式、合作模式或两者结合（即竞合），以及在不同需求状态下或随时间推移在这些模式之间切换，或推迟这些选择直到获得更多信息。因此，我们将专利视为一系列实物期权的组合，这不仅使企业能够管理承诺与灵活性之间的经典张力，还能管理竞争与合作之间的张力。我们构建理论和命题，以预测面对现有企业的创新者会选择哪种专利部署模式，这取决于技术实力、市场或议价能力以及其他市场条件，特别是市场需求的水平和波动性。

Over the years since the pioneering studies by Teece (1986) and Levin, Klevorick, Nelson, and Winter (1987) posed the fundamental question of how firms appropriate returns from their innovation efforts, the research literature has recognized that an innovator can choose to do so in ways that put it either in competition or in some form of cooperation with established rivals (Gans & Stern, 2003). For example, Netscape initially considered cooperatively licensing its technology to Microsoft in 1995 before choosing to compete directly via its own independent cross-platform browser. The deployment of a firm’s patents is part of that core choice since patents can be deployed in various ways that differ in their implications for the firm’s competitive or collaborative relationship with established firms—a phenomenon known as “strategic patenting” (Arundel & Patel, 2003). In addition to their conventional use based on exclusivity for protecting existing products or processes from direct imitation by competitors, an innovative firm may also deploy its patents strategically in two key ways: 自Teece（1986）以及Levin、Klevorick、Nelson和Winter（1987）开创性研究提出企业如何从创新成果中获取回报这一根本问题以来，多年间的研究文献已认识到，创新者可选择以竞争或与现有竞争对手合作的方式实现这一目标（Gans & Stern，2003）。例如，网景公司（Netscape）在1995年最初考虑与微软合作授权其技术，后来却选择通过自身独立的跨平台浏览器直接展开竞争。企业专利的部署是这一核心选择的一部分，因为专利可通过多种方式部署，而这些方式对企业与现有企业的竞争或合作关系的影响各不相同——这种现象被称为“战略性专利布局”（Arundel & Patel，2003）。除了基于排他性保护现有产品或流程免受竞争对手直接模仿的常规用途外，创新型企业还可能通过两种关键方式战略性地部署其专利：

1. For competitive purposes that benefit the firm at the expense of rivals, either offensively or defensively. Offensive aims include blocking competitors’ ability to expand their future technology beyond what they have already patented (Ceccagnoli, 2009; Granstrand, 1999), sometimes called “bracketing,” or obstructing their rival’s ability to conduct research and development (R&D) activities via patents on research tools or techniques (Heller & Eisenberg, 1998). Defensive tactics include protecting a firm’s own technology from competition by substitute technologies (Hounshell & Smith, 1988) or preventing competitors from blocking the firm’s own ability to expand its future technology beyond what it has already patented (Granstrand, 1999; Reitzig, 2004a), sometimes called “patent walls” or “patent fences,” or preventing competitors from obstructing the firm’s ability to conduct R&D activities by patenting research tools or techniques before they do (Walsh, Arora, & Cohen, 2003).
2. 为了竞争目的而损害竞争对手利益以惠及本公司，无论是进攻性还是防御性手段。进攻性目标包括阻止竞争对手将其未来技术发展超越其已获专利的范围（Ceccagnoli, 2009; Granstrand, 1999），有时称为“围堵”，或者通过对研究工具或技术申请专利来阻碍竞争对手开展研发（R&D）活动（Heller & Eisenberg, 1998）。防御性策略包括通过替代技术保护公司自身技术免受竞争（Hounshell & Smith, 1988），或者防止竞争对手阻碍公司将未来技术发展超越其已获专利的范围（Granstrand, 1999; Reitzig, 2004a），有时称为“专利壁垒”或“专利围栏”，或者在竞争对手之前对研究工具或技术申请专利，以防止其阻碍公司开展研发活动（Walsh, Arora, & Cohen, 2003）。

. For cooperative purposes that benefit a group of firms jointly. Such cooperation includes beneficent aims that allow rival firms to collectively create more value for their customers, such as cross-licensing their patented technologies to each other to achieve “design freedom” (Grindley & Teece, 1997), set common standards, or avoid a patent thicket or anticommons, or unilateral licensing to obtain royalty revenue from other firms, including rivals that can make better use of the technology (Gans & Stern, 2003; Rivette & Kline, 200ob). Cooperative deployment may also include collusive-type tactics that allow rival firms to collectively capture more value from their customers, such as attaining a stable cartel using licensing to disincentivize rivals from competing aggressively (Arora, 1997; Rubinfeld & Maness, 2005), or raising entry barriers to protect industry incumbents from external threats (Calabrese, Baum, & Silverman, 2000; Reitzig, 2004b). . 为了实现对一组企业共同有利的合作目的。这种合作包括善意的目标，使竞争企业能够共同为客户创造更多价值，例如相互交叉许可其专利技术以实现“设计自由”（Grindley & Teece，1997）、制定共同标准，或避免专利丛林或反公地问题，或者单方面许可以从其他企业（包括能够更好利用该技术的竞争对手）获得特许权使用费收入（Gans & Stern，2003；Rivette & Kline，200ob）。合作部署还可能包括勾结式策略，使竞争企业能够共同从客户那里攫取更多价值，例如通过许可来阻止竞争对手激烈竞争以形成稳定的卡特尔（Arora，1997；Rubinfeld & Maness，2005），或者提高进入壁垒以保护行业现有企业免受外部威胁（Calabrese，Baum，& Silverman，2000；Reitzig，2004b）。

PATENT DEPLOYMENT: ANSWERED AND UNANSWERED QUESTIONS

专利部署：已解答和未解答的问题

Why, and under what conditions, do innovators choose different patent deployment strategies that involve different types of competitive or cooperative relationships with established firms? Research on this strategic patenting phenomenon has focused predominantly on supply-side factors as drivers of the firm’s choices of patent deployment strategy, such as the characteristics of the firm, the technologies, the rivals, the industry, and the appropriability environment. For example, it is widely recognized that cooperation in the form of cross-licensing is more common in “complex product” industries that require combining multiple technologies whose patents are owned by different firms, which would otherwise leave these firms vulnerable to holdup problems (Cohen, Nelson, & Walsh, 2000; Grindley & Teece, 1997; Hall & Ziedonis, 2001). By contrast, patent deployment in the form of a competitive defense is more common in “discrete product” industries where such technology combinations are not required (Reitzig, 2004a). Likewise, an offensive competitive stance to patent deployment is more common in industries that exhibit incremental innovation than in those involving more radical innovation (Ceccagnoli, 2009). Another key supplyside factor is the inherent degree of imitability or substitutability in the product market (lack of product differentiation), which encourages patent deployment that is more proprietary and aggressive (Polidoro & Toh, 2011) and less cooperative (Arora & Fosfuri, 2003; Hill, 1992). The appropriability environment also influences firms’ patent deployment strategies (Gans, Hsu, & Stern, 2002; Gans & Stern, 2003; Teece, 1986). At the firm level, specialized complementary assets increase a firm’s propensity to pursue a more proprietary or competitive approach to patent deployment (Arora & Ceccagnoli, 2006). Firms are more likely to license patents cooperatively when aiming to influence industry standards (Shapiro & Varian, 1999), but are less likely to license patents on core technologies that are central to their efforts to build and sustain competitive advantage (Somaya, 2003). Mihm, Sting, and Wang (2015: 2666) predicted firms’ patent deployment strategies as a function of supply-side “internal contingencies (i.e., the firm’s own R&D strategy) and environmental contingencies (i.e., the competitors’ patenting and R&D strategies, industry clock speed and complexity).” 创新者为何以及在何种条件下会选择不同的专利部署策略，这些策略涉及与现有企业建立不同类型的竞争或合作关系？对这种战略性专利现象的研究主要聚焦于供给侧因素作为企业选择专利部署策略的驱动因素，例如企业自身的特征、技术、竞争对手、行业以及专利可获性环境。例如，众所周知，交叉许可形式的合作在“复杂产品”行业更为常见，这类行业需要结合由不同企业拥有专利的多项技术，否则这些企业将容易受到敲竹杠问题的影响（Cohen, Nelson, & Walsh, 2000；Grindley & Teece, 1997；Hall & Ziedonis, 2001）。相比之下，竞争性防御形式的专利部署在“离散产品”行业更为常见，在这些行业中不需要此类技术组合（Reitzig, 2004a）。同样，进攻性竞争姿态的专利部署在以渐进式创新为主的行业中更为常见，而不是在涉及更多激进创新的行业中（Ceccagnoli, 2009）。另一个关键的供给侧因素是产品市场中固有的可模仿性或可替代性程度（即产品差异化不足），这会鼓励专利部署更加专有化和具有侵略性（Polidoro & Toh, 2011），并减少合作（Arora & Fosfuri, 2003；Hill, 1992）。专利可获性环境也会影响企业的专利部署策略（Gans, Hsu, & Stern, 2002；Gans & Stern, 2003；Teece, 1986）。在企业层面，专门化的互补性资产会增加企业采取更专有或竞争性专利部署策略的倾向（Arora & Ceccagnoli, 2006）。当企业旨在影响行业标准时，它们更有可能合作许可专利（Shapiro & Varian, 1999），但不太可能许可对其构建和维持竞争优势至关重要的核心技术专利（Somaya, 2003）。Mihm, Sting, 和 Wang (2015: 2666) 预测企业的专利部署策略取决于供给侧的“内部突发事件（即企业自身的研发战略）和环境突发事件（即竞争对手的专利和研发战略、行业迭代速度和复杂性）”。

Hoffmann, Lavie, Reuer, and Shipilov (2018) highlighted the interplay of competition and cooperation and balancing the associated tension between value creation and appropriation as a key research gap in strategic management. The current article aims to help fill this research gap in the context of patent deployment strategies by focusing on two key intertwined aspects of this interplay: incorporating what Brandenburger and Nalebuff (1996) called “coopetition,” and considering switching over time. First, although much is already known about how a firm chooses between deploying patents in competitive versus cooperative ways, less is known about why a firm may do both simultaneously by aggressively commercializing the technology itself while also concurrently licensing it to competitors—that is, the patent-deployment version of coopetition. Second, even less is known about why firms sometimes switch their patent deployment between the competition, cooperation, and coopetition modes under different demand conditions or over time. Both aspects are included in the analysis to provide related but complementary perspectives on this interplay as part of a comprehensive framework of analysis—as coopetition represents a third (hybrid) alternative to competition and cooperation that a firm may switch to over time under certain conditions. A number of unique insights emerge from addressing both aspects of this interplay (discussed in Propositions 2.2 and 3). As we explain below, these two intertwined aspects of the main research gap concerning the interplay of competition and cooperation are partly due to the fact that prior studies have ignored product-market demand as a key factor to help explain patent deployment strategies. To clarify this connection, consider the limitations that have constrained extant research on both the independent- and the dependent-variable sides of the equation. Hoffmann、Lavie、Reuer 和 Shipilov（2018）强调，在战略管理中，竞争与合作的相互作用以及价值创造与攫取之间相关张力的平衡是一个关键研究空白。本文旨在通过聚焦这种相互作用的两个关键交织方面，在专利部署策略的背景下填补这一研究空白：一是纳入Brandenburger和Nalebuff（1996）所谓的“竞合”（coopetition）概念，二是考虑随时间推移的策略转换。首先，尽管我们已较为了解企业如何在竞争性或合作性专利部署之间做出选择，但对企业为何会同时以两种方式部署专利——即一方面积极商业化自身技术，另一方面又同时向竞争对手许可该技术——的原因却知之甚少，这正是专利部署视角下的竞合现象。其次，对企业为何在不同需求条件下或随时间推移，在竞争、合作与竞合模式之间切换专利部署策略的了解则更少。这两个方面均纳入分析，以提供关于这种相互作用的补充视角，共同构成综合分析框架的一部分——因为竞合代表了竞争与合作之外的第三种（混合）选择，企业在特定条件下可能会随时间转向这种模式。通过探讨这两个相互作用的方面，我们得出若干独特见解（详见命题2.2和3）。如下文所述，关于竞争与合作相互作用的主要研究空白的这两个交织方面，部分原因在于先前研究忽视了产品市场需求作为解释专利部署策略的关键因素。为阐明这一联系，我们需考虑限制现有研究的因素，这些因素既存在于方程的自变量一侧，也存在于因变量一侧。

Limitations of Prior Research on the Independent-Variable Side

自变量方面先前研究的局限性

All of the aforementioned explanatory factors are focused solely on the supply side of the product market, with little or no consideration given to the role that product-market demand may play in explaining firms’ patent deployment strategies. Yet expert practitioners have observed that “the phase of industry growth—for example, is it an emerging or a maturing business?—also shape[s] the IP strategy needs of each business unit” (Rivette & Kline, 2000b: 69). Mihm et al. (2015: 2670) also acknowledged this omission, as well as its significance, noting that the payoffs to various patent deployment strategies “are also affected by other factors … such as demand uncertainty.” However, we are aware of only two instances where research studies have considered demand-side factors in explaining patent deployment strategies. On the theoretical side, a pair of formal models (Bar-Ilan & Strange, 1998; Weeds, 1999) consider the influence of demand on the timing of patent deployment (i.e., when the patent is deployed, but not how it is deployed). On the empirical side, Polidoro and Toh (2011) examined the level of product-market demand as a contingency factor influencing patent deployment strategy. 所有上述解释性因素都仅聚焦于产品市场的供给方，对产品市场需求在解释企业专利部署策略中可能发挥的作用几乎未加考虑。然而，业内专家已观察到，“行业增长阶段——例如，是新兴业务还是成熟业务？——也会影响每个业务单元的知识产权战略需求”（Rivette & Kline，2000b：69）。Mihm等人（2015：2670）也承认了这一遗漏及其重要性，并指出，各种专利部署策略的回报“还受到其他因素……如需求不确定性的影响”。不过，我们仅发现两项研究在解释专利部署策略时考虑了需求侧因素。在理论层面，一组正式模型（Bar-Ilan & Strange，1998；Weeds，1999）考察了需求对专利部署时机（即何时部署专利，而非如何部署）的影响。在经验层面，Polidoro和Toh（2011）研究了产品市场需求水平作为影响专利部署策略的权变因素。

Using product-market demand to predict patent deployment strategy may not, by itself, generate significant new insights, as it is known that demand is an important determinant of market entry (see, e.g., Cabral, 2000), and a patent deployment choice (commercializing the technology oneself versus licensing it out to rivals) is in some ways similar to an entry decision. However, more interesting than accounting for demand’s main direct effect is the question of how patent deployment choices are influenced by the interaction of both demand-side and supply-side factors. For example, given that a firm would behave less cooperatively for a patent that offers stronger technological advantages (Somaya, 2003), is this effect strengthened or weakened as demand conditions become more favorable? On the one hand, stronger demand raises the amount of value a firm can capture in the product market when keeping its competitive advantage proprietary by using its patents in an exclusionary way. However, on the other hand, a stronger demand also raises the likelihood of a rival also producing profitably, which in turn increases the amount of licensing revenue a firm can capture in the market for technology by using its patent in a cooperative way. As a patent’s technological strength increases, under what conditions will the product-market benefit of stronger demand for the innovator, which requires more aggressive deployment, outweigh the rival entry and technology-market benefits, which favor cooperative deployment? Answering these questions requires addressing the interaction between factors on both the demand and supply sides. 利用产品市场需求来预测专利部署策略本身可能不会产生显著的新见解，因为众所周知，需求是市场进入的重要决定因素（例如，参见Cabral，2000），而专利部署选择（自行商业化技术与授权给竞争对手）在某些方面类似于进入决策。然而，比考虑需求的主要直接影响更有趣的问题是，专利部署选择如何受到需求方和供给方因素相互作用的影响。例如，鉴于企业对于技术优势更强的专利可能会采取更少的合作行为（Somaya，2003），随着需求条件变得更加有利，这种效应是增强还是减弱？一方面，更强的需求会提高企业通过排他性地使用专利来保持竞争优势时，在产品市场中能够攫取的价值量。然而，另一方面，更强的需求也会增加竞争对手也能盈利生产的可能性，这反过来又会增加企业通过合作方式使用专利在技术市场中能够攫取的许可收入。随着专利技术强度的增加，在何种条件下，对创新者而言更强需求带来的产品市场收益（这需要更积极的部署）会超过有利于合作部署的竞争对手进入和技术市场收益？回答这些问题需要解决需求方和供给方因素之间的相互作用。

An additional benefit of incorporating demandside factors as independent variables to explain patent deployment strategies is that they tend to be more dynamic in nature and therefore offer a greater potential to explain why firms choose different patent deployment strategies under different demand conditions, or switch their patent deployment strategies over time. Most of the supply-side factors that prior research has considered tend to be fairly static over time; if they change at all, they usually do so in slow, gradual, incremental ways. For example, a firm’s stock of complementary assets (Arora & Ceccagnoli, 2006) cannot change quickly due to timecompression diseconomies (Dierickx & Cool, 1989). An industry’s rate of innovation (Ceccagnoli, 2009), its degree of product substitutability or differentiability (Arora & Fosfuri, 2003; Hill, 1992; Polidoro & Toh, 2011), and the discrete versus complex nature of the technology (Cohen et al., 2000; Grindley & Teece, 1997; Reitzig, 2004a) are fairly stable characteristics at the industry level. Although the appropriability environment (Gans et al., 2002; Gans & Stern, 2003; Teece, 1986) may sometimes shift due to occasional changes in patent law and related practices, it varies much more cross-sectionally among jurisdictions than it does longitudinally over time. We are aware of only a few cases where supply-side factors have been shown to affect patent deployment strategies in a dynamic way. First, when biology and genetics replaced chemistry as the main basis for drug discovery, the pharmaceutical industry shifted from a discrete product technology to a complex one, making licensing more common (Clark & 将需求侧因素作为自变量纳入以解释专利部署策略的另一项额外优势在于，这些因素本质上往往更具动态性，因此更有可能解释企业为何在不同需求条件下选择不同的专利部署策略，或随时间改变其专利部署策略。先前研究考虑的大多数供给侧因素往往随时间变化较为稳定；即便发生变化，通常也是缓慢、渐进、增量式的。例如，企业的互补资产存量（Arora & Ceccagnoli，2006）由于时间压缩不经济（Dierickx & Cool，1989）而无法快速变化。一个行业的创新率（Ceccagnoli，2009）、产品可替代性或差异化程度（Arora & Fosfuri，2003；Hill，1992；Polidoro & Toh，2011），以及技术的离散性与复杂性特征（Cohen et al.，2000；Grindley & Teece，1997；Reitzig，2004a）在行业层面都是相当稳定的特征。尽管专利可获得性环境（Gans et al.，2002；Gans & Stern，2003；Teece，1986）有时可能因专利法及相关实践的偶然变化而发生转变，但它在不同司法管辖区之间的横向差异远大于其随时间的纵向变化。我们仅发现少数案例表明供给侧因素会以动态方式影响专利部署策略。首先，当生物学和遗传学取代化学成为药物研发的主要基础时，制药行业从离散产品技术转向复杂产品技术，使得许可更为普遍（Clark &

