Competing or coopeting? Italian banking rivalry in Egypt, 1924 – 1940

Purpose – The purpose of this paper is to investigate the key drivers of coopetition by adopting a managerial and economic framework. A case of coopetition failure is investigated by means of a historical example focused ontheEgyptian adventure oftheItalian banksin the ﬁ rstdecades of theXXcentury. Design/methodology/approach – The authors adopt a qualitative case study and a ﬂ exible pattern matching approach to develop theoretical ideas. Because the ﬂ exiblepattern matching approach was adopted, the authors build the analysis on a tentative analytical framework speci ﬁ ed a priori to provide guidance and focus. This approach allows a theory-driven research paradigm. The historical case study is mainly grounded onoriginalsources drawnfromsome majorbanking andinstitutionalarchives. Findings – While several scholars emphasised the relevance of external drivers, the literature has paid less attention to how relational and internal drivers combine. The historical case suggests that key mechanisms supporting the success of coopetition concern planning of common goals, con ﬂ icts management, alignment and formalisation of governance. In this vein, internal and relational dimensions seem to be more relevant than environment-context dimension. In addition, the historical example shows that an intra- and inter- ﬁ rm alignment is required to pursue the implementation of a coopetitive strategy. This suggests the relevance of a holistic approach to investigate coopetition. Further evidence con ﬁ rms the role of governance mechanisms for thesuccess of coopetition. Originality/value – The main contribution of this study is the re-consideration of the drivers of coopetition. In particular, the role of coopetition drivers has been investigated using a historical event: the Italian multinationalbanking in Egypt intheinterwaryears.


Introduction
Strategic interaction between firms might take different forms. Following a standard "competitive" paradigm, firms fight against each other and aggressiveness is perceived to be necessary to survive in the market (Porter, 1980). On the opposite, following a "cooperative" approach, the creation of strategic relationships with other firms benefits each firm, so that value creation is not related to the defeat of rivals but rather to the value cocreation between partners (Yami and Le Roy, 2007). However, firms are often characterised by a mixed attitude in their relations with the other firms, thus locating somewhere between the two extreme paradigms of competition and cooperation. In the case of a mixture of competition and cooperation, firms are said to engage in "coopetition" (Nalebuff and Brandenburger, 1996). This topic is receiving growing attention from scholars (Lascaux, 2020;Gernsheimer et al., 2021).
There is a vast economic and managerial literature focusing on the characteristics of coopetition (Walley, 2007;Gernsheimer et al., 2021). In particular, an increasing debate among scholars and practitioners focuses on the drivers of coopetition. Even if there is no consensus yet about the definition of coopetition or the drivers of coopetition, the existence of such a vibrating debate makes coopetition a hot topic nowadays (Yamy et al., 2010, Chen et al., 2021. However, while the economic and managerial literature on coopetition has grown significantly in the past two decades, historical examples of successful or unsuccessful coopetition have been rarely explored. In addition, as emphasised by some researchers "an empiric understanding about how contingencies influence the ways firms are managing coopetition over time has not yet been developed" (Gernsheimer et al., 2021, p. 131).
The aim of this paper is to investigate the key drivers of coopetition by adopting a managerial and economic framework. In particular, a case of coopetition failure is investigated by means of a historical example focused on the Egyptian adventure of the Italian banks in the first decades of the XX century. Indeed, the history of multinational banking seems to offer a promising field of investigation for coopetition theory: while the relationships between banks of different nationalities have been studied quite frequently in the literature, more attention still needs to be paid to the relationships between compatriot institutes on foreign markets (Austin and Uche, 2007). Besides, banking often represented one of the main fields of investment abroadthat was surely the case for Italy in Egypt, for exampleand the one able to engender spill-over effects over other sectors.
In general, it seems that the local market conditions are important in explaining whether coopetition arises or not. However, the characteristics of the banking sector in the country of origin are also relevant. The historical case we consider in this paper can help shed light on some of these issues precisely because of the peculiar characters of the local market (a formally independent country but in the British sphere of control, with a particularly numerous banking population) and the involved institutions. In this perspective, our hypothesis is that the Egyptian adventure of the Italian banks should be interpreted as a prominent example of coopetition failure.
The entry of the first Italian bank (Banco di Roma) in Egypt took place in 1905. After a period in which Banco di Roma remained the unique Italian bank with Egyptian subsidiaries, in 1924, another Italian bank (Banca Commerciale Italiana) entered the Egyptian market by establishing a subsidiary of its own. Therefore, starting from 1924, two Italian banks owned Egyptian subsidiaries [1]. Despite the competition from numerous banks of different nationalities, the Italian banks, whose business was mainly focused on financing cotton production, performed well during the 1920s, thanks to the positive trend of cotton prices. However, when the Great Depression broke out in the early 1930s and the cotton market went into crisis, Egyptian subsidiaries' business went into trouble. On top of this, the growing Egyptian nationalism made it more difficult for foreign firms, including the Italian ones, to operate in the Egyptian context. Lastly, the Second World War outbreak temporarily stopped the experience of the Italian banks in Egypt.
The Egyptian adventure of the Italian banks provides an interesting historical example to test coopetition theory across different geographical markets. Indeed, during the interwar period, the Egyptian environment became increasingly hostile for the Italian banks. This framework suggests an opportunity for the Italian banks to cooperate in their Egyptian business while still competing in the Italian banking industry, thus giving rise to coopetition. For example, a cooperative attitude would probably have improved their ability to face international competition in the cotton sector, which was the primary Egyptian production. However, this did not occur, with banks still competing both in Italy and Egypt.
Our analysis of such a historical experience aims to explain why coopetition failed to occur in the light of coopetition theory. In particular, we argue that while some factors, mainly related to the structure of the Egyptian banking environment and the Italian banks' characteristics, were conducive to inter-market coopetition, other factors, mainly related to individuals' characteristics, worked in the opposite direction, with the latter eventually prevailing over the former. In particular, the lack of flexibility and the absence of a cooperative attitude of the management of one of the Egyptian affiliates and the climate of mistrust and suspicion between the two Egyptian subsidiaries made it impossible to establish a cooperative agreement between the Italian banks in Egypt. Therefore, the interactive framework involving the Italian banks in their business activity in Egypt may be interpreted through the lens of potential coopetition.
The rest of the paper proceeds as follows. Section 2 provides a theoretical model of coopetition, focusing on the drivers of coopetition at the environment, firms and individuals' level. In Section 3, we briefly describe the historical background by focusing on the activity of the Italian banks in Egypt in the first decades of the XX century. In Section 4, we analyse the failure of coopetition in the Egyptian adventure of the Italian banks in the light of the interpretative model developed in Section 2. Lastly, Section 5 discusses and concludes.

