2.1 Stock Overvaluation and Corporate Investment

This study pertains to two primary research areas. First, it is rooted in the same motivations that underlie studies examining the relationship between stock overvaluation and corporate investment decisions. The Q theory of investment states that firms increase investment when Tobin’s Q is high, assuming that market valuations accurately reflect future growth opportunities (Tobin, 1969). Although managers frequently have superior internal information, they may also incorporate market signals to complement their private assessments when making investment decisions. This tendency becomes stronger when external signals are deemed informative (Bond et al., 2012; Kau et al., 2008; Lee et al., 2023).

Behavioral finance research challenges the notion that Tobin’s Q reliably captures fundamentals. A growing body of research has shown that asset prices are influenced not only by fundamentals but also by sentiment, cognitive biases, and interpretive noise (Blanchard et al., 1993; Boulton et al., 2020; Campello and Graham, 2013). These distortions may cause stock prices to deviate from their fundamental values, leading to periods of misvaluation. Consequently, investment decisions may deviate from their optimal levels, particularly when Tobin’s Q reflects mispricing rather than genuine investment opportunities (Dong et al., 2012; Polk and Sapienza, 2009). Empirical studies show that stock overvaluation frequently leads to excessive investment. For example, Gilchrist et al. (2005), Mrad et al. (2024), and Polk and Sapienza (2009) demonstrate that firms increase capital expenditures during periods of overvaluation, often choosing to expand despite weak fundamentals. These findings imply that managers may respond to inflated market valuations, even when such valuations are unsupported by underlying performance.

Recent research has broadened this line of inquiry to include R&D investment. R&D is a forward-looking activity with uncertain outcomes and high information asymmetry, making it difficult to assess in real time (Aboody and Lev, 2000; Hall, 2002; Hottenrott et al., 2016). As a result, R&D is especially sensitive to managerial interpretation and external influence. Managers may interpret high stock prices as reflecting strong market expectations, even in the absence of changes in internal fundamentals. For example, Dong et al. (2021) and Shen et al. (2021) discover that overvalued firms allocate more resources to R&D, interpreting investor sentiment as a demand for innovation. Similarly, Li (2026) demonstrates that market optimism correlates with increased R&D investment and enhanced innovation outcomes. According to these findings, stock overvaluation can affect both tangible capital investment and intangible, innovation-driven initiatives.

2.2 CEO age and R&D Investment

Second, this study builds on a significant stream of analysis on CEO characteristics and their impact on innovation-related decisions. Existing research presents two competing perspectives on how CEO age influences innovation behavior. One view suggests that younger CEOs are more likely to take bold or unconventional actions to demonstrate their competence to the labor market (Prendergast and Stole, 1996). Because their career horizon is longer, these CEOs may perceive risky investments, such as mergers and acquisitions (M&A) or R&D, as opportunities to develop their reputation early on (Gao, 2010; Li et al., 2017; Yim, 2013; Yim and Kang, 2024; Zou et al., 2024). This signaling motive may increase their responsiveness to market sentiment and encourage them to meet investor expectations when the firm is overvalued.

Conversely, younger CEOs may be more cautious because their future compensation and career prospects are more sensitive to early performance outcomes (Holmström, 1999). Thus, they may avoid uncertain investments that could jeopardize short-term results, instead prioritizing performance stability to mitigate the risk of early dismissal or reputational damage (Bliss and Rosen, 2001). With limited reputational capital, they encounter greater penalties for failure and are thus less willing to engage in uncertain or unconventional investments (Hirshleifer and Thakor, 1992; Scharfstein and Stein, 1990). Rather than pursuing innovative projects that may be difficult to assess externally, they may prefer safer investments that are easier to justify (Zwiebel, 1995).

These behavioral differences align with the upper echelons theory, which postulates that top executives’ attributes, such as age, experience, and mindset, influence how they make strategic decisions and interpret external signals (Hambrick and Mason, 1984). As older CEOs approach retirement, they are more likely to exhibit risk-averse behavior and shorter planning horizons, making them less willing to take on long-term projects such as R&D. In contrast, younger CEOs may perceive innovation as career-enhancing, especially in contexts where investor expectations are heightened (Barker and Mueller, 2002).

