The effect of strategic positioning within
strategic groups. An extension of the
strategic balance theory.
Master Thesis, Strategic Innovation Management
L.R. van der Meer, S2908573 Rijksuniversiteit Groningen
Faculty of Economics and Business, Duissenberg building Groningen, Groningen, The Netherlands
e-mail: l.r.van.der.meer@student.rug.nl Supervisor: dr. C. Carroll Co-assessor: dr. J.D. van der Bij
June 24th, 2019
*Acknowledgement: I would like toexpress my gratitude towards my supervisor, dr. Charlie Carroll, for the effortful
3
ABSTRACT
5
T
ABLE OF CONTENTS
ABSTRACT 3 1. INTRODUCTION 6 2. LITERATURE REVIEW 10 2.1 Strategic groups 102.2 Differences between groups 12 2.3 Core and secondary firms 13 2.4 Strategic positioning within groups 15
2.5 Hypotheses 19
3. METHODOLOGY 20
3.1 Banking industry 20
3.2 Variables 20
3.3 Identification of strategic groups 23
3.4 Analyses 23
4. RESULTS 25
4.1 Results between-group analysis 33 4.2 Results within-group analysis 36
4.3 Limitations 41
5. DISCUSSION 42
6. FUTURE RESEARCH 46
6
1. INTRODUCTION
Although there is no universally accepted definition of strategic groups, it continues to be commonly defined as subsets of firms within the same industry that pursue relatively similar strategies (McGee and Thomas, 1986). Strategic groups can differ significantly in conduct and group dynamics (e.g. perfect competition versus oligopoly), which in turn generate different strategic actions and performance effects (Carroll, 2019). The research in determinants of superior group performance has interested scholars since the early
developments of strategic management theory (Mehra, 1996). Literature has extensively theorized on the roots for these intra-industry differences, yet most studies have produced inconsistent empirical results. However, two concepts in the literature best explain the cause of these intra-industry performance differences. Firstly, the possession of strategically
relevant resources naturally plays an important role in the profitability of competing firms and groups in an industry (Barney, 1991). These resource differences are reflected in the groups’ strategies and protected by mobility barriers amongst groups (Mascarenhas and Aaker, 1989). Since strategic groups differ in their resource endowments, some groups possess more
negotiation power towards customers and suppliers which give them a competitive advantage over the other groups (Mascarenhas and Aaker,1989).
Secondly, initial research from the industrial organization economics (IO) proposed that industry structure determines the collective performance of firms but firms possess the ability to collude with each other through strategic groups to face different conditions from other groups (Reger and Huff, 1993). These interactions effects relate to the
7
While initial strategic groups literature mostly focused of the effects group
membership and the central effect of mobility barriers on firm performance, some researchers argued that these concepts inadequately explained performance differences among firms in the industry (Cool and Schendel, 1988). Porter (1979) was one of the first to state that the height of the ‘entry barriers’ contained insufficient explanatory power for performance differences among industry participants. Porter suggested that an intermediate analysis based on firm-specific characteristics could significantly enhance the explanatory power of the Industrial Organization model.
In general, research in strategic groups and its performance implications has produced indecisive results. The reason for these equivocal findings might relate to Porter’s (1979) arguments for the need of a more inclusive IO model. The omission of firm-specific factors namely excludes the effect of within-group differences, while in some cases this effect could be of significant influence. Several studies namely found that within-group performance was greater than between-group performance studies due to the relative differences with member firms (e.g. Lewis and Thomas, 1990; Cool and Schendel, 1988). If this phenomenon is truly at work, this might explain why many prior studies failed to find consistent evidence for
performance differences among strategic groups (Cool and Schendel, 1988).
8
be classified to the degree to which they are similar with their strategic group, such that some firms follow the group strategy closely (i.e. core firms) and others follow it less closely (i.e secondary firms) (Reger and Huff, 1993). With core firms being characterized as firms that share and define the competitive position of the group, while secondary firms are still linked to the key dimensions but differentiated from the group’s mean, which is reflected in its strategic variables (Ketchen et al., 1993). Research on the strategic positioning within a strategic group and its performance has advocated either strategic position, but lack consistent findings for the literature to accept the arguments advocated for a certain position. Therefore, research in a firm’s strategic position (core vs secondary), and its rivalrous interactions, in a strategic group can offer great insights for strategic groups research and managerial
implications since the degree of differentiation between members firms might significantly influence firm performance (McNamara et al., 2003). In sum, strategic groups are clusters of firms together based on their strategic similarity, thereby classifying relatively heterogenous firms into relatively homogenous groups. This research will analyze the effect of relative dissimilarity between member firms on performance, which means that I analyze the effect of the relative distance within a homogenous group.
A study by Deephouse (1999) extensively researched this strategic trade-off between conformity and differentiation in the banking industry, which complements the debate of relative competitive positioning within strategic groups. This strategic trade-off was captured in his strategic balance theory which synthesized the differentiation and conformity
9
This research aims to provide an extension of strategic groups literature by filling in the literature gap proposed by Deephouse (1999). This will be done by researching the strategic relative position of a firm and its performance implications within a strategic group setting by examining firms in the Dutch banking industry. Thus, this study will first try to explain and discover significant differences between groups. Then, this study sets out to examine the relationship between the strategic position of a firm within a strategic group and the associated performance implications for that firm. By doing so, I will ultimately attempt to answer the following research questions:
RQ1: “Do strategic groups significantly differ in financial performance from each other?”
RQ2: “Is the strategic distance of a firm from the center of a strategic group related to the relative performance of that firm?”
Most strategic groups studies are snapshots in time that do not serve to forecast future strategic directions beyond the research setting itself. However, the merit of these studies and the contribution to the literature is the recognition of the existence of differences within strategic groups, and that they are in part the result of the deliberate outcome of decisions made by firms in an industry (McGee and Thomas, 1986). In this case, decisions regarding the strategic position within the group itself.