Konrad, 2008). Second, voice-recognition patents were licensed more when the technology’s strength became clearer to potential licensees (Marx, Gans, & Hsu, 2014). Yet, both of these instances represent unidirectional shifts toward greater cooperation via licensing over time; no supply-side factors have been shown to shift patent deployment strategies toward more competition over time. 康拉德，2008）。其次，当技术优势对潜在被许可方变得更加清晰时，语音识别专利的许可量更大（马克思、甘斯和许，2014）。然而，这两个例子都代表了随着时间推移通过许可向更深入合作的单向转变；没有证据表明供应方因素会随着时间推移将专利部署策略转向更具竞争性。

Demand-side factors tend to be more dynamic than supply-side factors, for two reasons. First, industries typically experience a somewhat predictable product life cycle in which the level of demand for a product grows, matures, and eventually declines (Levitt, 1965; McGahan, 2004). Second, demand may also shift in stochastic or unpredictable ways due to fickle consumer preferences, economic shocks affecting customers’ incomes, cultural fads and fashions, and many other reasons. Insofar as stochastic demand changes may affect the optimal use of a patent, a firm may benefit from retaining flexibility to adjust patent deployment choices at a point in time as demand shifts or across time—for example, from commercialization to licensing, or vice versa. In this sense, patent deployment choices can be sensitive not only to the level of product-market demand but also to its volatility. Hence, strategic patent deployment has a real option value that has not previously been analyzed.1 需求侧因素往往比供给侧因素更具动态性，原因有二。首先，行业通常会经历一个相对可预测的产品生命周期，在此周期中，产品的需求水平会增长、成熟并最终下降（Levitt，1965；McGahan，2004）。其次，由于消费者偏好多变、经济冲击影响客户收入、文化潮流时尚等诸多原因，需求也可能以随机或不可预测的方式发生转变。只要随机的需求变化可能影响专利的最优使用，企业就可能从保持灵活性中受益，以便在需求变化或跨时间（例如，从商业化转为许可，或反之亦然）时调整专利部署选择。从这个意义上说，专利部署选择不仅可能对产品市场需求水平敏感，还可能对其波动性敏感。因此，战略性专利部署具有以前未被分析过的实物期权价值。1

Limitations of Prior Research on the Dependent-Variable Side

因变量方面既有研究的局限性

Most prior research has treated the strategic patent deployment phenomenon as a discrete choice between incompatible alternatives—that is, either competitive or cooperative. Yet firms often have more complicated relationships with each other that blend competitive and cooperative elements in what Brandenburger and Nalebuff (1996) called “coopetition.” With regard to patents, a firm may both exploit its patented technology for its own commercial use in the product market while also concurrently licensing the technology to direct competitors.2 For example, Valeo, a French producer of vehicle parking sensor (ultrasonic) technology licenses and supplies its sensor technology to a main European original equipment manufacturer (OEM) auto producer while it concurrently sells parking sensors directly in the auto parts aftermarket and to independent operators. LG, which has developed the superior OLED TV technology, supplies OLED panels to many of its rivals in the consumer electronics market, such as Sony and Panasonic, while simultaneously competing with them in the global TV market. Tesla is currently facing the choice of keeping its electric car battery and driverless technologies for proprietary use or collaborating (e.g., with firms like Panasonic on the car battery) and 大多数先前的研究将战略性专利布局现象视为在相互排斥的替代方案之间的离散选择——即要么具有竞争性，要么具有合作性。然而，企业之间的关系往往更为复杂，它们融合了竞争与合作的元素，这就是Brandenburger和Nalebuff（1996）所说的“竞合”。就专利而言，一家企业可能既将其专利技术用于自身在产品市场的商业用途，同时又将该技术许可给直接竞争对手。例如，法国车辆泊车传感器（超声波）技术生产商法雷奥（Valeo）向欧洲一家主要原始设备制造商（OEM）汽车生产商许可并供应其传感器技术，同时它还直接在汽车零部件售后市场以及向独立运营商销售泊车传感器。LG开发了卓越的OLED电视技术，向消费电子市场中的许多竞争对手（如索尼和松下）供应OLED面板，同时在全球电视市场与它们展开竞争。特斯拉目前面临的选择是将其电动汽车电池和无人驾驶技术用于专有用途，还是进行合作（例如，与松下等企业合作生产汽车电池），以及

2 For the purposes of this analysis, we treat licensing a technology as a cooperative deployment mode and selfcommercialization as a competitive deployment mode. While this may hold in most cases, we nevertheless acknowledge that there may be occasional exceptions. For example, if a firm faces severe capacity constraints (e.g., a startup with limited capital), then self-commercialization may be much less competitive than it would otherwise seem. Conversely, licensing a technology (even on royaltyfree terms) can be less “cooperative” if the license is given with strings attached"—for example, see the restrictive requirements for device manufacturers using Google’s Android operating system. For simplicity, we abstract from such situations as they are the exceptions from the rule. 2 就本次分析而言，我们将技术许可视为合作型部署模式，自主商业化视为竞争型部署模式。尽管这一划分在大多数情况下适用，但我们也承认可能存在偶尔的例外情况。例如，如果一家企业面临严重的产能限制（如资金有限的初创公司），那么自主商业化的竞争力可能会远低于其表面看起来的水平。相反，如果许可附带附加条件（例如，对使用谷歌安卓操作系统的设备制造商提出严格要求），即使是免版税的技术许可，其“合作性”也可能更低。为简化分析，我们暂不考虑这类例外情况，因为它们属于规则的特例。

supplying rival automakers with these technologies. Moreover, coopetition and hybrid strategies may shift over time, as illustrated by the examples in Table 1. Such examples include Genentech’s licensing of its patented synthetic insulin to Eli Lilly in the 1970s, Philips’s and Sony’s collaboration on the CD player standard in 1979, and Dell’s licensing its design technology to IBM and concurrently buying parts from it, while competing head on in the PC market. Similarly, Arora, Fosfuri, and Gambardella (2004: 172173) observed: 为竞争的汽车制造商提供这些技术。此外，竞合与混合策略可能会随时间变化，如表1中的示例所示。这些例子包括：1970年代基因泰克（Genentech）向礼来公司（Eli Lilly）授权其获得专利的合成胰岛素技术；1979年飞利浦（Philips）与索尼（Sony）在CD播放器标准方面的合作；以及戴尔（Dell）向IBM授权其设计技术，同时又从IBM采购零部件，而在个人电脑（PC）市场上又与之直接竞争。类似地，Arora、Fosfuri和Gambardella（2004：172-173）指出：

DuPont and Dow Chemicals, two chemical firms with a long tradition of exploiting technology in-house, have started to license their technology very actively. Indeed, in 1994 DuPont created a division with the specific task of overseeing all technology transfer activities … Dow Chemicals has also long had a reputation for “never licensing breakthrough technology, and there was an emotional bias against licensing” (Ed Gambrell, V.P., Dow Chemicals). In 1995, it formed a licensing group with the purpose to “create more value” from its technology. 杜邦和陶氏化学（Dow Chemicals）是两家长期以来在内部开发技术的化工企业，如今已开始积极对外授权技术。事实上，1994年杜邦成立了一个部门，专门负责监督所有技术转让活动……陶氏化学也长期以“从不授权突破性技术，且对技术授权存在情感偏见”而闻名（陶氏化学副总裁埃德·甘布勒（Ed Gambrell）语）。1995年，该公司成立了一个授权小组，旨在从其技术中“创造更多价值”。

The above study provides numerous reasons why a firm may use a patented technology to compete in the same market while it simultaneously licenses the same technology to a competitor: to create cartellike outcomes, commit to continue innovating, create demand, influence industry standards, deter entry by others, or dissuade rivals from conducting related R&D projects (Arora et al., 2004: 175178, 234). 上述研究提供了众多原因，解释企业为何可能在同一市场中使用专利技术进行竞争，同时又将相同技术许可给竞争对手：以形成类卡特尔的结果、承诺持续创新、创造需求、影响行业标准、阻止其他企业进入市场，或劝阻竞争对手开展相关研发项目（Arora等人，2004：175-178、234）。

This coopetition phenomenon can best be understood via a framework focused on how different patent deployment choices are interdependent: The pros and cons of licensing a patented technology depend critically upon whether the innovator also intends to commercialize the technology for its own use in the product market. Conversely, the pros and cons of commercializing a technology depend upon whether it will also be licensed out. These interrelations suggest that patent deployment choices should be made jointly, and have organizational implications for firms that delegate responsibility for different decisions to separate departments—for example, where engineering, finance, and marketing managers make commercialization choices while lawyers drive licensing choices. In addition, in the interplay between competition and cooperation, the benefit that a firm gains from deploying its patents competitively or cooperatively (or using both concurrently), and shifting between them over time may depend upon the corresponding deployment choices of its rivals. In this regard, the phenomenon of strategic patent deployment involves strategic interactions among rival firms in a game-theoretic sense that also embed various patent use options. 这种竞争合作现象可以通过一个侧重于不同专利部署选择如何相互依存的框架来最好地理解：许可一项专利技术的利弊在很大程度上取决于创新者是否还打算将该技术用于自身产品市场的商业化。相反，一项技术商业化的利弊取决于它是否也会被许可出去。这些相互关系表明，专利部署选择应共同做出，并且对那些将不同决策的责任委托给不同部门的企业具有组织意义——例如，工程、财务和营销经理做出商业化决策，而律师推动许可决策。此外，在竞争与合作的相互作用中，企业通过竞争性或合作性部署其专利（或同时使用两者）以及随时间在两者之间转换所获得的收益，可能取决于其竞争对手的相应部署选择。在这方面，战略性专利部署现象涉及竞争对手之间在博弈论意义上的战略互动，同时也嵌入了各种专利使用选项。

Addressing Both Limitations Together

同时解决这两个局限性

The issues on the dependent-variable side of the equation cannot easily be separated from those on the independent-variable side. After all, the level and volatility of market demand can influence the incentives of rival firms to compete, cooperate, do both concurrently, switch modes over time, as well as to enter or exit (Cabral, 2000). So it is necessary to address these two issues together, rather than separately. As a first step toward mitigating the limitations of extant patenting research on both the dependent- and independent-variable sides of the equation, we develop a theory that integrates dynamic demand-side factors and allows for coopetition in patent deployment choices as well as for shifting among competition and cooperation modes over time to better capture the interplay between competition and cooperation in the patent deployment context. Recognizing that both the strategic interaction among competitors and the flexibility to respond to demand volatility affect rival firms’ patent deployment choices, we use a real-options game framework (Smit & Trigeorgis, 2004, 2017) as the basis of our theory. Our analysis further considers that both competition and cooperation can be implemented in multiple ways. Competition can take the form of a defensive or offensive patent deployment strategy. The first strategy involves clustering or building a “patent wall” of related patents surrounding one’s own core patent (Arundel & Patel, 2003; Rv , i y involves “bracketing” or surrounding a rival’s core patent with related patents of one’s own (Cohen et al., 2000; Reitzig, 2004b; Rivette & Kline, 2000a). Likewise, cooperation may involve unilateral licensing of one’s technology, as in the case of Genentech licensing its synthetic insulin patent to Eli Lilly, or bilateral cross-licensing of patents, as in the case of IBM and Dell in 1999, GSK and Nuevolution in 2009, or Google and Samsung in 2014 (see Table 1 for details). 方程中因变量侧的问题难以与自变量侧的问题轻易分离。毕竟，市场需求的水平和波动性会影响竞争对手的竞争、合作、同时进行这两种行为、随时间切换模式以及进入或退出（Cabral，2000）的动机。因此，有必要同时处理这两个问题，而不是分别处理。作为减轻现有专利研究在方程的因变量和自变量两侧局限性的第一步，我们构建了一个理论，该理论整合了动态需求侧因素，并允许在专利部署选择中进行竞合，以及随时间在竞争和合作模式之间转换，以更好地捕捉专利部署环境中竞争与合作之间的相互作用。认识到竞争对手之间的战略互动以及应对需求波动的灵活性都会影响竞争对手的专利部署选择，我们以实物期权博弈框架（Smit & Trigeorgis，2004，2017）为理论基础。我们的分析进一步考虑到，竞争和合作都可以通过多种方式实施。竞争可以采取防御性或进攻性专利部署策略的形式。第一种策略涉及围绕自身核心专利进行聚类或构建“专利墙”（Arundel & Patel，2003；Rv , i y 涉及“包围”或用自身相关专利围绕竞争对手的核心专利（Cohen et al.，2000；Reitzig，2004b；Rivette & Kline，2000a）。同样，合作可能包括单方面许可自己的技术，例如基因泰克将其合成胰岛素专利许可给礼来公司，或者专利的双边交叉许可，例如1999年IBM和戴尔、2009年葛兰素史克（GSK）和Nuevolution，或2014年谷歌和三星（详情见表1）。

Our analysis sheds light on the conditions under which patent deployment strategies may optimally switch between competition and cooperation or coopetition at a point in time depending on the level of market demand (i.e., across demand states) or across time periods depending on the volatility of market demand, given the strength of the innovator’s patented technology, the bargaining power of the innovator versus the incumbent in licensing negotiations, and other factors. A main implication of our analysis is that the impact of patent deployment mode on rivalry in industry equilibrium can have counterintuitive effects on the patent’s value. For example, it is possible that a higher patent technological strength might actually backfire and reduce the patent’s value by pushing the industry into a higher-rivalry state where firms deploy their patents in more aggressive competitive ways. This is a key prediction that impacts directly on the interplay between competition and cooperation and the balance between value creation at the industry level and firm-level appropriation: deploying a competitive advantage residing in patent technological strength aggressively in pursuit of value appropriation may invite rival responses that damage profit value creation and value maximization. The form of competition may also become more offensive (moving from patent wall to bracketing) as demand increases when the technology is strong. More interestingly, at medium technological strength, hybrid patent deployment strategies may result, for example, in involving competition via offensive bracketing at high demand and a switch to cooperation via licensing at low or medium demand, which implies a higher option value due to the ability to switch deployment modes—especially when market volatility and the innovator’s bargaining power are high. Further, the theory predicts that the value of the option to switch patent deployment modes is greater under the combination of three factors: when the innovator’s technology is moderate in strength, its bargaining power is high, and market volatility is high. Under low market volatility, the current patent deployment mode is likely to persist across time periods. In addition, under the above conditions licensing is more likely to occur when demand is either very high (possibly used in combination with self-commercialization) or low (used by itself), but not medium, suggesting that extreme demand volatility favors cooperation via licensing. That is, high demand volatility, increasing the range of extreme high and low demand states, makes licensing more beneficial. Conversely, low demand volatility would likely favor (game-theoretic) competitive patent deployment to exploit own technological advantage. 我们的分析揭示了在给定创新者专利技术强度、创新者与在位者在许可谈判中的议价能力以及其他因素的情况下，专利部署策略可能会根据市场需求水平（即不同需求状态）或随时间变化（取决于市场需求波动）在某一时刻最优地在竞争、合作或竞合之间切换的条件。我们分析的一个主要含义是，专利部署模式对行业均衡中竞争的影响可能会对专利价值产生反直觉的影响。例如，较高的专利技术强度实际上可能适得其反，通过将行业推向更高竞争状态（企业以更激进的竞争方式部署其专利）而降低专利价值。这是一个关键预测，直接影响竞争与合作之间的相互作用以及行业层面价值创造与企业层面获取之间的平衡：为追求价值获取而积极部署专利技术优势中的竞争优势，可能会引发对手的回应，损害利润价值创造和价值最大化。当技术实力较强时，随着需求增加，竞争形式可能变得更具进攻性（从专利壁垒转向围堵）。更有趣的是，在中等技术强度下，混合专利部署策略可能会出现，例如，在高需求时通过进攻性围堵进行竞争，在低或中等需求时通过许可转向合作，这意味着由于能够切换部署模式而具有更高的期权价值——尤其是当市场波动性和创新者的议价能力较高时。此外，该理论预测，在以下三个因素的组合下，切换专利部署模式的期权价值更大：创新者的技术强度中等、其议价能力高且市场波动性高。在市场波动性较低时，当前的专利部署模式可能会在不同时间段持续存在。此外，在上述条件下，当需求非常高（可能与自我商业化结合使用）或低（单独使用）时，许可更有可能发生，而在中等需求时则不太可能，这表明极端需求波动有利于通过许可进行合作。也就是说，高需求波动增加了极端高和低需求状态的范围，使许可更有利。相反，低需求波动可能更有利于（博弈论意义上的）竞争性专利部署，以利用自身技术优势。