Lifespan developmental psychology provides a complementary explanation for age-related differences in innovation behavior. According to research in this tradition, future time perspective and motivational priorities shift with age: younger individuals emphasize exploration and growth, whereas older individuals increasingly value stability and loss avoidance (Baltes, 1987). Socioemotional selectivity theory supports this view by demonstrating that younger adults prioritize information seeking and long-term rewards, whereas older adults focus on maintaining current performance and limiting downside exposure (Carstensen, 1995). In managerial contexts, these patterns imply that younger CEOs, who have broader future horizons and stronger growth-oriented motivations, may regard rising stock prices as an incentive to undertake ambitious activities such as R&D. Older CEOs, whose goals tend to emphasize stability and risk containment, may instead respond to the same signals more conservatively (Kooij et al., 2011).

These contrasting incentives imply that CEOs may interpret stock price signals in systematically distinct ways depending on their traits (Gao, 2010). When stock prices deviate from firm fundamentals, some CEOs may view this as a transient mispricing, whereas others may perceive it as important information about investor expectations. This variation partly stems from differences in managerial horizon and behavioral disposition, which are frequently associated with observable characteristics such as age (Bantel and Jackson, 1989). Thus, CEO age provides a tractable and interpretable proxy for examining how managerial traits influence firm responses to stock overvaluation. Because R&D project outcomes are uncertain and difficult to assess in the short term, CEOs must rely more heavily on subjective interpretation, which creates greater scope for heterogeneity in how market signals are integrated into R&D decisions.

2.3 Differences from Prior Research

Recent studies have examined the relationship between CEO age and corporate innovation, typically focusing on age as a predictor of firms' innovation outcomes. For example, Yim and Kang (2024) show that younger CEOs in Korean IT firms exhibit higher innovation productivity and a broader technological scope, demonstrating a static association in which CEO age directly influences innovation levels. Although such research advances our understanding of demographic patterns in innovation, it does not address how CEOs of various ages adjust their innovation decisions amidst shifting market conditions.

In contrast, this study examines a dynamic mechanism in which CEO age moderates the responsiveness of R&D investment to stock overvaluation. This approach fundamentally departs from the traditional risk-preference perspective, which interprets age as reflecting inherent risk-taking tendencies and thus predicts differences in baseline innovation levels rather than in responses to valuation shocks. This study instead relies on a market-accommodating interpretation, in which valuation deviations signal investor expectations and managers adjust real decisions accordingly. According to this framework, CEO age influences how strongly managers interpret and react to such valuation signals, rather than acting as a direct driver of innovation intensity. This distinction positions our study within the broader literature on misvaluation-driven investment behavior and highlights interpretive heterogeneity as the primary channel that distinguishes our dynamic approach from static age-innovation analyses.

2.4 Hypothesis Development

Stock overvaluation can influence how CEOs evaluate R&D investment. The key question is not whether younger CEOs unconditionally prefer R&D investment. R&D is a long-horizon investment with uncertain outcomes, significant information asymmetry, and limited short-term verifiability (Aboody and Lev, 2000; Hall, 2002; Hottenrott et al., 2016). These features may render R&D sensitive to managerial interpretations and external market signals (Lee et al., 2023).

The baseline relation between younger CEOs and R&D investment is theoretically ambiguous. Younger CEOs may value R&D because it promotes long-term growth and demonstrates their willingness to pursue innovation. Their longer managerial horizons may make long-horizon projects more attractive. Prior studies also suggest that younger CEOs are more likely to take risks and engage in innovative activities (Li et al., 2017; Prendergast and Stole, 1996; Yim and Kang, 2024; Yim, 2013; Zou et al., 2024). This viewpoint is consistent with lifespan developmental research, which indicates that younger individuals prioritize exploration and long-term rewards (Baltes, 1987; Carstensen, 1995; Kooij et al., 2011). Simultaneously, younger CEOs may avoid R&D because it is costly, uncertain, and difficult to justify in the short term. Early performance outcomes may influence their future compensation and career prospects. Limited reputational capital can also increase the perceived cost of failed investment decisions (Bliss and Rosen, 2001; Hirshleifer and Thakor, 1992; Holmström, 1999; Scharfstein and Stein, 1990; Zwiebel, 1995). Thus, in the absence of valuation pressure, younger CEOs do not necessarily invest more or less in R&D.

