Master Thesis Organizational Management & Control

(1)

Master Thesis

Organizational Management & Control

Operationalizing culture in business research.

A comparison between Hofstede and the GLOBE study

.

(2)

2 Abstract

The usefulness of Hofstede‟s and GLOBE‟s cultural dimensions to examine the cultural impact on business has been discussed in prior research. However, the impact of the choice of theory is little discussed. The findings of a comparison between the use of Hofstede and Globe in a research in the field of working capital management to determine what the impact of culture is on the working capital management practices will be presented. By executing empirical analyses there is found that culture has a moderating effect on the relation between liquidity, measured by the cash conversion cycle, and firm performance, measured by the return on equity. However, this moderating effect deviates by the way culture is

operationalized. This confirms that the view on cultures can influence the outcomes of cross-cultural research.

(3)

3

countries become stronger. Despite this, national culture is often used to explain differences, not only in the field of business but also in other fields of research. Several persons have tried to define and operationalize national cultures of which Hofstede‟s cultural dimensions are the most widely used. But there are cultural theories which differ from Hofstede‟s theory. Trompenaars and Hampden-Turner (1997) give an alternative operationalization of culture as well as Hall (1997). These cultural theories are less used to study the influence of national culture but what if this was not the case? Do the relations found and explained using

Hofstede‟s cultural dimensions still hold if one of the other theories was used? For example, Chang and Noorbakhsh (2009) found a relationship between corporate cash holdings and liquid balances in countries where the people tend to avoid uncertainty, are culturally more masculine, and have longer term orientation. In their study they used Hofstede‟s cultural dimensions, but does the relation between national culture and international corporate cash holdings found by Chang and Noorbakhsh (2009) still hold if they used another

operationalizing of culture?

In the field of finance, the influence of culture on different topics is often researched. Within the management disciplines, the causal–national–culture accepting literature justifies its reliance on the notion of national culture by citing approvingly the work of Geert Hofstede, who claims to have successfully „uncovered the secrets of entire national cultures‟.

(McSweeney, 2002)

(5)

5 McSweeney (2002) does not present an alternative concept of culture but critzices the emperical study Hofstede performend to develop his cultural dimensions. He questions the plausibility of systematically causal national cultures. He also discusses the methodology and the assumptions Hofstede used to derive his theory.

Hofstede and Trompenaar have also critizesed each other‟s work. In a paper Hofstede published in 1993 he examines the methodology and results Trompenaars used for his book

Riding The Waves Of Culture and concludes that only two of the seven dimensions can be

proven statistically. Otherwise, Trompenaars et al. (1996) suggest that the dimensions defined by Hofstede as individualism-collectivism and power distance may be better defined as representing varying orientations toward continuity of group membership (loyal involvement/ utilitarian involvement) and varying orientations toward the obligations of social relationship (conservatism/egalitarian commitment). In 1991, House conceived his Global Leadership and Organizational Behavior Effectiveness study. In 2004, its first comprehensive volume on "Culture, Leadership, and Organizations: The GLOBE Study of 62 Societies" was published, based on results from about 17,300 middle managers from 951 organizations in the food processing, financial services, and telecommunications services industries. GLOBE's major premise (and finding) is that leader effectiveness is contextual, that is, it is embedded in the societal and organizational norms, values, and beliefs of the people being led. GLOBE empirically established nine cultural dimensions that make it possible to capture the

similarities and/or differences in norms, values, beliefs, and practices, among societies. They build on findings by Hofstede (1980), Schwartz (1994), Smith (1995), Inglehart (1997), and others. (Hoppe, 2007)

This research focuses on getting an insight of the influence the choice of operatinalizing culture can have on the outcomes of a research. To get an insight, the influence of culture on the relation between liquidity and firm performance will be explored. By examining this relation with differend ways to operationalize culture, an insight can be given.

The main research question of this thesis will be:

(6)

6 The purpose of this research is to contribute to the discussion of how to operationalize culture especially in the field of finance. Secondly, it will contribute to the understanding of the relation between culture and practices in the field of working capital management although this is not the main goal of this research. This research can be relevant for

researchers who are active in making cross-cultural comparisons in the field of finance. There will be three matching dimensions from Hofstede and the GLOBE study used. Because there is no correlation found between the matching dimension of Hofstede and GLOBE this can give some interesting results. For managers this study will be relevant because it will give an insight in how culture influences their working capital management practices. Studies have found relations between working capital practices and firm performance. These studies were often on national level. (Deloof, 2003; Raheman and Nasr, 2007) In this study, three relevant cultural dimensions of Hofstede and its counterparts of GLOBE‟s study will be used. Via several hypotheses based on the relation between liquidity and firm performance a comparison will be made.

An empirical analysis will be made with the data from more than 25,000 companies from 21 countries over the period of 2006 -2010.

(7)

7

2. Literature review

To answer the research question we first need to define several variables and relationships between the variables. In this part we will describe performance and in which wayss it can be measured. The ultimate goal of the most firms is maximizing of shareholder profit. There will be a small introduction in how this can be measured. Thereafter liquidity will be discussed. Liquidity can be approached in different ways. Several ways will be discussed and the choice for the cash conversion cycle will be motivated. Then the concept of culture will be explained and diverse theory‟s like Hofstede‟s (1993), Trompenaars and Hampden-Turner‟s (1998) and the one from Hall (1976) will be debated. In the last sections of this part the relations between these variables will be discussed and the hypotheses will be presented.

2.1 Performance

It is generally agreed that the primary objective of for-profit organizations is to maximize the value of the firm (or ownership wealth), firm value for short. Performance can be

measured with financial and non-financial measures. Financial measures can be categorized by several different types of measures. These are summary (single-number and aggregate), and bottom-line financial measures. Ratio terms like return on investment, return on equity and return on net assets do also belong to the summary financial measures group. (Merchant et al., 2007)

Financial measures (ratios) will be used to evaluate firm performance in this study. Using a ratio has some significant advantages. First, a ratio can be measured on a timely basis,

relatively precisely and objectively. Secondly, compared to other measures that can be measured on a timely basis, precisely and objectively ratios are relatively congruent with the organizational goal of maximizing profit. (Merchant et al., 2007) Third, accounting measures are understandable (Merchant et al., 2007) and used in financial statements which are used by managers, creditors, stockholders, potential investors and regulatory agencies. (Edmonds et al., 2009) Because of these advantages the comparison of different companies, based on ratios, is reliable.

(8)

8 the ROA (return on Assets) is the influence of the capital structure, which is only visible in the return on equity (Jose et al., 1996). Because the expected influence of culture on the capital structure of a firm, return on equity will be used in this study.

