How Do Performance Shortfalls Shape on Entrepreneurial Orientation? The Role of Managerial Overconfidence and Myopia

Abstract

In an era of rapid technological advancement—particularly with the accelerated development of artificial intelligence and digital technologies—entrepreneurship enables firms to dynamically adjust their strategies in response to environmental uncertainty and helps them maintain sustainable competitive advantages over time. As a key concept in entrepreneurship research, entrepreneurial orientation (EO) has long attracted scholarly attention. However, existing studies on EO have primarily focused on its specific outcomes, while insufficient attention has been paid to its antecedents from the perspective of internal threats. Under the threat of performance shortfalls, firms’ strategic choices are influenced not only by resource constraints but also by managerial cognitive biases. Drawing on Behavioral Theory of the Firm, we explore the moderating roles of managerial overconfidence and myopia in the relationship between performance shortfalls and EO. This study aims to uncover the cognitive “black box” behind why some firms are more likely to trigger entrepreneurial behavior in adverse situations. Based on panel data from 2822 A-share listed companies in China spanning the period from 2009 to 2020, and using a fixed-effects regression model, our findings indicate that both historical and social performance shortfalls have significant positive effects on EO. Further analysis reveals that the positive impact of performance shortfalls on EO is attenuated under conditions of heightened managerial overconfidence and myopia. By enriching the boundary conditions of EO from a cognitive perspective, this study provides a theoretical explanation for how firms can engage in entrepreneurial behavior under threat by reducing cognitive biases, thereby offering both theoretical and managerial insights into how firms can maintain sustainable development under crisis conditions.

1. Introduction

The existing literature suggests that firms may adopt either defensive or offensive strategies when confronted with a crisis [1]. Defensive actions include tighter budgets, cost cutting, and increased accountability [2]. Instead, when dealing with offensive options, firms focus on opportunities for innovation [3]. Crises can also create new opportunities [4], and innovative and entrepreneurial companies can cultivate markets by developing new products or introducing new technologies [5]. Ebersberger and Kuckertz (2021) [6] examined how quickly different types of innovators responded to the opportunities and challenges posed by the COVID-19 crisis. They found that innovative companies are the quickest to react to the changing business environment. These companies actively explore opportunities during a crisis and are not constrained by a specific business model. In contrast, the history and path dependence of conservative companies may limit their search options.

Entrepreneurial orientation (EO), conceptualized as a risk-infused strategic disposition [7], manifests through three constitutive dimensions: (a) innovation, (b) proactivity, and (c) risk-taking [8]. Meta-analytic evidence confirms EO’s measurable influence on organizational growth trajectories, accounting for 18–22% of performance variance across industries [9,10]. This strategic posture operates as both a growth accelerator and performance differentiator, particularly in dynamic environments [11]. Despite established insights into EO’s consequences, its antecedents in crisis contexts remain theoretically underspecified. As Wales et al. (2013) [12] critically observes, “the genesis and sustenance mechanisms of entrepreneurially oriented organizations constitute persistent blind spots in strategic entrepreneurship research.” This gap becomes particularly salient when juxtaposed with behavioral agency theory’s proposition that performance nadirs—operationalized as attainment discrepancies relative to aspiration levels [13]—may paradoxically trigger risk-laden strategic actions.

In developed economies such as the United States and Germany, firms have demonstrated greater agility and strategic adaptability in response to performance shortfalls, particularly under crisis conditions such as the 2008 global financial crisis and the COVID-19 pandemic. Prior studies suggest that enterprises in these contexts often adopt proactive measures—including increased investments in innovation, digital transformation, and organizational renewal—to mitigate environmental threats and maintain competitive viability [14]. In contrast, firms in emerging markets are frequently subject to heightened institutional volatility and resource constraints, which may reshape the mechanisms through which performance feedback influences strategic decision-making processes [15]. As the largest emerging economy, China presents a distinct institutional configuration and a rapidly evolving digital infrastructure, offering a representative and contextually rich setting to investigate how firms formulate strategic responses to performance shortfalls. Accordingly, this study contributes to a deeper understanding of how firms across different national environments develop crisis-responsive strategic behaviors, thereby enriching the comparative literature on decision-making under environmental adversity.

Organizational performance relative to desired levels can elucidate how organizations initiate or terminate a wide range of strategic activities [16,17]. In the long run, consistently achieving performance above desired levels is essential for organizations to survive and thrive, while consistently falling below desired levels can pose significant challenges. Consequently, performance that is below aspirations typically motivates organizations to seek activities that can enhance performance. In other words, it triggers a problemistic search to address underperformance. Scholars have examined how organizations adjust the intensity of R&D, product innovation, investment, growth, and other strategic changes in response to negative performance feedback [18]. However, EO has received relatively little attention to date. It has been suggested that when companies do not meet performance aspiration, they may place greater emphasis on enhancing their EO. In such cases, organizations recognize the need for change, leading to a new approach to business [19]. This shift can accelerate risk-taking and experimentation in search of innovative solutions.

Despite general empirical support for firms’ responses to performance feedback, recent research has begun to recognize that a deeper understanding of the impact of performance feedback on firm behavior requires consideration of moderating factors. Audia and Greve (2006) [20] show that negative achievement differentials reduce risk-taking behavior in small firms while increasing it in large firms. Vissa et al. (2010) [21] found that organizational form influences the response to performance feedback, depending on the type of search field. These studies suggest that richer theoretical insights into the link between performance feedback and firm decision outcomes can be gained by examining the moderators. However, the role of managerial cognitive biases in this context has yet to be explored. Given that behavioral theory emphasizes the risky choices made by firm decision-makers, and that firms’ risky behavioral activism may be constrained by management’s cognition [22], we argue managerial cognitive biases are critical factors to consider. The contribution of our study to the existing literature is twofold. First, it addresses the call for performance feedback as an antecedent of EO [12,23]. In doing so, this study also contributes to and extends the literature on the dynamic nature of EO [24,25], representing a novel application of behavioral theory of the firm. Second, overconfidence and myopia are recognized as the most pervasive and influential managerial biases in existing research. It significantly shapes how individuals perceive and interpret information, ultimately influencing their decision-making behavior. By incorporating them into the performance feedback literature as a previously underexplored factor, we offer fresh perspectives on how performance shortfalls are interpreted. More broadly, it contributes to a deeper understanding of the conditions under which the behavioral theory of the firm’s traditional predictions about firm risk-taking are upheld or potentially challenged.

2. Theoretical Framework

2.1. Firms’ Response to Threats

An insider threat is defined as “any form of dramatic reduction in financial and/or reputational well-being” [26]. However, some recent studies have moved away from this strict definition, arguing that any decline in a company’s performance can serve as a stimulus for a “glass cliff” [27]. Returning to the initial conceptualization of threat, this study defines a firm’s insider threat as the gap between its actual performance and its desired performance [25,26]. The literature suggests that firms may alter their typically conservative tendencies when their growth status is threatened. This change occurs to avoid remaining passive in the face of a larger predicament, prompting firms to adapt their existing practices [28], such as experimenting with new technologies or markets [29], investing more in R&D [30], or reorganizing the firm [31]. Additionally, the literature indicates that the pursuit of expected financial returns becomes more appealing to adversarial firms, as potential improvements in financial returns can help preserve the firm’s sustainable competitive advantage [32].

Behavioral theory posits that comparisons of gain and loss outcomes are always made relative to a reference point [33]. A gain frame-when performance is above the reference point-prompts risk-averse strategic postures aimed at protecting gains [28]. In contrast, a loss frame—when performance is below the reference point—encourages risk-seeking strategic postures [34]. Consistent with the existing literature [24,35], this paper treats entrepreneurship as a practical resource allocation decision in the face of uncertain outcomes. Previous studies have characterized various strategic choices, including venture and divestment [34], capital investment [35], new market entry [36], mergers and acquisitions [36], and investment in innovation [37]. This study uses EO as a proxy for risky strategic posture, as it can lead to uncertain outcomes [9]. EO is defined as “the organizational processes, methods, and styles that firms use to conduct entrepreneurial activities” [38] and is regarded as a strategic posture and behavior that involves innovation, proactivity, and risk-taking [39], with the aim of creating a competitive advantage. While the external environment influences EO [40], it is typically associated with a firm’s strategic responsiveness—its ability to readjust its strategy to align with the environment. Internal factors, such as the business conditions of the firm, also play a crucial role in shaping EO [12].

2.2. Performance Shortfalls and EO

Kahneman and Tversky (1979) [41] criticized the traditional assumption of rationality in classical economics by highlighting that individuals exhibit different attitudes towards risk depending on the context. This distinction clarifies the observation that people tend to be risk-averse in positive situations and risk-seeking in negative situations. These opposing preferences are referred to as “reflexive effects” [33]. Consequently, managers facing threats may exhibit both risk-seeking and risk-averse behaviors. In their study on management’s attitudes toward risk, March and Shapira (1987) [42] argued that managers tend to be risk-averse when organizational performance is strong. Conversely, they are more inclined to make riskier choices in response to organizational failures. Arguments from behavior theory suggest that when faced with a negative achievement gap, decision-makers recognize that the company is not keeping pace with the competition and, therefore, engages in problem searches aimed at closing the performance gap. Consistent with most research in the field of firm behavior, this study focuses on firms’ responses to below-aspiration performance. Specifically, a lack of clarity regarding the reasons for negative achievement gaps may encourage decision-makers to commit to broad strategic activities that involve some degree of risk, such as venture capital investments.

