International Review of Financial Analysis 91 (2024) 103020

Contents lists available at ScienceDirect

International Review of Financial Analysis

journal homepage: www.elsevier.com/locate/irfa

Firm-level political risk and intellectual capital investment: Does

managerial ability matter?
Nhan Huynh a, *, Quynh Nga Le b, Quang Thien Tran c
a
Macquarie Business School, Macquarie University, Sydney, Australia
b
Macquarie Group Limited, Sydney, Australia
c
Faculty of Public Relations and Communication, Van Lang University, Ho Chi Minh City, Viet Nam

A R T I C L E I N F O A B S T R A C T

JEL classifications: This paper examines the impacts of firm-level political risks on intellectual capital investment decisions and how
G18 managerial ability adjusts this relationship. Using a broad sample of U.S firms from 2002 to 2021, our results
G30 show that firms with higher political risks reduce their investment in intellectual capital. This impact is more
G31
prominent for high-tech firms and firms with high financial distress, external financial dependence, and lower
G38
institutional ownership. Further, we find supportive evidence that managerial ability can prevent a substantial
Keywords:
dimension (around 20%–40%) of the destructive impact of political risk on intellectual capital investment, which
Firm-level political risk
Intellectual capital
is also driven by firm-specific characteristics. These findings hold after a battery of robustness tests.
Managerial ability
Financial constraint

1. Introduction uncertainty on managerial decisions and financial outcomes, there is a

handful of papers that consider the impacts of political risks and un­
In the modern economy and integrated financial market, intellectual certainty on corporate investment (Azzimonti, 2018; Chen, Cihan, Jens,
capital (also described as intellectual property) is progressively playing & Page, 2023; Gulen & Ion, 2016; Handley & Limao, 2015; Julio & Yook,
a crucial role in determining the effectiveness of businesses as well as the 2012). By mainly focusing on general capital investment, prior studies
overall economic, management, technological, and sociological de­ suggest that firms decrease their overall investment during heightened
velopments (Mouritsen & Larsen, 2005; Oliveira, Lima Rodrigues, & uncertainty with preventative interruptions in capital expenditure,
Craig, 2010). In recent years, the focus has altered from the capital- ascribed to investment irreversibility (Gulen & Ion, 2016). Therefore,
intensive to the information and knowledge-intensive industries, understanding how firms alter their investment policies on intellectual
which is accredited to the augmented importance of intellectual capital capital in responding to political uncertainty remains limited. Further,
for both practitioners and academics (Alvino, Di Vaio, Hassan, & Pal­ the impact of political uncertainty on total corporate investment has
ladino, 2021; Su, 2014). Given the strand of literature on the phenom­ been well documented at the aggregate levels. Prior studies utilize the
enon of knowledge management, the concepts of intellectual capital and aggregate measures of political uncertainty in terms of country-level
intangible assets are extensively linked to each other (Hussi, 2004; uncertainty indices (such as Economic Policy Uncertainty – EPU and
Osinski, Selig, Matos, & Roman, 2017).1 In other words, intellectual Geopolitical Risk Index (GRI) or major political events (such as Presi­
capital is the primary strength of innovative knowledge and constant dential Election periods). However, those aggregate measures could not
effectiveness of businesses (Inkinen, Kianto, Vanhala, & Ritala, 2017) capture the cross-sectional differences in firm-level exposure to political
and shareholders' practical concerns (Tan, Plowman, & Hancock, 2008). risk (Ahmad, Aziz, El-Khatib, & Kowalewski, 2023; Hassan, Hollander,
Considering the ongoing debates around the impacts of political Van Lent, & Tahoun, 2019). Although there is a wide body of literature

* Corresponding author at: Macquarie Business School, Macquarie University, Sydney, Australia.
E-mail address: [email protected] (N. Huynh).
1
In a broader form, intangible assets are the firms' assets with no physical form, which can include knowledge related to legal ownership like patents, trademarks,
copyrights, trade secrets, registered designs, goodwill, computer software, contracts, and databases (Marcelin, Stephen, Fanta, & Tecklezion, 2019). In this study, we
focus on the indicator of intellectual investment proposed by Sydler et al. (2014) rather than exhaustively identifying the differences between the two concepts.

https://doi.org/10.1016/j.irfa.2023.103020
Received 12 June 2023; Received in revised form 27 August 2023; Accepted 25 October 2023
Available online 28 October 2023
1057-5219/© 2023 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).
N. Huynh et al. International Review of Financial Analysis 91 (2024) 103020

on the drivers behind investment in intellectual capital, which is critical overcome the barriers of macro uncertainty measures observed in prior
to firms' long-term development, the relationship between the studies. Further, we also document that the relationship between firm-
firm-specific political risks and intellectual capital investment is a level political risk and intellectual capital investment is shaped by
relatively unexplored question. several firm characteristics. As such, this study can extend the current
Given the unique characteristics of intellectual capital, such as dif­ literature on how political uncertainty/risk influences corporate in­
ficulties in measuring, long payback period, and higher risks (D'Amato, vestment as well as corporate decisions (Azzimonti, 2018; Choi, Chung,
2021). A comprehensive insight into the connection between firm-level & Wang, 2022; Gulen & Ion, 2016; Gyimah, Danso, Adu-Ameyaw, &
political risks and intellectual capital and the factors involved in such a Boateng, 2022).
relationship is critical to a corporate strategy and has significant im­ Second, this paper differs from prior studies in that we focus on the
plications for policymakers and stakeholders in the financial markets. impacts of managerial ability on moderating the nexus between political
Hence, in this study, we seek to fill this gap in the extant literature by risk and intellectual capital investment. Given the extant literature, few
examining how idiosyncratic firm-level political risks affect firm in­ studies have utilized the Demerjian et al. (2012) measure to examine
vestment in intellectual capital. In addition, inspired by prior studies on whether managerial ability can moderate the adverse effects of un­
political uncertainty and managerial ability, we also explore how certainties. For instance, prior studies mainly focus on the impacts of
managerial ability shapes this relationship between firm-level political managerial ability on corporate performance and activities, such as
risk and intellectual investment. Previous studies have shown that earnings quality (Demerjian, Lev, Lewis, & McVay, 2013), financial
managerial ability is deemed a strategic intangible asset of organiza­ performance (Cheung, Naidu, Navissi, & Ranjeeni, 2017), or payout
tions, given its intense impact on structural effectiveness (Chemmanur, policies (Guan, Li, & Ma, 2018). Two studies of particular interest and
Paeglis, & Simonyan, 2009; Demerjian, Lev, & McVay, 2012). Firms close to our research are Phan, Tran, Nguyen, and Le (2020) and Kumar
with better managers can identify beneficial investment prospects, have and Zbib (2022), which consider the association between Demerjian
a superior capacity to estimate demand, choose quality projects, un­ et al. (2012) managerial ability measure and financial performance
derstand risk, and thus can maximize investment efficiency (Gan, 2019; during periods of oil price uncertainty and COVID-19 crisis, respec­
Lee, Wang, Chiu, & Tien, 2018). In other words, managerial ability is tively. Therefore, this study can enrich the current literature on the
appreciably crucial in seeking better investment opportunities during importance of managerial ability associated with performance, invest­
uncertain periods (Andreou, Karasamani, Louca, & Ehrlich, 2017; Has­ ment decisions, and survival of firms at times of uncertainty. From the
san et al., 2019; Kumar & Zbib, 2022). starting point of this study, several practical implications can be pro­
Using a broad sample of the 3688 U.S publicly listed firms covering posed for practitioners, shareholders, and policymaking in evaluating
the period from 2002 to 2021, we examine the effect of firm-level po­ capital investment opportunities, especially during the increasing un­
litical risks proposed by Hassan et al. (2019) on corporate intellectual certainty stages. The practical implications for firms are also proposed
capital investment and the moderated function of managerial capability by drawing their attention to the associations between intellectual
in this relationship. Our analyses produce the following main findings. capital, political uncertainties, and the power of managerial capability.
First, firm-level political risk, proxied by political exposure and risk, Not only highlighting the importance of the development of intellectual
negatively and significantly impact firms' intellectual capital investment capital, but our results also provide suggestions for shareholders in
in a subsequent year. Second, our results show that firms with higher considering top executives' professional experiences and expensing their
financial distress and external finance dependence exhibit a more sig­ compensation.
nificant reduction in intellectual capital investment. The extent of this The remainder of the paper is organized as follows. Section 2 sum­
negative impact also depends on firm-level governance, as it is less marizes related literature on political risks, investment, and managerial
significant for firms with higher institutional ownership. Third, utilizing ability. Section 3 describes the data, variable constructions, and baseline
the managerial ability scores developed by Demerjian et al. (2012), we methods. Section 4 presents the main results. The robustness and addi­
find that the negative impact of political risk on intellectual capital in­ tional tests are reported in Section 5, and Section 6 concludes the study.
vestment is negative and statistically significant regardless of the
managerial ability scores. However, the results also confirm that firms 2. Related literature
with higher managerial ability scores are less concerned about political
risk than those with lower scores. The differences in economic magni­ 2.1. Political risk, corporate investment policies, and intellectual capital
tude further reveal that a higher level of managerial ability enables firms investment
to lower a substantial proportion (around 20% to 40%) of the unfa­
vorable influence of firms' political risk on their intellectual capital in­ Political risk or uncertainty is described as risks generated from
vestment. In addition, the moderated impact of managerial ability also political events, such as wars, terrorism, inter-nation conflicts, or un­
depends on firms' levels of financial distress, external finance depen­ stable political environment (Caldara & Iacoviello, 2022; Vu, Huynh,
dence, institutional holding, and analyst coverage. Further, we also Phan, & Hoang, 2023). At the aggregate level, political risks negatively
verify that the impacts of political risk are less visible for non-high-tech impact economic development, employment, investment, and financial
firms as they have fewer intellectual assets to be concerned about. market stability (Acemoglu & Robinson, 2001; Bekaert, Harvey, Lund­
However, the power of managerial ability is not sensitive to blad, & Siegel, 2016; Huynh, Dao, & Nguyen, 2021). A rich strand of
technological-related factors. Our results remain unchanged under a studies considers the impacts of political risks on corporate activities and
battery of robustness checks using alternative variables and sensitivity performance at the industry and firm levels. The increase in political
analyses to control potential endogeneity concerns. uncertainty entails firms implementing more cautionary decisions by
Our study contributes to the extant literature in the following ways. increasing cash holding (Duong, Nguyen, Nguyen, & Rhee, 2020),
First, to the best of our knowledge, this is the first study to consider the reducing financing (Çolak, Durnev, & Qian, 2017; Dai & Ngo, 2021; Lee
impacts of firm-level political risk on corporate intellectual capital in­ et al., 2018), financial performance (Joshi, Cahill, Sidhu, & Kansal,
vestment, which has yet to draw much attention. Complementing prior 2013; Keillor, Wilkinson, & Owens, 2005), and corporate innovation
studies, our findings show that firm-level political risk has non-trivial (Ellis, Smith, & White, 2020). Also, considering the unfavorable condi­
impacts on intellectual capital investment above and beyond that of tion generated by political uncertainty, King, Loncan, and Khan (2021);
economywide political risk, as reported by Hoang and Tran (2022) and Le and Tran (2021); Alam, Houston, and Farjana (2023); Gulen and Ion
Chen et al. (2023). Using firm-level political risk constructed from the (2016); and Caldara and Iacoviello (2022) confirm that higher level of
textual search approach by Hassan et al. (2019), we can better capture political risk lowers corporate investment.
the idiosyncratic exposure to political ambiguity across firms and On the one hand, prior literature on the impacts of political risk on