TABLE 1
表1

TABLE 1 (Continued)
表1（续）

The remainder of this article is organized as follows: We first derive our theory’s basic insights via a baseline numerical illustration using real option games. This approach aims to provide broad accessibility for a general audience and to incorporate a wide range of possible patent deployment options, such as commercializing the technology for own exclusive use, licensing the technology, crosslicensing patents, building a defensive patent wall, or engaging in an offensive bracketing war. The numerical analysis focuses on switching patent deployment modes across states (levels) of market demand at a given time. The next section supplements this inquiry with an extended analysis focused on switching patent deployment modes across time periods. To make the analysis tractable, we focus on a narrower range of just three deployment modes: commercializing the technology oneself (competition), licensing the patent to an incumbent (cooperation), or both concurrently (coopetition). The extended analysis is general in terms of method and parameter values, thus providing more rigorous sensitivity analysis and helping alleviate any concerns that our baseline numerical insights might be an artifact of specific parameter value choices. The final section discusses the broader implications of our study for theory, empirics, and practice. Details of our baseline numerical illustration and of our extended analysis are provided in Online Appendices A and B, respectively. 本文其余部分的组织方式如下：我们首先通过使用实物期权博弈的基准数值说明来推导我们理论的基本见解。这种方法旨在为普通受众提供广泛的可及性，并纳入各种可能的专利部署选项，例如将技术商业化以独家自用、许可技术、交叉许可专利、建立防御性专利壁垒，或参与进攻性包围战。数值分析侧重于在给定时间点跨市场需求状态（水平）切换专利部署模式。下一部分通过扩展分析补充这一研究，该扩展分析侧重于跨时间段切换专利部署模式。为了使分析易于处理，我们将范围缩小到仅三种部署模式：自行将技术商业化（竞争）、向在位者许可专利（合作），或同时进行这两种方式（竞合）。扩展分析在方法和参数值方面具有一般性，因此提供了更严格的敏感性分析，并有助于缓解我们的基准数值见解可能是特定参数值选择的产物这一担忧。最后一部分讨论了我们的研究对理论、实证和实践的更广泛意义。我们的基准数值说明和扩展分析的详细内容分别在在线附录A和B中提供。

BASELINE NUMERICAL ILLUSTRATION: STRATEGIC PATENT DEPLOYMENT

基准数值说明：战略性专利部署

In this section we provide a baseline numerical illustration of the patent deployment strategies of a technological innovator with a new patented technology facing an incumbent with an existing patented technology. Our analysis here focuses on switching between competition and cooperation modes depending upon the state of demand and the strength of the innovator’s technology. By patent deployment we mean an innovative firm’s set of choices about how to profit from its innovation and specifically which mode, if any, it chooses to deploy its patent—for example, competition (via defensive patent wall or offensive patent bracketing), cooperation (via licensing the technology or cross-licensing patents), or delaying its choice of competition versus cooperation until more information about market demand is revealed. We examine conditions where patent deployment strategies involve switching among these modes, as a function of demand shifts, demand volatility, and the relative strength of the innovator’s new technology. We abstract away from other considerations, such as the need to prove the technology before licensing or facing high upfront and later declining technology integration costs that may justify pivoting from competition to licensing (Marx et al., 2014). 在本节中，我们提供一个基准数值示例，说明拥有新技术专利的技术创新者面对拥有现有技术专利的在位企业时的专利部署策略。我们的分析重点是根据需求状态和创新者技术的强度，在竞争和合作模式之间进行切换。所谓专利部署，我们指的是创新型企业关于如何从其创新中获利的一系列选择，具体而言，它选择以何种模式（如果有的话）部署其专利——例如，竞争（通过防御性专利壁垒或进攻性专利包围）、合作（通过技术许可或交叉许可专利），或者在更多市场需求信息披露后再决定竞争还是合作。我们研究专利部署策略涉及根据需求变化、需求波动性以及创新者新技术的相对强度在这些模式之间切换的情况。我们暂不考虑其他因素，例如在许可前需要证明技术可行性，或面临高昂的前期成本以及后期下降的技术整合成本（这些成本可能证明从竞争转向许可是合理的）（Marx等人，2014）。

Assumptions

假设

We consider a situation where two firms, firm $T$ that innovates a new patented technology and incumbent firm I with an existing patented technology, are involved in a two-period strategic patent deployment game. For simplicity, we here assume the innovator’s patented technology results in one of three possible levels of technology improvement, labeled as weak, medium, or strong technology, where a strong technology can improve the firm’s competitiveness sufficiently to give it a monopoly position. The timing of the game is as follows: 我们考虑这样一种情况：两家企业参与了一个两期战略专利部署博弈，其中企业T创新出一项新的专利技术，而在位企业I拥有一项现有的专利技术。为简化分析，我们假设创新者的专利技术会带来三种可能的技术改进水平，分别标记为弱技术、中等技术或强技术，其中强技术能够充分提升企业的竞争力，使其获得垄断地位。博弈的时间安排如下：

1. In period 0, innovative firm $T$ patents a new superior technology. Assuming equal prior market power, this gives firm $T$ an asymmetric cost advantage over firm I.

2. 在第0期，创新企业T申请了一项新的先进技术专利。假设之前市场势力相同，这使得企业T相对于企业I拥有不对称的成本优势。

3. In period 1, the market experiences an initial demand shock that shifts demand either up or down.

4. 在第1期，市场经历了初始需求冲击，导致需求向上或向下移动。

5. In period 2, the market experiences a second demand shock that again shifts demand up or down. The cumulative effect of the two demand shocks leaves the industry in one of three possible demand states at time 2: low demand (two downward shocks), medium demand (one downward and one upward shock), or high demand (two upward shocks). Then each firm chooses its optimal patent deployment strategy (competing, cooperating, or waiting), depending on innovative firm $T \boldsymbol s$ relative technology strength (weak, medium, or strong) and the observed realization of market demand (low, medium, or high). A final demand shock occurs at time 3, with demand thereafter assumed to remain in steady state, so any firm that choses the waiting option (patent sleep) during period 2 receives a continuation option value $C$ .

6. 在第2期，市场经历第二次需求冲击，再次使需求上升或下降。两次需求冲击的累积效应使该行业在第2期处于三种可能的需求状态之一：低需求（两次向下冲击）、中等需求（一次向下和一次向上冲击）或高需求（两次向上冲击）。然后，每家企业根据创新企业T的相对技术实力（弱、中或强）和观察到的市场需求实现情况（低、中或高），选择其最优专利部署策略（竞争、合作或等待）。第3期会发生最终需求冲击，此后需求被假定保持稳态，因此在第2期选择等待选项（专利休眠）的任何企业将获得持续期权价值C。

Our baseline numerical analysis focuses on the Nash equilibrium deployment choices made by each firm in period 2 based on option games analysis (Chevalier-Roignant & Trigeorgis, 2011; Smit & Trigeorgis, 2004, 2017). We assume that both firms’ technologies are publicly known and that patent rights are perfectly enforced. Expected cash flows are discounted across periods to determine current market value (the underlying “asset” for the options) by using the risk-adjusted cost of capital $\rho$ while expected option values in the option games are based on risk-adjusted (or risk-neutral) probabilities and discounted at the risk-free interest rate $r$ (Chevalier-Roignant & Trigeorgis, 2011; Smit & Trigeorgis, 2004, 2017). 我们的基准数值分析聚焦于各企业在第二阶段基于期权博弈分析（Chevalier-Roignant & Trigeorgis, 2011; Smit & Trigeorgis, 2004, 2017）所做出的纳什均衡部署选择。我们假设两家企业的技术均为公开信息，且专利权得到完美保护。预期现金流会跨期折现，以确定当前市场价值（即期权的“标的资产”），折现时采用风险调整后的资本成本$\rho$；而期权博弈中的预期期权价值则基于风险调整（或风险中性）概率，并以无风险利率$r$进行折现（Chevalier-Roignant & Trigeorgis, 2011; Smit & Trigeorgis, 2004, 2017）。

TABLE 2 Parameters and Values Used in the Baseline Numerical Illustration
表2 基准数值说明中使用的参数和取值

Table 2 lists the parameters and parameter values used in our baseline numerical analysis. $V$ is the total market value for the industry, computed as the (gross) present value of expected cash inflows from commercializing the patented technology. Its initial (current) value $V _ { O }$ is the total market value at period 0, before any demand shocks, which we assume is $$ 1000$ Parameters $u$ and $d$ are the multiplicative factors specifying the magnitude of upward and downward movements for each demand shock (and hence the range of demand), with $\sigma = \ln u$ being the demand volatility, while $p$ and $1 - p$ are the riskadjusted (or risk-neutral) probabilities of an upward and downward demand shock movement, respectively. $V ^ { + + }$ , $V ^ { + - }$ , and $V ^ { — }$ are, respectively, the three possible market values observable at time 2 after either two upward demand shocks (high demand), one upward and one downward shock (medium demand), or two downward shocks (low demand). Exercise mode factor $m$ is a multiplier by which total market value is enlarged if the firms cooperate, or reduced if they compete aggressively. Under cooperation, we assume $m = c = 1 . 2$ , where $c$ refers to “cooperation.” Under competition, we assume $m = f = 0 . 7$ , where $f$ refers to “fight.” In terms of value capture, parameters $s$ (for firm T), and ${ 1 - s }$ (for firm I), represent the firms’ respective shares (based on relative market power) of the total market value $V$ or the continuation option value $C$ which represents the expected payoff from waiting until one more period (and settling in steady state thereafter). As explained in Online Appendix A, these market power shares depend on the degree of firm $T \boldsymbol s$ relative technology advantage. 表2列出了我们基准数值分析中使用的参数和参数值。$V$ 是行业的总市值，计算为商业化专利技术的预期现金流入的（总）现值。其初始（当前）值 $V_0$ 是第0期的总市值，在任何需求冲击之前，我们假设该值为1000美元。参数 $u$ 和 $d$ 是指定每次需求冲击的向上和向下变动幅度（以及需求范围）的乘数因子，其中 $\sigma = \ln u$ 是需求波动率，而 $p$ 和 $1-p$ 分别是需求冲击向上和向下变动的风险调整（或风险中性）概率。$V^{++}$、$V^{+-}$ 和 $V^{–}$ 分别是在经历两次向上需求冲击（高需求）、一次向上和一次向下冲击（中需求）或两次向下冲击（低需求）后，在时间2可观测到的三种可能的市值。行权模式因子 $m$ 是一个乘数，如果企业合作，总市值将被放大，而如果企业进行激烈竞争，总市值将被缩小。在合作的情况下，我们假设 $m = c = 1.2$，其中 $c$ 指“合作”。在竞争的情况下，我们假设 $m = f = 0.7$，其中 $f$ 指“竞争”。在价值捕获方面，参数 $s$（代表企业T）和 $1-s$（代表企业I）表示企业各自的市场价值份额（基于相对市场力量），即总市值 $V$ 或延续期权价值 $C$，其中 $C$ 表示等待到下一个时期（之后进入稳态）的预期收益。如在线附录A中所述，这些市场力量份额取决于企业T的相对技术优势程度。

Implementing a patent deployment strategy requires incurring an investment cost. Parameter $I$ $\scriptstyle ( = $ 80\mathrm { m } )$ is the baseline investment cost (e.g., for entry or integration) for a firm to commercialize its technology. (Under low demand, the cost is $\overset { \vartriangle } { \boldsymbol { I } } = $ 400$ ) This cost is shared if the two firms cooperate, is increased if they compete aggressively, or is delayed if they let their patents sleep. As noted, cooperation may take the form of either patent 实施专利部署策略需要投入成本。参数 \( I \) \( \scriptstyle ( = \$ 80\mathrm { m } ) \) 是企业将其技术商业化的基准投资成本（例如，用于市场进入或整合）。（在需求较低的情况下，成本为 \( \overset { \vartriangle } { \boldsymbol { I } } = \$ 400 \)）。如果两家企业合作，该成本将由双方分担；如果它们进行激烈竞争，成本将增加；如果它们让专利闲置，成本将被推迟。如前所述，合作可能采取专利

DEMAND 需求

FIGURE 1 Patent Deployment Strategies (Compete or Cooperate) Contingent on Technology Strength and State of Demand (for Base Uncertainty $\mathbf { \sigma _ { 0 } } = \mathbf { 6 0 % }$ in the Numerical Illustration 图1 专利部署策略（竞争或合作）取决于技术实力和需求状态（数值示例中基础不确定性 $\mathbf{\sigma_0} = \mathbf{60%}$）

licensing or cross-licensing, while competition may involve a defensive patent wall or offensive patent bracketing. Both competition modes increase the cost for the rival firms due to costly multiple patent filings, via a cost multiplier, while a patent wall also strengthens one’s market position, as reflected in a market value $V$ increase implemented through a multiplier m. Each payoff at (maturity) time 2 thus takes the form of a conditional net present value (NPV) of the general form $m V - z I$ ,where $m$ and $z$ are the value and cost multipliers, respectively. 许可或交叉许可，而竞争可能涉及防御性专利壁垒或进攻性专利包围。两种竞争模式都会因昂贵的多重专利申请而增加竞争对手的成本（通过成本乘数），同时专利壁垒还会增强一方的市场地位，这反映在通过乘数m实现的市场价值$V$增加中。因此，在到期时间2时的每个收益都采用一般形式的条件净现值（NPV）：$m V - z I$，其中$m$和$z$分别是价值乘数和成本乘数。

The three possible patent technology strengths (weak, medium, and strong) combined with the three possible demand levels (low, medium, and high) at time 2 yield nine possible option-game scenarios, shown in Figure 1. For these nine scenarios, Figure 2, Panel A shows the nine corresponding subgames, numbered 1 through 9, that the two firms might play. In each subgame, each firm chooses whether to invest in deploying its patent or to delay this investment, with the latter choice being labeled as “sleep.” An illustration of patent sleep is the case of Endovascular Technologies (EVT) and Guidant shown in Table 1. Keeping one’s patent sleeping amounts to delaying the deployment mode decision until next period (time 3). Holding a sleeping patent is a wait-and-see option that has more value when demand is volatile—specifically, continuation value $C$ represents a call option on the period 3 market value. If both firms let their patents sleep, they appropriate portions of $C$ in proportion to their market power shares, $s$ and ${ 1 - s }$ If a firm does not let its patent sleep, it can make four possible types of investment—cooperative investment via patent licensing (unilateral) or cross-licensing (bilateral) versus competitive investment via a patent wall (defensive) or bracketing (offensive). It is assumed that cooperation results in enlarged market value pie (here by $2 0 %$ ) as the collaborators capture and share monopoly-like profits by maintaining high prices. By contrast, aggressive competition (defensive or offensive) results in a reduced market value pie (here by $3 0 %$ )due to the ensuing product market rivalry and costly patent wars. 时间2时，三种可能的专利技术强度（弱、中、强）与三种可能的需求水平（低、中、高）组合，产生了九种可能的选项博弈场景，如图1所示。针对这九种场景，图2的A面板展示了两家企业可能进行的九个相应子博弈，编号为1至9。在每个子博弈中，每家企业选择是否投资部署其专利或延迟该投资，后者选择被标记为“休眠”。专利休眠的一个例子是表1中展示的Endovascular Technologies（EVT）和Guidant的案例。让专利处于休眠状态相当于将部署模式决策推迟到下一期（时间3）。持有休眠专利是一种观望期权，在需求波动时更具价值——具体而言，延续价值\( C \)代表对第3期市场价值的看涨期权。如果两家企业都让其专利休眠，它们将根据其市场势力份额\( s \)和\( 1 - s \)按比例分配\( C \)的部分。如果一家企业不让其专利休眠，它可以进行四种可能类型的投资——通过专利许可（单边）或交叉许可（双边）的合作投资，以及通过专利壁垒（防御性）或 bracketing（进攻性）的竞争性投资。假设合作会扩大市场价值蛋糕（此处扩大20%），因为合作者通过维持高价来捕获和分享类似垄断的利润。相比之下，激进竞争（防御性或进攻性）会因随之而来的产品市场竞争和昂贵的专利战而减少市场价值蛋糕（此处减少30%）。

FIGURE 2 Panel A. Subgame Payoffs (Panel A) and Nash Equilibria (Panel B) Depending on Demand and Technology Strength 图2 面板A. 子博弈收益（面板A）和纳什均衡（面板B）取决于需求和技术优势

Investing under a cooperative mode involves either unilaterally licensing out one’s patent to the rival, or cross-licensing patents with the rival when both firms invest. Under cross-licensing, both firms may still compete with each other in the product market while behaving cooperatively in the technology market. Thus, in a way cross-licensing may be interpreted as a form of coopetition as it combines elements of both cooperation and competition. The driving force of the sharing terms of collaboration among the firms is the relative market power based on the advantage of firm T’s patented technology relative to that of firm I. For example, if firm $T \boldsymbol s$ technology is strong, it appropriates most $( s = 7 5 %$ of conditional period 2 NPV (of the form $m V - z I )$ or $C$ while firm I gets the remainder. If firm $T s$ technology is medium, it gets $6 0 %$ while firm I gets $4 0 %$ . If firm $T$ has a weak technology advantage over firm I’s existing patented technology, sharing is assumed to be 5050. Parameter $F$ (expressed as a percentage of market value) is the licensing fee that one firm pays to the other when licensing its technology $\boldsymbol { \cdot } \boldsymbol { F } ^ { \prime }$ is the licensing fee under low market demand). 在合作模式下进行投资，涉及单方面向竞争对手许可自身专利，或在双方企业均进行投资时与竞争对手交叉许可专利。在交叉许可模式下，双方企业在产品市场上仍可能相互竞争，但在技术市场上则采取合作行为。因此，从某种意义上说，交叉许可可被视为一种竞合（coopetition）形式，因为它结合了合作与竞争的要素。企业间合作条款共享的驱动力是基于企业T的专利技术相对于企业I的专利技术所具有优势的相对市场力量。例如，如果企业T的技术较强，它将获得大部分（s = 75%的条件期2净现值（NPV，形式为mV - zI）或C），而企业I获得剩余部分。如果企业T的技术处于中等水平，它将获得60%，而企业I获得40%。如果企业T相对于企业I现有的专利技术具有较弱的技术优势，则假设共享比例为50:50。参数F（以市场价值的百分比表示）是一方企业在许可其技术时向另一方支付的许可费。F’是低市场需求下的许可费。

FIGURE 2 Panel B
图2 面板B

Note: This panel uses parameter values from Table 2 applied to the equations of Panel A of Figure 2. 注：此面板使用表 2 中的参数值，应用于图 2 面板 A 的公式。