2.2 Liquidity

Liquidity can be measured in several ways, traditionally it is done with the current ratio, quick ratio, net working capital and the ratio of net working capital to current liabilities. These ratios are based on static balance sheet entries. These ratios reflect the extent to which the firms currently maturing liabilities are covered by currently maturing assets. (Jose, 2006) Logue and Merville (1972) consider the current ratio as an important measure for liquidity because it is widely understood and has more intuitive appeal than other measures. However, a generalization like this should be viewed with caution according to Richards and Laughlin (1980). Logue and Merville (1972) fail to recognize that the basic liquidity protection against unanticipated discrepancies in the amount and timing of operating cash in- and outflows is provided by a firm‟s cash reserve investments in combination with its unused borrowing capacity rather than by total current asset coverage of outstanding current liabilities. (Richards and Laughlin (1980)

Another ratio to measure liquidity is the cash conversion cycle. The CCC reflects the time span between disbursements and collections of cash. It is measured by estimating the

inventory conversion period and the receivable conversion period, less the payables conversion period. (Ashraf, 2012)

The cash conversion cycle, the measure of ongoing corporate liquidity, is calculated on the basis of the receivables conversion period plus inventory conversion period minus the

payables conversion period. The receivables conversion period is calculated by dividing accounts receivable by the average number of days of sales. (Accounts receivable/ (sales/365)) It expresses the average number of days of within receivables of sales are

(9)

9 days of cost of goods sold. (Accounts payable/ (cost of goods sold/365)) It expresses the average number of days between the purchase of goods and the payment of those goods and will be referred to as the days in accounts payables. (Jose, 2006, Lyroudi, 2000)

CCC= Receivables conversion period + inventory collection period + payable conversion period

= (AR/ (sales/365) + (Inv./(cost of goods sold/365)- (AP/(cost of goods sold/365)

Figure 2.1: The cash conversion cycle (Jose et al., 1996)

(10)

10

2.3 Relation between liquidity and performance

Deloof (2003) stated that it can be expected that the way working capital management is managed will have a significant impact on the profitability of a firm. For many firms working capital management is a very important component of their financial management. In his study of Belgian firms he finds that the way working capital is managed has a significant effect on the profitability of firms. Managers can create value for their shareholders by reducing the number of days in accounts receivable and the inventory conversion period to a reasonable minimum. The negative relation between accounts payable and profitability is consistent with the view that less profitable firms wait longer to pay their bills. (Deloof, 2003)

Shin and Shoenen (1998) examined the relation between working capital management and firm profitability using data of almost 60.000 listed American firms. They found a strong negative association between the firm‟s Net trade Cycle (cash conversion cycle expressed as a percentage of sales) and its profitability.

Raheman and Nasr (2007) studied the effects of working capital management on liquidity and as well on the profitability of a firm. They studied the effect of different variables of working capital management including the average collection period, inventory turnover days, average payment period, CCC and current ratio on the net operating profitability of Pakistani firms. They also found a strong negative relation between the variables of working capital management and the profitability of the firm.

Also the studies of Lazaridis (2006) and Eljelly (2004) indicated that there is a strong significant negative relation between liquidity and firm profitability.

(11)

11 The above presented analysis results in the following main hypothesis which consists of three sub-hypotheses.

Hypothesis 1: There is a negative relation between the CCC and firm performance.

Hypothesis 1A: There is a negative relation between the days in accounts receivable and firm performance.

Hypothesis 1B: There is a positive relation between the days in accounts payable and firm performance.

Hypothesis 1C: There is a negative relation between the days in inventory conversion period and firm performance.

In the above stated hypotheses firm performance is defined as return on equity. With a negative relation it is implied that the shorter the CCC, the higher the performance of the firm.

2.4 Culture

National culture is widely used in research in the fields of e.g. accounting, finance, and sociology to explain differences all over the world. In the course of time, there are several frameworks, measurements and definitions developed to define the concept of national culture. Hofstede (1993), Trompenaars and Hampden-Turner (1998), Hall (1976), Schwartz (1994) and the GLOBE (1991) study made a substantially effort to define national culture. Trompenaars and Hampden-Turner and Schwartz based their studies on the dimensions developed by Hofstede.

2.4.1 Hofstede’s cultural dimensions (Hofstede, 1993)

Hofstede developed four dimensions in the mid-1980s. The four dimensions described by Hofstede are Power Distance, Individualism, Masculinity and Uncertainty avoidance.

Power distance expresses the degree to which the less powerful members of a society accept and expect that power is distributed unequally. The way a society handles the

(12)

12 which needs no further justification. In societies with low power distance, people strive to equalize the distribution of power and demand justification for inequalities of power.

Individualism refers to the degree people in a society prefer to be focused on their self and have loose social connections. The high side of this dimension, referred to as individualism, can be defined as a preference for a social framework in which individuals are expected to take care of themselves and their immediate families only. Members of a collective society have a preference for a tight framework in society in which individuals can expect their relatives or members of a particular in-group to look after them in exchange for unquestioning loyalty. People‟s self-image in an individualistic society is reflected in terms of “I” whereas people in a collective society reflect to their self-image in terms of “we”.

Masculinity represents the preference in a society for achievement, heroism, assertiveness and material reward for success. Its opposite, femininity, stands for a preference for

cooperation, modesty, caring for the weak and quality of life.

Uncertainty avoidance expresses the degree to which the members of a society feel uncomfortable with uncertainty and ambiguity. Societies with a strong uncertainty avoidance life by rigid rules of belief and behaviours and avoid unorthodox behaviour and ideas. A weak uncertainty avoidant society is less concerned about the future, practice counts more than principles.

Later a fifth dimension was added, namely long-term versus short-term orientation. This dimension was added due to a research of M.H. Bond. On the long-term side one finds values oriented towards the future, like thrift (saving) and persistence. On the short-term side one finds values rather oriented towards the past and present, like respect for tradition and fulfilling social obligations (Hofstede, 1993).

(13)

13

2.4.2 Hall’s cultural dimensions

Hall introduced his distinction between high- and low- context cultures in 1976. By observing cross cultural differences in fields like social interactions, communication and problem solving he made this distinction. Hall (1976) takes several factors into account to determine his cultural dimensions. By observing cross cultural differences in fields like social interactions, communication and problem solving he developed the concepts of high and low context cultures. The concept of high and low context cultures can be seen as a continuum with on the high end of the continuum countries such as China, Korea and Japan and on the low end of the continuum Scandinavian countries like Norway and Sweden. (Onkvisit, 2009, p. 187)

Social orientation is one of the most differentiating aspects of high context cultures. People in such societies are deeply involved in each other and create strong bonds between family members, friends and colleagues. In a high-context culture a structure of social

hierarchy exists, individual inner feelings are kept under strong self-control and information is widely shared trough simple messages with deep meaning (Kim et al, 1998).

A low context culture is typically more individual based; their bonds are less likely to add up to a large and interconnected network which would influence their actions. As a result, people in low context societies tend to make their decisions based on personal judgment and according to their own interest. In a low-context culture social hierarchy and society in general imposes less on individuals‟ lives. Communication between people is more explicit and nonpersonal (Kim et al, 1998).