Stock overvaluation may render R&D investment more appealing to younger CEOs.

One possible channel is equity financing. Overvaluation reduces the relative cost of issuing equity and enables firms to raise external capital on favorable terms (Baker et al., 2003; Dong et al., 2012; Kim, 2021; Stein, 1996). This financing advantage is significant for R&D, which is difficult to collateralize and frequently encounters external financing frictions. When younger CEOs’ firms are overvalued, they may find it easier to finance long-horizon R&D projects that funding frictions would otherwise limit.

Catering is another possible channel. Overvalued stock prices may reflect heightened investor expectations for future growth. R&D provides a visible way for managers to respond to such expectations by demonstrating their commitment to future technological development. Prior studies demonstrate that overvalued firms tend to increase R&D investment, implying that managers may interpret investor optimism as a demand for innovation (Dong et al., 2021; Shen et al., 2021; Li, 2026). Younger CEOs may be especially responsive to this signal because their longer managerial horizons render visible growth-oriented investment more appealing (Gao, 2010). According to this viewpoint, stock overvaluation enhances the incentive for younger CEOs to expand R&D investment because R&D helps firms align their actions with market expectations.

Although the underlying logic of these channels differs, both imply that younger CEOs may increase R&D investment more aggressively when their firms are overvalued. This reasoning yields the following hypothesis.

Hypothesis 1: Younger CEOs are more likely to increase R&D investment in response to stock overvaluation.

The opposite prediction is also plausible. The same factors that render R&D appealing when overvalued can also make it risky (Luo et al., 2022). If managers believe that high valuations are only temporary, they may become more cautious because overvaluation creates unrealistic expectations (Jensen, 2005). The same characteristics that make R&D sensitive to market signals also make it difficult to defend against overvaluation because R&D outcomes are opaque and delayed (Aboody and Lev, 2000; Hall, 2002). If market sentiment later reverses, delayed and hard-to-verify R&D outcomes may expose managers to criticism for expanding uncertain, long-horizon investments.

This risk-aversion logic may be stronger among younger CEOs. Younger CEOs may have longer managerial horizons, but they may also lack established reputational capital. Their future compensation and career prospects may be more dependent on early performance outcomes (Bliss and Rosen, 2001; Holmström, 1999). In addition, managers with limited reputational capital may avoid projects that are difficult to justify and have uncertain outcomes, as failed projects can be interpreted as poor judgment (Hirshleifer and Thakor, 1992; Zwiebel, 1995). According to this view, overvaluation does not make R&D more appealing to younger CEOs. Instead, it heightens the downside risk of expanding into uncertain, long-horizon investments. Younger CEOs may thus respond to stock overvaluation by reducing R&D investment or demonstrating a weaker R&D response.

Hypothesis 2: Younger CEOs are more likely to exhibit a weaker or negative R&D response to stock overvaluation.

3.1 Sample

The sample comprises non-financial firms listed on the Korea Exchange (KRX) between 2012 and 2022. Financial and insurance firms (two-digit Korean Standard Industrial Classification (KSIC) codes 64-66) and utility firms (two-digit KSIC codes 35-39) are excluded because their accounting structures and regulatory environments vary from those of other listed firms.

The study begins by creating a panel dataset to estimate stock overvaluation. The initial sample includes all firms listed on the KRX from 2012 to 2022, with fiscal years ending in December, and uses Data Guide 5.0 to estimate stock overvaluation. Observations are excluded if they lack the market value of equity or have a book value of equity of zero or negative. This filtering yields an unbalanced panel of 22,289 firm-year observations. To guarantee reliable estimation, any industry-year (based on two-digit KSIC codes) with fewer than ten firms is excluded (e.g., Kim, 2021; Kim, 2024). Subsequently, 21,142 firm-year observations remain for estimating stock overvaluation.

