Resource Allocation in Strategic Factor Markets: A Realistic Real Options Approach to Generating Competitive Advantage

Factor Markets and Competitive Advantage

The strategic factor market theory of competitive advantage explores mechanisms by which competitive heterogeneity results from firm behavior and market conditions. Our conception of competitive advantage is the relative difference between willingness to pay and cost (e.g., Hoopes, Madsen, & Walker, 2003: 892). That is, the ability to create more economic value than close competitors in a given transaction (Peteraf & Barney, 2003: 314). The factor market literature acknowledges three approaches to generating heterogeneity in factor markets compatible with a real options model.

First, firms may be differentially endowed with productive assets (e.g., Barney, 1991; Peteraf, 1993). This idea has been extended into the real options literature by work that assumes firms are differentially endowed with real options to acquire productive assets in factor markets (Kogut & Kulatilaka, 2001; Maritan & Alessandri, 2007; Myers, 1977). Kogut and Kulatilaka, for example, argue, “A core competence is a scarce factor as Barney (1986) defines it, that embeds complex options on future opportunities” (2001: 747). Firms with different competencies possess different sets of real options.

Second, firms may be differentially endowed with productive complementarities (e.g., Adegbesan, 2009; Lieberman et al., in press; Lippman & Rumelt, 2003). In the context of an option, the presence of the complementary asset increases the value of the focal asset at the time of option execution. Adegbesan argues, “Those with greater complementarity can outbid firms with lesser complementarity, and at least some of the acquiring firms will retain part of the surplus they help create” (2009: 463). For instance, it is possible the option to enter into the smartphone industry during the late 2000s was more valuable to Samsung than rivals, such as Nokia, as a result of Samsung’s complementary (and superior) LCD display technology.

Finally, firms may be differentially endowed with information about the value of assets traded in strategic factor markets (Barney, 1986; Makadok & Barney, 2001). For example, Cisco has superior information about the future value of new technologies as compared to rivals, in part, because their status as an industry leader exposes them to new technology, creating opportunities to interact and to share technological information with many start-up firms. Armed with more information than its rivals, Cisco is able to bid more intelligently for patents. When this sort of differential information endowment is extended to real options, one may consider why, at the time an option is initiated, some firms have better insight into the future value of the asset. In such situations, Maritan and Florence (2008) assert that an auction process will enable firms to acquire assets at less than their value in use.

These three approaches, summarized in Table 1, highlight factor market theories of competitive advantage compatible with real options theory. However, there is nothing inherent in the option itself that generates competitive advantage. The arguments underlying these approaches emphasize market failures that arise from factors external to resource allocation decisions: differential access to options, differential complementary assets, or differential information. In these existing approaches, firms are assumed to be equally effective at executing real options or acquiring assets in factor markets (see Leiblein, 2011, for a discussion concerning the focus on a priori resource endowments in resource-based views of competitive advantage).

OPEN IN VIEWER

Table 1

Existing Approaches to Generating Heterogeneity in Factor Markets That Are Compatible With Real Options Theory

Differential Endowments in Real Options	How Performance Heterogeneity Emerges
Endowments in options (Barney, 1991; Kogut & Kulatilaka, 2001; Myers, 1977)	Differential access to options allows a focal firm to acquire assets in factor markets that other firms cannot acquire.
Productive complementarities (Adegbesan, 2009; Lieberman, Lee, & Folta, in press; Lippman & Rumelt, 2003)	Unique complementarities allow the firm to acquire assets in factor markets because those assets are more valuable to the focal firm than to other firms.
Information (Barney, 1986; Makadok & Barney, 2001; Maritan & Florence, 2008)	Superior information allows the firm to more accurately price an option on an asset and to generate advantage in an otherwise complete market through bidding or exercise.