A Theoretical model of coopetition 2.1 A definition of coopetition
Coopetition is a relatively new concept that has been introduced into the economic and managerial literature by Nalebuff and Brandenburger (1996), as a win-win relationship between competitors instead of competition, which is instead interpreted as a win-lose relationship [2]. Lado et al. (1997), even if not using explicitly the term coopetition, note that firms commonly combine aggressive and cooperative attitudes in firms interactions operating in the same environment. Furthermore, Lado et al. (1997) propose to interpret competition and cooperation as the two extremities of a continuum of interaction strategies between firms. As depicted by Bengtsson and Kock (2000), coopetition is a relationship between two or more firms that cooperate in some activities and competing in other activities.
Building on these seminal definitions of coopetition, Dagnino and Padula (2002) deepen the conceptualisation of coopetition by defining a coopetitive strategy as "a kind of interfirm strategy which consents the competing firms involved to manage a partially convergent interest and goal structure and to create value by means of coopetitive advantage" (p. 13).
Coopetition differs from both competition and cooperation. In pure competition, the interests of the firms are entirely opposed. Indeed, each firm tries to win at the expense of the rival. Thus, by using standard game theory terminology, competition is a zero-sum game (or, according to Nalebuff and Brandenburger, 1996, taxonomy, a win-lose relationship),

Italian banking rivalry
where the defection of the rival necessarily accompanies the success of one firm. On the opposite side of the Lado et al. (1997) spectrum, cooperation is a situation where the interests of the interacting firms are wholly aligned because the success of one firm depends on the other firm's success. That is, cooperation is a positive-sum game. Therefore, while the interacting firms are "rivals" in a competitive framework, in a cooperative framework, they are "co-operators" [3]. As Dagnino and Padula (2002) argue, coopetition is a strategic relationship that maintains some of both competition and cooperation elements. For instance, the firms cooperate on some aspects of their business activity while maintaining a competitive stance on others. Under coopetition, there is not a complete alignment of the interests of the firms. Indeed, a firm might benefit more than the other, then trying to exploit the partial cooperation to maximise its benefit at the detriment of the other firm involved in the agreement. In other words, while in pure cooperation the two firms act as they would be a unique organisation, in coopetition the two firms remain two separate entities in their actions so that cooperation in one or more activities results from their interest maximisation. However, coopetition is also distinguished from competition. Indeed, in a purely competitive setup, the maximisation of the own profits implies the minimisation of the rival's ones, which is not the case under coopetition. That is, under competition, there is an unrecoverable contrast between the firms involved in the interaction. Instead, under coopetition, there is partial convergence of the interests of the firms. According to Dagnino and Padula (2002), "coopetition is a matter of incomplete congruence of interests and goals concerning firms' interdependence" (p. 3).
In what follows, we propose a simple two-dimensional taxonomy of coopetition. The first dimension captures the distinction between coopetition arising within the same market and coopetition arising across different markets, considering a geographic perspective. In the first case, where the two firms compete and cooperate within the same market, we refer to intra-market coopetition. In the second case, where the two firms compete in one market and cooperate in another market, we refer to inter-market coopetition. In particular, we will refer to the market where the firms compete as the "competitive" market, whereas we will refer to the market where the firms cooperate as the "cooperative" market [4].
The second dimension of the taxonomy is related to the level at which coopetition occurs. Even if coopetition has been mainly investigated at an inter-organisational level, i.e. by focusing on the relationships between two or more firms mixing competition and cooperation (Lorenzoni and Lipparini, 1999), coopetition may occur at higher or lower levels. For example, coopetition between districts, industries and consortia is possible. In this case, the literature refers to coopetition in networks (Bengtsson and Kock, 1999). However, coopetition might exist at lower levels, such as between groups or departments belonging to the same firm, or even between individuals operating within the same institution (intraorganisational or micro-level, coopetition, according to Bengtsson et al., 2010). Cross-level interactions are also possible, as the processes and outcomes at one level of coopetition are expected to impact the processes and the outcomes at another level of coopetition.

Coopetition in the business networks literature: the cooperation-competition interplay
In the business networks literature, coopetition has been investigated by some main schools of tought. According to , the actor school of thought describes coopetition as a situation where a mixture of cooperation and competition among actors arises (Nalebuff and Brandenburger, 1996). For instance, a focal firm cooperates with some actors and competes with others in the marketplace (Afuah, 2000). Following the business network approach, heterogeneous actors are interconnected by business relationships in a network, and such an interconnection allows them to share resources (Håkansson et al., 2009). Lastly, the activity school of thought focuses on coopetitive relationships where firms simultaneously cooperate in some activities (i.e. upstream activities such as research and development (R&D), purchasing and processing of raw material) and compete in others (i.e. downstream activities such as service/product development and marketing) (Bengtsson and Kock, 2000). In this case, Bengtsson and Kock (2000, p. 412) define coopetition as the: Dyadic and paradoxical relationship that emerges when two firms cooperate in some activities, such as in a strategic alliance, and at the same time compete with each other in other activities.
Several researchers focused their attention on coopetition related to horizontal relationships considering competitors belonging to the same industry (Dahl et al., 2016), while other scholars investigated coopetition in vertical relationships, such as between buyer and seller (Lacoste, 2012). In addition to this, coopetition has been investigated by means of multilevel analyses. From this perspective, relational drivers at external and internal levels may influence coopetition on the inter-firm, intra-firm, dyadic, network and inter-network levels Raza-Ullah, 2016, Resende et al., 2018).
There is a wide consensus about the potential benefits of coopetition. Coopetition allows to share resources and skills, supports innovation development (Yami and Nemeh, 2014;Xu et al., 2021) and reduces the monetary loss in case of failure (Yami and Nemeh, 2014;Schiavone and Simoni, 2011). Coopetition improves value creation in the supply chain (Resende et al., 2018) and allows to obtain technological benefits from competitors while protecting market shares at the same time.
Even if the benefits of coopetition could be very high, coopetition might not always lead to high performances for the participants (Crick and Crick, 2021). This happens, for example, when it is difficult to protect the knowledge which is shared across coopetitors (Le Roy et al., 2021;Gast et al., 2019;Fernandez et al., 2018).
Researchers identified some drivers, mainly related to inter-relationship factors and internal factors (Resende et al., 2018), that might influence the coopetitive performance of business networks. The inter-relationship factors concern the interaction between the companies and the network, such as trust and commitment, and the cooperative relationships between network companies, such as complementarity and reciprocity. In addition, these factors concern the internal cohesion of the network, such as exchange of experiences and learning, history and identity. These factors also consider sharing and equity, management of conflicts and incompatibilities, control and standardisation, adaptability and alignment, inter-dependence and governance. On the other hand, internal factors are identified in strategy and management, production competence and innovation competence (ability to innovate and meet the market demand). These factors also relate key resources, such as financial resources, people management and intangible resources.

