Journal of Business Research 139 (2022) 918–934

Contents lists available at ScienceDirect

Journal of Business Research

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

Generation effects and managerial risk taking

Siew-Boey Yeoh, Chee-Wooi Hooy *
School of Management, Universiti Sains Malaysia, Malaysia

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

Keywords: This study examines the generation effects on managerial risk-taking by decomposing between-generation
Managerial risk-taking variation and within-generation variation. Using 6,169 firm-year observations in Malaysia data from 2009 to
CEO generation 2017, for the between-generation effect, we find that Pre-War II and Pre-Independence CEOs undertake lower
Between-generation
levels of risk-taking while Post-Independence and Development-Era CEOs undertake higher levels of risk-taking.
Within-generation
For the within-generation effect, we find that CEOs who have experienced the traumatic event during their
impressionable year are relatively risk averse, which implies that among Pre-War II, Pre-Independence and Post-
Independence CEOs, those who were pre-adulthood at the time of the event will be the most risk adverse. The
cumulative impression, however, only significantly affects the Pre-War II CEOs because the layers of history may
erode the traces of older time periods, while more recent cohort of CEOs have fewer new layers that can
attenuate the strength of imprints over time; hence, significant effects were not observed.

1. Introduction born during the first five years of this 20-year time span may experience
a traumatic event during their early adulthood, while CEOs who are
Upper echelon theory has provided a substantial theoretical base for born during the last five years of this generation have yet to enter
research on the profile of corporate elites, such as gender, age and social adolescence when the event occurs. The age at which the event occurs
class, in determining strategic choices and organisational outcomes serves as an important determinant of the impression of the event.
(Hambrick & Mason, 1984). One stream of upper echelon research has Impression year is a sensitive transition period during which individuals
suggested that background and life experiences can have a profound demonstrate high susceptibility to external influences and develop im­
effect on individuals’ managerial styles and decision making (Chen, Fan, prints, and these characteristics reflect prominent features of the envi­
Yang, & Zolotoy, 2021; Malmendier, Tate, & Yan, 2011; Zhang, 2017). ronment (Marquis & Tilcsik, 2013). If the event occurs during an
To our knowledge, Malmendier et al. (2011) presented the first study to impressionable year for a CEO or during the coming-of-age period be­
link the early-life experience of business leaders to corporate financial tween 18 and 25, it is expected to have a significant effect on developing
policies. They use Depression-baby CEOs and identify the two biggest the imprinting characteristics that may affect the CEO’s risk-taking
shocks that represent the formative experiences of these CEOs, namely, preference at the firm level years later. Thus, we examine the hetero­
the Great Depression and military service, with the former group relying geneity within a generation, which we name the “within-generation”
more on internal finance and the latter choosing more aggressive pol­ effect, to perform a more refined investigation of generation effects on
icies. Despite this extensive empirical evidence, little has been done to managerial risk-taking. Grounded on the premise that persistence does
examine whether behaviours differ across different generations. Thus, not imply permanence (Marquis & Tilcsik, 2013), we investigate
our study aims to investigate the differences in risk-taking behaviours whether imprints may change over time. Research has shown that
across different generations of CEOs. We shall call this the between- socialised values can erode with time or in response to new events
generation effects, henceforth. In addition, we delve into variations in (Akerlof, 1983). The incremental effect of new shocks may depreciate
the early life experience within a generation, which we termed as when experience gradually accumulates throughout the life of an indi­
within-generation effect. vidual (Carlsson & Karlsson, 1970). Thus, we further examine how
A generation ranges across a time span of 15 to 20 years, and the subsequent interactions with new events may result in possible imprint
intensity of experience will differ across CEOs in the same generation decay using the aggregate outcome of events and analyse the effect of
depending on the age of the CEO at the time of the event. CEOs who are the cumulative impression of events throughout life on the risk

* Corresponding author at: School of Management, Universiti Sains Malaysia, 11800 Penang, Malaysia.
E-mail address: [email protected] (C.-W. Hooy).

https://doi.org/10.1016/j.jbusres.2021.09.063
Received 28 April 2020; Received in revised form 24 September 2021; Accepted 27 September 2021
Available online 26 October 2021
0148-2963/© 2021 Elsevier Inc. All rights reserved.
S.-B. Yeoh and C.-W. Hooy Journal of Business Research 139 (2022) 918–934

behaviour of CEOs. military service and combat exposure during World War II, focuses on
the imprinting aspect of individuals’ early life. The former research does
“Each generation is profoundly influenced by the times in which it grows
not account for the imprinting effect because these studies do not cap­
up – the music, movies, politics, and defining events of that period …
ture the age at which CEOs experience the Great Depression and Great
Members of a cohort [generation] share the same major culture, political,
Famine. For the latter research, these studies have not captured the
and economic experiences. They have similar outlooks and values.” –
socialisation process. Thus, whether similar behaviour is demonstrated
Kotler & Keller (2006)
by CEOs with recent military service in the Gulf War relative to CEOs
The above quotation implies that the generation effect has been used with World War II military experience cannot be clarified. We have
in marketing as an explicit approach to tap into shared social values. moved a step forward to integrate both the socialisation and imprinting
Generation is one type of subculture for understanding the evolutionary aspects using within-generation analysis and demonstrate how social­
process of change that has taken place in a country and influences in­ isation process and age at which external environmental stimuli occur
dividuals’ preference and behaviour (Egri & Ralston, 2004). While lead to different risk-taking behaviours of CEOs from different
previous studies have shown that early life events affect CEOs’ corporate generations.
risk-taking behaviour, they do not account for the potential dissimilar­ While intergenerational values leave a long-lasting effect throughout
ities among each generation. The dissimilarities may be attributed to the the lifetime of an individual (Inglehart, 1997), we demonstrate that
nature of historical events and scarcity aspects, which vary between information dynamics may affect the original imprint. The interplay
generations. Drawing on the theory of changes in intergenerational between different generations of imprints layered upon one another may
values by Inglehart (1997), we conduct a between-generation effect cause imprint erosion. Previous scholars suggested that imprints are
analysis to perform the comparison among generations. We argue that dynamic (Simsek, Fox, & Heavey, 2015), and our study also provides
individuals from a respective generation develop a distinctive set of evidence that imprints may change over time, especially for World War
relatively stable values that are different from that of other generations. II CEOs. Examining this issue is important because the extent to which
These distinctive sets of values emanate from the scarcity aspect during CEOs’ imprint decays can significantly affect the strategic decisions by
a generation’s adolescence (Schwartz & Sagie, 2000) and affect CEOs’ their firms.
decision-making through perceptual filtering (England, 1967). In addi­
tion, heterogeneity within a generation has remained virtually unex­ 2. Theory and hypothesis development
amined. Thus, we explore the within-generation effect by examining the
age at which an individual is exposed to the traumatic event and the 2.1. Theoretical background and segmentation of generation
influence of cumulative event impressions on the risk behaviour of
CEOs. Upper echelon theory explicitly argues that an organisation’s acts
This study examines the generation effect in the context of Malaysia. can, to some extent, be predicted from the background of its top exec­
Previous studies mainly labelled a generation based on the birth group utives (Hambrick & Mason, 1984). Upper echelon theory is built on the
generation in the United States (Strauss & Howe, 1991; Thau & Heflin, principle of bounded rationality, which suggests that informationally
1997). However, not all countries share the same generation effect as the complex and uncertain situations are indeed interpretable. Based on this
United States because exogenous macroeconomic events vary across argument, Hambrick & Mason (1984) developed two interconnected
countries (Schewe & Meredith, 2004), and performing similar profiling central ideas, namely, psychological properties and observable experi­
of these countries may produce misleading results. Malaysia experi­ ence. The former idea suggests that the actions of executives are based
enced turbulent political changes before independence and a relatively on their personal interpretation of the situation, while the latter idea
stable political environment but volatile economic conditions after in­ devotes that these personal interpretations are a function of their
dependence. Using a mix of political and economic factors as the experience, values and personalities. As such, executives’ construals of
defining moments, we create four generations of CEOs – Pre-War II, Pre- reality shape their strategic choices, which involve taking risks (Child,
Independence, Post-Independence and Development-Era. We identify 1974).
five major traumatic events that occur during the formative year of each Major traumatic events during formative years result in a distinctive
generation and analyse how scarce aspects during these events create set of values for a generation, as predicted by Inglehart’s (1997) theory
intergenerational value. of intergenerational value change, which is based on two hypotheses:
Our study makes several contributions. Previous studies tend to use the socialisation hypothesis and the scarcity hypothesis. The social­
age and generation interchangeably (Chen, Wong, & Lee, 2001). Age isation hypothesis predicts that adults’ basic values reflect the socio­
effects are developmental in nature and caused by internal biological or economic condition of their formative years, and this value remains
psychosocial ageing (Rhodes, 1983); however, generation effects are the unchanged throughout their lifetime (Inglehart, 1997; Strauss & Howe,
result of the influence of the external environment on all members in the 1991). The scarcity hypothesis proposes that the greatest value is placed
same cohort regardless of age (Yao, Sharpe, & Wang, 2011). If differ­ on the socioeconomic environmental aspects that are in short supply
ences in behaviours are caused by age effects, then younger individuals during a generation’s adolescence. Inglehart’s (1997) theory proposes
would likely become more similar to older adults as they age. However, that generations growing up during periods of socioeconomic and
if the differences are caused by generational effects, then these behav­ physical scarcity learn conformity and respect for authority while gen­
iours and the differences between age cohorts would likely remain erations growing up during socioeconomic security emphasise individ­
relatively stable (Rhodes, 1983). By examining how the generational ualism and openness-to-change.
value created from the collective experiences of events results in stable Values can be expected to affect CEO choices through two conduits,
behaviour throughout CEOs’ lifetimes, we explore how CEO generation namely, directly through behaviour channelling or indirectly through
effects explain organisational outcomes beyond CEO age. perceptual filtering, which is considered far more prevalent (Chin,
Early life research, such as by Malmendier & Nagel (2011) and Zhang Hambrick, & Treviño, 2013). Through “motivated cognition”, which
(2017), employs the Great Depression and Great Famine as macroeco­ psychologists refer to as the interpretive process, individuals tend to see
nomic events to deduce the risk-taking strategy of CEOs born during what they want to see and hear what they want to hear (Weick, 1979).
those generations. This body of research has featured the socialisation This limited cognitive ability and human need for simplification has
aspect of generation. Another stream of research, such as earlier works motivated executives to adopt perceptual filtering by selectively
by Malmendier et al. (2011) and Benmelech & Frydman (2015), and filtering information that is not aligned with their value (Weber &
recent work by Guo, Zan, Sun, & Zhang (2020) analyses how CEOs’ Milliman, 1997). Risk perception is therefore shaped by subjective
judgement bounded by the cognitive constraints of an individual