Research in the field of intrapreneurial venture capital indicates that the poor performance in a firm’s core business can motivate the firm to seek growth opportunities in more attractive market segments. Similarly, underperformance relative to peers may drive firms to search for new technologies or capabilities that could enhance their competitive efforts [43]. In both cases, the broad scanning nature of entrepreneurial behavior enables firms to respond effectively to the challenge of relative underperformance. Cyert and March (1963) [19] suggested that search activities are driven by specific problems related to underperformance. According to the behavior theory, firms evaluate their performance against desirable reference points, which include both historical ambitions—reflecting the firm’s past financial performance—and the firm’s current performance relative to its peers [19]. It has been argued that performance below historical aspiration (HPA) or societal aspiration (SPA) triggers problematic exploration among decision-makers who seek solutions to address performance shortfalls, leading to increased risk-taking behavior. Conversely, when performance exceeds aspiration, decision-makers tend to be more risk-averse. This higher performance often results in a “complacency” or “status quo effect,” making firms less likely to engage in risky activities [44,45].

Previous research has highlighted the connection between performance shortfalls and organizational change [46], international entrepreneurship [18], and strategic repositioning. When confronted with significant performance shortfalls, managers recognize that the current strategy is failing to deliver the required level of performance. However, due to a lack of clarity regarding the reasons behind the company’s underperformance, they often resort to risk-oriented strategic activities [44]. Consequently, companies begin to alter their strategic posture to address issues in their current operations [18]. This exploration uncovers strategic options for diversification and innovation [46]. In other words, when performance falls short, one approach is to modify internal activities in search of new products, markets, and strategies. While the initial search may focus on the current situation, subsequent efforts may lead the firm into new business areas, both of which align with a higher EO. Thus, the greater the performance shortfalls, the more likely the firm will broaden its search for business solutions, demonstrating an elevated EO.

Based on this, the following hypotheses are proposed:

H1a. 

Historical performance shortfalls promote EO.

H1b. 

Social performance shortfalls promote EO.

2.3. The Moderating Effects of Managerial Overconfidence

Overconfidence is a significant cognitive bias [47]. Generally, individuals demonstrate the highest levels of overconfidence in actions to which they are deeply committed [48] and tend to believe that outcomes are under their control. This phenomenon is particularly relevant for managers who are highly motivated to achieve positive results for their firms and possess considerable discretion over financial strategies. Importantly, like other cognitive biases, the extent and nature of overconfidence can vary significantly among individuals. Numerous empirical studies have consistently linked managerial overconfidence to various firm outcomes, including acquisitions [49], innovation, and corporate social responsibility [50]. Furthermore, prior theoretical and empirical studies have established that overconfidence exerts a substantial influence on individuals’ processing and interpretation of information pertaining to their past performance, thereby impacting subsequent decision-making [51]. This highlights the critical role of overconfidence in responding to performance feedback [52]. An increasing number of researchers are now delving into the cognitive mechanisms that underlie entrepreneurial behavior. For example, Gudmundsson and Lechner (2013) [53] investigated the effects of various factors, including overconfidence, optimism bias, and trust bias, on the survival of startups. Additionally, Wong et al. (2017) [54] explored the implications of CEO overconfidence for innovation duality.

Due to their exaggerated perceptions of managerial capabilities, overconfident managers are likely to interpret the firm’s anticipated future performance as more favorable than it objectively is, regardless of the level of performance feedback. Overconfident managers confronted with negative feedback may erroneously believe that remaining in a low-risk strategy, which effectively “locks in” expected returns, is the optimal approach to achieve the desired performance outcomes. Consequently, they may perceive less urgency to escalate entrepreneurial activities. In essence, overconfident managers are inclined to pursue innovative and risky strategies only in the presence of significant negative performance feedback; otherwise, they may mistakenly assume that reliance on established operational procedures will suffice to meet their performance objectives. Therefore, this study posits that overconfident managers are likely to perceive negative performance feedback as less severe than their non-overconfident counterparts, resulting in a diminished impact of such feedback on their decision-making processes.

Accordingly, we propose the following hypothesis:

H2a. 

The relationship between EO and historical performance shortfalls is moderated by managerial overconfidence, such that the impact of historical performance shortfalls on EO will be less positive for overconfident managers.

H2b. 

The relationship between EO and social performance shortfalls is moderated by managerial overconfidence, such that the impact of social performance shortfalls on EO will be less positive for overconfident managers.

2.4. The Moderating Effects of Managerial Myopia

Temporal orientation is an unconscious cognitive process that reflects the extent to which individuals prefer past, present, and future considerations in strategic decision-making [55]. Specifically, a myopic orientation among managers refers to their predominant focus on immediate concerns [56]. This inclination towards myopia thinking often results in myopic decision-making, wherein managers prioritize current benefits at the expense of the firm’s long-term development. From a social psychology perspective, temporal orientation is viewed as an intrinsic and stable personal characteristic [57]. When management displays myopic cognitive biases, the firm’s strategies and managerial practices are likely to become overly focused on short-term gains, thereby potentially jeopardizing its long-term viability.

EO is inherently characterized by high levels of investment, risk, and uncertainty. While a robust entrepreneurial orientation can catalyze innovation and growth, it also demands significant resources and can divert managerial attention. Furthermore, short-term performance pressures can adversely affect the firm’s stock price. In order to protect their interests and mitigate potential declines in stock value, myopic management teams may feel compelled to utilize their authority to obstruct the development of an entrepreneurial orientation. Although performance shortfalls can trigger problem searches that encourage a firm’s entrepreneurial orientation, short-sighted managers may resist or even impede the implementation of such initiatives when viewed through the lens of their self-interests. This myopia ultimately reduces the firm’s willingness to engage in entrepreneurial activities during periods of adversity, thereby diminishing the overall level of entrepreneurial orientation.

Based on this analysis, the study proposes the following hypothesis:

H3a. 

The relationship between EO and historical performance shortfalls is moderated by managerial myopia. Specifically, the impact of historical performance shortfalls on EO is less positive for myopic management.

H3b. 

The relationship between EO and societal performance shortfalls is moderated by managerial myopia. Specifically, the impact of societal performance shortfalls on EO is less positive for myopic management.

Our hypothesized relationships are graphically summarized in Figure 1, which outlines the study’s conceptual framework. In this model, performance shortfalls—both historical performance shortfalls and social performance shortfalls—are proposed to promote EO of the firm. These direct positive relationships correspond to Hypothesis 1a and 1b. Additionally, Figure 1 illustrates the proposed moderating roles of two key managerial traits: managerial overconfidence and managerial myopia. As indicated by the downward arrows from these moderator variables to the main effect path, we expect that these traits weaken the positive impact of performance shortfalls on EO. In other words, Hypotheses 2a and 2b posit that when a CEO is overly confident, the motivating effect of historical or social shortfalls on EO will be less positive. Similarly, Hypotheses 3a and 3b posit that a myopic manager will also attenuate the positive relationship between performance shortfalls and EO. This framework sets the stage for our hypothesis tests, with Figure 1 visually depicting H1a–H3b.

3. Research Methodology and Data

3.1. Sample and Data

We employ a sample of A-share companies listed on the Shenzhen and Shanghai Stock Exchanges from 2009 to 2020. The sample selection process was conducted according to the following clearly defined criteria to ensure the accuracy and credibility of the research: (1) Companies within the financial and insurance sectors were excluded to ensure the focus remains on non-financial firms. (2) Observations with a debt-to-equity ratio greater than 1 and companies with a board size of zero were removed from the sample. (3) Companies that were labeled as ST (Special Treatment), * ST, or those flagged with delisting risk during the sample period were excluded to avoid potential biases in performance assessment. (4) Any companies with incomplete data were also excluded from the analysis. After applying these filtering steps, the study ultimately identified a total of 2822 valid publicly listed companies, resulting in a total of 17,304 firm-year level observations. The primary data sources for this research are the CSMAR (China Securities Market and Accounting Research) database and the CNRDS (China National Research Database System). For instances of missing data, supplementary information was gathered from reliable sources, including annual reports of listed companies, the RESSET database, and the Wind database. Additionally, all continuous variables were Winsorized at the 1st and 99th percentiles to mitigate the impact of outliers on the research findings.