The drivers of coopetition: a tentative managerial and economic analytical framework
In this section, we propose a theoretical framework of the drivers of coopetition, which is based on the characteristics of the competitive environments, the firms and the individuals.
Let us consider first the role of the competitive environment in determining whether coopetition is likely to arise or not. As coopetition is an intermediate form of strategic Italian banking rivalry interaction between pure competition and pure cooperation where elements of both competition and cooperation are mixed, various degrees of coopetition are possible, depending on where coopetition is located along the spectrum of Lado et al. (1997). According to the neoclassical theory of markets (Williamson, 1985), pure (or perfect) competition is more likely to arise when there are many identical firms, no one of which is endowed with market power. In addition, the widespread diffusion of relevant information about products and prices makes the market close to the paradigm of perfect competition [6].
Overall, atomism is a requisite for perfect competition. Let us consider now the case of a cooperative market. It is commonly argued that cooperation is more accessible when the number of firms is limited. Indeed, a vast number of firms makes cooperation more difficult to create and maintain. The first problem relates to cooperation costs (Thomadsen and Rhee, 2007;Colombo, 2013), sustained by the firms when they build up the agreement and during the agreement's lifetime. Indeed, both when the agreement is explicit as in the case of legal cooperation and when it is implicit as in the case of illegal cooperation (i.e. collusion), the firms need to find a common set of objectives and behaviours that all cooperating firms should agree. Therefore, a bargaining process is needed, which is likely to be more complicated the greater is the number of participants to the agreement. The cooperation costs emerge both before the agreement is constituted (i.e. preliminary meetings to explore the possibility of the agreement) and during the life of the agreement (i.e. ongoing meetings between the participants to evaluate the agreement's progress). The second problem relates to the possibility of defection from the agreement. Indeed, an agreement has not only to be created, but none of the participants must deviate from the partnership once it exists. When the number of participating firms is large, detecting the defection of a firm is difficult, thus enlarging the incentive to defect [7]. Secondly, when considering the transparency of information, it should be observed that private information is a valuable asset for each firm. As profitable cooperation is more likely to occur when firms have something to exchange, private information is more conducive to cooperation than competition. With this regard, the possibility of a profitable exchange of private information can open the way to cooperative attitudes by firms. In other words, homogenously dispersed information is more conducive to a purely competitive framework rather than to a cooperative one.
Third, a competitive environment characterised by severe instability might induce firms to cooperate to overcome market riskiness. Indeed, as suggested by Bakshi and Kleindorfer (2009), when two firms face a common threat (e.g. demand instability, but also terrorist attacks or natural hazards) on one market, they are more prone to cooperate on that market to share the risks and losses or to mitigate the probability of the negative outcome (e.g. by sharing relevant information). Conversely, a relatively safe and easily predictable environment reduces the need for cooperation, hence the likelihood of coopetition.
Lastly, government and regulators play a crucial role (Carles Maixé-Altés, 2010) [8]. Indeed, such external players often design the "rules of the game", i.e. the rules governing the competitive environment where the firms interact (Depeyre and Dumez, 2010). One possibility is that government and antitrust authorities shape an environment more prone to cooperation through weak antitrust policies or resorting to active policies (e.g. by using incentives or law reforms) to encourage cooperation between firms. Notably, the government might also force the firms to cooperate. On the opposite, a strong attitude by government and antitrust authorities favouring pure competition might undermine any attempt by the firms to cooperate (Wu et al., 2010). Some empirical studies confirm that the external players might instigate coopetition. For example, Mariani (2007) shows that a regional policymaker induced cooperation among three competing opera houses in Italy.
Summarising, when considering the impact of the competitive environment characteristics on inter-market coopetition, we claim that coopetition is more likely to emerge when one market (the "competitive" market) is characterised by a vast number of firms, high transparency, high stability and a negative stance of the authorities toward firms' cooperation, whereas the other market (the "cooperative" market) is characterised by a small number of firms, low transparency, high instability and a positive stance of the authorities towards firms' cooperation.
Now we turn to consider those characteristics of the firms that are more conducive to coopetition. Firstly, coopetition is a dynamic process that requires time to be built. As coopetition entails elements of cooperation, trust is fundamental. Therefore, reciprocal knowledge that might consolidate trust in the partners is a necessary element for coopetition. As trust is the outcome of a long-time process, the coopeting firms should be in the market for a long time, so reciprocal trust is possible. According to Zineldin (2004), coopetition can be compared to marriage. As the bride and groom need to know each other before getting married, in coopetition the firms must know each other before starting the partnership. Indeed, like in marriage, reciprocal knowledge allows anticipating and coping with the conflicts that might arise during the partnership. Therefore, long-time knowledge is a necessary (even if not sufficient) condition for cooperation in a market [9].
Secondly, a certain degree of product homogeneity is required. Indeed, if the products or the services offered by the two firms are very different, it might be difficult for them to find common ground for coopetition. Indeed, partial convergence of interests requires some similarities between the products commercialised by the two firms. In other words, too much difference between the firms might yield an excess of goal dissimilarities, making cooperation difficult, if not impossible, to reach (Chen and Fan, 2006). At the same time, complete product homogeneity might disrupt the potential for coopetition. Indeed, the two firms coopete as long as each of them takes advantage of coopetition. A source of benefit from coopetition is the possibility to learn from the rival/partner. Therefore, a moderate product differentiation might help to create the benefits of coopetition.
Quite related to this point, the similarity/dissimilarity of human, financial and technical resources at the disposal of the firms might enlarge or reduce the scope of coopetition [10]. When there is a strong asymmetry between the firms regarding resources, an agreement might be difficult to reach. At the same time, too much similarity of the resources might reduce the benefits expected from coopetition. Therefore, a moderate asymmetry between the firms might open the way for competition in one market and cooperation in another market, i.e. coopetition. For example, according to Sahaym et al. (2007), firms from different industries might have more incentive to coopete, as they can integrate their technological skills and knowledge to develop new ideas or products [11].
A third firms' characteristic that impacts the profitability of coopetition is related to the weakness of the firms aiming to coopete relatively to the other firms in the markets. Indeed, when a group of firms is weak, it might find it difficult to survive in a hostile environment characterised by strong competitors. In this case, the only chance of survival for the weak firms comes from sharing the resources (i.e. joining the forces) in the hostile market, thus coopeting. Therefore, the relative weakness of the firms in one market might enlarge the scope for cooperation in that market. On the other hand, by the same line of reasoning, if there is no weakness in the other market, the firms might compete in that market (i.e. the incentive to cooperate is reduced).
When considering the impact of the firms characteristics on inter-market coopetition, we claim that: Italian banking rivalry long-time knowledge between firms is a necessary albeit not sufficient condition for coopetition to occur; coopetition is more likely to emerge when the firms are moderately similar in terms of products and internal resources; and coopetition is more likely to emerge when the firms are sufficiently strong relative to the competitors in one market (the "competitive" market) and sufficiently weak relative to the competitors in the other market (the "cooperative" market).
Lastly, we consider which characteristics of the individuals working in the firms might drive coopetition. Firstly, coopetition is the result of a change in the dynamic interaction between the firms. Coopetition typically emerges when two competing firms cooperate in some markets or relatively to some aspects of the production process [12]. Therefore, coopetition induces a relevant change in the intra-firm organisational routines. Moving from a competitive approach to a cooperative one requires the individuals more directly involved in switching from competition to cooperation a radical change when approaching the relationship with the other firms, which is no more a competitor but rather a partner. This change might be tricky if the individuals are not sufficiently flexible. Note that flexibility refers to both the analytical and the executive capability to face the change (De Graaf, 2018). Indeed, as pointed out by Gnyawali et al. (2016), coopetition might be realised in practice only when the individuals can understand, accept and adopt the (partial) switch from competition to cooperation (i.e. analytical flexibility) and only when they can create new routines and execute coopetition day by day (i.e. executive flexibility). Therefore, analytical and executive flexibility at the individual level is necessary for coopetition to be effectively realised [13]. Secondly, the existence of personal values consistent with the coopetitive approach is crucial for coopetition to exist. Indeed, as Bengtsson et al. (2010) pointed out, individuals' values related to competition are prestige and pride, whereas individuals' values related to cooperation are a sense of commitment and trust. Therefore, inter-market coopetition requires that those working in the cooperative market posit more importance to the partnership's success than to the firm's success; on the opposite, those working in the competitive market should consider the firm's success as their primary objective. A crucial role is played by reciprocal trust between the more directly involved individuals in the cooperative part of the coopetitive process. Indeed, as Park et al. (2014) argued, reciprocal and individual-level trust are needed to allow learning and knowledge sharing among coopetitors and limit the fear of partner's opportunism.
Similarly, Ritala et al. (2009) notice that trust in the other party's ability and goodwill to cooperate is necessary to maintain sufficient flexibility in the coopetitive relationship between the firms, thus avoiding that a formal and rigid contract entirely regulates the firms' interaction. At the same time, trust should not be excessive. Indeed, as empirically demonstrated by Wu (2014), there is an inverse U-shaped relationship between trust and firms' performance within a coopetitive relationship: for low levels of trust, a greater trust allows more cooperation and information sharing, thus pushing up the performance of the firms, but for high levels of trust, a greater trust might induce opportunistic behaviours by the partner, with a negative impact on the outcome.
When considering the impact of the individuals' characteristics on inter-market coopetition, we claim that: both analytical and executive flexibility of the individuals are necessary for coopetition to occur; and coopetition is more likely to emerge when there is a moderate and reciprocal trust between the individuals. Figure 1 summarises the analytical framework we adopt to interpret the historical experience of the Italian banks in Egypt in the first decades of the XX century, as described in Section 4. The methodology is illustrated in the following section.