In the discussion below, we begin to outline mechanisms by which firms may be differentially effective at executing real options on productive assets in factor markets. An important critique of strategic factor market theory is that it “reduces the role of managers in creating heterogeneous resource positions to gathering information . . . and downplays the complex relationship between managers and resources” (Maritan & Peteraf, 2011: 1382). For real options to address issues of competitive advantage, we must bring managers and organizational processes underlying decision making back into the discussion of option execution decisions. The consideration of organizational decision-making processes, moreover, implies an opportunity to explore how nontradable resources (Dierickx & Cool, 1989), such as organizational processes that are built rather than bought, foster the development of heterogeneity and advantage.

Grafting Strategic Factor Markets and Real Options

Grafting real options theory and strategic factor market theory suggests we must recognize that firms face not one but two types of uncertainty. Prospective uncertainty, which relates to the future value of an asset, is fundamental to a real option. Contemporaneous uncertainty, which relates to the current value of an asset, arises because most real assets are not traded in complete markets, and, as such, there is no guaranteed ability to sell an asset at a known market price (Posen, Leiblein, & Chen, 2016).

Recognizing these two forms of uncertainty requires confronting the tension between strategic factor market theory and real options theory. Factor market theory claims that firms “can only obtain greater than normal returns” when they “create or exploit competitive imperfections in strategic factor markets” (Barney, 1986: 1232). An implication of market incompleteness is that markets may poorly provision information on asset values (Denrell, Fang, & Winter, 2003; Lippman & Rumelt, 2003). The real options valuation approach, however, is premised on the assumption that markets are complete: “There is a market for every good” (Flood, 1991: 32). The real options literature recognizes that most real assets are not traded in spot markets (Trigeorgis, 1996). While the literature points to the use of proxy assets and tracking portfolios as a substitute for spot markets (Amram & Kulatilaka, 1999; Lander & Pinches, 1998; Merton, 1998; Trigeorgis, 1996), even these alternatives assume the firm perfectly knows the current value of the asset. The lack of complete information about contemporaneous asset values and the existence of contemporaneous uncertainty are two sides of the same coin—the existence of one implies the presence of the other. ²

The primary contribution, central mechanisms, and existing limitations of the strategic factor market and real options theories are highlighted in Table 2. As noted in the table, real options and strategic factor market theory have complementary contributions and limitations.

OPEN IN VIEWER

Table 2

Extant Theories That Are Grafted Onto Our Behavioral Real Option Theory

	Real Options		Strategic Factor Markets	Feedback Learning
	Valuation	Reasoning
Primary Contribution	Quantifies the value of managerial flexibility.	Highlights the value of managerial flexibility.	Information imperfections can lead to competitive heterogeneity.	A behaviorally realistic model of decision making.
Central Mechanism	Newly arrived information resolves prospective uncertainty that existed in the past.	Newly arrived information resolves prospective uncertainty that existed in the past.	Imperfect nature of information regarding the current value of assets.	The ability to learn (resolve uncertainty) as information arrives that updates asset value.
Existing Limitation	Assumption of perfect information on the contemporaneous value of an asset.	Inability to compare or ordinally rank the payoffs to alternative projects.	Does not explicitly recognize the arrival of new information over time, and learning is assumed to be trivial.	Does not (formally) model changes in asset value.

Grafting Feedback Learning and Real Options

Above, we argue that the incomplete markets assumption inherent in strategic factor market theory can be incorporated into real options by assuming the existence of contemporaneous uncertainty. A key implication is that the exercise/termination decision in the context of a real option is not as simple as comparing a known asset value to its exercise price. Instead, undertaking a real option may require substantial learning—the ability to integrate new information to exercise a contingent claim on an asset in a factor market. In this section, we graft feedback learning and real option theories. We begin by discussing the relationship between contemporaneous uncertainty and information processing and then discuss belief updating.

Realistic Real Options—A Link to Competitive Advantage

Our argument begins by defining a realistic real option as the right but not the obligation to acquire an asset when there is both prospective and contemporaneous uncertainty. Thus, realistic real options may differ in the magnitude of total uncertainty as well as the ratio of contemporaneous to prospective uncertainty. As with traditional options logic, the existence of prospective uncertainty implies that managerial flexibility is valuable. Our addition of contemporaneous uncertainty implies managerial decision making within a realistic option is nontrivial. Firms may be differentially effective in executing realistic real options in factor markets—they may differ in the number and magnitude of the errors they experience, executing options that should have been terminated or terminating options that should have been executed. Our argument proceeds in three steps.