2
N. Huynh et al. International Review of Financial Analysis 91 (2024) 103020

corporate investment is mainly drawn from aggregate levels of political 2.2. The roles of managerial ability
risk on major political events. One of the main shortfalls of this approach
is the assumption that macro-level political risks homogeneously affect In the word of the upper echelon theory, corporate executives
cross-sections of companies. Indeed, a specific political event cannot maintain decisive roles in firms' operations such as investment policies
affect the whole economy, and emanated risks are not equivalently (Bertrand & Schoar, 2003; Welch & Yoon, 2022; Pan, Wang and Weis­
circulated across firms. To overcome those drawbacks of the aggregate bach, 2016) and risk-taking (Lewellyn & Muller-Kahle, 2012; Pathan,
measures, Hassan et al. (2019) quantify the firm-level exposure to po­ 2009). Hence, the essential managerial attribute predominantly
litical risks and uncertainties using the computational linguistics stanches from a manager's understanding of the market to propose
method. Using the textual analysis of quarterly earnings conference-call corporate plans and technology (Kor, 2003). A handful of studies backed
transcripts, the Hassan et al. (2019) firm-level political risk measure can this view by acknowledging the economically significant roles of
capture more precise and relevant information on firms' political-related managerial ability in their firms' strategies and performance. A study by
hiring, investing, lobbying, and donating activities (Ahmad et al., 2023). Bertrand and Schoar (2003) confirms a substantial divergence in in­
Furthermore, the firm-level political risk can effectively depict corporate vestment, financial, and organizational decisions, which the managing
self-awareness, which is influential for their decision-making on known styles can explain. Chang, Dasgupta, and Hilary (2010) suggest a pro­
information (Citroen, 2011). Subsequent studies confirm the impacts of nounced connection between firm performance managers' traits or ex­
firm-level political risk on overall corporate investment (Choi et al., periences. Considering the IPO performance, Chemmanur and Paeglis
2022), financing decisions (Gyimah et al., 2022; Pan, Wang, & Yang, (2005) confirm that more able managers can obtain more profitable
2019), and corporate tax avoidance (Liu, Jin, Zhang, & Zhao, 2022). projects to proceed with IPO, improving the overall financial perfor­
Also falling within the scope of corporate investment policies, po­ mance. Similarly, findings by Cheung et al. (2017), Banker, Darrough,
litical uncertainty can significantly impact firms' investment decisions Huang, and Plehn-Dujowich (2013), and Chemmanur, Paeglis, and
on intellectual capital. Generally, intellectual capital investment com­ Simonyan (2010) also support the positive effects of management
prises innovation capital, human capital, and interpersonal capital in­ quality and corporate performance.
vestments that can be used to create wealth (Oliveira et al., 2010; Sydler, In addition, high-quality managers are better at understanding risks
Haefliger, & Pruksa, 2014). As such, intellectual capital – not tangible and overall market dynamics, more precisely predicting the product
assets or even financial capital is crucial tactical property as it is valu­ marketplace, and administering human resources compared with their
able, unique, and challenging to reproduce and thus a basis of compet­ counterparts (Demerjian et al., 2012). In other words, more able man­
itive advantage (Joshi et al., 2013) and the concept of long-term value agers can obtain more accurate information regarding investment
creation (Lerro, Linzalone, & Schiuma, 2014; Zhou & Fink, 2003). As prospects, granting firms better investment decisions and an elevated
intangible assets, prior studies have confirmed the critical relationship likelihood of success (Hasan, Alam, Paramati, & Islam, 2022). Regarding
between intellectual capital and firm performance (Maditinos, Chat­ risk management, managerial ability also exert positive impacts on
zoudes, Tsairidis, & Theriou, 2011), business reputation (Ginesti, Cal­ corporate investment during the crisis period (Andreou et al., 2017).
darelli, & Zampella, 2018), and financial management (D'Amato, 2021). Firms with higher managerial ability are expected to respond better to
Intellectual capital is a unique set of intangible assets which can economic pressures, competitive market, and environmental un­
contribute to a company's bottom line (Boekestein, 2006). Further, in­ certainties by better utilizing organizational resources and grasping
vestment decisions on intellectual capital are risky investments with promising investment prospects (; Demerjian et al., 2013). Hence,
large investments, longer payback periods, and high failure rates higher-quality firm management can enhance firms' capability to
(D'Amato, 2021; Sydler et al., 2014). Likewise, intellectual capital is execute more innovative policies, shaping risk-taking behavior and
significantly distinctive from physical assets because it is distinguished responding to unexpected uncertainties (Andreou et al., 2017; Chen,
by higher firm specificity and human capital intensity (Dierickx & Cool, Tseng, & Hsieh, 2015; Yung & Chen, 2018). In addition, the prominent
1989). As a result, investment decisions on intellectual capital are ex­ roles of manager ability also exhibit in capital management and raising
pected to be significantly driven by political uncertainty. From the capital finance. For instance, more able managers can do better in
viewpoint of the real option theory, during the higher uncertainty negotiating or dealing to get better external financing sources (Chem­
stages, firms tend to reduce costly oversights by dropping their invest­ manur et al., 2009). In other words, higher managerial ability can
ment disbursement or postponing investment decisions to “wait” for reduce information asymmetric between firms and creditors; therefore,
more favorable conditions at some point (Dixit & Pindyck, 2012). firms can obtain lower cost of debts (Owusu, Kwabi, Ezeani, & Owusu-
Hence, when facing higher political risks, firms may reduce intellectual Mensah, 2022).
capital investment to control the total risk level and sacrifice their The preceding debate directs us to suggest that managerial ability is a
long-term development potential. Further, the confounding uncertainty valuable factor in firms' policies for intellectual capital investments,
can exacerbate corporate financial constraints and increase the costs of especially for firms with higher revelation to political risk. Less able
financing, which are hypothetical reasons to justify investment reduc­ management teams tend to switch funds towards short-term and less
tion (Hu & Gong, 2019). Recent studies examine how environmental risky investments during more uncertain periods (Nadeem, Zaman,
guidelines impact corporate intellectual capital (Trevlopoulos et al., Suleman, & Atawnah, 2021). In other words, more able managers are
2021) and how macroeconomic factors affect the connection between expected to manage better intellectual capital investments, usually long-
corporate innovation and intellectual capital (Ren & Song, 2021). At the term investments with unreliable settlements (D'Amato, 2021). Taken
aggregate level, a recent study by Hoang and Tran (2022) confirms the together, more able managers can maintain investment scales, including
negative impacts of the Economic Policy Uncertainty Index (EPU) on the intellectual capital, when firms are more exposed to political un­
intellectual investment of UK firms. These conclusions confer cues to certainties, owing to better financial resources management and
examine the unclear relationship between intellectual capital invest­ reducing underinvestment puzzles. As such, we hypothesize that
ment and firm-level political risk. However, given the current literature, managerial ability, reflected in a more genuine managerial capacity
the impacts of political risk on corporate investment have not attracted (Demerjian et al., 2012), is a crucial channel to modify the impacts of
attention in prior studies. Based on the abovementioned arguments, we firm-level political risks on intellectual capital investment.
hypothesize that firms with higher political risks are more likely to
reduce their intellectual capital investment than their counterparts.

3
N. Huynh et al. International Review of Financial Analysis 91 (2024) 103020

3. Data and method Table 1

Variable definitions.
3.1. Data Variable Definition Source

Intellectual capital
To inspect the connection between firm-level political risk and in­ The natural logarithm of the ratio
tellectual capital (IC) investment, we utilize the sample of all U.S listed of intellectual capital (IC)
firms from CRSP/Compustat Merged database, which provides the investment on lagged total assets.
annual financial data from 2002 to 2021. The firm-level political risk Intellectual capital IC investment is calculated by Compustat and
investment (ICI) employing the IC accumulation authors' calculations
measures are obtained from the dataset of Hassan et al. (2019),2 which is rate (α) and the amortization rate
standardized by deducting the sample mean and dividing by the stan­ (δ) proposed by Sydler et al.
dard deviation. Those measures can directly quantify the cross-sectional (2014).
and time varying in firms' exposure to related political and regulatory Firm Political Risk
The natural logarithm of the
uncertainty and political events. In this study, we employ two indicators
average of the transcript-based
of Political Risk and Political Exposure to capture political risk at the firm counts of the number of political
level. Political Risk is constructed by counting all political bigrams, such bigrams of firms' conference call
as “political risk”, “politically risky”, “politically uncertain”, or “political reports by using the computational
uncertainty” from firms' earnings conference call transcripts. Political linguistic procedure to evaluate
Political Risk the contents. The counts focus on Hassan et al. (2019)
Exposure is constructed by counting all political bigrams by excluding risk or uncertainty bigrams, such
the terms “risk” or “uncertainty”. In addition, we also utilize eight as “political risk”, “politically
components of the Political Risk, as in Hassan et al. (2019), to further risky”, “politically uncertain”, or
consider the impacts of firm-level political risks on intellectual capital “political uncertainty” from firms'
earnings conference call
investment.3 Hassan et al. (2019) also posit that, overtime and across
transcripts.
industries, there is only a small portion of fluctuations in firm-level The natural logarithm of the
measures can be justified by changes in the aggregate indices. There­ average of the transcript-based
fore, this firm-level measures confines important statistics not attained counts of the number of political
by aggregate indices employed in earlier studies. As Hassan et al. (2019) bigrams of firms' conference call
reports by using the computational
provide a dataset on a quarterly basis, we compute the annual indicators Political Exposure Hassan et al. (2019)
linguistic procedure to evaluate
of political risk by taking the average of all quarters' values in a given the contents. The counts are not
fiscal year. conditioned for risk or uncertainty
We also collect the managerial ability (MA) scores constructed by bigrams by excluding the terms
“risk” or “uncertainty”..
Demerjian et al. (2012) to differentiate the potential impacts of firm-
level political risk on IC investment. To capture the levels of efficiency
credited to the firm's managers, Demerjian et al. (2012) construct the Components of Political Risk
The economic policy-specific
MA scores by using a two-step process, including frontier analysis and
component of political risk equals
data envelopment analysis (DEA) and regression to efficiency scores on Political Risk - the ratio of quarterly earnings
Hassan et al. (2019)
firm-specific attributes.4 This measure is assembled based on consis­ Economics conference calls of specific firms
tently accessible financial data overtime and comprises less noise than devoted to economic policy-
other managerial ability indicators (Demerjian et al., 2012). Therefore, related political risk.
The environment-specific
this measure of managerial ability has been extensively utilized and component of political risk equals
well-renowned in finance literature (Andreou et al., 2017; Demerjian Political Risk - the proportion of quarterly
Hassan et al. (2019)
et al., 2013; Kumar & Zbib, 2022; Lee et al., 2018; Phan et al., 2020). Environment earnings conference calls of
We also apply the standard data filtering approaches: (1) removing specific firms devoted to
environment-related political risk.
firm-year observations for financial institutions (SIC 6000–6999),
The trade policy-specific
regulated utilities (SIC 4900–4999), and unclearly-defined industries, component of political risk equals
(2) excluding firms with missing data for all variables used in the main Political Risk - the ratio of quarterly earnings
Hassan et al. (2019)
models, (3) including only firms with at least three years of data and (4) Trade conference calls of specific firms
winsorizing all continuous variables at 1% level on both sides. Our final devoted to trade policy-related
political risk.
sample of non-missing information of 29,504 firm-year observations The institutions and political
from 3688 unique firms spanning from 2002 to 2021. The descriptions process-specific component of
of all variables are reported in Table 1. political risk equal the proportion
Political Risk -
of quarterly earnings conference Hassan et al. (2019)
Institutions
calls of individual firms devoted to
the institutions and political
process-related political risk.
The health policy-specific
2
This dataset provides the political risk and exposure data at the firm level component of political risk equals
rather than the aggregate level. The data for this variable is publicly available Political Risk - the ratio of quarterly earnings
Hassan et al. (2019)
Health conference calls of specific firms
at: https://www.policyuncertainty.com/firm_pr.html
3 devoted to healthcare policy-
These eight constituents are constructed by quantifying specific political related political risk.
risk topics related to institutions, economics, trade, technology, taxes, health, The security and defense policy-
environment, and security. specific component of political risk
4
As the data from Demerjian's database are available through 2020, we equals the proportion of quarterly
Political Risk -
compute the MA score for all firms in our sample in 2021 by utilizing the simple earnings conference calls of Hassan et al. (2019)
Security
moving average from MA during the past three years. This approach can be individual firms devoted to
rationalized by long-term managerial ability, as in Doukas and Zhang (2021). security and defense policy-related
political risk.
For full details about this measure, please refer to Demerjian et al. (2012). The
data for this variable is publicly available at: https://peterdemerjian.weebly. (continued on next page)
com/managerialability.html