Research design 3.1 Research approach
We adopt a qualitative case study (Yin, 2009) and a flexible pattern matching approach to develop theoretical ideas (Bouncken et al., 2020;Sinkovics, 2018). To explore the paradoxical aspects of coopetition, a qualitative and explanatory case study has been recommended for example by Chiambaretto et al. (2019) and Chiambaretto and Fernandez (2016).

Data collection and analysis
As we adopt the flexible pattern matching approach, we build our analysis on a tentative analytical framework specified a priori to provide guidance and focus (Section 2). This approach allows a theory-driven research paradigm (Le Roy et al., 2021). In what follows, we provide a tentative explanation of coopetition failure by analysing the Egyptian adventure of the Italian banks in the interwar period within the theoretical framework of coopetition we have developed in Section 2 to characterise the preconditions for inter-market coopetition (competition in Italy between parent banks, cooperation in Egypt between affiliates). The case has been selected because of its relevance. Therefore, the paper adopts a qualitative methodology and a case study approach (Beverland and Lindgreen, 2010).
Due to the largely missing literature on these issues (historiography on Italian banking overseas proves rather scattered), the case study is mainly grounded on original sources drawn from some major banking and institutional archives, such as the Historical Archives of UniCredit, Milan; the Historical Archives of Intesa Sanpaolo, Milan; the Historical Archives of the Bank of Italy, Rome; the Historical Diplomatic Archives, Rome; the Central State Archive, Rome.