2.4.3 Trompenaars and Hampden-Turner cultural dimensions

Trompenaars and Hampden-Turner (1998) give an alternative operationalization of culture, although their basic assumptions are somewhat similar to the values in the model of Hofstede. They give seven dimensions for their model of culture. The seven dimensions are:

(14)

14

2.4.4 The Global Leadership and Organizational Behavior Effectiveness (GLOBE)

In 1991 Robert J. House developed the GLOBE research program. In 2004 he published its first volume of “Culture, Leadership and Organizations: The GLOBE Study of 62 Societies”. It was based on results from about 17300 middle managers from 951 organizations in the food processing, financial services and telecommunications services industries. In 2007, a second volume, “Culture and Leadership across the World: The GLOBE Book of In-Depth studies of 25 Societies” was published. In this second volume, the findings from the first volume with in-country leadership literature analyses, interview data, focus group discussions, and formal analyses of printed media provide in depth descriptions of leadership theory and leader behaviour in those 25 cultures. (Hoppe, 2007)

House established nine cultural dimensions that make it possible to capture leader effectiveness across cultures. These dimensions made it possible to capture the similarities and/or differences in norms, values, beliefs and practices among societies. The study is built on findings from other researchers who developed their own theories like Hofstede, Schwartz, Smith, and Inglehart. (Hoppe, 2007)

(15)

15 Figure 2.2: Cultural dimension comparison. (Scheffknecht, 2011)

2.5 Hofstede and the GLOBE study compared

Hofstede‟s study is criticized in the academic literature for several reasons. The criticism is often provided by proponents of other theories/studies. Chui et al. (2002) prefer Schwartz‟s theory instead of that of Hofstede because it contains comparable dimensions with those of Hofstede‟s study but is seen as more relevant for the world today. This is because the data used for the sample is collected more recent, and several socialist countries have also been included. Holden (2002) argues that the data Hofstede collected about 30 years ago is out-dated, using a theory developed with data that old will result in false conclusions he states.

(16)

16 In the existing literature the study of Hofstede and the GLOBE study have been compared several times. Shi et al. (2011) provide an overview (Figure 2.3) of differences that can be found easily after a quick comparison of the studies. It is clearly that although some of the dimensions are named the same the studies differ immensely.

Figure 2.3: Comparison between Hofstede and the GLOBE study (Shi et al., 2011)

Besides this superficial comparison there have been some more in-depth comparisons. Smith (2006) argues that the methodological problems Hofstede faced remain salient to all cross-cultural researches. Despite that the methods used by the GLOBE researchers address these problems in a somewhat different way and benefit of the greater power of recently developed procedures for statistical analysis, they also contain contingent risks and

ambiguities. The use of High minus Low (HML) analysis to split samples from within each nation to avoid common method bias and the theory driven design of measurement scales are mentioned as improvements compared to Hofstede. (Smith, 2006)

(17)

17 Hofstede study collected data from seven matched groups of which 2 were managerial and five were non-managerial. Hofstede chose this approach because he had found dramatic differences in statements about the former‟s leadership in earlier studies. Fourth, theory-driven versus action theory-driven. The development and analysis of the GLOBE questionnaire was theory-driven, based on existing literature, including a book of Hofstede. The questionnaire for the Hofstede study was action-driven to address issues that the IBM employees considered relevant in their working situation. Fifth, US inspired versus decentred. Hofstede states that although the GLOBE study tries to liberate organizational behaviour from the US hegemony, 25 editors and authors held degrees from US universities which points to the contrary.

Hofstede points on the fact that he recruited local researches with local degrees to contributed to the questionnaires and interpreting the results. Sixth, organizational culture as similar or different in nature to/from social culture. GLOBE asked its respondents questions in two formats: “in this society” and “in this organization”. The first one was labelled societal culture and the second organizational culture. In most cases they correlated closely and not treated separately in the GLOBE book according to Hofstede. He also studied organizational cultures and concludes that organizational and societal culture are phenomena of different order in such a way that calling them both “cultures‟ is somewhat misleading. The last difference Hofstede mentions is the influence of national wealth on culture. Wealth supports

individualism and is related to other dimensions. Although this relation is recognized by the authors of the GLOBE study it does not influence their interpretation of culture.

(18)

18

2.6 The influence of culture on the liquidity-performance relation

For this study three common dimensions of the Hofstede and GLOBE study will be used to define national culture. The hypothesis will be explained on the basis of the differences between Japan, the Netherlands and the United States but will be tested for a larger sample of countries.

Countries differ from each other with respect to the cultural dimensions of which I believe have an impact on the CCC and the performance of a firm. For instance, Hofstede et al. (2002) state that the US and the Netherlands scored high on the dimension individualism, the US had even the highest score. Japan scored about average on this dimension. An explanation given by Hofstede et al. (2002) for this difference can be the traditional family structure in the various countries. In the western countries the people traditional live in „core‟ families, that is: father, mother and children. In Japan the people live more in lineal families, that is

grandparents, oldest son, daughter-in-law, and children. Hechter et al. (1993) have examined the visibility of behaviour in Japanese schools, offices, factories, company houses and families. They found that the Japanese are used to the fact that they always have to take responsibility for their behaviour towards others. In the Netherlands and other Western societies this is not so strong. Managers in countries with high levels of individualism are more emphasized on self-sufficiency and self-interest (Ramirez and Kwok, 2009). Gleason et al. (2000) found that there is a strong relation between the cultural dimension individualism and the amount of debt in an organization.

Given this differences I will expect that individuality has a negative moderating effect on the relation between the CCC and ROE. In other words, the level of individuality will

influence the relation between the cash conversion cycle and firm performance, For example, an aggressive liquidity management can be driven by high level of individuality. It could be that firms in a high individualistic culture will bother less of what their reputation is. For testing this relation with the dimension from Hofstede the hypotheses will be:

Hypothesis 2.1: Individualism has a negative moderating effect on the relation between the cash conversion cycle and firm performance.

(19)

19

Hypothesis 2.1B: Individualism has a negative moderating effect on the relation between the days in account payable and firm performance.

Hypothesis 2.1C: Individualism has a negative moderating effect on the relation between the inventory conversion period and firm performance.

These relations above will also be tested for the cultural dimension institutional collectivism from the GLOBE study. These will be the hypotheses 2.2 to 2.2C.

The second dimension which I will use is uncertainty avoidance. Japan scores on this dimension fairly high whereas the USA and the Netherlands score weak on uncertainty avoidance. The Japanese society can be considered to be “rigid”. “Rigid” means that

behaviour in Japan is quite strictly prescribed, whereas at the same time people behave more nervously than in most other countries. Uncertainty avoiding societies are not kind to people who deviate from their rules. In weak uncertainty avoidance societies, on the contrary, the prevailing feeling would rather be that “what is different is curious”. US citizens attribute high importance to the reputations of the people they deal with, whereas Japanese people prefer to interact with “connections”, i.e. people they really know. This statement about the US is likely to apply equally well to the Dutch people (Hofstede et al., 2002). A strong relation between uncertainty avoidance and the amount of debt in an organization is found by Gleason et al. (2000). A similar relation was found by Chang et al. (2009). They found that countries with a higher degree of cultural risk aversion was directly associated with higher level of corporate cash holdings.