Under competition, investing might involve either defensively building a wall of related patents surrounding one’s own core patent to protect against encroachment, or offensive patent bracketing surrounding the rival’s patent with one’s own related patents to constrain the rival. To illustrate a patent wall, see the examples of Xerox versus IBM and Gillette versus BIC in Table 1. To illustrate bracketing, see Yamaha versus Bombardier in Table 1. In our numerical analysis, the main difference between these two forms of competition is that a patent wall provides both costs and benefits, while a bracketing war incurs costs without tangible benefits. With a patent wall, the investment cost is higher, multiplied by $W _ { T } \left( = 1 . 2 \right)$ for firm $T$ and by $W _ { I } \left( = 1 . 3 \right)$ for firm I, while the market value is also higher, multiplied by $W _ { T } ^ { \prime } \left( = 1 . 2 \right)$ for firm $T$ and by $W _ { I } ^ { \prime } ( = 2 . 2 )$ for firm I. If the two firms compete by engaging in an offensive patent bracketing war, their respective investment costs are increased to $b I _ { \mathrm { i } }$ where $b \left( = 1 . 3 \right)$ is the bracketing cost multiplier, without any compensating tangible benefits. 在竞争中，投资可能包括两种策略：一种是防御性地围绕自身核心专利构建相关专利壁垒，以抵御侵犯；另一种是进攻性地用自身相关专利包围竞争对手的专利，以限制对手。为说明专利壁垒，可参考表1中施乐（Xerox）与IBM以及吉列（Gillette）与比克（BIC）的案例；为说明包围策略，可参考表1中雅马哈（Yamaha）与庞巴迪（Bombardier）的案例。在我们的数值分析中，这两种竞争形式的主要区别在于：专利壁垒既能带来成本也能带来收益，而包围战只会产生成本，无实质性收益。对于专利壁垒，投资成本更高，企业T的成本乘以\(W_T(=1.2)\)，企业I的成本乘以\(W_I(=1.3)\)；同时市场价值也更高，企业T的市场价值乘以\(W_T'(=1.2)\)，企业I的市场价值乘以\(W_I'(=2.2)\)。如果两家企业通过发起进攻性专利包围战竞争，它们各自的投资成本将增加到\(bI_i\)，其中\(b(=1.3)\)是包围成本乘数，且没有任何补偿性的实质性收益。

In each of the nine subgames in Figure 2, some of these investment types are dominated by other types. So, to keep Figure 2, Panel A as simple as possible, we omit the dominated investment types. Thus, most of the subgames show only two possible types of investments—one when only one firm invests, and another when both invest. For example, in subgame 3, when firm T’s technology is strong and demand is high, the relevant investment is in a defensive patent wall if one firm invests, or in offensive bracketing if both invest—with both types being competitive investments. By contrast, at the opposite extreme, in subgame 7 where firm T’s technology is weak and demand is low, the relevant investment is in licensing if one firm invests, or in cross-licensing if both invest—with both types being forms of cooperative investment. The most interesting case, as we analyze later, is in the middle, where the “hybrid” subgame 5 includes more than two possible investment modes that span across both competitive and cooperative investment types (bracketing when both firms invest and licensing when only one firm invests). The payoff formulas shown in Figure 2, Panel A are explained and derived in Online Appendix A (see Table A1). Each firm’s payoff represents the value of net cash flows (or option value for patent sleep) generated by its chosen deployment strategy when met with its rival’s strategy. We use these payoffs to determine what deployment strategies (competition, cooperation, or wait-and-see) the firms optimally pursue, depending on the strength of the innovator’s new technology and the state and volatility of demand. 在图2的九个子博弈中，每种子博弈里都存在一些投资类型被其他类型所主导。因此，为了让图2的A面板尽可能简洁，我们省略了被主导的投资类型。因此，大多数子博弈只显示两种可能的投资类型——一种是只有一家企业投资，另一种是两家企业都投资。例如，在子博弈3中，当企业T的技术实力较强且需求较高时，若只有一家企业投资，相关投资是建立防御性专利壁垒；若两家企业都投资，则是进攻性交叉许可——这两种类型都是竞争性投资。相比之下，在另一端的极端情况中，子博弈7里企业T的技术实力较弱且需求较低，若只有一家企业投资，相关投资是许可；若两家企业都投资，则是交叉许可——这两种类型都是合作性投资的形式。正如我们稍后分析的那样，最有趣的情况是处于中间位置的子博弈5，它包含了两种以上的可能投资模式，这些模式横跨竞争性和合作性投资类型（两家企业都投资时采用交叉许可，只有一家企业投资时采用许可）。图2的A面板中显示的收益公式在在线附录A中进行了解释和推导（见表A1）。每家企业的收益代表了当面对竞争对手的策略时，其选定的部署策略所产生的净现金流价值（或专利休眠的期权价值）。我们利用这些收益来确定企业根据创新者新技术的强度以及需求的状态和波动性，会最优地采取哪些部署策略（竞争、合作或观望）。

Overview of Method

方法概述

For the remainder of this baseline numerical analysis, we substitute the parameter values from Table 2 into the payoff formulas in Figure 2, Panel A, which yields the numerical payoffs shown in Figure 2, Panel B that we use hereafter. The numerical analysis proceeds in several steps: First, innovative firm T’s patent is valued using a conventional NPV analysis. Second, based on the strategic interactions between the two firms, the full value (strategic NPV [S-NPV]) of strategically exploiting the patent on the new technology is obtained via the option games approach using binomial trees, based on the Nash equilibrium for each of the nine subgames (i.e., with each firm optimizing its choice conditional upon the other firm doing so as well). Based on these industry equilibria for each of the nine subgames in period 2 prevailing in different states of demand under different cases of technology strength, the initial (i.e., period 0) value of the patent strategy for innovator firm $T$ is then determined by “averaging out” the equilibrium payoffs across different demand scenarios and “folding back” using option pricing based on the notion of S-NPV, a valuation construct developed by Smit and Trigeorgis (2017). Beyond accounting for the traditional source of committed value captured in an NPV analysis (i.e., the value of expected discounted cash flows from preset operations and strategies), S-NPV captures the interplay among sources of value from commitment under competition (based on game theory) and flexibility under uncertainty (based on real options), yielding a more complete representation of the various value components and trade-offs from competing courses of action. Subsequently, sensitivity and robustness analyses are presented and discussed. 在本基准数值分析的其余部分中，我们将表2中的参数值代入图2 Panel A中的收益公式，从而得到图2 Panel B中所示的数值收益，我们今后将使用这些收益。数值分析分以下几个步骤进行：首先，利用传统的净现值（NPV）分析对创新企业T的专利进行估值。其次，基于两家企业之间的战略互动，通过二项树期权博弈方法，基于九个子博弈中的每一个的纳什均衡（即每家企业在另一方也进行优化选择的条件下优化自身选择），获得战略性开发新技术专利的全部价值（战略净现值[S-NPV]）。在不同技术实力情况下，针对不同需求状态下在第2期存在的九个子博弈中的每个子博弈的行业均衡，创新企业T的专利策略的初始（即第0期）价值随后通过“平均化”不同需求情景下的均衡收益，并基于S-NPV（由Smit和Trigeorgis（2017）开发的估值结构）的期权定价进行“回溯”来确定。除了考虑NPV分析中捕获的传统承诺价值来源（即预设运营和策略的预期贴现现金流的价值）外，S-NPV还捕获了竞争下承诺价值来源（基于博弈论）与不确定性下灵活性来源（基于实物期权）之间的相互作用，从而更完整地表示了不同行动方案的各种价值组成部分和权衡。随后，我们将呈现并讨论敏感性和稳健性分析。

Analysis

分析

We start with the conventional patent valuation for innovator firm $T .$ Based on NPV, the patent’s value for firm $T$ is obtained by discounting its expected cash flows (net of “entry” cost $I$ of $$ 800$ back to time $t = 0$ using the cost of capital $( \rho = 2 0 %$ . Expectations are taken by assigning appropriate probabilities to each demand scenario at period 2, resulting in a gross present value of the patent cash flows $V$ of $$ 1000$ .The static NPV of the patent is thus $$ 200\ 0$ . This static analysis ignores the options embedded in the patent deployment game, which we address next. 我们从创新企业$T$的传统专利估值开始。基于净现值（NPV），专利对企业$T$的价值是通过将其预期现金流（扣除“进入”成本$I$（800美元）后）使用资本成本（$\rho=20%$）折现回时间$t=0$得到的。通过在第2期为每个需求情景分配适当的概率来获得预期，这使得专利现金流的总现值$V$为$1000$美元。因此，专利的静态NPV为$2000$美元。这种静态分析忽略了专利部署博弈中嵌入的期权，我们接下来将对此进行探讨。

The patent will have higher value if it can be used strategically, either against or to also benefit the competitor. In terms of the Nash equilibria of each subgame, Online Appendix A derives the firms’ optimal patent deployment strategies at time 2 for each of the nine technology-demand combinations in Figure 2, Panel B. Three observations are noteworthy from this analysis. First, when there is weak technology advantage and rivals are symmetric in market power, cooperation is a natural equilibrium outcome: here, firms cooperate via licensing or cross-licensing—see subgames 1, 4, and 7 in Figures 1 and 2. This is consistent with prior literature, which has suggested that cross-licensing is appropriate when patent portfolios and players are similar in strength. Mihm et al. (2015) showed that this holds when both firms are R&D leaders or both are R&D imitators. Such collaboration under symmetry is also observed in practice, as illustrated by the examples of IBM-Dell or Google-Samsung in Table 1.3 如果专利能被战略性地运用，无论是用来对抗竞争对手还是使其受益，它的价值都会更高。就每个子博弈的纳什均衡而言，在线附录A针对图2 Panel B中的九种技术需求组合，推导出了企业在时间2时的最优专利部署策略。从这一分析中可以注意到三个观察结果。首先，当存在较弱的技术优势且竞争对手在市场力量上对称时，合作是一种自然的均衡结果：此时，企业通过许可或交叉许可进行合作——见图1和图2中的子博弈1、4和7。这与先前的文献一致，该文献表明，当专利组合和参与者实力相当时，交叉许可是合适的。Mihm等人（2015）表明，当两家企业都是研发领先者或都是研发模仿者时，这一情况成立。这种在对称条件下的合作在实践中也有观察到，如表1中IBM-Dell或Google-Samsung的例子所示。

The second observation is that, at the opposite extreme, if the innovator’s technology advantage is strong, the competition mode is more likely (subgames 3, 6, and 9 in Figures 1 and 2). As shown in Online Appendix A, the precise type of competitive patent strategy may differ across demand regimes— ranging from a wait-and-see strategy (patent sleep) in low demand to aggressive competition (e.g., patent wall or bracketing) in medium or high demand. This idea that the equilibrium changes when demand changes is also found in Cabral’s (2000) textbook example of simultaneous entry with fixed cost. We view this result as an interesting case of that general principle—interesting because the form of competition becomes more offensive (i.e., moving from patent wall to bracketing) as demand increases under strong technology, as in the example of Yamaha and Bombardier in Table 1.4 Intuitively, when demand is intermediate and there is a risk that one of the firms will be preempted, firms follow a defensive competitive patent strategy by building a patent wall in the hope of basic survival; but at higher levels of demand when both firms can readily survive, competition becomes more offensive (via bracketing) in an effort to attain higher duopoly gains. 第二个观察结果是，在另一端，若创新者的技术优势较强，竞争模式更有可能出现（图1和图2中的子博弈3、6和9）。正如在线附录A所示，竞争专利策略的具体类型可能因需求状态而异——从低需求时的观望策略（专利休眠）到中等或高需求时的激进竞争（例如，专利壁垒或分档竞争）。卡布拉尔（2000）教科书中关于固定成本下同时进入的例子也体现了均衡会随需求变化而改变这一观点。我们将这一结果视为该一般原则的一个有趣案例——有趣之处在于，在技术优势较强的情况下，随着需求增加，竞争形式会变得更具进攻性（即从专利壁垒转向分档竞争），如表1.4中亚马逊和庞巴迪的例子所示。直观地说，当需求处于中间水平且存在一方企业被先发制人的风险时，企业会采取防御性的竞争专利策略，通过建立专利壁垒来争取基本生存；但在需求更高的情况下，当两家企业都能轻松生存时，竞争会通过分档竞争变得更具进攻性，以获取更高的双寡头收益。

The third, and most interesting, observation is that under medium technological strength, hybrid patent deployment strategies may arise—for example, involving competition via offensive bracketing at 第三个，也是最有趣的观察结果是，在技术实力中等的情况下，可能会出现混合专利部署策略——例如，涉及通过进攻性包夹进行竞争，在

4 As demand for personal watercraft grew rapidly, Yamaha bracketed its rivals by obtaining 100 patents on feature improvements around its designs. In 2001, it claimed that Bombardier’s new products infringed on its patents. 4 随着个人水上摩托需求的快速增长，雅马哈通过围绕其设计获得100项功能改进专利，领先于竞争对手。2001年，该公司声称庞巴迪的新产品侵犯了其专利。

high demand and a switch to cooperation via licensing at low or medium demand. In case of medium technology strength under high demand (subgame 2 in Figures 1 and 2), each firm has a dominant strategy to invest involving a competition mode (again, see Yamaha vs. Bombardier in Table 1). However, under medium or low demand firms may switch to a cooperative mode via licensing (subgames 5 and 8 in Figures 1 and 2), as in the example of Genentech versus Eli Lilly in Table 1.5 At medium demand the incumbent pursues a wait-and-see strategy, with the innovator licensing its moderately better technology to the incumbent for a fee while also producing for itself (a case of coopetition). We summarize the above analyses as follows: 高需求时以及在低或中等需求下转向通过许可进行合作的情况。在高需求且技术实力中等的情况下（图1和图2中的子博弈2），每家企业都有通过竞争模式进行投资的主导策略（同样，见表1中的雅马哈与庞巴迪案例）。然而，在中等或低需求时，企业可能会转向通过许可进行合作模式（图1和图2中的子博弈5和8），如表1中的基因泰克与礼来公司案例所示。5在中等需求下，在位企业采取观望策略，创新者通过许可将其稍好的技术授权给在位企业并收取费用，同时自身也进行生产（这是一种竞合案例）。我们将上述分析总结如下：

Proposition 1 (form of competition). Under strong technology, competition prevails but the form of competition becomes more offensive (moving from patent wall to bracketing) as demand increases. 命题1（竞争形式）。在技术优势强劲的情况下，竞争依然存在，但随着需求增长，竞争形式会变得更具进攻性（从专利壁垒转向围堵）。

Proposition 2.1 (switching based on state of demand). Although an innovator will prefer the competition mode if its technology is strong and the cooperation mode if its technology is weak (unless market demand is very high), if its technology strength is medium the innovator will prefer a hybrid strategy that involves switching between competition at high market demand and cooperation at lower or medium demand levels. 命题2.1（基于需求状态的切换）。虽然创新者若技术实力较强会偏好竞争模式，技术实力较弱时（除非市场需求极高）会偏好合作模式，但当技术实力处于中等时，创新者会倾向于采用混合策略：在市场需求较高时选择竞争，在需求较低或中等时转用合作。

Corollary 2.1a (licensing based on state of demand). An innovator will likely license its technology to an incumbent when neither its technology nor market demand are strong. 推论 2.1a（基于需求状态的许可）。当创新者自身的技术和市场需求都不强劲时，其可能会向现有企业许可技术。

The circumstances around hybrid strategies are a novel aspect of this article as they give rise to deployment mode switching between competition and cooperation as a result of demand-side factors (shifts in market demand and, by extension, market volatility), independent of the supply-side considerations that prior research has focused on, such as uncertainty about the value of the technology and the (declining) cost of its integration (Marx et al., 2014) 混合策略的相关情况是本文的一个新颖方面，因为它们会因需求侧因素（市场需求的变化以及由此延伸的市场波动性）而导致竞争与合作之间的部署模式切换，这与先前研究关注的供给侧因素（如技术价值的不确定性以及其整合成本的下降趋势，Marx et al., 2014）无关。

that may justify pivoting from competition to licensing or cooperation. Thus, in our model, switching can be in the reverse direction (i.e., from a cooperative to a competitive mode) if demand or volatility shifts are opposite, which is not explainable under the prior supply side—based theory. 这可能证明从竞争转向许可或合作是合理的。因此，在我们的模型中，切换可以是反向的（即从合作模式转向竞争模式），如果需求或波动性变化相反的话，而这在之前以供应方为基础的理论中是无法解释的。

We next show how the option-games valuation of the patent via S-NPV differs from a static NPV. Option-games valuation of firm $T \boldsymbol s$ patent deployment strategy depends on the equilibrium outcomes for each of the subgames composing the overall options game. The equilibrium outcome values under high, medium, and low demand $\left( E _ { H } , E _ { M } , \right.$ and $E _ { L } )$ provide the payoffs in the end-of-period nodes that are then folded back in a binomial option tree. Following Smit and Trigeorgis (2017), the period 2 equilibrium payoffs in each state of demand for a given technology strength (weak, medium, or strong) are then weighted by their respective (risk-adjusted) probabilities and discounted back to the present $\left( t = 0 \right)$ at the riskless interest rate $r$ within the backward binomial process. 我们接下来展示通过S-NPV对专利进行期权博弈估值与静态NPV的不同之处。企业T的专利部署策略的期权博弈估值取决于构成整体期权博弈的每个子博弈的均衡结果。高、中、低需求下的均衡结果值（\( E_H \)、\( E_M \)和\( E_L \)）提供了期末节点的收益，这些收益随后在二项式期权树中反向折叠。根据Smit和Trigeorgis（2017）的研究，在给定技术强度（弱、中或强）的每种需求状态下，第2期的均衡收益随后按其各自的（风险调整后）概率加权，并以无风险利率\( r \)折现回现值（\( t=0 \)），这一过程在反向二项式过程中进行。