10
2. LITERATURE REVIEW 2.1 Strategic groups
The term ‘strategic group’ has first been coined by Hunt (1972) in his identification of groups in the U.S home appliance industry. This term was quickly picked up by the academic world and further developed upon and has since become a pervasive branch of strategic management literature (Ketchen et al., 1997). Strategic groups are subsets of firms within the same industry that pursue strategies that are similar to each other (McGee and Thomas, 1986). These strategies consists of several strategic dimensions of the firm, including: scale of the firm, market segments, firm’s assets, price and quality (Gulati et al., 2000). Strategic groups are an extension of the Industrial Organization paradigm (IO), which states that industry structure is the main determinant of conduct and behavior of firms within a given industry, which in turn impacts the collective performance of those firms in the industry (Bain, 1968). IO economics assumes homogeneity within an industry and is required in order for the industry structure to determine the conduct and behavior and subsequently the performance for the entire set of homogenous firms (Porter, 1980). However, Hunt (1972) observed that within this homogenous set of firms, groups started to form groups which had an influence on performance in the industry. In this way, strategic groups can be seen as an extension of the IO paradigm, in which strategic groups reflect heterogeneity within an industry, while homogeneity still exist within each group (Ketchen et al., 1993).
The formation of strategic groups helps member firms to adopt similar strategic postures by appearing to be in possession of similar strategic identities to the external environment (Deephouse and Ferguson, 2000). Strategic group membership resembles this strategic identity and therefore it has been a focal theme in strategic group literature to demonstrate the existence of a relationship between strategic group membership and
11
has been extensively documented, strategic group membership performed poorly as an indicator of firm performance (Cool and Dierickx, 1993). A study by Nair and Kotha (2001) perfectly captured this shortcoming of strategic group membership as their results showed a positive direct relationship between strategic group membership and organizational
performance in the first seven years of their research, ceteris paribus, but a negative
association in the last five years. This study pointed out the context- and time-specificity of strategic groups research, which has been a limitation for the generalization of produced results in the literature. However, strategic groups research has consistently pointed out the competitive advantages that strategic groups can offer through the collective tacit collusion between member firms (e.g. Newman, 1978).
One of these collusion advantages is the creation of barriers to imitation, also known as ”mobility barriers”. Mobility barriers have been a prominent subject in strategic groups research and, in many cases, form the base of competitive advantages within industries (Mascarenhas and Aaker,1989). Mobility barriers exist between strategic groups to inhibit entrants into groups, which in turn protect the strategic group members against profit losses (Porter, 1979). Firms are hindered in their ability to switch between strategic groups due to the presence of these mobility barriers (Leask, 2003). A high degree of mobility barriers in an industry is associated with higher expected costs of attempts to switch strategic group
12
2.2 Differences between groups
With regard to the between group differences, there are two concepts that together offer the best theory for variations in intra-industry profitability. Firstly, the resource-based view places emphasis on the possession of strategically relevant resources to achieve a
competitive advantage in an industry (Barney, 1991). The Resource-based view is popularized by an article by Barney (1991) which argued that sustained competitive advantage is derived from a firm’s controllable resources and capabilities that satisfy the VRIN-characteristics (Valuable, Rare, Inimitable, Non-substitutable). This new perspective explained firm
performance through resource heterogeneity of firms, and the groups they classify in, and has evolved itself into one of the most influential theories for strategic management literature (Barney et al., 2001). Within an industry, this perspective helps explain the performance difference between strategic groups as the use of resource endowments enables the firm to acquire a unique resource position and realize a competitive advantage through increased negotiation power (Hatten and Hatten, 1987). An industry consists of different groups that each possess and offer unique resources that define their strategic fit with the industry. Since some groups have a better strategic fit with the industry and therefore achieve greater returns, (asymmetrical) mobility barriers are necessitated for financial gains to be consistently realized (Hatten and Hatten, 1987). The combination of superior resources and barriers to imitation give strategic groups more bargaining power towards customers and suppliers which in turn create performance differences between groups (Mascarenhas and Aaker,1989).
13
competition within an industry. (McGee and Thomas, 1986). The
structure-conduct-performance (S-C-P) paradigm stems from IO theory as it dictates that structure determines the conduct of the firms, which in turn determines the performance (Cool and Schendel, 1987). The S-C-P model has been commonly referred to in strategic groups literature as many studies have relied on its existence for legitimacy purposes (McGee and Thomas, 1986). The S-C-P model has predictive power for the interaction effects between firms scattered across strategic positions in an industry. Attractive strategic positioning would form densely populated positions in an industry, which creates interdependence between these proximate strategic groups (Carroll, 2019). Though, sufficient isolation from other groups could contain competition within the group, it is argued that ‘interdependent’ strategic groups in general interact with other strategic groups, either cooperatively or competitively. This pattern of rivalry is determined by the different conducts of strategic groups, which influences
performance (Mas-Ruiz and Ruiz-Moreno, 2011). These S-C-P effects in combination with the competitive market forces through possession of strategically relevant resources form the basis of performance differences between groups.
2.3 Core and secondary firms
14
Moreover, strategic groups research has proven the existence of rivalrous behavior among member firms which contribute to an increase in within-firm performance differences (e.g. Smith et al., 1997; Mas-Ruiz and Ruiz-Moreno, 2011). However, an extra dimension of complexity is introduced, since it has been empirically proven that core firms compete more intensely with one another but less intensely with periphery firms, and vice versa (DeSarbo et al., 2008). The notion of intra-firm rivalry emphasizes on the importance of the strategic position of a firm within groups as colluding too closely with firms exposes a firm to risks of opportunistic behavior of its peers (Cool and Dierickx, 1993).
15
2.4 Strategic positioning within strategic groups
While Deephouse (1999) found significant results for his study, he highlighted the opportunity to test his strategic balance theory in a strategic group setting. These stated opportunities revolve around the interaction effects a firm can have with the core recipe of a given strategic group. It is namely stated that group membership fosters legitimacy of individual member firms (Peteraf and Shanley, 1997). Therefore, it can be argued that group members that follow the core recipe of the group tightly have greater legitimacy than those operating at the fringes of the group, the periphery firms (McNamara, 2003). Strategic balance theory addresses relations among strategic similarity, legitimacy, competition and associated performance. Several theories explain how a firm’s strategic position, relative to that of competing firms, is able to affect performance, though terms do differ. This reasoning is the backbone of the strategic balance theory. Moreover, every individual firm's strategic position is supported by both the resources and capabilities that firm possess, reflecting the idea that these are the two sides of the same coin (Wernerfelt, 1984). Thus, the strategic balance theory addresses how the strategic position is challenged by the conflicting pressures of conformity and differentiation and how research can provide managerial implications for the firm by finding the optimal strategic balance point (Deephouse, 1999). This optimal point is also known as the ‘competitive cusp’, which has been a prominent subject in the work of Porac et al. (1989).