The panel dataset generated for estimating stock overvaluation is then compared to the firm-level data used in the empirical analysis, which includes CEO age, R&D expenditures, and other control variables. CEO age data are gathered from TS-2000, while R&D expenditure and firm-level control variable data are obtained from Data Guide 5.0. After merging these datasets, the empirical analysis’s final sample comprises 17,551 firm-year observations.

3.2 Variables

3.3 Descriptive Statistics

Table 2

Regarding firm characteristics, the mean values of firm size (FirmSIZE) and firm age (FirmAGE) are 19.15 and 3.18, respectively. For the other variables, the average tangibility (TANG) is 29.78%, and the average debt ratio (LEV) is 42.50% of total assets. Furthermore, a sample firm has, on average, a return on assets (ROA) of 3.00% and an annual sales growth rate (GROWTH) of 13.22%. Additionally, foreign ownership (FOR) averages 6.35%, and the ownership share held by the largest shareholder and affiliated parties (LARGE) is 39.60%, indicating a relatively high level of ownership concentration among sample firms. The mean value of BETA is 0.8527.

This table shows the mean, standard deviation (SD), and selected percentiles (1st, 25th, 50th, 75th, and 99th) for the variables. All continuous variables are winsorized at the 1st and 99th percentiles. Table A1 provides the variable definitions.

This table presents the Pearson correlation coefficients for the main variables in the analysis. All continuous variables are winsorized at the 1st and 99th percentiles. Table A1 provides the variable definitions. ** indicates statistical significance at the 5% level.

Table 3

4.1 Univariate Results

Table 4

This table presents the univariate and pairwise comparisons of R&D intensity. Panels A and B compare the average and median R&D intensity between firms led by younger and older CEOs, as well as between overvalued and undervalued firms. Panels C and D show the pairwise comparisons of average and median R&D intensity across four subsamples, characterized by the combination of CEO age (younger vs. older) and valuation status (overvalued vs. undervalued). The differences in means (medians) are evaluated using t-tests (Wilcoxon rank-sum tests). ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

Panel B reports the median comparisons and demonstrates a similar trend. The median RDI is significantly higher for firms with younger CEOs (0.60%) than for those with older CEOs (0.39%). In addition, overvalued firms (0.85%) exhibit higher median RDI than undervalued firms (0.28%).

Panels C and D present pairwise comparisons of both CEO age and stock misvaluation status (i.e., (a) Younger CEO and Overvalued, (b) Older CEO and Overvalued, (c) Younger CEO and Undervalued, and (d) Older CEO and Undervalued). In Panel C, the average RDI is 4.14% for overvalued firms and 2.10% for undervalued firms among firms with younger CEOs, yielding a difference of 2.03 percentage points. For older CEOs, the average RDI is 2.40% for overvalued firms and 1.66% for undervalued firms, with a smaller difference of 0.70 percentage points. Regarding the median values in Panel D, the median RDI is 0.98% for overvalued firms and 0.32% for undervalued firms among firms led by younger CEOs. The difference of 0.68 percentage points is statistically significant at the 1% level. In comparison, firms led by older CEOs exhibit a smaller and statistically insignificant difference, with median RDIs of 0.38% and 0.27% in overvalued and undervalued firms, respectively. Overall, these results indicate that the positive association between stock overvaluation and R&D investment is more pronounced when younger CEOs lead firms.

4.2 Baseline Regressions

To test the hypotheses, this study estimates the following regression model:

where i and t denote firm and year, respectively. The dependent variable is R&D investment, proxied by R&D intensity in the subsequent year. The explanatory variables include a dummy for younger CEOs (YoungerCEO), an indicator for stock overvaluation (Overvalued), and their interaction term (YoungerCEO × Overvalued). X denotes a vector of control variables. Firm and year fixed effects (FIRM and YR) control for unobserved heterogeneity across firms and years. To mitigate potential simultaneity concerns, the explanatory variables are measured in year t, whereas the dependent variable is measured in year t+1. Standard errors are clustered at the firm level.

The main coefficient of interest is $\beta_{YO}$, which captures whether younger CEOs respond differently to stock overvaluation in their R&D investment decisions. A positive (negative) $\beta_{YO}$ indicates that younger CEOs increase (decrease) R&D investment more (less) than their older counterparts when their firms are overvalued. Accordingly, hypothesis 1 predicts $\beta_{YO} > 0$, whereas hypothesis 2 predicts $\beta_{YO} < 0$.

