Sustainability Assessment of the Performance of Parmigiano Reggiano PDO Firms: A Comparative Analysis of Firms’ Legal Form and Altitude Range

Abstract

Geographical indications (GIs), protected by the European Union with the collective marks of PDO (protected designation of origin), PGI (protected geographical indication), and TSG (traditional specialty guaranteed), play an important role in the social and economic system. They not only guarantee food needs, but promote correct consumer information, protect local food, and play a role in the environmental and social sustainability of rural areas. In Italy, Parmigiano Reggiano (PR-RE) PDO cheese is ranked second in foods with the GI protection mark by turnover. This research aims to assess the financial sustainability of the firms registered in the PR-RE PDO consortium using financial statement (FINSTAT) analysis. Financial ratios (FR) and the EM-Score were applied to assess firms’ performance, financial risk, and credit score. The analysis distinguished firms by legal form, cooperative and non-cooperative, and altitude range—plain hill and mountain. The main findings of the research were as follows: (1) a better performance of lowland non-cooperative firms and lower financial risk, (2) a longer duration of the inventory cycle of cooperative firms, and (3) a greater financial risk in mountain cooperatives. The results provide indications for improving firms’ performance and for designing financial instruments for the sector. To our knowledge, this is the first research to carry out an analysis of all the available FINSTATs of firms in the PR-RE PDO sector.

1. Introduction

1.1. Research Premise

Agri-food production plays a central role in social, economic, and environmental sustainability; in fact, it not only guarantees the population’s food needs, but also has a significant impact on its health. Agri-food production plays a role in economic development, in terms of employment, investments, and innovation, and this also happens in disadvantaged areas.

Regarding this topic, in recent years, there has been a theoretical and operational interest in the application of sustainable practices in business management [1,2,3,4]. The objective is to increase environmental and social responsibility and, at the same time, increase the ability of firms to create value, not only for shareholders, but for all the stakeholders, internal and external, that are involved in firm processes [5]. Referring to the definition of the sustainable business model (SBM) proposed by Stubbs and Cocklin [6], these authors represented the SBM approach in corporate operational activity, composed of three components, namely, economic, environmental, and social sustainability. In the SBM context of agri-food production, the geographical indication (GI) of the European Union exists. GI foods are protected with the collective marks of PDO (protected designation of origin), PGI (protected geographical indication), and TSG (traditional specialty guaranteed). These food products play an important role in ensuring the sustainability of local rural areas, also playing an important role from a social and environmental point of view [7,8,9,10] in promoting the transition toward a circular economy model [11].

Numerous protection rules have followed one another over time, starting from the European Union (EU), which regulated the protection of GI products with several legislative interventions [12]. In fact, GI products have various positive effects: (a) on the market, on the demand side, GI products make it possible to reduce information asymmetry; in fact, consumers can purchase GI products with greater awareness, as they are traceable products and follow a mandatory production specification subject to a control system [13,14,15]; (b) GI products allow producers to use a collective mark that differentiates the product from competitors’ and links the product to the territory. This particularly favors smaller producers who are unable to develop a corporate brand recognized by consumers [16,17,18,19,20]. Regarding typical production, we observed that there are some potentially critical factors; in fact, compliance with production regulations can lead to the maintenance of inefficient production techniques or the failure to adopt product or process innovations. Furthermore, the adoption of a collective brand can disadvantage the adoption or establishment of corporate brands. GI products, therefore, play a role in the field of sustainability, both from a social point of view, for their positive impacts on the reduction in information asymmetry and on employment, and from an environmental point of view, for their positive impacts on the conservation of the rural environment [21,22,23]. Furthermore, the production of renewable energy deriving from production waste allows the transition to forms of production typical of the circular economy, which aim to promote animal welfare, healthy production, consumer health, and a reduction in polluting emissions [24,25,26,27,28].

In Italy, at the end of 2022, 853 PDO, PGI, and TSG products (3151 in Europe) were registered (of which 527 were wines), involving 195,407 operators organized in 296 consortia for the protection of typical products, with an overall value of the production stage of 20.2 billion euros. Italy is the leading European country in the number of PDO, PGI, and TSG foods, followed by France (713 productions) and Spain (357 productions) [29]. Parmigiano Reggiano PDO cheese (PR-RE PDO) is the second-ranked Italian cheese in terms of value at the production stage, with 1.72 billion euros, after Grana Padano PDO cheese, with 1.734 billion euros [29].

The PR-RE PDO production area includes the provinces of Parma, Reggio Emilia, Modena, and Bologna to the west of the Reno River and Mantua to the south of the Po River. The PR-RE PDO supply chain involves 3726 agricultural firms that supply bovine milk produced for processing in dairies, which operate as PR-RE PDO milk-processing firms [30]. In 2023, the cheese production was approximately 162 thousand tons of PR-RE PDO divided into 4,014,300 individual PR-RE forms, with an average weight of approximately 40 kg [30]. PR-RE PDO activates multiplicative effects in the local context, not only due to the significant turnover and direct employment, but also for related activities, such as the activities of technical services, construction, and veterinary medical services, and activities related to credit, the chamber of commerce system, and the PDO product certification system [31,32,33].

A recent study [34] highlighted several positive elements of the PR-RE PDO: (1) the product has a long tradition; in fact, the first product traces back to the 13th century. (2) The PR-RE PDO presents a premium price compared to its competitors in terms of consumers’ willingness to pay, even if this price differential is not as high as expected. (3) PR-RE PDO has high sustainability in terms of water consumption and a low carbon footprint; however, this advantage tends to decrease when the product is exported. Regarding this issue, it should be noted that the high production volume makes the PR-RE PDO a GI product with a high supply volume, for which consumption is necessary outside the production area, even in foreign countries, since local demand would not be able to sustain the consumption of all production. Furthermore, PR-RE PDO has long since achieved international fame among consumers, with an increased effect on the willingness to pay (WTP) among consumers [35] and it has received international awards [36]. On the topic of the production quality of PR-RE PDO, there have been several recent studies that have confirmed its role in the human diet as a functional food [37] and its appreciation by consumers [38], while other authors have highlighted that PR-RE PDO also has a positive effect on the development of incoming tourism, with the attraction of tourists for food and wine being linked to this product’s existence in the production area, with the use of augmented reality and artificial intelligence technologies also playing a role [39,40,41,42,43,44].

The PR-RE PDO sector has had to face difficulties due to market price fluctuations and competition from similar products, even those with names and brands that mislead consumers with incorrect information. Furthermore, the increase in production costs, including energy costs, and the increase in the costs of bovine feeding along the supply chain have led to several business crises. Starting from 2022, an increase in interest rates has been observed, which disadvantages capital-intensive firms, such as those of the PR-RE PDO. As a result, the number of dairies has decreased from 330 in 2018 to 292 in 2023 [30]. The PR-RE PDO production dairies have different legal forms, and this is possible because the legislation does not impose constraints on this aspect.

PR-RE PDO is regulated by production specification (PS), or Disciplinare di Produzione (in Italian) which is published on 13 April 2018 in the Official Journal of the European Union [45]. The PS establishes that PR-RE PDO is “a hard cheese made from raw cow’s milk… The milk must not undergo any heat treatment and has to come from cows fed primarily on fodder obtained in the area of origin. … The cheese must be matured for at least 12 months. … The minimum weight of each cheese is 30 kg. …… At least 75% of fodder dry matter must be produced within the geographical area. Feedstuffs may make up at most 50% of the dry matter of the feed ration. The use of silage of any kind is banned. … The milk is from cows reared in the defined geographical area”.

The PS imposes some constraints on the production of PR-RE PDO, which have an effect on the management: (1) PR-RE PDO production must be entirely carried out, in all its production phases including the production of milk for processing, in a limited territory in the northern part of Italy, defined as a production area by PS, as shown in Figure 1.

For PR-RE PDO cheese, PS therefore imposes [46] a production technique that affects production costs. Furthermore, as some authors have already highlighted in the case of Parma PDO Ham [47] and PR-RE PDO, PS has effects on the investments and financial structure of firms; the firms must mature PR-RE cheese according to PS, an activity that involves an investment in working capital and requires financial support. The topic of access to credit for agricultural businesses and agri-food processing has been the subject of several studies; a recent study highlighted the difficulty of accessing credit in particular for SMEs and for firms in the start-up phase [48]. A significant amount of research highlights the difficulty in the relationship between agri-food firms and banks, with reference to the issue of farmer training and information asymmetry in Italy [49] and in other European countries [50,51,52,53]. Other studies focus in particular on the management of working capital, of which particularly concerns agri-food firms, both in the maturation phase of agri-food products and in the management of the collection of trade receivables [54,55,56]. The debate on new forms of financing for agri-food firms, especially SMEs, has also been the subject of growing debate in recent years, in particular due to the increase in financing through fin-tech operations, crowdfunding, and private equity [57].

We can then consider some difficulties that PR-RE PDO firms find themselves facing, which motivate our study: (1) difficulties linked to production costs, due to an increase in energy and feeding costs, which have caused an increase in the cost of raw milk; (2) competition, including in the form of unfair competition, from competing cheeses and a reduction in the purchasing power of high levels of consumers, following inflationary phenomena and economic difficulties due the COVID-19 pandemic; (3) an increase in interest rates, which penalize investments both in fixed capital and in working capital in the cheese maturing cycle, with a consequent increase in barriers to the entry of capital into the sector.

In the PR-RE PDO sector, the majority of dairies take the form of cooperatives; in fact, in 2023, 174 of the 292 total dairies were cooperative [58]. The cooperative enterprises are owned by agricultural entrepreneurs who produce bovine milk which is given to the cooperative for subsequent transformation into PR-RE PDO cheese. Regarding this topic, the theory of cooperation, as illustrated by Zamagni [59], underlines how cooperatives can be a legal form through which to support the aggregation of small producers operating in different sectors, as in the case of PR-RE PDO, where there are high barriers to the entry of capital, or the activity is carried out in economically disadvantaged areas, such as mountainous areas. This aggregative phenomenon is particularly relevant in the Italian agri-food sector, where cooperation has historically been a determining factor for the survival and success of many businesses, especially in Southern Italy and in the rural regions of the North [60,61] and in several European countries [62,63].

Recent data from the European Union have highlighted that in the European Union there are approximately 250,000 cooperatives, owned by approximately 163 million citizens; in agriculture, the market share of cooperatives is 83% in the Netherlands, 79% in Finland, 55% in Italy, and 50% in France [64]; these data confirm the importance of a legal form analysis that considers cooperatives.

The effect of legal form on the performance of firms, on access to the capital market and, in a word, on economic and social sustainability in the short and long term has been, and still is, the subject of significant debate [65,66,67]. Research from various authors has focused on the comparison between legal forms of business activity, differentiating between cooperatives (COOPs) and non-cooperatives, i.e., investor-owned firms (IOFs), on the topic of access to the capital market and, consequently, on the constraints on growth and performance comparison between IOFs (limited company or joint stock company form) and COOPs [68,69,70,71]. Even in the Italian case, the distinction between cooperative firms and limited and joint stock firms is sanctioned by legal provisions, in this case, contained in the Italian civil code and in the special laws on cooperatives and joint-stock firms in general [72].

The role of the cooperative legal form is particularly significant in the food sector; the cooperative form in fact favors access to the market and the overcoming of size-based barriers to entry for investments, particularly in the case of small-sized operators [73].

For the legal form of cooperatives, some authors have coined the “user-owner principle” [74,75,76]; this approach expresses that COOPs are capitalized by those who use their services and not by external investors, who take on a passive role in management, such as banks, leasing firms or, in general, financial intermediaries who do not participate in management choices, such as, appoint managers. The members of the cooperative therefore take on a triple role: users, suppliers of capital, and recipients of residual profits. According to this principle, the members of the cooperative are mainly interested in the services offered rather than in the remuneration of the equity invested; in this way, the operational, investment, and financing strategies of the COOPs are not guided by the creation of value, as happens in the IOFs [77,78,79,80].

The cooperative legal form helps to better manage information asymmetries between operators along the supply chain and therefore power imbalances in buyer–seller negotiations [81]; for example, agricultural cooperatives allow individual producers to reach higher price levels than those obtainable by selling individually on the market [82]. The benefits that cooperative members can derive from this legal form are broader, since cooperatives can also improve the quality of production by purchasing specialized consultancy and training services for their members on the market, in addition to hiring highly qualified personnel [83]. Even the PR-RE PDO sector, it is possible that IOFs can solicit investors without the obligation of purchasing products or delivering raw materials; IOFs are not obliged to self-finance capital through the accumulation of profits, but can request capital from the market [84]. COOPs have less incentive to reinvest profits, because this can give the COOP an advantage by reducing debt, but this does not necessarily favor or satisfy members; it follows that COOPs are financially constrained and are incentivized to be more indebted than IOFs.