16
the literature on, it is required to tap into adjacent niches in order to dissect this broad concept. Therefore, I will now follow up by elaborating the two main concepts of the strategic balance theory, namely the concept of differentiation and conformity. Though these concepts reflect opposing sides of the spectrum, relative strategic positioning possesses a fluid form which can take any intermediate form across the spectrum.
Firstly, strategic conformity to the core strategy of a group is mainly driven by different facets of isomorphism (Barreto and Baden‐Fuller, 2006). Isomorphism refers to the process in which firms that operate in the same environment become constrained and inclined to strategically resemble each other (Hawley, 1968). This can further be divided into
institutional isomorphism, which relates to the derived homogeneity between peers through legitimation mechanisms, and competitive isomorphisms, which refers to the economically derived homogeneity of firms based on performance evaluations (Fennell, 1980). The concept of isomorphism can be used to explain the social and economic fitness of member firms (DiMaggio and Powell, 1983). The subsequent increased fitness between a member firm and its strategic group in turn increases the legitimacy of that firm. The increased legitimacy from intentional strategic decisions of the firm to identify itself closely to the strategic group have significant benefits for the firm (McNamara et al., 2003). Firms that conform to the
prototypical ingroup norm, which therefore identify as a core firm, are able to acquire
17
within-group distance from the from the focal group’s core would be negative for the member firm and yield lower profits.
18
This reasoning explains how secondary firms in a group try to differentiate themselves from the core firms in an attempt to achieve better performance. This concept has been
19
2.5 Hypotheses
The literature on strategic positioning within a strategic group provides interesting and legitimate reasoning for both sides of the strategic trade-off. In order to take a more
explorative approach to the theory in matter, competing hypotheses are developed to retain the explorative nature of this research, while maintaining the ability to perform hypothesis testing by satisfying the quantitative research requirements.
Drawing upon Deephouse’s strategic balance theory, the following competing hypotheses are formulated:
H1: Within a strategic group, the relationship between a member firm’s strategic distance to the group centroid is negatively related to its financial performance, meaning that a lower
strategic distance from the group’s centroid results in higher relative financial performance of the firm.
H2: Within a strategic group, the relationship between a member firm’s strategic distance to the group centroid is positively related to financial performance, meaning that a higher
20
3. METHODOLOGY 3.1 Banking industry
To maintain consistency with several of the reviewed literature, in specific Deephouse (1999) his research, the banking industry will be the most suitable industry to test the
assertions of this research. The banking industry has rich availability of archival data, complete set of competitors and acknowledged competitiveness (Reger and Huff, 1993). Moreover, the banking industry serves as a feasible industry for core-periphery research because identifying strategic groups in the banking industry is significant because the industry represents a turbulent environment and fierce competition (Fiegenbaum and Thomas, 1993). Furthermore, the financial crisis highlighted the need to better understand financial market structure and interconnectedness within the industry and groups (Craig and von Peter, 2014). Although, it is evident that the banking industry faces strict regulation procedures, individual firms retain the ability to determine strategic direction by choosing several competitive positions e.g.; strategic pricing/margins, financial leverage and service offerings (McNamara et al., 2003).
3.2 Variables
21
The strategic variables required for the cluster analysis are of importance because it defines the strategic groups and the associated outcomes. In total there are 11 variables included in the dataset. All value of all variables have been taken from the year 2015.
The following variables have been selected and will be used as input for the cluster analysis. These strategic variables either relate to the transaction patterns with suppliers and customers or the asset quality of the banks. These were also both important factors for the defining variables in the strategic groups study in the banking industry of Mehra (1996).
Table 1 – Strategic variables for the cluster analysis
Variable Absolute/ Percentage Transformed into Log. variable? Short description
Total assets Absolute Yes Sum of assets
Equity to assets Percentage Yes Total equity divided by total assets Loan loss reserves to
risk weighted assets
Absolute Yes Measures how well bank can cover losses on loans
Interest margin Percentage Yes Measures how successful bank is at investing its funds
Efficiency Percentage Yes Measures how well bank uses its assets
Interbank assets to liabilities
Absolute Yes Measures how much the bank borrows and lends with other banks
Liquid assets to total assets
Percentage Yes Total liquid assets divided by total assets
Net loans to total assets
Percentage Yes Net loans divided by total assets
Capital adequacy ratio Percentage No Measures financial strength and stability
Interest income to risk
weighted assets Absolute No Measures how much a bank makes
relative to its risk weighted assets Loan loss reserves to
impaired loans
22
Important to note is that all of the strategic variables have been transformed into Z-scores to standardize the values in the data to reflect the relativity to each other. The Z-score transformation and subsequent use of transformed data to classify subsets of firms is a familiar statistical method in strategic groups research. The use of standardization of numerical variables in a cluster analysis is advocated by Milligan and Cooper (1988). 3.2.1 Independent variables: Strategic distances
In order to analyze the strategic position of firms in strategic groups and its performance effects, it is required to convert this concept into a measurable variable. Therefore, this research employs strategic distance to evaluate the strategic position of the firms relative to the group’s average. Because this research uses multiple strategic variables for the cluster analysis, the strategic distance has to be calculated for each strategic variable to find the strategic positioning of each firm relative to its peers (Cool and Dierickx, 1993). This is done by calculating the absolute distance from the centroid of each strategic variable through subtracting the value of the original Z-scores in the sample with the group’s average. This method has been used before in strategic groups by Schimmer and Brauer (2012).