The period 2 equilibrium payoffs used for valuing the innovator’s patent deployment strategy are those highlighted as shaded boxes in Figure 2, Panel B (first entry in each cell for the innovator firm $T$ : in case of weak technology (subgames 1, 4, and 7) $E _ { H } = 1 5 4$ (under high demand), ${ \cal E } _ { M } = 2 5$ (under medium demand), and $E _ { L } = 2$ (under low demand); for medium technology (subgames 2, 5 and 8) $E _ { \mathrm { H } } = 6 1 $ (under high demand), ${ { E } _ { M } } = 4 6$ (under medium demand), and $E _ { L } = 1 2$ (under low demand); for strong technology (subgames 3, 6, and 9) $E _ { H } = 9 2 $ (under high demand), ${ \cal E } _ { \mathrm { M } } = 3 4$ (under medium demand), and $E _ { L } = 9$ (under low demand). As shown in Online Appendix A, the value of the patent deployment strategy, based on the S-NPV (see Smit & Trigeorgis, 2004, 2017) for innovator firm $T _ { i }$ is $$ 320$ in case of weak technology (following a cooperative licensing mode), $$ 310$ in case of medium technology (following a hybrid strategy of competing under high demand, and cooperating via licensing under medium or low demand), and $$ 290$ in case of strong technology (using bracketing under high demand, raising a patent wall under medium demand, and sleeping under low demand). 用于评估创新者专利部署策略的周期 2 均衡收益是图 2，面板 B 中以阴影框突出显示的那些（创新企业 T 的每个单元格中的第一个条目：在技术薄弱的情况下（子博弈 1、4 和 7），高需求下 \( E_H = 154 \)，中需求下 \( \mathcal{E}_M = 25 \)，低需求下 \( E_L = 2 \)；在技术中等的情况下（子博弈 2、5 和 8），高需求下 \( E_H = 61 \)，中需求下 \( E_M = 46 \)，低需求下 \( E_L = 12 \)；在技术强大的情况下（子博弈 3、6 和 9），高需求下 \( E_H = 92 \)，中需求下 \( \mathcal{E}_M = 34 \)，低需求下 \( E_L = 9 \)）。如在线附录 A 所示，基于创新企业 \( T_i \) 的 S-NPV（见 Smit & Trigeorgis, 2004, 2017）的专利部署策略价值，在技术薄弱的情况下（采用合作许可模式）为 320 美元，在技术中等的情况下（采用高需求下竞争、中或低需求下合作许可的混合策略）为 310 美元，在技术强大的情况下（高需求下采用括号策略、中需求下筑专利墙、低需求下休眠）为 290 美元。

These results are summarized in Table 3. Compared to the static NPV of $$ 200$ (which assumes investing now while foregoing the option to exploit the interplay between competition and collaboration after observing actual demand), firm T’s patent option portfolio (estimated as S-NPV static NPV) is worth $$ 123$ under the cooperative strategy when the technology is weak (symmetry), $$ 110$ under the hybrid strategy when the technology is medium, and $$ 90$ under a competitive mode when the technology is strong. The competitive mode in this case, despite firm T’s technological advantage, results in lower value due to value dissipation from the ensuing patent war. 这些结果总结在表3中。与静态净现值（NPV）200美元（假设现在投资，同时放弃在观察到实际需求后利用竞争与合作之间相互作用的选择权）相比，在技术薄弱（对称）时，采用合作策略的T公司专利期权组合（估计为S-NPV 静态NPV）价值123美元；技术中等时，采用混合策略价值110美元；技术强大时，采用竞争模式价值90美元。在这种情况下，尽管T公司具有技术优势，但竞争模式因随后的专利战导致价值耗散，从而产生较低的价值。

In case of strong technology, firm $T$ realizes that under high demand $( H )$ it might be better off to cooperate (e.g., via cross-licensing), obtaining a smaller $( 5 0 % )$ share of a $( 2 0 % )$ larger market pie (resulting in 154 as in subgame 1 in Figure 2, Panel B), rather than compete offensively shouldering higher bracketing costs to obtain a higher share $( 7 5 % )$ of a fiercely contested and $( 3 0 % )$ smaller pie (resulting in value of 92 as in subgame 3 in Figure 2, Panel B). Such a switch from a competitive to a cooperative mode implies effectively replacing the equilibrium competitive payoff of 92 by the equilibrium cooperative payoff of 154 under strong technology. Such a hybrid patent strategy, switching from a compete mode (via defensive patent wall) in medium demand with room for just one firm to a cooperative mode in high demand results in a higher S-NPV of $$ 380$ (up from $$ 290$ , which nearly doubles the value of the patent option portfolio to $$ 18 m$ Such cooperative colicensing of an industry standard is illustrated by the example of Philips and Sony in Table 1. This hybrid patent strategy under strong technology is more valuable $( $ 380$ than the cooperative strategy under weak technology involving symmetric firms $\left( $ 3200\right)$ or the hybrid strategy under medium technology $\left( $ 3100\right)$ . Figure 3 shows how the value of the patent strategy (S-NPV) varies with technological strength, as reflected by market concentration (Herfindahl-Hirschman Index [HHI]) at different levels of demand volatility $( \sigma )$ under competitive, cooperative, and hybrid patent strategies. The cooperation and hybrid strategies are at a higher (elevated) value level. 在技术实力强劲的情况下，企业T意识到，在高需求（H）下，与其通过进攻性竞争承担更高的跨层成本以获取一个竞争激烈且规模较小（30%）的市场份额（带来92的价值，如子博弈3所示，见图2 Panel B），不如选择合作（例如通过交叉许可），获取一个规模较大20%的市场的较小份额（50%）（带来154的价值，如子博弈1所示，见图2 Panel B）。这种从竞争模式转向合作模式的转变，在技术实力强劲的情况下，实际上将平衡竞争收益92替换为平衡合作收益154。这种混合专利策略在中等需求下通过防御性专利壁垒采用竞争模式（仅容一家企业生存），在高需求下切换至合作模式，带来的战略净现值（S-NPV）更高（380美元，较之前290美元增长，专利期权组合价值接近翻倍至1800万美元）。表1举例说明了飞利浦与索尼在行业标准合作许可方面的案例。在技术实力强劲情况下，这种混合专利策略的价值更高（380美元），优于技术较弱时涉及对称企业的合作策略（3200美元）或技术中等时的混合策略（3100美元）。图3展示了专利策略价值（S-NPV）随技术实力（以市场集中度，即赫芬达尔-赫希曼指数[HHI]衡量）、不同需求波动率（σ）以及竞争、合作和混合专利策略的变化趋势。合作策略和混合策略的价值处于更高（提升）水平。

TABLE 3 Value of Patent Strategy (S-NPV) in the Numerical Illustration
表3 数值示例中专利策略（S-NPV）的价值

If ignore competition; invest now 如果忽略竞争；现在投资

FIGURE 3 Value of Patent Strategy (S-NPV) in the Numerical Illustration for Varying Degrees of Technology Strength (Proxied by the Market Concentration Index HHI) and Different Volatility, under Competitive, Cooperative, and Hybrid or Flexible Strategy (Symmetric Duopoly) 图3 专利策略（S-NPV）在不同技术实力（由市场集中度指数HHI代理）、不同波动性、竞争、合作及混合或灵活策略（对称双寡头）下的数值示例中的价值

For robustness, we also consider how these numerical results shift in response to two main changes: First, in Online Appendix A, we examine how the results shift under the assumption of asymmetric prior market power in favor of incumbent firm I, which offsets the technological advantage of innovator firm T. We find that cases that were previously symmetric, resulting in cooperation, now become asymmetric (due to prior advantage of the incumbent) switching to competition, and vice versa. What matters, therefore, is the relative overall competitive advantage (or firms’ asymmetry), not just the technological advantage of the innovator. Second, we next examine the impact of changing the level of demand volatility. Figure 4 shows sensitivity of S-NPV to demand volatility $( \sigma )$ under weak, medium, and strong technological advantage. The base level of $\sigma = 6 0 %$ repeats the S-NPV values shown in the middle column of Table 3. The conflict between competition and cooperation in high demand states leads to a value discontinuity (gap) between the rigid competitive strategy and the flexible cooperative (switch) strategy S-NPV under strong technology. S-NPV values decline at lower volatility levels, as expected by real options theory. At $\sigma = 1 5 %$ , values coincide with those shown in the left S-NPV column in Table 3. An interesting discontinuity in the S-NPV values is noted around a critical volatility level of about $\sigma ^ { * } = 3 8 %$ . This discontinuity arises due to a shift in certain subgame equilibria as volatility declines below a critical threshold level. Under strong technology, in low demand the equilibrium strategy is to sleep (wait) under high volatility; but as $\sigma$ declines below $\sigma ^ { * }$ , the value of the wait-and-see option declines while the appeal for the technologically advantaged firm $T$ to compete (and drive the rival out) given low demand and recovery prospects rises. However, at very high demand cooperation is more attractive under high volatility, partly deriving from the option to jointly appropriate value and optimize under demand uncertainty, avoiding the prisoners’ dilemma of both firms investing prematurely under the pressure of competitive rivalry. As volatility drops below a certain level, however, there is a switch from cooperative to competitive equilibrium (involving a shift from the sleep mode to a compete mode under low demand, and from the cooperate mode to a compete mode under high demand). Furthermore, under medium or high demand if technology is medium and volatility is high, future high rewards may induce the incumbent to compete aggressively and aSometimes cooperate (e.g., cross-licensing against third rivals). 为确保稳健性，我们还考虑了这些数值结果如何响应两个主要变化而变化：首先，在在线附录A中，我们研究了在假设在位企业I具有不对称先验市场势力（有利于在位企业）的情况下结果如何变化，这抵消了创新企业T的技术优势。我们发现，之前对称的情况（导致合作）现在因在位企业的先验优势而变得不对称（转向竞争），反之亦然。因此，重要的是相对整体竞争优势（或企业的不对称性），而不仅仅是创新企业的技术优势。其次，我们接下来研究需求波动水平变化的影响。图4显示了在弱、中、强技术优势下，S-NPV对需求波动率（σ）的敏感性。基准波动率σ=60%重复了表3中间列中显示的S-NPV值。高需求状态下竞争与合作之间的冲突导致强技术优势下刚性竞争策略与灵活合作（转换）策略S-NPV之间出现价值不连续性（差距）。根据实物期权理论，S-NPV值在较低波动率水平下会下降。当σ=15%时，数值与表3左侧S-NPV列中的数值一致。在约σ*=38%的临界波动率水平附近，观察到S-NPV值的一个有趣不连续性。这种不连续性是由于当波动率下降到临界阈值以下时，某些子博弈均衡发生了转变。在强技术优势下，低需求时高波动率的均衡策略是“休眠”（等待）；但当σ下降到σ*以下时，“观望”期权的价值下降，而技术优势企业T在低需求和复苏前景下进行竞争（并驱逐对手）的吸引力上升。然而，在高需求时，高波动率下合作更具吸引力，部分源于共同获取价值和在需求不确定性下优化的期权，避免了两家企业在竞争对抗压力下过早投资的囚徒困境。然而，当波动率下降到一定水平以下时，会出现从合作均衡向竞争均衡的转变（包括低需求下从“休眠”模式转向“竞争”模式，以及高需求下从“合作”模式转向“竞争”模式）。此外，在中或高需求下，如果技术水平为中等且波动率较高，未来的高回报可能会促使在位企业积极竞争，有时会进行合作（例如，针对第三方对手的交叉许可）。

FIGURE 4 Sensitivity of S-NPV to Volatility under Weak, Medium, or Strong Technology in the Numerical Illustration 图4 数值说明中弱、中或强技术下S-NPV对波动性的敏感性

FIGURE 5 Summary and Extension of Patent Deployment Strategies in the Numerical Illustration for a Broader Range of Demand and Uncertainty (Under Strong Technology)—Symmetry Case 图5 数值示例中更广泛需求和不确定性（强技术情境下）的专利部署策略总结与扩展——对称情形

FIGURE 6 Extension of Compete versus Cooperate Strategies in the Numerical Illustration (for a Broader Range of Demand/Uncertainty) under Weak, Medium, or Strong Technology 图6 数值说明中竞争与合作策略的扩展（针对更广泛的需求/不确定性范围）——在技术薄弱、中等或强大的情况下

enter a bracketing war; as volatility declines, however, the possibility of high rewards declines and the incumbent may face a patent wall by the innovator or shift to cooperation. Figure 5 confirms, in the case of strong technology, that at low $\sigma$ a rigid, competeonly strategy may be best. However, as the cone of uncertainty rises a wider menu of strategic deployment choices opens up, including sleep (and possibly exit) at the low end and cooperate at high (and possibly at middle) demand. At high volatility $( \sigma = 9 0 %$ patent deployment strategies thus span the whole range, including sleep or exit, compete, as well as cooperate strategy modes. 进入一场括号战争；然而，随着波动性下降，高回报的可能性也下降，在位者可能面临创新者的专利壁垒，或转向合作。图5证实，在技术实力强劲的情况下，当σ较低时，纯粹竞争的刚性策略可能是最佳选择。然而，随着不确定性锥度上升，战略部署的选择范围扩大，包括在低端休眠（并可能退出）以及在高（并可能在中）需求时合作。在高波动性（σ=90%）下，专利部署策略涵盖了整个范围，包括休眠或退出、竞争以及合作策略模式。

Figure 6 provides an extension (including the case of weak and medium technology) of various cooperative versus competitive patent deployment strategies that may be optimal when a broader (more extreme) spread of demand states is possible. The case of strong technology (rightmost column) corresponds to the high volatility case (rightmost column) of Figure 5. Here, however, higher demand volatility allows adding extreme very high (VH) and very low $( V L )$ demand states at the two tails, besides high $\left( H \right)$ , medium $( M )$ and low $\left( L \right)$ demand as in the base case. As previously, in determining the equilibria for each of the various cooperate or compete subgames, each firm would select the type of patent deployment strategy and associated options to optimally exercise depending on different market demand (and volatility) conditions and its relative technological strength. Under strong technology, patent deployment by firm $T$ may span the entire menu of available options depending on prevailing market demand conditions: exit when demand is very low; sleep or “wait and see” when demand is low; expand or strengthen the patent to compete (e.g., through a patent wall) at medium demand, while at times cooperating with the rival (e.g., to preempt third entrants); compete offensively (e.g., via bracketing) in high demand; and potentially switch to a cooperate mode at very high demand, allowing both rivals to profit. Under strong technology the optimal patent strategy of firm T may vary or switch among defer (or exit) and compete or cooperate modes at a given time depending on the level of demand and other conditions, such as volatility and industry dynamism. Under volatile conditions, patent deployment should be more flexible, able to adapt and switch among competitive, cooperative, or even coopetition modes, including delaying the commitment decision via a sleep mode. Patent deployment strategy is generally more hybrid (allowing more switching possibilities) when the innovator’s technology advantage is medium, with even small variations in demand (e.g., from high to medium) bringing about a strategy switch from a competitive to a cooperative mode involving licensing. The above insights can be summarized as follows: 图6提供了各种合作与竞争专利部署策略的扩展（包括弱技术和中等技术的情况），当需求状态的分布更广泛（更极端）时，这些策略可能是最优的。强技术的情况（最右列）对应图5的高波动性情况（最右列）。然而，在这里，除了基准情形中的高（H）、中（M）和低（L）需求外，更高的需求波动性允许在两个尾部添加极端的极高（VH）和极低（VL）需求状态。与之前一样，在确定各种合作或竞争子博弈各自的均衡时，每个企业将根据不同的市场需求（和波动性）条件及其相对技术实力，选择最优实施的专利部署策略类型及相关选项。在强技术条件下，企业T的专利部署可能根据当前市场需求条件涵盖所有可用选项：需求极低时退出；需求较低时休眠或“观望”；需求中等时扩张或加强专利以竞争（例如，通过专利壁垒），同时有时与对手合作（例如，以阻止第三方进入）；需求较高时进行进攻性竞争（例如，通过包围策略）；并可能在极高需求时切换到合作模式，使双方竞争对手都能获利。在强技术条件下，企业T的最优专利策略可能会随时间变化或在给定时间在推迟（或退出）、竞争或合作模式之间切换，具体取决于需求水平和其他条件，如波动性和行业动态。在波动条件下，专利部署应更具灵活性，能够适应并在竞争、合作甚至竞合模式之间切换，包括通过休眠模式推迟承诺决策。当创新者的技术优势为中等时，专利部署策略通常更具混合性（允许更多切换可能性），即使需求有微小变化（例如，从高到中）也会导致策略从竞争模式切换到涉及许可的合作模式。上述见解可总结如下：

Corollary 2.1b (licensing in volatile regimes). Licensing can prevail in more volatile regimes, even when the innovator’s technology is strong, under very high demand when the (smaller) share of (larger) joint benefits from cooperation exceeds the dominant share of a reduced market pie from a costly patent war. High demand volatility, increasing the range of extreme high and low demand states, makes licensing more beneficial. 推论 2.1b（波动环境下的许可）。在需求波动较大的环境中，即使创新者的技术实力强劲，许可也可能占据主导地位——当合作带来的（较大）联合收益中（较小）的份额超过因专利战代价高昂而缩小的市场规模中占主导地位的份额时，且需求处于极高水平。需求波动性高会增加极端高需求和低需求状态的范围，这使得许可更具优势。

Give-up (cooperate) strategies initially adopted at low or medium demand may switch to competition at high demand and then, at higher demand in volatile regimes, to cooperation. Volatility exacerbates these switching patterns between competition and cooperation. 初始在低或中等需求下采用的放弃（合作）策略，在高需求时可能转为竞争，而在波动状态下更高需求时，又会转为合作。波动性加剧了竞争与合作之间的这些转换模式。

EXTENDED ANALYSIS: PATENT DEPLOYMENT MODE SWITCHING OVER TIME

扩展分析：专利部署模式随时间的切换

In this section, we extend the above baseline analysis of the tradeoff between competition and collaboration in patent deployment strategies to explicitly account for coopetition and for switching among deployment modes over time. To focus on these aspects and to keep the analysis tractable, we set aside some of the other features of the baseline numerical analysis. In the extension, only the innovator has a patented technology, eliminating the possibility of firms cross-licensing patents to each other. Consequently, we also set aside any notion of patent bracketing or patent wall. Further, to keep the analysis simple, we assume here that the incumbent always remains in the market. Finally, we make a specific assumption about the structure of the product market, namely firms simultaneously choose their output levels, as in Cournot quantity competition (focusing on process innovation). Details and derivations for this extended analysis are given in Online Appendix B. 在本节中，我们将上述关于专利部署策略中竞争与合作权衡的基准分析扩展到明确考虑竞合（coopetition）以及随时间在不同部署模式间切换的情况。为聚焦这些方面并保持分析的可处理性，我们暂不考虑基准数值分析中的其他一些特征。在扩展分析中，仅创新者拥有专利技术，从而排除了企业间交叉许可专利的可能性。因此，我们也暂不考虑专利组合或专利壁垒的概念。此外，为简化分析，我们在此假设在位企业始终留在市场中。最后，我们对产品市场的结构做了一个特定假设，即企业同时选择其产出水平，这类似于古诺（Cournot）产量竞争（聚焦于工艺创新）。该扩展分析的细节和推导见在线附录B。