919
S.-B. Yeoh and C.-W. Hooy Journal of Business Research 139 (2022) 918–934

(Hambrick & Mason, 1984). Era (1980–2000) individuals were born in the era of globalisation and
According to Sitkin & Pablo (1992), the risk perception of an indi­ grew up during a period of advancements in technology. The country
vidual is shaped by the salience of loss and gain outcomes. If an indi­ marked an important milestone under the leadership of Tun Dr Maha­
vidual focuses more on the loss outcome, then the subjective estimate of thir, the father of modernisation for Malaysia. Privatisation and a
the probability of the loss event becomes higher. As a result, he or she structural shift towards the manufacturing sector have contributed
will perceive higher risk (Sitkin & Weingart, 1995). In contrast, if an significantly to the unprecedented economic growth of the country.
individual focuses on gain outcomes, then the upside possibilities Post-Independence (1961–1979) CEOs represent the earliest gener­
become salient and he or she will perceive the event as lower risk (March ation that experienced economic and political stability, but they expe­
& Shapira, 1987). Greater salience of gains to firm performance and to rienced two major crises during their formative years due to the
CEO option wealth leads to a lower risk perception and greater risk- openness of the Malaysian economy. The first crisis was the commodity
taking tendency which is able to weaken the negative effect of equity shock from 1985 to 1986, which resulted in a collapse of worldwide
risk bearing on strategic risk taking (Benischke, Martin, & Glaser, 2019). commodity trade. The second crisis was the Asian financial crisis from
Each generation that has experienced different life events develops 1997 to 1998, when the collapse of the Thai baht caused a domino effect
different values that are part of their cognitive base. These generational in the region. As this generation grew up during periods of economic
values are imprinted on the individuals, which will then be reflected in restructuring with better job opportunities, they are likely to place
their decision making over time. These values are translated into CEOs’ greater importance on career success, individualism and self-
attention, selection, and interpretation of stimuli, which explains why enhancement (even with financial crisis experience). We propose that
some generations are more likely to take risks than others. this generation is expected to perceive a lower risk of failure and un­
dertake higher risks.
2.2. Comparison among generations (between-generation effect) Compared to the Post-Independence CEOs, Pre-War II CEOs experi­
enced severe physical and socioeconomic insecurity because they lived
We examine the between-generation effect based on the notion that through a period of poverty, war and political instability. Pre-War II was
generational value change is an ongoing evolutionary process that leads the only generation in Malaysia that experienced war. Individuals who
to changes in the priorities emphasised during a particular historical grow up during war are viewed as supportive of conformity and having
period that has taken place in a country (Egri & Ralston, 2004). Schewe the deepest concern for security (Strauss & Howe, 1991) and the desire
& Meredith (2004) suggest using defining moments to create genera­ to avoid risk (Egri & Ralston, 2004), and they weigh loss outcomes much
tions with different dates, lengths and value systems. Previous studies more heavily than gain outcomes. They are likely to perceive higher risk
often define a moment using a single factor, such as political ideology for a strategic choice and thus undertake a lower risk. Exposure to war
(Egri & Ralston, 2004), economic condition (Schewe & Noble, 2000) and recollection of fear from violence can result in risk aversion even
and historical milestones (Cox, Hannif, & Rowley, 2014). three to five decades later (Kim and Lee, 2014). Thus, we propose the
We create generations in Malaysia using a unique mixture of political following hypothesis:
and economic conditions. Before the formation of Malaysia, the country
Hypothesis 1a.. In comparison to Post-Independence CEOs, Pre-War
underwent turbulent political changes with British colonisation, World
II CEOs undertake the lowest levels of managerial risk-taking.
War II and Japanese occupation. After the country’s independence in
1957, the political environment was more stable under the ruling of the Pre-Independence CEOs have struggled through colonisation and
United Malays National Organisation (UMNO), which was operated as a strived for independence. They expected a better life after independence
single-party-dominant regime for 61 years until 2018. However, the because the economic conditions appeared conducive to sustained
openness of the Malaysian economy has made it vulnerable to economic growth with well-developed infrastructure inherited from the colony.
shocks. We segment the generations of Malaysia into Pre-War II, Pre- However, economic expansion has failed to make a substantial contri­
Independence, Post-Independence and Development-Era and consider bution to employment and income distribution. Dissatisfaction due to
the impact of five major historical events: (1) World War II 1941–1945, interethnic income inequalities resulted in the May 1969 outbreak
(2) the May 1967–1969 riots, (3) the commodity shock 1985–1986, (4) (Athukorala, 2010).
the Asian financial crisis 1997–1998 and (5) the global financial crisis The salience of loss outcomes is greater for Pre-Independence CEOs
2007–2008. because they experience distress and an absence of optimism on their
Pre-War II (born before 1945) individuals were born during British future even though they have made a great effort to achieve indepen­
colonisation and Japanese invasion in World War II (1941–1945). They dence. They are more likely to perceive greater firm risk and reduce
witnessed extensive destruction of the infrastructure and severe eco­ strategic risk-taking. In contrast, Post-Independence CEOs grew up
nomic dislocation and experienced life threats and extreme poverty, during economic restructuring and have a higher propensity to take risk.
especially during Japanese occupation. Pre-Independence (1946–1960) Therefore, we hypothesise that Pre-Independence CEOs may take fewer
individuals were born after the Japanese occupation. The returning risks relative to their Post-Independence counterparts.
British rulers have rebuilt the economy and rationalised fragmented
Hypothesis 1b.. Relative to Post-Independence CEOs, Pre-
administrative structure. Compared with Pre-War II, this generation
Independence CEOs undertake a lower level of managerial risk-taking.
grew up in a better environment despite the high unemployment and
unstructured education system. Income inequalities between ethnic Development-Era CEOs grew up in periods of robust economic
classes have increased due the wealth and educational level disparities growth. This generation experienced the global financial crisis of
between urban and rural areas. The growing disillusion culminated in 2007–2008. This crisis affected the subprime mortgage sector in the
the 1969 racial riots (Jomo, 1990) and was further sparked by United States and subsequently disrupted the global economy. However,
communist insurgency. Many hundreds of people were killed, and the unlike the Asian financial crisis of 1997–1998, Malaysia had negligible
riots were followed by the declaration of a state of emergency and exposure to securities linked to US subprime loans. In addition, financial
eventually the implementation of the New Economic Policy. institutions in Malaysia were in better shape than they were during the
Post-Independence (1961–1979) individuals were born in a newly Asian financial crisis (Bank Negara, 2008).
independent nation. Given the establishment of the New Economic Under the modernisation, rapid economic growth and weaker global
Policy in 1970 and liberalisation of trade, this generation grew up financial crisis impacts in Malaysia, this generation is expected to
during a new phase of economic growth (Yusoff, Hasan, & Jalil, 2000). embrace openness-to-change. Gain outcomes are more salient to them,
This generation experienced better job opportunities and improved and the perceived risk in decisions is likely to be lower. De Meuse,
welfare and quality of life from the urbanisation process. Development- Bergmann, & Lester (2001) argue that generations growing up during

920
S.-B. Yeoh and C.-W. Hooy Journal of Business Research 139 (2022) 918–934

periods of robust economy exhibit a high level of entrepreneurial risk- Drawing on the impressionable year hypothesis, we contend that
taking. Hence, our next specific hypothesise is as follows: even if a generation experiences the same historical event, the impact of
the event may vary from one individual to another, as the event may or
Hypothesis 1c.. Relative to Post-Independence CEOs, Development-
may not occur during his or her coming-of-age. Universal agreement has
Era CEOs undertake the highest levels of managerial risk-taking.
not been reached on the most impressionable years. Studies on psy­
Table 1 demonstrates the Malaysian generation segmentation and chology development suggest that values crystallise during the coming-
summarises the socioeconomic conditions and major societal events that of-age of between 18 and 25 years old (Krosnick & Alwin, 1989). Po­
created the value orientation and risk perception of the respective litical socialisation research found evidence of significant socialisation
generations. between 18 and 25 years old (Newcomb,1943). Economics literature has
also documented attitudes are durably formed during age of 18 to 25
(Eichengreen, Aksoy, & Saka, 2021). Thus, we also analyse whether the
2.3. Heterogeneity within a generation (within-generation effect) event occurs during the coming-of-age period from 18 to 25 years of age.
As shown in Fig. 1, the dark grey area indicates the impressionable
To better understand the generation effect, we examine the within- years of the individuals (age 18 to 25) when the traumatic event occurs.
generation effect by analysing the impression of the event and the cu­ For instance, among the Pre-War II generation, CEOs born in 1917–1927
mulative impression of events throughout the life span of CEOs. experienced World War II, and CEOs born in 1942–1945 experienced
1969 riots during their coming-of-age. Relative to other Pre-War II
2.3.1. Impression of the event CEOs, these two birth groups are expected to undertake lower risk. The
Perceptual filtering indicates that information processing is the same empirical analysis extends to other generations. Hence, we suggest
mechanism through which value shapes choice (England, 1967). the following hypothesis:
Imprinting is a process in which an entity develops characteristics that
reflect prominent features of the environment during a brief period of Hypothesis 2.. Within the same generation, CEOs who have experi­
susceptibility. According to Marquis & Tilcsik (2013), imprinting enced the traumatic event during their coming-of-age period undertake
(experience of an event) serves as an important filter that shapes in­ a lower level of managerial risk-taking than those who have experienced
dividuals’ information processing. Upper echelon theory has devoted the event during their non-coming-of-age.
personal interpretation as a function of observable experience (Ham­
brick & Mason, 1984). Similarly, experiences during this sensitive 2.3.2. Cumulative impression of events
period shape the interpretation of what constitutes appropriate behav­ Scholars have argued that persistence does not imply permanence
iour later in life. Characteristics formed during this period persist despite because there are forces that can attenuate the strength of imprints over
subsequent environmental changes (Marquis & Tilcsik, 2013). time (Simsek et al., 2015). When imprints of the past clashes with the
Sensitive periods typically occur during early adulthood (Kish- prevailing ideology of the present, they suffer transformation and are
Gephart & Campbell, 2015). Early adulthood years are the impression­ replaced with imprints that are opposite of them (Albu, Albu, Apostol, &
able years or coming-of-age years of an individual (Giuliano & Spi­ Cho, 2020). Although information filters are fairly stable, information-
limbergo, 2009). The impressionable years hypothesis states that events processing systems are also adaptive and capable of changing in­
occurring during the coming age of an individual have a more profound dividuals’ behaviour over time through learning (Simon, 1978). Thus,
impact on his or her thinking (Cutler, 1974). Coming of age is a time of imprints may also change over time because exposure to new informa­
identity formation (Erikson, 1950) and time of their generation (Schu­ tion from different channels can affect the original imprint (Marquis &
man & Scott, 1989). According to psychology research, it is a special Qiao, 2020). Political socialisation research also proposes that as prior
time for autobiographical memory encoding (Rubin, 1988). Schuman & experience gradually accumulates, the incremental impact of new
Scott, 1989 found that individuals who were 16–24 years old during shocks declines (Carlsson & Karlsson, 1970). Routinisation of attitudes
World War II and 15–27 years old during the Vietnam War were sub­ occurs during early adulthood and is characterised by a cut-off point at
stantially more likely to recall these events. Individuals who did not which firm attitudes are established. From then on, it becomes expo­
come of age during these events did not have such strong recall. nentially harder for new events to update one’s previous events (Alwin

Table 1
Malaysian generation labels.
Pre-War II Pre-Independence Post-Independence Development-Era

Year of birth Before 1945 1946–1960 1961–1979 1980–2000

Socioeconomic ● Strive for survival ● Struggle through colonisation ● Economic restructuring ● Globalisation
condition ● Greatest hardship and seeking independence ● Urbanisation ● Economic growth
● Lack of shelter ● High unemployment and low ● Equitable and fair income ● Technology advancement
● Inadequate infrastructure wages distribution
● Extreme poverty ● Unstructured education ● Better job opportunities
system
Political condition ● British colonisation ● Post-British colonisation ● Relatively stable political ● Political stability under single-party rule and
● Japanese occupation (World condition under single-party leadership by the “father of development,”
War II, 1941 to 1945) rule Tun Mahathir
Economic ● Dependence on agriculture ● Dependence on agriculture ● New Economic Policies ● Dependence on manufacturing
condition and mining and mining ● Import substitution and ● International trade expansion
● Economic activities in the ● Small-scale and family-based export-oriented strategy ● Privatisation Master Plan
hand of leading colonial local business enterprise
industries
Historical event ● World War II (1941–1945) ● May 1969 riots (1967–1969) ● Commodity Shock ● Global Financial Crisis (2007–2008)
● May 1969 riots (1967–1969) (1985–1986)
● Asian Financial Crisis
(1997–1998)
Value orientation ● Conservation ● Self-transcendence ● Self-enhancement ● Openness-to-change
● Individualism
Risk perception ● Perceive highest risk ● Perceive higher risk ● Perceive lower risk ● Perceive lowest risk

921
S.-B. Yeoh and C.-W. Hooy Journal of Business Research 139 (2022) 918–934

Fig. 1. Historical macroeconomic events and intergenerational effects influencing different generations in Malaysia. Note: This figure shows the impressionable year
of the CEO according to their birth year during the event. For example, for a Pre-War II CEO born from 1917 to 1927 (y-axis), the impressionable years (dark grey
area) occurred during the World War II event (x-axis). Similarly, for a Pre-War II CEO born from 1942 to 1945, the impressionable years were during the May
1969 Riots.