3.2.2 Dependent variables: Performance measures
The measure of performance consists of a combination of three financial ratios, namely: Return on average assets (ROAA)%, Return on average equity (ROAE)% and Growth in Assets (GIA)%. ROAA and ROAE are common measures for bank performance and have been used in the measurement of financial banking performance in most studies (e.g. Lawless et al., 1989; Mehra, 1996). Moreover, ‘Growth in Assets’ has been used as a
performance measure by Gordon and DiTomaso (1992). These dependent variables are not converted in Z-scores and also did not need rescaling through log transformation. The
23
3.3 Identification of strategic groups
A cluster analysis is performed with significance testing based on a permutation test and a Monte Carlo test. The role of cluster analysis in this research is to discover whether the collected data contains sub-groups among the set of firms. Ward’s minimum variance method is the most used hierarchical algorithm in strategic research (Ketchen et al., 1996) and also used in this research. Ward’s method identifies different clusters based on the strategic variables selected by the researcher and then merges the identified clusters together to look at the aggregate deviation of the formed clusters. Groups are then formed based on the minimum sum of squared deviations of all samples from the created centroid (Reger and Huff, 1993). Due to the recent advancements in strategic groups research, the significance of these identified clusters will simultaneously be tested through an intensive computing process and produce a permutation and Monte Carlo statistics as output. These significance tests display the probabilities for significance for the entire range of possible clusters. The final amount of clusters, and thus strategic groups, is selected by the researcher through various supporting tools.
3.4 Analyses
24
3.4.1 Performance differences across groups
A MANOVA will be performed to determine if the strategic groups are significantly different from each other in terms of performance. If the four cluster solution is significant on the MANOVA tests then the null hypothesis is rejected and we conclude that the clusters are significantly different from each other on the selected dependent variables. Subsequently, individual ANOVA analyses and independent t-tests will be conducted as follow-up tests.
3.4.2 Performance differences within groups
25
4. RESULTS
For the determination of the most appropriate amount of clusters, a combination of three main instruments were employed to determine the optimal amount of groups, namely: significance testing statistics, kink heuristics (values and kinks) and logical reasoning. The four cluster model ended up providing the best fit for this research. Firstly, the significance testing statistics from the cluster analysis are given below.
Permutation significance test
Groups Criteria Prob Wilks Prob Hotell Prob 1 297.000 1.000 . . . . 2 247.840 .030 .114 .332 7.743 .332 3 205.051 .003 .014 .374 15.602 .458 4 165.387 .001 .000 .002 78.747 .001 5 140.267 .001 .000 .003 83.874 .002 6 116.745 .001 .000 .011 92.167 .006 7 97.250 .001 .000 .002 120.892 .005 8 79.651 .001 .000 .001 147.941 .009 9 63.027 .001 .000 .001 207.096 .008 10 55.422 .001 .000 .001 236.588 .011 11 47.879 .001 .000 .001 276.477 .019 12 41.299 .001 .000 .003 320.771 .044
Permutation significance test for kinks
27
Simulation significance test
Groups Criteria Prob Wilks Prob Hotell Prob 1 297.000 .948 . . . . 2 247.840 .984 .114 .158 7.743 .158 3 205.051 .975 .014 .030 15.602 .072 4 165.387 .919 .000 .001 78.747 .001 5 140.267 .873 .000 .001 83.874 .001 6 116.745 .712 .000 .001 92.167 .001 7 97.250 .487 .000 .001 120.892 .002 8 79.651 .212 .000 .001 147.941 .001 9 63.027 .028 .000 .001 207.096 .001 10 55.422 .023 .000 .001 236.588 .004 11 47.879 .015 .000 .001 276.477 .009 12 41.299 .007 .000 .001 320.771 .020
Simulation significance test for kinks
29
The four cluster solution is significant on criteria for the permutation test, but not for the Monte Carlo simulation test. However, the four cluster solution indicated significant differences for the Wilk’s Lambda and Hotelling’s Trace. Furthermore, based on the
associated coefficients, kink heuristics across the range of possible clusters clearly depict the strength of a certain amount of clusters. There are six pairs of charts that depict the graph of values and kinks for the criteria values under each other. The charts are organized in such a way that you can easily connect the values of the significance test with the lines and peaks in the kink heuristics. A sharp turn (elbow) in the line of values and a peak (or inverted peak) in the graph of kinks is what you should be looking for to find the optimal amount of clusters. You can clearly see that the graphs react heavily on the 4 clusters solution, although, the nine cluster solution also shows strong results, especially in the kinks graphs. However, due to the goal of this research, namely finding a significant relationship between strategic positioning within a group and performance effects, it is preferable to test this assertion on a lower
amount of clusters with more firms in each cluster. Therefore, after a descriptive classification of the different clusters and its idiosyncrasies, the four cluster model will be employed in the remainder of this research.
30
Table 2 – The research sample classified in the four cluster solution Cluster number: Specialisation Bank:
Cluster 1: Real Estate & Mortgage bank Rabo Vastgoedgroep Holding N.V.-Rabo Real Estate Group
Cluster 2: Commercial Bank Natwest Markets N.V.-RBS NV
Commercial Bank BinckBank NV
Commercial Bank Deutsche bank
Cooperative bank Oikocredit, Ecumenical Development Co-Operative Society U.A.
Commercial Bank LeasePlan Corporation NV
Cluster 3: Commercial Bank Delta Lloyd Bank NV Real Estate & Mortgage bank Achmea Bank NV
Commercial Bank AEGON Bank NV
Commercial Bank NIBC Bank NV
Commercial Bank ING Bank NV
Commercial Bank GarantiBank International NV
Commercial Bank Nationale-Nederlanden Bank NV
Cooperative bank Cooperatieve Rabobank U.A.
Commercial Bank Yapi Kredi Bank Nederland N.V
Commercial Bank ABN AMRO Bank NV
Commercial Bank De Volksbank N.V.
Commercial Bank Triodos Bank NV
Commercial Bank Anadolubank Nederland NV
Commercial Bank Demir-Halk Bank (Nederland) N.V-DHB Bank
Commercial Bank Credit Europe Bank N.V.