The extended analysis considers an industry previously served only by an incumbent monopolist, firm I, over two subsequent time periods, 1 and 2, with linear demand for an undifferentiated product. Demand may shift between these two periods due to a demand shock, with equal probability of moving either up or down (as well as some probability of remaining at the same level). Here, overall demand volatility is influenced by two factors: the likelihood of a demand shift occurring—that is, the tail probability—and the magnitude of the shift if it occurs— that is, the demand spread. Innovator firm $T$ patents a new cost-saving technology whose strength reflects the magnitude of the reduction in the marginal cost. In each period $\dot { t } = 1$ or 2), firm $T$ can deploy its patent in one of three ways: (a) It can pay a fixed capital cost to obtain the commercialization capability to produce at the lower marginal cost and engage in asymmetric Cournot competition with incumbent firm I, which must produce at its higher marginal cost (the “competition” deployment mode, abbreviated as $P$ for “produce”). (b) It can license out its technology to the incumbent in a “gain sharing” licensing deal, allowing firm I to also reduce its marginal cost by paying firm $T$ a licensing fee consisting of a share of the incumbent’s incremental profit gain from using the licensed technology (a “cooperation” deployment mode, abbreviated as $L$ for “license” only). (c) In a hybrid arrangement, firm $T$ can both produce and license concurrently, with both firms enjoying the same lower marginal cost, while firm $T$ receives a licensing fee from the incumbent (the “coopetition” deployment mode, abbreviated as $B$ for “both” producing and licensing concurrently). The innovator firm’s bargaining power in licensing negotiations is captured by the portion of the incumbent’s profit gain from using the superior technology that must be paid to the innovator as a licensing fee, under either the cooperation $\left( L \right)$ or coopetition $( B )$ deployment modes. The innovator firm $T$ chooses whether to produce output in each period, whereas the incumbent firm I remains in the market for both periods. 扩展分析考虑了一个此前仅由在位垄断企业（企业I）服务的行业，在随后的两个时间段（时期1和时期2）内，该行业的产品为无差异产品且需求呈线性。由于需求冲击，两个时期之间的需求可能发生变化，上升或下降的概率相等（也有一定概率保持不变）。在此，总体需求波动受两个因素影响：需求变化发生的可能性（即尾部概率）以及若发生变化时的变化幅度（即需求扩散度）。创新企业T为一项新的成本节约技术申请了专利，该技术的强度反映了边际成本降低的幅度。在每个时期（t=1或2），企业T可以通过三种方式之一部署其专利：(a) 它可以支付固定资本成本以获得商业化能力，从而以较低的边际成本生产，并与必须以较高边际成本生产的在位企业I进行非对称古诺竞争（“竞争”部署模式，缩写为P，代表“生产”）。(b) 它可以在“收益共享”许可协议中将技术许可给在位企业，允许企业I也通过支付许可费（该许可费由企业I使用许可技术获得的增量利润收益的一部分构成）来降低其边际成本（“合作”部署模式，缩写为L，仅代表“许可”）。(c) 在混合安排中，企业T可以同时生产和许可，此时两家企业均享有相同的较低边际成本，而企业T从在位企业获得许可费（“竞合”部署模式，缩写为B，代表同时生产和许可）。在合作（L）或竞合（B）部署模式下，创新企业在许可谈判中的议价能力由在位企业使用先进技术获得的利润收益中必须支付给创新企业的部分来体现。创新企业T选择每个时期是否生产产品，而在位企业I在两个时期均留在市场中。

TABLE 4 Summary of Innovative Firm’s Ex Post Choice of Patent Deployment Mode (Competition, Cooperation, or Coopetition Dependingon the Strengt o Technological Advantage and the Levelf Demand, under Moderate (Panel A) or High (Panel B) Bargaining Power for the Innovator
表4 创新企业事后专利部署模式选择总结（根据技术优势强度和需求水平，在创新者议价能力中等（面板A）或高（面板B）的情况下，竞争、合作或竞合）

As derived in Online Appendix B, Table 4 shows conditions under which innovator firm $T$ will choose each of the three patent deployment modes (competition, cooperation, or coopetition) in a given time period, as a function of its technological advantage, the realized demand level, and its bargaining power in licensing negotiations. Table 4, Panel A represents our base case, predicated on the common situation where the innovator has moderate bargaining power, similar to Figure 1 from the baseline numerical illustration. In this case, cooperative deployment (i.e., licensing without producing) tends to prevail when the innovative firm’s technology is weak and market demand is not high, but the innovator will produce itself under the competitive mode when its technological advantage is strong or market demand is high, which makes it worthwhile for the innovator to pay the fixed capital cost for producing in that period. The pattern in Table 4, Panel A is broadly in line with Figures 1 and 6 (where competition prevails under very high demand even under weak technology). Thus, Proposition 2.1 and Corollary 2.1a from the baseline numerical analysis are reaffirmed. 如在线附录B中推导所示，表4显示了创新企业T在给定时间段内选择三种专利部署模式（竞争、合作或竞合）的条件，这些条件取决于其技术优势、实现的需求水平以及其在许可谈判中的议价能力。表4的A部分代表我们的基准案例，基于创新者具有中等议价能力的常见情况，类似于基准数值说明中的图1。在这种情况下，当创新企业的技术较弱且市场需求不高时，合作部署（即无生产许可）往往占主导地位，但当创新者的技术优势较强或市场需求较高时，创新者会在竞争模式下自行生产，这使其有必要在该时期支付生产的固定资本成本。表4的A部分中的模式与图1和图6大致一致（在非常高的需求下，即使技术较弱，竞争也占主导地位）。因此，基准数值分析中的命题2.1和推论2.1a得到了再次确认。

Table 4, Panel B shows a less common case where the innovator has higher bargaining power in licensing negotiations, capturing more than half of the incumbent’s profit gain as a licensing fee. In this case, the coopetition mode (i.e., both producing and licensing concurrently) can prevail if market demand is sufficiently high. Otherwise, for lower levels of demand, the results of the high bargaining power case in Panel B are similar to the moderate bargaining power case from Panel A. This difference occurs because coopetition requires a trade-off in which the innovator sacrifices some product-market profits (i.e., giving up its asymmetric competitive advantage by allowing the incumbent rival to also use its superior technology) in exchange for licensing fees, and this trade-off is only worthwhile when the innovator enjoys strong enough bargaining power to extract a majority of the incumbent’s gains from licensing. As a result, when the innovator has high bargaining power, growth in demand can precipitate a switch from competition to coopetition, as in the case of LG’s multiple relationships with Sony and Panasonic in the OLED TV market.6 We organize these ex post (period 2) results as follows: 表4的面板B展示了一种较不常见的情况，即创新者在许可谈判中拥有更高的议价能力，将超过一半的在位企业利润增长作为许可费收入囊中。在这种情况下，如果市场需求足够高，竞合模式（即同时生产和许可）可能会占主导地位。否则，对于较低的需求水平，面板B中高议价能力案例的结果与面板A中中等议价能力案例的结果相似。这种差异的产生是因为竞合需要一种权衡：创新者需要牺牲部分产品市场利润（即通过允许在位竞争对手也使用其优势技术，放弃自身的非对称竞争优势），以换取许可费，而这种权衡只有在创新者拥有足够强的议价能力，能够从许可中获取在位企业大部分收益时才值得进行。因此，当创新者拥有高议价能力时，需求的增长可能会促使竞争转向竞合，例如LG在OLED电视市场与索尼和松下的多重合作关系。6我们将这些事后（第2期）结果整理如下：

Proposition 2.2 (switching based on state of demand—high bargaining power case). When an innovator’s bargaining power is high, higher demand may induce switching from competition to coopetition if the innovator’s technology is strong, or from cooperation to coopetition if its technology is weak. For medium-strength technology, patent deployment mode may switch from initial give-up strategies involving cooperation at low demand, to competition at medium demand, and then switching to coopetition at higher demand. 命题2.2（基于需求状态的切换——高议价能力情形）。当创新者的议价能力较高时，若其技术实力较强，更高的需求可能会促使竞争转向竞合；若其技术实力较弱，则可能促使合作转向竞合。对于中等强度的技术，专利部署模式可能从初始的低需求下涉及合作的放弃策略，转变为中等需求下的竞争，进而在更高需求下转向竞合。

Bearing in mind that an innovator can license either with or without competing against the licensee—that is, either in cooperation or in coopetition—the above proposition also implies the following corollary: 考虑到创新者可以选择在不与被许可方竞争的情况下进行许可（即合作），也可以选择在竞争的情况下进行许可（即竞合），上述命题还隐含了以下推论：

Corollary 2.2a (licensing based on state of demand— high bargaining power case). When an innovator’s bargaining power is high, it will prefer to license a weak technology always, a strong technology only when demand is high, and a medium-strength technology at the extremes of demand—that is, when demand is either low (cooperation) or high (coopetition). 推论 2.2a（基于需求状态的许可——高议价能力情形）。当创新者的议价能力较高时，它会始终倾向于许可弱技术，仅在需求高时许可强技术，而在需求的极端情况下（即需求低（合作）或高（竞合））许可中等强度技术。

The above ex post analysis examined conditions for actual switching among deployment modes at a given time in order to maximize the innovator’s realized profit, as a function of the level of demand prevailing at that time—as well as other parameters, such as the innovator’s technological advantage and bargaining power. Next, we consider an ex ante analysis of conditions that increase the probability of switching at a future time to maximize the innovator’s expected profit as a function of the volatility of demand and other parameters. Table 5 shows how a marginal increase in the probability of tail outcomes affects the ex ante probability of innovative firm $T$ switching between different patent deployment modes (i.e., competition, cooperation, or coopetition) across time (i.e., between periods 1 and 2), depending on the spread of demand outcomes, the innovator’s bargaining power, the strength of its technology, and the baseline demand. Panel A of Table 5 considers the base case with moderate bargaining power for the innovator, where increasing the probability of tail outcomes increases the probability of switching between cooperation and competition when the technology is weak or medium and the spread of demand outcomes is high. Panel B considers the less common case of high bargaining power where, under high demand spread, increasing the probability of tail outcomes increases the probability of switching either between cooperation and coopetition if the technology is weak, or between competition and coopetition if the technology is strong, or in more complicated ways if the technology strength is medium, as we discuss next. 上述事后分析研究了在给定时间点实际切换部署模式以最大化创新者实际利润的条件，这一条件取决于当时普遍存在的需求水平以及其他参数，如创新者的技术优势和议价能力。接下来，我们考虑事前分析，即分析哪些条件会增加未来时间点切换的概率，以最大化创新者的预期利润，这一概率取决于需求波动性和其他参数。表5展示了尾部结果概率的边际增加如何影响创新企业T在不同专利部署模式（即竞争、合作或竞合）之间随时间（即第1期和第2期之间）切换的事前概率，具体取决于需求结果的分布范围、创新者的议价能力、技术实力以及基准需求。表5的A部分考虑了创新者议价能力适中的基准情形，其中当技术实力较弱或中等且需求结果分布范围较高时，尾部结果概率的增加会提高合作与竞争之间切换的概率。B部分考虑了议价能力较高这一较少见的情形，其中在需求分布范围较高的情况下，当技术实力较弱时，尾部结果概率的增加会提高合作与竞合之间切换的概率；当技术实力较强时，会提高竞争与竞合之间切换的概率；而当技术实力处于中等水平时，则会以更复杂的方式影响切换概率，下文将对此进行讨论。

Figure 7 provides a more in-depth graphical representation of the ex ante results from Table 5, focusing on cases where the technology strength is medium. We reduce the parameter space to just two dimensions for graphing purposes by setting the baseline demand and capital cost at reasonable typical levels. Along the bottom of the graph, where demand spread is zero, we find the innovator’s initial (period 1, preshock) starting point (deployment mode), as a function of the technology advantage along the horizontal axis. When demand spread is zero, the outcome is nonswitching, i.e., staying with the same deployment mode choice that the innovator makes in period 1. In period 1, the innovator starts in cooperation mode where the horizontal axis is in a red region, in competition mode where it is in a yellow region, and in coopetition where it is in a blue region. If the demand spread is close enough to zero in period 2 to remain in the red, yellow, or blue region, then no switching occurs among the two periods. When demand spread is sufficiently high in period 2 to move up into a different colored region, switching becomes possible. We start at the bottom of any dashed line, where no switches occur, and then follow the dashed vertical line upward to the point at which we reach the threshold level of demand spread where switches can happen. 图7更深入地以图形方式展示了表5中事前结果，重点关注技术实力为中等的情况。为了便于绘图，我们将参数空间简化为仅两个维度，将基准需求和资本成本设定在合理的典型水平。在图表底部，需求波动为零的位置，我们可以找到创新者的初始（第1期，冲击前）起点（部署模式），作为沿水平轴的技术优势的函数。当需求波动为零时，结果为非切换，即创新者在第1期做出的部署模式选择保持不变。在第1期，创新者从合作模式开始，此时水平轴处于红色区域；竞争模式下处于黄色区域；竞合模式下处于蓝色区域。如果第2期的需求波动足够接近零，仍处于红色、黄色或蓝色区域，则两期之间不会发生切换。如果第2期的需求波动足够高，进入不同颜色的区域，则切换成为可能。我们从任何虚线的底部开始，在那里没有切换发生，然后沿着虚线垂直向上追踪，直到达到需求波动的阈值水平，此时切换成为可能。

For the moderate bargaining power case in Panel A of Table 5 (corresponding to the baseline numerical illustration in the previous section), switching is relatively simple because there are only two possible types of switches, between competition and cooperation. On the left side, where the technology advantage is weak (below about 0.09), the red vertical dashed line shows that when demand spread is low enough to stay in the red region, no switching occurs away from cooperation, but a positive probability of switching to competition (orange region) occurs when demand spread crosses into the orange region. On the right side, where technology advantage is medium or strong (above about 0.09), the yellow vertical dashed line shows that when demand spread is low enough to stay in the yellow region no switching occurs from competition, but a positive probability of switching to cooperation occurs when demand spread crosses into the orange region. The situation is somewhat more involved in the less common case of Panel B, where the innovator has higher bargaining power, as there are now six dashed vertical lines. The ones with very strong or weak technology involve simpler switches between two modes, while those in the middle region of technology strength involve possible switches among all three modes.7 对于表5 Panel A中的中等议价能力案例（对应上一节中的基准数值说明），切换相对简单，因为只有两种可能的切换类型：竞争与合作之间的切换。在左侧（技术优势较弱，低于约0.09），红色垂直虚线显示，当需求扩散足够低以保持在红色区域时，不会发生脱离合作的切换，但当需求扩散进入橙色区域时，存在切换到竞争的正概率（橙色区域）。在右侧（技术优势为中等或较强，高于约0.09），黄色垂直虚线显示，当需求扩散足够低以保持在黄色区域时，不会发生从竞争的切换，但当需求扩散进入橙色区域时，存在切换到合作的正概率。在Panel B这种较不常见的案例中，情况更为复杂，此时创新者具有更高的议价能力，共有六条垂直虚线。技术非常强或非常弱的情况涉及两种模式之间的简单切换，而技术强度处于中间区域的情况则可能涉及三种模式之间的切换。7

TABLE 5 SuaryxtceasgheroabilTailOtc hexpecatioichim ent Deployment Modes across Periods 1 and 2 Depending on the Demand Spread, Baseline Demand, and the Strength of Technological Advantage, under Moderate (Panel A) or High (Panel B) Bargaining Power for the Innovator
表5 不同需求分布、基准需求及技术优势强度下，在创新者议价能力为中等（面板A）或高（面板B）时，第1和第2阶段的部署模式（注：原标题存在拼写错误，此处按原样翻译）

Although Figure 7 provides a useful graphical illustration, it is limited to a single value of the baseline demand, which dictates a particular starting point in terms of deployment mode in period 1. So, to generalize to other possible values of the baseline demand, which would allow other starting points for the deployment mode, we turn our attention to Tables 4 and 5 (see also Tables B4 and B5 in Online 虽然图7提供了有用的图形说明，但它仅限于基准需求的单一值，这在第1期的部署模式中规定了一个特定的起点。因此，为了推广到基准需求的其他可能值（这将允许部署模式的其他起点），我们将注意力转向表4和表5（另见在线附录中的表B4和B5）。

Appendix B). From these tables, we make the following observations for the two cases. 附录 B）。从这些表格中，我们对两种情况得出以下观察结果。

In the more common base case, when an innovator’s bargaining power is moderate: 在更常见的基本情况下，当创新者的议价能力适中时：

1. If the innovator starts in cooperation, which can occur when technology is not strong and baseline demand is not high, it will stay there in the next period if the demand spread is sufficiently low, but if demand spread is high it may switch to competition.
   如果创新者从合作开始（这种情况可能发生在技术实力不强、初始需求不高时），那么在下一阶段，如果需求扩散程度足够低，它会继续保持合作状态；但如果需求扩散程度高，它可能会转向竞争。

2. If the innovator starts in competition, which can occur under strong technology or high baseline demand, it will stay there if demand spread is sufficiently low, but if demand spread is high it may switch to cooperation.