Given the theoretical premises, it is relevant to analyze whether there are differences between COOPs and IOFs to test in the PR-RE PDO sector and what previous authors have inferred and assumed regarding the differences between cooperatives and IOFs. Therefore, the research has two main research questions (RQs):

-

(RQ1) By applying financial ratios (FRs) and credit scoring (CS) to dairies operating in the PR-RE DOP sector, are the performances of cooperative dairies (COOPs) and investor-owned firms (IOFs) statistically different? The first RQ takes into consideration the legal form of carrying out the firms’ activity; therefore, considering two groups of firms (COOPs and IOFs);

The second RQ replicates the analysis of RQ1 by providing the further segmentation of firms, not only on the basis of their legal form but also on the basis of the altitude range in which the dairy activity is carried out; thus

-

(RQ2) By applying financial ratios (FRs) and credit scoring (CS) to dairies operating in the PR-RE PDO sector, are the performances of cooperative dairies (COOPs) and investor-owned firms (IOFs) statistically different taking into account, in addition to legal form, the altitude ranges in which the firms’ activity are carried out (plain or hills and mountains)? The second RQ takes into consideration four groups of firms, dividing the data in groups according to their legal form (COOPs and IOFs) and altitude range (plain or hills and mountains).

RQs may have general interest and applied utility for some reasons: (1) PR-RE PDO is one of the most important cheeses at the national and European level, and the research aims to investigate the firms’ performance in the sector and their capacity to access credit ensuring financial sustainability; (2) the RQs have the objective of differentiating firms’ performances by legal form and altitude and this reflection is part of a broad literary debate with the aim of comparing the performances, differentiating the legal forms, and differentiating the altitude at which the activity takes place.

1.2. Theoretical Background

Cooperative enterprises are essentially of two types, namely worker cooperatives and contribution cooperatives [85]. In worker cooperatives, the cooperative members carry out their work in the cooperative and receive compensation that is determined by taking into account the profit of the cooperative; very often a minimum wage is paid throughout the year. In contribution cooperatives, individual members each own an agricultural firm that produces agricultural products, and they contribute these products to the cooperative. The cooperative then carries out the transformation and marketing phase of the agricultural raw materials supplied by the members. The cooperative is financed with the capital of the farmer members, who must therefore find join together at their convenience for the transformation phase by establishing the cooperative, and maintaining this investment over time [86,87,88]. The cooperative members then share the profits from the transformation, which varies according to the efficiency of the cooperative. If the value of the remuneration of the contributions is higher than the market price at which the member could sell the product, the member has the advantage of conferring the agricultural production to the cooperative; otherwise, there is no convenience [89].

Dairy farmers, given the specificity of the investments necessary for milk processing, find themselves in a weak contractual position, and the cooperative phenomenon can be analyzed by applying the theory of transaction costs. The Coase’s theory of transaction costs [90] starts from the observation that the existence of the firm is explained by the fact that it is more expensive to purchase a service on the market rather than integrate the production process in an organization with joint administration, which is precisely the firm. The firm is therefore created with the aim of vertically or horizontally internalizing some production phases, which are less expensive to bring within the organization rather than purchasing on the market. Williamson’s [91] seminal paper indicates that in cases of the specificity of the asset, and uncertainty and frequency of the relationship, firms consider the process of integrating the production phase (making the case) preferable to purchasing it on the market (buying the case); this integration strategy allows one to reduce the transaction costs associated with the exchange. The case that our research engages with is that of the sector of the transformation of milk into PR-RE PDO cheese, in which there are many cooperative enterprises that belong to the second type, i.e., cooperatives of agricultural entrepreneurs associated in a cooperative for the transformation of bovine milk. The cooperative members are remunerated by paying for the milk delivered at a price at least equal to the market price of the milk; this is in fact possible when the cooperative has lower production costs than competing processing firms, as it is exposed by the theory of transaction costs.

Saatz [92] observed that the phenomenon of downstream vertical integration, with the formation of DCs, applies to transaction cost theory; in fact, the costs associated with the formation of a new cooperative entity are lower, in terms of transaction costs, than purchasing a milk processing service on the market [93]. This approach also applies to the case of milk transformation into PR-RE PDO through the use of milk processing cooperatives in the production area defined by the specification. In fact, the formation of DCs allows for the aggregation of smaller agricultural firms, helping to overcome the dimensional capital barrier of the initial investment.

On the topic of cooperative governance, recently, some authors [94] have theorized the primacy of some cooperative stakeholders over others, and the risk of the primacy of the management board and some members over others; this proposes the application of the principal agent approach to the case of cooperation, and the presence of conflicts and negotiations between the stakeholders of the cooperative. Other authors have therefore highlighted that cooperative members hold property rights and are also users of the cooperative’s services [95]; the members of the cooperative have the rights to participate in the assembly, divide profits, and appoint managers; this principle occurs according to the democratic criterion according to which each shareholder is given a vote, regardless of the share of capital he or she owns. There is therefore a disincentive for investing in cooperative shares by non-member investors, because (a) they are not interested in the services provided by the cooperative, but only in the return that the participation share can offer; (b) governance is based on the cooperative criterion and not on the capitalist criterion of the majority. Precisely in the case of milk-processing cooperatives (dairy cooperatives, or briefly, DCs) some authors have demonstrated [96] that the attractiveness of capital from external investors is limited because investments do not have the objective of efficiency, but rather the farmers who are members of the cooperative have the objective of controlling the DCs’ management and consequently maximizing the profit of the firms supplying the milk rather than maximizing the profit of the transformation that takes place in the cooperative. This benefits the farms that deliver the milk and discourages DC investors.

The seminal research of Fulton [97] highlighted the difficulty of COOPs in collecting equity for two main reasons, due to their legal form: (1) COOPs only collect equity from members, as they are the firms that have the characteristic of being user-owned; (2) COOPs are obliged to reimburse members’ equity when requested and, therefore, long-term investments are more difficult because stable sources of financing are fewer than IOFs.

In recent years, a growing debate [98,99,100] has developed on the role of cooperation in the development of local economic systems and in the aggregation of farmers, after the seminal works of Cook [101] in which five main problems are listed in the cooperative: freeriding, horizon, portfolio, control, and influence cost. As FAO research has highlighted [102], agricultural cooperatives play an important role in rural areas in promoting, in particular, for producers (1) access to natural resources; (2) the circulation of information and knowledge; (3) access to technical means; in particular, those for which there are barriers to entry due to the purchase cost; and (4) access to the market and the definition of commercial strategies.

The PR-RE PDO sector presents a case of sunk costs, analyzed by Tirole in the context of transaction cost theory [103]; in fact, the degree of specificity of the investments necessary for dairy farming is high. Therefore, it is difficult to transfer investments to other firms given the asset specificity in dairy farming and the potentially higher sunk costs of investment failure [104]. Investments in the sector are largely irrecoverable, as the equipment for the production of PR-RE PDO is largely specific, and only in very limited parts can it be used for production differentiation or be reused in the case of the reconversion of industrial activity. This is, in particular, the case for the boilers where PR-RE PDO is processed, the sizing of which is functional to the production of a wheel of cheese. Even the product-maturing structures, organized in vertical shelving, and assisted by robots for moving the wheels, represent a sunk cost case.

Recent research on the topic of cooperative financial constraints addressed the comparison between cooperatives (COOPs) and investor-owned enterprises (IOFs) in accessing the capital market, with research pertaining to the agricultural reference sector in the USA [105]; this research had interesting results, as no difficulties emerged in accessing credit in the short term by COOPs compared to IOFs; however, the authors found that there is a capital constraint in COOPs to finance long-term investments. Although short-term financing allows firms to support the payment of current debts and achieve financial sustainability in the short term, long-term investments are essential for the remuneration of share capital, for attracting stable investments and, essentially, for a firm, which promotes sustainable management over time and is capable of investing in the sunk costs of specific assets, as in the case of DCs.

Other authors [106] have recently applied the stakeholder theory approach to cooperation, highlighting that it is a different legal form compared to private owned firms because participation in the shareholders’ meeting takes place on a democratic basis, as each member holds only one vote, independent from their share of capital held, and the objective of the cooperative is to provide services to its members. The application of the stakeholder theory to the case of firms considers the involvement in the firm’s objectives of external stakeholders, not limited only to the members or managers of the cooperative, but also to subjects in society, such as suppliers, customers, employees, banks, the State and, in general, the economic and social environment in which the cooperative operates [107,108,109]. This relationship favors innovation, the growth of intellectual capital and the durability of the cooperative; there is thus evidence of the usefulness of applying stakeholder theory to cooperation to understand the transition to managerial approaches based on sustainability [110,111,112].

The legal structure of cooperatives, and the consequent contractual form that binds cooperative members to each other, the mutual obligations between members and cooperatives, and the mechanism of voting and the election of managers, have been highlighted by some authors [113,114] who showed how the cooperative’s transaction costs increase through internalization. This also affects DCs where the returns of some specific tangible assets, investments in RD, or simply investments to increase efficiency, may cause discouragement because they exceed the expected duration of farmers’ membership in the DC; this discourages investment in long-term investments in the cooperative and farmers therefore push managers to choose a less risky investment portfolio [98,115,116,117]; this problem increases when the composition of the cooperative is predominantly elderly members [118]; this case also concerns the DCs who work with PR-RE PDO, considering the aging of the farmers also in the production areas. From this point of view, the illiquidity of the circulation of cooperative shares discourages investments made with equity if some of the cooperative’s members are close to retiring; this creates a lower incentive to finance long-term investments, both with equity and debt, because these benefits effect those who will use the cooperative in the future, as these benefits cannot be monetized by current members through the negotiation of the cooperative’s shares [119,120,121,122].

We can observe that some large cooperatives raise, even in Italy, capital from private investors, in the form of loans and, therefore, there is, to a certain extent, a return on capital within the cooperative world in favor of the interested parties and, to a marginal extent, to the services offered by the cooperative (the case of social loans is emblematic of this, regulated in Italy by Law 17 February 1971, n. 127). Furthermore, in the PR-RE DOP sector, the first mini-bond issue was carried out by a cooperative firm, with subscription by subjects including non-members [123].

Furthermore, we note that Italian legislation does not allow for the distribution of assets in the cooperative among the members at the time of liquidation; this represents a further disincentive to investment by private investors who are not members. As a result, COOPs, despite being so important for the social and economic development of rural areas, face the difficulty of raising capital in the market to finance investments, particularly investments in highly sector-specific assets, and suffer from difficulties in their relationship with the capital market, and several studies have highlighted how cooperatives are characterized by undercapitalization and have difficulty accessing credit [124,125,126,127].

Some authors have pointed out that COOPs have less incentive to reinvest profits, because this can give the COOPs an advantage by reducing debt, but this does not necessarily favor or satisfy members [128]; it follows that COOPs are financially constrained and are incentivized to be more indebted than IOFs. According to this approach, several studies have highlighted that the accumulation of equity in COOPs is hindered because the equity, in a traditional cooperative, is non-marketable and non-transferable and its value is not modified by changes in value on an exchange market [129,130,131,132,133]. The same author [129] also highlighted that there is, for cooperative members, a competition between investments in the cooperative and investments in their own business; consequently, the opportunity cost of investing in the cooperative may increase, due to the member’s budget constraint, and consequently, investments in the cooperative may decrease and there may be opportunistic behavior, typical of free riding.

Even if susceptible to opportunistic behavior, a large part of the literature agrees in stating that cooperation favors the sustainability of the growth of agricultural firms, in smaller sized parcels and in the start-up phase [134,135,136,137]. Given the consensus present in the literature on the importance of the cooperative legal form, the research analyzes the performances of COOPs compared to IOFs in the PR-RE DOP sector to test any differences, as highlighted by the financial ratios and credit scoring illustrated and detailed in Materials and Methods.

2. Materials and Methods

2.1. Financial Ratio Analysis (FRs)

We can observe that the Italian civil code, in Article 2423 and those that follow, illustrate the need to draw up financial statements (FINSTATs) for joint-stock firms; the national legislation for Italy applies the Legislative Decree of 139 18 August 2015, which transposes the European Directive 2013/34/EU into national law [138]; the firms obliged to prepare the FINSTATs are cooperatives, limited liability firms, and joint-stock firms. The obligation to draw up the FINSTATs has the aim of providing information to all the stakeholders, internal and external, on the progress of the firm activity [139,140]; this legal provision is mandatory for all the cases in which there is perfect patrimonial separation between the assets of the firm and the assets of the shareholders and, therefore, in the event of bankruptcy, only the firm, which has the assets is liable for debt, and not the shareholders. The obligation to inform stakeholders therefore concerns all firms, including agricultural firms, if the business activity is carried out in the form of a joint-stock company. In Italian legislation, the definition of agricultural enterprise contained in Article 2135 of the Italian Civil Code refers to the particular connection of agricultural enterprises’ activity to the land and to the transformation, conservation, and marketing activity that the agricultural entrepreneur carries out on the goods produced in the firm, in addition to any related activities. Sole proprietorships or partnerships are not required to prepare FINSTATs. This definition is important because it defines the parameters of agricultural activity, as an activity not only of production, but also of the transformation and marketing of the firm’s production, as well as the carrying out of a series of services, including those of tourism and hospitality conducted on the farm. The Italian legislator provides forms of relief for agricultural firms, and in fact, Italian regulations exempt agricultural firms from bankruptcy [141]. Furthermore, income taxation does not occur on the basis of the profit calculated in firm FINSTATs, is increased by non-deductible costs, and is reduced by non-taxable revenues according to tax legislation, as happens for non-agricultural joint-stock firms [142]. Italian tax law states that agricultural firms are subject to taxation on a cadastral basis [143]. The reason for these favorable regulatory provisions regarding bankruptcy and taxation are due to (1) the importance of the agricultural sector for the nation in ensuring food self-sufficiency; (2) the presence of additional climactic and biological risks of which agricultural entrepreneurs are subject to and of which non-agricultural (commercial) entrepreneurs are not, in addition to the traditional business risks; agricultural entrepreneurs are also subject to the meteorological and biological risks that are typical of agricultural enterprise and, therefore, other things of equal conditions are more subject to uncontrollable risks and therefore require, under the Legislator, a particular form of protection given by their exemption from bankruptcy proceedings.