Investment bank Promsvyaz Capital
Cluster 4: Commercial Bank MUFG Bank (Europe) NV
Commercial Bank Mizuho Bank Europe NV
Commercial Bank Bank Mendes Gans NV
Savings Bank Algemene Spaarbank voor Nederland - ASN Bank NV
Commercial Bank Van Lanschot N.V.
32
The individual interpretations of the four strategic banking groups will now be given.
Cluster 1: Non-commercial orientation group. This group only consists of one firm. It is clearly reflected in the strategic values that this bank is not a typical commercial bank as it is the only bank that offers a negative interest margin, and in general it scores very low on capital variables, such as Capital Adequacy, Net loans ratio and Liquid assets ratio, in comparison with the rest of the banks. This indicates that this group (firm) is very
differentiated from the other Dutch banks in the sample. Cluster 2: Low-risk, solvent group. This group consist of five firms, which score on average
high on the strategic variables “Equity to assets”, “Efficiency”, “Capital Adequacy” and “Loan loss reserves to impaired loans”. The relatively high scores relative on these distinct variables, relative to the dataset, indicates that this group of firms is more likely to be able to
withstand financial setbacks due to their financial configurations. Cluster 3: Mainstream, commercially-oriented banks. This is the biggest cluster of firms,
containing 16 firms. This cluster has moderate scores for all variables (-0.5 < score < 0.5). Since this is the largest group in the sample, it is assumed that there will be higher levels of
competition in this group. Cluster 4: Less secure banks. This group distinguishes itself from the other groups on several
financial configurations. They have relatively low scores on most financial variables that involve risk taking. Mainly on the variables “Loan loss to risk weighted Assets”, “Capital adequacy” and “Loan loss to impaired loans” these firms score relatively low compared to the other banks. This indicates that these banks are less secure and are more vulnerable to failure and cannot easily cover poor assets such as impaired loans or financial obstacles in general.
33
4.1 Results between-group analysis
Table 4 – Results of the MANOVA test
A MANOVA test was used to answer the first research question and find whether differences in performance across groups exists. The results indicated that there were significance
34
Subsequently, independent one-way ANOVA analyses on the performance variables “ROAA, ROAE and Growth in assets” were conducted as follow-up tests.
Table 5 – Results of the ANOVA test ANOVA N=28 Sum of Squares df Mean Square F Sig. Return On Avg Assets (ROAA) Between Groups 2,192 3 0,731 1,916 0,154 Within Groups 9,150 24 0,381 Total 11,342 27 Return On Avg Equity (ROAE) Between Groups 700,844 3 233,615 4,341 0,014 Within Groups 1291,684 24 53,820 Total 1992,528 27 Growth in total assets Between Groups 1046,948 3 348,983 2,219 0,112 Within Groups 3774,349 24 157,265 Total 4821,297 27
There was a significant effect of “Return of Average Equity” on cluster differences for the three conditions [F(3, 24) = 4.341, p = 0.014]. However, “Return on Average Assets” and “Growth in Assets” were not significant. Taken together, the results suggest there was a significant multivariate interaction and main effect for ROAE, but not ROAA and GIA.
Then, an independent-samples T-test was conducted to compare ROAE (because it was the only significant performance measure on the ANOVA test) in two different clusters at a time. Cluster 1 is excluded in this independent T-test because as it only consists of one firm and is therefore less representative for an independent T-test.
Table 6 – Group statistics for Return on Average Equity Group Statistics for Return On Avg Equity
(ROAE)
Ward Method N Mean Std. Deviation
Cluster 1 1 -20,07 0
Cluster 2 5 2,57 9,39057
Cluster 3 16 6,49 3,87926
Cluster 4 6 6,21 11,94339
35
Table 7 – Independent sample t-test between Cluster 2 and 3 Independent Samples Test
Levene's Test for Equality of
Variances t-test for Equality of Means
F Sig. t df Sig. (2-tailed) Mean Difference Std. Error Difference 95% Confidence Interval of the Difference Lower Upper Return On Avg Equity (ROAE) Equal variances assumed 4,827 0,041 -1,386 19 0,182 -3,91875 2,82699 -9,83572 1,99822 Equal variances not
assumed
-0,909 4,435 0,410 -3,91875 4,31012 -15,43688
7,59938
Table 8 – Independent sample t-test between Cluster 2 and 4 Independent Samples Test
Levene's Test for Equality of
Variances t-test for Equality of Means
F Sig. t df Sig. (2-tailed) Mean Difference Std. Error Difference 95% Confidence Interval of the Difference Lower Upper Return On Avg Equity (ROAE) Equal variances assumed 0,198 0,667 -0,552 9 0,595 -3,63500 6,58997 -18,54255 11,27255 Equal variances not
assumed
-0,565 8,987 0,586 -3,63500 6,43511 -18,19532
10,92532
Table 9 – Independent sample t-test between Cluster 3 and 4 Independent Samples Test
Levene's Test for Equality of
Variances t-test for Equality of Means
F Sig. t df Sig. (2-tailed) Mean Difference Std. Error Difference 95% Confidence Interval of the Difference Lower Upper Return On Avg Equity (ROAE) Equal variances assumed 7,434 0,013 0,087 20 0,932 0,28375 3,28007 -6,55835 7,12585 Equal variances not assumed 0,057 5,401 0,957 0,28375 4,97138 -12,21565 12,78315
36
4.2 Results within-group analysis
Table 10 - Pearson correlations between the aggregate strategic distance and performance Pearson Correlations Matrix N=28
Return On Avg Assets (ROAA) Return On Avg Equity (ROAE) Growth in tot assets (GIA) Distance_Own_Centroid Pearson Correlation 0,310 0,201 0,024 Sig. (2-tailed) 0,108 0,305 0,904
Moving on to the within-group analysis which includes the hypothesis testing part. A bivariate correlation analysis was conducted as a supporting analysis to check for any