3.2. Description of Related Variables

3.2.1. Dependent Variable—Entrepreneurial Orientation

EO is widely conceptualized as a multidimensional construct comprising innovativeness, proactiveness, and risk-taking behavior [58]. According to the studies by Miller and Le Breton-Miller (2011) [58] and Kreiser et al. (2020) [25], financial indicators are utilized to measure EO. Traditionally, EO has been assessed using survey-based data; however, this approach poses challenges for longitudinal tracking of EO due to issues such as self-reporting bias and response inconsistency. In contrast, financial indicators such as R&D intensity offer a consistent, annual, and observable metric of how firms allocate resources toward technological innovation and exploratory activities. Financial metrics provide insight into how firms utilize their resources, capturing tangible actions and outcomes on an annual basis [25]. Thus, using R&D intensity allows us to capture the tangible and investment-driven aspect of innovativeness within the broader EO framework. In alignment with Kreiser et al. (2020) [25], this study posits that financial ratios are the most effective means of capturing EO behavior at the firm level. In this study, we measure innovativeness using R&D intensity, calculated as research and development expenditures divided by total assets. This approach aligns with the prior empirical literature, where R&D intensity is commonly used as a robust and objective proxy for firm-level innovativeness, especially in archival studies relying on longitudinal data [27]. R&D intensity reflects the extent of a firm’s investment in new technologies and the degree to which it fosters exploration through the absorption of new knowledge.

Proactiveness is measured as the percentage of annual earnings reinvested into the company, calculated by dividing retained earnings by total assets [58]. Consistent with the methodology proposed by Miller and Le Breton-Miller (2011) [58], this study computes proactiveness by subtracting the industry average from the reinvestment percentage of each company. This metric serves as a comprehensive indicator of a firm’s pursuit of opportunities and is adjusted for industry factors that influence profit reinvestment. It symbolizes the overall initiative of the firm in establishing a sustainable long-term business.

Risk-taking is commonly assessed using various indicators, including the volatility of return on assets (ROA) [59], stock return volatility, and debt ratios [60]. Given the unique institutional and policy environment in China, the survival conditions for Chinese publicly listed firms may be influenced by government policies. Consequently, this study employs the volatility of ROA as a measure of risk tolerance among Chinese firms. ROA is defined as the earnings before interest and taxes divided by total assets at year-end. To mitigate the effects of industry and temporal variations, this research follows the approach of John et al. (2008) [61] by subtracting the annual industry average from ROA to obtain the adjusted ROA (Adj_ROA) as shown in Equation (3). Subsequently, using a three-year observation period (from year t − 2 to year t), the standard deviation and range of the industry-adjusted ROA are calculated based on Equations (1) and (2). Building on this, the results are multiplied by 100 to derive two risk-taking indicators: Risk1 and Risk2, which serve as measures of the firm’s risk-taking propensity.

$${Risk1}_{i,t}=\sqrt{{\displaystyle \frac{1}{T-1}}\sum _{t=1}^T({Adj\_{Roa}_{i,t}-{\displaystyle \frac{1}{T}}\sum _{t=1}^{T}Adj\_{Roa}_{i,t})}^2}|T=3$$

(1)

$${Risk2}_{i,t}=Max\left({Adj\_Roa}_{i,t}\right)-Min\left({Adj\_Roa}_{i,t}\right)$$

(2)

$$Adj\_Roa={\displaystyle \frac{{EBIT}_{i,t}}{{ASSET}_{i,t}}}-{\displaystyle \frac{1}{X}}\sum _{k=1}^X{\displaystyle \frac{{EBIT}_{i,t}}{{ASSET}_{i,t}}}$$

(3)

Finally, the standardized values of the three dimensions are summed to compute a comprehensive index of EO. To mitigate potential endogeneity issues, the dependent variable in the model utilizes the value from the subsequent period (t + 1). This multidimensional approach follows established practices in EO research and enhances the validity of our dependent variable.

3.2.2. Independent Variable—Performance Shortfalls

Performance shortfalls are operationalized based on the performance feedback theory, which emphasizes the gap between a firm’s actual performance and its aspiration levels [62]. In line with the prior literature [46], this study employs return on assets (ROA) as the primary indicator of firm performance, given its robustness, availability, and comparability across time and industries. In studies related to performance feedback and problem search, the expected levels typically have two sources: historical aspiration and social aspiration [46]. Historical aspiration refers to the anticipated levels based on the company’s past performance history, while social aspiration pertains to the expected levels considering the performance of competitors within a specific industry. To model historical aspiration, this study adopts the approach proposed by Greve (2003) [46], which defines historical aspiration gap (HPA) as a weighted moving average of the previous period’s performance. The formula is as follows:

$$A_{i,t}={\alpha }_1{P}_{i,t-1}+\left(1-{\alpha }_1\right)A_{i,t-1}$$

(4)

In this context, $A_{i,t}$ represents the historical aspiration of company i in year t, while $P_{i,t-1}$ denotes the actual performance of company i in year t − 1. The parameter ${\alpha }_1$ indicates the relative weight assigned to the previous year’s performance level. This study estimates the weight ${\alpha }_1$ by testing different values ranging from 0 to 1 with an increment of 0.1, selecting the weight that maximizes the model’s log-likelihood. After this search, a weight of ${\alpha }_1$ = 0.6 is determined to be allocated to the previous year’s historical aspiration. The historical performance feedback is then calculated as the difference between a company’s actual performance and its historical aspiration. A value less than zero indicates a performance shortfall. It is important to note that $A_{i,0}$ for company i is replaced with the actual performance in year 0. Additionally, following the method outlined by Deb et al. [63], the historical performance shortfalls is truncated: when $P_{i,t}-A_{i,t}$ < 0, it is set to 1; and when $P_{i,t}-A_{i,t}$ ≥ 0, it is set to 0. For the purpose of subsequent analysis, the absolute value of the historical performance shortfalls is taken.

The measurement of social performance shortfalls (SPA) follows a similar methodology. The specific calculation formula is as follows:

$${IE}_{i,t}={\alpha }_{2}I{P}_{i,t-1}+\left(1-{\alpha }_2\right){IE}_{i,t-1}$$

(5)

where ${IE}_{i,t}$ represents the industry performance aspiration for firm i in year t, $I{P}_{i,t-1}$ denotes the median actual performance of firms within the same industry (defined as those sharing the same two-digit SIC code) for year t − 1, and ${\alpha }_2$ represents the relative weight, taking values in the range of [0, 1]. Through a search process to determine the specific value of the weight, the research finds that ${\alpha }_2$ = 0.6 leads to the maximum log-likelihood of the model, which is consistent with the historical performance shortfalls. Based on this, the same approach is applied for trimming and taking the absolute value, ensuring the measurement’s validity and reliability.

3.2.3. Moderating Variable—Managerial Overconfidence

Malmendier and Tate (2008) [49] demonstrated that overconfident executives tend to engage in more M&A as well as investments. Thus, following the approaches of Schrand and Zechman (2012) [64] and Ahmed and Duellman (2013) [65], this study measures overconfidence based on executives’ investment decisions. A Model (6) is established, where the dependent variable is the growth rate of total assets and the independent variable is the growth rate of operating revenue. The residuals obtained from estimating Model (6) are adjusted by subtracting the industry median residuals. If the result is greater than zero, it indicates overconfidence, and a value of 1 is assigned; otherwise, it is assigned a value of 0 [63].

$$y_{i,t}={\beta }_0+{\beta }_1{\left(SalesGrowth\right)}_{i,t}+{\epsilon }_{i,t}$$

(6)

3.2.4. Moderating Variable—Managerial Myopia

Based on the research by Chen et al. (2015) [51], this study uses the tendency of management to reduce R&D expenditures to achieve short-term profit objectives as a proxy for their short-sighted behavior. Specifically, $∆$R&D is calculated as the difference between the R&D expenditure in year t + 1 and that in year t, divided by the total assets at the end of year t, and then multiplied by 100. Management faces immense pressure to avoid losses and declines in earnings, and R&D is often characterized by uncertain outcomes and difficult to measure returns. In this context, short-sighted managers may choose to cut discretionary R&D investments to boost current reported earnings. This behavior has received empirical support from scholars. Therefore, the reduction in R&D expenditure serves as an appropriate proxy variable for managerial myopia.

3.2.5. Control Variables

We control for a range of variables that may influence a firm’s EO. First, we account for firm size (Lnasset), measured by the natural logarithm of total assets. The company’s leverage (Leverage) is included to control for financial risk, as a higher financial leverage may hinder the firm’s ability to engage in entrepreneurial orientation; leverage is measured using the debt ratio. In line with the views of Helmers et al. (2017) [66], this study also controls for board size (Bsize), defined as the total number of members on the board of directors. Organizational redundancy, which represents the available assets within the company, could significantly impact entrepreneurial orientation. Five performance indicators are controlled for in this study: Tobin’s Q (TobinQ), earnings per share (EPS), sales growth rate (Growth), and return on equity (ROE). The level of industry competition is assessed using structural characteristics of the industry, such as product differentiation, industry concentration, average industry profitability, and the number of firms in the industry. The Herfindahl-Hirschman Index (HHI) is employed to measure the degree of industry competition. Additionally, this study controls for variables that may affect entrepreneurial orientation, including ownership concentration (Shrholder), the proportion of shares held by directors (Direct), the shareholding ratio of institutional investors (Insti), and the proportion of independent directors (Indep).

Table 1. Variable descriptions.