4
N. Huynh et al. International Review of Financial Analysis 91 (2024) 103020

Table 1 (continued ) this approach utilizes the intellectual capital-creating expenses (IE) by
Variable Definition Source adding up the three proxies: labour expenses (HE), R&D expenditures
(RE), and advertising expenses (AE), for intellectual capital. The IC in­
The tax policy-specific component
of political risk equals the
vestment is computed using the IC accumulation rate (α) and the
proportion of quarterly earnings amortization rate (δ). Sydler et al. (2014) also assume that the growth
Political Risk - Tax Hassan et al. (2019)
conference calls of individual factor (g) is constant for IC-creating expenses (IE), which is the risk-free
firms devoted to tax policy-related rate of the average six-month US treasury bill, as suggested by Ballester,
political risk.
Livnat, and Sinha (2002). The value of the IC at the end of the period is
The technology policy-specific
component of political risk equals computed as follows:
the proportion of quarterly
Political Risk - ICt = α(HEt + REt + AEt ) + (1 − δ)(ICt− 1 ) = α(IEt ) + (1 − δ)(ICt− 1 ) (1)
earnings conference calls of Hassan et al. (2019)
Technology
individual firms devoted to ICt is the intellectual capital at time t, α is the accumulation rate (0 <
technology and infrastructure
policy-related political risk.
α < 1), and δ is the amortization rate (0 < δ < 1). It means that we can
capture the levels of IE accumulated in IC during the current financial
year - α(IEt ) and levels of prior year IC investment remains after
Firm-level control variables
The natural logarithm of total considering amortization (1 − δ)(ICt− 1 ). Then, we apply the value of
Firm Size assets computed over a firm's Compustat ICt− 1 recursively results in ICt in the following equation:
financial year. [ ( )2 ( )t ] [ ]
The ratio of the market value of 1− δ 1− δ 1− δ 1+g
the common equity and its balance ICt = α(IEt ) 1 + + +…+ = α(IEt )
Market to Book 1− g 1− g 1− g δ+g
sheet value of the common equity Compustat
ratio
computed over a firm's financial (2)
year.
The leverage ratio is the ratio of With the model of Sydler et al. (2014), we obtain the values of 0.752
Leverage total debts and assets computed Compustat for α and 0.0124 for δ in our U.S sample. It indicates that 75.2% of
over a firm's financial year. existing IE is accrued to generate IC on a yearly basis, while total IC
The ratio of total earnings to total amortizes at the rate of 1.24% from prior year values. Then, we apply
Profitability assets is computed over a firm's Compustat
financial year.
Eq. (2) to calculate the value of IC investment for each firm in a given
Measured by total cash plus fiscal year. After obtaining the annual levels of IC, we construct the
marketable securities to lagged primary dependent variable using the log-transformed ratio of IC in­
Cash holdings Compustat
total assets computed over a firm's vestment to the lagged total assets (TAi,t− 1 ). In other words, our primary
financial year.
dependent variable is the intellectual capital investment - ICIi,t , which
The annual growth in total sales is ( )
ICI
Sales growth computed over a firm's financial Compustat equals to ln TAi,t−i,t 1 . To strengthen our findings, we also employ two
year.
The buy-and-hold stock return of alternative measures of ICI, including modification of Sydler et al.
Stock return the financial year is computed Refinitiv (2014) approach and the value-added intellectual coefficient (VAIC)
over a firm's financial year. approach proposed by (Pulic, 2000), which are reported in Section 4.1.
The fraction of shares institutional
Institutional
investors own in a firm's financial Refinitiv
ownership
year. 3.3. Bassline models
Natural logarithm of the total
Board Size Boardex
number of directors on the board.
To estimate the impacts of firm-level political risk and IC investment,
The binary variable takes the value
CEO Duality one if the CEO also chairs the Boardex the regression model takes the following form:
board and 0 otherwise. ∑
Managerial ability ICI i,t = α0 + β1 PRi,t− 1 + δk Controlsi,t− 1 + εi,t− 1 (3)
(MA)
The dummy variable equals one if where, ICIi,t is our main variable of interest - the intellectual capital
the firms' managerial ability scores
investment of firm i in year t. PR is the firm-level Political Risk and Po­
MA_HIGH in year t are higher than the whole
sample mean over the examined litical Exposure. Controls include firm-specific variables that are standard
Demerjian's database
period or equals zero. for the IC and corporate investment literature (Gulen & Ion, 2016;
and authors'
The dummy variable equals one if
calculations
Hoang & Tran, 2022; Sydler et al., 2014). We also control for firm-fixed,
the firms' managerial ability scores industry-fixed, and time-fixed effects in all equations, and all standard
MA_LOW in year t are lower than the whole
sample mean over the examined
errors are clustered at the firm level. The descriptions of all variables are
period or equals zero. reported in Table 1.
Following Phan et al. (2020) and Demerjian et al. (2013), we divide
This table describes the definitions of variables used in the analyses with the
the sample into high-score firms (MA_HIGH) and low-score firms
data sources.
(MA_LOW) by comparing the annual firms' MA scores to the whole
sample mean over the examined period. Then, we utilize the following
regression model:
3.2. Intellectual investment measures
ICI i,t = α0 + α1 Political Riski,t− 1 + β1 Political Riski,t− 1 × MA HIGH i,t− 1
∑
Following prior literature on intellectual capital, we utilize the +β2 Political Riski,t− 1 × MA LOW i,t− 1 + δk Controlsi,t− 1 + εi,t− 1
adjusted residual income model proposed by Sydler et al. (2014).5 As the (4)
intellectual capital level is not reported in the firm's financial reports,
where, MA_HIGH (MA_LOW) is a dummy variable, which equals one if
the MA score of firm i in year t is higher (lower) than the mean of all
5
For recent literature on intellectual capital that utilized the method of firms over the sample period or equals zero otherwise. The set of main
Sydler et al. (2014), see Osinski et al. (2017), Hoang and Tran (2022), Kweh, variables, controls, and fixed effects are identical to Eq. (3). This model
Lu, Tone, and Nourani (2022), and others. allows us to directly capture the economic significance for each MA

5
N. Huynh et al. International Review of Financial Analysis 91 (2024) 103020

Table 2
Summary statistics and univariate results.
Panel A. Descriptive statistics

Variable Mean S.D. Min Median Max N

Intellectual capital investment

Intellectual capital investment (ICI) 3.549 1.721 0.145 3.126 8.040 29,504
Firm political risk
Political Risk − 0.075 0.856 − 0.535 − 0.294 − 0.002 29,504
Political Exposure − 0.084 0.832 − 0.599 − 0.324 − 0.002 29,504
Firm-level control variables
Firm Size 7.864 1.732 5.713 7.381 9.901 29,504
Leverage 0.222 0.186 0.000 0.192 0.454 29,504
Profitability 0.034 0.108 − 0.045 0.040 0.125 29,504
Cash holdings 0.186 0.172 0.012 0.090 0.387 29,504
Sales growth 0.081 0.187 − 0.106 0.049 0.235 29,504
Stock return 0.118 0.345 − 0.267 0.069 0.432 29,504
Institutional ownership 0.509 0.448 0.000 0.487 0.908 29,504
Board Size 1.966 1.730 0.000 1.879 3.503 29,504
CEO Duality 0.169 0.158 0.000 0.000 1.000 29,504

Panel B. Univariate analyses

Univariate analyses Political Exposure Political Risk

High Low Difference High Low Difference

Intellectual capital 2.662 4.472 ¡1.810*** 2.538 4.564 ¡2.026***

Firm Size 8.068 7.667 0.401** 8.336 7.471 0.865***
Leverage 0.221 0.223 − 0.002 0.219 0.224 − 0.005
Profitability 0.029 0.039 − 0.010** 0.028 0.040 − 0.012***
Cash holdings 0.196 0.182 0.014* 0.202 0.156 0.046**
Sales growth 0.071 0.082 − 0.011*** 0.069 0.085 − 0.016***
Stock return 0.116 0.121 − 0.006* 0.113 0.126 − 0.012**

This table presents the summary statistics and univariate analysis in Panel A and B, respectively. The sample consists of 3688 U.S firms between 2002 and 2021. The
firm-level political risk, proxied by political exposure and political risk, is collected from Hassan et al. (2019). The intellectual capital investment is computed using the
approach of Sydler et al. (2014). The firm-specific variables include Firm Size, Market to Book ratio, Leverage, Profitability, Cash holdings, Sales growth, Stock return,
Institutional ownership, Board size, and CEO duality. All continuous variables are winsorized at the 1st and 99th percentiles. Panel B shows the mean differences for the
subsamples with high and low firm-level Political Risk and Political Exposure using a two-sample t-test. ***, **, and * indicate significance at the 1%, 5%, and 10% levels,
respectively. The descriptions of all variables are reported in Table 1.

group and compute the magnitude of differences. level political risk and IC investment.6
We present the baseline results in Table 3 for the impacts of firm-
4. Baseline results level PR, proxied by Political Risk and Political Exposure, on firms' in­
tellectual capital investment. Utilizing the multivariate regression from
4.1. Firm-level political risk and IC investment Eq. (1), we perform two modified models with and without industry-
fixed effects in (1) and (2), respectively.7 In Column (1), the estimated
Table 2 presents the summary statistics and univariate analyses of coefficients of Political Exposure and Political Risk are all negative
selected variables in this study. We only report the statistics for variables (− 0.132 and − 0.194) and statistically significant at a 1% level, indi­
in our main models for brevity. In Panel B, to perform our univariate cating the negative impact of PR on ICI. With the industry fixed effect in
analyses, we split our sample into High and Low subsamples based on the Column (2), the estimated coefficients for Political Risk and Political
means of firm-level Political Risk and Political Exposure. Across two in­ Exposure remain negative and statistically significant at the 1% level.
dicators of PR, we observe that firms with high PR are highlighted with a Given the economic significance of two PR indicators, the impacts of
lower IC investment compared to those with lower PR. The mean dif­ Risk are relatively stronger compared to Exposure, suggesting that firms'
ferences of ICI for high and low PR are all statistically significant at a 1% intellectual capital investment decisions are more sensitive to political
level, suggesting that high-PR firms are more likely to reduce their in­ hazards than basic exposure to political uncertainties. For the control
vestment in ICI than low-PR firms. Similarly, considering the statisti­ variables, profitability and sales growth also positively and significantly
cally significant test-of-difference values for other variables, our results impact intellectual capital investment.8 In Appendix B1, we also
indicate that high high-PR firms, when compared to low-PR firms, (1)
are comparatively larger, (2) are less profitable, (3) hold more cash, (4)
have lower sales growth, and (5) have lower annual stock returns. 6
In untabulated results, our main variables do not suffer from the multi­
Overall, our preliminary subsample analyses suggest that firms with collinearity issues supported by the VIF tests.
7
high PR have considerably different firm characteristics compared to For all following tables, we perform all regressions with industry fixed effect
low-PR firms, especially the firms' intellectual capital investment levels. as all modified models provide statistically similar outcomes.
8
In Appendix A, we also present the Pearson correlations in Table A for all For robustness, we have re-estimated our main estimations from Equation
(3) and (4) by including one- to three-year lags of the dependent variable and
main and control variables. Overall, the correlation coefficients between
models without and with lagged control variables. Overall, the untabulated
ICI and PR indicators are negatively significant, indicating that firms
results remain qualitatively unchanged that firm-level political risk has nega­
facing higher levels of political risk reduce their IC investment. This tive impacts on IC investment. Those additional tests also suggest that our re­
lends early support to our prediction for the relationship between firm- sults are robust to within-firm autocorrelation.