& McCammon, 2003). If this is the case, then we would expect that CEOs considered to be less influential. The impacts of the three financial crises
who have accumulated more event experience tend to undertake a in Malaysia are different due to the different policy responses and re­
higher level of risk-taking. covery processes. Malaysia has managed the commodity shock well by
the relocation of production bases and promotion of foreign direct in­
Hypothesis 3.. Within the same generation, CEOs who have accu­
vestment. The Asian financial crisis was the most impactful because the
mulated more traumatic events undertake a higher level of managerial
sharp depreciation of the ringgit against the US dollar resulted in a deep
risk-taking.
economic recession. The global financial crisis is the least impactful
We analyse the cumulative impression of CEOs by considering the given the sound banking system and reforms of financial regulations in
internal aspect and external aspect of each experience across the life the aftermath of the Asian financial crisis (Athukorala, 2013). Thus, the
span of the CEOs. We analyse the internal aspect by identifying the level severity scale of events that occurred in Malaysia are arranged in such a
of impression of the events. According to Schuman & Scott, 1989, events way: most severe for World War II, which was followed by the May 1969
coming from the age range of 17–25 are viewed as crucial; nevertheless, riots, Asian financial crisis, commodity shock and global financial crisis.
events from other periods also have effects. Before the age of early Here, we present the conceptual framework (See Fig. 2).
adulthood, individuals lack the cognitive abilities to learn about com­
plex events and lack awareness of the world beyond the family until age 3. Method
17 (Schuman & Scott, 1989). Significant socialisation occurs between
ages 18 and 25, and people are most susceptible to attitude changes; 3.1. Data
susceptibility drops off immediately after age 25 and remains low
throughout the remainder of life (Krosnick & Alwin, 1989). Thus, we The rich societal events with several major political shocks and
label events at ages from 18 to 25 as the most impactful, events before 17 economic crises throughout the history of Malaysia have provided a
as less influential and events after 25 as the least impactful. good setting to test the generation effect. In addition, the wide ranges of
We analyse the external aspect of experience based on the severity CEO birth years from 1917 to 1989 allow for an examination of the effect
level of the event. Prior studies have demonstrated that traumatic ex­ between generations and a more refined investigation on heterogeneity
periences such as war and turmoil have a long-term impact. Exposure to within a generation. We test our predictions using a sample of publicly
war influences individuals to be more averse to risk for decades (Kim listed firms on the main market in Bursa, Malaysia. We exclude the ACE
and Lee, 2014). Unlike the detrimental physical and mental effects due market because it mainly consists of newly listed firms that might not
to war (Akbulut-Yuksel, 2014), the effect of a financial crisis is have large capital. We also exclude finance companies, real estate

Fig. 2. Conceptual framework.

922
S.-B. Yeoh and C.-W. Hooy Journal of Business Research 139 (2022) 918–934

investment trusts (REITs), closed-end funds and exchange-traded funds which the CEO experiences each event. The index reflects the effect of
because of different accounting practices and statutory requirements. imprint/impression of the event, with low score indicating less
Our sample frame covers nine financial years from 2009 to 2017. We impression of the event as the event is less severe and/or the event does
choose 2009 as the starting point to avoid any disturbance in risk from not occur during the early adulthood of the CEOs. Fewer impression
the 2007–2008 financial crisis. Malaysia underwent political stability events will be remembered and thus have less impact on behaviour later
after independence from the UMNO as a single-party-dominant regime in life. In contrast, a high score reflects a more profound adverse effect
until 2018. Because the political environment is used as part of our because the traumatic event is more severe and occurs during the
generation segmentation criteria, we choose the ending point of 2017 to impressionable years of the CEOs. The index of each event is then added
minimise any possible aftermath of political instability that may disturb to compute the aggregate outcome of all events throughout the life of
our segmentation. CEO and board characteristics are hand-collected CEOs. With this computation, we can examine the incremental effect of
from the profile of the board of directors in the annual report, and traumatic experience accumulated throughout the whole life of a CEO.
financial data are obtained from DataStream. After removing firms with We present the cumulative impression index for each birth year in
incomplete data, the final panel consists of 6,169 firm-year Table 2, Column 5, and provide explanations on how this cumulative
observations. index is computed in Table 3. Using 3 different birth years of Pre-War
CEOs and 1 birth year of Development-Era CEOs, we demonstrate in
3.2. Main variables and empirical model Table 3 a varying aggregate index for CEOs. For example, in Panel A, for
a CEO born in 1917, the age of the CEO ranges from 24 to 28 during
The dependent variable in this study is managerial risk-taking, and it event 1 (World War II), and has reached 50 for event II. At the age of 24,
is measured by two variables: financial leverage (Leverage) and capital the Impression score is 3 because it is at the stage of early adulthood (age
expenditure (Capex). These variables are widely used to examine CEOs’ 18–25), and the Event score is 5 because World War II is the most severe
risk-taking behaviour (Hoskisson, Chirico, Zyung, & Gambeta, 2017; event. Thus, the impression index gained at this point is 15 (Event score
Serfling, 2014). Leverage is a riskier financial choice because a high × Impression score = 5 × 3). By adding the indices, the cumulative
level of debt financing increases the risk of default (Serfling, 2014). impression index for the CEO is 69. CEOs who were born in 1921 have a
Capital expenditures involve high costs in the year they are incurred and cumulative index of 99, which is mainly because they experienced
may not lead to revenue for several years in the near future (Lu & Wang, World War II for 5 years at the age of 20–24. CEOs who were born in
2018). Following prior studies, financial leverage is measured as long- 1928 did not experience traumatic events at the age of 18–25; thus, the
term debt over the book value of total assets (Hall, 2012) and capital cumulative index was lower at the aggregate score of 74. Development-
expenditure is measured as capital expenditure divided by the book Era CEOs born in 1989 have a relatively low index of 18 because they
value of total assets (Coles, Daniel, & Naveen, 2006). experienced only 2 financial crises. We illustrate the composite of the
To investigate the between-generation effect of the CEO on mana­ cumulative impression index of each birth year for the respective gen­
gerial risk-taking (RiskTaking), we estimate the following regression erations in Fig. 3.
model. Dummy variables are constructed based on the category of in­ We normalise the cumulative impression index (Cumulative) using
dividuals who were born before 1945 (DPre-War II); born between 1946 distancing from the minimum value technique (Freudenberg, 2003) by
and 1960 (DPre-Independence); born between 1961 and 1979 (DPost-Indepen­ dividing the actual value by the minimum value and estimate the
dence); and born between 1980 and 2000 (DDevelopment-Era). The coefficients
following equation:
attached to these dummy variables are known as the differential inter­ ∑
RiskTakingit = αit + βk Controlit + β1 RiskTakingi,t− 1 + β2 Cumulativeit + εit
cept coefficients. We assign Post-Independence CEOs as the reference
category, and all comparisons are made in relation to this benchmark (3)
category. Based on previous studies on managerial risk-taking (Lu & Wang,
∑ 2018; Rashad Abdel-Khalik, 2014; Serfling, 2014), we control for a se­
RiskTakingit = αit + βk Controlit + β1 RiskTakingi,t− 1 + β2 DPre− WarII,it
ries of factors that are potentially correlated with risk-taking. We control
+ β3 DPre− Independence,it + β4 DDevelopment− Era,it + εit (1) for CEO tenure (Tenure, measured as the number of years the CEO has
To analyse the within-generation effect, we examine whether the been with the firm), CEO age (CEOAge, measured as the age of the CEO),
event occurs during an impressionable year and the cumulative event firm size (LogFirmSize, measured as the natural logarithm of the assets),
impressions of CEOs. For the former, we treat each generation as a firm age (LogFirmAge, measured as the natural logarithm of the number
subsample and use DImpression, a dummy variable that equals one if the of years from the establishment of the firm to the year of observation),
CEO aged 18–25 at the time of the traumatic event and zero otherwise. cash holdings (Cash, measured as the cash and cash equivalents divided
The tabulation of the impression dummy for each birth year is presented by the assets), sales growth (SalesGrowth, measured as the annual rate of
in Table 2, Column 4. We estimate the following model and run the growth of sales), board independence (IndBoard, measured as the pro­
regression for each generation subsample. portion of independent directors on the board) and board size (Board­
∑ Size, measured as the total number of directors on the board).
RiskTakingit = αit + βk Controlit + β1 RiskTakingi,t− 1 + β2 DImpression,it + εit
(2)
3.3. Estimations
For the latter, we construct an index to test the cumulative impres­
sion effect. Based on the discussion in Section 2.3.2, the severity levels of The basic logic of upper echelon theory is that executives take ac­
the events are scaled as follows: most severe for World War II (score 5), tions based on their personalised construal of the situation. However,
May 1969 riots (score 4), Asian financial crisis (score 3), commodity Hambrick (2007) suggest that there are at least two reverse logics at
shock (score 2) and global financial crisis (score 1). We label events at play. First, it is possible that executives are drawn to and advance within
18–25 years of age as most impactful (score 3), events before age 17 as settings that suit their profiles. For instance, while upper echelon theory
less influential (score 2) and events after age 25 as least impactful (score may predict that executives with technology expertise tend to invest
1). The event scores are then multiplied by the impression scores to more in R&D, one can also expect that executives might be drawn to
compute an impression index, which is then summed to attain the cu­ R&D-intensive companies. This issue can be solved by incorporating
mulative impression index. temporal lags for prior states of independent and dependent variables.
The multiplication of the event scores and impression scores allows The second type of reverse causality is the problem of endogeneity. In
us to capture the interaction of the severity of the event and the age at such cases, executives indeed take actions that mirror their profile,

923
S.-B. Yeoh and C.-W. Hooy Journal of Business Research 139 (2022) 918–934

although it is not because of their information processing bias but rather that fulfil the condition of relevance and exogeneity (Wintoki, Linck, &
is due more to the mandate of the board or hiring body, who believes Netter, 2012).
that the executive has the precise characteristics needed for the situation The system GMM estimator offers a good approach for mitigating
at hand. This behaviour can be addressed with the use of instrumental these issues. It is one of the endogeneity remedies that correct causality
variables to decompose the variation in the endogenous independent issues. The GMM estimator uses lagged values of explanatory variables
variable. However, it is difficult to identify valid instrumental variables as instruments for the current explanatory variables. Any information

Table 2
Impression dummy and corresponding cumulative impression index for each birth year.