The application of FRs allows the performance of firms to be evaluated over time, considering, not the pure numbers expressed in money, but the relationships between FINSTATs’ values; FRs use data from the firm’s FINSTATs; it is therefore a public database, which joint-stock firms must draw up and make public as required by law [144,145,146]. FRs use FINSTATs as databases and are therefore influenced by accounting standards [147]. The application of FRs has also been widely used over time to analyze agri-food firms and dairy firms [148,149,150,151]. The application of FRs began in the financial offices of the multinational firm DuPont [152]; the index system developed by DuPont, used as a reporting system to evaluate the performance of managers and the return on capital, is still widely applied today, both in professional activity [153,154,155] and in academic research [156,157,158,159,160,161].

The analysis of the investments and sources of financing uses the balance sheet format; in any given fiscal year, investments are given by FA (fixed assets), WCiⁱ (working capital, assets, inventories of stock), WCarⁱ (working capital, assets, account receivables), WCoaⁱ (working capital, assets, other assets), L (liquidity); the sources of capital, to ensure investments’ coverage, are given by E^sc (share capital), E^r (reserves), Π^pT (retained earnings), WCap^s (working capital, liabilities, account payables), WCol^s (working capital, liabilities, other liabilities), FD^M<12 (financial debt due within 12 months), and FD^M>12 (financial debts due after 12 months). We can summarize the sources of capital as follows: equity capital (E^T = E^sc + E^r + Π^pT) and debt capital (D^T = WCap^s + WCol^s + FD^M<12 + FD^M>12) where WCap^s + WCol^s + FD^M<12 = D^M<12 is short term debts. We can reclassify the balance sheet with the expression of the net financial position (NFP):

FD^M<12 + FD^M>12 − L = NFP

(1)

We then consider the net investment in working capital (NWC) that expresses the absorption of financial resources as a result of the acquisition cycle, processing, and sale that is expressed as follows [162]:

(WCiⁱ + WCarⁱ + WCoaⁱ) − (WCap^s + WCol^s) = WC^iT − WC^sT = NWC

(2)

In Equation (2), in any given fiscal year, WC^iT is an investment of capital, WC^sT is a source of capital; NWC quantifies the net resources generated by (NWC < 0) or absorbed (NWC > 0) from working capital management and it is therefore useful to express such reclassification in the balance sheet with the functional form [162]:

FA + NWC = E^T + NFP

(3)

Equation (3) directly expresses NWC and highlights the second member, expressing funding sources, with NFP and E^T sources of capital. We can have three other cases of Equation (3): (3a) FA + NWC + NFP = E^T; (3b) FA + NFP = E^T + NWC; and (3c) FA = E^T + NWC + NFP. In (3a) FA, NWC and NFP are all assets and E^T is the unique source of capital; in (3b), FA and NFP are assets where E^T and NWC are sources of capital; and, at the end, (3c) has FA as the only asset, while E^T + NWC + NFP are all sources of capital. Some authors have pointed out that the case in which NWC > 0 is called the conservative management of working capital [163,164]. The case of NWC < 0 is called the aggressive management of working capital [165,166] and is considered to be directly related to the risk of financial distress. In order to quantify the duration of the NWC financial cycle, five main financial ratios for the NWC duration are frequently applied in the agri-food firms’ FRs analysis [167,168,169,170,171]. The duration in days of NWC is expressed applying five main financial ratios, respectively, INV_DAYS, AR_DAYS, OA_DAYS (in the asset side of the NWC), and AP_DAYS, (in the source side of the NWC). Respectively, we can express this as follows:

INV_DAYS = WCiⁱ × 365/S

(4)

AR_DAYS = WCarⁱ × 365/S

(5)

OA_DAYS = WCoaⁱ × 365/S

(6)

AP_DAYS = WCap^s × 365/S

(7)

OL_DAYS = WCol^s × 365/S

(8)

In the above Equations (4)–(8), S are sales in the fiscal year, INV_DAYS expresses the length of inventory rotation in days, AR_DAYS expresses the length of the payment deferral given for accounts receivable in days, OA_DAYS expresses the length of the payment deferral given by working capital and other assets, represented in days, all in the asset-side of the balance sheet. AP_DAYS expresses the length of the payment deferral given for accounts as payable in days and OL_DAYS expresses the length of the payment deferral given working capital and other liabilities in days, all in the source side of the balance sheet. To measure the duration of working capital in days, many researchers apply measures based on the duration of NWC in days, as follows:

NWC_DAYS = INV_DAYS + AR_DAYS + OA_DAYS − AP_DAYS − OL_DAYS

(9)

Several authors define NWC_DAYS as the Cash Conversion Cycle (CCC), expressed in days [172,173,174]. Applying FRs with the aim of evaluating the performance of firms, we can express the overall performance of the firm as follows:

$$\mathrm{ROE}=\left[\mathrm{ROA}+(\mathrm{ROA}-\mathrm{ROD}){\displaystyle \frac{\mathrm{NFP}}{\mathrm{E}}}\right]\left(1-{\mathrm{T}}_{\mathrm{m}}\right)$$

(10)

Equation (10) is the well-known approach from the seminal work of Modigliani and Miller and its long-lasting subsequent debate [175,176,177,178,179,180], even in relation to the topic of the capital structure for the agri-food firms [181,182,183,184,185]. ROE is applied to calculate shareholder value and can be expressed as the ratio of net income (П) to equity (E), as follows: ROE = П:E. Another widely applied FR is the return on assets (ROA), often applied to compare the operating income to the total capital invested in the business; ROA is the ratio of earnings before interest and taxes (EBIT) to total assets (TA) expressed as TA = FA + WCiⁱ + WCarⁱ + WCoⁱ + L, ROA = EBIT:TA; ROA expresses the annual percentage return of each unit of capital invested in a firm, before the cost of debt and tax payment. In (10), ROD expresses the cost of debt (precisely, the return on debt), which is calculated as follows: ROD = SF: PFN, where SF is the result of financial operations, therefore, expressing the cost of debt; ROD expresses the cost of net financial debt (NFP) as a percentage in a given fiscal year; the relation between ROA and ROD is applied to test financial leverage (ROA > ROD). The turnover ratios of the revenues are used for the multiplicative decomposition of the ratios of profitability; the ROA is decomposed, applying a multiplicative equation [162], as follows:

$$\mathrm{ROA}={\displaystyle \frac{\mathrm{EBIT}}{\mathrm{S}}}\cdot {\displaystyle \frac{\mathrm{S}}{\mathrm{TA}}}=\mathrm{ROS}\cdot \mathrm{T}$$

(11)

In Equation (11), the return on invested capital (ROA) is broken down through a multiplicative relationship between the return on sales (ROS) and the turnover (T). The ROS expresses the operating profitability of sales, while T expresses the capital turnover in the year. Firms in which a margin strategy prevails over sales will have a prevalence of the ROS factor, while firms that favor a capital rotation strategy will have a prevalence of the factor T. The multiplicative decomposition of ROA shows how firms can generate revenue by using two strategies. In the first strategy, the firms operate with high-volume production in relation to the capital invested, then with high turnover. These are firms that can operate with low margins on sales (ROS) per unit compared to the high turnover of capital; in this case we can consider, for example, the large distribution firms that operate with high volumes of sales and, at the same time, can reduce the margin unit on sales applying a commercial strategy; contrarily, there are firms with low turnover that need to have a high level of ROS to achieve a satisfactory return on their invested capital.

FRs, by their nature, however, present some characteristics that limit their application [186,187]:

• The consideration of economic and financial values is affected by the legal provisions on accounting matters: accounting principles determine the potential underestimation of income, at least in the short term, and does not allow for evidence of latent capital gains on fixed assets, as well as the accounting of values of intangible assets;
• The moment of manifestation of the financial flows is not considered, resulting in situations in which a valuation with an economic approach, despite showing a rate of return considered adequate by the shareholders, suffers from a lack of liquidity and the impossibility of distributing dividends, even with positive profit;
• The values of the flows in the numerator are related to the values of the shares in the denominator for all economic indices; the numerator considers the flow values, formed during the reference financial year, from the beginning to the end of it, as an algebraic sum of the positive and negative components of the income, while the denominator considers securities values that have instant quantification; the ratios have the greatest distortion in the case of evaluations linked to highly seasonal activities, where the quantification of the capital stock at the end of the period is not highly expressive of the average equity capital (or debt).

2.2. Credit Scoring Analysis

As shown by a significant amount of research, FRs reduce information asymmetry in markets, improving their efficiency [188,189,190]. FRs are also applied by banks in assessing creditworthiness, also with reference to the Basel interbank stability agreements [191,192,193]. The application of FRs allows for comparisons between different firms, as a various forms of research have demonstrated for the agri-food system [194,195,196,197,198,199,200,201].

In recent years, after the subprime mortgage crisis in particular, there has been a radical change in the way creditworthiness is assessed. In fact, banks’ analysis of creditworthiness moved from an approach traditionally based on real asset guarantees, as collateral for the granting of credit, to evaluations based on credit scoring and the quantification of cash flows to support debt service [202]. In this context of change, smaller firms have not always been aware of the change in perspective in assessing creditworthiness, as it is needed to access financial markets [203].

FRs are widely applied in the construction of credit scoring systems; the first works in this area date back to Beaver’s seminal works [204], Altman [205], and Ohlson [206]; these authors applied the FRs with multivariate discriminant analysis (MDA) techniques, dividing insolvent firms from solvent ones and analyzing the values that the FRs have in the two groups of firms, with the aim of predicting the probability of default. The applications of MDA also involved agricultural firms in that period, with satisfactory results [207]. Subsequent studies have applied logit and probit models to credit scoring [208,209,210]. More recently, with the increase in computing power, more powerful techniques derived from the application of neural network analysis have been applied that do not require the definition of a specific probability of default; these techniques apply big data and artificial intelligence techniques [211,212,213,214]. The research of these authors confirms the usefulness of the application of FRs in credit scoring systems. Credit scoring estimates the risk in lending operations, with the aim of: (1) reducing the information asymmetry between firms and banks, to avoid moral hazard and help financial operators to refuse businesses that are deemed to risky from the taking out a loan; (2) correctly define, in a risk–return relationship, the correct pricing of financial operations, also considering that in the event of an increase in debt risk, financiers must cover their risk by increasing the pricing of financial operations. Credit scoring favors the functioning of the capital market, as it reduces the information asymmetry between the borrowers and lenders of funds [215,216]. Riskier firms obtain loans at a higher price, or are not financed; on the other hand, less risky firms obtain financing at lower prices and in greater quantities. The precision of credit scoring systems makes it possible to reduce errors of the first type (the solvent firm is classified as an insolvent, and credit is refused) and of the second type (the insolvent firm is classified as a solvent, and credit is granted to the firm, as it is destined for default).

A widely applied credit scoring ratio is the Altman Z-Score model [217]. The Z-Score has been variously and subsequently modified over time by Altman himself [218] and by other scholars and professionals [219,220,221,222,223,224,225,226,227,228,229] for application to unlisted firms. The revision of the model gave rise to the credit scoring system called EM-Score, formulated as follows:

EM-Score = 6.56 × (X₁) + 3.26 × (X₂) + 6.72 × (X₃) + 1.05 × (X₄) + 3.25

(12)

In Equation (12), X₁ is the ratio between NWC and the total asset (TA), expressed as follows: TA = FA + WCiⁱ + WCarⁱ + WCoⁱ + L. X₂ is the ratio between E^T and TA, X₃ is the ratio between earnings before interest and tax (EBIT) and TA, and X₄ is the ratio between E^T and total debt (D^T). The EM-Score is applied with an equivalence scale to evaluate the debt risk that a given investor faces, as in the case of a financial intermediary. An EM-Score index is applied to assess the financial sustainability of the management cycle, using decreasing risk classes (

Table 1. EM-Score scoring table, elaboration by the authors from [218].

EM-Score	Rating
EM ≥ 8.15	AAA
7.60 ≤ EM < 8.15	AA+
7.30 ≤ EM < 7.60	AA
7.00 ≤ EM < 7.30	AA−
6.85 ≤ EM < 7.00	A+
6.65 ≤ EM < 6.85	A
6.40 ≤ EM < 6.65	A−
6.25 ≤ EM < 6.40	BBB+
5.85 ≤ EM < 6.25	BBB
5.65 ≤ EM < 5.85	BBB−
5.25 ≤ EM < 5.65	BB+
4.95 ≤ EM < 5.25	BB
4.75 ≤ EM < 4.95	BB−
4.50 ≤ EM < 4.75	B+
4.15 ≤ EM < 4.50	B
3.75 ≤ EM < 4.15	B−
3.20 ≤ EM < 3.75	CCC+
2.50 ≤ EM < 3.20	CCC
1.75 ≤ EM < 2.50	CCC−
EM < 1.75	D

), from D (default) to AAA+ (lower risk of insolvency) which therefore express progressively better credit scoring. EM-Score values above 3.75 are “investment grade” ratings (B- and higher rating classes); EM-Score values equal to or lower than 3.75 are “non-investment grade” ratings (CCC+ and lower rating classes); and EM-Score values less than or equal to 1.75 are at high risk of default (class D of the equivalence scale). The EM-Score has some advantages and for this reason we apply it in our research: (1) the EM-Score is a widely known test, used in research and business practice; (2) the EM-Score has been extensively tested and considered reliable, in particular for the risk analysis of SMEs, even those not listed on the financial markets, which is the main case of our research.

