MODELING PUBLIC COLLABORATIVE PROCESSES: THE CASE OF SAFEGUARDING FINANCIAL STABILITY
Peter Sarlin1 Goethe University Frankfurt Henrik J. Nyman Åbo Akademi University
Abstract This paper introduces the notion of public collaborative processes (PCPs) to the field of Business Process Management (BPM). PCPs involve multiple organizations with a common objective, where a number of dispersed organizations cooperating under various unstructured forms take a collaborative approach to reaching the final goal. This paper exemplifies the analysis of PCPs with the case of safeguarding financial stability, particularly macroprudential oversight in Europe. Following the literature on BPM, this paper models the macroprudential oversight process as an event-driven process chain, including activities, responsible entities, as well as inputs and outputs. We view the process from two directions. As an initial step, we document a high-level representation of the process (i.e., without specifying responsible entities) as depicted in the literature. Moreover, we show the increase in complexity when moving to a low-level description of the process (i.e., specifying responsible entities). Along these lines, we put forward a matrix view consisting of five necessary steps to managing PCPs. Thus, in terms of managing the macroprudential oversight process, this motivates further work in collaboration with experts to document, analyze and improve the macroprudential oversight process at a more detailed level. Keywords: Business Process Management, Public collaborative processes, Macroprudential oversight, Event-driven process chain
1
Corresponding author: Peter Sarlin, Center of Excellence SAFE at Goethe University and RiskLab at IAMSR, Åbo Akademi University and Arcada University of Applied Sciences. Postal address: Goethe University, SAFE, Grüneburgplatz 1, 60323 Frankfurt am Main, Germany. Email:
[email protected]. Tel: +358 40 5727670.
1
1
Introduction
Business process management (BPM) concerns the analysis and improvement of current processes through systematic representations of organizations' processes (e.g., Segatto et al., 2013). BPM provides means not only for representing current processes, but also for reaching consensus on what form the processes should take and possible areas of improvement (e.g., Hammer, 2010; Eikebrokk et al., 2011). Yet, this describes the functioning and optimization of processes from the viewpoint of a single organization. In the public sector, processes may be of a different nature. They may involve multiple organizations with a common objective, where the actors take a collaborative approach to reaching the final goal. We denote these as public collaborative processes (PCPs). Beyond the common setting in BPM that oftentimes also involves interaction among multiple organizations, such as the case of supply chains, PCPs involve a number of dispersed organizations cooperating under various unstructured forms. For instance, public health at the national level may involve activities ranging from school gymnastics to nutritional health advice to services of health care centers. While being of different nature, PCPs can still be examined through the lens of BPM. This paper exemplifies PCPs through the case of safeguarding financial stability. The current financial crisis has highlighted the importance of a system-wide, or macroprudential, approach to safeguarding financial stability, rather than one being only concerned with the stability of individual financial institutions (Borio, 2011). Macroprudential oversight can be viewed as a process, which is characterized by vast amounts of data, inherent complexities related to temporal and crosssectional dimensions, decision-making based upon numerical methods and expert judgment, and a few central parties orchestrating an interconnected system of actors. As such, the characteristics of the macroprudential oversight process are complex. In the case of macroprudential oversight in Europe, the system of financial supervisors include a large number of actors both at the European and nationallevel, with a common goal of safeguarding financial stability (EU Commission, 2009a; 2009b). As the structure of the European system of financial supervisors is currently in the making, now is the time to disentangle activities at the level of responsible entities and understand how the actors interact (as initiated by Hartmann (2013)). In this vein, this paper applies a process view to analyze macroprudential oversight in Europe. The macroprudential oversight process, as depicted by ECB (2010), comprises the following highlevel tasks: risk identification, risk assessment and risk communication, as well as the assessment and implementation of policies. While being relatively well-defined at this level, the tasks still involve a number of uncertainties and limitations that challenge the functioning of the process. In particular, the lack of accurate, complete, frequent and timely data, challenges in moving from systemic risk measurement to policy implementations, and the high degree of dependence between and within the tasks in the process. In particular, understanding dependence within the tasks becomes crucial when moving towards a more detailed low-level representation of the process. Given the aim of coordinating