(continued on next page)

924
S.-B. Yeoh and C.-W. Hooy Journal of Business Research 139 (2022) 918–934

Table 2 (continued )

Table 3
Explanation of the computation of the cumulative impression index.
Event 1 Event 2 Event 3 Event 4 Event 5 Cumulative impression
World War II May 1969 riots Commodity Asian Global index
(1941–1945) (1967–1969) Shock Financial Financial
(1985–1986) Crisis Crisis
(1997–1998) (2007–2008)

Year of event 1941 1942 1943 1944 1945 1967 1968 1969 1985 1986 1997 1998 2007 2008

Event score 5 5 5 5 5 4 4 4 2 2 3 3 1 1
Panel A: CEO born in 1917
Age at time of event 24 25 26 27 28 50 51 52 68 69 80 81 90 91 69
Impression score 3 3 1 1 1 1 1 1 1 1 1 1 1 1
Impression index 15 15 5 5 5 4 4 4 2 2 3 3 1 1
Panel B: CEO born in 1921
Age at time of event 20 21 22 23 24 46 47 48 64 65 76 77 86 87 99
Impression score 3 3 3 3 3 1 1 1 1 1 1 1 1 1
Impression index 15 15 15 15 15 4 4 4 2 2 3 3 1 1
Panel C: CEO born in 1928
Age at time of event 13 14 15 16 17 39 40 41 57 58 69 70 79 80 74
Impression score 2 2 2 2 2 1 1 1 1 1 1 1 1 1
Impression index 10 10 10 10 10 4 4 4 2 2 3 3 1 1
Panel D: CEO born in 1989
Age at time of event 8 9 18 19
Impression score 2 2 3 3
Impression index 6 6 3 3 18

Note: Event score represents the severity level of the event, and Impression score represents the impression level of the event (explanations on assignment of scores are
provided in Section 2.3.2). The Impression index is calculated as follows: Event score × Impression score. The cumulative impression index is the sum of the impression
indices. For example, in Panel A, for a CEO born in 1917, the age of the CEO ranged from 24 to 28 for Event 1 (World War II), and for Event 2, the age of the CEO
reached 50. At the age of 24, the Impression score is 3, as it is the stage of early adulthood (age 18–25), and the event score is 5 as World War II is the most severe event.
Thus, the Impression index at this point is 15 (Event score × Impression score = 5 × 3).

925
S.-B. Yeoh and C.-W. Hooy Journal of Business Research 139 (2022) 918–934

Fig. 3. Illustration of the composite of the cumulative impression index of each birth year for the respective generation.

from the past is impounded into current expected risk-taking with p time
Table 4
periods. Therefore, p risk-taking lags are sufficient to capture the in­
Descriptive statistics for variable characteristics.
fluence of the firm’s past on the present. If the current CEO trades off the
anticipated costs and benefits of a particular generation, then the cur­ Mean Median S.D. Min Max

rent shock to risk-taking must have been unanticipated when the CEOs Leverage 0.087 0.044 0.111 0 0.591
were chosen. Any information from the firm’s past history that is older Capex 0.035 0.021 0.043 0 0.263
Pre-War II 0.075 0 0.263 0 1
than that event has no direct effect on current risk-taking activities.
Pre-Independence 0.563 1 0.496 0 1
Thus, the firm’s history beyond t-p should be exogenous with respect to Post-Independence 0.344 0 0.475 0 1
any shocks to risk-taking in the current or future periods (Wintoki et al., Development-Era 0.018 0 0.133 0 1
2012). We apply the xtabond2 command in Stata to run the GMM esti­ Tenure 10.549 9 8.531 1 46
mator. The fixed effects are part of the error term of the untransformed CEOAge 54.737 55 9.131 23 96
LogFirmSize 12.957 12.811 1.451 10.005 17.726
level model under this command (Blundell and Bond, 1998). To account
FirmAge 25.607 21 16.219 1 107
for these effects, we choose instrumental variables that are uncorrelated Cash 0.139 0.100 0.126 0.001 0.660
with these unobserved time-invariant effects. SalesGrowth 0.078 0.032 0.460 − 0.916 3.956
IndBoard 0.466 0.444 0.127 0.111 0.857
BoardSize 7.918 8 2.144 3 21
4. Empirical results
Edu 0.632 1 0.482 0 1

4.1. Descriptive statistics and correlations

proportion, which shows that CEOs from the early generation dominate
Table 4 presents the descriptive statistics for the dependent and publicly listed firms in Malaysia. Post-Independence and Development-
control variables used in this study. We show the mean, median, stan­ Era CEOs have shorter tenures compared with the early generation,
dard deviation, minimum and maximum values of the variables. The although the firms that they manage are long-established with an
mean values of leverage and capital expenditure are 8.7% and 3.5% of average age of more than 20 years. This finding indicates that most
total assets, respectively, indicating that Malaysian firms generally un­ Development-Era CEOs are the successors of the firm and inherit the
dertake a low level of corporate risk-taking. business from their predecessors. In the last column of Table 5, we
Table 5 presents the descriptive statistics of the four generations of present the mean test to determine whether the means of the four groups
CEOs. Among these CEOs, 7.5% were born in the Pre-War II period, are different from each other. The results show that the means for the
56.3% were born in the Pre-Independence period, 34.4% were born in nine variables are significantly different from each other. Table 6 pre­
the Post-Independence period, and 1.8% were born in the Development- sents the Pearson’s correlation matrix of the variables used in this study.
Era period. Pre-War II CEOs have the longest tenure, indicating that
many of them still hold their position because they are generally foun­
ders of firms. In addition, pre-independence CEOs made up the largest

926
S.-B. Yeoh and C.-W. Hooy Journal of Business Research 139 (2022) 918–934

Table 5
Comparison of descriptive statistics among generations of CEOs.
Pre-War II Pre-Independence Post-Independence Development-Era Mean

Mean Median Mean Median Mean Median Mean Median test

No. of obs. 460 3473 2125 111

Leverage 0.102 0.043 0.109 0.074 0.099 0.058 0.089 0.042 3.13**
Capex 0.028 0.015 0.037 0.022 0.036 0.021 0.025 0.007 11.78***
Tenure 18.626 16 11.777 10 7.174 5 3.234 3 208.85***
CEOAge 71.137 70 58.390 58 46.416 47 31.802 32 1876.45***
LogFirmSize 13.001 12.860 13.081 12.926 12.746 12.627 12.949 12.923 75.59***
FirmAge 31.024 30 24.744 20 25.537 21 31.505 24 206.2***
Cash 0.155 0.116 0.142 0.104 0.134 0.094 0.108 0.052 22.71***
SalesGrowth 0.062 0.012 0.096 0.031 0.155 0.036 0.072 0.050 2.26*
IndBoard 0.428 0.4 0.463 0.444 0.480 0.455 0.481 0.455 3.18**
BoardSize 8.030 8 8.021 8 7.742 7 7.595 7 1.45

Note: ***, ** and * denote significance at the 1%, 5% and 10% levels, respectively.

4.2. Results Cumulative is positively associated with leverage (β = 0.155) and capital
expenditure (β = 0.009) for Pre-War II CEOs. However, as shown in
To reduce potential biases in the results from extreme values, we columns 3 to 6, Cumulative is negatively related to leverage (β = − 0.020)
winsorize sales growth in the 1st and 99th percentiles. The results of the and capital expenditure (β = − 0.045) for Pre-Independence. A similar
Arellano-Bond and Hansen tests are reported. The insignificant figures negative relationship is also associated with Post-Independence. There is
(at 5%) of the second-order serial correlation (AR(2)) indicate that there no significant relationship between cumulative impression and risk-
is no autocorrelation in the first difference, and the insignificant figures taking for Development-Era CEOs. This finding indicates that as the
(at 5%) of the Hansen test indicate the validity of the instruments. Pre-War II CEOs’ societal experience accumulates, the negative impact
Hypothesis 1 predicts the relationship between the generation effect of new shocks declines and thus leads to higher levels of risk-taking.
and risk-taking. We compare CEOs from other generations with Post- However, the reverse results are found for the Pre-Independence and
Independence CEOs and show the results in Table 7. The relationship Post-Independence CEOs with lower levels of risk-taking.
between Pre-War II and risk-taking is significant and negative (Leverage:
β = − 0.206; Capex: β = − 0.069). This result shows that relative to Post- 4.3. Robustness check with alternative measure of firm risk
Independence CEOs, Pre-War II CEOs are 20.6% less likely to undertake
financial leveraging and 6.9% less likely to spend on capital expendi­ The extent to which risky firm activities actually reflect risk remains
tures. These negative coefficient estimates are the largest among other an open question. According to Bromiley (1991), if a firm’s actions are
generations and support Hypothesis 1a that Pre-War II CEOs undertake legitimate forms of corporate risk-taking, engaging in these activities
the lowest levels of managerial risk-taking. increases firm risk (Bromiley, 1991). Because capital expenditure and
The relationship between Pre-Independence and risk-taking is sig­ financial leverage are discretionary-based firm activities, we conduct a
nificant and negative (Leverage: β = − 0.092; Capex: β = − 0.034). robustness check using firm risk to test the validity and reliability
Therefore, relative to Post-Independence CEOs, Pre-Independence CEOs measures of the construct. We rerun all regressions using firm risk,
are 9.2% and 3.0% less likely to undertake leverage and capital ex­ which is measured as the annualised standard deviation of daily stock
penditures, respectively. This finding also supports Hypothesis 1b, returns during the fiscal year (FirmRisk). Firm risk captures the overall
which states that Pre-Independence CEOs undertake a lower level of risk taken by the firm (Li, Low, & Makhija, 2017). As shown in Table 10
managerial risk-taking compared with Post-Independence. Column 1, the findings demonstrate that Pre-War II CEOs are associated
The relationship with Development-Era is statistically positive, with with the lowest level of firm risk; Development-Era CEOs are associated
a coefficient estimate of 0.089 for Leverage and 0.030 for Capex. This with the highest level of firm risk; and Pre-Independence CEOs are
finding shows that Development-Era CEOs are 8.9% and 3.0% more associated with a lower level of firm risk relative to Post-Independence
likely to undertake leverage and capital expenditures, respectively, CEOs. We next tested the heterogeneity within the same generation. As
compared with Post-Development CEOs. This coefficient is the only shown in Columns 2–5, firms with CEOs who have experienced trau­
positive coefficient estimate, indicating that Development-Era CEOs matic events during their coming-of-age are associated with lower levels
undertake the highest levels of managerial risk-taking, supporting Hy­ of firm risk, except for Development-Era CEOs with an insignificant
pothesis 1c. relationship. We also examine the relationship between the cumulative
Hypothesis 2 predicts the effect of impression of the event. Table 8 event impressions of CEOs and firm risk. The Cumulative is positively
shows that firms managed by CEOs who were between 18 and 25 years associated with Pre-War II CEOs and negatively associated with Pre-
old at the time of the event tend to exhibit a significant negative rela­ Independence and Post-Independence CEOs; moreover, significant re­
tionship with leverage and capital expenditure. As shown in Columns 1 lationships are not found for Development-Era CEOs. These findings are
and 2, Pre-War II CEOs who experienced World War II and May 1969 consistent with our original result.
riots during their coming-of age undertake lower levels of leverage (β =
− 0.021) and capital expenditure (β = − 0.027). Similar results are found
in Pre-Independence CEOs who have experienced May 1969 riots and 4.4. Robustness check to address sample selection bias
Post-Independence CEOs who have experienced Commodity Shock and
Asian Financial Crisis. However, Development-Era CEOs exhibit a Firms that experience CEO turnover may systematically select CEOs
nonsignificant relationship between experience of Global Financial with certain characteristics to execute a strategy. We follow the study by
Crisis during their coming-of-age and level of risk-taking. Thus, our re­ Li et al. (2017) to address this selection bias by deleting newly hired
sults show that all coefficient estimates have expected signs and are CEOs. The intuition is that at the time of the hire, the CEO and the firm
statistically significant, except for Development-Era CEOs. are most well matched. The longer the CEO stays with the firm, the more
Hypothesis 3 predicts the impact of cumulative event impressions. likely it is that the firm’s needs could deviate from the CEO’s expertise so
The findings are reported in Table 9. As shown in columns 1 and 2, that matching between the CEO and the firm’s needs becomes less
perfect. We conduct analysis by restricting the CEOs with tenure of at

927
S.-B. Yeoh and C.-W. Hooy Journal of Business Research 139 (2022) 918–934

least three years. As the Development-Era CEOs are mostly young CEOs

1.000
with tenure of less than three years, we have excluded this category of

14
CEOs from the sample. Due to space limitations, we report only the
coefficients of the main variables. The results remain similar to our
original findings (See Table 11).