Given these results I expect that uncertainty avoidance will have a positive moderating effect on the relation between liquidity and firm performance. This can mean that firms from highly uncertainty avoidance cultures are less willing to take the risk of stock outs due to a short inventory conversion period and lost sales due to short credit period. For testing this relation with the dimension from Hofstede the hypotheses will be:

Hypothesis 3.1: Uncertainty avoidance has a positive moderating effect on the relation between the cash conversion cycle and firm performance.

(20)

20

Hypothesis 3.1B: Uncertainty avoidance has a positive moderating effect on the relation between the days in accounts payable and firm performance.

Hypothesis 3.1C: Uncertainty avoidance has a positive moderating effect on the relation between the inventory conversion period and firm performance

These relations above will also be tested for the cultural dimension uncertainty avoidance from the GLOBE study. These will be the hypotheses 3.2 to 3.2C.

Finally, long-term orientation / future orientation will be examined. According to research done by Hofstede et al. (2002), Japan scores high on long-term orientation. The US scores medium low on this dimension. A survey done by Hofstede and Bond (1988) pointed out that US and most other Western students, scored tradition and respecting social obligations more important than thrift and perseverance with regard to this dimension. This difference between Japan and the US can also be confirmed by the distinction between the US orientation towards profit maximization versus the Japanese preference towards “empire building” (Thurow, 1993). The Netherlands score average on this dimension. Hofstede refers to a study of Benedict from 1944 and a more recent study done by Visser and Hemelrijck in 1994 to illustrate that the Dutch tendency to save is not something from the last few years but is part of their culture. Chang et al. (2009) found also that countries that score high on long term orientation hold larger cash and liquid balances because they are not continuously pressured by shareholders to demonstrate short-term positive returns. Given the fact that the Cash Conversion Cycle is focused on the short-term I think that long-term orientation has a positive moderating effect on the relation between liquidity and firm performance. The hypotheses for testing the dimensions of Hofstede are:

Hypothesis 4.1: Long-term orientation has a positive moderating effect on the relation between the cash conversion cycle and firm performance.

Hypothesis 4.1A: Long-term orientation has a positive moderating effect on the relation between the days in account receivable and firm performance.

(21)

21

Hypothesis 4.1C: Long-term orientation has a positive moderating effect on the relation between the inventory conversion period and firm performance.

These relations above will also be tested for the cultural dimension future orientation from the GLOBE study. These will be the hypotheses 4.2 to 4.2C. 1

Power distance, masculinity and indulgence from Hofstede‟s dimensions are not explored in this study. The dimensions from GLOBE not explored in this study are power distance, in-group collectivism, gender egalitarianism, human orientation, assertiveness and performance orientation. These dimensions are not explored due to the fact that some of the dimensions from the GLOBE study do not have a matching dimension with the dimensions from Hofstede. Secondly, culture is shared but also situational; people do not use all five dimensions in every situation they face. Not all dimensions of culture need to affect debt ratios (Ramirez and Kwok, 2009). I think that this also applies for this research. Masculinity and power distance are helpful in explaining other business practices, such as differences in organizational structures in cross-country studies, but they do not have a clear theoretical implication for capital structure (Ramirez and Kwok, 2009).

Overview of hypotheses

In figure 2.4 an overview of the hypothesis presented in this chapter can be found.

Figure 2.4: Graphical overview of hypotheses.

1_{The hypotheses and supporting text are based on the paper “Cross culture research of the relationship}

between liquidity and performance”. The paper was written by J. van Baal, J. Golbach, L. ten Kate and L.

(22)

22

3. Method

In this study, an attempt is made to explore the differences in test results when using Hofstede‟s or the GLOBE study. To explore these differences, the influence of culture on the liquidity-performance relation will be examined by conducting an empirical examination of the formulated hypotheses. Before this empirical examination takes place the sample used will be discussed and the concepts will be operationalized.

3.1 Data sample

The financial data used is extracted from the database Orbis. The data will be collected for the period 2006– 2010 and only data from firms that are classified as a NACE Rev. 2. Main section C (manufacturing) firm are used to exclude these influences. This selection is made because capital intensity, product durability, production process and competitive forces influence the relationship between the CCC and profitability of a firm. (Jose et al., 1996) Making this distinction should reduce the influence of differences in characteristics of industries. Because the ROE is used as a measure of firm performance only listed firms were selected with a minimum of 1 million US$ in total assets. Based on these selection criteria a dataset consisting data from 25.826 firms from 21 countries was constructed.

3.2 Concepts

3.1.1 Performance

Firm performance will be evaluated by return on equity (ROE). This ratio is one of the best-known and most-widely used of all financial ratios. The ROE is a measure of how

shareholders fared during the year. The ROE is calculated by dividing the net income by the total equity of a firm. (Hillier et al., 2010) The data needed for the ROE is collected from the database Orbis and is expressed in percentages.

3.1.2 Liquidity

The cash conversion cycle is used to measure liquidity. The cash conversion cycle, the measure of ongoing corporate liquidity, is calculated on the basis of the receivables

(23)

23 calculated by dividing the inventory by the average days of the cost of goods sold.

(Inventory/(cost of goods sold/365)) It expresses the average number of days from the acquisition to the sale of the good. The payables conversion period is calculated by dividing accounts payable by the average days of cost of goods sold. (Accounts payable/(cost of goods sold/365)) It expresses the average number of days between the purchase of goods and the payment of those goods. (Jose, 2006, Lyroudi, 2000) The data needed to calculate the CCC can be collected from the database Orbis and is expressed in number of days.

3.1.3 Culture

Hofstede‟s study provides scores for 72 countries and the GLOBE study for 62 countries. Only countries for which both studies provide scores are selected. The scores of Hofstede represent each countries position from each other on scale of 0 to 118, which is the highest known value. The scores of the GLOBE study represent each country‟s relative position along a 0 to 7 point scale. The GLOBE scores produced two scores for each dimension. One score represents the judgment on items in the form of What should be. This emphasizes on values grown out of an anthropologic tradition of culture assessment. The other score is measuring indicators assessing What is, or What are, common behaviours, institutional practices, proscriptions en prescriptions. This approach to the assessment of culture grows out of a psychological/behavioural tradition, in which it is assumed that shared values are enacted in behaviours, policies and practices. (House et al. 2002) The scores representing the practices will be used because they display the common behaviours.

3.1.4 Control variables

(24)

24

Variable Measure

Dependent Performance Return on equity (ROE)

Independent Liquidity Cash Conversion Cycle (CCC)

Days is accounts receivable (COP) Days in accounts payable (PAY) Inventory conversion period (ICP) Moderaters Culture (Hofstede) Individualism (HIND)

Uncertainty avoidance (HUA) Long-term orientation (HLTO) Culture (GLOBE) Uncertainty avoidance (GUA)

Institutional Collectivism (GIC) Future Orientation (GFO)

Control Size Ln(sales) (LNS)

Sales growth LN (Sales growth) (LNSG) Financial debt Debt ratio (DEBR)

(25)

25

3.3 Statistical method

The hypothesis will be tested with a multiple regression analyses. In order to use this analyses there are several assumptions that have to be met. First of all there has to be a linear correlation. Secondly, the variables have to be normal distributed. Thirdly, there have to be homoscedasticity of the variance. Finally, there may be no multicollinearity of the variance.