activities, so that disperse processes performed by many different partners act as one seamless process, concerns have been raised about the functioning of the European macroprudential oversight process, particularly the interplay among actors at the national and European levels (e.g., De Grauwe, 2013). Documenting and analyzing the macroprudential oversight process on a detailed level, supports involved entities working together to achieve a mutual understanding of what the process entails, a shared goal, and a clear division of work. This paper documents the macroprudential oversight process as an Event-driven Process Chain (EPC), including activities, inputs and outputs. There is a need to view the process from two perspectives. As an initial step, we document a high-level representation of the process (i.e., without specifying responsible entities) as depicted in the literature. Then, we show the increase in complexity when moving to a low-level description of the process (i.e., specifying responsible entities). While illustrating overall challenges in managing PCPs, this motivates further work in collaboration with
2
experts to document, analyze and improve the macroprudential oversight process at a more detailed level. The paper is structured as follows. First, we briefly discuss BPM and PCPs. Then, we present the case of safeguarding financial stability as a PCP. Finally, before concluding, we discuss the overall implications of PCPs, provide a brief guideline for analyzing PCPs and put forward an agenda for future research on low-level modeling of the macroprudential oversight process.
2
Process modeling
This section discusses the main concepts we draw upon in this paper, starting out with BPM. Following this, we examine the particularities of PCPs from a process modeling perspective.
2.1
Business Process Management
Broadly speaking, an organizational process can be defined as a chain of activities that make use of an input to produce an output. This implies that any organization, public or private, has processes. The raison d'être for documenting and understanding organizational processes on the other hand goes back to performance. This can imply that there is a desire for improvement, or a desire to overcome shortcomings in performance. Absent any common understanding of a process, it becomes difficult or impossible to meet or exceed any performance targets. While a common understanding of how a process works (and how the activities therein interact) can be implicitly understood by those working on the task at hand, more complex processes that span across organizational entities create a need for more formalized communication. As such, a more structured approach to gain a common understanding of how a process works, and what the associated performance targets are, is needed. To define a mature, high-performance process, one can assess process-specific characteristics. Hammer (2010) defines five critical enablers for well-functioning processes: (i) process design, (ii) process metrics, (iii) process performers, (iv) process infrastructure, and (v) process owners. The design of the process goes back to the specification of what tasks are to be performed, by whom, and what the associated inputs and outputs are. This also involves an understanding of the organization(s) involved and how the activities interact to produce a desired outcome. The metrics of the process detail the desired outcome in more specific terms. Performance needs to be monitored against targets such as speed or quality. With process design and metrics at hand, process performance can be simulated (in relation to, for example, time, cost, and resources needed), allowing for a better understanding of potential improvement areas. This approach has been taken in many different contexts, such as hospitals and banks (Islam and Ahmed, 2012; Shim and Kumar, 2010). The performers of the process are those that perform the process activities. The persons involved should not only be well versed in their particular activity or part of the process, but also be aware of the endto-end design and metrics of the whole process. In addition, understanding who should be involved in process improvement efforts is crucial. For instance, a panel of experts, consisting of those responsible for running the process, as well as academics and a representative of a supervisory body, was put in place to better understand the processes at a surgical ward in Italy (Bertolini et al., 2011). This was the starting point for identifying several process bottlenecks, to be later addressed by the process performers. A suitable infrastructure also needs to be in place for the process performers to be able to conduct their work. Typically, this involves IT systems with the appropriate access to information they need in order to accomplish their task. For instance, a process improvement effort at a Swiss bank revealed many improvement areas that were specifically related to the IT systems that were used to support the process (Küng and Hagen, 2007). Last but not least, a process owner is to be nominated. This is an executive level person that has the authority to oversee the process as a whole. Without a process owner, improvements and changes that span across different involved organizations becomes difficult or impossible.