− 0.253*
1.000
13

5. Discussion and implication

Our findings reveal that Pre-War II CEOs with war experience

1.000
0.000
0.011
represent the most conservative generation in Malaysia. Pre-
12

Independence CEOs who have struggled through colonisation and

interethnic tension are more focused on loss outcomes and thus under­
take a lower level of risk-taking. In contrast, despite their financial crisis

− 0.057*
0.034*
1.000
0.004
experiences, Post-Independence CEOs who grew up during economic
11

growth perceive lower risk and undertake a higher level of risk-taking.

Development-Era CEOs who experienced robust economic conditions
uphold the openness-to-change value and demonstrate the greatest level
0.096*
0.043*
1.000
0.005
0.018

of risk-taking.
10

Our results show an evolutionary process that has not been addressed
in prior works. We demonstrate how an evolving set of shared beliefs
and values provides cognitive maps for individuals within a generation
0.298*
− 0.050*
0.056*
− 0.056*
0.340*
1.000

to perceive, think and behave. As time passes, new cohorts are added,
9

and new styles of behaviour continue to emerge. Significant shifts in

intergenerational values in terms of risk behaviour are found in the
Note: This table reports the correlations of the variables used in a multivariate analysis; * denotes significance at the 5% level (two-tailed).

emergence of Malaysia as a nation and its subsequent development over

0.086*
0.085*
0.058*
− 0.029*
− 0.082*
0.058*
1.000

time. These values have formed varying orientations and shaped

different risk-taking behaviours among generations of CEOs in Malaysia.
8

Malaysia has gone through different phases of its development

experience with socioeconomic instability before independence, struc­
0.445*
0.058*
0.150*
0.049*
− 0.033*
− 0.083*
− 0.044*
1.000

tural transformation after independence and vulnerability to external

shocks due to open trade policy. By using the rich political and economic
7

environmental events to segment the generation, we unearth a more

precise between-generation effect to explore the dissimilar risk behav­
− 0.116*
− 0.340*

0.049*
− 0.035*
1.000

− 0.001

− 0.002
0.015
− 0.020

iours among CEOs’ generations. Our findings also demonstrate that

political instability and physical insecurity have a more profound
6

adverse effect on the formation of CEOs’ value relative to economic

instability, which is evident in our results that CEOs who have lived in
0.098*
0.287*
0.661*
0.105*

0.031*
0.035*
0.076*
− 0.060*

fear and physical insecurity during political turbulence exhibit a lower

1.000

0.003

level of risk-taking.
5

Our study has important implications for strategic management

−
−
−
−
−
−

research. In the context of risk-taking, the identification of generational

values is relevant for other emerging countries that are in the midst of
− 0.823*
− 0.154*
0.164*
0.454*
0.096*
− 0.060*

− 0.031*
0.055*
1.000

0.020
− 0.022

transformational change. This study’s findings demonstrate that the

risk-taking behaviour of CEOs of earlier generations is essentially lower
4

than that of CEOs of more recent generations. Pre-War II and Pre-

Independence CEOs in Malaysia are mainly the founders of the firm.
− 0.322*
− 0.206*
− 0.038*
0.269*
0.510*

0.095*
0.036*

− 0.086*
1.000

0.009

− 0.022

0.015

They have gone through the traumatic event of World War II and show
conservative business strategies. For example, the founding CEO of
3

Genting Malaysia is well known among his conservative business empire

for rooting the business in one location, namely, Genting Highlands. The
− 0.052*
0.032*

− 0.034*
− 0.041*
− 0.030*
0.122*
− 0.092*

0.047*
− 0.066*
0.104*
1.000

0.005

− 0.015

subsequent CEO, who was born after World War II, undertook a much
more aggressive approach by expanding the business to the overseas
2

market, thus creating a new milestone for the company. Our study ex­
Pearson correlation of explanatory variables.

plains the organisational pursuit behind his management style, and it is

0.114*
0.033*
0.061*
0.043*

0.033*

0.425*
0.053*
0.191*
0.066*

0.145*

consistent with the premise of upper echelon theory that the outcome of
1.000

0.009

0.006

0.002

an organisation can be partially predicted by the characteristics of the

1

−
−
−
−

CEO. From a managerial standpoint, this premise implies that organi­

sations that wish to effect changes in their strategies can derive signif­
icant benefits through the selection of CEOs with demographic attributes
Post-Independence
Pre-Independence

Development-Era

associated with greater openness to change.

SalesGrowth
LogFirmSize

By examining the heterogeneity within a generation, we extend our

Pre-War II

BoardSize
IndBoard
Leverage

FirmAge
CEOAge
Tenure

understanding of early life experience. On the one hand, we complement

Capex

Cash

the stream of studies that focus on the socialisation aspect (but not the
imprinting effect) and another stream of studies that emphasise the
Table 6

imprinting effect (but not the socialisation aspect). On the other hand,
10
11
12
13
14
1
2
3
4
5
6
7
8
9

we advance these prior studies by integrating both aspects. Consistent

928
S.-B. Yeoh and C.-W. Hooy Journal of Business Research 139 (2022) 918–934

Table 7
CEO generation and managerial risk-taking (between-generation effect).
(1) (2)
Leverage Capex

Coefficient p-value 95% confidence interval Coefficient p-value 95% confidence Interval

Lagged 0.708*** 0.000 0.624 0.793 0.253*** 0.000 0.232 0.275

Tenure 0.000 0.377 − 0.001 0.000 0.000 0.982 0.000 0.000
CEOAge 0.006*** 0.001 0.003 0.010 0.002*** 0.000 0.001 0.004
LogFirmSize 0.009*** 0.000 0.006 0.013 0.004*** 0.000 0.003 0.006
LogFirmAge − 0.004 0.148 − 0.010 0.001 − 0.005*** 0.001 − 0.007 − 0.002
Cash − 0.004 0.927 − 0.083 0.076 − 0.007 0.295 − 0.020 0.006
SalesGrowth 0.007** 0.022 0.001 0.013 0.003*** 0.002 0.001 0.004
IndBoard − 0.003 0.859 − 0.032 0.027 − 0.005 0.441 − 0.017 0.007
BoardSize 0.000 0.420 − 0.001 0.002 0.000 0.289 0.000 0.001
DPre-War II − 0.206*** 0.000 − 0.310 − 0.103 − 0.069*** 0.000 − 0.105 − 0.032
DPre-Independence − 0.092*** 0.001 − 0.148 − 0.037 − 0.034*** 0.000 − 0.052 − 0.015
DDevelopment-Era 0.089*** 0.002 0.033 0.146 0.030*** 0.003 0.010 0.051
Constant − 0.359*** 0.000 − 0.135*** 0.000 − 0.200 − 0.069
Industry fixed effects Yes Yes
Year fixed effects Yes Yes
Number of observations 5191 5309
Number of groups 769 769
AR(1) test − 6.921** 0.000 − 7.708** 0.000
AR(2) test − 0.551 0.582 0.641 0.522
Hansen test 73.027 0.839 122.009 0.886

Note: p-values are calculated based on two-step standard errors that adjusted for finite-sample correction to the two-step covariance matrix derived by Windmeijer
(2005). The symbols ***, ** and * denote significance at 1%, 5% and 10%, respectively.

with the impressionable year hypothesis, our findings show that having societal events during their coming-of-age period (Impression dummy of
lived through a traumatic event during the coming-of-age period results 1) and thus undertake lower levels of risk-taking. These findings relate
in a distinctive imprint on the individual’s risk behaviour. This finding back to the underlying impressionable year hypothesis in which events
implies that within the same generation of CEOs, age cannot be a occurring during coming-of-age periods are especially influential to an
judging factor on risk-taking per se because our results show that the individual.
CEOs may or may not carry the generational values, which depends on The findings of this study need to be viewed in the context of certain
whether they experienced the societal event during their impressionable limitations, which in turn suggest some avenues for future research.
period. A younger CEO born in the May 1969 riots may not be as risk Researchers such as Pfeffer (1983) and Finkelstein (1988) strongly
averse as those who are already in their early adulthood during the advocate for the use of demographic data due to objectivity and data
event. availability. Consistent with our reliance on demographic proxies, our
We further test how imprinting unfolds over time because we paper relies exclusively on secondary data sources. Future research that
contend that exposure to new information from different channels can seeks to examine executive cognitions more directly needs to employ a
affect the original imprint. The cumulative event impression results method that can tap more directly into the underlying psychological
showed that imprinting dynamics are supported only by the higher risk- mechanism. Primary data sources, such as questionnaire surveys or in­
taking tendency of Pre-War II CEOs. The imprinting ability of traumatic terviews, should facilitate examinations of how executives influence
events is significantly tempered by subsequent events, which may act as organisational outcomes.
a counterbalancing force, and in our case, the strength is so great that it Our inferences are based on Malaysia, and our findings may only be
reverses the risk-taking propensity from risk aversion to risk taking. The generalised to Southeast Asian countries because of their proximity in
superposition of imprints by which layers of history are deposited in the terms of historical events. We encourage future research to examine this
individual at a few specific points in time (Marquis & Tilcsik, 2013) may generation effect in other countries. In addition, a cohort analysis can
erode the traces of older periods. However, this superposition of im­ also be conducted on other top management team (TMT) members
prints is less salient in Pre-Independence and Post-Independence CEOs because a firm’s corporate policies may not be determined by the CEO
because they have fewer new layers. The events they have experienced alone but by the team (Serfling, 2014). This approach would allow for a
are fewer than those experienced by Pre-War II CEOs, with these richer theoretical and empirical study of executive generation
younger CEOs only experiencing riots and financial crises but not major orientation.
traumatic events, such as war; thus, the cumulative effect may not be as
apparent. In this way, this finding also partly attempts to respond to calls 6. Conclusion
for research on imprint decay at the individual level (Marquis & Tilcsik,
2013). Prior research attempted to associate early-life experience with
Alternatively, Table 2 may illustrate another explanation. Pre-War II managerial decisions by examining one extreme event experienced by a
CEOs who have accumulated stronger impressions, particularly for CEOs specific birth cohort (Chen et al., 2021; Malmendier et al., 2011; Zhang,
who were born from 1928 to 1940 (which accounts for the majority of 2017) and captured the generation effect of value crystallisation but not
Pre-War II CEOs), did not experience the traumatic event during their on the aspect of differentiation between generations. Our findings speak
coming-of-age period (with a higher cumulative impression index of 74 to the importance of generation in ascertaining CEOs’ personal value
and impression dummy of 0), whereas CEOs who were born from 1942 orientations and firms’ subsequent risk-taking. Major shifts in inter­
to 1945 (with a lower cumulative impression index of 48 to 62 but an generational value orientations are found in Malaysia, with high
impression dummy of 1) did experience this trauma. Because the soci­ conservatism in the early generations and high openness-to-change in
etal events did not happen during their impressionable period, they the later generations. In addition, the historical events experienced
undertake higher levels of risk-taking. Similarly, Pre-Independence and during impressionable years are found to have a major impact on risk-
Post-Independence CEOs with a higher impression index experienced taking behaviour. Last, the cumulative effect is only significant for