FRs and the EM-Score are applied in this research with the aim of analyzing performance and access to credit, comparing the data of the agri-food firms included in the groups, to verify if there are significant differences.

3. Results and Discussion

3.1. Research Plan

This research performs an analysis of all the available FINSTATs of the firms registered in the PR-RE PDO cheese consortium. To our knowledge, this is the first piece if research that has considered this database and carried out the complete analysis of the FINSTATs of these firms. To answer the RQs, the following research plan is prepared and carried out:

• There are 292 firms operating in the PR-RE PDO cheese sector, registered in the PR-RE PDO Cheese Consortium [58], which make publicly available the names of the firms, the registered office address, and the VAT number of each firm. In Italy, only firms (cooperatives, limited firms, and joint stock firms, respectively, known as the società a responsabilità limitata e società per azioni, in Italian) are obliged to deposit FINSTATs’ data in the Companies Register; sole partnerships (imprese individuali, in Italian) and proprietorships (società semplici, società in nome collettivo or società in accomandita semplice, in Italian) do not present their FINSTATs. The research revealed that 75 firms are sole proprietorships or partnerships; the FINSTATs data of these firms are not mandatorily filled in the Company Register and are therefore not available to the public, so it is not possible to insert them into the search database;
• Of the 217 firms for which FINSTAT is available, the data are extracted from the AIDA database which collects public information and allows faster processing by extracting data from groups of firms and for different years. The same data can be freely consulted upon request at the Company Registry Office at the Chambers of Commerce. The data available in this case, however, concerns each individual firm for each individual year; this data extraction method requires a considerable data entry effort; furthermore, each data request requires a fee. The AIDA database is made available free of charge for authors by the University of Parma; in the data set extraction, three firms have been eliminated from the database, because their FINSTATs do not provide the minimum information necessary to carry out the requested analysis and the data are evidently tainted by errors. In particular, two of the three canceled firms present different years of financial statements in which the assets are different from the liabilities, and this inconsistency indicates that the data are not reliable, while one of the three canceled firms presents deficiencies in different financial statements in the mandatory details required by the Italian Civil Code. Therefore, the data of these three firms cannot be used in the research and has therefore been deleted. The elimination of the three firms from the database has no effect on the final results because (a) the number of deleted firms is very small, i.e., 3 out of a total of 217; (b) it is not possible to construct a robustness test considering the data of the three canceled firms because deleted firms present incomplete data and, therefore, it is not possible to expose a complete database with the data of all 217 firms. Data extraction covered a 10-year series, which is the longest series available in the database; the series covered the years from 2013 to 2022 and it is therefore affected, for the 2020–2022 three-year period, by the effects of the COVID-19 pandemic; the use of a ten-year series, the longest available in the AIDA database, has the aim of broadening the breadth of the data analyzed as much as possible to capture long-term trends in the sector and reduce the effect of market fluctuations in the short term. Some firms do not have a ten-year series available, so the database observations are not 2140, which would be equivalent to 214 firms in 10 years, but only 2062. The AIDA database is widely used in research activities in Italy for the analysis of the survival of firms [230], the analysis of the performance of manufacturing sectors [231], and the analysis of the startup phase of firms [232]. Furthermore, the AIDA database also finds wide application in research in the field of performance the analysis of agri-food firms [233,234];
• For this research, two groups of firms were first created, the first composed of non-cooperative firms (limited firms and joint-stock firms, i.e., the IOFs (G1_IOFs), and the second composed of cooperative firms (G2_COOPs). After this first subdivision, the two groups were further divided with the criterion of the altitude range; IOFs were divided between IOFs in plains and hills (IOFsPH) and mountain IOFs (IOFsM), and applying the same criterion, we have plain and hill cooperatives (COOPsPH) and mountain cooperatives (COOPsM); we therefore have one sample, divided into two groups, further divided into four subgroups. The attribution of the altitudinal bands of the individual firms was carried out by taking into account the attribution of the Italian Institute of Statistics (ISTAT) to the plain, hill, or mountain band belonging to the municipality in which the firm is based. These data were exposed in the AIDA database used for data extraction. It must be considered that it is (1) possible that some firms have their operational headquarters in a municipality with a different altitude than that of their registered office; (2) some firms may have multiple factories in which the operational activity is carried out, even in municipalities with different altitudes. To solve these potential flaws, an additional investigation was carried out using the data contained in the AIDA database and, in case of doubt, the optical FINSTAT of the firm was viewed and the Chamber of Commerce certificate was extracted from the Company Register. The analysis of these documents allowed the previous points (1) and (2) to be resolved, and one cooperative, attributed in a prior phas, to the plain and hill, is now correctly attributed, because the firm formal office is in the plain area, while the operating activity is carried out in mountainous area;
• All the data used for this research is therefore public, and the research is replicable. Data extraction can occur online in an electronic format or in a spreadsheet database. Data analysis was performed with IBM™ SPSS Statistics version 29.

The research plan must consider that (a) the sample of firms considered the firms currently registered with the PR-RE PDO Cheese Consortium; firms that are canceled were not considered, either due to voluntary resignations or because they ceased due to liquidation or bankruptcy; (b) the FINSTAT data included in the research have not been revalued for inflation. The values indicated are in euros, which is the currency in which FINSTATs are expressed; (c) data extraction is not able to distinguish whether firms carry out activities other than the sole production of PR-RE PDO; considering the regulatory constraints, the dairy-processing cooperatives included in the sample mainly carry out the transformation of PR-RE PDO; (d) the classification of businesses cannot divide businesses into managerial businesses or family businesses, nor be vertically or horizontally integrated, nor belong to the business groups of firms. Some limitations need to be considered when applying the financial statements extracted from the AIDA database: (1) the financial statement data are affected by the application accounting principles and the rules of the Civil Code regarding bookkeeping; therefore, the results of the calculations are affected, in particular, by the application of the principles of accounting conservatism and accrual accounting contained in Article 2323-bis of the Italian Civil Code [235,236]; (2) the financial statements contained in the database and used in the research are not subject to verification for accounting errors or fraud, because this verification is not possible without the investigation and acquisition of documents reserved for the judicial authority [237,238,239,240,241]. Given the duration of the 10-year series used in this research, it can be assumed that the firms present in this database, even if they present accounting errors in their financial statements, do not present errors of a nature that would concern the firms.

3.2. Descriptive Statistics of FRs and EM-Score Values

The database data are processed in the following way to answer the RQs: (1) first, a descriptive statistical approach was applied on the FINSTAT values of the firms and the related FRs; (2) other FRs are calculated, applying a multiplicative decomposition approach, to which descriptive statistics were applied; (3) credit scoring and financial sustainability calculation with an EM-Score calculation for all database observations were performed; and (4) an analysis of the significant differences between the mean/median values was performed. The sample showed a divergent distribution compared to the normal distribution; therefore, a non-parametric approach was applied.

Research confirms that the sector is characterized by the presence of dairies that operate in the legal form of cooperatives; the sector includes 292 dairies and, of these, the available data of 214 firms were analyzed, of which 42 are IOFs (365 observations) and 172 are COOPs (1697 observations). Observations relating to the IOFs are included in Group 1 (G1_IOFs) while observations relating to the cooperative firms are included in Group 2 (G2_COOPs). The data were analyzed by firstly applying a descriptive statistical approach. The preliminary findings of the research, considering all the samples of the 214 firms, for a total of 2064 observations (every observation is a fiscal year of FINSTATs data), can be summarized as follows (

Table 2. Descriptive statistics’ FRs—total sample PR-RE PDO firms (214 firms, 2064 observations).

Financial Ratio ID	Mean	Median	St. Dev. Sample	Skewness (g1)	Kurtosis (g2)
Sales (EUR)	17,249,449	3,531,659	74,204,918	9.63 ***	108.66
EBITDA (EUR)	922,545	110,854	4,724,239	10.46 ***	127.04
EBITDA:Sales (%)	4.93	3.17	6.74	127.16 ***	3725.21
Net Profit (EUR)	90,069	0	4,812,446	−32.39 ***	1268.19
Total Asset (EUR)	20,758,785	6,230,940	72,771,531	9.78 ***	114.08
Turnover (T)	0.70	0.66	0.31	5.78 ***	74.08
Equity Capital (EUR)	4,188,433	172,197	22,872,553	6.94 ***	72.07
NFP:Equity Ratio (DER)	101.87	4.84	584.13	6.35 ***	59.07
Return on Sales (ROS) (%)	1.30	0.90	41.06	−11.65 ***	695.99
Return on Asset (ROA) (%)	1.14	0.58	5.01	−17.92 ***	558.42
Return on Debt (ROD) (%)	0.95	0.75	0.01	1.45 ***	2.91
Return on Equity (ROE) (%)	1.24	0.00	10.40	−1.97 ***	54.97
INV_DAYS (Duration of Inventories)	399.55	388.29	239.83	10.93 ***	206.00
AR_DAYS (Duration of Acc. Receivable)	91.03	73.44	82.34	5.69 ***	62.25
AP_DAYS (Duration of Acc. Payable)	377.61	365.79	275.61	7.53 ***	116.88
CCC_DAYS (Cash Conversion Cycle)	112.97	88.78	226.86	0.97 ***	82.45
EM-Score	7.00	4.25	43.03	19.39 ***	400.70

Shapiro–Wilk (SW) test on the distributions’ normality; *** SW is significant at the 0.001 level.

), allowing us to respond to RQ1:

• The values relating to sales and income flows (sales, EBITDA, EBITDA: sales%, net profit) indicate that the sector has a median business turnover of around EUR 3.5/million, which falls within the small- and medium-sized enterprises. The median firm size is that of small businesses, and this is relevant for access to credit and the capital market, for the capacity to invest in RD, for the capacity to access national and foreign markets, and for the capacity to attract of qualified operational and managerial personnel; the data are asymmetric because the average turnover is EUR 17.2/million and this highlights the presence of some large firms by turnover, which determines the asymmetry of the data. Operating profitability before amortization (EBITDA) is confirmed asymmetric, both in absolute value (average is EUR 922,545 and median is EUR 110,854) and as a percentage of sales (EBITDA: sales average is 3.17 and median is 4.93), as also highlighted by skewness and kurtosis;
• The profitability of the firms in the sector is therefore low as a percentage of sales compared to other sectors [242] and this is one of the reasons for the research and the insights in the following paragraphs analyze this result by dividing the firms by legal form and altitude range. Net profit also has a noteworthy value; in fact, the average value is EUR 90,069 while the median value is EUR 0; there are at least two observations of this result: (1) the profitability of the firms is low in relation to the value of their sales; (2) the median value is affected by the cooperative legal form, as detailed in the following paragraphs;
• Another important piece of data concerns the total assets invested; the average of the total assets is EUR 20,758,785 with a median value of EUR 6,230,940. This allows us to observe that (1) even the investment in total assets, like the profit margins analyzed in the previous point, are asymmetric (skewness 9.78 and kurtosis 114.08); (2) the turnover of the firms in this sector is lower than threshold value one, with an average value of 0.70 and a median value of 0.66. It is therefore confirmed that the firms in the PR-RE PDO sector are also capital-intensive, as highlighted in other studies for agricultural and agri-food activities [243], in GI production and, in particular, for firms operating in PDO transformation [244];
• The fact that firms in the PR-RE PDO sector are capital-intensive raises the need to verify how investment coverage is carried out in terms of financial structure. For this reason, the analysis of the contribution of E and the relationship between NFP and E (DER) is required. The value of E is EUR 4,188,433 on average and EUR 172,197 as the median value. Notable asymmetry (6.94) and kurtosis (72.07) are also observed for this value. It is also observed that the ratio between E and TA is 2.76%, and these data highlight that a very low share of TA is financed with E^T; the net financial debt (NFP) is higher than with E, as highlighted by the DER, which has a mean value of 101.87 and a median value of 4.84; financial debt therefore has a greater weight among the sources of financing compared to E^T among the firms in the sector. This ratio is also asymmetric (5.32) and has kurtosis divergent from the normal distribution (30.59);
• Regarding the profitability FRs, very relevant results emerge; the operating profitability of sales (ROS) has a low value, 1.30% average, and 0.90% median (standard deviation is 41.06%). The ROA has a greater central tendency, with a mean of 1.13% and a median of 0.58% (standard deviation is 5.01%). The cost of debt (ROD) has an average value of 0.95% and a median of 0.75%. The ROE has an average value of 1.24% and a median of 0.00 (standard deviation 10.40). The joint reading of these FRs allows us to highlight that (a) the profitability of sales in the sector is low; (b) the operating profitability of the assets is equally low; (c) the cost of debt (ROD) is low and lower than the return on invested capital (ROA), even if with a low difference between the median values of the indices, which makes it possible to use leverage, even if with a low margin between the ROA and ROD;
• Regarding the FRs, which concern the duration of the financial cycle, it is observed in the total sample that the value of the warehouse duration (INV_DAYS) is 399.55 average days and 388.29 median days. It is therefore confirmed that the effect of the PS for PR-RE PDO cheese, which imposes a minimum maturation of the product of at least 12 months, has an effect on the duration of the warehouse cycle of the firms in the sample. We observe that this effect is differentiated by the legal form and by altitude range, considering that this has an effect on the firms’ strategy in terms of product differentiation, even beyond just the production of PR-RE PDO, producing or marketing food products with shorter inventories’ cycles. In the sector as a whole, there are also AR_DAYS with an average value of 91.03 and a median of 73.44, and this highlights that the collection of credits occurs in a relatively short duration. For the firms in the sample, the CCC_DAYS has a positive value, namely a 112.97 average value and median value of 88.78.