One, firms differ in the accuracy of their subjective beliefs about the value of the asset at the time of the option exercise decision. As noted in our discussion grafting feedback learning to real options, at least two classes of mechanisms exist by which firms endowed with identical options and receiving the same data may differ in their subjective beliefs. First, they may differ in their ability to transform data into information—to manage the information processing activity demanded by realistic real options. Second, given contemporaneous uncertainty, firms may differ in their ability to manage the belief updating process on the basis of noisy new information. Taken together, these mechanisms underlie a learning ability in the context of a realistic real option—the ability to integrate new information to exercise a contingent claim on an asset.

Two, differences in the accuracy of subjective beliefs lead to differences in option execution decisions and, therefore, option execution errors. Heterogeneity in option exercise decisions emerge when subjective beliefs deviate (1) across firms and (2) from the (true) latent value of an asset. For instance, consider a firm that exercises a realistic real option to acquire an asset in a factor market when, in fact, it should have terminated the option. This occurs when a firm’s subjective belief exceeds the exercise price while the true value of the asset does not. Alternatively, the firm might terminate the option and forego asset acquisition in the factor market when it should have exercised the option. ⁴

Three, in the resource allocation process, option execution errors lead to the emergence of competitive heterogeneity and, possibly, competitive advantage. The logic is as follows. Consider two firms, Alpha and Beta, that are equally endowed with shared realistic real options to acquire productive assets in factor markets. In each period, managers make decisions on exercising some of these options. The assets reduce a firm’s cost by a fixed amount. A firm should exercise its option only if the exercise price is less than the cost reduction from owning a particular asset. Exercise errors in realistic options can thus give rise to competitive advantage. For example, because they hold different beliefs on asset value, Alpha might exercise its realistic option to acquire the asset in the factor market when it should have terminated, while Beta terminates. In this case, Beta would have a competitive advantage over Alpha.

Propositions Derived From Realistic Real Options

These differences in information processing and belief updating give rise to differences in option execution decisions in strategic factor markets. These differences in option execution in turn may lead to competitive advantage. We state this basic logic as a series of general propositions that result from our realistic real options theory.

First, realistic real options are subject to not only prospective but also contemporaneous uncertainty. The latter arises because a realistic real option provisions data about the value of the asset rather than unambiguous information. A firm must engage in an information processing task to transform the data provisioned by the option into actionable information. For a given body of data, more effective information processing leads to lower contemporaneous uncertainty. That is, the information about the value of the asset at the time of the option execution decision is less noisy. It is likely that firms vary in their information processing capacity in the context of a realistic real option as a result of differences in top management team (TMT) characteristics, organizational structure, problem formulation routines, or problem solving routines, for example. Thus, a firm with superior information processing capacity will realize comparatively low contemporaneous uncertainty, which suggests the following proposition.

Second, firms are intendedly rational (in an adaptive sense). They transform their new data into information and seek to effectively employ the new information to make their exercise/terminate decision. A key idea underlying the notion of Bayesian learning, as we noted earlier, is that the firm updates its beliefs about the value of the asset in a proportional manner. Its posterior belief about the value of the asset is intermediate to its prior belief and the noisy new information. The optimal rate (i.e., extent) of belief updating is a function of the noisiness of the new information (i.e., level of contemporaneous uncertainty).

The organizational challenge is three-fold. First, the organization must assess the level of contemporaneous uncertainty. Second, the organization must align its belief updating rate to match the level of contemporaneous uncertainty. Third, the organization must form an updated belief that appropriately accounts for the extent of contemporaneous uncertainty. These challenges are likely to be a function of factors including managerial confidence, overoptimism, self-efficacy, and locus of control. Consequently, firms are likely to differ in the rate at which they update their current beliefs to the new information in a real option. Deviations from the Bayesian rate of updating will lead to greater errors in the execution of real options in factor markets, which suggests the following proposition.