929
 S.-B. Yeoh and C.-W. Hooy
Table 8
CEO generation and managerial risk-taking: Impression of the event (within-generation effect).
Pre-War II Pre-Independence
(1) (2) (3) (4)
Leverage Capex Leverage Capex

Coefficient p-value 95% confidence Coefficient p-value 95% confidence Coefficient p-value 95% confidence Coefficient p-value 95% confidence
interval interval interval interval

Lagged 0.650*** 0.000 0.605 0.695 0.073 0.219 − 0.044 0.190 0.736*** 0.000 0.661 0.812 0.379*** 0.003 0.125 0.633

Tenure 0.000 0.475 0.000 0.000 0.001** 0.033 0.000 0.002 0.000 0.453 0.000 0.001 0.001 0.345 − 0.002 0.004
CEOAge 0.000 0.400 − 0.001 0.000 − 0.001*** 0.005 − 0.002 0.000 0.000 0.689 − 0.001 0.001 − 0.001 0.432 − 0.002 0.001
LogFirmSize 0.014*** 0.000 0.009 0.019 − 0.018*** 0.007 − 0.032 − 0.005 0.009*** 0.000 0.006 0.012 0.012 0.126 − 0.003 0.028
LogFirmAge − 0.008* 0.087 − 0.017 0.001 − 0.01 0.487 − 0.038 0.018 − 0.002 0.632 − 0.008 0.005 − 0.066** 0.015 − 0.120 − 0.013
Cash − 0.047*** 0.000 − 0.065 − 0.029 − 0.087*** 0.007 − 0.149 − 0.024 0.152** 0.036 0.010 0.293 0.053 0.389 − 0.068 0.175
SalesGrowth − 0.003** 0.018 − 0.006 − 0.001 − 0.01 0.126 − 0.022 0.003 0.012*** 0.002 0.005 0.020 − 0.005 0.772 − 0.036 0.027
IndBoard − 0.058*** 0.000 − 0.082 − 0.033 0.034 0.309 − 0.031 0.098 0.011 0.440 − 0.017 0.038 0.231*** 0.003 0.077 0.385
BoardSize − 0.001*** 0.000 − 0.002 − 0.001 0.002 0.110 0.000 0.004 0.000 0.757 − 0.001 0.002 0.008** 0.046 0.000 0.015
DImpression − 0.021*** 0.0000 − 0.031 − 0.012 − 0.027*** 0.007 − 0.046 − 0.007 − 0.015* 0.092 − 0.031 0.002 − 0.024** 0.020 − 0.044 − 0.004
Constant − 0.071** 0.028 − 0.135 − 0.008 0.349*** 0.000 0.156 0.541 − 0.087** 0.033 − 0.167 − 0.007 − 0.103 0.326 − 0.308 0.102
Industry fixed effects Yes Yes Yes Yes
Year fixed effects Yes Yes Yes Yes
Number of observations 365 378 2907 2953
Number of groups 80 80 513 513
AR(1) test − 2.854** 0.004 − 2.430** 0.015 − 5.813** 0.000 − 3.780** 0.000
AR(2) test 0.729 0.466 − 1.2 0.230 − 0.164 0.870 0.427 0.669
Hansen test 42.702 0.398 20.242 0.855 53.309 0.159 29.797 0.627
930

Post-Independence Development-Era
(5) (6) (7) (8)
Leverage Capex Leverage Capex
Coefficient p-value 95% confidence Coefficient p-value 95% confidence Coefficient p-value 95% confidence Coefficient p-value 95% confidence
interval interval interval interval

Lagged 0.692*** 0.000 0.647 0.736 0.201*** 0.000 0.157 0.245 0.684*** 0.000 0.360 1.008 0.618*** 0.005 0.187 1.050

Tenure 0.001** 0.019 0.000 0.002 0.000 0.524 − 0.001 0.002 − 0.011 0.259 − 0.031 0.008 0.003 0.562 − 0.008 0.014
CEOAge − 0.001* 0.088 − 0.001 0.000 0.002** 0.049 0.000 0.003 0.000 0.949 − 0.006 0.007 − 0.001 0.628 − 0.004 0.003
LogFirmSize 0.009*** 0.000 0.005 0.012 0.005*** 0.000 0.003 0.007 0.001 0.967 − 0.028 0.029 − 0.014 0.254 − 0.037 0.010
LogFirmAge 0.000 0.945 − 0.006 0.005 − 0.006** 0.029 − 0.011 − 0.001 − 0.026 0.150 − 0.061 0.009 − 0.01 0.524 − 0.042 0.021
Cash − 0.029*** 0.003 − 0.047 − 0.010 − 0.007 0.877 − 0.097 0.083 − 0.017 0.845 − 0.191 0.156 − 0.083 0.661 − 0.452 0.287
SalesGrowth 0.001 0.516 − 0.003 0.005 0.001 0.597 − 0.003 0.005 0.019 0.237 − 0.013 0.051 − 0.008 0.481 − 0.030 0.014
IndBoard 0.059*** 0.001 0.023 0.096 0.005 0.769 − 0.031 0.042 − 0.152** 0.013 − 0.273 − 0.032 − 0.101 0.204 − 0.256 0.055

Journal of Business Research 139 (2022) 918–934

BoardSize 0.003*** 0.000 0.001 0.004 0.001 0.159 0.000 0.002 − 0.009* 0.093 − 0.019 0.001 0.000 0.915 − 0.007 0.007
DImpression − 0.014* 0.100 − 0.030 0.003 − 0.019*** 0.005 − 0.032 − 0.006 0.034 0.464 − 0.057 0.125 0.006 0.892 − 0.077 0.089
Constant − 0.014 0.585 − 0.066 0.037 − 0.065* 0.087 − 0.139 0.009 0.151 0.627 − 0.458 0.759 0.326 0.189 − 0.161 0.812
Industry fixed effects Yes Yes Yes Yes
Year fixed effects Yes Yes Yes Yes
Number of observations 1818 1872 101 106
Number of groups 341 345 29 30
AR(1) test − 5.012** 0.000 − 4.882** 0.000 − 2.570** 0.010 − 2.544** 0.011
AR(2) test − 1.679 0.093 0.799 0.424 − 1.32 0.187 − 0.728 0.467
Hansen test 67.41 0.163 57.287 0.223 4.612 1.000 5.398 1.000

Note: p-values are calculated based on two-step standard errors that adjusted for finite-sample correction to the two-step covariance matrix derived by Windmeijer (2005). The symbols ***, ** and * denote significance at
1%, 5% and 10%, respectively.
 S.-B. Yeoh and C.-W. Hooy
Table 9
CEO generation and managerial risk-taking: Cumulative impression of events (within-generation effect).
Pre-War II Pre-Independence
(1) (2) (3) (4)
Leverage Capex Leverage Capex
Coefficient p-value 95% confidence Coefficient p-value 95% confidence Coefficient p-value 95% confidence Coefficient p-value 95% confidence
interval interval interval interval

Lagged 0.403*** 0.000 0.248 0.558 0.162*** 0.000 0.116 0.207 0.788*** 0.000 0.751 0.826 0.230*** 0.001 0.095 0.365

Tenure − 0.001** 0.028 − 0.002 0.000 0.000* 0.057 0.000 0.001 0.000 0.476 − 0.001 0.001 0.000 0.167 0.000 0.000
CEOAge − 0.014*** 0.000 − 0.021 − 0.008 − 0.001*** 0.002 − 0.002 0.000 0.000 0.427 − 0.001 0.002 0.001 0.16 − 0.001 0.003
LogFirmSize 0.042*** 0.002 0.016 0.068 − 0.009*** 0.000 − 0.013 − 0.004 0.002 0.239 − 0.002 0.006 0.003*** 0.004 0.001 0.005
LogFirmAge − 0.070*** 0.005 − 0.119 − 0.021 − 0.015** 0.014 − 0.027 − 0.003 0.009 0.432 − 0.014 0.032 − 0.001 0.569 − 0.005 0.003
Cash − 0.138*** 0.009 − 0.242 − 0.034 0.006 0.566 − 0.015 0.028 − 0.018 0.660 − 0.097 0.061 − 0.017 0.434 − 0.061 0.026
SalesGrowth 0.000 0.979 − 0.007 0.006 0.000 0.933 − 0.006 0.007 − 0.003 0.748 − 0.019 0.014 0.004** 0.04 0.000 0.009
IndBoard − 0.127*** 0.006 − 0.217 − 0.037 0.126*** 0.000 0.105 0.146 − 0.029 0.453 − 0.105 0.047 − 0.008 0.291 − 0.022 0.006
BoardSize 0.007*** 0.004 0.002 0.012 − 0.001 0.220 − 0.002 0.000 0.002 0.312 − 0.002 0.007 0.001** 0.027 0.000 0.002
Cumulative 0.155*** 0.000 0.082 0.228 0.009** 0.029 0.001 0.018 − 0.020* 0.099 − 0.043 0.004 − 0.045*** 0.007 − 0.077 − 0.012
Constant − 0.102 0.580 − 0.465 0.260 0.138*** 0.000 0.061 0.216 − 0.021 0.532 − 0.086 0.044 − 0.020 0.715 − 0.129 0.088
Industry fixed effects Yes Yes Yes Yes
Year fixed effects Yes Yes Yes Yes
Number of observations 365 378 2907 2953
Number of groups 80 80 513 513
AR(1) test − 2.007** 0.045 − 2.470** 0.013 − 7.031** 0.000 − 5.363** 0.000
AR(2) test − 0.603 0.546 − 1.529 0.126 0.053 0.958 − 0.442 0.659
Hansen test 18.357 0.685 43.732 0.999 43.564 0.212 43.248 0.160
931

Post-Independence Development-Era
(5) (6) (7) (8)
Leverage Capex Leverage Capex
Coefficient p-value 95% confidence Coefficient p-value 95% confidence Coefficient p-value 95% confidence Coefficient p-value 95% confidence
interval interval interval interval

Lagged 0.476*** 0.000 0.367 0.585 0.208*** 0.000 0.140 0.276 0.705*** 0.000 0.540 0.870 0.547*** 0.000 0.317 0.778
Tenure − 0.001 0.352 − 0.004 0.001 0.002 0.205 − 0.001 0.004 0.003 0.592 − 0.008 0.013 − 0.003 0.640 − 0.014 0.009
CEOAge 0.005** 0.016 0.001 0.009 0.003* 0.067 0.000 0.006 0.003 0.496 − 0.005 0.011 0.001 0.733 − 0.004 0.005
LogFirmSize 0.012*** 0.001 0.005 0.019 0.008 0.286 − 0.007 0.023 0.01 0.487 − 0.018 0.038 0.006 0.282 − 0.005 0.016
LogFirmAge − 0.007 0.222 − 0.018 0.004 − 0.005 0.600 − 0.025 0.014 − 0.007 0.842 − 0.075 0.061 0.011 0.497 − 0.020 0.042
Cash − 0.059*** 0.009 − 0.103 − 0.014 0.05 0.390 − 0.064 0.164 − 0.117 0.388 − 0.382 0.149 0.01 0.742 − 0.049 0.069
SalesGrowth 0.003 0.322 − 0.003 0.010 − 0.004 0.362 − 0.012 0.004 0.006 0.796 − 0.038 0.049 − 0.008 0.238 − 0.021 0.005
IndBoard 0.004 0.913 − 0.074 0.083 0.038 0.295 − 0.033 0.110 − 0.083 0.365 − 0.264 0.097 − 0.027 0.810 − 0.244 0.191
BoardSize 0.002 0.106 0.000 0.005 0.000 0.979 − 0.004 0.004 0.001 0.801 − 0.008 0.010 − 0.001 0.883 − 0.010 0.008