6
N. Huynh et al. International Review of Financial Analysis 91 (2024) 103020

Table 3 surrounding political or policy environment (Mishra, 2023). Hence, the

Firm-level political risk and intellectual capital investment. nexus between firm-level PR and ICI possibly reflects the impacts of such
Dependent variable: Political Exposure Political Risk aggregate political uncertainties. To confirm that our results do not
Intellectual Capital carry the effect of aggregate political risk in prior studies, we further
(1) (2) (1) (2)
Investment (ICI) perform horse race regressions by imitating the baseline model- Eq. (3)
Political Exposure − 0.132*** − 0.125*** in Table 4. In this analysis, we further control for two economywide
(0.012) (0.011) political uncertainty indices of Geopolitical Threats (GT) and Geopolit­
Political Risk − 0.194*** − 0.191*** ical Risk Index (GRI) developed by Caldara and Iacoviello (2022).9 As
(0.009) (0.008)
Firm Size − 0.138** − 0.126* − 0.040 − 0.053
the Caldara and Iacoviello (2022) data is on a monthly basis, the GT and
(0.142) (0.165) (0.231) (0.228) GRI are computed from 12 monthly values captured from the financial
Leverage − 0.239** − 0.166 − 0.201 − 0.145 year-start month to the financial year-end month. In Table 4, the esti­
(0.050) (0.259) (0.161) (0.234) mated coefficients on GT and GRI are negatively significant. Further,
Profitability 0.322** 0.420*** 0.389*** 0.410***
upon controlling for the aggregate political uncertainty indices, the
(0.045) (0.012) (0.020) (0.018)
Cash holdings 0.018 0.017 0.023 0.026 Political Risk and Political Exposure continue to load with positive and
(0.109) (0.100) (0.099) (0.057) significant coefficients at the 1% level. As such, the impact of firm-level
Sales growth 0.167*** 0.124*** 0.175*** 0.212** PR on intellectual capital investment is unique and is in addition to the
(0.009) (0.013) (0.008) (0.005) aggregate political risks.10
Stock return 0.229* 0.206 0.232* 0.211
(0.072) (0.227) (0.080) (0.209)
Having established a robust relationship between a firm's political
Institutional ownership − 0.016 − 0.012 0.015 0.012 risks and intellectual capital investment, we seek to validate the cross-
(0.131) (0.145) (0.198) (0.197) sectional heterogeneity by which PR affects ICI in Table 5. First, we
Board Size 0.364 0.388 0.294 0.273 create dummy variables by utilizing six firm-specific factors, including
(0.287) (0.128) (0.374) (0.392)
financial distress - FD (KZ index11 and Altman' (1968) Z-score12) (Chen
CEO Duality − 0.121 − 0.124 − 0.131 − 0.116
(0.115) (0.104) (0.251) (0.357) & Wang, 2012; Phan et al., 2020), dependence on external finance -
Constant 1.113*** 2.463*** 1.755** 2.318*** EFD13 (Rajan & Zingales, 1998), institutional ownership, HHI14 (Her­
(0.057) (0.020) (0.121) (0.057) findahl–Hirschman Index of Institutional Ownership) (Choi et al., 2022),
S.E. clustered by Firm Yes Yes Yes Yes and information efficiency (IE) proxied by Analyst Coverage (Chang,
Firm-fixed effect Yes Yes Yes Yes
Dasgupta, & Hilary, 2006). For each firm-a firm-specific factor, we
Time-fixed effect Yes Yes Yes Yes
Industry-fixed effect No Yes No Yes create a dummy variable that equals one if the value of firm i in year t is
Observation 29,504 29,504 29,504 29,504 higher (lower) than the mean of all firms over the sample period or
Adjusted R-squared 0.702 0.770 0.805 0.823 equals zero otherwise. Then, we interact two proxies of PR with the
This table presents the results of OLS regression models investigating the impact identified dummy variables. As expected, all interaction terms in Panel
of firm-level political risk and corporate intellectual capital investment (ICI). A are negative and statistically significant for both FD and EFD. These
The sample consists of 3668 U.S firms between 2002 and 2021. The firm-level results suggest that the adverse effects of political risk on intellectual
political risk, proxied by political exposure and political risk, is collected from capital investment are more prominent for firms encountering higher
Hassan et al. (2019). The intellectual capital investment is computed using the financial constraints and external financing. This finding is in line with
approach of Sydler et al. (2014). We control for industry, firm, and time-fixed prior corporate investment literature that firms need to ignore invest­
effects in all specifications. Standard errors are clustered at the firm level and ment opportunities due to higher transaction costs (both debt and equity
are presented in parentheses under the associated coefficients. ***, **, and *
issuance) (Faulkender & Wang, 2006) and potential cash flow problems
indicate significance at the 1%, 5%, and 10% levels, respectively. The de­
(Paquette, Huynh, & Vu, 2022).
scriptions of all variables are reported in Table 1.
In Panel B, we investigate other channels through which impacts of
firm-level political risk on intellectual capital investment can be miti­
examine the impacts of eight political risk category-based indicators.
gated. Considering the interaction terms of institutional ownership and
The results indicate that only Economic Policy, Environment, and Tax
Policy risks exert adverse and substantial effects on firms' IC investment,
which aligns with the findings of Hassan et al. (2019). First, by following 9
The indices are publicly available at: https://www.matteoiacoviello.co
the approach of Phan et al. (2020), we utilize the sub-sample analysis for m/gpr.htm
eight sectors in Appendix B2, including Consumer staples, Consumer 10
In unreported results, we also include GT and GRI indices as control vari­
Discretionary, Health care, Information technology, Commercial ser­ ables for all other tables (5 to 11). Overall, the results are strongly consistent
vices, Industrials, Energy, and Materials. Overall, our results are sig­ with our current reported results. The results are available on request.
nificant for all eight sectors, confirming our main findings. 11
Following a study by Baker, Stein, and Wurgler (2003), we compute the KZ
Given the attention of political uncertainties, higher PR means that Index as − 1.002 × Cashflow − 39.368 × Dividends − 1.315 × Cash +3.139 ×
firms suffer both underlying costs (reducing reputation and competitive Leverage.
12
advantage) (Hassan et al., 2019; Keillor et al., 2005) and explicit costs Following Chen and Wang (2012), we compute the Altman Z-score as:Z −
(increase in the cost of financing, reducing firms' asset returns and cash score = 1.20X1 + 1.40X2 + 3.30X3 + 0.60X4 + 0.999X5 , with X1 to X5 are
working capital to book value of assets ratio, retained earnings to book value of
flows) (Brogaard & Detzel, 2015; Huynh, 2023; Mishra, 2023). There­
assets ratio, EBIT to book value of assets ratio, the market value of equity to
fore, firms with high PR are more likely to reduce their investment,
total liabilities ratio, and net sales to book value of assets ratio, respectively.
namely intellectual capital, due to profitability uncertainties and po­ The dummy variable equals to 1 for firms with higher financial distress if the
tential default on financial commitments. Further, firm-level PR in­ value of the Altman Z-score is lower than 1.81.
creases managerial conservatism as managers become more risk-averse 13
In this study, we compute the level of dependence on external finance (EFD)
during high-uncertainty periods (Hasan et al., 2022; Panousi & Papa­ by following Rajan and Zingales (1998). The EFD is the difference between
nikolaou, 2012). As a result, firms maintain more extensive cash re­ capital expenditures and cash flow from operations scaled by capital
serves and reduce their investment. Overall, our findings link and expenditures.
14
expand prior literature on the impacts of uncertainties on corporate The HHI - Herfindahl–Hirschman Index is computed from the sum of
investment (Choi et al., 2022; Handley & Limao, 2015), human capital squares of the proportions of the firm's shares held by the top five institutional
(Naidenova, 2022), and corporate innovation (Huang & Yuan, 2021). investors. The higher values of HHI indicate that the ownership is assembled
within a handful of large institutional investors (Choi et al., 2022; Ferreira,
It is noticeable that firm-level PR also imitates the tremors from the
Matos, Pereira, & Pires, 2017).

7
N. Huynh et al. International Review of Financial Analysis 91 (2024) 103020

Table 4 for low-MA firms, for two indicators of PR. In the last four columns, we
Firm-level political risk and intellectual capital investment - controlling for also report the actual and percentage differences in ICI between two
aggregate political risk. clusters of MA scores. Higher MA helps firms reduce the negative im­
Dependent variable: ICI Geopolitical Threats (GT) Geopolitical Risk Index pacts of Political Exposure on ICI by 31.82% compared to low-MA firms.
(GRI) For the Political Risk, the divergence between firms with low and high
(1) (2) (3) (4) MA is higher, with a 34.44% reduction in intellectual capital investment.
Briefly, the above findings imply that better managerial teams can
Political Exposure − 0.112*** − 0.195***
(0.012) (0.008) moderate the destructive impacts of political risks on firms' intellectual
Political Risk − 0.137*** − 0.201*** capital investment opportunities (up to 35%), which collaborates with
(0.009) (0.007) prior findings of Lee et al. (2018) and Gan (2019) on corporate invest­
GT − 0.051* − 0.060** ment. In other words, firms with higher managerial ability are more
(0.088) (0.076)
GP − 0.144** − 0.132**
capable of realizing risk and aligning resources for corporate invest­
(0.041) (0.069) ment, such as intellectual capital, when exposed more to unexpected
Controls Yes Yes Yes Yes political uncertainties (Andreou et al., 2017; Demerjian et al., 2013).
S.E. clustered by Firm Yes Yes Yes Yes To better identify the effect of managerial ability on the PR-ICI
Firm/Time/Industry-fixed
Yes Yes Yes Yes nexus, we rely on cross-sectional comparisons in Table 7. We consider
effects
Observation 29,504 29,504 23,603 23,603 four firm-specific factors of cross-sectional heterogeneity (See Table 4):
Adjusted R-squared 0.495 0.665 0.667 0.762 Financial distress (KZ index), External financial dependence (EFD),
Institutional ownership (IO), and Analyst Coverage (AC). We divide our
This table presents the results of OLS regression models investigating the impact
of firm-level political risk and corporate intellectual capital investment (ICI).
sample into High and Low groups using the sample mean and then
The sample consists of 3668 U.S firms between 2002 and 2021. The firm-level utilize Eq. (4) for each sub-sample. Panel A and B report the results for
political risk, proxied by political exposure and political risk, is collected from Political Exposure and Political Risk, respectively. Regarding the coeffi­
Hassan et al. (2019). Two economywide political uncertainty indices of cient differences, the absolute value of β1 is necessarily lower than that
Geopolitical Threats (GT) and Geopolitical Risk Index (GRI), developed by of β2 across all sub-samples, which reconfirms that managerial ability
Caldara and Iacoviello (2022). We control for industry, firm, and time-fixed can lower the adverse effect of PR on IC investment. Turning to the
effects in all specifications. Standard errors are clustered at the firm level and economic significance, the differences also reveal the reliability with our
are presented in parentheses under the associated coefficients. ***, **, and * prior outcomes that high MA scores can help decrease the harmful ef­
indicate significance at the 1%, 5%, and 10% levels, respectively. The de­ fects of political risks from 20% to 40%. However, when we compare the
scriptions of all variables are reported in Table 1.
different levels between high versus low sub-samples, the modified ef­
fect of MA also depends on firm firm-specific factors. Specifically, the
HHI, we observe that the mitigation effects of institutional holdings on effect is more significant for firms with higher FD and EFD and lower IO
political exposure and risk are positive and statistically significant. In and AC than their counterparts (about 24% to 39% compared to 18% to
other words, firms with higher institutional holdings, a proxy for good 28%). Overall, our results corroborate that the significant role of
firm-level governance, can offset the effects of political exposure and managerial ability is more valuable for firms with poorer firm-level
risk on ICI. This governance mechanism efficiently enhances the firm's governance, greater financial constraints, and information asymmetry.
value and alleviates agency problems (Mishra, 2023). The interaction
terms of IE are all positive but statistically insignificant, indicating that 5. Robustness checks and additional analyses
Analyst Coverage does not exert significant impacts on reducing the
adverse effect of political risk on their intellectual capital. 5.1. Heterogenous impacts of political risks across sectors

4.2. Role of managerial ability and political risk-intellectual capital nexus In the first additional analysis, we extend our line of research by
considering the sector-level heterogeneity in the relationship between
Given that crude firm-level political risk harms the firm's intellectual political risk, intellectual capital investment, and managerial ability.15
capital investment, we instantaneously investigate the role of manage­ One scenario is that the negative impacts of PR on ICI are more visible
rial capability in this connection. Using the mean of firms' MA scores to for high-tech industry firms as they tend to rely more on intellectual
split the sample into high-score firms (MA_HIGH) and low-score firms capital than their counterparts. Therefore, we expect that high-tech
(MA_LOW), we utilize Eq. (4) to consider the modified impacts of MA in firms will drive our results as they need to promptly alter their invest­
Table 6. In Panel A, we report the coefficients (β1 and β2 ) of the inter­ ment strategies in IC when faced with PR. In this section, we follow the
action term between Political Risk and Political Exposure and MA score approach of Loughran and Ritter (2004) to split our sample into two
dummy variables. In line with our prior findings, the effect of PR on ICI groups of high-tech firms and non-high-tech firms.16 We then re-
is negative and statistically significant. The results from the Wald test estimate the results using Eq. (4) and report the results in Table 8.
further confirm that the absolute values of β1 are significantly lower Consistent with our prediction, we observe that high-tech firms are in
than that of β2 , indicating that firms with lower scores of MA are more the driver's seat. For both indicators of PR, high-tech firms are more
significantly impacted by political risks than those with higher MA sensitive to the change of political risk regarding more significant re­
scores. In other words, we can confirm that managers with better ductions in ICI, as evidenced by chi-square test comparisons on α1 .
managerial ability can help lessen the harmful impacts of PR on ICI.
In Panel B, we further validate our results by considering the eco­
nomic significance. We compute the economic impacts of one standard 15
deviation change in firm-level Political Risk and Political Exposure on ICI In Appendix B, we have confirmed the impacts of PR and ICI across eight
sectors. In unreported results, we also utilize Equation (4) to consider the
for the whole sample and two sub-sample of high and low MA scores. We
modified impacts of MA for those sectors, with the significant difference be­
report the changes in ICI all in percentage points. On average, a one
tween high vs low MA groups. However, we do not find significant differences
standard deviation increases in Political Risk and Political Exposure re­ between those sectors regarding the power of MA.
sults in a 2.267 and 3.095 percentage points drop in ICI in the subse­ 16
High-tech firms are classified as those in 4-digit SIC codes of 3571, 3572,
quent year. We also confirm that the heterogeneity for firms is clustered 3575, 3577, 3578, 3661, 3663, 3669, 3671, 3672, 3674, 3675, 3677, 3678,
based on their MA scores. The IC investment declines 4.178 and 2.868 3679, 3812, 3823, 3825, 3826, 3827, 3829, 3841, 3845, 4812, 4813, 4899,
percentage points for high-MA firms, while it is 2.739 and 1.927 points 7371, 7372, 7373, 7374, 7375, 7378, and 7379.