The case of coopetition
The local community of expatriates long established in Egypt took the initiative to found the first "Italian" bank in the country [14]. As a result, the Cassa di Sconto e di Risparmio was established in Alexandria in 1887, five years after the occupation of Egypt by British troops. The bank primarily addressed the Italian colony, made up of small traders, entrepreneurs and workers [15]. Therefore, it occupied a relatively marginal stance compared to the market leaders, above all British and French banks [16].
3.3.1 1905 -First World War. In 1905, an Italy-based bank set up its branches in Egypt, with the Banco di Roma landing on Alexandria and then expanding into Cairo, Port Said and the primary centres in the interior (Banco di Roma, 1942, pp. 205-211). The Banco di Roma carried out an unscrupulous policy in this country, aiming to acquire customers among Italians and the indigenous population.
Furthermore, during the First World War, the Banco extended its activity to the commercial and industrial fields. In doing so, it took up risks that went well beyond the pure banking business. In November 1917, the director of the Egyptian office, Riccardo Interdonato, sent a large investment project to the steering committee in Rome. He intended to create a holding company that would control a group of shareholdings consisting of an import-and-export company, a warehouse company, a river transport company, a shipping company and an export cotton company (De Rosa, 1983). This was too large a project for an Italian institute operating in a country controlled by the British, politically allied and a rival for economic dominance in the Middle East.
3.3.2 First World War -1924. This ambitious project eventually failed. It brought about heavy losses and made an overall reorganisation of the institution in Egypt necessary by creating a subsidiary controlled by the Banco [17]. Economic and political reasons were at the base of this decision. On the one hand, reducing the bad credits accumulated over the years was urgent, transferring only the most liquid assets to the new institute and gradually giving up the others. On the other, the mounting nationalism (in 1922, Great Britain formally recognised Egyptian independence, albeit with substantial limitations) and fears of a possible abolition of capitulations were also pressing to make the Italian presence more embedded in the local economy [18] Thus, in February 1921, the Banco del Levante, Fondazione del Banco di Roma was founded (De Rosa, 1983, pp. 271-275).
From the very beginning, the new institute took advantage of a favourable situation in the cotton market, which was the most valuable "money crop" and Egypt's main export. However, in the meanwhile, Banco di Roma went into crisis in Italy, and its bankruptcy was averted only thanks to state intervention (Confalonieri, 1994, pp. 266-282;Cafaro, 2014, p. 120). Against this backdrop, Banco di Roma took the opportunity to dilute its stake in the Egyptian affiliation. A first option was to share the burden with the other three leading Italian banks, the Banca Commerciale Italiana (Comit), the Credito Italiano (Credit) and the Banca Nazionale di Credito (Bnc). However, because Comit attempted to impose its control on the nascent institution, an agreement was finally reached in 1924 among Banco di Roma, Bnc and Credit, reserving the latter a 40% stake and 30% each for the other two institutions. The new bank was named Banco Italo-Egiziano (Italegi) (De Rosa, 1983, pp. 422-428). Comit, for its part, finally decided to establish a subsidiary of its own under the name Banca Commerciale Italiana per l'Egitto (Comitegit) (Petricioli, 2007, pp. 139-140).
The two Italian banks were to operate in a country populated by several credit institutions of different nationalities, and they also had to face the unfriendly attitude of the British diplomatic authoritieswhich had suspected the Italians of acting against British interests in the Middle East since before the war (Cassar, 2016, pp. 196-197). Environment hurdles notwithstanding, in the 1920s, they managed to conquer a prominent place in the Egyptian banking market and collect an amount of deposits that, although not comparable to that of the leading institutes, allowed an appropriate base of operations ( Figure 2). Apart from the Egyptian branches of British and French large banks such as Barclays DCO and Crédit Lyonnais, the Italian institutions differed much less in capital invested and deposits raised than other foreign commercial banks.
3.3.3 1925 -Great Depression. How did they manage to reach this relative development? Both Italian banks mainly focused on financing cotton production, in which most commercial banks were engaged. This choice was probably explained by the need to make up for the reduced margins of activity still free in other sectors of the Egyptian economy, such as public worksthe prerogative of English, French and Belgian capitaland the industrial sector, the development of which was, however, still minimal. Besides, the demand for cotton by the Italian industry (in the second half of the 1920s, raw cotton accounted for 10.3% of total Italian imports) (Federico et al., 2011, pp. 19-21) and the centrality of this commodity in the Egyptian economy (in 1930-1931, cotton represented ) foreshadowed some positive developments in the years to come. Therefore, cotton rising prices allowed Italian banks to prosper in the 1920s. However, their close dependence on the performance of this commodity on international markets turned out to be their primary source of immobilisation once the Great Depression broke out in the early 1930s, and the cotton market went into trouble.
The general crisis soon involved the Italian banks as well. Indeed, precisely cottonexporting countries such as Egypt were among the most affected by the crisis, both because this product was the most valuable agricultural commodity in international trade and because demand for it was relatively elastic.
Egyptian exports collapsed by 60%-65% in value between 1928-1929 and 1932-1933, while in August 1931, the price of cotton in Liverpool was a third of that in August 1929 (Tignor, 2017, p. 113). By 1933, Egyptian cotton prices had fallen to a sixth of 1929 prices, which impliedceteris paribusreducing bank turnover [20].
Consequently, in the early 1930s, the Egyptian government was forced to intervene. It did so in two ways. Firstly, it bought domestic cotton at a price higher than the world price in a vain attempt to support its quotations. Secondly, it tried to make farmers' debts more sustainable after the prices fell. The underlying meaning was to support debtors at the expense of creditors by stopping land expropriation by the banks, which were now immobilised by the number of mortgage loans granted in the previous decade. These provisions went against the interests of foreign banks that had to face increasing difficulties to operate in a country characterised by marked nationalism (Petricioli, 2007, p. 160).
Comitegit was probably the most affected, while the archival sources are less explicit regarding Italegi. In fact, due to the fall in cotton prices, the banks that financed the cotton production had their profits cancelled out and many credits granted in the past were no longer sufficiently guaranteed. As a result, no longer able to recover part of these funds. Comitegit sought to transform the credits previously granted into covered loans and obtain guarantees against the remaining uncovered after the cotton sale. As most of the assets owned by farmers and cotton merchants consisted of land, Comitegit needed to constitute a mortgage branch [21]. Italegi also generally managed to cover itself with mortgage guarantees [22]. In this way, however, some cotton credits were transformed into mortgage procedures, "that is, into the longest, most quarrelsome, random procedures in which a foreign bank could be involved" [23].
Therefore, two main factorsthe cotton falling prices and the slow recovery of mortgage loansundermined the solidity of the two Italian institutions, which from that moment entered a phase of reorganisation and progressive narrowing of their business. As a result, by the eve of the war, the two banks had closed the whole structure of cotton agencies to focus on Alexandria, Cairo and Port Said. However, how they achieved this identical result differed substantially because of diverse views and strategies.
3.3.4 1934-1939. Faced with a gradual loss of its capital (sources estimate a loss of fourfifths of the paid-up capital [24]), Comitegit rapidly converted its business model through a progressive reduction of the cotton work until the complete closure of all the agencies in the interior in 1936. This plan allowed most of the bad credits to be reabsorbed, from 1.1 million Legs (Egyptian pounds) in June 1934 to 274,000 in December 1937. Italegi, for its part, persevered in cotton financing until 1939. Only the outbreak of the Second World War convinced its director, Enrico Biagi, to limit the activity to the Alexandria, Cairo and Port Said offices, but always in the hope of being able to reconstitute the organisation in the interior once a normal situation returned [25]. Indeed, Biagi did not consider Italy's growing political and financial isolation after the invasion of Ethiopia, which would cause much trouble to the Italian banking abroad. Especially a major country such as Egypt would be immediately involved in case of an Anglo-Italian war.
The lack of resources to continue the financing of cotton production was one of the main reasons for the Italegi's increasingly marginal position in this field to benefit British banks, which ended up almost monopolising this activity (Berbenni, 2022). The definitive loss of influence in the cotton financing sector ultimately deprived the Italian banks of the only field of activity in which they had previously managed to exert a strong presence at the expense of the competition.
3.3.5 1940. In 1940the situation was not significantly different in previous yearsthere were 21 banks, attributable to British, French, Belgian, Italian, Greek, German, Turkish, Japanese and Egyptian capital [26]. Immediately after Italy entered the Second World War, the Egyptian government seized both Italian banks, and this experience would only be resumed at the end of hostilities.

Results
The Italian colony had been particularly influential in Egypt during the XIX century, and a large community of Italians resided in that country still in the early XX century. However, the local environment became increasingly hostile to the Italian presence in the interwar. The English distrust of any Italian economic initiative in the Middle East was a primary cause, adding to Egyptian nationalism during the 1930s. Nationalists were increasingly hostile to foreign capital, even though they had been exploited instrumentally by the Italians to weaken the British leadership after the First World War (De Rosa, 1983;Petricioli, 2007, p. 160). Furthermore, the two Italian banks had to compete in a market populated by a relatively high number of foreign banks of different sizes and nationalities.
Therefore, it would have been better for the Italian banks to cooperate rather than compete in the Egyptian market. Moreover, in the aftermath of the First World War, the Italian authorities pushed for a joint initiative of some leading Italian banks in the Middle East [27]. So, why did not they cooperate?