Variable Types	Variable Name	Abbreviation	Description
Dependent Variable	Entrepreneurial Orientation	EO	Sum of the standardized values of innovativeness, proactiveness, and risk-taking behavior; measured using data from the subsequent period (t + 1).
Independent Variable	Historical performance shortfalls	HPA	A weighted moving average of the previous period’s performance.
	Social performance shortfalls	SPA	The absolute difference between a firm’s actual ROA and the industry median ROA.
Moderating Variable	Overconfidence	OC	A dummy variable indicating whether the adjusted residual—calculated as the difference between the firm’s asset growth residual and the industry median residual from a regression of asset growth on revenue growth.
	Managerial Myopia	Short	R&D expenditure in year t minus R&D expenditure in year t − 1 divided by total assets at the end of year t, multiplied by 100.
Control Variable	Firm Size	Lnasset	The natural logarithm of total assets.
	Leverage	Leverage	The debt ratio.
	Return on Equity	ROE	A profitability ratio calculated as net income divided by shareholders’ equity.
	Sales Growth Rate	Growth	The year-over-year percentage increase in a company’s sales revenue.
	Tobin’s Q	TobinQ	The ratio of market value to book value of assets.
	Earnings Per Share	EPS	Net profit for the current period/Paid-in capital at the end of the current period.
	Ownership Concentration	Shrholder	The sum of the shareholding ratios of the company’s top five shareholders.
	Industry Concentration	HHI	The squared sum of the ratio of each company’s main business revenue to the total main business revenue of the industry.
	Organizational Redundancy	Slack	(Selling expenses + Administrative expenses + Financial expenses)/Operating income.
	Institutional Shareholding Ratio	Insti	The proportion of shares held by institutional investors to the total shares of the listed company.
	Director Shareholding Ratio	Direct	The ratio of the number of shares held by the board of directors to the total number of shares of the company.
	Board Size	Dsize	The number of board members at the end of the year.
	Independent Director Ratio	Indep	The ratio of the number of independent directors to the total number of board members.

provides an overview of the main variables discussed in this section.

3.3. Analytical Model

To test Hypotheses 1a and 1b, this study constructs the following Models (7) and (8), while controlling for individual, industry, and year fixed effects. To examine the moderating effects of managerial overconfidence and managerial short-sightedness, and to test Hypotheses 2a and 2b, Hypotheses 3a and 3b, the study categorizes the moderating variables into groups and then conducts regressions using Models (9) and (10).

$${EO}_{i,t+1}={\alpha }_0+{\alpha }_1{HPA}_{i,t}+{\alpha }_2{Control}_{i,t}+Fixedeffects+\epsilon$$

(7)

$${EO}_{i,t+1}={\alpha }_0+{\alpha }_1{SPA}_{i,t}+{\alpha }_2{Control}_{i,t}+Fixedeffects+\epsilon$$

(8)

$${EO}_{i,t+1}={\alpha }_0+{\alpha }_1{HPA}_{i,t}+{\alpha }_{2}OC+{\alpha }_{3}Short+{\alpha }_{4}HPA\times OC+{\alpha }_{5}HPA\times Short+{\alpha }_{6}Control+Fixedeffects+\epsilon$$

(9)

$${EO}_{i,t+1}={\alpha }_0+{\alpha }_1{SPA}_{i,t}+{\alpha }_{2}OC+{\alpha }_{3}Short+{\alpha }_{4}SPA\times OC+{\alpha }_{5}SPA\times Short+{\alpha }_{6}Control+Fixedeffects+\epsilon$$

(10)

4. Results

4.1. Descriptive Statistics

Table 2. Descriptive statistics.

VarName	Obs	Mean	Std. Dev.	Min	Median	Max
EO	17,304	−0.213	1.156	−1.641	−0.488	5.070
HPA	17,304	0.019	0.039	0.000	0.003	0.252
SPA	17,304	0.019	0.041	0.000	0.000	0.264
OC	17,304	0.266	2.193	−6.444	0.579	4.699
Short	17,304	0.438	0.972	−1.718	0.183	5.028
Lnasset	17,304	22.229	1.238	20.079	22.041	26.105
Leverage	17,304	0.415	0.193	0.059	0.409	0.858
ROE	17,304	0.058	0.122	−0.605	0.066	0.300
Growth	17,304	0.298	0.641	−0.606	0.135	3.942
TobinQ	17,304	2.090	1.622	0.227	1.623	9.269
EPS	17,304	0.375	0.576	−1.305	0.274	2.840
Shrholder	17,304	0.522	0.148	0.203	0.522	0.876
HHI	17,304	0.119	0.116	0.022	0.079	0.676
Slack	17,304	0.182	0.124	0.023	0.149	0.648
Insti	17,304	0.420	0.247	0.003	0.438	0.901
Direct	17,304	0.130	0.184	0.000	0.008	0.652
Dsize	17,304	8.572	1.624	5.000	9.000	14.000
Indep	17,304	0.375	0.053	0.333	0.357	0.571

summarizes the basic statistical characteristics of the main variables in this study. From the descriptive statistics of the independent variables, the mean of the historical performance shortfalls for the sampled listed companies is 0.019, with a minimum value of 0.000, a maximum value of 0.252, and a standard deviation of 0.039. Similarly, the mean of the social performance shortfalls is also 0.019, with a minimum value of 0.000, a maximum value of 0.264, and a standard deviation of 0.041. These results indicate significant variation in the performance shortfalls among the listed companies. Regarding the dependent variable, the mean EO is −0.213, with a minimum value of −1.641, a maximum value of 5.070, and a standard deviation of 1.156. This also reflects considerable differences in EO among the firms.

4.2. Correlation Analysis

Table 3. Correlation Analysis.

	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18
1. EO	1
2. HPA	0.416 ***	1
3. SPA	0.356 ***	0.793 ***	1
4. OC	−0.047 ***	0.169 ***	0.322 ***	1
5. Short	0.257 ***	−0.072 ***	−0.131 ***	−0.116 ***	1
6. Lnasset	−0.234 ***	−0.069 ***	−0.081 ***	0.101 ***	−0.035 ***	1
7. Leverage	−0.160 ***	0.036 ***	0.171 ***	0.579 ***	−0.058 ***	0.506 ***	1
8. ROE	−0.216 ***	−0.682 ***	−0.863 ***	−0.397 ***	0.186 ***	0.101 ***	−0.176 ***	1
9. Growth	0.023 ***	−0.045 ***	−0.035 ***	0.037 ***	0.095 ***	−0.025 ***	0.016 **	0.027 ***	1
10. TobinQ	0.256 ***	−0.009	−0.068 ***	−0.173 ***	0.161 ***	−0.437 ***	−0.437 ***	0.158 ***	0.065 ***	1
11. EPS	−0.109 ***	−0.439 ***	−0.591 ***	−0.416 ***	0.184 ***	0.249 ***	−0.099 ***	0.750 ***	0.000	0.115 ***	1
12. Shrholder	−0.119 ***	−0.067 ***	−0.146 ***	−0.174 ***	0.028 ***	0.160 ***	−0.026 ***	0.169 ***	−0.025 ***	0.027 ***	0.206 ***	1
13. HHI	−0.091 ***	0.014 *	0.007	0.019 **	−0.086 ***	0.127 ***	0.085 ***	−0.009	−0.018 **	−0.066 ***	0.030 ***	0.084 ***	1
14. Slack	0.240 ***	0.177 ***	0.214***	−0.055 ***	0.061 ***	−0.254 ***	−0.245 ***	−0.132 ***	0.059 ***	0.285 ***	−0.117 ***	−0.098 ***	−0.122 ***	1
15. Insti	−0.167 ***	−0.092 ***	−0.102 ***	−0.009	−0.008	0.442 ***	0.222 ***	0.140 ***	−0.022 ***	−0.093 ***	0.207 ***	0.504 ***	0.113 ***	−0.149 ***	1
16. Direct	0.116 ***	0.024 ***	−0.063 ***	−0.181 ***	0.104 ***	−0.338 ***	−0.292 ***	0.061 ***	0.011	0.198 ***	0.033 ***	0.115 ***	−0.067 ***	0.112 ***	−0.679 ***	1
17. Dsize	−0.110 ***	−0.043 ***	−0.038 ***	0.037 ***	−0.012	0.282 ***	0.180 ***	0.038 ***	−0.020 ***	−0.159 ***	0.048 ***	0.027 ***	0.055 ***	−0.088 ***	0.247 ***	−0.211 ***	1
18. Indep	0.025 ***	0.011	0.024 ***	0.012	−0.007	0.012	−0.013 *	−0.014 *	0.017 **	0.052 ***	−0.001	0.045 ***	−0.006	0.040 ***	−0.065 ***	0.075 ***	−0.503 ***	1

Note: *, **, and *** indicate significance at the 10%, 5%, and 1% levels, respectively (this applies to subsequent tables).

presents the Pearson correlation coefficients among the independent variables, dependent variable, moderating variables, and control variables. The results indicate that all correlation coefficients are below 0.8, and the variance inflation factors (VIF) are below the threshold of 10, suggesting that there are no serious multicollinearity issues among the variables. Furthermore, the results demonstrate that both historical performance shortfalls and social performance shortfalls exhibit a positive and highly significant correlation with EO, thereby providing preliminary support for Hypotheses H1a and H1b.