significant correlations between the calculated distances and the three dependent performance variables. The ‘Distance_own_centroid’ measures the difference between the aggregate of all strategic values and the centroid of the cluster for the same values. There were no significant Pearson correlations found between the variables. “Distance from own centroid” and
38
Table 11 – Test for dimensionality Canonical Correlations Canonical functions Correlation Eigenvalue
Wilks Statistic F Num D.F Denom D.F. Sig. 1 0,823 2,104 0,102 1,488 33,000 41,951 0,111 2 0,776 1,512 0,316 1,167 20,000 30,000 0,343 3 0,453 0,259 0,794 0,460 9,000 16,000 0,881
Table 12 – The first canonical function Canonical correlation analysis, function 1, p=0,111
Strategy variable: Coefficients Loadings
Cross-Loadings Absolute_Distance_assets 0,103 0,133 0,109 Absolute_Distance_equity_to_assets 0,609 0,300 0,247 Absolute_Distance_loan_loss_to_RWA -0,081 -0,106 -0,087 Absolute_Distance_interest_margin 0,263 0,019 0,016 Absolute_Distance_efficiency -0,306 -0,335 -0,276 Absolute_Distance_interbankratio 0,288 -0,116 -0,095 Absolute_Distance_liquid_assetratio 0,383 0,022 0,018 Absolute_Distance_netloansratio 0,235 0,179 0,147 Absolute_Distance_capital_adequacy -0,741 -0,127 -0,105 Absolute_Distance_interest_income_to 0,190 0,102 0,084 Absolute_Distance_loanloss_to_impaired -0,931 -0,598 -0,493
Performance variable: Coefficients Loadings
Cross-Loadings
Relative_Distance_ROAA -0,391 0,788 0,416
Relative_Distance_DistanceROAE 0,720 0,240 0,571
39
Table 13 – The second canonical function Canonical correlation analysis, function 2, p=0,343
Strategy variable: Coefficients Loadings
Cross-Loadings Absolute_Distance_assets -0,485 -0,263 -0,204 Absolute_Distance_equity_to_assets 0,923 0,640 0,497 Absolute_Distance_loan_loss_to_RWA -0,283 0,414 0,321 Absolute_Distance_interest_margin -0,445 -0,069 -0,054 Absolute_Distance_efficiency 0,004 0,322 0,249 Absolute_Distance_interbankratio 0,334 0,510 0,396 Absolute_Distance_liquid_assetratio -0,289 0,174 0,135 Absolute_Distance_netloansratio 0,383 0,421 0,327 Absolute_Distance_capital_adequacy -0,33 0,190 0,148 Absolute_Distance_interest_income_to -0,302 0,297 0,230 Absolute_Distance_loanloss_to_impaired 0,383 0,473 0,367
Performance variable: Coefficients Loadings
Cross-Loadings
Relative_Distance_ROAA 0,164 -0,353 -0,274
Relative_Distance_DistanceROAE -1,071 -0,679 -0,527
Relative_Distance_Asset_growth 0,749 0,442 0,343
Table 12 – The third canonical function Canonical correlation analysis, function 3, p=0,881
Strategy variable: Coefficients Loadings
Cross-Loadings Absolute_Distance_assets 0,227 0,387 0,176 Absolute_Distance_equity_to_assets 0,704 0,518 0,235 Absolute_Distance_loan_loss_to_RWA -0,636 0,176 0,080 Absolute_Distance_interest_margin 0,234 0,433 0,196 Absolute_Distance_efficiency 0,450 0,390 0,177 Absolute_Distance_interbankratio -0,703 -0,088 -0,040 Absolute_Distance_liquid_assetratio -0,261 0,300 0,136 Absolute_Distance_netloansratio 0,175 0,132 0,060 Absolute_Distance_capital_adequacy 0,828 0,467 0,212 Absolute_Distance_interest_income_to -0,563 0,061 0,028 Absolute_Distance_loanloss_to_impaired 0,177 0,130 0,059
Performance variable: Coefficients Loadings
Cross-Loadings
Relative_Distance_ROAA 1,594 0,788 0,357
Relative_Distance_DistanceROAE -0,941 0,240 0,109
40
Table 11 contains the results of the dimensionality tests for the canonical functions. The number of possible canonical variates, or also known as canonical dimensions, is equal to the amount of dependent variables in the smaller set of variables. In this case, the smaller set of variables consist of the three performance variables; ROAA, ROAE and Growth in Assets. No significant canonical dimensions resulted from the analysis This can be interpreted in the following way. The composite between the absolute distances from the centroid of strategic variables and the relative distances of the performance variables is not significantly correlated with each other. The default null hypothesis of the canonical correlation analysis, which tested whether the correlations between the two sets of variables were zero, is supported. Therefore, we can conclude that strategic distance from the centroid, either far or close, in a strategic group does not influence performance. Since none of the canonical dimensions are significant, the underlying coefficients and loadings of the individual variables cannot be interpreted.
In sum, the set of within-group analyses looked for positional effects of strategic positioning in a variety of different ways. The Pearson correlations and scatterplots looked for the possibility of linear and curvilinear effects, respectively, between distance and
41
4.3 Limitations
There are several shortcomings to the empirical analyses of this study. Firstly, the research sample is the biggest limitation of this study. The statistical analyses showed no patterns of association in the sample, which made it unable to test the hypotheses further on. The proposed arguments should be tested in a new research setting where performance differences between member firms are bound to exist. Moreover, a combination of quantitative and qualitative data is often the standard and preferred for research in the
strategic positioning within groups. The measurement of strategic similarity should therefore be improved as it now insufficiently reflects the competitive and institutional pressures of the strategic balance theory. Archival data often has insufficient explanatory power to precisely classify a firm’s relative position within a strategic group. Further research should therefore refine the operationalization and include more accurate measurements of strategic similarity to adequately match the proposed theory with the study’s methods. Lastly, this study
examined the effect of a group structure at a single point in time with a constrained sample. Naturally, this serves as a limitation as the results are highly subjective to the research setting and, as cautioned before, findings cannot be interpreted beyond this research setting.