3. 如果创新者从竞争开始（这种情况可能在技术壁垒高或基准需求高时出现），如果需求扩散程度足够低，它会一直保持竞争状态；但如果需求扩散程度高，它可能会转向合作。

In the less common case when an innovator’s bargaining power is high: 在创新者议价能力较强的较不常见情况下：

1. If the innovator starts in cooperation, which can occur when technology is not strong and baseline demand is not high, it will stay there in the next period if demand spread is sufficiently low, but if demand spread is high it may switch to either coopetition or competition depending on the technology strength, as follows: a. For weak technology, it may switch to coopetition. b. For medium-strength technology, it may switch to competition under moderately low baseline demand, or to coopetition under higher baseline demand.
2. 如果创新者从合作开始（这种情况可能在技术不强且基准需求不高时发生），如果需求扩散足够低，它在下一阶段会保持合作；但如果需求扩散高，它可能会根据技术实力转向竞合或竞争，具体如下：a. 对于技术薄弱的情况，它可能转向竞合。b. 对于中等技术实力，在基准需求中等偏低时可能转向竞争，在基准需求较高时可能转向竞合。

FIGURE 7 Switching Probabilities as a Function of Technology Advantage and Demand Spread, Under Either Moderate (Panel A) or High (Panel B) Innovator Bargaining Power 图7 技术优势和需求扩散下的切换概率，在创新者议价能力中等（面板A）或高（面板B）的情况下

2. If the innovator starts in coopetition, which can occur under high baseline demand, it will stay there in the next period if demand spread is sufficiently low, but if demand spread is high it may switch to either cooperation or competition depending on the technology strength, as follows:
3. 如果创新者从竞合状态开始（这种情况可能出现在高基准需求下），那么在需求分散程度足够低的情况下，下一阶段会继续保持该状态；但如果需求分散程度高，它可能会根据技术实力转向合作或竞争，具体如下：

a. For weak technology, it may switch to cooperation.
a. 对于技术薄弱的企业，可能会选择合作。

b. For strong technology, it may switch to competition.
b. 对于强大的技术，它可能会转向竞争。

C. For medium-strength technology, it may switch to competition under somewhat lower (moderately high) baseline demand, or to cooperation under lower baseline demand. C. 对于中等强度的技术，它可能会在略低（中等偏高）的基准需求下转向竞争，或者在较低的基准需求下转向合作。

3. If the innovator starts in competition, which can occur either when technology is strong and demand is not high or when technology and demand are both medium, it will stay there in the next period if demand spread is sufficiently low, but if demand spread is high it may switch to either cooperation or coopetition depending on the technology strength, as follows:
4. 如果创新者从竞争开始（这种情况可能发生在技术强但需求不高，或者技术和需求都处于中等水平时），如果需求扩散足够低，它在下一阶段将继续保持竞争状态；但如果需求扩散高，它可能会根据技术实力转向合作或竞合，具体如下：

a. For strong technology, it may switch to coopetition, as in the case of the 2004 formation of a joint venture between rivals Samsung and Sony.8
a. 为了增强技术实力，企业可能会转向竞合关系，例如2004年竞争对手三星和索尼成立合资企业的案例。8

b. For medium-strength technology, it may switch to either cooperation or coopetition, as in the case of Yahoo! and Facebook.9 b. 对于中等强度的技术，它可能会转向合作或竞合，雅虎和Facebook的情况就是如此。9

We can summarize the essence of the results in these tables as follows: 我们可以将这些表格中结果的要点总结如下：

Proposition 3 (demand volatility and switching across time). Under high demand volatility, an increased probability of extreme tail outcomes increases the probability of switching across time periods among competition, cooperation, or coopetition. Under moderate bargaining power for the innovator (and weak or medium technology), it increases switching among cooperation and competition. When bargaining power is high, it increases switching between cooperation and coopetition under weak technology and between competition and coopetition under strong technology. When technology is medium (and bargaining power is high), it increases switching from any of the three deployment modes available to either of the other two modes. Switches can occur in either direction depending on technology strength and demand conditions. 命题3（需求波动性和跨时间切换）。在高需求波动性下，极端尾部结果的可能性增加会提高在竞争、合作或竞合之间跨时间段切换的概率。当创新者的议价能力处于中等水平（且技术较弱或中等）时，会增加在合作与竞争之间的切换。当议价能力较高时，在技术较弱的情况下会增加合作与竞合之间的切换，在技术较强的情况下会增加竞争与竞合之间的切换。当技术处于中等水平（且议价能力较高）时，会增加从三种可用部署模式中的任何一种切换到其他两种模式中的任意一种。根据技术实力和需求条件，切换可以向任一方向发生。

The above ex ante results also provide insights for the flexibility versus commitment trade-off, since they have implications for the conditions that affect the value of the option to switch deployment modes across time periods, rather than committing irrevocably to a specific deployment mode in period 1. The value of this kind of flexibility is illustrated by Motorola’s failure when it irrevocably committed to licensing prior to a volatility increase.10 This flexibility would be worthless in conditions where the innovator would never choose to switch, regardless of the actual demand realized in period 2 (anywhere in the red, yellow, and blue regions in Figure 7). That is, flexibility would have no value if there were no switches desirable to the innovator in period 2. At the opposite end, flexibility would have the most value if there were multiple different types of switches that the innovator might want to pursue in period 2 (e.g., anywhere in the brown region of 上述事前结果还为灵活性与承诺之间的权衡提供了见解，因为它们影响着随时间推移调整部署模式的期权价值的条件，而非在第 1 期不可撤回地承诺采用特定部署模式。这种灵活性的价值通过摩托罗拉在波动率上升前不可撤回地选择授权而失败的案例得到说明。10 在创新者无论第 2 期实际实现的需求如何都绝不会选择转换的情况下（如图 7 中的红色、黄色和蓝色区域内的任何位置），这种灵活性将毫无价值。也就是说，如果创新者在第 2 期没有任何希望进行的转换，灵活性就没有价值。而在另一端，如果创新者在第 2 期希望进行多种不同类型的转换（例如，在棕色区域内的任何位置），灵活性的价值将最大。

Figure 7, Panel B, where the innovator can switch from competition to either cooperation or coopetition, depending on actual demand realized in period 2).11 This outcome occurs under high innovator bargaining power, medium technology strength, and high volatility; thus: 图7，面板B，其中创新者可以根据第2期实现的实际需求从竞争转向合作或竞合。11 这种结果出现在创新者议价能力高、技术实力中等且波动性高的情况下；因此：

Proposition 4 (value of option to switch modes across time). The value of the innovator’s option to switch patent deployment modes across time is greater when it enjoys high bargaining power in licensing negotiations, when its technology is of medium strength, and when demand volatility is high. 命题4（跨时间切换模式的期权价值）。创新者跨时间切换专利部署模式的期权价值，在其在许可谈判中拥有较高议价能力、技术强度为中等以及需求波动性较高时更大。

IMPLICATIONS AND CONCLUSIONS

意义与结论

Choices about how to deploy a patent involve the two classic trade-offs of competition versus cooperation (Brandenburger & Nalebuff, 1996) and flexibility versus commitment (Ghemawat & del Sol, 1998). Our analysis indicates that although stronger technology makes an innovator more likely to compete, this competition may be blended with cooperation (i.e., coopetition) when market demand and the innovator’s bargaining power are both high. Likewise, although weaker technology makes an innovator more likely to cooperate via licensing, high demand may induce it to compete—again leading to coopetition when it enjoys high bargaining power. With mediumstrength technology, an innovator may switch between cooperation and competition or do both concurrently, depending on demand. Indeed, an innovator with a medium-strength technology has a particularly strong incentive to be flexible in its patent deployment strategy when its bargaining power is high. In such a case, all three modes of competition, cooperation, and coopetition can be optimal at different levels of demand, and licensing is preferred at the extremes of demand—either low demand (in cooperation) or in high demand (in coopetition, while also competing against the licensee). Since this case is where the greatest variety of optimal patent deployment modes can occur, it is also where volatility has the greatest impact on the option value of the innovator’s flexibility to switch modes over time. Thus, our analysis supports the logic of Lado, Boyd, and Hanlon (1997: 123) that successful firms “possess enhanced strategic flexibility by either holding or striking a wide variety of strategic options.” 关于如何部署专利的选择涉及竞争与合作（Brandenburger & Nalebuff，1996）以及灵活性与承诺（Ghemawat & del Sol，1998）这两个经典权衡。我们的分析表明，尽管更强的技术会使创新者更有可能竞争，但当市场需求和创新者的议价能力都很高时，这种竞争可能会与合作（即竞合）相结合。同样，尽管较弱的技术会使创新者更有可能通过许可进行合作，但高需求可能会促使其竞争——当它拥有高议价能力时，再次导致竞合。对于中等强度的技术，创新者可能会根据需求在合作与竞争之间切换或同时进行两者。事实上，当创新者的议价能力高时，拥有中等强度技术的创新者会特别有动力在其专利部署策略中保持灵活性。在这种情况下，竞争、合作和竞合这三种模式在不同的需求水平下都可能是最优的，而许可在需求的极端情况下（要么是低需求时的合作，要么是高需求时的竞合且同时与被许可方竞争）会被优先选择。由于这种情况是最优专利部署模式种类最多的情况，因此也是波动性对创新者随时间切换模式的灵活性的期权价值影响最大的情况。因此，我们的分析支持Lado、Boyd和Hanlon（1997：123）的逻辑，即成功的企业“通过持有或制定各种战略选择来增强战略灵活性”。

Several of our above findings contribute to research on patent deployment strategies. First, although strong technology makes an innovator more likely to pursue a competitive patent deployment mode, it may combine competition with cooperation via licensing—that is, deploy a coopetition strategy—when market demand and the innovator’s bargaining power in licensing negotiations are high. Similarly, although weak technology makes an innovator more likely to pursue cooperation via licensing, sufficiently high demand may also induce it to compete—also leading to coopetition when the innovator enjoys high bargaining power. Second, the form of competition may also become more offensive (moving from patent wall to bracketing) as demand increases when the technology is strong. Third, we find that an innovator with medium-strength technology may switch between cooperation and competition, or do both concurrently, depending on the state of market demand and other conditions, such as entry cost and relative bargaining power. Specifically, when bargaining power is moderate, it would likely switch from cooperation at low demand to competition at higher demand levels, and vice versa, but when bargaining power is high, any of the three deployment modes may prevail—cooperation for low demand, competition for medium demand, and coopetition for high demand. Thus, if bargaining power is high, we may observe switching between cooperation and coopetition when the innovator’s technological advantage is weak, between competition and coopetition when the advantage is strong, and between all three deployment modes when the technology advantage is medium. 我们上述的几项研究发现对专利部署策略的研究有所贡献。首先，尽管强大的技术使创新者更有可能采取竞争型专利部署模式，但当市场需求以及创新者在许可谈判中的议价能力较高时，它可能会通过许可将竞争与合作结合起来——即采取竞合策略。同样，尽管弱势技术使创新者更有可能通过许可寻求合作，但当需求足够高时，也可能促使其进行竞争——当创新者拥有较高议价能力时，这也会导致竞合。其次，当技术实力较强时，随着需求增加，竞争形式可能会变得更具进攻性（从专利壁垒转向包围）。第三，我们发现，技术实力中等的创新者可能会根据市场需求状况以及其他条件（如进入成本和相对议价能力）在合作与竞争之间切换，或者同时采取这两种策略。具体而言，当议价能力适中时，它可能会从低需求时的合作转向高需求时的竞争，反之亦然；但当议价能力较高时，三种部署模式中的任何一种都可能占主导地位——低需求时的合作、中等需求时的竞争以及高需求时的竞合。因此，如果议价能力较高，当创新者的技术优势较弱时，我们可能会观察到合作与竞合之间的切换；当优势较强时，会观察到竞争与竞合之间的切换；而当技术优势处于中等水平时，则会观察到三种部署模式之间的切换。

Finally, our analysis also has implications for the conditions affecting the option value of the flexibility to switch among deployment modes across time periods and for the impact of demand volatility on the probability of such switches. Generally, the value of the innovator’s option to switch is greatest when it may have an incentive to switch into multiple patent deployment modes. Such switching is more likely when the innovator enjoys high bargaining power, when technology strength is medium, and when demand volatility is high. Higher demand volatility, increasing the range of extreme low and high demand states, exacerbates these switching patterns between competition and either cooperation or coopetition, rendering licensing more beneficial in both very high and low demand states. The above contributions complement and contrast with work in the innovation literature (e.g., Marx et al., 2014) that has focused on supply-side drivers of patent deployment strategy. 最后，我们的分析还对影响跨时间段在部署模式间切换的灵活性的期权价值的条件，以及需求波动性对这种切换概率的影响具有启示意义。一般而言，创新者的切换期权价值在其有动机切换到多种专利部署模式时最大。当创新者拥有较高的议价能力、技术实力处于中等水平且需求波动性较高时，这种切换更有可能发生。更高的需求波动性增加了极端低需求和高需求状态的范围，加剧了竞争与合作或竞合之间的这些切换模式，使得在极高和极低需求状态下许可都更有利。上述贡献与创新文献（例如Marx等人，2014）中的研究互补且形成对比，后者侧重于专利部署策略的供给方驱动因素。

Implications for Future Research

对未来研究的启示

The result that cooperation or coopetition may prevail in high-demand conditions may help explain why such approaches to commercializing a strong technological innovation might prevail in dynamic environments that entertain the prospect of very high levels of demand. Our analysis suggests that if the firm follows a cooperation or coopetition strategy, the joint benefits from cooperation enlarging the market pie may exceed the value from a higher share of a smaller market pie from competing aggressively. A strong technological advantage may also induce patent sleeping (or rival exit) under very low demand, as in EVT’s selloff of its sleeping patented coronary stent technology to Guidant. All of these results merit further consideration and additional research. 在需求旺盛的情况下，合作或竞合可能占主导地位，这一结果或许有助于解释为何将强大技术创新商业化的此类方法在充满极高需求前景的动态环境中可能占主导地位。我们的分析表明，如果企业采取合作或竞合战略，合作带来的扩大市场规模的联合收益可能超过在较小市场中通过激烈竞争获得更高份额的价值。在需求极低的情况下，强大的技术优势也可能导致专利沉睡（或竞争对手退出），例如EVT将其沉睡的冠状动脉支架专利技术出售给Guidant。所有这些结果都值得进一步考虑和深入研究。

Our analysis also reveals severe limitations of traditional NPV approaches that treat the size of the market pie as given, and suggests that future research should adopt S-NPV analysis instead. In optiongames analysis, firm decisions are contingent on both market demand and rivals’ reactions to the firm’s own patent deployment moves. The size and sharing of the market pie are a function of competition, cooperation, or hybrid coopetition strategies pursued by the firm and its rivals, moderated by the level and volatility of demand. Pursuing a competitive strategy may potentially lead to lower overall value due to ensuing patent wars, despite a considerable technological advantage. Hence, the value of a patent deployment strategy may be enhanced by a combination of favorable market conditions and via a cooperation or coopetition approach under high demand and volatility conditions. Even in low demand with a weak or medium technology advantage, the value of the associated patent strategy may be enhanced via licensing in light of future collaboration. 我们的分析还揭示了传统净现值（NPV）方法的严重局限性，这类方法将市场规模视为既定条件，同时表明未来的研究应采用战略净现值（S-NPV）分析。在期权博弈分析中，企业决策取决于市场需求以及竞争对手对企业自身专利部署行动的反应。市场规模及其分配是企业与其竞争对手所采取的竞争、合作或混合竞合策略的函数，同时受需求水平和波动性的调节。尽管拥有相当大的技术优势，但追求竞争策略仍可能因随之而来的专利战而导致整体价值降低。因此，专利部署策略的价值可能会通过有利的市场条件以及在高需求和高波动性条件下采用合作或竞合方法而得到提升。即使在需求较低且技术优势较弱或中等的情况下，考虑到未来的合作，通过许可授权也可能提升相关专利策略的价值。

Another insight worthy of future examination is that demand volatility and extreme outcomes can be value-enhancing as they not only increase growth option value but can also induce firms to switch to a cooperation or coopetition strategy, which reduces the incentive to win at any cost. This hidden potential from high demand volatility often occurs when firms are roughly symmetric with equivalent technologies (e.g., Google and Samsung). High demand volatility and extreme tail outcomes can motivate a richer set of switching patterns between competition and cooperation or coopetition, which is particularly relevant in emerging or dynamic technology industries (Ang, 2008). Additional research is needed to test these predictions and better understand their boundary conditions. 另一个值得未来研究的洞察是，需求波动和极端结果可能具有价值增值作用，因为它们不仅会提高增长期权价值，还能促使企业转向合作或竞合战略，从而降低企业不惜一切代价取胜的动机。这种来自高需求波动的潜在价值通常出现在企业大致对称且技术相当的时候（例如谷歌和三星）。高需求波动和极端尾部结果可以激发竞争与合作或竞合之间更丰富的转换模式，这在新兴或动态技术行业中尤为重要（Ang，2008）。需要进一步研究来验证这些预测，并更好地理解其边界条件。

Implications for Practice

对实践的启示

Our analysis also offers practical implications to intellectual property () managers for a more adaptive patent deployment strategy by identifying conditions that justify pivoting among competition, cooperation, and coopetition modes. For example, a hybrid coopetition mode makes sense if market demand and the innovator’s bargaining power are very high. Under high enough bargaining power to appropriate most of the gains from innovation, licensing out the technology makes sense when man is combined with self-commercialization); thus, extreme demand volatility favors licensing. More generally, the use of option games in strategic decisions gives managers a tool to assess the value of patents and other strategic investments under both demand and strategic uncertainty. S-NPV analysis enables managers to address concurrently the competition versus cooperation and the flexibility versus commitment trade-offs within a single holistic framework and obtain insights on strategy mode switching and path dependency. Such analysis depends on some well-known factors, such as the strength of the new technology, relative market or bargaining power and anticipated rival reactions, but also on some less obvious factors, such as the demand and volatility regimes, which may help to explain more subtle differences in strategic investment behavior. Our results also highlight the imporand attracting the right collaborations (Grindley & Teece, 1997). Creating and managing cooperative relationships, leveraging (“borrowing”) resources outside firm boundaries (Capron & Mitchell, 2012) within a broder alliance portfolio that is “evolving from adapting to shaping and exploiting, according to the state of strategic uncertainty” (Hoffmann, 2007: 849), is a critical dynamic capability. Management should thus be flexible enough to dynamically switch among wait-and-see, competition, cooperation, or coopetition strategy modes as market circumstances warrant. 我们的分析还为知识产权（）管理者提供了实际启示，使其能够通过确定在竞争、合作与竞合模式之间进行战略调整的条件，制定更具适应性的专利部署策略。例如，如果市场需求和创新者的议价能力都非常高，混合竞合模式就会很合适。当创新者拥有足够高的议价能力以获取大部分创新收益时，在自主商业化的同时进行技术许可是合理的；因此，极端的需求波动有利于技术许可。更广泛地说，在战略决策中运用期权博弈为管理者提供了一种工具，使其能够在需求和战略不确定性下评估专利及其他战略投资的价值。S-NPV分析使管理者能够在单一整体框架中同时处理竞争与合作、灵活性与承诺之间的权衡，并深入了解战略模式转换和路径依赖。此类分析取决于一些众所周知的因素，如新技术的强度、相对市场或议价能力以及预期的竞争对手反应，还取决于一些不太明显的因素，如需求和波动状态，这些因素可能有助于解释战略投资行为中更细微的差异。我们的研究结果还强调了吸引合适合作伙伴的重要性（Grindley & Teece, 1997）。建立和管理合作关系、利用公司边界之外的资源（Capron & Mitchell, 2012），并在一个“根据战略不确定性状态从适应转变为塑造和利用”的更广泛联盟组合中进行运作（Hoffmann, 2007: 849），是一项关键的动态能力。因此，当市场环境需要时，管理层应具备足够的灵活性，在观望、竞争、合作或竞合等战略模式之间进行动态切换。

REFERENCES

参考文献

Ang, S. H. 2008. Competitive intensity and collaboration: Impact on firm growth across technological environments. Strategic Management Journal, 29: 1057 1075.
Ang, S. H. 2008. 竞争强度与合作：对不同技术环境下企业增长的影响。《战略管理杂志》，29：1057-1075。

Arora, A. 1997. Patents, licensing, and market structure in the chemical industry. Research Policy, 26: 391403.
Arora, A. 1997. 化学工业中的专利、许可与市场结构。《研究政策》，26: 391-403.