8
N. Huynh et al. International Review of Financial Analysis 91 (2024) 103020

Table 5
Cross-sectional heterogeneity: firm-level political risk and intellectual capital investment.
Panel A: Financial Distress (FD) and External Finance Dependence (EFD)

Dependent variable: ICI KZ index (high) Altman Z-score (Low) EFD (High)

(1) (2) (1) (2) (1) (2)

Political Exposure − 0.144*** − 0.103***

(0.018) (0.028)
Political Exposure × FD − 0.123** − 0.097**
(0.031) (0.043)
Political Risk − 0.129*** − 0.183***
(0.011) (0.011)
Political Risk × FD − 0.175*** − 0.124***
(0.010) (0.018)
Political Exposure − 0.103**
(0.025)
Political Exposure × EFD − 0.141***
(0.007)
Political Risk − 0.122***
(0.011)
Political Risk × EFD − 0.125***
(0.005)
Controls Yes Yes Yes Yes Yes Yes
S.E. clustered by Firm Yes Yes Yes Yes Yes Yes
Firm/Time/Industry-fixed effects Yes Yes Yes Yes Yes Yes
Observation 29,504 29,504 29,504 29,504 22,128 22,128
Adjusted R-squared 0.437 0.460 0.482 0.505 0.439 0.464

Panel B: Institutional ownership (IO) and Information efficiency (IE)

Dependent variable: ICI Institutional ownership (High) Herfindahl index (High) Analyst Coverage (Low)

(1) (2) (1) (2) (1) (2)

Political Exposure − 0.182*** − 0.114**

(0.012) (0.024)
Political Exposure × IO 0.492** 0.120**
(0.022) (0.024)
Political Risk − 0.253*** − 0.260***
(0.010) (0.004)
Political Risk × IO 0.207** 0.108**
(0.018) (0.034)
Political Exposure − 0.208***
(0.006)
Political Exposure × IE 0.087
(0.198)
Political Risk − 0.211**
(0.018)
Political Risk × IE 0.017
(0.389)
Controls Yes Yes Yes Yes Yes Yes
S.E. clustered by Firm Yes Yes Yes Yes Yes Yes
Firm/Time/Industry-fixed effects Yes Yes Yes Yes Yes Yes
Observation 29,504 29,504 29,504 29,504 20,948 20,948
Adjusted R-squared 0.463 0.497 0.395 0.462 0.442 0.374

This table presents the results of OLS regression models investigating the impact of firm-level political risk and corporate intellectual capital investment (ICI) by
considering the cross-sectional heterogeneity. The sample consists of 3668 U.S firms between 2002 and 2021. The firm-level political risk, proxied by political exposure
and political risk, is collected from Hassan et al. (2019). The intellectual capital investment is computed using the approach of Sydler et al. (2014). The Dummy factors
are identified by firm-specific factors, which are interacted with the main indicators of firm-level political risks. In Panel A, the dummy equals one if firms are
financially distressed, as defined by the KZ index, Altman Z-score, and External financial dependence (EFD). In Panel B, the dummy is equal to 1 if firms with high
Institutional ownership, Herfindahl index, and low Analyst coverage. We control for industry, firm, and time-fixed effects in all specifications. Standard errors are
clustered at the firm level and are presented in parentheses under the associated coefficients. ***, **, and * indicate significance at the 1%, 5%, and 10% levels,
respectively. The descriptions of all variables are reported in Table 1.

Regarding the results on MA, our results further confirm that firms with two alternative proxies of IC investment as the dependent variables.
more able managers can considerably reduce the impact of PR on ICI. Following a study by Hoang and Tran (2022), we compute an alternative
Notably, the economic magnitude for MA between two subgroups of indicator as the new IC assets created during a fiscal year (A-ICI), which
high-tech and non-high-tech firms are insignificantly different, which is the amount of IE accumulated into IC assets. A-IC is the log-
surrounds 30%. In other words, the modified impacts of MA on the PR- transformed fraction between the product of IE and α scaled by the
ICI nexus are not heterogeneous across different sectors. one-year lagged total assets. The second measure is the value-added
intellectual coefficient (VAIC) approach developed by (Pulic, 2000),
which is widely employed by prior studies (Nadeem et al., 2021; Soe­
5.2. Alternative indicators of intellectual capital investment warno & Tjahjadi, 2020). The VAIC approach implicitly determines in­
tellectual capital as the sum of capital employed efficiency (VACA),
To strengthen our findings, we employ various robustness tests with

9
N. Huynh et al. International Review of Financial Analysis 91 (2024) 103020

human capital efficiency (VAHU), and structural capital efficiency Table 6

(STVA). Consequently, the VAIC approach provides information about Managerial ability: firm-level political risk and intellectual capital investment.
the efficiency of both the tangible (capital employed) and intangible Panel A: High versus low managerial ability
(human and structural capital) assets of a firm. Therefore, a higher value
Dependent variable: ICI Political Dependent variable: Political
of VAIC highlights a better utilization of firms' resources for value- Exposure ICI Risk
creation processes. The VAIC is computed as follows:
(1) (2)
VA VA Political Exposure − 0.131*** Political Risk − 0.173***
VACA = ; VAHU = ; STVA
Capital Employed Human Capital (0.007) (0.023)
Structural Capital Political Exposure × Political Risk ×
− 0.107* − 0.113**
= MA_HIGH (β1 ) MA_HIGH (β1 )
VA
(0.105) (0.019)
Where, Value Added (VA) = Amortization + Depreciation + Operating Political Exposure ×
− 0.283**
Political Risk ×
− 0.269***
Profit (OP) + Employees Cost (EC); Capital Employed = Total Assets - MA_LOW (β2 ) MA_LOW (β2 )
(0.038) (0.023)
Intangible Assets; Human Capital = Total employees' salaries and wages
Controls Yes Controls Yes
which were paid annually; and Structural Capital = Value Added - Human S.E. clustered by Firm Yes S.E. clustered by Firm Yes
Capital. We use the baseline model- Eq. (3) for two alternative IC in­ Firm/Time/Industry- Firm/Time/Industry-
Yes Yes
vestment proxies and reported the results in Table 9. Overall, the main fixed effects fixed effects
Observation 29,504 Observation 29,504
findings remain robust under two alternative proxies that firms with
Adjusted R-squared 0.542 Adjusted R-squared 0.676
higher PR significantly reduce their ICI in the following year. β1 − β2 (Wald test) 0.176*** β1 − β2 (Wald test) 0.156***
In the next robustness check, we apply Eq. (4) for two alternative ICI (p-value) 0.000 (p-value) 0.000
proxies and the managerial ability scores and report the results in
Table 10. In Panel A, the Wald test results confirm that absolute values of
β1 are substantially lower than that of β2 , demonstrating that firms with Panel B: Economic significance.
lower MA scores are less affected by political risks than those with better
Dependent All Low managerial High managerial
MA scores. The economic significance in Panel B further confirms that variable: ICI firms ability firms ability firms
firms with more able managers can lessen the harmful impacts of po­
(1) (2) (3)
litical risks (around 22% to 27%) on firms' intellectual capital invest­
ment. Overall, we find consistency with our prior results on the power of Political Exposure − 2.267 − 2.868 − 1.927
Political Risk − 3.095 − 4.178 − 2.739
managerial ability when utilizing the alternative indicators of intellec­
tual capital investment.
Difference (3)–(2)
Political Exposure 0.941
5.3. The impacts of U.S presidencies and elections % − 32.81%
Political Risk 1.439
Prior studies confirm that political risk in the U.S is relatively lower % − 34.44%
during the presidency of the Democratic party (Blinder & Watson, 2016; This table presents the results of OLS regression models that investigate the ef­
Santa-Clara & Valkanov, 2003) and higher during election years fect of managerial ability (MA) on the relationship between firm-level political
(Marshall, Nguyen, Nguyen, & Visaltanachoti, 2018). Prior studies by risk and corporate intellectual capital investment (ICI) by using Eq. (4). The
Julio and Yook (2012) and Jens (2017) confirm that firms lessen their sample consists of 3668 U.S firms between 2002 and 2021. The firm-level po­
investment in the election years in relation to non-election years. In this litical risk, proxied by political exposure and political risk, is collected from
section, we further consider an additional analysis of how the U.S Hassan et al. (2019). The intellectual capital investment is computed using the
presidency and elections modify the nexus between firm-level political approach of Sydler et al. (2014). The Dummy of MA_HIGH (MA_LOW) is created
uncertainties and intellectual capital investment. We utilize the inter­ using Demerjian's managerial ability score higher (lower) than the average score
of all firms over the sample period or equal zero otherwise, which interacted
action terms between political risk and exposure and two dummies of U.
with the main indicators of firm-level political risks. Panel A reports the dif­
S presidencies (Democratic versus Republican) and Election years. The
ferences from regression with the Wald test for the difference between β1 and β2.
dummy of Democratic equals to one if the U.S president is a Democrat in Panel B economic significance of one standard deviation change in the political
the given financial year and zero otherwise. Similarly, the dummy of risk on IC for firms with low and high managerial ability, and its percentage of
Election to one if the given financial year are election years, and zero differences. We control for industry, firm, and time-fixed effects in all specifi­
otherwise. Then, we utilize the baseline model (Eq. 3) by adding two cations. Standard errors are clustered at the firm level and are presented in
aforementioned dummies and reported the results in Table 11. In parentheses under the associated coefficients. ***, **, and * indicate significance
addition, we also document the potential impacts of the political envi­ at the 1%, 5%, and 10% levels, respectively. The descriptions of all variables are
ronment by considering the U.S presidencies (Democratic versus reported in Table 1.
Republican) and Election years. The estimated coefficients of the
interaction terms of PR × Democratic are negative and significant, 5.4. Alternative econometric approaches and sensitivity analyses
indicating that the impacts of firm-level political risk are more visible
during election years due to higher levels of political uncertainty Furthermore, we apply several alternative econometric approaches
(Marshall et al., 2018). However, we observe that the modification to alleviate endogeneity concerns regarding potential omitted variables
impacts of Democratic presidencies on political exposure and risk are that simultaneously impact both firm-level political risk and intellectual
primarily insignificant. capital investment in Table 12. First, we confirm our findings and
address the model's dynamic by employing the two-step system

10
N. Huynh et al. International Review of Financial Analysis 91 (2024) 103020

Table 7
Cross-sectional heterogeneity: managerial ability, firm-level political risk, and intellectual capital investment.
Panel A: Political Exposure.

Firm Regression results Economic significance

characteristics
Political Exposure × Political Exposure × β1 − β2 High managerial ability Low managerial ability High- %
MA_HIGH (β1 ) MA_LOW (β2 ) firms firms Low

High FD − 0.293*** − 0.412*** 0.119***

− 2.409 − 3.299 0.891 − 27.00%
(0.029) (0.011) (0.007)
Low FD − 0.166** − 0.213*** 0.047**
− 1.789 − 2.350 0.561 − 23.88%
(0.075) (0.049) (0.014)
High EFD − 0.367*** − 0.501*** 0.140***
− 2.529 − 4.124 1.595 − 38.68%
(0.010) (0.003) (0.000)
Low EFD − 0.185*** − 0.234*** 0.049**
− 1.741 − 2.240 0.499 − 22.27%
(0.059) (0.026) (0.027)
High IO − 0.311*** − 0.457*** 0.145***
− 2.404 − 3.172 0.768 − 24.20%
(0.018) (0.007) (0.000)
Low IO − 0.181** − 0.272*** 0.091***
− 1.994 − 2.443 0.449 − 18.37%
(0.103) (0.032) (0.006)
High AC − 0.249*** − 0.382*** 0.133**
− 2.165 − 3.044 0.879 − 28.88%
(0.025) (0.013) (0.034)
Low AC − 0.209*** − 0.355*** 0.146***
− 2.092 − 2.901 0.809 − 27.89%
(0.036) (0.015) (0.004)

Panel B: Political Risk

Firm Regression results Economic significance

characteristics
Political Risk × MA_HIGH Political Risk × MA_LOW β1 - β2 High managerial ability Low managerial ability High- %
(β1 ) (β2 ) firms firms Low

High FD − 0.320*** − 0.459*** 0.139**

− 3.050 − 4.470 1.420 − 31.76%
(0.006) (0.000) (0.016)
Low FD − 0.107** − 0.266*** 0.159***
− 1.987 − 2.582 0.595 − 23.04%
(0.052) (0.015) (0.007)
High EFD − 0.403*** − 0.542*** 0.140**
− 3.307 − 5.271 1.964 − 37.25%
(0.000) (0.000) (0.042)
Low EFD − 0.214*** − 0.299*** 0.085*
− 1.862 − 2.609 0.747 − 28.63%
(0.014) (0.011) (0.098)
High IO − 0.414*** − 0.622*** 0.208***
− 2.993 − 4.561 1.567 − 34.37%
(0.000) (0.000) (0.001)
Low IO − 0.148** − 0.294*** 0.146**
− 2.168 − 2.846 0.678 − 23.81%
(0.075) (0.020) (0.045)
High AC − 0.227*** − 0.375*** 0.148**
− 2.681 − 3.601 0.920 − 25.56%
(0.003) (0.001) (0.041)
Low AC − 0.246*** − 0.359*** 0.113*
− 2.755 − 3.583 0.828 − 23.11%
(0.002) (0.001) (0.087)

This table presents the results of OLS regression models that investigate the effect of managerial ability (MA) on the relationship between firm-level political risk and
corporate intellectual capital investment (ICI) by considering the cross-sectional heterogeneity and using Eq. (4). The sample consists of 3668 U.S firms between 2002
and 2021. The firm-level political risk, proxied by political exposure and political risk, is collected from Hassan et al. (2019). The intellectual capital investment is
computed using the approach of Sydler et al. (2014). The Dummy of MA_HIGH (MA_LOW) is created using Demerjian's managerial ability score higher (lower) than the
average score of all firms over the sample period or equal zero otherwise, which interacted with the main indicators of firm-level political risks. All firms are grouped
into high and low Financial distress (KZ index), External financial dependence (EFD), Institutional ownership (IO), and Analyst Coverage (AC). We also report the
differences from regression with the Wald test for the difference between β1 and β2. Panel A and B report the results for Political Exposure and Political Risk, respectively.
We control for industry and time-fixed effects in all specifications. Standard errors are clustered at the firm level and are presented in parentheses under the associated
coefficients. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively. The descriptions of all variables are reported in Table 1.