After having analyzed the FRs in the previous points, we can now analyze the EM-Score result for the total sample; in fact, the EM-Score value is calculated on the basis of Equation (12), in accordance with the work of Altman [218], using the data from FINSTATs and FRs. The average EM-Score value is 7.00 and the median value is 4.25, the standard deviation is 43.03, the skewness is 19.39, and the kurtosis is 400.70; the analysis of the EM-Score for the total sample allows us to observe that (a) on average, the scoring of the firms is high; in fact, a value of 7.00 corresponds to the AA-scoring class, which is in the investment range grade (the EM-Score lower risk band has a threshold value greater than 8.15 while the non-investment grade values are EM-Score < 3.75); (b) the median EM-Score value is much lower than the average, and is 4.25, corresponding to scoring class B, slightly higher than the lower limit of the investment grade; (c) the dispersion of the EM-Score values is high, and this is highlighted by the standard deviation (43.03), the skewness (19.39), and the kurtosis (400.70). We therefore observe a distribution of values that is highly asymmetric and divergent compared to the normal kurtosis. The outcome of the EM-Score confirms the need for a subdivision of the sample into further groups, to verify, as per the RQs and title of our research, whether the legal form (COOPs or IOFs) and the altitude at which the firm are discriminating elements in the values of both FRs and EM-Score. For completeness, we can observe (Table 2) that the normality analysis of the distribution of FRs and the EM-Score has highlighted how all the relative distributions, applying the Shapiro–Wilk statistics, with a maximum error not exceeding 0.001, diverge from the normal distribution and, consequently, a non-parametric statistics approach must be applied to subsequent processing; this result is consistent with what has already been observed by many other authors [245,246].

3.3. Comparison for Legal Form

After having analyzed the data of the total firms in the sample, we restrict the analysis of the FRs and EM-score to the IOFs only. The data of these firms are shown in

Table 3. Descriptive statistics FRs—G1_IOFs (42 firms, 365 observations).

Financial Ratio ID	Mean	Median	St. Dev. Sample	Skewness (g1)	Kurtosis (g2)
Sales (EUR)	72,663,023	6,166,468	164,999,561	3.92 ***	17.06
EBITDA (EUR)	4,285,416	736,912	10,562,433	4.32 ***	20.53
EBITDA:Sales (%)	10.79	8.82	9596.65	−1686.84 ***	31,814.21
Net Profit (EUR)	460,099	205,422	11,428,194	−13.81 ***	227.16
Total Asset (EUR)	75,162,372	14,411,127	160,465,964	4.11 ***	18.93
Turnover (T)	0.79	0.67	0.65	3.17 ***	19.27
Equity Capital (EUR)	21,356,954	5,189,094	50,762,208	2.54 ***	10.48
NFP:Equity Ratio (DER)	2.67	0.89	5.23	3.31 ***	16.59
Return on Sales (ROS) (%)	−2699.55	4.45	51,371.43	−19.10 ***	364.99
Return on Asset (ROA) (%)	3.50	2.84	10.20	−11.00 ***	180.20
Return on Debt (ROD) (%)	1.09	0.90	0.01	1.03 ***	0.85
Return on Equity (ROE) (%)	8.07	4.62	0.41	7.26 ***	92.50
INV_DAYS (Duration of Inventories)	313.54	280.15	313.91	6.59 ***	81.31
AR_DAYS (Duration of Acc. Receivable)	120.11	93.72	131.78	5.90 ***	43.77
AP_DAYS (Duration of Acc. Payable)	223.81	135.57	352.92	5.29 ***	34.66
CCC_DAYS (Cash Conversion Cycle)	209.84	211.70	427.42	−0.02 **	35.10
EM-Score	20.23	6.92	101.17	8.06 ***	68.08

Shapiro–Wilk (SW) test on the distributions’ normality. *** SW is significant at the 0.001 level; ** SW is significant at the 0.01 level.

, for the same FRs and EM-Score already analyzed in Table 2 relating to the total sample; there are 42 IOFs for 365 annual FINSTATs observations available. For a better understanding of the research results, we comment on the results of the IOFs together with the results of the COOPs (172 firms, 1697 observations) which are shown in

Table 4. Descriptive statistics FRs—G2_COOPs (172 firms, 1697 observations).

Financial Ratio ID	Mean	Median	St. Dev. Sample	Skewness (g1)	Kurtosis (g2)
Sales (EUR)	5,330,796	3,355,654	5,703,529	3.56 ***	18.86
EBITDA (EUR)	199,240	95,111	408,961	6.83 ***	68.14
EBITDA:Sales (%)	2.34	2.80	5892.45	27.42 ***	793.63
Net Profit (EUR)	9281	0	137,136	8.52 ***	238.34
Total Asset (EUR)	9,057,365	5,543,739	11,085,809	4.74 ***	37.53
Turnover (T)	0.68	0.66	0.17	2.41 ***	16.16
Equity Capital (EUR)	495.734	113,330	2,099.916	14.84 ***	291.99
NFP:Equity Ratio (DER)	123.21	6.58	641.88	5.72 ***	47.98
Return on Sales (ROS) (%)	1.71	0.73	24.57	40.77 ***	1674.03
Return on Asset (ROA) (%)	0.63	0.47	2.57	−22.28 ***	639.91
Return on Debt (ROD) (%)	0.89	0.70	0.01	1.64 ***	4.06
Return on Equity (ROE) (%)	−3.38	0.00	1.86	−39.02 ***	1592.02
INV_DAYS (Duration of Inventories)	418.05	396.96	216.24	14.09 ***	303.33
AR_DAYS (Duration of Acc. Receivable)	84.77	66.76	65.43	2.56 ***	13.55
AP_DAYS (Duration of Acc. Payable)	410.69	397.28	243.58	10.30 ***	202.49
CCC_DAYS (Cash Conversion Cycle)	92.13	71.06	144.18	1.26 ***	2.90
EM-Score	4.15	4.05	1.53	0.18 **	4.44

Shapiro–Wilk (SW) test on the distributions’ normality. *** SW is significant at the 0.001 level; ** SW is significant at the 0.01 level.

. In fact, as highlighted by Bertolini and Giovannetti [247], cooperative firms play a central role in the PR-RE PDO sector. Cooperative legal form requires the processing of milk supplied mainly by members and not purchased on the market; the contributing members are agricultural entrepreneurs that carry out dairy cattle breeding in the PR-RE PDO production area, as indicated in the PS. These are therefore processing cooperatives specialized in the production of PR-RE PDO, which is essential in single production (except for the modest quantities of other minor productions derived from milk or marketed), with a membership base restricted to agricultural firms involved in the transformation.

The data of the two groups of firms, IOFs and COOPs. highlight some relevant data and allow some observations:

• The average sales of the IOFs are EUR 72,663,023 (median value of EUR 6,166,468) while in the COOPs average sales are EUR 5,330,796 (median value of EUR 3,355,654). This first result allows us to observe that (a) the size of IOFs is approximately twice as large as that of COOPs firms and this is observed in the median value; (b) in the IOFs there are numerous large firms, as emerges from the average value of sales and the standard deviation EUR 164,999,561). The data therefore confirms the presence, in the IOFs, of some large groups of capitalist firms. Even if the FINSTATs data do not provide this information, it is possible to hypothesize that the larger IOFs also implement a strategy of differentiating production and marketing, while the COOPs, also due to legal and statutory constraints, appear concentrated in the production of PR-RE PDO only;
• Other relevant observations, which concern the legal form, are possible by analyzing the intermediate income margins. The data highlights that EBITDA in IOFs has an average value of EUR 4,285,416 and a median value of EUR 736,912; in COOPs, it has an average value of EUR 199,240 and a median value of EUR 95,111. It follows that EBITDA, as a percentage of sales in IOFs, has an average value of 10.79 and a median value of 8.82; in COOPs, it has an average value of 2.34 and a median value of 2.80. Again, regarding net profit, in the IOFs the average value is EUR 460,099 and the median value is EUR 205,422; in the COOPs, it has an average value of EUR 9281 and a median value of EUR 0. These data allow us to observe that joint-stock firms express a higher profitability than cooperatives if read through economic margins. It must be noted, however, that less profit generation emerges for reinvestment. These results then have an effect on the EM-Score, as we highlight later;
• The analysis of invested assets also has an important result. In fact, the total assets in the IOFs have an average value of EUR 75,162,372 and a median value of EUR 14,411,127; in the COOPs it has an average value of EUR 9,057,365 and a median value of EUR 5,543,739. The research therefore allows us to highlight that (a) as already observed for sales, even for total assets the dimensions of IOFs are larger than COOPs and there is, in the former, a greater asymmetry and standard deviation of values; (b) IOFs make greater investments than COOPs, and, consequently, we can hypothesize that these investments make it possible to overcome barriers to entry, make investments in RD, and carry out acquisition and integration strategies in a horizontal and vertical sense;
• The capital intensity of firms can be analyzed through turnover (T); this index, which, as is known, expresses the speed of the turnover of capital, relating sales (S) to invested capital (TA), indicates whether a firm is, or is not, capital-intensive. In the case of firms in the PR-RE PDO sector, the turnover in the IOFs has an average value of 0.79 and a median value of 0.67; in the COOPs, it has an average value of 0.68 and a median value of 0.66; the values are therefore similar between the two groups of firms and it is confirmed that the firms in the PR-RE PDO sector, whether they operate as IOFs or as COOPs, are in any case capital-intensive, having a T lower than the unit threshold value;
• Relevant, to answer RQ1, for the financial structure of firms, in absolute values, the equity contribution is greater in IOFs than in COOPs, but this result is rationally expected, given that investments in TA in IOFs are greater than in COOPs. We can observe that the ratio between the median values of E^T and TA in the IOFs is 36.01% and, in the COOPs, is 2.04%; these data express the evident undercapitalization of COOPs compared to IOFs in terms of capacity to collect equity; in this case, both among the members of the COOPs and among any external financiers, this result confirms other studies on COOPs [98,104,248]; to answer RQ1 regarding the capital structure, we also have interesting results analyzing the DER, calculated as the ratio between NFP and E^T; DER has a mean value of 2.67 and a median of 0.89 in the IOFs and has a mean value of 123.21 and a median of 6.58 in the COOPs; the DER also highlights a high financial debt (NFP) in the COOPs in relation to E^T, even regardless of the average value, which is influenced by asymmetry in the distribution and the probable presence of outliers; it should be noted that, in absolute value, NFP is not high, and the median value of this index (6.58) should be read considering that E^T in COOPs is very low (2.04% of TA). In response to RQ1, it is therefore confirmed that COOPs depend on financial leverage, even if (as will be seen in the course of the exposition) the credit scoring (EM-Score) of the COOPs is worse than that of the IOFs;
• The analysis of the FRs relating to profitability (ROS, ROA, ROD and ROE) allows us to answer RQ1; in the IOFs, the median values of the FRs are as follows: ROS = 4.45%, ROA = 2.84%, ROD = 0.90%, and ROE = 4.62%; in COOPs, the median values of FRs are as follows: ROS = 0.73%, ROA = 0.47%, ROD = 0.70%, and ROE = 0.00%. The data confirms that COOPs have lower income index values than IOFs, for ROS, ROA, and ROE also in the PR-RE PDO sector. We also observe that for IOFs they can use leverage positively, given that the ROA is greater than the ROD; on the other hand, COOPs have a ROA lower than ROD and therefore cannot use leverage because the cost of debt is greater than the return on capital, and this is evident even if the ROA is decreased by remuneration to the cooperative members for the milk deliveries made; this represents a limit to the ability of COOPs to access the debt capital market and therefore highlights that they are financially constrained firms. The zero ROE level confirms that the COOPs of the sector, in these budget conditions, are not able to attract capital from non-user subjects, such as capital-only shareholders, and this represents a constraint on access to the capital market of risk;
• A further analysis to answer RQ1 concerns the CCC expressed in days (CCC_DAYS) and, consequently, the monetary value (EUR) assumed by NWC in the FINSTATs of firms in the PR-RE PDO sector. In this analysis, we have developed three FRs, namely INV_DAYS, AR_DAYS and AP_DAYS, which are used in the calculation formula of CCC_DAYS. These indices have the following median values in the IOFs expressed as usual in days: INV_DAYS = 280.15, AR_DAYS = 93.72, and AP_DAYS = 135.57. Consequently, the median value of CCC_DAYS = 211.70; the median values in the COOPs expressed as usual in days are INV_DAYS = 396.96, AR_DAYS = 66.76, AP_DAYS = 397.28; consequently the median value of CCC_DAYS = 71.06. The data relating to AP_DAYS is very interesting, which highlights an average of 377.61 and a median of 365.79; these results are of great interest because they highlight how, in COOPs, the duration of the warehouse cycle (INV_DAYS) is almost entirely financed by AP_DAYS. The data from this analysis allow us to observe, in response to RQ1, that the duration of INV_DAYS in the COOPs is approximately 117 days longer than in the IOFs; this result highlights that the COOPs matured the PR-RE PDO cheese for a longer time than to the IOFs or that the IOFs, in addition to the PR-RE PDO, produce or market other cheeses with shorter maturing duration, obtaining a lower use of capital and improving the turnover of the invested capital. The shorter duration of the INV_DAYS cycle allows the IOFs to use the leverage of the commercial credit, given by the duration of AR_DAYS, granting extensions for approximately 27 more days, and this represents a significant competitive leverage that the IOFs can use compared to the COOPs. Furthermore, the data highlights a difference of approximately 262 days in the duration of AP_DAYS between IOFs (135.57 days) and COOPs (397.28 days); these data are very important in the response to RQ1 because they highlight that the capital structure of the COOPs is unbalanced on the AP_DAYS, which are deferred to over a year, and in this class of values we also have the trade credits of the cooperative members for the milk delivered; it therefore emerges that the members of the COOPs perform the function of financiers of the COOPs, not only as members, but also as suppliers, and this evidence in the PR-RE PDO sector confirms other research [249,250,251]. A reflection on this point is necessary, which opens up the following in-depth analysis of the data in PR-RE PDO COOPs. In fact, it frequently happens, and the data seems to confirm it, that the COOP members who deliver the milk that is sold to the cooperative to be transformed into cheese, grant payment extensions to the cooperative in the meantime as the cheese is matured and subsequently sold; therefore, the trade credits and, consequently, the trade debts, of the cooperative members are formed (which are part of the AP_DAYS) in the cooperative’s balance sheet. It should be noted that the members of COOPs intervene, in addition to commercial deferrals, also with loans of a financial nature, to support the investment needs of the cooperative, both in FA and in the maturation of the PR-RE DOP (INV_DAYS); these loans are classified in the balance sheet as financial debts and therefore become part of the NFP.