4.3. Regression Results

To empirically validate our hypotheses, we conducted fixed-effects panel regression analyses. The detailed results are presented in

Table 4. Regression results: Entrepreneurial orientation and HPA (fixed effects).

	(1)	(2)	(3)	(4)
	EOt+1	EOt+1	EOt+1	EOt+1
HPA		7.915 ***	8.948 ***	8.204 ***
		(17.429)	(17.079)	(17.465)
HPA × OC			−3.410 ***
			(−5.192)
HPA × Short				−0.647 **
				(−2.084)
OC	−0.083 ***	−0.082 ***	−0.035 **	−0.082 ***
	(−4.922)	(−5.054)	(−2.020)	(−5.115)
Short	0.093 ***	0.088 ***	0.087 ***	0.100 ***
	(10.018)	(9.771)	(9.757)	(9.837)
Lnasset	−0.125 ***	−0.194 ***	−0.186 ***	−0.193 ***
	(−2.887)	(−4.668)	(−4.517)	(−4.636)
Leverage	0.258	0.333 **	0.313 **	0.329 **
	(1.608)	(2.177)	(2.058)	(2.150)
ROE	−2.362 ***	−0.043	0.206	0.028
	(−11.560)	(−0.182)	(0.842)	(0.118)
Growth	−0.012	0.005	0.006	0.006
	(−0.733)	(0.304)	(0.339)	(0.390)
TobinQ	0.021	0.007	0.004	0.005
	(1.637)	(0.593)	(0.304)	(0.460)
Stock	0.233 ***	0.235 ***	0.216 ***	0.229 ***
	(5.323)	(5.661)	(5.216)	(5.505)
Shrholder	−1.070 ***	−1.019 ***	−0.977 ***	−1.017 ***
	(−4.318)	(−4.325)	(−4.173)	(−4.316)
HHI	0.289	0.293	0.290	0.290
	(1.401)	(1.443)	(1.431)	(1.430)
Slack	0.953 ***	0.624 ***	0.623 ***	0.626 ***
	(3.897)	(2.730)	(2.753)	(2.742)
Insti	0.107	0.052	0.023	0.049
	(0.582)	(0.301)	(0.132)	(0.286)
Direct	−0.646 ***	−0.668 ***	−0.687 ***	−0.673 ***
	(−3.160)	(−3.462)	(−3.576)	(−3.494)
Dsize	0.015	0.017	0.016	0.017
	(1.010)	(1.271)	(1.150)	(1.275)
Indep	−0.412	−0.358	−0.370	−0.341
	(−1.304)	(−1.195)	(−1.239)	(−1.139)
Cons	1.783	2.994 **	2.837 **	2.965 **
	(1.300)	(2.307)	(2.203)	(2.285)
Year	Yes	Yes	Yes	Yes
Industry	Yes	Yes	Yes	Yes
N	13,877	13,877	13,877	13,877
adj. R2	0.166	0.213	0.217	0.214

Note: t statistics in parentheses ** p < 0.05, *** p < 0.01.

and

Table 5. Regression results: Entrepreneurial orientation and SPA (fixed effects).

	(1)	(2)	(3)	(4)
	EOt+1	EOt+1	EOt+1	EOt+1
SPA		9.132 ***	9.498 ***	9.362 ***
		(14.917)	(15.379)	(15.208)
SPA×OC			−4.648 ***
			(−6.229)
SPA × Short				−0.973 ***
				(−3.282)
OC	−0.083 ***	−0.075 ***	−0.021	−0.077 ***
	(−4.922)	(−4.550)	(−1.153)	(−4.633)
Short	0.093 ***	0.089 ***	0.088 ***	0.102 ***
	(10.018)	(9.732)	(9.595)	(9.996)
Lnasset	−0.125 ***	−0.105 **	−0.094 **	−0.103 **
	(−2.887)	(−2.534)	(−2.270)	(−2.491)
Leverage	0.258	−0.014	−0.025	−0.026
	(1.608)	(−0.091)	(−0.158)	(−0.169)
ROE	−2.362 ***	0.576 **	0.626 **	0.625 **
	(−11.560)	(2.167)	(2.381)	(2.362)
Growth	−0.012	−0.004	−0.004	−0.003
	(−0.733)	(−0.248)	(−0.254)	(−0.205)
TobinQ	0.021	0.001	−0.001	−0.001
	(1.637)	(0.083)	(−0.120)	(−0.078)
Stock	0.233 ***	0.178 ***	0.154 ***	0.171 ***
	(5.323)	(4.624)	(4.061)	(4.475)
Shrholder	−1.070 ***	−0.931 ***	−0.913 ***	−0.929 ***
	(−4.318)	(−3.951)	(−3.884)	(−3.936)
HHI	0.289	0.303	0.300	0.306
	(1.401)	(1.486)	(1.489)	(1.506)
Slack	0.953 ***	0.537 **	0.581 **	0.542 **
	(3.897)	(2.353)	(2.561)	(2.373)
Insti	0.107	0.068	0.041	0.065
	(0.582)	(0.403)	(0.246)	(0.388)
Direct	−0.646 ***	−0.582 ***	−0.606 ***	−0.589 ***
	(−3.160)	(−3.047)	(−3.183)	(−3.078)
Dsize	0.015	0.010	0.010	0.011
	(1.010)	(0.741)	(0.734)	(0.754)
Indep	−0.412	−0.535 *	−0.527 *	−0.497
	(−1.304)	(−1.756)	(−1.729)	(−1.636)
Cons	1.783	1.472	1.224	1.429
	(1.300)	(1.167)	(0.977)	(1.133)
Year	Yes	Yes	Yes	Yes
Industry	Yes	Yes	Yes	Yes
N	13,877	13,877	13,877	13,877
adj. R2	0.166	0.199	0.203	0.200

Note: t statistics in parentheses. * p < 0.1, ** p < 0.05, *** p < 0.01.

. We first examine the direct effects (H1a and H1b), followed by the moderating effects (H2a, H2b, H3a, H3b), as explained below:

Main Effects (H1a and H1b): Table 4 examines the relationship between historical performance shortfalls and EO, while Table 5 focuses on the relationship between social performance shortfalls and EO. In Model (2) of Table 4, the direct impact of historical performance shortfalls on EO is assessed. The coefficient for historical performance shortfalls is positive (β = 7.915, p < 0.01), indicating a statistically significant relationship when predicting EO, thus supporting Hypothesis 1a. Similarly, Model (2) in Table 5 provides empirical evidence supporting Hypothesis 1b, showing that the coefficient for social performance shortfalls is also positive (β = 9.132, p < 0.01) and highly significant in relation to EO.

Moderating Effects (H2a, H2b, H3a, and H3b): In building upon the validated model regarding the impact of performance shortfalls on EO, this section further examines the moderating roles of managerial overconfidence and managerial myopia in the relationship between performance shortfalls and EO. The regression results are presented in Table 4 and Table 5. In Table 4, Model (3) examines the moderating effect of managerial overconfidence on the relationship between historical performance shortfalls and EO, with an interaction term coefficient of −3.410, which is significant at the 1% level, thus validating Hypothesis 2a. Model (4) in Table 4 explores the moderating effect of managerial short-termism on the relationship between historical performance shortfalls and EO, yielding an interaction term coefficient of −0.647, also significant at the 1% level, validating Hypothesis 3a. In Table 5, Model (3) assesses the moderating effect of managerial overconfidence on the relationship between social performance shortfalls and EO, revealing an interaction term coefficient of −4.648, significant at the 1% level, thus validating Hypothesis 2b. Model (4) in Table 5 examines the moderating effect of managerial short-termism on the relationship between social performance shortfalls and EO, yielding an interaction term coefficient of −0.973, which is significant at the 1% level, thereby validating Hypothesis 3b.

Overall, the regression results in Table 4 and Table 5 align well with our theoretical expectations. Hypotheses 1a and 1b are fully supported by the data, as both historical and social performance shortfalls show a significant positive relationship with EO. Furthermore, the moderation tests generally support Hypotheses 2a, 2b, 3a, and 3b: in both domains (historical and social), higher managerial overconfidence or greater managerial myopia attenuate the positive effect of performance shortfalls on EO. These findings reinforce our argument that while performance shortfalls push firms toward greater entrepreneurial initiatives, this push is significantly tempered by cognitive traits of the top management. Managers who are overconfident or fixated on short-term results do not respond as strongly to performance gaps, leading to a less pronounced increase in EO under such conditions. This pattern of results is exactly in line with the moderating relationships hypothesized in H2a, H2b, H3a, and H3b, thereby providing empirical support for the full theoretical model illustrated in Figure 1.