Generalized Method of Moments (S-GMM) estimations (Blundell & the Oster (2019). Overall, our results support that the inferences from
Bond, 1998). Prior studies confirm that the S-GMM approach can our models are highly unlikely to suffer from an omitted variable bias.
partially reduce the potential heterogeneity and omitted variable bias In Panel B, we further another endogenous-treatment method using
(Trinh, Aljughaiman, & Cao, 2020). In Panel A of Table 12, the findings the Two-Stages Least Square/Instrumental Variable Analysis (2SLS) to
are consistent with our main results that political risk and exposure address the possible endogeneity of firm-level political risk measure­
negatively and significantly impact firms' intellectual capital invest­ ment error. By following prior studies (Ahmad et al., 2023; Azzimonti,
ment. The subsequent identification test is a Placebo test by replacing 2018), we utilize the Partisan Conflict Index as our instrumental variable
the political risk measures of a given firm in a given year with a (IV).17 The Partisan Conflict Index is the level of political polarization or
randomly drawn value from the sample and then re-estimate the base­ disagreement among politicians in the U.S, which is obtained from the
line regression (Eq. 3). Overall, we can further reinforce the robustness Federal Reserve Bank of Philadelphia. This index can impact the firm-
of our baseline results as the estimated coefficients of the independent
variable are statistically insignificant. In other words, this tests lend
further support to reject potential endogeneity issues vis-à-vis the like­ 17
The Partisan Conflict Index has been extensively used as an instrumental
lihood of having time-varying omitted variables that can instanta­
variable in the economic policy and political uncertainty literature, such as by
neously modify both firms' IC investment and their specific political risk. Gulen and Ion (2016); D'Mello and Toscano (2020); Pan et al. (2019); among
In Appendix C, we also address the omitted variables bias by employing others.

11
N. Huynh et al. International Review of Financial Analysis 91 (2024) 103020

Table 8 Table 9
Heterogenous impacts of political risks across high-tech and non-high-tech Robustness checks: Alternative measures for intellectual capital investment and
firms. political risk.
Panel A: Political Exposure Dependent variable: A-ICI VAIC
Alternative IC
Firm Regression results (1) (2) (3) (4)
characteristics
Political Political Political β1 − β2 Political Exposure − 0.189*** − 0.587**
Exposure Exposure × Exposure × (Wald (0.016) (0.045)
(α1 ) MA_HIGH (β1 ) MA_LOW (β2 ) test) Political Risk − 0.224** − 0.792***
(0.008) (0.007)
High-tech firms − 0.299*** − 0.189** − 0.329*** 0.140***
Controls Yes Yes Yes Yes
(0.006) (0.018) (0.008) (0.022)
S.E. clustered by Firm Yes Yes Yes Yes
Non-high-tech
− 0.155** − 0.196*** − 0.316*** 0.120** Firm/Time/Industry-fixed
firms Yes Yes Yes Yes
effects
(0.025) (0.015) (0.010) (0.026)
Observation 29,504 29,504 23,603 23,603
Wald test on α1
0.000 Adjusted R-squared 0.401 0.539 0.652 0.617
(p-value)
This table presents the results of OLS regression models investigating the impact
of firm-level political risk and corporate intellectual capital investment (ICI).
Economic significance The sample consists of 3668 U.S firms between 2002 and 2021. The firm-level
political risk, proxied by political exposure and political risk, is collected from
Firm High managerial Low managerial High- % Hassan et al. (2019). In Panel A, the intellectual capital investment is computed
characteristics ability firms ability firms Low
by using a modification from the approach of Sydler et al. (2014) and VAIC
High-tech firms − 2.107 − 3.128 1.021 − 32.63% approach (Public, 2002, 2004). In Panel B, we further control for two econo­
Non-high-tech mywide political uncertainty indices of Geopolitical Threats (GT) and Geopo­
− 2.156 − 3.097 0.941 − 30.38%
firms litical Risk Index (GRI) developed by Caldara and Iacoviello (2022). We control
for industry, firm, and time-fixed effects in all specifications. Standard errors are
clustered at the firm level and are presented in parentheses under the associated
Panel B: Political Risk coefficients. ***, **, and * indicate significance at the 1%, 5%, and 10% levels,
respectively. The descriptions of all variables are reported in Table 1.
Firm Regression results
characteristics
Political Political Risk Political Risk β1 − β2
Risk × MA_HIGH × MA_LOW (Wald test) level political risks but is not probable to impact firms' IC investment. In
(β1 ) (β2 ) the first stage, we estimate firm-level political risk and exposure using
High-tech firms − 0.355*** − 0.204*** − 0.387*** 0.183***
the Partisan Conflict Index. The estimated coefficients are positive and
(0.002) (0.014) (0.008) (0.007) statistically significant, indicating that higher political conflicts give rise
Non-high-tech to higher levels of firm-level political risk. Also, the results for conven­
− 0.195*** − 0.212*** − 0.403*** 0.191***
firms tional tests (F-test, test of under-identification, and weak instruments)
(0.009) (0.012) (0.006) (0.006)
further confirm the Partisan Conflict Index as an appropriate IV. In the
Wald test on α1
(p-value)
0.000 second stage, we estimate the results using instrumented firm-level po­
litical risk/exposure as the explanatory variables. We consistently obtain
the negative and sizeable influence of firm-level political risk and
Economic significance exposure on intellectual capital investment. Generally, the results are
consistent with our baseline findings when using IV-2SLS to control for
Firm High managerial Low managerial High- %
characteristics ability firms ability firms Low
the endogeneity concerns.

High-tech firms − 2.347 − 3.352 1.005 − 29.99%

6. Conclusion
Non-high-tech
− 2.438 − 3.497 1.059 − 30.29%
firms
This paper intends to enrich the understanding of how firm-level
This table presents the results of OLS regression models that investigate the ef­
exposure to political uncertainty affects corporate investment in intel­
fect of managerial ability (MA) on the relationship between firm-level political
lectual capital and the role of managerial ability in this relationship. We
risk and corporate intellectual capital investment (ICI) by considering the het­
erogeneity between high-tech and non-high-tech firms and using Eq. (4). The
investigate these effects using a U.S sample of 29,504 firm-year obser­
sample consists of 3668 U.S firms between 2002 and 2021. The firm-level po­ vations from 2002 to 2021. Our results confirm that firms' political risk
litical risk, proxied by political exposure and political risk, is collected from is a destructive determining factor of corporate intellectual capital in­
Hassan et al. (2019). The intellectual capital investment is computed using the vestment. The extent of this adverse effect depends on firms' financial
approach of Sydler et al. (2014). The Dummy of MA_HIGH (MA_LOW) is created distress, external finance dependence, and institutional holdings.
using Demerjian's managerial ability score higher (lower) than the average score Further, the intellectual capital investments of high-tech firms are more
of all firms over the sample period or equal zero otherwise, which interacted sensitive to their political risk due to the nature of their businesses. Our
with the main indicators of firm-level political risks. All firms are grouped into results also confirm that the adverse effect of firms' political risk on
high-tech and non-high-tech firms using the classification of Loughran and Ritter intellectual capital investment is significantly driven by managerial
(2004). We also report the differences from regression with the Wald test for the
ability, where firms with lower managerial ability scores are concerned
difference between α1 of two subgroups and β1 and β2. Panel A and B report the
more than those with higher scores. Regarding the economic magnitude,
results for Political Exposure and Political Risk, respectively. We control for in­
dustry and time-fixed effects in all specifications. Standard errors are clustered a higher level of managerial ability facilitates firms to reduce a sub­
at the firm level and are presented in parentheses under the associated co­ stantial proportion (approximately 20% to 40%) of the political risk-
efficients. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, intellectual capital investment nexus. Further, we find that moderated
respectively. The descriptions of all variables are reported in Table 1. role of managerial ability is more significant for firms with higher
financial distress and external finance dependence and lower institu­
tional ownership and analyst coverage. Our findings remain unchanged
under several robustness checks and sensitivity analyses to control the
endogeneity issue. Overall, this study proposes valuable implications for

12
N. Huynh et al. International Review of Financial Analysis 91 (2024) 103020

Table 10
High versus low managerial ability with the alternative intellectual capital measure.
Panel A: High versus low managerial ability

Dependent variable: A-ICI Political Exposure Dependent variable: A-ICI Political Risk

(1) (2)

Political Exposure − 0.154*** Political Risk − 0.198***

(0.011) (0.001)
Political Exposure × MA_HIGH (β1 ) − 0.097** Political Risk × MA_HIGH (β1 ) − 0.171***
(0.066) (0.011)
Political Exposure × MA_LOW (β2 ) − 0.176*** Political Risk × MA_LOW (β2 ) − 0.252***
(0.018) (0.006)
Controls Yes Controls Yes
S.E. clustered by Firm Yes S.E. clustered by Firm Yes
Firm/Time/Industry-fixed effects Yes Firm/Time/Industry-fixed effects Yes
Observation 29,504 Observation 29,504
Adjusted R-squared 0.442 Adjusted R-squared 0.590
β1 − β2 (Wald test) 0.079** β1 − β2 (Wald test) 0.081**
(p-value) 0.037 (p-value) 0.022

Panel B: Economic significance

Dependent variable: A-ICI All firms Low managerial ability firms High managerial ability firms

(1) (2) (3)

Political Exposure − 2.643 − 3.000 − 2.339

Political Risk − 3.165 − 3.687 − 2.690
Difference (3)–(2) (3)–(1)
Political Exposure 0.661
% − 22.03%
Political Risk 0.997
% − 27.04%

This table presents the results of OLS regression models that investigate the effect of managerial ability (MA) on the relationship between firm-level political risk and
corporate intellectual capital investment (ICI) by using Eq. (4). The sample consists of 3668 U.S firms between 2002 and 2021. The firm-level political risk, proxied by
political exposure and political risk, is collected from Hassan et al. (2019). The intellectual capital investment is computed using a modification from the approach of
Sydler et al. (2014). The Dummy of MA_HIGH (MA_LOW) is created by using Demerjian's managerial ability score higher (lower) than the average score of all firms
over the sample period or equals zero otherwise, which are interacted with the main indicators of firm-level political risks. Panel A reports the differences from
regression with the Wald test for the difference between β1 and β2. Panel B economic significance of one standard deviation change in the political risk on IC for firms
with low and high managerial ability, and its percentage of differences. We control for industry, firm, and time-fixed effects in all specifications. Standard errors are
clustered at the firm level and are presented in parentheses under the associated coefficients. ***, **, and * indicate significance at the 1%, 5%, and 10% levels,
respectively. The descriptions of all variables are reported in Table 1.

Table 11
The impacts of U.S presidencies and elections.
Dependent variable: ICI Political Exposure Dependent variable: ICI Political Risk

(1) (2) (1) (2)

Political Exposure − 0.133*** − 0.115*** Political Risk − 0.114*** − 0.122***

(0.011) (0.012) (0.011) (0.010)
Election 0.264 Election 0.189
(0.067) (0.051)
Political Exposure × Election − 0.076** Political Risk × Election − 0.102***
(0.012) (0.010)
Democratic − 0.027 Democratic − 0.032
(0.018) (0.015)
Political Exposure × Democratic 0.088 Political Risk × Democratic 0.065
(0.025) (0.026)
Controls Yes Yes Controls Yes Yes
S.E. clustered by Firm Yes Yes S.E. clustered by Firm Yes Yes
Firm/Time/Industry-fixed effects Yes Yes Firm/Time/Industry-fixed effects Yes Yes
Observation 29,504 29,504 Observation 29,504 29,504
Adjusted R-squared 0.396 0.365 Adjusted R-squared 0.506 0.348

This table presents the results of OLS regression models investigating the impact of firm-level political risk and corporate intellectual capital investment (ICI) by
considering the impacts of U.S presidencies and elections. The sample consists of 3668 U.S firms between 2002 and 2021. The firm-level political risk, proxied by
political exposure and political risk, is collected from Hassan et al. (2019). The intellectual capital investment is computed using the approach of Sydler et al. (2014).
The Dummy factors are identified by the U.S presidencies (Democratic versus Republican) and Election years, which are interacted with the main indicators of firm-
level political risks. We control for industry, firm, and time-fixed effects in all specifications. Standard errors are clustered at the firm level and are presented in
parentheses under the associated coefficients. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively. The descriptions of all variables are
reported in Table 1.