After commenting on the absolute values of the FRs, to answer RQ1, we carried out a comparison between the median values of the FRs of the G1_IOFs and G2_COOPs groups; preliminarily, we carried out a check on the normality of the distribution of all the FRs and the EM-Score belonging to the two groups of FRs and the EM-Score by applying Shapiro–Wilk statistics. The observed result is that all distributions diverge from the normal distribution, with a maximum error not exceeding 0.01 (Table 3 and Table 4). Accordingly, we applied the Mann–Whitney U test; this is the non-parametric alternative to the t-test, to be used in case of analysis on unpaired samples with a distribution divergent from the normal distribution. The U test is used to test the null hypothesis that two samples come from the same population and, therefore, have the same median; the decision, for the purposes of this research, was, therefore, to use the U test to verify whether the observations between the two samples being researched (G1_IOFs and G2_COOPs) have medians that are different from each other in a statistically significant way, therefore coming from a different statistical universe, and, therefore, it can be stated that the median values of the FRs and EM-Score of the two samples are statistically different from each other. The analysis tested the following 17 null hypotheses:

H1. 

Sales (EUR) in G1_IOFs and G2_COOPs have equal medians.

H2. 

EBITDA (EUR) in G1_IOFs and G2_COOPs have equal medians.

H3. 

EBITDA:Sales (%) in G1_IOFs and G2_COOPs have equal medians.

H4. 

Net Profit (EUR) in G1_IOFs and G2_COOPs have equal medians.

H5. 

Total Asset (EUR) in G1_IOFs and G2_COOPs have equal medians.

H6. 

Turnover (T) in G1_IOFs and G2_COOPs have equal medians.

H7. 

Equity Capital (EUR) in G1_IOFs and G2_COOPs have equal medians.

H8. 

NFP:Equity Ratio (DER) in G1_IOFs and G2_COOPs have equal medians.

H9. 

Return on Sales (ROS) (%) in G1_IOFs and G2_COOPs have equal medians.

H10. 

Return on Asset (ROA) (%) in G1_IOFs and G2_COOPs have equal medians.

H11. 

Return on Debt (ROD) (%) in G1_IOFs and G2_COOPs have equal medians.

H12. 

Return on Equity (ROE) (%) in G1_IOFs and G2_COOPs have equal medians.

H13. 

INV_DAYS (Duration of Inventories) in G1_IOFs and G2_COOPs have equal medians.

H14. 

AR_DAYS (Duration of Acc. Receivable) in G1_IOFs and G2_COOPs have equal medians.

H15. 

AP_DAYS (Duration of Acc. Payable) in G1_IOFs and G2_COOPs have equal medians.

H16. 

CCC_DAYS (Cash Conversion Cycle) in G1_IOFs and G2_COOPs have equal medians.

H17. 

EM-Score in G1_IOFs and G2_COOPs have equal medians.

We carried out the calculations of the Mann–Whitney U test (

Table 5. Comparison FRs—G1_IOFs and G2_COOPs.

Financial Ratio ID	Median G1_IOFs	Median G2_COOPs	Null Hypotheses	Accept/Reject
Sales (EUR)	6,166,468	3,355,654	H1	Reject ***
EBITDA (EUR)	736,912	95,111	H2	Reject ***
EBITDA:Sales (%)	8.82	2.80	H3	Reject ***
Net Profit (EUR)	205,422	0	H4	Reject ***
Total Asset (EUR)	14,411,127	5,543,739	H5	Reject ***
Turnover (T)	0.67	0.66	H6	Accept
Equity Capital (EUR)	5,189,094	113,330	H7	Reject ***
NFP:Equity Ratio (DER)	0.89	6.58	H8	Reject ***
Return on Sales (ROS) (%)	4.45	0.73	H9	Reject ***
Return on Asset (ROA) (%)	2.84	0.47	H10	Reject ***
Return on Debt (ROD) (%)	0.90	0.70	H11	Accept
Return on Equity (ROE) (%)	4.62	0.00	H12	Reject ***
INV_DAYS (Duration of Inventories)	280.15	396.96	H13	Reject ***
AR_DAYS (Duration of Acc. Receivable)	93.72	66.76	H14	Reject **
AP_DAYS (Duration of Acc. Payable)	135.57	397.28	H15	Reject ***
CCC_DAYS (Cash Conversion Cycle)	211.70	71.06	H16	Reject ***
EM-Score	6.92	4.05	H17	Reject ***

*** The relation is significant at the 0.001 level (two-tailed). ** The relation is significant at the 0.01 level (two-tailed).

) and it emerged that (1) we can reject the null hypothesis (p < 0.001 significance, two-tailed test) for the tests concerning H1, H2, H3, H4, H5, H7, H8, H9, H10, H12, H13, H15, H16, H17; (2) we can reject the null hypothesis (p < 0.01 significance, two-tailed test) for the test concerning H14; (3) we can accept the null hypothesis (p > 0.05 significance, two-tailed test) for the tests concerning H6 and H11 (Table 5).

The comparison of the data therefore allows us to conclude that there is a statistical significance in the differences between IOF and COOP firms; therefore, by the legal form of the firms, in the PR-RE PDO sector, this difference concerns both aspects of the firm size, in terms of sales and turnover, and in terms of the FRs that express the profit margins (ROA, ROS, ROE, and EBITDA) as the percentage of sales. The levels of indebtedness and the use of risk capital between the sources of financing between IOFs and COOPs are also differentiated, and this is a further relevant response to RQ1 of the research, regarding the capital structure.

The duration of the CCC expressed in days (CCC_DAYS) is also statistically different between IOFs and COOPs, and this difference is due to the difference found between the groups in INV_DAYS and AP_DAYS; the AR_DAYS difference is also significant, although less significant (p < 0.05). It is therefore observed that the duration of CCC_DAYS is greater in IOFs than in COOPs (211.70 compared to 71.06), but the duration of INV_DAYS is greater in COOPs than in IOFs (396.96 compared to 280.15). It is therefore highlighted that the COOPs have a longer duration of their maturing cycle, and this is due to a greater specialization of the COOPs in the production of PR-RE PDO compared to the IOFs that also produce other less mature cheeses and market-ready matured products, such as can be read in the supplementary notes of the FINSTATs of the firms included in this research. The research allows us to conclude on the subject of the duration of INV_DAYS that the legal form of the firms and the PS of the PR-RE PDO is relevant, and this answers the RQ1 regarding the capital structure. A further statistically significant difference is also observed in the extension granted by non-financial creditors, expressed by AP_DAYS, which is approximately five times the duration that the IOFs in the sample grant to their customers (AR_DAYS).

In response to RQ1, we can conclude that the EM-Score value is statistically different in IOF and COOP firms; this result is significant because it allows us to observe that the IOFs and the COOPs have budget parameters regarding profitability, the CCC, and the financial structure in terms of E and NFP, as expressed by the FRs, which determines a different result in terms of credit scoring. On this point, however, it emerges that the ROD between IOFs and COOPs is not statistically different, and therefore we cannot conclude that the difference in credit scoring determines an increase in debt pricing for COOPs compared to IOFs; on this point, it should however be noted that, although the use of financial debt is greater in the COOPs than in the IOFs, as emerges from the DER, however, the COOPs use commercial debts (AP_DAYS) a lot to finance themselves, as well as financing from members. In this way, they reduce the use of financial debt and it is rational to hypothesize that this could reduce the pricing of loan operations, as well as allow COOPs to overcome financial contracts. The presence of collateral to guarantee the loans granted to the COOPs can be further investigated, including the pledge on the cheese as a guarantee, pursuant to the rotating pledge on PDO and PGI food products established by Law no. 27 of 24 April 2020, which some banks propose as a specific line of financing for PR-RE PDO firms, both IOFs and COOPs. It is also possible that banks, in the case of COOPs, as well as IOFs, ask for surety signatures from members as collateral for their financing lines, and this could align the pricing of the financing between IOFs and COOPs. This topic could be the subject of further research.

To answer RQ1, we applied the following regression model:

ROE_t = α + β₁(ROA_t − ROD_t) + β₂T + β₃DER + β₄CCC_DAYS

(13)

Model (1) allows us to highlight some aspects that are useful for theoretical reflection, in fact, (1) β₁ is the coefficient of the difference between the operating profitability (ROA) and the cost of debt (ROD) of Equation (10); (2) β₂ is the coefficient of T, which expresses the efficiency of capital use through rotation; (3) β₃ is the coefficient of financial leverage (DER); (4) β₄ is the coefficient of the duration of the monetary conversion cycle in days (CCC_DAYS). The independent variable, which expresses the firm’s performance, is given by ROE. All variables were considered in the reference year (t). For the application of (1), the observation of the data revealed for ROE is the presence of numerous outliers, with two tails. Outliers greater or less than, in absolute value, 1.5 times the interquartile range (Q3_ROE–Q1_ROE), given by the difference between the observed third quartile value (Q3_ROE) and the first quartile value (Q1_ROE), were then eliminated. The topic of outliers in the distribution of FRs has been widely observed in the literature starting from the seminal works of Frecka and Hopwood [252], Watson [253] and So [254]; the topic was then explored in depth regarding the issues of identifying outliers [255,256]. Interesting developments have been made by Coenders et al. [257,258], who have proposed a normalization method of FRs through the application of the statistical analysis of compositional data. There have also been recent applications in the field of credit scoring and bankruptcy prediction [259].

A stepwise regression procedure followed that allowed (a) eliminating the independent variables characterized by multicollinearity on the basis of the Variance Inflation Factors (VIF) statistic and are found to be higher than the threshold value of two; (b) eliminating the independent variables, whose contribution to the model is not statistically significant (p-value). It is therefore observed that the independent variables ROA, sales, TA, ROA, ROS, INV_DAYS, AP_DAYS, and AR_DAYS are excluded from the applied model. The model is applied to analyze the performance of IOFs and COOPs, to complete the answer to RQ1. Regarding IOFs, we have

IOF_ROE_t = α + β₁(IOF_ROA_t − IOF_ROD_t) + β₂IOF_T + β₃IOF_DER + β₄IOF_CCC_DAYS

(14)

In

Table 6. Regression model (14).

ANOVA	DF	SS	MS	F	p-Value	Sig.
Regression	4	0.9380	0.2345	111.6331	0.0000 ***	yes
Residual	310	0.6512	0.0021
Total	314	1.5891
Regression model	Coeff.	Std. Error	t-stat	p-value	lower	upper	VIF
Intercept	0.0160	0.0052	3.0552	0.0024 **	0.0057	0.0262
ROA–ROD	1.4073	0.0714	19.7056	0.0000 ***	1.2668	1.5478	1.1632
T	0.0063	0.0044	1.4198	0.1567	−0.0024	0.0150	1.1126
DER	0.0022	0.0007	3.0733	0.0023 **	0.0008	0.0036	1.1263
CCC_DAYS	−0.0000	0.0001	−1.7845	0.0753	−0.0000	0.0000	1.0146
Regression analysis	Overall Fit
Multiple R	0.7683
R square	0.5902
Adjusted R Square	0.5849
Standard Error	0.0458
Q1_ROE	0.0113
Q3_ROE	0.1052
Observations	315

*** The relation is significant at the 0.001 level (two-tailed). ** The relation is significant at the 0.01 level (two-tailed).