Journal of Business Research 139 (2022) 918–934

Cumulative − 0.045*** 0.009 − 0.079 − 0.011 − 0.021** 0.043 − 0.041 − 0.001 − 0.085 0.516 − 0.343 0.172 0.021 0.753 − 0.111 0.153
Constant − 0.222*** 0.006 − 0.381 − 0.063 − 0.090 0.578 − 0.405 0.226 − 0.372 0.506 − 1.466 0.723 − 0.134 0.387 − 0.438 0.170
Industry fixed effects Yes Yes Yes Yes
Year fixed effects Yes Yes Yes Yes
Number of observations 1818 1872 101 106
Number of groups 341 345 29 30
AR(1) test − 5.206** 0.000 − 6.121** 0.000 − 1.775** 0.076 − 1.665** 0.096
AR(2) test − 1.72 0.085 1.152 0.249 0.319 0.750 − 0.24 0.810
Hansen test 48.541 0.451 39.578 0.489 5.989 1.000 5.409 1.000

Note: p-values are calculated based on two-step standard errors that adjusted for finite-sample correction to the two-step covariance matrix derived by Windmeijer (2005). The symbols ***, ** and * denote significance at
1%, 5% and 10%, respectively.
 S.-B. Yeoh and C.-W. Hooy
Table 10
CEO generation and managerial risk-taking: Using firm risk (FirmRisk) as a robustness test to reproduce the results in Table 7, Table 8 and Table 9.
(1) (2) (3) (4) (5)

FirmRisk Pre-War II Pre-Independence Post-Independence Development-Era

Coefficient p-value 95% confidence Coefficient p-value 95% confidence Coefficient p-value 95% confidence Coefficient p-value 95% confidence Coefficient p-value 95% confidence
interval interval interval interval interval
Lagged 0.283*** 0.000 0.215 0.351 0.108** 0.046 0.002 0.214 0.132*** 0.003 0.043 0.220 0.062** 0.034 0.005 0.120 0.535 0.162 − 0.214 1.285
Tenure 0 0.997 0.000 0.000 0.000 0.785 0.000 0.000 0.000 0.287 0.000 0.000 0.000 0.945 0.000 0.000 − 0.001 0.674 − 0.005 0.003
CEOAge 0.000* 0.091 0.000 0.001 0.000 0.977 0.000 0.000 − 0.000* 0.092 0.000 0.000 0.000* 0.097 0.000 0.001 0.000 0.154 0.000 0.001
LogFirmSize − 0.004*** 0.000 − 0.004 − 0.003 − 0.011*** 0.000 − 0.016 − 0.007 − 0.004*** 0.000 − 0.005 − 0.004 − 0.006*** 0.000 − 0.009 − 0.003 − 0.003 0.331 − 0.009 0.003
LogFirmAge 0.001 0.347 − 0.001 0.002 0.013*** 0.003 0.004 0.021 0.000 0.573 − 0.001 0.002 0.001 0.590 − 0.002 0.003 0.004 0.196 − 0.002 0.011
Cash − 0.014*** 0.007 − 0.025 − 0.004 − 0.012 0.304 − 0.035 0.011 − 0.012 0.104 − 0.027 0.003 − 0.054*** 0.002 − 0.088 − 0.021 0.014 0.776 − 0.082 0.110
SalesGrowth − 0.001*** 0.001 − 0.002 − 0.001 0.002 0.136 − 0.001 0.004 − 0.003*** 0.000 − 0.004 − 0.002 0.000 0.732 − 0.001 0.001 0.004 0.329 − 0.004 0.012
IndBoard 0.009* 0.052 0.000 0.019 0.007 0.432 − 0.011 0.025 0.000 0.937 − 0.006 0.006 0.019*** 0.002 0.007 0.031 0.014 0.443 − 0.021 0.048
BoardSize 0 0.940 0.000 0.000 0.000 0.481 − 0.001 0.001 0.000 0.713 0.000 0.000 0.000 0.546 0.000 0.001 0.000 0.831 − 0.002 0.001
DPre-War II − 0.019* 0.091 − 0.041 0.003
DPre-Independence − 0.006* 0.099 − 0.014 0.001
DDevelopment-Era 0.009** 0.099 0.001 0.017
DImpression − 0.009** 0.019 − 0.016 − 0.001 − 0.005*** 0.001 − 0.008 − 0.002 − 0.006** 0.037 − 0.011 0.000 − 0.002 0.833 − 0.023 0.019
Constant 0.048*** 0.000 0.023 0.073 0.136*** 0.000 0.072 0.201 0.094*** 0.000 0.076 0.113 0.088*** 0.000 0.056 0.119 − 0.060 0.544 − 0.252 0.133
Industry fixed effects Yes Yes Yes Yes Yes
Year fixed effects Yes Yes Yes Yes Yes
No. of observations 5275 377 2936 1857 105
Number of groups 766 79 510 345 30
AR(1) test − 5.054** 0.000 − 2.435** 0.015 − 3.755** 0.000 − 3.731** 0.000 − 2.031** 0.042
AR(2) test 1.081 0.280 − 0.433 0.665 − 0.08 0.936 − 0.367 0.714 0.692 0.489
932

Hansen test 83.203 0.294 25.79 0.934 57.729 0.339 52.763 0.331 2.657 1.000

(6) (7) (8) (9)

Pre-War II Pre -Independence Post-Independence Development-Era

Coefficient p-value 95% confidence interval Coefficient p-value 95% confidence interval Coefficient p-value 95% confidence interval Coefficient p-value 95% confidence interval
Lagged 0.292*** 0.000 0.242 0.342 0.449*** 0.000 0.363 0.535 0.045 0.134 − 0.014 0.103 0.412 0.350 − 0.451 1.275
Tenure 0.000 0.675 0.000 0.000 0.000 0.612 0.000 0.000 0.000 0.571 0.000 0.001 0.000 0.944 − 0.002 0.002
CEOAge − 0.000*** 0.007 − 0.001 0.000 0.000 0.671 0.000 0.000 0.000* 0.063 0.000 0.001 − 0.001 0.734 − 0.004 0.003
LogFirmSize − 0.012*** 0.000 − 0.014 − 0.010 − 0.003*** 0.000 − 0.004 − 0.002 − 0.004** 0.016 − 0.008 − 0.001 − 0.003 0.702 − 0.016 0.011
LogFirmAge 0.010*** 0.000 0.007 0.014 0.001 0.171 0.000 0.002 0.000 0.980 − 0.003 0.003 − 0.003 0.683 − 0.017 0.011
Cash − 0.037*** 0.000 − 0.052 − 0.022 − 0.037*** 0.000 − 0.055 − 0.018 − 0.036*** 0.006 − 0.062 − 0.011 − 0.008 0.879 − 0.112 0.096
SalesGrowth 0.002** 0.013 0.000 0.003 − 0.003*** 0.000 − 0.004 − 0.002 0.000 0.915 − 0.001 0.001 0.005 0.274 − 0.004 0.013
IndBoard 0.015*** 0.007 0.004 0.026 − 0.002 0.312 − 0.007 0.002 0.020*** 0.003 0.007 0.033 0.041 0.121 − 0.011 0.092
BoardSize 0.000 0.112 0.000 0.001 0.000 0.142 0.000 0.000 0.000 0.57 − 0.001 0.001 0.000 0.974 − 0.001 0.001
Cumulative 0.005*** 0.005 0.002 0.008 − 0.012** 0.017 − 0.022 − 0.002 − 0.007*** 0.004 − 0.011 − 0.002 0.025 0.582 − 0.065 0.115

Journal of Business Research 139 (2022) 918–934

Constant 0.148*** 0.000 0.106 0.189 0.088*** 0.000 0.067 0.109 0.070*** 0.001 0.030 0.110 0.255* 0.061 − 0.011 0.522
Industry fixed effects Yes Yes Yes Yes
Year fixed effects Yes Yes Yes Yes
No. of observations 377 2936 1857 105
Number of groups 79 510 345 30
AR(1) test − 2.935** 0.003 − 4.086** 0.000 − 3.473** 0.001 − 2.748** 0.006
AR(2) test − 1.043 0.297 0.914 0.360 − 0.273 0.785 1.512 0.131
Hansen test 42.569 0.656 55.433 0.161 59.237 0.150 7.137 1.000

Note: p-values are calculated based on two-step standard errors that adjusted for finite-sample correction to the two-step covariance matrix derived by Windmeijer (2005). The symbols ***, ** and * denote significance at
1%, 5% and 10%, respectively.
S.-B. Yeoh and C.-W. Hooy Journal of Business Research 139 (2022) 918–934

Table 11
CEO generation and managerial risk-taking: robustness test to address selection bias.
(1) (2)
Leverage Capex

Coefficient p-value 95% confidence Coefficient p-value 95% confidence

interval interval

Between-generation effect
DPre-War II − 0.590* 0.056 − 1.195 0.014 − 0.113** 0.012 − 0.201 − 0.025
DPre-Independence − 0.451* 0.061 − 0.923 0.021 − 0.051*** 0.007 − 0.088 − 0.014
DDevelopment-Era 1.729 0.305 − 1.575 5.034 0.095 0.674 − 0.347 0.536
Within-generation effect (Impression of the event)
Pre-War II
DImpression − 0.036*** 0.000 − 0.050 − 0.023 − 0.008*** 0.001 − 0.012 − 0.003
Pre-Independence
DImpression − 0.018*** 0.000 − 0.027 − 0.009 − 0.010** 0.025 − 0.019 − 0.001
Post-Independence
DImpression − 0.026* 0.073 − 0.055 0.002 − 0.035** 0.037 − 0.068 − 0.002
Within-generation effect (Cumulative impression of events)
Pre-War II
Cumulative 0.075*** 0.000 0.038 0.113 0.026** 0.026 0.003 0.049
Pre-Independence
Cumulative − 0.034*** 0.007 − 0.059 − 0.010 − 0.053* 0.096 − 0.115 0.009
Post-Independence
Cumulative − 0.044** 0.017 − 0.080 − 0.008 − 0.026* 0.058 − 0.053 0.001

Note: p-values are calculated based on two-step standard errors that adjusted for finite-sample correction to the two-step covariance matrix derived by Windmeijer
(2005). The symbols ***, ** and * denote significance at 1%, 5% and 10%, respectively.