Finally, we suggest an interaction between the challenges of information processing and belief updating. When contemporaneous uncertainty is at either extreme, low or high, the decision as to the appropriate rate of belief updating is trivial. With no contemporaneous uncertainty (e.g., as in a financial option), decision makers have a noise-free metric of the asset value (e.g., a Reuters terminal that lists the current market price). Therefore, the optimal rate of belief updating is full updating, and regardless of one’s prior belief, the posterior belief should be the observed asset value at the time of option exercise. At the other extreme, with very high contemporaneous uncertainty, the signal at the time of the exercise decision is pure noise; there is no information in the signal, and the optimal rate of belief updating is no updating. New information should be ignored, and the posterior belief about asset value, which is used to make the option execution/termination decision, should be unchanged from the prior belief. At an intermediate level of contemporaneous uncertainty, which results when firms are moderately effective at information processing, belief updating will be most challenging, which suggests the following proposition.

Broader Implications of Realistic Real Options

Recognition of the role that contemporaneous uncertainty, information processing, and belief updating play in exercise decisions in realistic real options suggests opportunities to strengthen linkages between behavioral approaches to understanding the resource allocation process and economic approaches to understanding the factors that lead to competitive advantage. In this section, we suggest promising linkages between our theory and extant research on cognition, organization structure, problem formulation, and problem solving. Consistent with the core premise underlying Bower (1970) and Burgelman (1983), our approach highlights how competitive advantage arises out of managerial and organizational processes.

The opportunity to link our propositions to extant behavioral work is evident in literature emphasizing the role of TMTs in organizational decision making. This literature discusses how TMT cognition and values affect attention to, perception of, and processing of data (Hambrick & Mason, 1984: 195). Early work argues that TMT demographics proxy for underlying cognitive processes and documented associations between TMT characteristics, such as age, education, and organizational tenure, and the propensity to enact discrete strategic changes (e.g., Bantel & Jackson, 1989; Tihanyi, Ellstrand, Daily, & Dalton, 2000; Wiersema & Bantel, 1992, 1993). Related work identified associations between measures of variance in these TMT characteristics and decision making (e.g., Lant, Milliken, & Batra, 1992; Wiersema & Bantel, 1992).

Our realistic real options theory suggests an opportunity to leverage this early work on cognition by exploring whether TMT demographics affect a firm’s ability to convert data into information, manage the belief updating process, and generate competitive advantage through option execution in strategic factor markets. For instance, future fieldwork may explore whether and how educational differences, such as whether one is educated in the fields of business, design, or engineering, affect the breadth and depth of information considered when determining to exercise or terminate an option. If educational fields that engender broader thinking are better suited to processing information in noisy environments, our work suggests that individuals educated in those fields will make fewer option execution errors as contemporaneous uncertainty increases. Similarly, if demographic factors such as age or organizational tenure are associated with inertial tendencies, older and longer tenured TMTs will be slower to update their beliefs and will make fewer option execution errors in noisy environments (where a slower updating rate is predicted to be preferred). In a laboratory setting, similar propositions may be tested more directly, for instance, by associating individual or group demographic characteristics with the outcomes of experiments aimed to evaluate information processing or belief updating.

While the use of TMT characteristics as proxies for cognitive processes is prevalent in early work, more recent studies recognize the benefit of more direct approaches to examining cognition within TMTs. Consider, for example, differences in the cognitive ability (e.g., J. P. Davis, Eisenhardt, & Bingham, 2009; Felin & Foss, 2005; Foss, 2011; Gavetti & Levinthal, 2000) of the decision-making team and whether these differences affect a firm’s ability to enact better decisions or lead to “behavioral failures.” Gavetti (2012: 270-271) defines “behavioral failures” as factors affecting firms’ ability to compete for opportunities and argues that these failures are driven by TMT members’ rationality, mental plasticity, and shaping ability.