, the results of the (14) model allow us to highlight that the model fits the explanation of ROE, with a F value of 111.6331 (p < 0.001). The independent variables that have the greatest explanatory power are ROA-ROD (p < 0.001) with a positive sign of the coefficient (1.4073) and DER (p < 0.01) with a positive sign of the coefficient (0.0022); the signs are consistent with the theory. The variables T (p = 0.1567) with a positive sign of the coefficient (0.0063) and CCC_DAYS (p = 0.0753) with a negative sign of the coefficient (0.0001) are also less significant, even if with a tendency towards significance and therefore, in the opinion of the authors, useful for the model; note that the sign of T is in accordance with the theory, while the sign of CCC_DAYS is debated, and in our research this has a negative value, although with only a tendency significance, highlighting that an increase in the duration of CCC_DAYS has a moderately negative effect on performance (ROE). The model explains 58.49% of the observed variability (Adjusted R Square), noting that there are 315 observations out of an original 365, given that 50 observations are eliminated as they are classified as anomalous values; this confirms that there is a high variability in the observed data for the ROE (compare the skewness result for the ROE already shown in Table 4, which already highlighted a high asymmetry of distribution.

Regarding COOPs, we then have

COOP_ROE_t = α + β₁(COOP_ROA_t − COOP_ROD_t) + β₂ COOP_T + β₃COOP _DER + β₄COOP _CCC_DAYS

(15)

In

Table 7. Regression model (15).

ANOVA	DF	SS	MS	F	p-Value	Sig.
Regression	4	1103.98	275.99	98.0164	0.0000	yes
Residual	1.692	4764.36	2.8158
Total	1.696	5868.35
Regression model	Coeff.	Std. Error	t-stat	p-value	lower	upper	VIF
Intercept	−0.9806	0.1824	−5.3759	0.0000	−1.3384	−0.6228
ROA-ROD	29.4588	1.5594	18.8908	0.0000 ***	26.4002	32.5174	1.0016
T	1.4086	0.2501	5.6317	0.0000 ***	0.9180	1.8991	1.0552
DER	−0.0000	0.0001	−0.5197	0.6033	−0.0002	0.0001	1.0311
CCC_DAYS	0.0008	0.0003	2.6102	0.0091 **	0.0002	0.0013	1.0838
Regression analysis	Overall Fit
Multiple R	0.4337
R square	0.1881
Adjusted R Square	0.1862
Standard Error	1.6780
Q1_ROE	0.0000
Q3_ROE	0.000
Observations	1.697

*** The relation is significant at the 0.001 level (two-tailed). ** The relation is significant at the 0.01 level (two-tailed).

, the results of the (15) model allow us to highlight that the model fits the explanation of ROE, with an F value of 98.0164 (p < 0.001). To understand the model, we recall that the remuneration of cooperative members often occurs, as exposed in the literature analysis, with refunds, that is, the remuneration of the raw material delivered, and not with the profits. The regression model, which is well-suited to the IOFs, must consider this premise. The independent variables that have the greatest explanatory power are ROA-ROD (p < 0.001) with a positive sign of the coefficient (29.4588) and T (p < 0.01) with a positive sign of the coefficient (1.4086); the signs are consistent with the theory. The variable CCC_DAYS (p = 0.1567) with a positive sign of the coefficient (0.0063) and CCC_DAYS (p = 0.0753) with a positive sign of the coefficient (0.0008) are also significant (p < 0.0091), while DER, considered in the (14) model here, is not significant (p = 0.6033). The model explains only 18.62% of the observed variability (Adjusted R Square). This result requires specifying that, out of 1697 observations, the ROE is negative in 15 cases, equal to zero in 1533 cases, and positive in 149 cases, Consequently, no outliers are eliminated with the interquartile difference methodology used for IOFs. It is therefore clear that the COOPs in the PR-RE PDO sector have followed a policy of zeroing the profits in the financial statements, with the consequent zeroing of the ROE in the majority of observations, and the remuneration of the contributions of the members through refunds, i.e., the remuneration of the raw milk material contributed by the members; data shows the undercapitalization of COOPs in the PR-RE PDO sector, since reinvested profits, present in only 149 cases out of 1697, are a minority share of the observations. The results therefore confirm that the FINSTATs of the COOPs require further specifications to be comparable to the FINSTATs of the IOFs, where the remuneration of the members occurs with the profits and not with the refunds; regarding this topic, the impact of raw materials on sales, among which the milk supplied by members for processing is the main one, has a differentiated value between COOPs and IOFs. In the COOPs, the incidence is 85.86% on average and 82.85% is the median value, while in the IOFs, the incidence drops to 70.89% on average and 73.21% for the median value. There is, therefore, a difference of approximately nine percentage points greater in COOPs compared to IOFs in the cost of raw materials on sales on which, by applying the Mann–Whitney U test, it is possible to reject the null hypothesis of the equivalence of the median values (p < 0.001 significance, two-tailed test). Remembering that the refunds are remunerated to the members in the COOPs with a greater value for the raw materials, we can conclude that one of the reasons for the greater incidence of the costs of the raw materials in the COOPs may be the effect of the refunds to the members and, consequently, the underestimation of ROE in the COOPs’ FINSTATs as the performance index confirmed; other factors may have an impact, such as lower average sales prices of the finished product in the COOPs, the costs of raw materials not present in the IOFs, and the budget policies of stabilizing the results through the overvaluation or undervaluation of the inventory stock. Even with these preliminary conclusions, the RQ can be the subject of further in-depth studies that require more mandatory details in the firms’ FINSTATs.

To answer RQ2, we divide the sample again into two further subgroups, therefore considering, in addition to the legal form (divided into IOFs and COOPs), the altitude ranges in which the dairy included in the research is located. The sample is then divided as follows: (1) the G1_IOFs sample, of 42 for 365 available annual FINSTATs observations, is divided into two, discriminating the firms by altitude range, and the G1.A_IOFsPH is obtained, made up of 40 firms with 353 observations and G1.B_IOFsM formed by 2 firms with 12 observations; (2) the G2_COOPs sample, of 172 for 1697 annual observations available, is divided into two, discriminating the firms by altitude, and we obtain the G2.A_COOPsPH, made up of 129 firms with 1267 observations and the G2.B_COOPsM made up of 41 firms with 410 observations. In Table 5, as is performed previously in the comparison between the IOFs and the COOPs, the division into four sub-samples is also subject to the verification of the equality between the median values of the FRs. The comparison is carried out this time by comparing the IOFs of the plains and hills (G1.A_IOFsPH) with those of the mountains (G1.B_IOFsM) and the COOPs of the plains and hills (G2.A_COOPsPH) with those of the mountains (G2.B_COOPsM), given that the comparison has already been made between the total of the observations relating to the IOFs and the total of those of the COOPs. To verify whether the difference between the median values is statistically significantly different from the sub-samples we applied the Mann–Whitney U statistic, responding to 34 null hypotheses. The first set of 17 null hypotheses is related to the relative comparison difference between the medians of the 17 FRs of G1.A_IOFsPH and G1.B_IOFsM; the second set of 17 null hypotheses is related to the relative comparison of difference between the medians of the 17 FRs of G2.A_COOPsPH and G2.B_COOPsM. To improve the exposure, the data are exposed by combining the data of the four sub-samples in

Table 8. Descriptive statistic and comparison FRs (4 sub-samples).

Financial Ratio ID	Sub-Sample G1.A_IOFsPH 40 Firms 353 Observ.	Sub-Sample G1.B_IOFsM 2 Firms 12 Observ.	Null Hypotheses	Accept/Reject	Sub-Sample G2.A_COOPsPH 129 Firms 1267 Observ.	Sub-Sample G2.B_IOFsM 43 Firms 430 Observ.	Null Hypotheses	Accept/Reject
Sales (EUR)	6,833,438	1,387,370	H1	Reject ***	3,625,536	2,950,453	H18	Reject **
EBITDA (EUR)	746,902	−6054	H2	Reject ***	91,620	96,546	H19	Accept
EBITDA:Sales (%)	8.50	−0.94	H3	Reject ***	2.50	3.38	H20	Reject ***
Net Profit (EUR)	228,305	−134,796	H4	Reject ***	0	0	H21	Accept
Total Asset (EUR)	14,685,978	5,394,740	H5	Reject ***	5,980,183	4,901,223	H22	Reject ***
Turnover (T)	0.62	0.25	H6	Reject ***	0.67	0.61	H23	Accept
Equity Capital (EUR)	5,626,657	272,584	H7	Reject ***	113,330	117,819	H24	Accept
NFP:Equity Ratio (DER)	0.72	0.00	H8	Reject ***	7.82	3.57	H25	Reject ***
Return on Sales (ROS) (%)	4.60	−3.24	H9	Reject ***	0.71	0.76	H26	Accept
Return on Asset (ROA) (%)	2.72	−0.85	H10	Reject ***	0.47	0.47	H27	Accept
Return on Debt (ROD) (%)	0.84	1.71	H11	Reject ***	0.67	0.74	H28	Accept
Return on Equity (ROE) (%)	4.62	−46.35	H12	Reject ***	0.00	0.00	H29	Accept
INV_DAYS (Duration of Inventories)	266.80	368.45	H13	Reject ***	404.36	376.09	H30	Accept
AR_DAYS (Duration of Acc. Receivable)	92.21	182.95	H14	Reject **	66.34	67.56	H31	Accept
AP_DAYS (Duration of Acc. Payable)	134.87	273.39	H15	Reject ***	390.29	415.71	H32	Accept
CCC_DAYS (Cash Conversion Cycle)	211.21	253.93	H16	Reject **	82.88	22.44	H33	Reject ***
EM-Score	6.99	3.65	H17	Reject ***	4.19	3.39	H34	Reject ***

*** The relation is significant at the 0.001 level (two-tailed). ** The relation is significant at the 0.01 level (two-tailed).

.

The data highlights that the IOFs in the plains and hills (G1.A_IOFsPH) have the largest size in terms of turnover and invested capital, and this difference is statistically significant (test of the null hypotheses from 1 to 17). The sample has an adequate size in the first sub-sample (G1.A_IOFsPH), in which we have 40 firms and 353 observations, while for the second sub-sample (G1.B_IOFsM) we only have 2 firms and 12 observations, and this allows us to have statistically significant conclusions only for the first sub-sample G1.A_IOFsP; the answer to the null hypothesis questions and the consequent conclusions in the answers to the RQ2 must consider the reduced number of the G1.A_IOFsM sub-sample. Given this premise, the plain and hill IOFs (G1.A_IOFsPH) are also able to generate a higher turnover compared to the mountain IOFs (G1.A_IOFsM) which; therefore, are firms with greater capital intensity and therefore with greater barriers to entry (p < 0.001). The lowland IOFs obtain better performances in terms of EBITDA, EBITDA as a percentage of sales and net profit compared to the mountain IOFs, and this finding is also new and statistically significant (p < 0.001). This result is also confirmed by ROS, ROA, ROD and ROE. It emerges that lowland IOFs manage to remunerate risk capital (ROE) with positive margins on sales (ROS) and on invested capital (ROA); the cost of debt (ROD) is lower in hill and plain IOFs than in mountain ones (p < 0.001). As for other GI food sectors analyzed in other research [260,261,262], it is highlighted that the firm strategy is based on margins (ROS) and to a lesser extent on turnover, which is lower than threshold value one, and therefore confirms that IOFs are capital intensive, although to a lesser extent the plain and hill IOFs compared to the mountain IOFs.

The IOFs in the plains and hills use risk capital (equity capital) to a greater extent to finance themselves, while the IOFs in the mountains attract little equity capital, around 5.05% of the financing needs (p < 0.001). This observation is also statistically significant (p < 0.001). The mountain IOFs use NFP to a small extent to finance themselves, while they use to a significant extent the suppliers’ deferral (AP_DAYS) which has a median duration of 273.39 days compared to the 134.87 days of the lowland IOFs; this difference is also statistically significant (p < 0.001). The duration of the maturing cycle in the mountain IOFs is higher (INV_DAYS = 368.45) than in the plain and hill IOFs (INV_DAYS = 266.80); this difference is also statistically significant (p < 0.001) and provides a relevant answer to RQ2 because it allows us to conclude that: (1) the INV_DAYS cycle is high in both sub-samples of the IOFs, and is affected by the minimum duration of the maturation of the PR- RE PDO mandatory for the PS for a minimum of 12 months; (2) despite this, IOFs in the plains, through the differentiation of production and trade, manage to reduce the average duration of the warehouse by approximately 102 days compared to mountain firms and this reduces financial needs and the consequent barriers to entry, it reduces the cost of capital and the turnover of invested capital, reduces the price risk of the finished product and the operational risk associated with maturing. An extension of the research will be able to investigate whether lowland or mountain firms that carry out long seasoning have a premium price on the finished product that allows them to compensate for the risks of prolonging the seasoning of PR-RE PDO and the related costs. Finally, regarding the EM-Score, the statistically significant difference (p < 0.001) between plain and hill IOFs (EM-Score = 6.99) and mountain IOFs (EM-Score = 3.65) is confirmed; this result is very relevant because the EM-score value 6.99 observed for the plain and hill IOFs is in the A+ class; therefore, investment grade, while the EM-score value 3.25 observed for the mountain IOFs is in the CCC+ class; therefore, non-investment grade.