5. Robustness Check

5.1. Dependent Variable Substitution

Based on existing research, we utilize the three aforementioned indicators of EO to reconstruct a comprehensive measure of entrepreneurial orientation intensity. These three indicators form different combinations within a three-dimensional probability space, with each combination corresponding to a distinct state of EO. This comprehensive indicator allows for a more thorough understanding of the complexity of EO and assists researchers and decision-makers in more accurately evaluating and analyzing the actual situation of EO. The specific calculation method is as follows: First, let x represent the proportion of research and development expenditures to total assets for the i-th company in year t within the two-dimensional probability space, y represent the percentage of annual earnings reinvested into the company (i.e., the proportion of retained earnings to total assets) for the i company in year t, and z represent the level of earnings volatility for the i-th company in year t. Thus, (x, y, z) reflects the EO state of the i-th company in year t. Second, calculate the Euclidean distance from the origin (0, 0, 0) to the entrepreneurial orientation (x, y, z) of the i-th company in year t. The square root of this distance is defined as the intensity of the company’s EO. The formula is as follows:

$$Entrepreneurial\_Orientation=\sqrt{(x_{i,t}-0)+(y_{i,t}-0)+(z_{i,t}-0)}=\sqrt{x_{i,t}+y_{i,t}+z_{i,t}}$$

The smaller the value of Entrepreneurial_Orientation, the lower the EO intensity; conversely, a higher value indicates stronger EO intensity. In other words, the closer the EO position is to (0, 0, 0) in the three-dimensional space, the weaker the EO; the farther the position is from (0, 0, 0), the stronger the EO. Regression analysis was conducted on the performance shortfalls and EO after replacing the measurement method. The results are shown in

Table 6. Regression results of HPA and EO.

	(1)	(2)	(3)	(4)
	EOt+1	EOt+1	EOt+1	EOt+1
HPA		0.306 ***	0.507 ***	0.369 ***
		(2.743)	(4.407)	(3.281)
HPA × OC			−0.717 ***
			(−4.857)
HPA × Short				−0.143 **
				(−2.416)
OC	−0.011 **	−0.011 **	−0.001	−0.011 **
	(−2.334)	(−2.301)	(−0.184)	(−2.338)
Short	0.002	0.002	0.002	0.005
	(0.922)	(0.859)	(0.790)	(1.631)
Lnasset	−0.057 ***	−0.059 ***	−0.058 ***	−0.059 ***
	(−4.824)	(−5.035)	(−4.890)	(−5.021)
Leverage	0.000	0.003	−0.002	0.002
	(0.000)	(0.068)	(−0.046)	(0.045)
ROE	−0.149 ***	−0.059	−0.011	−0.043
	(−3.780)	(−1.110)	(−0.203)	(−0.812)
Growth	0.006	0.007 *	0.007 *	0.007 *
	(1.490)	(1.647)	(1.674)	(1.716)
TobinQ	−0.002	−0.003	−0.003	−0.003
	(−0.762)	(−0.908)	(−1.175)	(−1.027)
Stock	0.040 ***	0.040 ***	0.036 ***	0.039 ***
	(3.544)	(3.579)	(3.228)	(3.468)
Shrholder	−0.136 **	−0.135 **	−0.126 **	−0.134 **
	(−2.308)	(−2.292)	(−2.153)	(−2.284)
HHI	0.254 ***	0.254 ***	0.254 ***	0.254 ***
	(3.061)	(3.059)	(3.055)	(3.053)
Slack	0.068	0.054	0.055	0.055
	(1.255)	(1.005)	(1.016)	(1.018)
Insti	0.077	0.075	0.069	0.075
	(1.635)	(1.603)	(1.475)	(1.597)
Direct	0.030	0.030	0.026	0.029
	(0.652)	(0.644)	(0.559)	(0.628)
Dsize	0.003	0.003	0.003	0.003
	(0.656)	(0.683)	(0.591)	(0.683)
Indep	0.056	0.058	0.055	0.062
	(0.588)	(0.611)	(0.581)	(0.651)
Cons	1.777 ***	1.824 ***	1.792 ***	1.818 ***
	(6.191)	(6.365)	(6.283)	(6.353)
Year	Yes	Yes	Yes	Yes
Industry	Yes	Yes	Yes	Yes
N	13,877	13,877	13,877	13,877
adj. R2	0.222	0.223	0.225	0.223

Note: t statistics in parentheses. * p < 0.1, ** p < 0.05, *** p < 0.01.

and

Table 7. Regression results of SPA and EO.

	(1)	(2)	(3)	(4)
	EOt+1	EOt+1	EOt+1	EOt+1
SPA		0.442 ***	0.502 ***	0.470 ***
		(2.904)	(3.280)	(3.065)
SPA × OC			−0.786 ***
			(−4.592)
SPA × Short				−0.120 *
				(−1.880)
OC	−0.011 **	−0.011 **	−0.002	−0.011 **
	(−2.334)	(−2.252)	(−0.286)	(−2.283)
Short	0.002	0.002	0.002	0.004
	(0.922)	(0.847)	(0.734)	(1.362)
Lnasset	−0.057 ***	−0.056 ***	−0.054 ***	−0.055 ***
	(−4.824)	(−4.762)	(−4.589)	(−4.743)
Leverage	0.000	−0.013	−0.015	−0.015
	(0.000)	(−0.349)	(−0.395)	(−0.388)
ROE	−0.149 ***	−0.007	0.001	−0.001
	(−3.780)	(−0.098)	(0.017)	(−0.009)
Growth	0.006	0.006	0.006	0.006
	(1.490)	(1.591)	(1.577)	(1.607)
TobinQ	−0.002	−0.003	−0.004	−0.003
	(−0.762)	(−1.089)	(−1.226)	(−1.164)
Stock	0.040 ***	0.037 ***	0.033 ***	0.036 ***
	(3.544)	(3.316)	(2.949)	(3.242)
Shrholder	−0.136 **	−0.129 **	−0.126 **	−0.129 **
	(−2.308)	(−2.198)	(−2.151)	(−2.194)
HHI	0.254 ***	0.255 ***	0.255 ***	0.256 ***
	(3.061)	(3.070)	(3.065)	(3.074)
Slack	0.068	0.048	0.055	0.049
	(1.255)	(0.884)	(1.023)	(0.894)
Insti	0.077	0.075	0.071	0.075
	(1.635)	(1.601)	(1.507)	(1.597)
Direct	0.030	0.033	0.029	0.032
	(0.652)	(0.721)	(0.637)	(0.703)
Dsize	0.003	0.003	0.003	0.003
	(0.656)	(0.608)	(0.602)	(0.612)
Indep	0.056	0.050	0.052	0.055
	(0.588)	(0.527)	(0.544)	(0.578)
Cons	1.777 ***	1.761 ***	1.719 ***	1.756 ***
	(6.191)	(6.193)	(6.052)	(6.176)
Year	Yes	Yes	Yes	Yes
Industry	Yes	Yes	Yes	Yes
N	13,877	13,877	13,877	13,877
adj. R2	0.222	0.223	0.224	0.223

Note: t statistics in parentheses. * p < 0.1, ** p < 0.05, *** p < 0.01.

. Table 6 presents the regression results of the historical performance shortfalls and EO, while Table 7 presents the regression results of the social performance shortfalls and EO. From the results, it can be observed that the findings remain robust after the replacement of the EO measurement method.

5.2. Controlling for Correlated Independent Variables

To further ensure the robustness of the research results, this study addresses the relatively high correlation between the independent variables—historical performance shortfalls and social performance shortfalls. Given the potential strong association between the two, this study takes further consideration when examining the relationship between performance shortfalls and EO. Specifically, when analyzing the impact of historical performance shortfalls, social performance shortfalls are included as a control variable in the model, and vice versa. The results continue to support the research hypotheses, confirming the robustness of the findings.

5.3. Endogeneity Analyses

Considering the potential endogeneity issues between performance shortfalls and EO, we employed the propensity score matching (PSM) method to conduct robustness tests on the hypotheses. We constructed a dummy variable based on the mean of the performance shortfalls; if the performance shortfalls were above the mean, the dummy variable was assigned a value of 1; otherwise, it was assigned a value of 0, thereby categorizing companies into an experimental group and a control group. Next, we used the dummy variable as the dependent variable and all control variables included in the main regression as independent variables, employing a logit model for regression analysis to calculate the propensity scores for the observed samples. Following this, we performed matching between the experimental and control groups based on a 1-to-4 nearest neighbor matching principle. After matching, the standardized bias of all covariates was less than 10%, and most covariates did not exhibit significant differences post-matching. Finally, we conducted regression analysis on the successfully matched samples. Specifically, the regression coefficients for both the interaction of historical performance shortfalls and overconfidence/myopia, and the interaction of social performance shortfalls and overconfidence/myopia were significantly positive. This indicates that after employing the propensity score matching method to enhance comparability between samples, the research findings remain robust.

Given that the study’s results may be biased due to omitted variables or measurement errors, leading to endogeneity issues, this research adopts the instrumental variable method to address this potential problem. This study uses the industry-year mean of performance shortfalls as the instrumental variable. This instrumental variable satisfies both relevance and exogeneity conditions. The study conducted tests for weak instruments and over-identification, both of which met the requirements for selecting the instrumental variable. The results indicate that after considering the potential endogeneity between performance shortfalls and corporate EO, both historical performance shortfalls and social performance shortfalls still have positive coefficients with significance at the 1% level, demonstrating that performance shortfalls can significantly promote corporate EO, consistent with previous findings. This further verifies the robustness of the results in this study.