These assumptions can be tested when the outliers are removed from the dataset. Outliers can be detected by using scatterplots or boxplots. Keller (2005) identifies outliers as points that lie outside the whiskers of a boxplot. These whiskers extend outward to the smaller of 1.5 times the interquartile range or to the extreme point that is not an outlier. Field (2005)

identifies outliers by using the standardized residuals, when the standardized residual is greater than 3.29 than it can be identified as an outlier. Keller (2005) uses a standardized residual of 2 to identify outliers. In this study, a standardized residual of 2.575829 will be used. This value marks the 1% outlying values which will be excluded. This method is also used in the study of Deloof (2003).

Normal distribution of the variables can be tested by examining the skewness and kurtosis of the variables. Al the variables have a skewness between -1 and 1 except sales, sales growth and days in inventory. Sales and sales growth are transformed with a natural log

transformation. Inventory conversion period transformed using a squared root transformation. ROE is also transformed using the square root transformation because it was positively skewed. By examining a histogram, scatterplot and P-P plot after the transformation the normal distribution of the return on equity is checked.

Homoscedasticity occurs when at each level of the predicting variables the variance of the residual terms is constant. (Field, 2005) Graphs and the Levene‟s test can be used to look at the homoscedasticity of the variances.

To execute a multiple regression analyses there should be no perfect mulitcollinearity, which means there should be no perfect linear relationship between two or more of the predictors. If there is perfect collinearity between predictors it becomes impossible to obtain unique estimates of the regression coefficients. (Field, 2005; Keller, 2005) The

(26)

26

3.3.1 Moderators

Because culture is seen as a moderator of the relation between the CCC and ROE, a moderator variable has to be computed. The moderating variables were obtained by

multiplying the independent variables with the cultural variables. Multiplying these variables result in a product term e.g. CCC*HIND which represents the moderating effect of

(27)

27

3.3.1 Regression models

In table 4.1 below the models used in the regression models can be found. Models 1 to 4 examine the relation between the cash conversion cycle and firm performance. Model 5 to 8 give an example of the models used to examine the same relation as in models 1 to 4 but moderated by a cultural variable. Model 5 to 8 are used to test the moderating effect of individualism on the relation between liquidity and firm performance. The hypothesis 2.2 to 4.2C will be tested by replacing the cultural and moderating variables in the models 5 to 8 by the ones tested in the hypotheses. (Hartmann et al. 1999)

Model Hypotheses Regression

1 H1 ROE=B0+B1CCC+B2LNS+B3LNSG+B4DEBR+e 2 H1A ROE=B0+B1COP+B2LNS+B3LNSG+B4DEBR+e 3 H1B ROE=B0+B1PAY+B2LNS+B3LNSG+B4DEBR+e 4 H1C ROE=B0+B1ICP+B2LNS+B3LNSG+B4DEBR+e 5 H2.1 ROE=B0+B1CCC+B2HIND+B3(CCC*HIND)+ B4LNS+B5LNSG+B6DEBR+e 6 H2.1A ROE=B0+B1COP+B2HIND+B3(COP*HIND)+ B4LNS+B5LNSG+B6DEBR+e 7 H2.1B ROE=B0+B1PAY+B2HIND+B3(PAY*HIND)+ B4LNS+B5LNSG+B6DEBR+e 8 H2.1C ROE=B0+B1ICP+B2HIND+B3(ICP*HIND)+ B4LNS+B5LNSG+B6DEBR+e

(28)

28 4.

Results

Linear correlation is shown with the correlation table. (Table 8.3) There is no perfect correlation (ρ=-1 or 1) between the variables so there is no perfect multicollinearity There is some evidence that there is no homoscedasticity of the variances. The scatterplot does not show an evenly spread of the residuals so the data shows signs of heteroscedasticity. In all the models the F-value was highly significant.

4.1 Descriptive statistics

Table 8.1 in the appendixes presents the financial descriptive statistics. The average ROE is 13.58%, the average cash conversion cycle is 82.76 days. Payments on sales are received after an average of 79.16 days. Firms pay for their purchases after an average of 43.76 days and it takes firms an average of 47.36 days to sell inventory. Mean sales are 81,685,000 USD and the firms have an average sales growth of 1.43%. The mean financial debt is 51.5% which means than on average the half of the liabilities of the firms exist out of equity.

In Table 8.2 in the appendixes the descriptive statistics for the cultural dimensions can be found. Hofstede‟s uncertainty avoidance scores an average of 57.71, Individualism scores 51,05 and long-term orientation scores 49.71 on average. The uncertainty avoidance dimensions that pertains to the GLOBE study scores 4.40, institutional collectivism has a mean of 4.42 and the mean of future orientation is 4.03.

4.2 Correlations

In Table 8.3 the correlation coefficients are presented for all the variables considered. There is a negative relation between the ROE and the cash conversion cycle and his separate parts (day‟s accounts payable, day‟s accounts receivable and days in inventory). This is consistent with the view that decreasing the time between the expenses for the acquisition of raw materials and receiving the payments for sold goods can increase profitability.

(29)

29 where Hofstede gives high scores the GLOBE study gives low scores for the comparable dimensions. Hofstede‟s uncertainty avoidance and individualism are negatively correlated with GLOBE‟s uncertainty avoidance and institutional collectivism. Hofstede‟s long-term orientation correlates positively with GLOBE‟s future orientation.

The correlations between the comparable cultural dimensions and the cash conversion cycle and his separate parts differ also. So is accounts payables similar correlated with Hofstede‟s and GLOBE‟s uncertainty avoidance (positive) but Hofstede‟s individualism and long-term orientation correlate positively where GLOBE‟s institutional collectivism and future orientation correlate negatively. Days in accounts receivable is positive correlated with Hofstede‟s uncertainty avoidance and individualism but negatively correlated with Hofstede‟s long-term orientation and all the three cultural dimensions of GLOBE‟s Study. Days in inventory also shows different correlations with comparable cultural dimensions. Where Hofstede‟s uncertainty avoidance is negatively correlated is that of GLOBE positively correlated with days in inventory. The opposite is the case for Hofstede‟s individualism (positive correlated with days in inventory) and GLOBE‟s institutional collectivism (negatively correlated with days in inventory). Hofstede‟s long-term orientation and GLOBE‟s future orientation are both negatively correlated.