42
5. DISCUSSION
This study provides new insights to the debate over the relationship between strategic positioning and performance within strategic groups. This study looked at two prominent issues from the literature which functioned as the red line throughout this paper. Firstly, this study sets out to find the existence of performance differences across group, which are based on the principles of the resource-based view and the structure-conduct-performance (S-C-P) link. This preliminary between-group analysis served as validation for the created clusters because the existence of significant performance differences were required to proceed with hypothesis testing. Subsequently, the within-group analysis was performed, in which the developed hypotheses are tested and insights are derived from. This study sets out to examine the performance effects due to the strategic positioning in a strategic group, measured by the strategic distance of firms from the centroid of their strategic group. Two competing
hypotheses were derived from the strategic balance theory as Deephouse (1999) emphasized on the opportunities to test his theory in a strategic groups setting in combination with reviewed literature that supported the same propositions. The findings of this study have limited implications for the competitive analysis of the focal firm but yield insights for the strategic groups literature.
43
the arguments found in the literature and there was a lack of evidence to support either of the developed hypotheses.
The findings of this study strongly relate to a hypothesis formulated in the work of Cool and Schendel (1988). The authors hypothesized that firms that belong to the same strategic group realize similar performance levels. While this hypothesis neglects the concept of strategic positioning, it argues that two factors may be responsible for the (lack of)
existence of within-group performance differences. Firstly, the input market and associated procurement of required assets could play an important role in the variation of ingroup performance as a multitude of costs and returns are tied to this process. Secondly, the
competition in output markets and their varying degrees of market power can either constrain or amplify performance differences.
44
The findings still reflect inconsistencies with most of reviewed literature regarding the existence of significant within-group performance differences. In specific, the results are inconsistent with McNamara et al. (2003), which conducted a similar study on the strategic positioning within groups. McNamara et al. namely found that the greater part of variation in firm performance was due to the firm differences within strategic groups as opposed to the systemic differences between different groups. This is interesting as the opposite holds true for this study. Furthermore, this study found results that are at first sight unrelated to the article by Deephouse (1999) and the propositions he made due to the lack of supporting evidence for either hypothesis. However, thus far there is no reason to invalidate Deephouse’s arguments in a strategic group setting and his propositions should therefore be tested in a sample that reflect high levels of within-group competition. Hence, the results are largely dependent on the competitive setting in question and more research is needed to capture the full potential of the strategic trade off between differentiation and legitimacy.
While strategic management research devotes much attention to the development of taxonomies and typologies, little attention is given to the question whether and why
performance differs among firms pursuing similar strategies (Cool and Schendel, 1988). The equivocal findings thus far indicate that more research attention should be devoted to establish more validity for the performance implications of within group research.
As aforementioned, the results remain highly subjective to the context of the research. The negative results for the canonical correlation analysis is the product of a snapshot in time for a certain industry in a certain year. Hence, the magnitude of the findings should be
45
In sum, strategic group members are commonly viewed as similar in both conduct and performance, which has led to an overemphasis on between-group performance differences and a lack of focus on the performance effects of within-group performance differences. Therefore, this study out to examine the performance effects due to the strategic positioning in a strategic group. This study found no significant relationship between the strategic distance from the group’s centroid and financial performance of the firm. Literature argues that the input and output markets of an industry are of significant influence on the rivalrous interactions between strategic group member firms, which in turn affect the within-group performance differences. Although, this partly justifies the lack of support for the hypotheses, the findings remain contradictory with most of the reviewed literature on this topic. This highlights the importance of the research setting in question and the need for further research in this field of strategic research management. Altogether, this study advocates for a collective shift in focus to these suppressed or even unexplored corners of this field as it might uncover fruitful new insights for theorists and strategists alike.
46
6. FUTURE RESEARCH
This study highlighted several opportunities for future research(es) to improve upon. These opportunities mostly run parallel with the limitations of this study, which are stated in the results section. As mentioned before, the biggest leap in improvement can be achieved by the inclusion of variables that clearly reflect high levels of within-group competition. Within-group performance variation seems to be dependent on the level of rivalrous interactions with each other. Therefore it is recommended to test to the strategic balance theory in a more competitive research setting to adequately test Deephouse’s (1999) propositions. Furthermore, a multi-layered research across different industries over a longer period of time should be conducted to increase the generalizability and validity for a theoretical contribution to this field. Moreover, it is likely that certain contingencies were of effect on the relationship between strategic positioning within a group and firm performance, as the banking industry often characterized by its strong institutional environment (Scott, 1995). So, while this study theorized upon the benefits for finding an optimal point between differentiation and similarity, future research could devote its efforts to exploring the degree to which this relationship is contingent on firm capabilities and industry characteristics.
In sum, future research should test the proposed arguments in a research setting that adequately reflect high levels of competitive and institutional pressures to ensure the
existence of within-group performance differences. Moreover, future research could attempt to control contingencies that may be off influence on the relationship between strategic positioning within groups and firm performance. Due to the limited timeframe for this research I was unable to adequately address these issues, neither could I extend this research in terms of sample size and complexity. Hopefully, this study encourages other researchers to build further on my findings and capitalize upon the proposed opportunities for future
47
7. REFERENCES
Bain, J. S. (1956). Barriers to new competition: their character and consequences in manufacturing industries (Vol. 329). Cambridge, MA: Harvard University Press.
Bain, J.S. Industrial organization (2nd ed.) New York: Wiley, 1968
Barney, J. B., & Hoskisson, R. E. (1990). Strategic groups: Untested assertions and research proposals. Managerial and decision Economics, 11(3), 187-198.
Barney, J. (1991). Firm resources and sustained competitive advantage. Journal of
management, 17(1), 99-120.
Barney, J., Wright, M., & Ketchen Jr, D. J. (2001). The resource-based view of the firm: Ten years after 1991. Journal of management, 27(6), 625-641.
Barreto, I., & Baden‐Fuller, C. (2006). To conform or to perform? Mimetic behaviour, legitimacy‐ based groups and performance consequences. Journal of management studies, 43(7), 1559-1581. Carroll, C., Pandian, J. R. M., & Thomas, H. (1994). Assessing the Height of Mobility Barriers: A Methodology and an Empirical Test in the UK Retail Grocery Industry 1. British Journal of
Management, 5(1), 1-18.
Carroll, C. (2006). Canonical correlation analysis: Assessing links between multiplex networks. Social Networks, 28(4), 310-330.
Carroll, C (2019). Strategic groups and significant clustering. (Work-in-progress)
Cool, K., & Dierickx, I. (1993). Rivalry, strategic groups and firm profitability. Strategic Management Journal, 14(1), 47-59.