6. Supplementary Analyses

6.1. Ownership

Previous studies have pointed out significant differences in operational objectives between state-owned and non-state-owned enterprises (NSOE), with the former typically possessing both political and economic identities [67]. Given this context, it is reasonable to expect that the boards of directors in non-state-owned and state-owned enterprises may play different roles in shaping entrepreneurial behavior. To further explore this issue, the research sample is divided into two subsamples based on whether the enterprise is state-owned. The regression results are presented in

Table 8. Heterogeneity analysis of ownership.

	(1)	(2)	(3)	(4)
	EOt+1	EOt+1	EOt+1	EOt+1
	SOE	SOE	NSOE	NSOE
HPA	6.694 ***		7.983 ***
	(9.477)		(12.658)
SPA		6.205 ***		9.376 ***
		(6.707)		(11.152)
Lnasset	−0.125 **	−0.083	−0.235 ***	−0.119 **
	(−2.118)	(−1.413)	(−4.010)	(−2.050)
Leverage	0.275	0.013	0.301 *	−0.004
	(1.495)	(0.074)	(1.654)	(−0.024)
ROE	0.353	0.451	−0.256	0.483
	(1.490)	(1.645)	(−0.722)	(1.147)
Growth	0.003	0.002	0.023	0.009
	(0.163)	(0.105)	(0.940)	(0.364)
TobinQ	0.030	0.024	0.023	0.020
	(1.578)	(1.200)	(1.558)	(1.335)
Stock	0.124 **	0.108 **	0.308 ***	0.210 ***
	(2.494)	(2.279)	(5.269)	(3.792)
Shrholder	−0.263	−0.279	−0.976 ***	−0.885 ***
	(−0.764)	(−0.783)	(−3.137)	(−2.816)
HHI	−0.039	−0.063	0.427	0.468
	(−0.175)	(−0.286)	(1.399)	(1.535)
Slack	0.855 ***	0.711 **	0.673 **	0.638 **
	(2.722)	(2.126)	(2.370)	(2.273)
Insti	−0.050	−0.008	−0.143	−0.127
	(−0.195)	(−0.030)	(−0.700)	(−0.635)
Direct	−1.310	−1.602	−0.538 ***	−0.456 **
	(−0.653)	(−0.805)	(−2.606)	(−2.240)
Dsize	−0.021	−0.028 *	0.046 **	0.040 *
	(−1.530)	(−1.850)	(2.150)	(1.809)
Indep	−0.535 *	−0.694 **	−0.066	−0.165
	(−1.693)	(−2.125)	(−0.142)	(−0.348)
Cons	1.232	0.729	3.542 ***	1.316
	(0.647)	(0.382)	(2.728)	(1.013)
Year	Yes	Yes	Yes	Yes
Industry	Yes	Yes	Yes	Yes
N	4728	4728	9149	9149
adj. R2	0.103	0.072	0.236	0.226

Note: t statistics in parentheses. * p < 0.1, ** p < 0.05, *** p < 0.01.

. The results for the non-state-owned enterprise sample indicate that the coefficients for historical performance shortfalls and social performance shortfalls are 7.983 and 9.376, respectively, both of which are significantly positive at the 1% level. In contrast, the results for the state-owned enterprise sample show that, unlike the non-state-owned enterprise sample, the performance shortfalls significantly influence EO in state-owned enterprises, but the coefficient is smaller than that of the non-state-owned enterprises. Furthermore, the inter-group coefficient difference test indicates a significant difference between the regression coefficients of the two groups. These results suggest that the impact of performance shortfalls on EO is more pronounced in non-state-owned enterprises. This may be attributed to the more uncertain external environment faced by non-state-owned enterprises. In situations where there is a threat of poor performance, non-state-owned enterprises need a stronger entrepreneurial orientation to turn around performance downturns and re-establish competitive advantages in market competition, thereby acquiring scarce resources to achieve growth in adversity.

6.2. Digital Transformation Degree

The development of digital technologies and infrastructures continues to provide opportunities for creating new businesses. Digital transformation refers to the transformation of business processes and value creation using digital technologies such as the Internet of Things, artificial intelligence, machine learning, and cloud computing [68]. It emphasizes how to make effective decisions in uncertain environments and has become the key to entrepreneurship. Digital transformation changes organizational forms and gives rise to new business models, opening up new growth points for entrepreneurial development. Thus, under different degrees of digital transformation, the EO intentions of enterprises will vary significantly. Based on the annual sample median for digital transformation, samples above the median are coded as 1, while those below the median are coded as 0.

From the regression results shown in

Table 9. Heterogeneity analysis of digital transformation degree.

	(1)	(2)	(3)	(4)
	EOt+1	EOt+1	EOt+1	EOt+1
	Digital = 1	Digital = 1	Digital = 0	Digital = 0
HPA	8.510 ***		6.688 ***
	(12.426)		(11.620)
SPA		9.057 ***		6.793 ***
		(10.572)		(7.327)
Lnasset	−0.307 ***	−0.201 ***	−0.196 ***	−0.127 *
	(−4.941)	(−3.294)	(−2.744)	(−1.775)
Leverage	0.255	−0.039	0.602 ***	0.347 *
	(1.092)	(−0.158)	(3.176)	(1.795)
ROE	0.414	0.846 **	−0.236	−0.010
	(1.292)	(2.317)	(−0.751)	(−0.026)
Growth	0.042 *	0.027	0.027	0.019
	(1.851)	(1.164)	(1.017)	(0.697)
TobinQ	−0.039 **	−0.046 ***	0.059 ***	0.057 ***
	(−2.337)	(−2.748)	(3.295)	(3.063)
Stock	0.200 ***	0.119 **	0.259 ***	0.206 ***
	(3.337)	(2.002)	(4.402)	(3.697)
Shrholder	−1.867 ***	−1.648 ***	−0.402	−0.399
	(−4.918)	(−4.345)	(−1.231)	(−1.205)
HHI	0.605 **	0.574 *	0.240	0.237
	(2.041)	(1.840)	(0.938)	(0.926)
Slack	0.850 ***	0.811 **	0.916 ***	0.828 ***
	(2.594)	(2.489)	(2.932)	(2.591)
Insti	0.362	0.351	0.001	0.022
	(1.292)	(1.267)	(0.004)	(0.094)
Direct	−0.810 ***	−0.655 **	−0.428	−0.410
	(−2.797)	(−2.307)	(−1.515)	(−1.427)
Dsize	0.034	0.026	−0.007	−0.011
	(1.604)	(1.229)	(−0.441)	(−0.626)
Indep	0.404	0.099	−1.106 ***	−1.188 ***
	(0.865)	(0.209)	(−2.801)	(−2.930)
Cons	7.521 ***	5.492 ***	2.681 *	1.601
	(5.650)	(4.240)	(1.655)	(0.980)
Year	Yes	Yes	Yes	Yes
Industry	Yes	Yes	Yes	Yes
N	7102	7102	6775	6775
adj. R2	0.235	0.219	0.158	0.138

Note: t statistics in parentheses. * p < 0.1, ** p < 0.05, *** p < 0.01.

, it can be observed that in different contexts of digitalization, the gap between historical and social performance shortfalls has a positive impact on EO, which is significant at the 1% level. Additionally, the inter-group coefficient difference test indicates a significant difference between the regression coefficients of the two groups. For enterprises with a higher degree of digital transformation, the gap between historical and social performance shortfalls has a greater impact on EO, suggesting that digital transformation can enhance the problem-searching capability of underperforming enterprises, thereby further promoting entrepreneurial behavior.

7. Conclusions, Policy Recommendations and Discussion

7.1. Conclusions and Contributions

This study set out to examine how internal organizational crises—specifically, performance shortfalls relative to aspirations—shape a firm’s entrepreneurial orientation (EO). Our findings confirm that internal performance threats can indeed act as a catalyst for entrepreneurial behavior. Firms operating below their aspiration levels tend to pursue more innovative, risk-taking, and proactive strategies, indicating that adversity within the organization can spur a search for new opportunities. This result extends the scope of EO research beyond external triggers (like market turbulence or disruptive events) to include internal triggers. In doing so, it demonstrates that EO is not solely a response to external pressures but also a deliberate strategic reaction to challenges arising from within the firm.

Our evidence aligns with core tenets of the Behavioral Theory of the Firm and prospect theory. When performance falls below targets, decision-makers appear more willing to deviate from the status quo, engage in problemistic search, and take strategic risks. This behavior supports the classic argument that organizations respond to performance shortfalls by seeking innovative solutions. By showing that firms increase EO when facing internal shortfalls, we provide empirical confirmation of this behavioral tendency within a strategic entrepreneurship context. Furthermore, we contribute to the theoretical debate on organizational responses to adversity. An alternative perspective, threat-rigidity theory, would predict that threats lead to conservatism and rigidity. In contrast, our results show a pattern of increased entrepreneurial action under internal threat. This suggests that, under the conditions we studied (moderate performance shortfalls rather than existential crises), the drive to recover and seize new opportunities outweighs any impulse to retreat. By suggesting that only more extreme threats trigger the paralysis described by threat-rigidity, this finding helps explain why some prior studies observed reduced risk-taking under severe threat while we observe the opposite in more moderate circumstances.