Arora, A., & Ceccagnoli, M. 2006. Patent protection, complementary assets, and firms’ incentives for technology licensing. Management Science, 52: 293308.
Arora, A.，& Ceccagnoli, M. 2006. 专利保护、互补资产与企业技术许可的激励。《管理科学》，52：293308。

Arora, A., & Fosfuri, A. 2003. Licensing the market for technology. Journal of Economic Behavior & Organization, 52: 277295.
Arora, A., & Fosfuri, A. 2003. 技术市场的许可。《经济行为与组织杂志》，52: 277-295。

Arora, A., Fosfuri, A., & Gambardella, A. 2004. Markets for technology: The economics of innovation and corporate strategy. Cambridge, MA: MIT Press.
阿罗拉，A.，福斯富里，A.，& 甘巴德拉，A. 2004. 技术市场：创新与企业战略的经济学。马萨诸塞州剑桥市：麻省理工学院出版社。

Arundel, A., & Patel, P. 2003. Strategic patenting. Background report for the Trend Chart Policy Benchmarking Workshop “New Trends in IPR Policy.” Luxembourg.
阿伦德尔（Arundel）, A.，& 帕特尔（Patel）, P. 2003. 战略性专利申请。趋势图政策基准研讨会“知识产权政策新趋势”的背景报告。卢森堡。

Bar-Ilan, A., & Strange, W. C. 1998. A model of sequential investment. Journal of Economic Dynamics & Control, 22: 437463.
巴-伊兰（Bar-Ilan），A.，& 斯特兰奇（Strange），W. C. 1998. 序贯投资模型。《经济动态与控制杂志》，22: 437-463。

Brandenburger, A., & Nalebuff, B. 1996. Co-opetition. New York, NY: Doubleday.
Brandenburger, A., & Nalebuff, B. 1996. 竞合. 纽约州纽约市: 双日出版社.

Cabral, L. M. B. 2000. Introduction to industrial organization. Cambridge, MA: MIT Press.
卡布拉尔，L. M. B. 2000. 产业组织导论。剑桥，马萨诸塞州：麻省理工学院出版社。

Calabrese, T., Baum, J. A. C., & Silverman, B. S. 2000. Canadian biotechnology start-ups, 19911997: The role of incumbents’ patents and strategic alliances in controlling competition. Social Science Research, 29: 503534.
卡拉布雷斯（Calabrese, T.）、鲍姆（Baum, J. A. C.）、西尔弗曼（Silverman, B. S.）. 2000. 1991-1997年加拿大生物技术初创企业：在位企业专利和战略联盟在控制竞争中的作用. 社会科学研究（Social Science Research）, 29: 503-534.

Capron, L., & Mitchell, W. 2012. Build, borrow, or buy: Solving the growth dilemma. Boston, MA: Harvard Business Review Press.
Capron, L., & Mitchell, W. 2012. 构建、借用或购买：解决增长困境。马萨诸塞州波士顿：哈佛商业评论出版社。

Ceccagnoli, M. 2009. Appropriability, preemption, and firm performance. Strategic Management Journal, 30: 8198.
Ceccagnoli, M. 2009. 可侵占性、抢先行动与企业绩效。《战略管理杂志》，30: 8198。

Chevalier-Roignant, B., & Trigeorgis, L. 2011. Competitive strategy: Options and games. Cambridge, MA: MIT Press.
谢瓦利埃-罗尼亚，B.，&特里吉奥吉斯，L. 2011. 竞争战略：选项与博弈。剑桥，马萨诸塞州：麻省理工学院出版社。

Clark, D. J., & Konrad, K. A. 2008. Fragmented property rights and incentives for R&D. Management Science, 54: 969981.
克拉克，D. J.，& 康拉德，K. A. 2008. 碎片化产权与研发激励。《管理科学》，54: 969981.

Cohen, W., Nelson, R., & Walsh, J. 2000. Protecting their intellectual assets: Appropriability conditions and why U.S. manufacturing firms patent (or not). NBER Working Paper Series No. 7552.
科恩（Cohen, W.）、纳尔逊（Nelson, R.）和沃尔什（Walsh, J.）。2000。保护其知识产权：可获得性条件以及为何美国制造企业申请（或不申请）专利。美国国家经济研究局（NBER）工作论文系列第7552号。

DeLacey, B. J., Herman, K., Kiron, D. J., & Lerner, J. 2006. Strategic behavior in standard-setting organizations. Harvard NOM Working Paper No. 903214. doi: 10.2139/ssrn.903214.
DeLacey, B. J., Herman, K., Kiron, D. J., & Lerner, J. 2006. 标准制定组织中的战略行为. 哈弗NOM工作论文第903214号. doi: 10.2139/ssrn.903214.

Dierickx, I., & Cool, K. 1989. Asset stock accumulation and sustainability of competitive advantage. Management Science, 35: 15041511.
Dierickx, I., & Cool, K. 1989. 资产库存积累与竞争优势的可持续性。《管理科学》，35: 1504-1511.

Gans, J. S., Hsu, D. H., & Stern, S. 2002. When does startup innovation spur the gale of creative destruction? RAND Journal of Economics, 33: 571586.
Gans, J. S., Hsu, D. H., & Stern, S. 2002. When does startup innovation spur the gale of creative destruction? RAND Journal of Economics, 33: 571586.

Gans, J. S., & Stern, S. 2003. The product market and the market for “ideas”: Commercialization strategies for technology entrepreneurs. Research Policy, 32: 333350.
Gans, J. S., & Stern, S. 2003. The product market and the market for “ideas”: Commercialization strategies for technology entrepreneurs. Research Policy, 32: 333350.

Ghemawat, P., & del Sol, P. 1998. Commitment versus flexibility? California Management Review, 40: 2642.
Ghemawat, P.，& del Sol, P. 1998. 承诺与灵活性？《加州管理评论》，40：2642。

Granstrand, O. 1999. The economics and management of intellectual property: Towards intellectual capitalism. Cheletenham, U.K.: Edward Elgar Publishing.
Granstrand, O. 1999. 知识产权的经济学与管理：迈向知识资本主义。英国切尔滕纳姆：爱德华·埃尔加出版社。

Grindley, P. C., & Teece, D. J. 1997. Managing intellectual capital: Licensing and cross-licensing in semiconductors and electronics. California Management Review, 39: 841.
Grindley, P. C., & Teece, D. J. 1997. 管理智力资本：半导体和电子行业的许可与交叉许可。《加州管理评论》，39：841。

Hall, B. H., & Ziedonis, R. H. 2001. The patent paradox revisited: An empirical study of patenting in the U.S. semiconductor industry, 19791995. RAND Journal of Economics, 32: 101128.
霍尔，B. H.，& 齐多尼斯，R. H. 2001. 重新审视专利悖论：1979-1995年美国半导体行业专利申请的实证研究。《兰德经济学杂志》，32：101-128。

Heller, M. A., & Eisenberg, R. S. 1998. Can patents deter innovation? The anticommons in biomedical research. Science, 280: 698701.
Heller, M. A., & Eisenberg, R. S. 1998. 专利能否抑制创新？生物医学研究中的反公地悲剧。《科学》, 280: 698701.

Hill, C. W. L. 1992. Strategies for exploiting technological innovations: When and when not to license. Organization Science, 3: 428441.
Hill, C. W. L. 1992. Strategies for exploiting technological innovations: When and when not to license. Organization Science, 3: 428441.

Hoffmann, W., Lavie, D., Reuer, J. J., & Shipilov, A. 2018. The interplay of competition and cooperation. Strategic Management Journal, 39: 30333052.
Hoffmann, W., Lavie, D., Reuer, J. J., & Shipilov, A. 2018. The interplay of competition and cooperation. Strategic Management Journal, 39: 30333052.

Hoffmann, W. H. 2007. Strategies for managing a portfolio of alliances. Strategic Management Journal, 28: 827 856.
霍夫曼，W. H. 2007. 联盟组合管理策略。《战略管理杂志》，28：827-856。

Hounshell, D., & Smith, J. 1988. Science and corporate strategy: Du Pont R&D, 19021980. Cambridge, MA: Cambridge University Press.
Hounshell, D., & Smith, J. 1988. Science and corporate strategy: Du Pont R&D, 1902-1980. Cambridge, MA: Cambridge University Press.

Lado, A. A., Boyd, N. G., & Hanlon, S. C. 1997. Competition, cooperation, and the search for economic rents: A syncretic model. Academy of Management Review, 22: 110141.
拉多，A. A.，博伊德，N. G.，& 汉隆，S. C. 1997. 竞争、合作与经济租金的探寻：一种融合模型。《管理学会评论》，22：110141。

Levin, R. C., Klevorick, A. K., Nelson, R. R., & Winter, S. G. 1987. Appropriating the returns from industrial research and development. Brookings Papers on Economic Activity, 18: 783831.
莱文，R. C.，克莱沃里克，A. K.，纳尔逊，R. R.，& 温特，S. G. 1987. 工业研究与开发的收益侵占。《布鲁金斯经济活动论文集》，18：783 - 831。

Levitt, T. 1965. Exploit the product life cycle. Harvard Business Review, 43: 8194.
Levitt, T. 1965. 利用产品生命周期。《哈佛商业评论》，43: 8194。

Marx, M., Gans, J. S., & Hsu, D. H. 2014. Dynamic commercialization strategies for disruptive technologies: Evidence from the speech recognition industry. Management Science, 60: 31033123.
马克思, M., 甘斯, J. S., & 徐, D. H. 2014. 颠覆性技术的动态商业化策略：来自语音识别行业的证据。《管理科学》, 60: 31033123.

McGahan, A. M. 2004. How industries evolve: Principles for achieving and sustaining superior performance. Boston, MA: Harvard Business School Press.
麦加汉，A. M. 2004. 产业如何演变：实现和维持卓越绩效的原则。马萨诸塞州波士顿：哈佛商学院出版社。

Mihm, J., Sting, F. J., & Wang, T. 2015. On the effectiveness of patenting strategies in innovation races. Management Science, 61: 26622684.
Mihm, J., Sting, F. J., & Wang, T. 2015. On the effectiveness of patenting strategies in innovation races. Management Science, 61: 26622684.

Polidoro, F., & Toh, P. K. 2011. Letting rivals come close or warding them off? The effects of substitution threat on imitation deterrence. Academy of Management Journal, 54: 369392.
Polidoro, F.，& Toh, P. K. 2011. 让竞争对手靠近还是击退他们？替代威胁对模仿威慑的影响。《管理学会期刊》，54: 369392。

Reitzig, M. 2004a. The private values of “thickets” and “fences”: Towards an updated picture of the use of patents across industries. Economics of Innovation and New Technology, 13: 457476.
Reitzig, M. 2004a. The private values of “thickets” and “fences”: Towards an updated picture of the use of patents across industries. Economics of Innovation and New Technology, 13: 457–476.

Reitzig, M. 2004b. Strategic management of intellectual property. MIT Sloan Management Review, 45: 3540.
Reitzig, M. 2004b. 知识产权的战略管理。《麻省理工学院斯隆管理评论》，45: 35-40.

Rivette, K., & Kline, D. 2000a. Discovering new value in intellectual property. Harvard Business Review, 78: 5466.
里维特（Rivette, K.）和克莱恩（Kline, D.），2000a. 发现知识产权中的新价值。《哈佛商业评论》，78：5466.

Rivette, K., & Kline, D. 200ob. Rembrandts in the attic: Unlocking the hidden value of patents. Boston, MA: Harvard Business School Press.
里维特（Rivette, K.）和克莱恩（Kline, D.）. 200ob. 阁楼里的伦勃朗：解锁专利的隐藏价值. 马萨诸塞州波士顿：哈佛商学院出版社.

Rubinfeld, D. L., & Maness, R. 2005. The strategic use of patents: Implications for antitrust. In F. Lévêque & H. A. Shelanski (Eds.), Antitrust, patents, and copyright: EU and US perspectives: 85102. Cheltenham, U.K. Edward Elgar Publishing.
Rubinfeld, D. L., & Maness, R. 2005. The strategic use of patents: Implications for antitrust. In F. Lévêque & H. A. Shelanski (Eds.), Antitrust, patents, and copyright: EU and US perspectives: 85102. Cheltenham, U.K. Edward Elgar Publishing.

Shapiro, C., & Varian, H. R. 1999. Information rules: A strategic guide to the network economy. Boston, MA: Harvard Business School Press.
夏皮罗（Shapiro, C.）和瓦里安（Varian, H. R.），1999年。《信息规则：网络经济的战略指南》。马萨诸塞州波士顿：哈佛商学院出版社。

Smit, H. T. J., & Trigeorgis, L. 2004. Strategic investment: Real options and games. Princeton, NJ: University Press.
Smit, H. T. J.，& Trigeorgis, L. 2004. 战略投资：实物期权与博弈论。新泽西州普林斯顿：大学出版社。

Smit, H. T. J., & Trigeorgis, L. 2017. Strategic NPV: Real options and strategic games under different information structures. Strategic Management Journal, 38: 25552578.
Smit, H. T. J. 和 Trigeorgis, L. 2017. 战略净现值：不同信息结构下的实物期权与战略博弈。《战略管理杂志》，38：2555-2578。

（注：原句中“25552578”应为页码范围，按学术规范补充分隔符“-”，使格式更准确。）

修正说明：根据学术引用格式，期刊页码通常为范围形式，原输入中“25552578”可能为排版错误，补充“-”后更符合规范，且未改变原意。

Somaya, D. 2003. Strategic determinants of decisions not to settle patent litigation. Strategic Management Journal, 24: 1738.
萨马亚，D. 2003. 专利诉讼中不达成和解的决策的战略决定因素。《战略管理杂志》，24：1738.

Teece, D. J. 1986. Profiting from technological innovation: Implications or integration, collboration, lising, and public policy. Research Policy, 15: 285305.
Teece, D. J. 1986. 从技术创新中获利：整合、合作、许可和公共政策的影响。《研究政策》，15：285-305。

Walsh, J. P., Arora, A., & Cohen, W. M. 2003. Effects of research tool patents and licensing on biomedical innovation. In W. M. Cohen & S. A. Merrill (Eds.), Patents in the knowledge-based economy: 285340. Washington, DC: National Academies Press.
沃尔什（J. P.）、阿拉（A.）和科恩（W. M.）。2003年。研究工具专利及许可对生物医学创新的影响。载于W. M. 科恩与S. A. 梅里尔（编），《知识经济中的专利：285340》。华盛顿特区：美国国家科学院出版社。

Weeds, H. 1999. Sleeping patents and compulsory licensing: An options analysis. University of Warwick Department of Economics Working Paper No. 577. 杂草，H. 1999年。《休眠专利与强制许可：一种选择分析》。华威大学经济系工作论文第577号。

Lenos Trigeorgis (lenos@ucy.ac.cy) is the Bank of Cyprus chair professor of finance at the University of Cyprus and has been a Marie Sklodowska-Curie visiting scholar at MIT. He earned his doctorate from Harvard University. His main research interests are corporate finance, real options and investment decisions under uncertainty, valuation of intangible assets, competition and business strategy. 莱诺斯·特里乔吉斯（lenos@ucy.ac.cy）是塞浦路斯大学塞浦路斯银行金融椅教授，曾是麻省理工学院（MIT）的玛丽·斯克洛多夫斯卡-居里访问学者。他拥有哈佛大学博士学位。他的主要研究兴趣包括公司金融、实物期权以及不确定条件下的投资决策、无形资产估值、竞争与商业战略。

Francesco Baldi (francesco.baldi@unito.it; fbaldi@luiss. it) is an associate professor of corporate finance at the University of Turin and adjunct professor of finance at LUISS Guido Carli University. He earned his PhD from 弗朗切斯科·巴尔迪（francesco.baldi@unito.it；fbaldi@luiss.it）是都灵大学的企业金融副教授，同时也是罗马路易斯大学（LUISS Guido Carli University）的金融兼职教授。他获得了博士学位。

University of Rome III. He has been Chazen visiting scholar at Columbia University and a visiting scholar at MIT. His main research interests are strategic finance, real options theory and its application to valuing intangibles. 罗马第三大学。他曾是哥伦比亚大学的查森访问学者，也是麻省理工学院的访问学者。他的主要研究兴趣是战略金融、实物期权理论及其在无形资产估值中的应用。

Richard Makadok (rmakadok@purdue.edu) is the Brock Family chair professor of strategic management at Purdue University’s Krannert School of Management. He earned his PhD at the University of Pennsylvania’s Wharton School of Business. His primary research interests are competitive advantage and applying economic modeling to business strategy. 理查德·马卡多克（rmakadok@purdue.edu）是普渡大学克兰纳特管理学院布洛克家族战略管理教授。他在宾夕法尼亚大学沃顿商学院获得博士学位。他的主要研究兴趣是竞争优势以及将经济模型应用于商业战略。

M M

Copyright of Academy of Management Review is the property of Academy of Management and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder’s express written permission. However, users may print, download, or email articles for individual use. 《管理学会评论》的版权归管理学会所有，未经版权所有者明确书面许可，其内容不得复制、通过电子邮件发送至多个网站或发布到邮件列表服务器。但是，用户可以为个人使用打印、下载或通过电子邮件发送文章。