13
N. Huynh et al. International Review of Financial Analysis 91 (2024) 103020

Table 12
Alternative econometric approaches.
Panel A: Alternative models

Variables Two-step S-GMM Placebo

(1) (2) (3) (4)

Political Exposure − 0.144*** − 0.073

(0.012) (0.940)
Political Risk − 0.211*** − 0.037
(0.007) (0.241)
Controls Yes Yes Yes Yes
S.E. clustered by Firm Yes Yes Yes Yes
Firm/Time/Industry-fixed effects Yes Yes Yes Yes
Observation 29,504 29,504 29,504 29,504
Adjusted R-squared 0.498 0.669 0.407 0.385

Panel B: Alternative models: 2SLS

Variables Political Exposure Political Risk

First-stage Second-stage First-stage Second-stage

Partisan Conflict Index 2.201** 3.456**

(0.022) (0.011)
Instrumented Political Exposure − 0.198**
(0.029)
Instrumented Political Risk − 0.304***
(0.000)
Joint test of excluded Instruments F-stat = 21.22*** F-stat = 18.44***
Test of under-identification 24.43*** 19.75***
Test of weak instruments 23.77 25.33
Controls Yes Yes Yes Yes
S.E. clustered by Firm Yes Yes Yes Yes
Firm/Time/Industry-fixed effects Yes Yes Yes Yes
Observation 29,504 29,504 29,504 29,504
Adjusted R-squared 0.548 0.562 0.573 0.517

This table presents the results of alternative approaches investigating the impact of firm-level political risk and corporate intellectual capital investment (ICI). The
sample consists of 3668 U.S firms between 2002 and 2021. The firm-level political risk, proxied by political exposure and political risk, is collected from Hassan et al.
(2019). The intellectual capital investment is computed using the approach of Sydler et al. (2014). Panel A reports the results from the two-step System Generalized
Method of Moments and Placebo analyses. Panel B reports the results from Two-Stages Least Square/Instrumental Variable (IV-2SLS) analysis with Partisan Conflict
Index as an IV. We control for industry, firm, and time-fixed effects in all specifications. Standard errors are clustered at the firm level and are presented in parentheses
under the associated coefficients. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively. The descriptions of all variables are reported in
Table 1.

academics, practitioners, stakeholders, and regulators in mitigating the Acknowledgements

political risk-induced capital investment possibility. As a final point, our
study certainly reveals limitations that could be an open door for future We would like to thank the Editor-in-chief - Professor Brian Lucey
research. Further study can focus on the impacts of geopolitical risks on and two anonymous reviewers for their constructive suggestions, which
specific outputs from IC investment, such as interpersonal, knowledge, greatly improve the scientific level of this paper. The authors are
and human factors, with more advanced measures. Further, a cross- indebted to Anh Pham, Carlos Burga, Lee Eun Kyung, and Kyle Paquette
country study with international data for firm-level political risk can for their valuable comments and suggestions. Quang Thien Tran also
provide miscellaneous standpoints of different institutional would like to acknowledge the financial support from Van Lang Uni­
backgrounds. versity, Ho Chi Minh City, Vietnam. The usual caveat applies.

Data availability

The authors do not have permission to share data.

Appendix A. Appendix

A.1. Correlation Matrix

14
 N. Huynh et al.
Table A1
Pearson correlation matrix.
Variables

(1) Intellectual c1apital investment 1.000 – – – – – – – – – – – – –

(2) Political Risk − 0.513*** 1.000 – – – – – – – – – – – –
(3) Political Exposure − 0.448*** 0.597*** 1.000 – – – – – – – – – – –
(4) Firm Size − 0.160** 0.027 0.024 1.000 – – – – – – – – – –
(5) Leverage − 0.105** 0.045* 0.036* 0.231*** 1.000 – – – – – – – – –
(6) Profitability 0.370*** − 0.374*** − 0.325*** 0.187** − 0.062** 1.000 – – – – – – – –
(7) Cash holdings 0.136** − 0.347*** − 0.413*** 0.401*** − 0.027 0.031 1.000 – – – – – – –
(8) Sales growth 0.483*** − 0.178*** − 0.158** 0.116** − 0.011 0.107** 0.009 1.000 – – – – – –
(9) Stock return 0.326*** 0.018 0.014 0.018 0.009 0.374*** 0.045* 0.285*** 1.000 – – – – –
(10) Institutional ownership − 0.079 0.009 0.007 0.223*** 0.053 0.231*** 0.080* 0.059 0.036 1.000 – – – –
15

(11) Board Size 0.138* 0.018 0.016 0.320*** 0.036 0.009 − 0.030 − 0.011 0.134* 0.303*** 1.000 – – –
(12) CEO Duality − 0.109* − 0.027 − 0.020 0.062* 0.059* 0.036 − 0.043 0.017 0.036 0.134** 0.223*** 1.000 – –
(13) MA_HIGH 0.235** 0.039 0.047 0.027 0.015 0.490*** 0.099* 0.250*** 0.117** 0.036 − 0.074* 0.045 1.000 –
(14) MA_LOW − 0.235** − 0.039 − 0.047 − 0.027 − 0.015 − 0.490*** − 0.099* − 0.250*** − 0.117** − 0.036 0.074* − 0.045 − 1.000*** 1.000

This table reports the Pearson correlations among the variables employed in this study. All continuous variables are winsorized at the 1st and 99th percentiles. ***, **, and * indicate significance at the 1%, 5%, and 10%
levels, respectively. The descriptions of all variables are reported in Table 1.

International Review of Financial Analysis 91 (2024) 103020

N. Huynh et al. International Review of Financial Analysis 91 (2024) 103020

Appendix B. Additional tests

Table B1
Firm-level political risk components and intellectual capital investment.

Dependent variable: ICI Political Risk

Economic Policy Environment Trade Institutions Health Security & Defense Tax Policy Technology

(1) (2) (3) (4) (5) (6) (7) (8)

Political Risk − 0.323*** − 0.118* − 0.039 − 0.079 − 0.017 − 0.040 − 0.287*** − 0.056
(0.000) (0.010) (0.204) (0.157) (0.450) (0.391) (0.000) (0.210)
Controls Yes Yes Yes Yes Yes Yes Yes Yes
S.E. clustered by Firm Yes Yes Yes Yes Yes Yes Yes Yes
Firm/Time/Industry -fixed effect Yes Yes Yes Yes Yes Yes Yes Yes
Observation 29,504 29,504 29,504 29,504 29,504 29,504 29,504 29,504
Adjusted R-squared 0.406 0.441 0.350 0.470 0.569 0.538 0.676 0.447
This table presents the results of OLS regression models investigating the impact of eight components of firm-level political risk and corporate intellectual capital
investment (ICI). The sample consists of 3668 U.S firms between 2002 and 2021. The firm-level political risk indicators are collected from Hassan et al. (2019). The
intellectual capital investment is computed using the approach of Sydler et al. (2014). We control for industry, firm, and time-fixed effects in all specifications.
Standard errors are clustered at the firm level and are presented in parentheses under the associated coefficients. ***, **, and * indicate significance at the 1%, 5%, and
10% levels, respectively. The descriptions of all variables are reported in Table 1.

Table B2
Sectoral analyses: Firm-level political risk and intellectual capital investment

Sectors Political Exposure Political Risk

Coefficient p-value Coefficient p-value

Energy − 0.128** 0.038 − 0.112** 0.042

Consumer discretionary − 0.114** 0.044 − 0.199*** 0.006
Consumer staples − 0.104** 0.048 − 0.124** 0.037
Communication services − 0.277*** 0.000 − 0.322*** 0.000
Materials − 0.097** 0.049 − 0.111** 0.042
Information technology − 0.220*** 0.001 − 0.295*** 0.001
Industrials − 0.141*** 0.007 − 0.136** 0.026
Health care − 0.184*** 0.003 − 0.191*** 0.005
This table presents the results of OLS regression models investigating the impact of firm-level political risk and corporate intellectual
capital investment (ICI) for eight sectoral panels. The sample consists of 3668 U.S firms between 2002 and 2021. The firm-level
political risk, proxied by political exposure and political risk, is collected from Hassan et al. (2019). The intellectual capital invest­
ment is computed using the approach of Sydler et al. (2014). We control for firm and time-fixed effects in all specifications. Standard
errors are clustered at the firm level and are presented in parentheses under the associated coefficients. ***, **, and * indicate sig­
nificance at the 1%, 5%, and 10% levels, respectively. The descriptions of all variables are reported in Table 1.

Appendix C. Oster (2019) tests for omitted variable bias

In this study, we employ the statistical test proposed by Oster (2019) to handle the potential omitted variable bias in our models. Oster (2019)
assumes that the power of coefficients linked to the R-squares from regressions with and without controls can be preserved to generate an identifiable
set. We exploit the Mian and Sufi (2014) assumptions of Oster (2019) to fabricate the lower and upper bounds of the identified set, with δ = 1 and Rmax
= min (2.2R̃,1) and the extreme ones from Oster's study of δ =1 and RMAX =1.

Table C1
Omitted variable bias – Oster (2019).

Oster Condition Dependent Variables Variable of interest Lower Bound Upper Bound Includes Zero?

Assume t = 1; RMAX = min(2.2R̃,1) Intellectual capital investment (ICI) Political Exposure 0.0362 0.0758 No
Assume t = 1; RMAX = min(2.2R̃,1) Intellectual capital investment (ICI) Political Risk 0.0408 0.0851 No
Assume t = 1; RMAX = 1 Intellectual capital investment (ICI) Political Exposure 0.0192 0.0426 No
Assume t = 1; RMAX = 1 Intellectual capital investment (ICI) Political Risk 0.0149 0.0391 No

References Alvino, F., Di Vaio, A., Hassan, R., & Palladino, R. (2021). Intellectual capital and
sustainable development: A systematic literature review. Journal of Intellectual
Capital, 22(1), 76–94.
Acemoglu, D., & Robinson, J. A. (2001). A theory of political transitions. American
Andreou, P. C., Karasamani, I., Louca, C., & Ehrlich, D. (2017). The impact of managerial
Economic Review, 91(4), 938–963.
ability on crisis-period corporate investment. Journal of Business Research, 79,
Ahmad, M. F., Aziz, S., El-Khatib, R., & Kowalewski, O. (2023). Firm-level political risk
107–122.
and dividend payout. International Review of Financial Analysis, 86, Article 102546.
Azzimonti, M. (2018). Partisan conflict and private investment. Journal of Monetary
Alam, A. W., Houston, R., & Farjana, A. (2023). Geopolitical risk and corporate
Economics, 93, 114–131.
investment: How do politically connected firms respond? Finance Research Letters,
Baker, M., Stein, J. C., & Wurgler, J. (2003). When does the market matter? Stock prices
53, Article 103681.
and the investment of equity-dependent firms. The Quarterly Journal of Economics,
Altman, E. I. (1968). Financial Ratios, Discriminant Analysis and the Prediction of
118(3), 969–1005.
Corporate Bankruptcy. The Journal of Finance, 23(4), 589–609.