Cool, K. O., & Schendel, D. (1987). Strategic group formation and performance: The case of the US pharmaceutical industry, 1963–1982. Management science, 33(9), 1102-1124.
Cool, K., & Schendel, D. (1988). Performance differences among strategic group members. Strategic
Management Journal, 9(3), 207-223.
Craig, B., & Von Peter, G. (2014). Interbank tiering and money center banks. Journal of Financial
Intermediation, 23(3), 322-347.
Deephouse, D. L. (1999). To be different, or to be the same? It’s a question (and theory) of strategic balance. Strategic management journal, 20(2), 147-166.
DeSarbo, W. S., & Grewal, R. (2008). Hybrid strategic groups. Strategic Management Journal, 29(3), 293-317.
DiMaggio, P. J., & Powell, W. W. (1983). The iron cage revisited: Institutional isomorphism and collective rationality in organizational fields. American sociological review, 147-160.
Dooley, R. S., Fowler, D. M., & Miller, A. (1996). The benefits of strategic homogeneity and strategic heterogeneity: Theoretical and empirical evidence resolving past differences. Strategic Management Journal, 17(4), 293-305.
Fennell, M.L., 1980. The effects of environmental characteristics on the structure of hospital clusters. Administrative Science Quarterly 25 (484–510)
48
Fiegenbaum, A., & Thomas, H. (1993). Industry and strategic group dynamics: competitive strategy in the insurance industry, 1970–84. Journal of Management Studies, 30(1), 69-105.
Galaskiewicz, J. (1985). Interorganizational relations. Annual review of sociology, 11(1), 281-304. Gimeno, J., & Woo, C. Y. (1996). Hypercompetition in a multimarket environment: The role of strategic similarity and multimarket contact in competitive de-escalation. Organization science, 7(3), 322-341.
Gordon, G. G., & DiTomaso, N. (1992). Predicting corporate performance from organizational culture. Journal of management studies, 29(6), 783-798.
Gulati, R., Nohria, N., & Zaheer, A. (2000). Strategic networks. Strategic management journal, 21(3), 203-215.
Hatten, K. and M. Hatten (1987). 'Strategic groups, asymmetrical mobility barriers and contestability', Strategic Management Journal, 8(4), pp. 329-34
Hawley, A., 1968. Human ecology. In: Sills, D.L. (Ed.), International Encyclopedia of the Social Sciences. Macmillan, New York.
Hunt, M. S. (1972). Competition in the major home appliance industry, 1960-1970. Harvard University.
Ketchen Jr, D. J., Thomas, J. B., & Snow, C. C. (1993). Organizational configurations and
performance: A comparison of theoretical approaches. Academy of management journal, 36(6), 1278-1313.
Lawless, M. W., Bergh, D. D., & Wilsted, W. D. (1989). Performance variations among strategic group members: An examination of individual firm capability. Journal of Management, 15(4), 649-661.
Leask, G. (2004). Is there still value in strategic group research?. Aston Business School Research Institute.
Lewis, P., & Thomas, H. (1990). The linkage between strategy, strategic groups, and performance in the UK retail grocery industry. Strategic Management Journal, 11(5), 385-397.
Mascarenhas, B., & Aaker, D. A. (1989). Mobility barriers and strategic groups. Strategic
Management Journal, 10(5), 475-485.
Mas‐Ruiz, F., & Ruiz‐Moreno, F. (2011). Rivalry within strategic groups and consequences for performance: the firm‐size effects. Strategic Management Journal, 32(12), 1286-1308.
McGee, J., & Thomas, H. (1986). Strategic groups: theory, research and taxonomy. Strategic
management journal, 7(2), 141-160.
McNamara, G., Deephouse, D. L., & Luce, R. A. (2003). Competitive positioning within and across a strategic group structure: the performance of core, secondary, and solitary firms. Strategic
Management Journal, 24(2), 161-181.
Mehra, A. (1996). Resource and market based determinants of performance in the US banking industry. Strategic Management Journal, 17(4), 307-322.
Miller, J. G., & Roth, A. V. (1994). A taxonomy of manufacturing strategies. Management Science, 40(3), 285-304.
49
Nair, A., & Kotha, S. (2001). Does group membership matter? Evidence from the Japanese steel industry. Strategic Management Journal, 22(3), 221-235.
Newman, H. H. (1978). Strategic groups and the structure-performance relationship. The Review of Economics and Statistics, 417-427.
Peteraf, M., & Shanley, M. (1997). Getting to know you: A theory of strategic group identity. Strategic Management Journal, 18(S1), 165-186.
Porac, J. F., Thomas, H., & Baden‐Fuller, C. (1989). Competitive groups as cognitive communities: The case of Scottish knitwear manufacturers. Journal of Management studies, 26(4), 397-416. Porter, M. E. (1979). The structure within industries and companies’ performance. Review of
economics and statistics, 61(2), 214-227.
Porter, M. E. (1981). The contributions of industrial organization to strategic management. Academy
of management review, 6(4), 609-620.
Reger, R. K., & Huff, A. S. (1993). Strategic groups: A cognitive perspective. Strategic management
journal, 14(2), 103-123.
Schimmer, M., & Brauer, M. (2012). Firm performance and aspiration levels as determinants of a firm’s strategic repositioning within strategic group structures. Strategic organization, 10(4), 406-435. Scott, W. R. (1995). Institutions and organizations. Foundations for organizational science. London: A Sage Publication Series.
Singh, J. V., Tucker, D. J., & House, R. J. (1986). Organizational legitimacy and the liability of newness. Administrative science quarterly, 171-193.
Smith, K. G., Grimm, C. M., Young, G., & Wally, S. (1997). Strategic groups and rivalrous firm behavior: Towards a reconciliation. Strategic Management Journal, 18(2), 149-157.
Snow, C. C., & Miles, R. E. (1983). The role of strategy in the development of a general theory of organizations. Advances in strategic management, 2(1), 213-259.
Suchman, M. C. (1995). Managing legitimacy: Strategic and institutional approaches. Academy of
management review, 20(3), 571-610.