This table presents the baseline regression results. The dependent variable is RDIt+1. YoungerCEO is a binary variable that equals 1 if the CEO is younger than 60 years and 0 otherwise. Overvalued is a binary variable that equals 1 for firms in the top tercile of the annual distribution of firm-specific misvaluation, and 0 otherwise. Table A1 provides the variable definitions. The t-statistics based on clustered standard errors at the firm level are presented in parentheses. ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

Table 5

The economic magnitude of this effect is also substantial. In column (2), the coefficient on Overvalued is 0.6227, implying that firms led by older CEOs increase RDI by approximately 0.62 percentage points when their stocks are overvalued. For firms led by younger CEOs, the response to overvaluation is 0.69 percentage points higher. Collectively, this implies that overvaluation correlates with an increase of approximately 1.31 percentage points in RDI for firms led by younger CEOs. Considering that the sample mean of RDI is approximately 2.7%, this represents a sizeable increase in innovation investment.

The coefficient on YoungerCEO itself is statistically insignificant across specifications. When stock overvaluation is considered, this result indicates that younger CEOs do not systematically invest more in R&D than older CEOs. This finding varies from prior studies that document a stronger propensity for innovation among younger CEOs (e.g., Yim, 2019; Yim and Kang, 2024; Zou et al., 2024). One possible interpretation is that the positive relationship between younger CEOs and innovation observed in previous studies may partly reflect their stronger response to periods of stock overvaluation. Consistent with this interpretation, once stock overvaluation is explicitly considered, the standalone effect of younger CEOs on R&D investment is no longer statistically distinguishable from zero. Instead, the findings indicate that younger CEOs’ higher R&D investment is primarily due to a stronger response to stock overvaluation.

Although the risk-aversion perspective hypothesizes that CEOs may reduce R&D investment when firms are overvalued, our results provide minimal support for this hypothesis. One possible explanation is that younger CEOs often experience stronger career concerns and incentives to demonstrate growth-oriented performance (Li et al., 2017; Yim, 2019; Yim, 2013). In the Korean corporate context, where technological competitiveness and innovation are extensively regarded as key drivers of long-term firm value, managers may encounter strong pressure to pursue visible investment activities such as R&D rather than adopt a conservative investment stance (Yim, 2019). In addition, periods of stock overvaluation can reduce the perceived downside risk associated with investment decisions by acting as a valuation buffer (Polk and Sapienza, 2009; Stein, 1996). Consequently, the risk-aversion viewpoint may be less likely to dominate managerial investment decisions in this context.

4.3 Endogeneity

Endogeneity is a potential source of concern for the baseline regression results, as the observed positive association between stock overvaluation and R&D investment may partly reflect reverse causality. For example, if investors expect a firm to increase its R&D investment, the stock market may factor this expectation into the firm’s stock price, leading to higher valuations (Dong et al., 2021; Li, 2026; Luo et al., 2022). In this case, causality would be from expected R&D investment to stock overvaluation, rather than the other way around.

To account for the possibility of reverse causality, this study estimates a dynamic panel model using the two-step system GMM (Arellano and Bover, 1995; Roodman, 2009). This method is well-suited for contexts where the dependent variable, RDI, is potentially endogenous and influenced by its past values. System GMM can help control for unobserved firm-specific effects and endogeneity in explanatory variables, including the interaction term, by incorporating the dependent variable’s lagged values as regressors and employing appropriate internal instruments. This approach enables us to obtain consistent estimates even when Overvalued, alongside its interaction with YoungerCEO, is potentially endogenous.

Table 6

Another potential concern is that firms led by younger CEOs and those that are overvalued may systematically differ from their counterparts in observable firm characteristics. If these differences are related to CEO age or stock overvaluation and firms’ R&D investment decisions, the baseline regression results may be influenced by selection bias. For instance, prior studies indicate that firms led by younger CEOs frequently differ from those led by older CEOs in several observable characteristics (Yim and Kang, 2024). Similarly, overvalued firms may exhibit distinct characteristics, such as size, profitability, and growth opportunities, which have been shown to influence corporate investment behavior.