Beyond the overall effect of performance shortfalls, we uncover important sources of heterogeneity in how firms respond. Managerial cognitive biases emerged as critical contingent factors. Overconfident top managers significantly dampen the firm’s tendency to respond entrepreneurially to performance shortfalls. These leaders, due to unwarranted faith in their current strategies, may discount or ignore negative performance feedback, and thus they initiate fewer innovative or risky changes than their less overconfident counterparts. While overconfidence is often associated with higher general risk-taking, our findings nuance this view by showing that overconfidence can impede the specific adaptive risk-taking triggered by performance feedback. Similarly, managerial myopia weakens the positive relationship between performance shortfalls and EO. Myopic managers may prefer quick wins or cost-cutting to improve short-term metrics, rather than investing in new ventures or creative projects that pay off in the longer run. Together, these findings contribute a cognitive perspective to EO theory: performance feedback will spur entrepreneurial action only if decision-makers are psychologically willing to heed the warning signs and take a long-range view. We thereby explain one reason why some firms fail to pivot or innovate when performance declines—the biases of their leaders act as brakes on the very behaviors that the Behavioral Theory of the Firm would predict.

In addition to managerial traits, organizational context conditions the strength of EO responses to internal crises. Our results show that ownership structure matters: non-state-owned enterprises exhibited a much stronger increase in EO following performance shortfalls than state-owned enterprises. This likely reflects differences in incentives: non-state-owned firms face stronger profit pressures and competition, prompting bold action when goals are missed, whereas state-owned firms often have broader objectives and some degree of support or protection, reducing their urgency to innovate in response to shortfalls. Thus, the entrepreneurial drive triggered by internal shortfall is somewhat muted in state-owned enterprises. We also find that digital transformation plays a role: firms with greater digital maturity respond more entrepreneurially to shortfalls. Digital capabilities likely enable faster pivots and opportunity-seeking in the face of setbacks, whereas firms low in digital maturity lack the agility to mount such responses. These contextual findings underscore that the link between internal crisis and EO is not uniform across all organizations. It depends on the firm’s environment and resources. By identifying ownership type and digital capability as key moderators, our study reinforces a contingency view in entrepreneurship theory: the same trigger (performance shortfall) can lead to divergent outcomes depending on organizational context.

In sum, our research contributes to a deeper theoretical understanding of entrepreneurial orientation by integrating internal performance feedback into its antecedents and revealing how cognitive and contextual factors shape entrepreneurial responses. First, we establish internal performance shortfalls as a significant driver of EO, expanding the entrepreneurial orientation discourse beyond external environmental factors. Second, we clarify why firms may respond differently to similar performance shortfalls: managerial overconfidence and myopia can restrain the adaptive, opportunity-seeking behaviors that performance shortfalls would otherwise stimulate, while factors like non-state-owned ownership and digital readiness strengthen those behaviors. This helps explain inconsistencies in prior findings and demonstrates the value of incorporating cognitive and contextual lenses into behavioral theories of the firm.

Overall, this study enriches the conversation on why and when firms embrace an EO, showing that it is a dynamic strategic posture shaped not only by external shocks but also by internal triggers, tempered by who leads the firm and what resources the firm can leverage.

7.2. Policy and Managerial Recommendations

Our findings yield several actionable insights for managers and policymakers seeking to sustain competitive advantage amid performance shortfalls. First, for corporate executives and board members, the results highlight the importance of treating internal performance shortfalls as a catalyst for entrepreneurial action rather than a trigger for panic or retrenchment. When a firm underperforms relative to its aspirations, managers should resist any instinct to hide the problem or resort solely to conservative cutbacks. Instead, they can frame the shortfall as an opportunity to search for new strategies—for example, by launching internal reviews or task forces to generate innovative solutions. This proactive stance is more likely to yield turnaround ideas aligned with an EO, consistent with our evidence that moderate internal setbacks spur innovation.

Crucially, executives must also recognize and mitigate cognitive biases that our study found can dampen the positive effects of adversity. Managerial overconfidence can lead leaders to dismiss negative feedback or cling to the status quo, while managerial myopia may push them to seek quick fixes (such as cost cuts or one-off gains) at the expense of longer-term innovation. To counter overconfidence, companies could implement checks and balances in decision-making—such as encouraging open debates in leadership teams or bringing in independent advisors who can provide critical perspectives when performance flags. Firms may establish a series of processes to challenge entrenched strategies, ensuring that early warning signs are acknowledged rather than ignored. To combat short-termism, firms should align incentives and evaluation systems with long-run performance. This might include tying a portion of executive compensation to multi-year innovation milestones or future-oriented metrics, such as successful new product launches or R&D outcomes, rather than focusing solely on the next quarter’s profits. Such incentive structures encourage managers to pursue bold, forward-looking projects even when immediate results are underwhelming. By cultivating an organizational culture that values learning from failure and long-term vision, top leaders can turn a performance shortfall into a springboard for entrepreneurial renewal.

Additionally, our study underlines the role of organizational capabilities—notably, digital transformation—in enabling an effective entrepreneurial response to adversity. Firms with greater digital maturity, such as advanced IT systems, data analytics capabilities, and digitally skilled talent, were significantly more likely to respond to performance shortfalls with increased EO. For practitioners, this implies that investing in digital tools and infrastructure is not just a technological upgrade but a strategic buffer against turbulence. A digitally agile firm can more rapidly diagnose the causes of a performance decline and can more flexibly pivot operations or business models in response. Managers should therefore champion digital transformation initiatives in their organizations before a crisis hits—such as implementing data-driven decision platforms, fostering a workforce adept at digital technologies, and establishing agile processes—so that when internal challenges arise, the firm is equipped to innovate its way out of trouble.

From a policy and regulatory perspective, our findings offer guidance to those aiming to stimulate corporate entrepreneurship in China’s listed firms. One notable insight is the difference we observed between state-owned enterprises and non-state-owned enterprises: state-owned enterprises showed a weaker entrepreneurial response to performance shortfalls. Policymakers can address this gap by reforming incentive structures and governance in the state sector to mimic competitive, innovation-driven dynamics. For example, regulators or state owners could revise performance evaluation criteria for state-owned enterprises executives to place greater weight on innovation outcomes and adaptive responses to performance issues. Rewarding managers for entrepreneurial behaviors and holding them accountable when complacent during downturns would send a strong signal that adaptability and innovation are expected, even for firms with government ownership.

Finally, policymakers should continue promoting digital transformation at the national and industry levels. Targeted support policies, such as subsidies or tax incentives for investments in digital infrastructure and employee training, can lower adoption barriers and enhance firms’ agility in responding to performance issues.

7.3. Limitations and Future Directions

Like any other research, this paper has some limitations that could serve as viable directions for future studies. First, this study acknowledges the limitations of using secondary data, such as financial data from companies, to measure EO, as EO includes cognitive factors to some extent that secondary data can not fully capture. And using financial proxies to measure EO offers objective, longitudinal metrics, these indicators have inherent limitations. Financial proxies indirectly reflect entrepreneurial behaviors, potentially raising construct validity concerns. Similarly, earnings reinvestment as a measure of proactiveness might be influenced by external factors such as profitability constraints or dividend policies. Additionally, ROA volatility used for risk-taking could represent market conditions or management inefficiency rather than deliberate risk-taking. Therefore, future research could explore other methodologies to address current measurement issues related to EO, aiming for more accurate and comprehensive identification of EO, such as experimental methods. In addition, consistent with mainstream literature, this paper excluded sample firms in the financial sector and ST * listed companies. However, this does not imply that these firms do not engage in entrepreneurial activities. Thus, future research could expand the sample to enhance the generalizability of the findings.

Second, in this study, the explanation regarding overconfidence still cannot clearly attribute the research findings to management’s overestimation of capabilities rather than to overestimation of other factors affecting company performance. Future attempts to directly explore this issue would be an interesting direction, such as adopting survey-based methods to collect primary data, which would allow researchers to measure both overconfidence and personality optimism simultaneously, thus attempting to unravel these two closely related phenomena. This study focused only on performance-based aspiration levels. However, companies typically pay attention not only to financial goals but also to other objectives, such as company growth and innovation. Decision-makers in organizations are not always satisfied with merely avoiding bankruptcy or achieving performance levels determined by average industry performance; rather, they strive to achieve more ambitious goals, such as attaining industry leadership. Expanding the model and empirical research to explore how these alternative objectives influence EO and interact with cognitive biases and other organizational dilemmas would be highly meaningful.

Third, successful companies demonstrate the ability to learn and adapt during their development, but they may overlook environments that could lead to low commitment. Future research should further consider the impact of the surrounding environment to clarify the relationships among entrepreneurial behavior, performance outcomes, and the dominant logic within firms. In explaining organizational responses to performance feedback, this paper primarily focuses on the problem-search mechanism. However, existing studies have identified several other theoretical mechanisms, such as loss aversion, threat rigidity, and adaptive aspiration. Through these mechanisms, performance feedback can be translated into organizational actions. Therefore, future research could investigate different theoretical mechanisms and specific causal chains to explain how organizations respond to performance feedback.