(30)

30

4.3 Results of the hypotheses

The results of the determinants of return on equity without the moderating of cultural dimensions can be found in table 8.4. Regression model (1) examines the relation between the CCC and the firm performance. The coefficient of the CCC is negative and highly significant, this implies that shortening the CCC will result in a higher ROE. The coefficients of the other variables are also highly significant. Return on equity decreases with firm size and rises when sales growth increases and the financial debt ratio lowers. In model 2, the coefficient of days in accounts receivable is negative and significant which means that with a decrease in

numbers in accounts receivable will result in an increase in the ROE. Model 3 shows a

significant negative relation between days in accounts payables and ROE. This relation can be explained by the fact that less profitable firms probably wait longer to pay for their purchases. Model (4) shows a very significant negative relation between the inventory conversion period and the ROE. According to the results of models (1) to (4), managers can increase the ROE of their firm by decreasing the number of day in receivable and the inventory conversion period.

In table 8.5 the moderating variable individualism of Hofstede and in table 8.6 the moderating variable institutional collectivism of the GLOBE study are included in the regression models.

Although there were already significant coefficients in the models 1 to 4, including these moderating variables result in a higher R squared for all the models (models 5 to 12) tested with these variables. This increase due to the moderating variables is higher for GLOBE‟s institutional collectivism than for Hofstede‟s individualism. The models 5, 6 and 8 have problems with variables, CCC*HIND, financial debt and ICP*HIND, which are not significant so the reliability of the models can be disputed.

In table 8.7 and 8.8 the moderating variable uncertainty avoidance from Hofstede‟s and the GLOBE study are added to the regression. The result of including these variables is a higher R squared for the models 13 to 20 in comparison to the models without these moderating variables. In these models the moderating variables are highly significant except for CCC*HUA, therefor the reliability of model (5) is questionable.

(31)

31

Hypotheses

Hypothesis 1 states that there is a negative relation between the cash conversion cycle and firm performance. The results of model (1) in table 8.4 show that there is a highly significant negative relation between these two variables so hypothesis 1 can be confirmed.

Hypothesis 1A states that there is a negative relation between the days in accounts receivable and firm performance. Model 2 in table 8.4 shows that there is a significant negative relation between days in accounts receivable and firm performance. Given these results hypothesis 1A is accepted.

Hypothesis 1B states that there is a positive relation between days in accounts payables and firm performance. Model 3 in table 8.4 shows that there is a significant negative relation between the two variables so hypothesis 1B is rejected. An explanation for this outcome can be that firms that have a low profitability have the tendency to wait longer to pay their bills. (Deloof, 2003)

Hypothesis 1C states that there is a negative relation between the days in inventory conversion period and firm performance. Model 4 shows that there is a negative relation between inventory conversion period and firm performance so hypothesis 1C is accepted.

To get an insight in how culture moderates the relation between the cash conversion cycle (and its separate parts) and firm performance scatterplots can be computed. By using a

scatterplot and adding a fit line to this plot the degree of influence on the relation between the independent variables and ROE can be shown. Before this fit line can be added to the plot the cultural dimension have to be separated in high and low scores. This is done by taking the mean of the scores for the cultural dimensions. For the hypotheses of which the one with cultural dimension of Hofstede as well the comparable one of GLOBE‟s study was accepted, a scatterplot whit fit line is computed. The results can be found in appendix. (Figure 8.4 to 8.23).

Hypothesis 2.1 states that individualism has a negative moderating effect on the relation between the CCC and firm performance. Model 5 in table 8.5 shows that there is a significant negative relation between the CCC and firm performance. However the moderator variable is not significant so therefor hypothesis 2.1 is rejected.

(32)

32 performance however this relation is not stronger for high individualism so hypothesis 2.1A rejected.

Hypothesis 2.1B states that there is a relation between days in account payable and firm performance which is negative moderated by the cultural dimension individualism. Model 7 in table 8.5 shows that there is a significant relation between these two variables which is enhanced by the moderating variable. The coefficient is become smaller and the R squared value had increased in comparison with model 8 in table 8.5 without the moderating variable. These results confirm hypothesis 2.1B. The stronger negative relation for high individualism cultures can be found in figure 8.6.

Hypothesis 2.1C states that individualism has a negative moderating effect on the relation between the inventory conversion period and firm performance. Model (7) shows in table 4 that the moderating variable and the other variables are significant. Besides this, the R squared has risen. Figure 8.8 does show a positive negative effect so hypothesis 2.1C is accepted.

Hypothesis 2.2 states that institutional collectivism has a negative moderating effect on the relation between cash conversion cycle and firm performance. Model 9 in table 8.6 shows that the R squared increased from 0.119 to 0.136 which indicates that there is a moderating effect. Given these results hypothesis 2.2 is accepted.

Hypothesis 2.2A states that institutional collectivism has a negative moderating effect on the relation between days in accounts receivable and firm performance. The results of model (10) in table 8.6 are significant and show a higher R squared than model (2) which indicates a negative moderating effect by institutional collectivism. Based on the results shown in figure 8.5 hypothesis 2.2A is confirmed.

Hypothesis 2.2B states that institutional collectivism has a negative moderating effect on the relation between days in accounts payable and firm performance. Model (11) in table 8.6 shows a significant negative relation between days in accounts payables and firm performance which is enhanced by the moderating variable. The coefficient is -0.003 whereas in the

regression without the moderating variable the coefficient is -0.002 and the R squared has increased from 0.087 to 0.96. Based on these results and the plot in figure 8.7, which shows a stronger negative relation for cultures with high institutional collectivism hypothesis 2.2B can be confirmed.

(33)

33 8.6) show a significant negative relation between the inventory conversion period and firm performance which is positively moderated (figure 8.9) by high collectivism variable. The R squared has increased from 0.118 to 0.132. Based on these results hypothesis 2.2C is

accepted.

Hypothesis 3.1 states that uncertainty avoidance has a moderating effect on the relation between the cash conversion cycle and firm performance. The results of model (13) can be found in table 8.7. As these results show the moderating variable is not significant and therefor hypothesis 3.1 is rejected.

Hypothesis 3.1A states that uncertainty avoidance has a positive moderating effect on the relation between accounts receivable and firm performance. The R squared has increased from 0.097(model 4) to 0.100 as a result of the moderating variable. The result of model (14) can be found in table 8.7. Figure 8.10 does not show a positive moderation effect (less

negative slope) for high uncertainty avoidance so hypothesis 3.1A is rejected.

Hypothesis3.1B states that uncertainty avoidance has a positive moderating effect on the relation between the days in accounts receivable and firm performance. Model (15) in table 8.7shows that there is a significant negative relation between the days in accounts payable and firm performance. This relation is changed due to the introduction of the moderating variable in the regression model. Also the R squared has been increased. Figure 8.12 shows a less negative slope. Given these results the hypothesis 3.1B is accepted.