To address this concern, this study employs PSM to generate more comparable samples. Separate matching procedures are used to match firms led by older CEOs with firms led by younger CEOs, as well as overvalued firms with non-overvalued firms. Because the number of firms with younger CEOs exceeds that of firms with older CEOs, firms led by older CEOs are regarded as the treatment group to ensure adequate matching quality. The propensity scores are estimated using the entire set of control variables included in the baseline regression model. To pair treated firms with observationally similar control firms, I perform one-to-one nearest-neighbor matching without replacement, applying a common-support restriction.

Before matching, significant differences in several covariates are observed between the treated and control groups. These differences are detailed in Table A3 of the Appendix. Following matching, covariate imbalances are significantly reduced. Figure 1

This study uses the matched samples to re-estimate the baseline regression model. Table 7

4.4 Robustness Checks

To ensure that variable definitions or specific modeling choices do not influence the baseline findings, this study performs several additional robustness checks.

Table 8

I also substitute the overvaluation indicator with the continuous FSE, as shown in column (3) of Panel B. The baseline results may be influenced by the tercile-based classification used to identify overvalued firms. Because the indicator variable considers only firms in the top tercile of the misvaluation distribution, it may overlook differences in the degree of stock misvaluation across firms. To address this concern, I re-estimate the model using the FSE measure and its interaction with YoungerCEO. The interaction term between YoungerCEO and FSE remains positive and statistically significant, thus indicating that the results persist when stock misvaluation is measured using a continuous specification.

Panel C presents robustness checks using alternative model specifications. First, younger CEOs may systematically differ from older CEOs in unobserved managerial characteristics, such as a stronger natural proclivity for innovation (Bostan and Mian, 2019; Yim, 2019; Yim and Kang, 2024). The estimated interaction between YoungerCEO and Overvalued may partly capture CEO-specific heterogeneity rather than a differential response to stock overvaluation. To address this concern, I incorporate CEO fixed effects (e.g., James et al., 2020; Yim and Kang, 2024), which absorb time-invariant CEO characteristics. Due to some CEOs in our dataset serving as CEOs in multiple firms within the same year, the CEO-year combination is not unique, hence precluding the use of a conventional panel estimator based on a unique CEO-year identifier. Consequently, I estimate the model using a high-dimensional fixed-effects estimator that aligns with the data structure (Correia, 2016; Gaure, 2013; Guimarães and Portugal, 2010). As reported in column (1), the coefficient on YoungerCEO × Overvalued remains positive and statistically significant, implying that unobserved CEO-specific traits do not influence the observed relationship.

Second, time-varying industry conditions may simultaneously impact both stock valuations and firms’ R&D investment. For example, technological opportunities, industry-wide innovation cycles, or sector-specific demand shocks may result in higher stock prices and increased R&D spending within the same industry-year (Agrawal et al., 2020; Videnord, 2025). If such industry-level dynamics influence both valuation and investment, the baseline results may reflect industry-specific shocks rather than firms’ differential responses to stock overvaluation. To address this concern, I incorporate industry-by-year fixed effects, which absorb all time-varying shocks common to firms within the same industry-year. As reported in column (2), the coefficient on YoungerCEO × Overvalued remains positive and statistically significant, implying that time-varying industry conditions do not account for the results.

4.5 Underlying Mechanisms

The baseline results indicate that younger CEOs significantly increase R&D investment when firms are overvalued. This finding supports Hypothesis 1, which states that younger CEOs exhibit a more pronounced response to stock overvaluation by increasing R&D investment. However, this result alone does not reveal the underlying mechanism. The positive association between stock overvaluation and R&D investment in firms led by younger CEOs is not unique to the catering explanation (Alzahrani and Rao, 2014; Kusnadi and Wei, 2017). The catering hypothesis predicts that younger CEOs increase R&D investment to meet investor expectations during periods of overvaluation, and a similar pattern may also emerge under the equity financing mechanism (e.g., Baker et al., 2003; Morck et al., 1990; Stein, 1996).