The competitive environment
We consider firstly the competitive environment. Concerning the number of companies, the Egyptian banking system differed from the rest of the colonial world (Egypt was formally independent but under British tutelage) as to the number of credit institutions and the variety of national interests represented (Crossley and Blandford, 1975, p.72).
In 1940the situation was not significantly different in previous yearsthere were 21 banks, attributable to British, French, Belgian, Italian, Greek, German, Turkish, Japanese and Egyptian capital. Intense banking competition led to a partial segmentation of the market. The Italian banks specialised in financing cotton production, although not exclusively and still in competition with banks of other nationalities (Berbenni, 2022). Therefore, a limited agreement between the two Italian institutions could also occur because both aimed at the same target of customers of average social extraction, belonging not only to the populous Italian community but also to other nationalities [28].
Despite these encouraging factors, reaching agreements proved easier with other foreign banks than between the Italian ones, as demonstrated by a banking cartel that Comitegit joined in Alexandria, but Italegi did not. Besides, Comitegit repeatedly accused Italegi of exercising unfair competition in terms of interest rates applied to customers, pursuing an aggressive hoarding policy that was not sustainable in the long term. The interest rate granted to depositors (in Cairo, Italegi granted 3%, while Comitegit could not exceed 2.5%) Italegi, in turn, repeatedly expressed his disappointment at the incorrect behaviour of Comitegit, e.g. by competing for the most prominent personalities of the Alexandria business circles to be co-opted into its board of directors [30]. Concerning information transparency, both institutes possessed information and experience that could be of mutual benefit. Comitegit had a privileged relationship with the Egyptian Royal House since its foundation, which also held a stake in the bank's share capital. In this regard, among the Comitegit's shareholders was a particularly influential personality with the king such as Ernesto Verrucci, an Italian architect who had been the chief architect of the royal palaces for almost all interwar. Due to his influence, the British suspected him of being a spy in the pay of the Italians (Petricioli, 2007, p. 378, 391). It is impossible to quantify how much this relationship influenced Comitegit's business, but it represented an informational and relational advantage. On the other hand, the strong expertise gained by Banco di Roma in the cotton sector represented a know-how asset inherited by Italegi, while Comitegit could not make use of it (as demonstrated by the more significant losses suffered by Comitegit in the cotton sector during the Great Depression). Thus, according to the coopetition theory, a basis would exist to share the respective information and relational sets.
The instability of the Egyptian market should also have induced the two Italian institutions to cooperate. From a macroeconomic perspective, the consequences of the Great Depression were particularly severe in developing countries, which specialised in the export of individual commodities, as was the case with Egyptian cotton (Foreman-Peck, 1995). It was precisely this economic inversion that made Italian banks more fragile because they had specialised in financing that commodity.
Political instability also made the Egyptian case especially noteworthy. One reason lies in the Egyptian nationalism, which Italian banks sought to exploit in the post-war period to attract even indigenous customers. However, growing nationalism in the 1930s turned out dangerous for all foreign institutions in the country, making it appropriate for them to include some Egyptian names on the board of directors to take on a more local character. Furthermore, British diplomacy proved wary towards any Italian economic interests in the Middle East. This mistrust grew to almost lead to war following the Italian invasion of Ethiopia in 1935 and condemned Italian banks to play an increasingly marginal role in the Egyptian economy.
Moreover, there appears to be no formal or informal prohibition against cooperative behaviour between credit institutions. Although a deeper investigation is required on this topic, the existing inter-bank agreements regarding the setting of interest rates suggested that authorities would not hinder, at least openly, such behaviour between firms.

Characteristics of the firms
Concerning the second driver of coopetitioni.e. the characteristics of the firmswith particular regard to mutual knowledge, there seems to be no doubt that this condition was fully satisfied, both regarding the parent banks and the two Egyptian banks. The former (Banca Commerciale Italiana, Credito Italiano and Banco di Roma) were, in fact, the three large mixed banks, which in the past had contended the Italian credit market and which, following the 1929 crisis, had entered a crisis from which they only emerged thanks to a public bail-out. As a result of this, in 1934, their control capital had passed into the hands of IRI, a public body created in the previous year to restore that part of the banking system that was most immobilised and to allow it to resume providing liquidity to the industrial system (Confalonieri, 1994). Therefore, they had a similar operational and managerial structure and were under public control in the second 1930s, although they kept acting formally as competitors. The Egyptian affiliates were also very similar institutions, both from an operational perspective (both focused on financing cotton production) and from an organisational point of view (their general managers resided in Alexandria, but the executive committee and the board of directors were in Milan). Therefore, they knew each other and monitored each other, but this mutual knowledge translated more into a climate of mistrust than of collaboration [31].
Regarding the product homogeneity/differentiation dimension, the Egyptian business of the two banks was similar, having opted for a specialisation in the financing of cotton. This resemblance made the two institutions more fragile when the Great Depression heavily hit this commodity and made them almost identical in terms of the type of customers they were targeting (they were also very similar in terms of deposits collected). It thus discouraged a cooperative attitude in a market like that of cotton in which Comitegit and Italegi were direct competitors (see Figure 3, which shows that the cotton agencies and offices of the two banks overlapped to a certain degree). The different strategy followed since the mid-1930s (Comitegit immediately abandoned the cotton business, Italegi maintained it until 1939) seems more the consequence of a lack of cooperation than the cause of a growing conflict between the two institutes [32].
Regarding the third dimensionthe weakness of the firmsit seems instead that the characteristics of the two banks should have induced them to cooperate. They operated in a politically and economically hostile market. The presence of numerous institutions of different nationalities represented a peculiarity of the Egyptian banking system, and the losses accumulated in the 1930s had made the Italian banks extremely precarious [33]. In 1938, Francesco Pomi (the director of Comitegit) reported that Enrico Biagi (the director of Banco Italo-Egiziano) had stated, evidently in a moment of despair, that "an Italian bank in Egypt is a 'nonsense' and two banks 'two nonsense'"! [34].

Characteristics of individuals
Lastly, we consider the characteristics of individuals. This aspect is more complex to verify from a historical perspective, but some personality traits seem not favourable to cooperative behaviour. In this regard, Biagi's figure appears emblematic. Biagi, who had led Italegi since the end of 1929, had always supported the convenience of keeping the cotton business alive, repeatedly showing a sort of impatience concerning the directives from Italy. Even in the face of the numerous bad credits that Italegi had accumulated in the cotton market, Biagi proved not very flexible and was reluctant to abandon this branch of activity, thus slowing down the bank's recovery.
Furthermore, he showed much strength of character and a tendency to operate independentlyvery different character traits compared to Pomi, who was more "disciplined" in the face of Comit's directives. Remarkably, Biagi was defined as a person "intolerant of controls and [who did not] want to be joined by someone with whom he has to share the responsibility of management" [35]. Therefore, the governance difficulties appear particularly serious in Italegi, probably contributing to explaining the non-cooperative attitude of the two Italian institutions. Indeed, the lack of flexibility of the Italegi Italian banking rivalry Figure 3. and Comitegit (1935) management in Egypt made it challenging to develop an agreement between the two banks. Biagi's personal traits also compromised relations with the executive committee, even though he was never dismissed from office (for the impact of management hubris on the likelihood of bank's failure, see also the historical case described by Hollow, 2014). In the latter's documents, a Biagi's profile was found that highlights his difficult temperament, exemplified by the:

Branches of Italegi
Not very considerate correspondence addressed by the same to the Committee and [by] the obstructionism that the same had revealed about the credit proposals of the cotton campaign that had been requested of him [35].
Furthermore, when the executive committee sent an official to oversee the organisation, inspectorate and personnel services, Biagi hindered his work, showing a marked unwillingness to share management responsibilities with others [35]. Against this backdrop, the relations with the other Italian subsidiary in Egypt and, specifically, with its director Pomi, should also have been somewhat complicated, although their actual nature must be inferred indirectly from the archival sources. Episodes such as the simultaneous opening of a branch of Italegi and Comitegit in a strategic place like Port Said in 1937 were undoubtedly an occasion for friction between the two managers. While Biagi accused Comitegit of unfair competition, Pomi did not hide his intention to absorb the other Italian bank, writing (6 October 1938) to the CEO of the Banca Commerciale Italiana, Raffaele Mattioli, that the time had come to do so, given the state of weakness of Italegi [36].
Writing a letter to the executive committee (19 March 1937), the Italegi's management in Alexandria used very harsh words that clarify the relationship between the two Italian banks: It is our precise impression that the bank in question [Comitegit] follows for some time now a particular policy that consists of grabbing customers at any cost, increasing its turnover even at very onerous conditions, creating a situationalbeit artificialthat can allow it to demonstrate its vitality in this country. What is the aim of this policy? [. . .] We must face a problem: how countering this policy and stemming the efforts of this competitor [emphasis added] who mainly targets our customers? [37] 5. Conclusions

Theoretical implications
The main contribution of this study is the re-consideration of the drivers of coopetition. In particular, the role of coopetition drivers has been investigated employing a historical event: the Italian multinational banking in Egypt in the interwar years.
On a broader perspective, literature has usually classified the drivers of coopetition in external drivers (environmental conditions), relational specific drivers (partners characteristics and relationship characteristics) and internal drivers (internal goals/ capabilities) . When considering a multi-level perspective, coopetition is related to individual, organisational, inter-organisational and network levels (Dagnino and Padula, Tidstrom and Rajala, 2016). While several scholars emphasised the relevance of external drivers, the literature has paid less attention to how relational and internal drivers combine.
We identify those factors that might facilitate the emergence of coopetition by adopting a managerial and economic approach focused on the characteristics of the competitive environment, the firms and the individuals. Then, we apply the theory developed so far to the historical event of the Italian multinational banking in Egypt in the inter-war years.

Italian banking rivalry
Applying the coopetition theory to the historical case of the Italian banks in Egypt leads to conflicting conclusions. While some factors were conducive to inter-market coopetition, other factors worked opposite (Table 1).
Our historical analysis shows the importance of the individuals' characteristics and other factors related to operational similarity to determine the success or failure of coopetition between firms. We argue that individuals' characteristics have been the leading cause of the coopetition failure, together with excessive homogeneity of the business models adopted by the banks. In particular, the prevailing distrust among the parent institutions and the lack of flexibility of the management of one of the Egyptian affiliates impeded the formation of a cooperative agreement in Egypt. Therefore, even if the competitive environment seemed strongly favourable to cooperation between the subsidiaries in Egypt (thus giving rise to inter-market coopetition), cooperation in Egypt did not arise due to the unfavourable attitudes of the individuals more directly involved in the firms' interactions.
From this perspective, the analysis of the coopetition failure of the Italian banks in Egypt suggests that key mechanisms supporting the success of coopetition concern planning of common goals, conflict management, alignment and formalisation of governance. In this vein, internal and relational dimensions seem to be more relevant than environment-context dimension. In addition, our historical example shows that an intra-and inter-firm alignment is required to pursue the implementation of a coopetitive strategy. This suggests the relevance of a holistic approach to investigate coopetition.
Recognising that strategic intent is an antecedent of coopetition (Seepana et al., 2020), our historical analysis also shows the key role of governance in order to develop a coopetitive strategy. In a general perspective, governance refers to organisational or structural arrangements designed to determine and influence the behaviour of organisation members (Das and Teng, 1998). Governance mechanisms are implemented by firms to govern interorganisational exchange, minimise exposure to opportunism and protect transactionspecific investments (Jap and Ganesan, 2000). Relational governance is based on shared goals, and it is used to maintain cooperation through relational norms (e.g. trust, commitment, cooperation) and joint actions (Heide and John, 1992). A relational governance founded on joint planning and joint problem-solving could allow the management of internal criticisms. In other words, governance allows to manage tension and contradictory logics of the interaction (Gernsheimer et al., 2021).
Further evidence confirms the role of governance mechanisms for the success of coopetition. In 1918, the Italian government supported a project to establish a single institution in the Middle East, which failed due to the lack of agreement between the leading banks [38]. In the subsequent years, these differences remained. It is interesting to note that even among the same shareholders of Italegi (Credit and Banco di Roma), the relations deteriorated rapidly during the 1930s due to the worsening conditions of the Egyptian subsidiary. Giuseppe Veroi (the managing director of Banco di Roma) identified the scapegoat in the director of Italegi, Enrico Biagi, whom the Credito Italiano had appointed. It thus allowed the Banco di Roma to ask the latter for a review of the agreements to rebalance the decision-making power of the two shareholders, hitherto strongly unbalanced in favour of Credito Italiano. Furthermore, after the Italian conquest of Ethiopia in 1936, the Banco di Roma expressed growing ambitions to become the government's favourite Italian institution to lead expansion into the colonial empire (De Rosa, 1984, p. 155, Tuccimei, 1999. However, this ambition alarmed both Credit and Comit, which negotiated to merge the two Egyptian banks to reduce the interference of the Banco di Roma in Italian East Africa [39].