In response to RQ2 we can conclude that the plain and hill IOFs of the PR-RE PDO sector have an average high credit score and therefore are not financially constrained, while the mountain IOFs of the PR-RE PDO sector have an average credit score low and therefore financially constrained; in fact, confirming this assessment, we read that these firms make little use of financial debt (NFP) and equity capital, as emerged from the analysis of the FRs. We must observe that the mountain IOFs are small in number (2) and the number of observations is equally small (12); it is therefore not possible to draw conclusions of general validity from the comparisons, but only statistically significant indications regarding the FRs and EM-Score values of the plain IOFs; this result is however relevant because, as far as we know, it is the first piece of research that has carried out this analysis in the PR-RE PDO sector. In management and policy terms, this observation has applicative relevance; in fact, in the financing of dairies, the following are particularly used: (1) collateral guarantees, also with the instruments of revolving pledge in agriculture. Loans to finance the warehouse of agri-food firms are guaranteed by a non-possessory pledge as real collateral; in Italy, the specific rule is Article 1 of Legislative Decree 59/2016, then amended and converted into Law no. 199/2016. The non-possessive pledge differs from the pledge provided by Article 2784 of the Civil Code because the firm can continue to dispose of the asset covered by the guarantee which therefore remains functional to the production cycle. The non-possessory pledge rule has also been extended to PDO and PGI products, including PR-RE PDO, and this has favored the financing of the working capital cycle in Italian agri-food firms [263]. (2) A growing number of firms are resorting to ESG-compliant emissions, which are becoming one of the main markets for the collection and use of capital in the agri-food sector [264].

We can now proceed with the analysis and comparison (null hypotheses from 18 to 34) of the plain and hill COOPs (G2.A_COOPsPH) and of the mountain (G2.A_COOPsM). This sample has an adequate size, as we have 129 firms and 1267 observations for G2.A_COOPsPH and 43 firms and 430 observations for G2.B_COOPsM, and this allows us to have statistically significant conclusions. In general, responses to the null hypotheses do not yield statistically significant results; in fact, we reject the null hypothesis of equality between the averages between sub-samples of the COOPs in 6 cases out of 17. In particular, (1) the COOPs in the plains and hills have an average firm size of sales expressed in euros and total assets compared to the mountain COOPs (p < 0.01); (2) the COOPs in mountain have an average profitability of sales expressed as a percentage of EBITDA on sales, compared to the plains and hills COOPs (p < 0.001); the COOPs in the mountain have an average profitability of sales expressed as a percentage of EBITDA upon sales, compared to the plains and hills COOPs; (3) in the COOPs of the plains and hills, there are greater investments than in the mountains, with a significance p < 0.001; to finance investment, data show a greater use of financial debt as a percentage of equity capital (DER) in plain and hill COOPs, with a significance of p < 0.001; (4) in the COOPs of the plains and hills, data show a higher duration of CCC_DAYS (cash conversion cycle), with a significance of p < 0.001; to finance investment, data show a greater use of financial debt as a percentage of equity capital (DER) in plain and hill COOPs, with a significance of p < 0.001; (5); the median credit score of the plain and hill COOPs is higher (4.19) than the mountain COOPs (3.39) and, even if the difference is not high in absolute value as instead found in the sub-samples of the IOFs, there is still a statistically significant difference (p < 0.001); and in the case of plain and hill COOPs we are in the investment grade range (B rating class), while in the case of mountain COOPs we are in the non-investment grade range (CCC+ rating class) even if at the upper limit of the class. The analysis of the data leads us to accept the null hypotheses of the remaining 11 comparisons, and, therefore, we can state that in the COOPs, there are no statistically significant differences in profit margins (EBITDA, net profit), in turnover (T), in the value of equity capital, in income FRs (ROS, ROA, ROE), and in the cost of debt (ROD). In response to RQ2, we can therefore observe that, within the COOPs, the statistically significant differences depending on the altitude range between plain and hilly COOPs and mountain COOPs is limited to some management parameters (6 out of 17). We can observe that many of the acceptances of the null hypotheses are related to FINSTATs’ parameters for which the legal form of the cooperative can have an effect, as highlighted by many studies on the sector [265,266] which we can therefore confirm; these parameters are, in particular, related to the EBITDA and net profit margins, and to the income FRs connected to them (ROA, ROE, ROS), given that the remuneration of the contributions of the members in the PR-RE PDO COOPs often occurs, not with the profits, but with the remunerating the milk paid by the cooperative to the contributing members with a premium price above the market price. The data confirms that mountain COOPs are financially constrained; in fact. they make less use of financial debt (DER) and have lower investments (TA) compared to lowland COOPs. Even the size in terms of sales (S) is smaller in the mountain COOPs compared to the plains and hills. The research allows for some policy notes: (1) mountain dairies, COOP and IOF, are both relevant for the protection of the environment, the territory, and local communities; an expansion of this research could concern the strategies of this cluster of firms; (2) COOPs are mainly single-product firms, while IOFs largely produce and trade a wide range of food products, mainly cheeses of various types, but also other livestock production; an extension of this research could concern the marketing strategies of single-production firms and multi-product firms in the dairy sector. Despite these constraints, the legal form of the cooperative has allowed, in particular small producers, the overcoming of barriers to entry for investments in production equipment, and achieve economies of scale necessary for the transformation and subsequent commercialization of food production [267].

A further topic of interest therefore concerns the location of dairies, be they COOPs or IOFs. In fact, a significant portion of dairies are located in the mountains; these firms suffer from some disadvantages: (1) they have higher milk collection and processing costs; (2) they have a smaller average firm size, with fewer production and commercial economies of scale, and with less use of joint production economies; (3) the small average size generally makes access to credit more difficult and more expensive. Furthermore, mountain dairies have been negatively affected by the reduction in mountain livestock farms, as highlighted by census data [268]; this has reduced milk production in the mountains, and, consequently, the social base of cooperative dairies and the milk available for processing. Despite these constraints, PR-RE DOP mountain cheese is appreciated by consumers [269] and has been the subject of a valorization policy by the PR-RE DOP Consortium [270] which has applied EU Regulation 1151/12, which puts forward the indication of “mountain product” as an optional quality mention to classify food products originating in mountainous areas of the European Union, including PR-RE DOP produced in the mountainous area within the production area, with a premium price recognized by the consumer [271].

Research has highlighted that IOFs are more capitalized in terms of equity than COOPs, and that COOPs are consequently more indebted. The EM-Score values confirm these data and highlight better results for IOFs; COOPs in the plains and hills have better EM-Scores than mountain COOPs and, in fact, have greater access to financial debt capital, as highlighted by the DER; this result allows us to answer RQ2 and fits into our line of research that has commented on the financial sustainability of COOPs.

Regarding the possibility of generalizing the research results, some limitations must be taken into consideration:

(1)

PR-RE PDO is a typical product, the production of which is linked to a limited territory in northern Italy. The social structure and the historical background, and the interaction between agriculture and other sectors have determined the typical nature of the structure of the firms in terms of size and peculiar contracts. Cooperative legal form is particularly present in the sector, and, for the processing of milk, the annual procurement contract is used in some provinces, in particular Parma, which provides that the processing phase is contracted out to an entrepreneur who carries out the transformation of milk in cheese, and not carried out by cooperative internal workers. Therefore, the typical features of the PR-RE PDO sector require, for the generalization of the research results, to also take into consideration the environmental, social and contractual context that is observed in the production area, as suggested by other authors for the GI food [272,273,274,275,276];

(2)

Given the limitation referred to in Point (1), the research approach used for the PR-RE PDO can be used for the analysis of other food sectors characterized by the presence of cooperative firms, both in the case of GI foods and in other cases. From this point of view, however, it is necessary to observe that the legislation for the protection of GI products has typical elements in the European Union and, therefore, generalizations of the approach and conclusions must also consider differences in the legislation for the protection of GI products in extra-UE countries [277,278,279,280].

At the end, we can state that research has an applicative interest in disseminating the results to stakeholders and trade associations, so that firms can adopt virtuous corporate behavior. It also seems useful for COOPs to provide legal provisions that favor policies for reinvesting members’ profits and attracting capital from the market; on this point, legislative interventions to modify current regulations may be useful.

4. Conclusions

The research we conducted represents, to our knowledge, the first application of FRs and credit scoring to dairy firms operating in the PR-RE DOP production sector; this research is interesting because GI food products have a significant impact in terms of economic, social, and environmental sustainability and, however, these firms face financial and market difficulties. This research allows some conclusions:

• Overall, the firms in the PR-RE PDO sector are characterized by significant investments in fixed and working capital, as highlighted by the FRs; these are firms that therefore have barriers to capital entry. This research confirms the effect of production regulations on firm performance, as highlighted by the FRs. This result concerns all firms in the sector, COOPs and IOFs, and therefore confirms that (a) firms in the sector are capital intensive; (b) the PS has a significant effect in expanding the working capital cycle and, consequently, the barriers to entry in terms of capital. The usefulness of the cooperative legal form is therefore confirmed to overcome the constraints on access to the sector for small agricultural producers who would not be able individually to provide for the transformation of bovine milk into PR-RE PDO;
• Non-cooperative firms (IOFs) have better performances in terms of profitability and have a higher turnover; they have better capitalization in terms of equity compared to cooperatives (COOPs). It follows that the IOFs have a better credit score (EM-Score), and it is confirmed that the COOPs, even in the PR-RE PDO sector, are financially constrained. The differences found between IOFs and COOPs are statistically significant and it is therefore confirmed that the legal form has an impact on FRs and credit scoring. It emerges that FINSTATs of the COOPs have specific characteristics that seem to disfavor capitalization in terms of equity and the exposure of the income results in an income statement; this represents a problem in communication to investors, in credit scoring, and in the ability of these firms to attract capital from the market. The financial statement of these cooperatives implements the remuneration of the contribution of the members through refunds, governed by Article 2545-sexies of the Italian Civil Code; this research finding needs to be explored with further research. In fact, the refund is the result of a typical institution in the cooperation that allows you to grant to members a deferred mutual benefit through the distribution of a share of the profit generated by the active members themselves of the economic relations that occurred during the financial year between each member and the cooperative;
• The IOFs operating in the plains and hills have larger dimensions and better performance than the mountain IOFs, and this is reflected both in the FRs and in credit scoring; this result confirms the literature analyses, and also confirms the statistical data which have highlighted, in the PR-RE PDO production areas, a significant reduction in livestock activity in recent decades. However, the sample size for IOFs in the mountainous area is very small and this result must be considered with this limitation. The plain and hill COOPs have few differences compared to the mountain COOPs; the former have larger dimensions and better credit scoring, with a statistically significant difference.

The research has an applied interest, because it can be useful to managers and operators in the sector to have useful information on the management of firms in the PR-RE PDO sector and improve the performance of the firms and the financial and social sustainability of these firms, thus important in rural areas. Furthermore, the research can be useful to guide financial intermediaries, who will be able to build financial products useful to firms in the PR-RE PDO sector, and to other sectors of GI products characterized by financial barriers to entry, to guarantee financial sustainability and overcome financial constraint. Furthermore, the conclusions we have reached on the undercapitalization of COOPs can be useful in two aspects: (1) activate legislative instruments that favor the capitalization of COOPs, for example, through the possibility of dividing the assets during the liquidation of the COOPs and encouraging the circulation of quotas; (2) activate legislative instruments that allow improving the FINSTAT information of the COOPs so that this information can be considered in the credit scoring issued by financial intermediaries. These two measures should improve the financial sustainability of COOPs and, consequently, improve social sustainability in the whole PR-RE PDO sector.

In understanding and evaluating this research, however, some limitations must be considered: (1) the FINSTATs of all IOFs and COOPs firms in the PR-RE PDO sector are considered, but not all firms in the sector are in these legal forms and therefore are not required to file FINSTATs, as is the case of legal forms of sole proprietorship and partnership; consequently, the data of these firms are not considered; (2) the data in the FINSTATs of the firms do not contain all the details of the financial and operational management, for example, relating to the sales volumes for individual products and the related margins. Without prejudice toward these limitations, the conclusions we have reached can be considered based on sufficiently reliable and complete data, and therefore can be used to promote new research in other sectors of GI food products, in Italy and in other countries; and (3) the number of firms enrolled in the research does not allow us to draw general conclusions comparing COOPs and IOFs; the conclusions of the research can only refer to the PR-RE PDO sector, for which all the firms registered in the Consortium participated in the research and for which a series of ten-year financial statements is available. In fact, the number of firms in the sample is low compared to the total number of firms operating in the dairy sector in Italy, and this limitation needs to be considered in the generalization of the research conclusions; (4) firms that operate in the legal form of sole proprietorship or partnership are not included in the sample, because a public financial statement is not available for these firms, but all the data are confidential and reserved for tax purposes; therefore, these are not accessible at the tax level.

Despite these limitations, the research methodology that has been applied can be extended to other sectors of the Italian agri-food sector, in particular, if the simultaneous presence of COOP and IOF firms is relevant to compare firms’ style of governance and performance. The expansions of the research can consider both GI food productions with a high supply volume, as in the case of Grana Padano PDO cheese, and productions with a low supply volume. Research extensions may also consider cross-sector comparisons of food and non-food firms, and cross-national comparisons. At the end, the dissemination of the results to operators in this sector becomes useful for improving the strategic conduct of firms.