If stock overvaluation reduces the cost of equity, firms may be able to raise equity capital more easily and increase R&D investment (e.g., Dong et al., 2021; Li, 2026; Shen et al., 2021), and younger CEOs may exploit this opportunity more aggressively. Therefore, the positive interaction observed in the baseline regressions may indicate either the catering or the equity financing mechanism. To distinguish between these competing explanations, I perform a series of mechanism tests.

5.1 Implications

These findings indicate that the real effects of stock misvaluation depend on CEOs’ interpretations of market signals. Although prior research frequently assumes that managers respond similarly to stock overvaluation (e.g., Dong et al., 2021; Li, 2026; Luo et al., 2022), the results indicate that such responses vary systematically with CEO characteristics. Particularly, younger CEOs are more likely to increase R&D investment when their firms’ stock prices exceed fundamental values. This perspective also provides a new interpretation of prior evidence linking CEO age to innovation outcomes. Rather than reflecting a purely intrinsic managerial preference for innovation, the results indicate that part of younger CEOs’ innovation behavior may stem from a stronger response to market valuation signals. Collectively, the findings highlight the significance of managerial interpretation in comprehending how stock market signals influence corporate innovation decisions.

Regarding practical implications, the findings propose that boards and investors should focus more on executive characteristics when evaluating corporate responses to market signals. The results indicate that younger CEOs are more inclined to increase R&D investment when their firms’ stock prices are overvalued, implying that managerial traits influence how firms interpret and respond to valuation signals. When temporary market conditions influence decisions, this responsiveness in innovation-intensive industries can either enhance long-term value creation or result in excessive investment.

The findings have important implications for innovation policy and R&D support programs. If R&D investment is partially influenced by stock market signals rather than solely by technological opportunities, policy frameworks that significantly rely on quantitative input measures, such as R&D spending levels, may not adequately reflect the underlying motivations for firms’ R&D investment. Incorporating qualitative assessment criteria, such as project continuity, technological coherence, and alignment with long-term development goals, may aid in the effectiveness of public R&D evaluation systems.

These implications are particularly significant in contexts such as Korea, where firms exhibit exceptionally high R&D intensity but uneven conversion of R&D inputs into innovation outcomes (Choi et al., 2024). In such contexts, strong performance pressure and relatively short CEO tenures may amplify managerial sensitivity to stock market signals. Recognizing how managerial characteristics interact with valuation signals may aid policymakers in designing governance mechanisms that mitigate valuation-driven distortions while promoting long-term innovation.

5.2 Limitations and Future Research

Despite its contributions, this study has several limitations that provide avenues for future research. First, CEO age represents only one dimension of managerial interpretation. Future research could examine other managerial characteristics that influence how executives respond to market signals, such as CEO tenure, compensation structure, career concerns, and individual risk preferences. Second, although this study indicates that younger CEOs respond more strongly to stock overvaluation by increasing R&D investment, the underlying reasons for this behavior remain unanswered. Younger CEOs may experience stronger career concerns, greater reputational incentives, or stronger pressure to meet market expectations. Future research could further explore these mechanisms by examining how managerial incentives and governance structures interact with stock overvaluation.

Third, although this study focuses on R&D investment decisions, it raises an important question concerning the efficiency of valuation-driven innovation. If R&D spending increases primarily in response to stock overvaluation, such investment may not always result in enhanced innovation outcomes. Future research could investigate whether valuation-induced R&D investment improves innovation efficiency or causes overinvestment in R&D. Fourth, although I employ a firm-specific measure of stock overvaluation, other approaches may provide complementary insights. For example, alternative approaches to measuring misvaluation include the residual income model or mutual fund flow measures (Bonaimé et al., 2014; Dong et al., 2021; Jin, 2022; Luo et al., 2022). However, the availability of detailed mutual fund flow data is limited in the Korean market, making this approach ineffective.

Lastly, because the analysis is based on Korean firms, caution should be exercised when generalizing the findings to other institutional contexts. Future research could benefit from applying this framework to firms operating in different institutional settings to gain a deeper comprehension of how governance structures and capital market conditions influence valuation-driven innovation behavior.