Managerial and policy implications
The analytical framework developed in this paper offers a tool that can be applied in the management of coopetition. In particular, it emphasises the relevant role of cultural dimensions of human resources, as they that can facilitate or limit a coopetitive strategy. In addition, coopetition requires a relational governance that might overcome the top-down perspective.
Even if coopetition has been mainly associated with dyadic relationship (Pellegrin-Boucher et al., 2013, and relationships between firms in a "value net" including suppliers, clients, competitors and complementors (Nalebuff andBrandenburger, 1996, Bouncken andFredrich, 2012), more recently new forms have been identified, such as international coopetition (Le Roy et al., 2021). In this vein, when considering the internationalisation strategy, new intermediaries are required to support the strategy of coopetition. Such new intermediaries could allow firms to better understand the international environment and the cultural business dimensions, as well as they could support the alignment of business goals.
Focusing on the policy implications, our historical example shows that the support provided by institutional actors to develop a coopetitive strategy might not be enough, and it must be supported by the commitment of each firm involved in the coopetitive interaction. Indeed, the competition among the parent institutions added to strained relations between Comitegit and Italegi due to widespread rumours of a government project to merge the two institutions to streamline the Italian banking presence abroad. The precarious financial condition of the two Egyptian subsidiaries cast doubts on the possibility of both returning to prosper in that market [39]. This project put them in competition to gain consensus before the Italian government and thus be the chosen institution to absorb the other. Therefore, the two banks set themselves conflicting objectives to seek a favourable position vis-à-vis the political authority (i.e. a negative-sum game). Not surprisingly, the most remarkable differences emerged when all the institutionsthe Italian parent banks and, indirectly, the Egyptian affiliateswent into trouble and came under public control. The nationalisation of the leading Italian banks following the Great Depression could theoretically have facilitated an agreement between the two Egyptian affiliates, but this did not happen. Instead, they continued to operate according to purely competitive logic aimed at capturing the public actor who should decide which of the two institutions would survive by absorbing the other. Therefore, the possibility of a cooperative behaviour on the Egyptian market was significantly reduced. Thus, even if the project ended in a stalemate, the survival of the two Italian banking rivalry banks became more and more a political issue, and no advantage could accrue from a cooperative approach.

Limitation and future research
This paper is subject to some limitations that allow suggestions for future research. In particular, the paper is based on a single case study from the banking industry. We recommend comparing strategies carried out by banks of other nationalities, highlighting differences and similarities. Furthermore, we recommend studying longitudinal cases from different industrial and geographical contexts. More generally, the analysis relates to a dyadic relationship; thus, future research could apply it to a network of participants in the coopetitive interaction. 2. However, according to Cherington (1976), Kirk S. Pickett, the owner of an oyster manufacturer, in 1911, described the relationships among his 35,000 oyster dealers by stating "You are only one of several dealers selling our oysters in your city. But you are not in competition with one another. You are co-operating with one another to develop more business for each of you. You are in coopetition, not in competition". This is probably the first time the term coopetition has been mentioned. Furthermore, combinations of competition and cooperation to create rents have always existed in history, e.g. trade guilds in the Middle Ages are ante-litteram examples of coopetition.
3. Other definitions of coopetition can be found for example in Kock (1999), Laine (2002) and Luo (2007). A very streamlined definition is provided by Eikerbokk and Olsen (2005), according to which coopetition is simply "simultaneous cooperation and competition".
4. An example of coopetition from the banking industry is the following. In 2019, the European Union sanctioned five global banks for collusion (Makortoff, 2019). In particular, they have been found guilty of collusive manipulation of the spot foreign exchange market for 11 currencies (including dollar). If we consider a geographical interpretation of the market, this is an example of inter-market coopetition: indeed, the banks cooperated in the European market, but not in the USA. However, if we consider a product interpretation of the market, this is an example of intramarket coopetition: indeed, the banks cooperated on the spot exchange market, but not on other products or services. Among other possible, even less recent examples, we could refer to the Inter-Alpha Banking Group, established in 1972, following an agreement reached by various European banks. The group intended to provide customers of each member bank with financial and operational assistance worldwide. However, each institution retained its own specific national characters and structural autonomy (Pino and Mignone, 2016).
5. The underlying assumption is that the interplay between competitive and cooperative stances could be analysed under different perspectives and, mutatis mutandis, adapted to different sectors. For example, Arranz Peña and Fern andez de Arroyabe (2002) carry out an empirical study focused on small and medium-sized Spanish firms to investigate the role of cooperation vs competition in the internalisation process of firms. Sarach (2015) empirically analyses the cooperating attitude of Russian firms operating in industrial clusters. The importance of cooperation for innovative industrial processes has been analysed by Stejskal et al. (2016), which focus on the Czech Republic's machinery industry. Belderbos et al. (2004) introduce the role of universities into the cooperative/competitive relationship between innovating firms, and find that cooperation with both universities and competitors push up the output levels. Lastly, Peterson and Behfar (2022) investigate the incentives to cooperate and/or compete with the colleagues, thus moving the analysis from the firm level to the individual level.
6. Of course, other characteristics of the competitive environment might be more/less conducive to a competitive setup rather than a cooperative one, as for example the existence or the absence of entry barriers. Dai (2008) claims that the reduction of the barriers to entry, by making competition fiercer, induces the firms to cooperate more as a reaction to greater competitive pressure.
7. The idea that more concentrated industries are more prone to cooperation can be found also in Dowling et al. (1996).
8. Government and authorities are sometimes indicated as the "architects" of the industry, as they shape the degree of competition/cooperation within the industry. See for example Jacobides (2007).
9. As emphasised by Luo (2007) when discussing international coopetition, weak interorganisational trust between two global players intensifies competition and delay cooperation, while a high level of trust allows them to create stronger ties.
10. Ahuja (2000) argues that the attractiveness of a firm for potential partnership depends on the value that it can provide to the partners in coopetition.
12. Less frequently, coopetition might emerge when two cooperating firms start to compete somewhere.
13. Quite connected to this point is the ability of high-level managers to communicate with middleand low-level employees. Indeed, as coopetition often requires a radical change in the intra-firm routines, a good top-down communication is needed to make such a change understandable and implementable (Lundgren-Henriksson and Kock, 2016).
14. On the Italian community in Egypt, see Rainero (1991). Its numerical consistency cannot be computed precisely (it was estimated at 23,000 in 1889), but it seemingly represented the second in size among the colonies of Europeans, second only to the Greek one.  29.4.1918, 8.5.1918, 17.6.1918, 15.7.1920, 26.11.1920Ibid, Busta 116, fasc. 3515. See Jones (1993), pp. 210-211;De Rosa (1983) 27. In July 1918, the president of the Jewish community of Alexandria in Egypt, the Italian Edgar Soarez, wrote to the Italian Minister of Foreign Affairs about the danger of British hegemony in the Middle East being established in the post-war period and the need therefore to give life to a "very strong banking institution", capable of effectively supporting Italian interests in that region. Soarez's proposal met with the favor of the Treasury Minister Nitti, who, however, drew attention to the advisability of a broader banking combination involving the main banks in the country, to nip in the bud any possible form of competition and to favour the highest national interest (ASU, Banco di Roma, Verbale CdA, 24 April, 1918). This wish could not be realised, most likely due to the existing conflicts between the Italian banks to be involved (De Rosa, 1983, pp. 198-202).
31. More and more frequent references to mutual control between the two banks with regard to the interest rates granted to customers, the opening of new branches or, more generally, the business model, emerged when the rumor of their possible merger spread since the mid-1930. See, for instance, ASI, Bci, AD 2, cart. 8, fasc. 9.
32. As early as 1934, Comitegit had tried to reach an agreement with Italegi to reduce the financing of cotton in a coordinated way. However, Italegi did not seem willing to take steps in this direction. ASU, Banco di Roma, Ufficio partecipazioni, Banco Italo-Egiziano, Prima sezione, fasc.