When the insights from these behavioral and cognitive approaches are linked to our propositions, it is possible to derive associations between scores on various mental tests, option execution errors, and the emergence of advantage. For instance, work in behavioral economics links calculative skills to the cognitive reflection test (Frederick, 2005). If TMT members who score higher on the cognitive reflection test (Frederick, 2005) are better able to “see through the mist” surrounding ambiguous and uncertain decisions, they may make more informed decisions, experience fewer option execution errors, and generate advantage. More generally, our work provides a circumstance-contingent explanation for observed differences between individual TMT members and firm value (e.g., Mackey, 2008).

Also consider how cognitive biases affect decisions regarding option execution in strategic factor markets. One research stream where these biases are particularly salient is the entrepreneurial confidence literature (Astebro, Herz, Nanda, & Weber, 2014; Busenitz & Barney, 1997; Camerer & Lovallo, 1999; Lowe & Ziedonis, 2006). A key finding in this literature is that confidence is positively associated with decisions such as investment, entry, and delayed exit after controlling for ability and context. This research suggests that cognitive biases, such as overconfidence, and the extent to which managers learn to overcome these biases, may have important effects on resource allocation and competitive advantage. If personality characteristics, such as one’s degree of confidence, affect belief updating, then our propositions imply that over- and underconfident managers should generate different types of option execution errors. This suggests an opportunity for experimental work to link responses to various surveys regarding individual bias (e.g., Cooper, Woo, & Dunkelberg, 1988) and to examine whether individuals prone to particular biases are more or less likely to make errors when deciding to exercise options in factor markets.

It has long been recognized that organizational structures vary in their ability to focus effort and process information. This work emphasizes how tasks and decision rights are allocated, incentives are aligned, and communications are structured jointly and interactively affect information processing (Daft & Lengel, 1986; Galbraith, 1974; Siggelkow & Rivkin, 2005; Tushman & Nadler, 1978). The effectiveness of a given organizational configuration is contingent on the task environment (Galbraith, 1974; Tushman & Nadler, 1978). Research suggests that the size, breadth, and depth of a hierarchical organization affect the nature and effectiveness of decisions (e.g., Christensen & Knudsen, 2010; Sah & Stiglitz, 1986). Likewise, structured processes that reduce biases may affect belief updating. For instance, Milkman, Chugh, and Bazerman discuss how a “System 2” strategy “involves taking an outsider’s perspective: trying to remove oneself mentally from a specific situation or to consider the class of decisions to which the current problem belongs (Kahneman & Lovallo, 1993)” (2009: 381).

Applying insights from work on organization structure to our theory suggests a number of promising avenues for future research. For instance, if delegation of decision making and splits in authority adversely affect information processing, then organizations exhibiting these characteristics should be more prone to make errors when deciding to exercise or terminate options in factor markets and, therefore, be less likely to generate competitive advantage through the exercise of real options. The efficacy of information processing may also be related to the number of levels or the number of distinct actors that data must traverse before a decision is made. Alternatively, if factors such as an organization’s vertical span of control reduce the rate of belief updating (so that it is less than optimal), our propositions suggest an interaction between the number of vertical levels in a hierarchy and the level of contemporaneous uncertainty would be positively related to the propensity to make errors in the execution of options in factor markets.

Finally, our work also has implications for the burgeoning problem formulation and problem solving literature. As noted by Nickerson and colleagues, problem framing skills have implications for how firms formulate (e.g., Baer, Dirks, & Nickerson, 2013; Nickerson, Yen, & Mahoney, 2012) and attempt to solve (e.g., Nickerson & Zenger, 2004) problems. The problem solving perspective emphasizes that organizations differ in their incentive intensity, communication codes, and dispute resolution regimes. These organizational dimensions affect the hazards of knowledge appropriation and strategic knowledge accumulation, with implications for the optimal form of search (Nickerson & Zenger, 2004). Our theory provides a means to leverage these concepts by testing associations between the use of different problem formulating processes, elements of organization structure, and execution errors in real options. For instance, the use of open, consensus-based problem frames may affect hazards of knowledge appropriation and accumulation and, thus, errors in option execution. Similarly, the nature of communication within a team may affect the transformation of data into information and, thus, decisions to execute or terminate real options.