16
N. Huynh et al. International Review of Financial Analysis 91 (2024) 103020

Ballester, M., Livnat, J., & Sinha, N. (2002). Tracks. Journal of Accounting, Auditing and Gyimah, D., Danso, A., Adu-Ameyaw, E., & Boateng, A. (2022). Firm-level political risk
Finance, 17(4), 351–373. and corporate leverage decisions. International Review of Financial Analysis, 84,
Banker, R. D., Darrough, M. N., Huang, R., & Plehn-Dujowich, J. M. (2013). The relation Article 102354.
between CEO compensation and past performance. The Accounting Review, 88(1), Handley, K., & Limao, N. (2015). Trade and investment under policy uncertainty: Theory
1–30. and firm evidence. American Economic Journal: Economic Policy, 7(4), 189–222.
Bekaert, G., Harvey, C. R., Lundblad, C. T., & Siegel, S. (2016). Political risk and Hasan, S. B., Alam, M. S., Paramati, S. R., & Islam, M. S. (2022). Does firm-level political
international valuation. Journal of Corporate Finance, 37, 1–23. risk affect cash holdings? Review of Quantitative Finance and Accounting, 59(1),
Bertrand, M., & Schoar, A. (2003). Managing with style: The effect of managers on firm 311–337.
policies. The Quarterly Journal of Economics, 118(4), 1169–1208. Hassan, T. A., Hollander, S., Van Lent, L., & Tahoun, A. (2019). Firm-level political risk:
Blinder, A. S., & Watson, M. W. (2016). Presidents and the US economy: An econometric Measurement and effects. The Quarterly Journal of Economics, 134(4), 2135–2202.
exploration. American Economic Review, 106(4), 1015–1045. Hoang, K., & Tran, T. T. (2022). Policy uncertainty and intellectual capital investment.
Blundell, R., & Bond, S. (1998). Initial conditions and moment restrictions in dynamic Applied Economics Letters, 29(15), 1369–1377.
panel data models. Journal of Econometrics, 87(1), 115–143. Hu, S., & Gong, D. (2019). Economic policy uncertainty, prudential regulation and bank
Boekestein, B. (2006). The relation between intellectual capital and intangible assets of lending. Finance Research Letters, 29, 373–378.
pharmaceutical companies. Journal of Intellectual Capital, 7(2), 241–253. Huang, Q., & Yuan, T. (2021). Does political corruption impede firm innovation?
Brogaard, J., & Detzel, A. (2015). The asset-pricing implications of government economic Evidence from the United States. Journal of Financial and Quantitative Analysis, 56(1),
policy uncertainty. Management Science, 61(1), 3–18. 213–248.
Caldara, D., & Iacoviello, M. (2022). Measuring geopolitical risk. American Economic Hussi, T. (2004). Reconfiguring knowledge management–combining intellectual capital,
Review, 112(4), 1194–1225. intangible assets and knowledge creation. Journal of Knowledge Management, 8(2),
Chang, X., Dasgupta, S., & Hilary, G. (2006). Analyst coverage and financing decisions. 36–52.
The Journal of Finance, 61(6), 3009–3048. Huynh, N. (2023). Unemployment beta and the cross-section of stock returns: Evidence
Chang, Y. Y., Dasgupta, S., & Hilary, G. (2010). CEO ability, pay, and firm performance. from Australia. International Review of Financial Analysis, 102522.
Management Science, 56(10), 1633–1652. Huynh, N., Dao, A., & Nguyen, D. (2021). Openness, economic uncertainty, government
Chemmanur, T. J., & Paeglis, I. (2005). Management quality, certification, and initial responses, and international financial market performance during the coronavirus
public offerings. Journal of Financial Economics, 76(2), 331–368. pandemic. Journal of Behavioral and Experimental Finance, 31, 100536.
Chemmanur, T. J., Paeglis, I., & Simonyan, K. (2009). Management quality, financial and Inkinen, H., Kianto, A., Vanhala, M., & Ritala, P. (2017). Structure of intellectual
investment policies, and asymmetric information. Journal of Financial and capital–an international comparison. Accounting, Auditing & Accountability Journal,
Quantitative Analysis, 44(5), 1045–1079. 30(5), 1160–1183.
Chemmanur, T. J., Paeglis, I., & Simonyan, K. (2010). Management quality and equity Jens, C. E. (2017). Political uncertainty and investment: Causal evidence from US
issue characteristics: A comparison of SEOs and IPOs. Financial Management, 39(4), gubernatorial elections. Journal of Financial Economics, 124(3), 563–579.
1601–1642. Joshi, M., Cahill, D., Sidhu, J., & Kansal, M. (2013). Intellectual capital and financial
Chen, S. S., & Wang, Y. (2012). Financial constraints and share repurchases. Journal of performance: An evaluation of the Australian financial sector. Journal of Intellectual
Financial Economics, 105(2), 311–331. Capital, 14(2), 264–285.
Chen, T. K., Tseng, Y., & Hsieh, Y. T. (2015). Real earnings management uncertainty and Julio, B., & Yook, Y. (2012). Political uncertainty and corporate investment cycles. The
corporate credit risk. The European Accounting Review, 24(3), 413–440. Journal of Finance, 67(1), 45–83.
Chen, Z., Cihan, M., Jens, C. E., & Page, T. B. (2023). Political uncertainty and firm Keillor, B. D., Wilkinson, T. J., & Owens, D. (2005). Threats to international operations:
investment: Project-level evidence from M&a activity. Journal of Financial and Dealing with political risk at the firm level. Journal of Business Research, 58(5),
Quantitative Analysis, 58(1), 71–103. 629–635.
Cheung, K. T. S., Naidu, D., Navissi, F., & Ranjeeni, K. (2017). Valuing talent: Do CEOs’ King, T., Loncan, T., & Khan, Z. (2021). Investment, leverage and political risk: Evidence
ability and discretion unambiguously increase firm performance. Journal of from project-level FDI. Journal of Corporate Finance, 67, Article 101873.
Corporate Finance, 42, 15–35. Kor, Y. Y. (2003). Experience-based top management team competence and sustained
Choi, W., Chung, C. Y., & Wang, K. (2022). Firm-level political risk and corporate growth. Organization Science, 14(6), 707–719.
investment. Finance Research Letters, 46, Article 102307. Kumar, S., & Zbib, L. (2022). Firm performance during the Covid-19 crisis: Does
Citroen, C. L. (2011). The role of information in strategic decision-making. International managerial ability matter? Finance Research Letters, 47, Article 102720.
Journal of Information Management, 31(6), 493–501. Kweh, Q. L., Lu, W. M., Tone, K., & Nourani, M. (2022). Risk-adjusted banks’ resource-
Çolak, G., Durnev, A., & Qian, Y. (2017). Political uncertainty and IPO activity: Evidence utilization and investment efficiencies: Does intellectual capital matter? Journal of
from US gubernatorial elections. Journal of Financial and Quantitative Analysis, 52(6), Intellectual Capital, 23(3), 687–712.
2523–2564. Le, A. T., & Tran, T. P. (2021). Does geopolitical risk matter for corporate investment?
Dai, L., & Ngo, P. (2021). Political uncertainty and accounting conservatism. The Evidence from emerging countries in Asia. Journal of Multinational Financial
European Accounting Review, 30(2), 277–307. Management, 62, Article 100703.
D'Amato, A. (2021). Does intellectual capital impact firms' capital structure? Exploring Lee, C. C., Wang, C. W., Chiu, W. C., & Tien, T. S. (2018). Managerial ability and
the role of firm risk and profitability. Managerial Finance, 47(9), 1337–1356. corporate investment opportunity. International Review of Financial Analysis, 57,
Demerjian, P., Lev, B., & McVay, S. (2012). Quantifying managerial ability: A new 65–76.
measure and validity tests. Management Science, 58(7), 1229–1248. Lerro, A., Linzalone, R., & Schiuma, G. (2014). Managing intellectual capital dimensions
Demerjian, P. R., Lev, B., Lewis, M. F., & McVay, S. E. (2013). Managerial ability and for organizational value creation. Journal of Intellectual Capital, 15(3), 350–361.
earnings quality. The Accounting Review, 88(2), 463–498. Lewellyn, K. B., & Muller-Kahle, M. I. (2012). CEO power and risk taking: Evidence from
Dierickx, I., & Cool, K. (1989). Asset stock accumulation and sustainability of the subprime lending industry. Corporate Governance: An International Review, 20(3),
competitive advantage. Management Science, 35(12), 1504–1511. 289–307.
Dixit, R. K., & Pindyck, R. S. (2012). Investment under uncertainty. Princeton University Liu, Y., Jin, J., Zhang, Z., & Zhao, R. (2022). Firm-level political sentiment and corporate
Press. tax avoidance. International Review of Financial Analysis, 84, Article 102358.
D'Mello, R., & Toscano, F. (2020). Economic policy uncertainty and short-term financing: Loughran, T., & Ritter, J. (2004). Why has IPO underpricing changed over time? Financial
The case of trade credit. Journal of Corporate Finance, 64, Article 101686. Management, 5–37.
Doukas, J. A., & Zhang, R. (2021). Managerial ability, corporate social culture, and Maditinos, D., Chatzoudes, D., Tsairidis, C., & Theriou, G. (2011). The impact of
M&as. Journal of Corporate Finance, 68, Article 101942. intellectual capital on firms' market value and financial performance. Journal of
Duong, H. N., Nguyen, J. H., Nguyen, M., & Rhee, S. G. (2020). Navigating through Intellectual Capital, 12(1), 132–151.
economic policy uncertainty: The role of corporate cash holdings. Journal of Marcelin, I., Stephen, S. A. K., Fanta, F., & Tecklezion, M. (2019). Political regimes,
Corporate Finance, 62, Article 101607. investment and electoral uncertainty. Research in International Business and Finance,
Ellis, J., Smith, J., & White, R. (2020). Corruption and corporate innovation. Journal of 47, 580–599.
Financial and Quantitative Analysis, 55(7), 2124–2149. Marshall, B. R., Nguyen, H. T., Nguyen, N. H., & Visaltanachoti, N. (2018). Politics and
Faulkender, M., & Wang, R. (2006). Corporate financial policy and the value of cash. The liquidity. Journal of Financial Markets, 38, 1–13.
Journal of Finance, 61(4), 1957–1990. Mian, A., & Sufi, A. (2014). What explains the 2007–2009 drop in employment?
Ferreira, M. A., Matos, P., Pereira, J. P., & Pires, P. (2017). Do locals know better? A Econometrica, 82(6), 2197–2223.
comparison of the performance of local and foreign institutional investors. Journal of Mishra, D. R. (2023). Firm-level political risk and implied cost of equity capital.
Banking & Finance, 82, 151–164. International Review of Finance. https://doi.org/10.1111/irfi.12411
Gan, H. (2019). Does CEO managerial ability matter? Evidence from corporate Mouritsen, J., & Larsen, H. T. (2005). The 2nd wave of knowledge management: The
investment efficiency. Review of Quantitative Finance and Accounting, 52(4), management control of knowledge resources through intellectual capital
1085–1118. information. Management Accounting Research, 16(3), 371–394.
Ginesti, G., Caldarelli, A., & Zampella, A. (2018). Exploring the impact of intellectual Nadeem, M., Zaman, R., Suleman, T., & Atawnah, N. (2021). CEO ability, career
capital on company reputation and performance. Journal of Intellectual Capital, 19(5), concerns, firms’ lifecycle and investments in intellectual capital. International Review
915–934. of Economics and Finance, 75, 237–251.
Guan, J. X., Li, O. Z., & Ma, J. (2018). Managerial ability and the shareholder tax Naidenova, I. (2022). Economic policy uncertainty and company's human capital.
sensitivity of dividends. Journal of Financial and Quantitative Analysis, 53(1), Journal of Economic Studies, 49(5), 902–919.
335–364. Oliveira, L., Lima Rodrigues, L., & Craig, R. (2010). Intellectual capital reporting in
Gulen, H., & Ion, M. (2016). Policy uncertainty and corporate investment. The Review of sustainability reports. Journal of Intellectual Capital, 11(4), 575–594.
Financial Studies, 29(3), 523–564.

17
N. Huynh et al. International Review of Financial Analysis 91 (2024) 103020

Osinski, M., Selig, P. M., Matos, F., & Roman, D. J. (2017). Methods of evaluation of Soewarno, N., & Tjahjadi, B. (2020). Measures that matter: An empirical investigation of
intangible assets and intellectual capital. Journal of Intellectual Capital, 18(3), intellectual capital and financial performance of banking firms in Indonesia. Journal
470–485. of Intellectual Capital, 21(6), 1085–1106.
Oster, E. (2019). Unobservable selection and coefficient stability: Theory and evidence. Su, H. Y. (2014). Business ethics and the development of intellectual capital. Journal of
Journal of Business & Economic Statistics, 37(2), 187–204. Business Ethics, 119, 87–98.
Owusu, A., Kwabi, F., Ezeani, E., & Owusu-Mensah, R. (2022). CEO tenure and cost of Sydler, R., Haefliger, S., & Pruksa, R. (2014). Measuring intellectual capital with
debt. Review of Quantitative Finance and Accounting, 59(2), 507–544. financial figures: Can we predict firm profitability? European Management Journal, 32
Pan, W. F., Wang, X., & Yang, S. (2019). Debt maturity, leverage, and political (2), 244–259.
uncertainty. The North American Journal of Economics and Finance, 50, Article Tan, H. P., Plowman, D., & Hancock, P. (2008). The evolving research on intellectual
100981. capital. Journal of Intellectual Capital, 9(4), 585–608.
Pan, Y., Wang, T. Y., & Weisbach, M. S. (2016). CEO investment cycles. The Review of Trevlopoulos, N. S., Tsalis, T. A., Evangelinos, K. I., Tsagarakis, K. P., Vatalis, K. I., &
Financial Studies, 29(11), 2955–2999. Nikolaou, I. E. (2021). The influence of environmental regulations on business
Panousi, V., & Papanikolaou, D. (2012). Investment, idiosyncratic risk, and ownership. innovation, intellectual capital, environmental and economic performance.
The Journal of Finance, 67(3), 1113–1148. Environment Systems and Decisions, 41, 163–178.
Paquette, K., Huynh, N., & Vu, P. (2022). To judge leviathan: Sovereign credit ratings, debt Trinh, V. Q., Aljughaiman, A. A., & Cao, N. D. (2020). Fetching better deals from
overhang, and corporate investment. Available at SSRN 4209496. creditors: Board busyness, agency relationships and the bank cost of debt.
Pathan, S. (2009). Strong boards, CEO power and bank risk-taking. Journal of Banking & International Review of Financial Analysis, 69, Article 101472.
Finance, 33(7), 1340–1350. Vu, P. T. T., Huynh, N., Phan, H., & Hoang, H. (2023). Financial earthquakes and
Phan, D. H. B., Tran, V. T., Nguyen, D. T., & Le, A. (2020). The importance of managerial aftershocks: From Brexit to Russia-Ukraine conflict and the stability of European
ability on crude oil price uncertainty-firm performance relationship. Energy banks. Journal of International Financial Markets, Institutions and Money, 88, 101830.
Economics, 88, Article 104778. Welch, K., & Yoon, A. (2022). Do high-ability managers choose ESG projects that create
Pulic, A. (2000). VAICTM–an accounting tool for IC management. International Journal of shareholder value? Evidence from employee opinions. Review of Accounting Studies,
Technology Management, 20(5–8), 702–714. 1–28.
Rajan, R., & Zingales, L. (1998). Financial dependence and growth. American Economic Yung, K., & Chen, C. (2018). Managerial ability and firm risk-taking behavior. Review of
Review, 88(3), 559–586. Quantitative Finance and Accounting, 51, 1005–1032.
Ren, S., & Song, Z. (2021). Intellectual capital and firm innovation: Incentive effect and Zhou, A. Z., & Fink, D. (2003). The intellectual capital web: A systematic linking of
selection effect. Applied Economics Letters, 28(7), 617–623. intellectual capital and knowledge management. Journal of Intellectual Capital, 4(1),
Santa-Clara, P., & Valkanov, R. (2003). The presidential puzzle: Political cycles and the 34–48.
stock market. The Journal of Finance, 58(5), 1841–1872.

18