Hypothesis3.1C states that uncertainty avoidance has a positive moderating effect on the relation between the inventory conversion period and firm performance. Model 16 in table 8.7 shows a significant negative relation between the inventory conversion period which is

stronger than the results of model (4) without the moderating variable. However figure 8.14 dos not show a less negative slope so hypothesis 3.1C is rejected

Hypothesis 3.2 states that uncertainty avoidance has a positive moderating effect on the relation between the cash conversion cycle and firm performance. Model 17 in table 8.8 shows a highly significant relation between the CCC and firm performance. Therefore hypothesis 3.2 will be accepted

(34)

34 performance which is moderated by the cultural dimension uncertainty avoidance. Besides this a less negative slope is shown in figure 8.11 so hypothesis 3.2A is accepted

Hypothesis 3.2B states that uncertainty avoidance positively moderates the relation between the days in accounts payables and firm performance. Model (19) in table 8.8 shows that there is a significant negative relation between accounts payables and firm performance. Due to the moderating variable the R squared is increased. Also figure 8.13 shows a positive moderating effect for hign uncertainty avoidance so hypothesis 3.2B will be accepted

Hypothesis 3.2C states that uncertainty avoidance positively moderates the relation between inventory conversion period and firm performance. In table 8.8 the results of

regression model 20 can be found. The results show that there is a significant positive relation between the inventory conversion period and firm performance. The R squared has increased from 0.118 to 0.132. However, figure 8.15 does not show a less negative slope for high uncertainty avoidance so this hypothesis will be rejected.

Hypothesis 4.1 states that Long-term orientation has a positive moderating effect on the relation between the CCC and firm performance. Model (21) in table 8.9 shows that there is a significant negative relation between the cash conversion cycle and firm performance. The R squared has risen from 0.119 to 0.124 and the coefficients for the variables are significant. Therefore, hypothesis 4.1 will be accepted. Figure 8.16 shows a less negative slope for high long-term orientation.

Hypothesis 4.1A states that long-term orientation has a positive moderating effect on the relation between the days in accounts receivable and firm performance. Model (22) in table 8.9 shows that there is a significant negative relation between days in accounts receivable and firm performance but it is not moderated by the cultural variable long-term orientation. The R squared has risen and the results are significant however figure 8.18 does not show a less negative slope for high long-term orientation so hypothesis 4.1A is rejected.

Hypothesis 4.1B states that long-term orientation has a positive moderating effect on the relation between days in accounts payable and firm performance. Model (23) in table 8.9 shows that there is negative relation between days in accounts payable and firm performance. The coefficient of the moderating variable is significant and the R squared has increased however figure 8.20 does not show a less negative slope for high long-term orientation so hypothesis 4.1B is rejected

(35)

35 model 24 in table 8 shows that this relation is positively moderated by the cultural variable long-term orientation, the coefficient has fallen from -0.088 to 0.089 so the influence is small but exists. Besides this the R squared has risen from 0.118 to 0.119. Given these results and the results in figure 8.20 this hypothesis 4.1C is accepted.

Hypothesis 4.2 states that future orientation has a positive moderating effect on the relation between the cash conversion cycle and firm performance. Model 25 in table 8.10 shows the results for this hypothesis. The results show that Future orientation has a moderating effect on the relation between the CCC and firm performance The R squared has risen. Given these results and figure 8.17 shows a less negative slope so hypothesis 4.2 is accepted.

Hypothesis 4.2A states that future orientation has a positive moderating effect on the relation between the days in accounts receivable and firm performance. Model (26) in table 8.10 shows a significant moderator variable and also the other variables are significant however there is no less negative slope for high future orientation cultures in figure 8.19 so the hypothesis will be rejected.

Hypothesis 4.2B states that future orientation has a positive moderating effect on the relation between the days in accounts payables and firm performance. model (27) in table 8.10 shows that the moderating variable made the coefficient of days in accounts payable drop from -0.002 (model 3) to The R squared has increased from 0.087 to 0.100. however there is no less negative slope for high future orientation cultures in figure 8.21 so hypothesis 4.2B will be rejected.

The last hypothesis states that future orientation has a positive moderating effect on the relation between the inventory conversion period and firm performance. Adding the moderating variable into the model results in a decrease of the coefficient of inventory conversion period from -0.088 to -0.100. The moderating variable let the R squared increase from 0.118 to 0.146. (table 8.10) Figure 8.23 shows a positive slope for high future

(36)

36

4.4 Results of Hofstede and GLOBE compared

To make a comparison between the results of the two theories the scatterplots will be compared. The comparison is done by comparing the direction and slopes of the fit lines in the scatterplots. Comparing the scatterplots should give in insight in the similarities and differences in the results

In figure 8.4 and 8.5 the scatterplots with fit lines can be found. These lines have the same direction for low and high values on the dimensions individualism and institutional

collectivism. This indicates that the variables moderate the relation between days in accounts receivable and firm performance in the same direction however with different strength. The moderation however differs between Hofstede and GLOBE, figure 8.4 shows a less negative slope for the high score than for the low score which means a positive moderation. For GLOBE the opposite accounts which means a negative moderation

In figures 8.6 and 8.7 there can be seen that the regression lines have different directions. In figure 8.6 the line for low and high individualism both have a negative slope whereas in figure 8.7 the slope for low institutional collectivism has a positive slope and for High Institutional Collectivism has a negative slope. The high level of individualism has a positive slope whereas the high level of Institutional collectivism has a negative slope. A negative moderation effect occurs for Hofstede as well GLOBE.

Figures 8.8 and 8.9 show the fit lines for the moderation individualism/ institutional collectivism on the relation between the inventory conversion period and the ROE. Both the low and high values of the cultural dimension have a negative slope. The moderation for Hofstede is negative (slope of high score is steeper than the one for the low score) and for GLOBE positive moderation occurs (slope of low score is steeper than the one for the high score).

In figures 8.10 and 8.11 the fit lines representing the moderating effect of uncertainty avoidance on the relation between the days in account receivable and ROE can be found. Figures 8.10 and 8.11 show that the moderating effect for high uncertainty avoidance for Hofstede is negative whereas the for GOBE the moderating effect is positive

In figure 8.12 and 8.13 show the slopes for the moderation of uncertainty avoidance on the relation between the days in account payables and ROE. The lines for both Hofstede as

(37)

37 uncertainty avoidance are almost the same. Because the slopes for the high scores on both theories are more positive than on the low scores they both moderated positively

Figure 8.14 and 8.15 show the moderation of uncertainty avoidance on the relation between the inventory conversion period and ROE. Both figures show negative slopes for both low and high uncertainty avoidance. Because the slopes for the high scores on both theories are more negatively than on the low scores they both moderated negatively.

In figure 8.16 and figure 8.17 the fit lines can be found for the moderators long-term orientation and future orientation. In both the scatterplots the fit lines have a negative direction and the slope of the low score on the cultural dimension has the steepest slope. Because the slopes for the high scores for both the theories in these figures are less negative than those of the low scores the moderated both positively.

This is also the case for figures 8.18 and 8.19 which show that both long-term orientation and future orientation have a negative coefficient i.e. they have same moderating effect on the relation between accounts receivable and ROE.

In figures 8.20 and 8.21 the fit lines can be found for the moderating effect of Long-term orientation and future orientation on the relation between accounts payables and the ROE.) The slopes for high long-term- and future orientation are both negative and have a more negative slope than the low score, which have a positive slope. So for this pair of hypotheses the moderation is positive.