CEOs that have gone through World War II during their impressionable Chen, Y., Fan, Q., Yang, X., & Zolotoy, L. (2021). CEO early-life disaster experience and
stock price crash risk. Journal of Corporate Finance, 68, 101928. https://doi.org/
year because new layers of history may erode the traces of old years. Our
10.1016/j.jcorpfin.2021.101928
study demonstrates that segmenting a generation using a mix of political Child, J. (1974). Managerial and organizational factors associated with company
and economic factors can be performed to identify the distinctive effect performance part I. Journal of Management Studies, 11(3), 175–189.
of both factors and thereby provide new insights into cross-country Chin, M. K., Hambrick, D. C., & Treviño, L. K. (2013). Political ideologies of CEOs: The
influence of executives’ values on corporate social responsibility. Administrative
research with significantly different political and economic histories. Science Quarterly, 58(2), 197–232.
Funding: Acknowledgement to Ministry of Higher Education Coles, J., Daniel, N., & Naveen, L. (2006). Managerial incentives and risk-taking. Journal
Malaysia for Research University (Individual) Grant with Project Code: of Financial Economics, 79(2), 431–468.
Cox, A., Hannif, Z., & Rowley, C. (2014). Leadership styles and generational effects:
1001/PMGT/8016091 Examples of US companies in Vietnam. The International Journal of Human Resource
Management, 25(1), 1–22.
Declaration of Competing Interest Cutler, N. E. (1974). Aging and generations in politics: The conflict of explanations and
inference. In A. R. Wilcox (Ed.), Public opinion and political attitudes. New York:
Wiley.
The authors declare that they have no known competing financial De Meuse, K. P., Bergmann, T. J., & Lester, S. W. (2001). An investigation of the
interests or personal relationships that could have appeared to influence relational component of the psychological contract across time, generation, and
employment status. Journal of Managerial Issues, 102–118.
the work reported in this paper. Egri, C. P., & Ralston, D. A. (2004). Generation cohorts and personal values: A
comparison of China and the United States. Organization Science, 15(2), 210–220.
References Eichengreen, B., Aksoy, C. G., & Saka, O. (2021). Revenge of the experts: Will COVID-19
renew or diminish public trust in science? Journal of Public Economics, 193, 104343.
https://doi.org/10.1016/j.jpubeco.2020.104343
Akbulut-Yuksel, M. (2014). Children of war the long-run effects of large-scale physical
England, G. W. (1967). Personal value systems of American managers. Academy of
destruction and warfare on children. Journal of Human Resources, 49(3), 634–662.
Management Journal, 10(1), 53–68.
Akerlof, G. A. (1983). Loyalty filters. The American Economic Review, 73(1), 54–63.
Erikson, E. H. (1950). Childhood and society. New York: Norton.
Albu, N., Albu, C. N., Apostol, O., & Cho, C. H. (2021). The past is never dead: The role of
Finkelstein, S. (1988). Managerial orientations and organizational outcomes: The moderating
imprints in shaping social and environmental reporting in a post-communist context.
roles of managerial discretion and power. PhD dissertation. New York: Columbia
Accounting, Auditing & Accountability Journal, 34(5), 1109–1136.
University.
Alwin, D. F., & McCammon, R. J. (2003). Generations, cohorts, and social change. In
Freudenberg, M. (2003). Composite indicators of country performance: a critical
J. T. Mortimer, & M. J. Shanahan (Eds.), Handbook of the life course (pp. 23–49). New
assessment. OECD Science, Technology and Industry Working Papers, 2003/16.
York: Kluwer Academic/Plenum.
OECD, Paris.
Athukorala, P. (2013). The Malaysian economy during three crises. In H. Hill, T. S. Yean,
Giuliano, P., & Spilimbergo, A. (2009). Growing up in a recession: Beliefs and the
& R. H. M. Zin (Eds.), Malaysia’s Development Challenges: Graduating from the middle
macroeconomy. National Bureau of Economic Research (NBER) Working Paper
(pp. 83–105). Abingdon: Routledge.
15321.
Negara, B. (2008). Financial Stability and Payments System Report 2007. Kuala Lumpur:
Guo, S., Zan, B., Sun, Y., & Zhang, M. (2020). Effects of top managers’ military
Bank Negara Malaysia.
experience on technological innovation in the transition economies of China.
Benischke, M. H., Martin, G. P., & Glaser, L. (2019). CEO equity risk bearing and strategic
Technological Forecasting and Social Change, 153, 119909. https://doi.org/10.1016/j.
risk taking: The moderating effect of CEO personality. Strategic Management Journal,
techfore.2020.119909
40(1), 153–177.
Hall, T. W. (2012). The collateral channel: Evidence on leverage and asset tangibility.
Benmelech, E., & Frydman, C. (2015). Military CEOs. Journal of Financial Economics, 117
Journal of Corporate Finance, 18(3), 570–583.
(1), 43–59.
Hambrick, D. C. (2007). Upper echelons theory: An update. Academy of Management
Blundell, R., & Bond, S. (1998). Initial conditions and moment restrictions in dynamic
Review, 32(2), 334–343.
panel data models. Journal of Econometrics, 87(1), 115–143.
Hambrick, D. C., & Mason, P. A. (1984). Upper echelons: The organization as a reflection
Bromiley, P. (1991). Testing a causal model of corporate risk taking and performance.
of its top managers. Academy of Management Review, 9(2), 193–206.
Academy of Management Journal, 34(1), 37–59.
Hoskisson, R. E., Chirico, F., Zyung, J.(., & Gambeta, E. (2017). Managerial risk taking: A
Carlsson, G., & Karlsson, K. (1970). Age, cohorts and the generation of generations.
multitheoretical review and future research agenda. Journal of Management, 43(1),
American Sociological Review, 35(4), 710. https://doi.org/10.2307/2093946
137–169.
Chen, R., Wong, K. A., & Lee, H. C. (2001). Age, period, and cohort effects on life
Inglehart, R. (1997). Modernization and postmodernization: Cultural, economic, and political
insurance purchases in the US. Journal of Risk and Insurance, 68(2), 303. https://doi.
change in 43 societies. Princeton, NJ: Princeton University Press.
org/10.2307/2678104
Jomo, K. S. (1990). Whither Malaysia’s new economic policy? Pacific Affairs, 63(4),
469–499. https://doi.org/10.2307/2759912

933
S.-B. Yeoh and C.-W. Hooy Journal of Business Research 139 (2022) 918–934

Kim, Y.-I., & Lee, J. (2014). The long-run impact of a traumatic experience on risk Schwartz, S. H., & Sagie, G. (2000). Value consensus and importance: A cross-national
aversion. Journal of Economic Behavior & Organization, 108, 174–186. study. Journal of Cross-Cultural Psychology, 31(4), 465–497.
Kish-Gephart, J. J., & Campbell, J. T. (2015). You don’t forget your roots: The influence Serfling, M. A. (2014). CEO age and the riskiness of corporate policies. Journal of
of CEO social class background on strategic risk taking. Academy of Management Corporate Finance, 25, 251–273.
Journal, 58(6), 1614–1636. Simon, H. A. (1978). Information-processing theory of human problem solving.
Kotler, P., & Keller, K. (2006). Marketing Management. Upper Saddle River, NJ: Prentice Handbook of learning and cognitive processes, 5, 271–295.
Hall. Simsek, Z., Fox, B. C., & Heavey, C. (2015). “What’s past is prologue” A framework,
Krosnick, J. A., & Alwin, D. F. (1989). Aging and susceptibility to attitude change. review, and future directions for organizational research on imprinting. Journal of
Journal of Personality and Social Psychology, 57(3), 416–425. Management, 41(1), 288–317.
Li, X., Low, A., & Makhija, A. K. (2017). Career concerns and the busy life of the young Sitkin, S. B., & Pablo, A. L. (1992). Reconceptualizing the determinants of risk behavior.
CEO. Journal of Corporate Finance, 47, 88–109. Academy of Management Review, 17(1), 9–38.
Lu, J., & Wang, W. (2018). Managerial conservatism, board independence and corporate Sitkin, S. B., & Weingart, L. R. (1995). Determinants of risky decision-making behavior: A
innovation. Journal of Corporate Finance, 48, 1–16. test of the mediating role of risk perceptions and propensity. Academy of Management
Malmendier, U., & Nagel, S. (2011). Depression babies: Do macroeconomic experiences Journal, 38(6), 1573–1592.
affect risk taking? The Quarterly Journal of Economics, 126(1), 373–416. Strauss, W., & Howe, N. (1991). Generations: The history of America’s future, 1584 to 2069.
Malmendier, U., Tate, G., & Yan, J. (2011). Overconfidence and early-life experiences: New York: William Morrow & Co.
The effect of managerial traits on corporate financial policies. The Journal of Finance, Thau, R. D., & Heflin, J. S. (1997). Generations Apart: Xers Versus Boomers Versus the
66(5), 1687–1733. Elderly. Amherst, NY: Prometheus Books.
March, J. G., & Shapira, Z. (1987). Managerial perspectives on risk and risk taking. Weber, E. U., & Milliman, R. A. (1997). Perceived risk attitudes: Relating risk perception
Management Science, 33(11), 1404–1418. to risky choice. Management Science, 43(2), 123–144.
Marquis, C., & Tilcsik, A. (2013). Imprinting: Toward a multilevel theory. Academy of Weick, K. E. (1979). The social psychology of organizing. New York, NY: Addison-Wesley.
Management Annals, 7(1), 195–245. Windmeijer, F. (2005). A finite sample correction for the variance of linear efficient two-
Marquis, C., & Qiao, K. (2020). Waking from Mao’s dream: Communist ideological step GMM estimators. Journal of Econometrics, 126(1), 25–51.
imprinting and the internationalization of entrepreneurial ventures in China. Wintoki, M. B., Linck, J. S., & Netter, J. M. (2012). Endogeneity and the dynamics of
Administrative Science Quarterly, 65(3), 795–830. internal corporate governance. Journal of Financial Economics, 105(3), 581–606.
Newcomb, T. M. (1943). Personality and Social Change: Attitude Formation in a Student Yao, R., Sharpe, D. L., & Wang, F. (2011). Decomposing the age effect on risk tolerance.
Community. New York: Dryden Press. The Journal of Socio-Economics, 40(6), 879–887.
Pfeffer, J. (1983). Organizational demography. In L. L. Cummings, & B. M. Staw (Eds.), Yusoff, M. B., Hasan, F. A., & Jalil, S. A. (2000). Globalisation, economic policy and
Research in Organizational Behavior (pp. 299–357). Greenwich, CT: JAI Press. equity: The case of Malaysia. Poverty and Income Inequality in Developing Countries: A
Rashad Abdel-Khalik, A. (2014). CEO risk preference and investing in R & D. Abacus, 50 Policy Dialogue on the Effects of Globalisation.
(3), 245–278. Zhang, L.e. (2017). CEOs’ early-life experiences and corporate policy: Evidence from
Rhodes, S. R. (1983). Age-related differences in work attitudes and behavior: A review China’s great famine. Pacific-Basin Finance Journal, 46, 57–77.
and conceptual analysis. Psychological Bulletin, 93(2), 328–367.
Rubin, D. C. (1988). Autobiographical memory. Cambridge, UK: Cambridge University
Siew-Boey, Yeoh is a PhD holder of Finance at the School of Management, Universiti Sains
Press.
Malaysia. Her research interest is in the areas of corporate finance and corporate
Schewe, C. D., & Meredith, G. (2004). Segmenting global markets by generational
governance.
cohorts: Determining motivations by age. Journal of Consumer Behaviour: An
International Research Review, 4(1), 51–63.
Schewe, C. D., & Noble, S. M. (2000). Market segmentation by cohorts: The value and Chee-Wooi, Hooy is a Professor of Finance at the School of Management, Universiti Sains
validity of cohorts in America and abroad. Journal of Marketing Management, 16(1-3), Malaysia. His current research interest is in international finance and corporate gover­
129–142. nance. He has publication in, among others, British Accounting Review, Emerging Market
Schuman, H., & Scott, J. (1989). Generations and collective memories. American Review, International Finance, Pacific Basin Finance Journal, and The North America Journal
Sociological Review, 54(3), 359–381. https://doi.org/10.2307/2095611 of Economics and Finance.

934