Figures 8.22 and 8.23 show that the correlations for high long-term orientation and future orientation are both less negative, in the case op future orientation positive, than the low scores on these dimensions. This implies that the moderating for Hofstede‟s dimension and the one of GLOBE is positively.

(38)

38

5. Discussion and conclusion

In this last section the outcomes of the analyses will be discussed thereafter a conclusion, limitations and ideas for further research will be provided.

5.1 Discussion

A total of twenty-eight hypotheses are tested in order to answer the main research question:

Does the chosen theory for operationalizing culture affects the influence of national culture on the relationship between liquidity and performance?

The results of the hypotheses will be discussed to answer the main research question. There is a significant negative relation found between the cash conversion cycle and the firm performance. Introducing the moderating variables resulted in insignificant coefficients in the models. Therefore, only two of the six hypotheses that examined the moderating effect on this relation could be accepted. The hypotheses that tested the moderation of institutional collectivism and long-term orientation were accepted. Given this result we can say that institutional collectivism and long-term orientation moderate the relation between the cash conversion cycle and the ROE.

The hypothesis examining the relation between days in accounts receivable and the ROE could not be confirmed because the model suffered with insignificant coefficients which make the results questionable.

Introducing the cultural dimensions as moderators in the relation between days in accounts receivable and ROE result in several significant results. These results indicate that

institutional collectivism, uncertainty avoidance (for both Hofstede and GLOBE), and future orientation moderate the relation between the days in accounts receivable and ROE. Based on the scatterplots with the fit lines this hypothesis can be confirmed

A significant negative relation between the days in accounts payables and the ROE is found. This indicates, despite of the small coefficient (ρ=-0.002) that the shorter this period is the higher the firm performance will be. For individualism, institutional collectivism,

(39)

39 These results suggest that cultural dimensions have influence on firm performance by moderating the relation between the cash conversion cycle and ROE. The relation between cash conversion cycle and ROE is moderated by long-term orientation and institutional collectivism. The relation between the separate parts of the cash conversion cycle and firm performance is moderated by all the cultural dimensions used in this study.

(40)

40

5.2 Conclusion

The results discussed in the previous section show results that prove that the choice of operationalizing culture with Hofstede or GLOBE will affect the way culture influences the relation between liquidity and firm performance. The strength of the relations differed when testing equal dimensions and in some cases the positive relation turned into a negative relation. Knowing these differences occur implies outcomes of research can be manipulated by choosing one of the two ways in operationalizing culture. Because culture is hard to grasp there will always be disagreements in which the best theory for culture is. Therefor it is important to underpin why a certain theory is chosen in a research. The fact that a theory is easy to use and widely spread should not be the only argument.

Although these strong results, this research has its limitations and therefor the results should be interpreted with care. The limitations can be found in the used data and results

In this study the matching dimensions from Hofstede and GLOBE are treated equally. This is done to explore the differences that will occur when the same relation is tested. Although the dimension is considered to represent the same values/behaviours of a culture they differ slightly in their origins. Hofstede‟s cultural dimensions are based on a research conducted under IBM employees in different countries. These dimensions should represent the national culture whereas the dimensions from the GLOBE study are based on a research on the values/behaviours in the workplace. The sample was larger than the one of Hofstede and consisted out of employees in lower and higher management. Roughly speaking, these dimensions could be considered as the dimensions of the organizational culture. This can be seen as a limitation of the research but it also could be seen as strength. Although, the results didn‟t show large differences between the results of Hofstede and GLOBE, the opposite results in some of the scatterplots may indicate that results in studies can be influenced by the choice of operationalizing culture.

(41)

41 research the R squared can be increased by adding more financial variables in the models but also non-financial measures like market share and customer satisfaction could be added.

Using the ROE, or a similar accounting based measure of firm performance has its limitations because it bears the risk of investment or operating myopia. When managers are hold accountable for short-term profits or returns they will be induced to reduce or postpone investments that promise pay-offs in future measurement periods. (Merchant, 2007) An alternative could be a performance measure based on operating income. The influence of the debt and assets structure will be eliminated and the distinction between listed and not listed firms will be irrelevant

Secondly, the low R squared linear values that were observed when examining the

scatterplots (figure 4 to figure 23) are a limitation. These values represent the slope of the fit line. The steepest slope that was observed can be found in figure 8.17 and has a value R squared = 0.087. In this scatterplot, the relation between the cash conversion cycle and ROE in countries with low future orientation was measured. The correlation between the two variables tested in that scatterplot is the squared root of 0.087 which is -0.294. (Minus sign due to the downwards slope) Most of the values were close to zero which means that the correlation will become close to zero. The explanation for these small values may lie in the fact that the cultural dimensions were only grouped in a low and high group. Dividing the cultural dimension in 3 or 4 groups will probably amplify the contrast between the low and high scores.

Future research could focus the differences of the origins of Hofstede‟s theory and the GLOBE study. Also the importance of corporate and national culture could be explored. Research could focus on the influence of organizational culture on firm performance or how the national culture influences the corporate culture. Also the influence of corporate culture on national culture can be examined. The fact that we live in a world were multinationals have a huge impact on our lives maybe results in changing our national cultures.

Third, in future research the operationalization of firm performance can be discussed, a closer look can be taken on the financial and non-financial performance measures.

Although this research has its limitations, an insight is given in how the choice of

(42)

42

6. References

Ashraf, C.K., 2012. The Relationship between Working Capital Efficiency and Profitability,

Advances in Management, Vol. 5, pp. 60-74

Chui, A., Lloyd, A. and Kwok, C., 2002.The Determination of Capital Structure: Is national Culture a Missing Piece to the Puzzle? Journal of International Business

Studies, 33, No. 1, pp. 99-127

Deloof, M., 2003. Does Working Capital Management Affect Profitability of Belgian Firms. Journal of Business Finance & Accounting, Vol. 30 No. 3/4, pp. 573-587

Edmonds, T.P., Tsay, B., Olds, P.R., 2009, Fundamental Managerial Accounting (5th edition), McGraw Hill Irwin, New York

Eljelly, A.M.A., 2004, Liquidity-profitability trade-off: an empirical investigation in an emerging market, International Journal of Commerce & Management, Vol. 14 No. 2, pp. 48-61

Field, A, 2005, Discovering Statistics Using SPSS (2nd edition), SAGE Publications, London

Gleason, K.C., Mathur, L.K, Mathur, I. 2000. The interrelationship between culture, capital structure and performance: evidence from European retailers. Journal of Business

Research, Vol. 50 No. 2, pp. 185-191

Gouveia, V.V., Ros, M., 2000. Hofstede‟s and Schwartz‟s models for classifying

individualism at the cultural level: their relation to macro-social and macro-economic variables. Psicothema, Vol. 12, pp. 25-33

Hartmann, F.G.H., Moers, F., 1999. Testing contingency hypotheses in budgetary research: an evaluation of the use of moderated regression analyses, Accounting, Organizations

and Society, Vol. 24, pp 291-315

Hechter, M., Kanazawa S., 1993, Group Solidarity and Social Order in Japan,. Journal of