3
Process design can perhaps be said to be the most central element of a process. This is also referred to as process modeling, whereby an external representation of the process is documented as a process model (Eikebrokk et al., 2011). This can be done on two different levels, as an as-is representation of the current status or a to-be representation that details a desired future state incorporating necessary improvements. It should be noted that process modeling is a collaborative effort (often between outside expertise in process modeling, the process owner, and those performing the process) that makes use of a standardized and jointly agreed method for documenting the process, with the aim of creating transparency. Everyone involved, from the performers to the process owner, should have a detailed understanding of what the process entails. To go beyond mere process modeling, business process management (BPM) takes a more holistic approach to managing an organization's business processes. Segatto et al. (2013) define BPM as a discipline focusing on gaining a common understanding of processes with the aim of continuously seeking improvement through a feedback cycle, while at the same time also aligning these processes to organizational strategies. They divide BPM to six distinct phases (adapted from ABPMP, 2009): 1. Planning: In this initial stage, executive sponsors, roles and responsibilities, goals and overall purpose of the BPM exercise are determined. 2. Analysis: This involves understanding of the current state of organizational processes. 3. Design and modeling: At this stage, a more detailed representation of as-is and/or to-be processes are created. Essentially, this phase focuses on gaining answers to questions such as what is done, by whom, when, how, and in which organizational entity. 4. Implementation: Here, activities in the organization are, if necessary, adapted to findings from previous stages. 5. Monitoring and control: Here, performance metrics are analyzed to see whether these give raise to further changes in the organization. 6. Refinement: After an analysis of process performance, further work in the analysis and design phases of the BPM cycle might be necessary. With an understanding of processes overall and BPM in particular, we can move forward to discussing PCPs.
2.2
Public collaborative processes
Societies are often organized around particular goals regarding the well-being of their citizens. These goals involve a disperse set of actors, often mostly comprising public institutions, but possibly also private institutions, all acting together for a common purpose. Examples of these societal goals include public safety, education and health, as well as a well-functioning financial system. Seen from an individual’s perspective, the actors involved in creating the necessary conditions for achieving these goals are often tied to various stages of a person’s life. Oftentimes, we also see actors taking on roles that are typically not associated with their primary goal. One such example could be schools educating children in traffic rules and fire hazards, and thus promoting public safety. The police might be offering lectures on drug abuse, and thus promoting public health. Thus, public institutions work together to build a well-functioning society, oftentimes taking on different roles depending on an individual’s age and other particular needs. Sickness, criminal behavior, age, disabilities, and personal interests are all relevant examples of conditions that can trigger a particular response from society. In some instances, the response is highly regulated and clearly defined, such as in the case of criminal misconduct, but oftentimes there is a high degree of informal co-operation and implicit understanding of how to promote a particular goal. The case of public health is such an example. If we disregard the most obvious example of an actor in this field, hospitals, which are
4
usually involved when a de facto health problem already exists, we see a host of actors involved in preventive measures to avoid health issues. As such, many different actors are then involved in the process of ensuring public health. This is highly dependent on specific structures in various countries, but it is not uncommon to see a diverse set of actors in fields like education, food safety, and sports promoting health, in addition to the explicit healthcare system. We denote the activities performed by various societal actors working together for a common purpose as a public collaborative process (PCP). This is defined as a set of public, private and volunteering actors, working together for a societal goal which is defined and possibly regulated on a governmental level. The complexity of these types of processes becomes apparent when one moves from a highlevel description to a lower level, where responsibilities and activities performed by individual organizations are disentangled. The high-level goals for PCPs are typically documented, and there is often a primary organizational entity that assumes the overall responsibility for the process. However, at a lower level, the activities performed by a dispersed set of secondary actors are oftentimes dependent on either informal agreements or personal contacts. Mostly, these activities are also coordinated on a high level, such as by a responsible ministry, in practice creating a gap between the operative and governing actors. From the above described example of public health, we move over to the an illustrative application to ensuring a well-functioning financial system.
3
The PCP of safeguarding financial stability
This section focuses on the analysis of safeguarding financial stability as a PCP. We first describe the system-wide macroprudential approach to oversight and supervision of the financial system, as well as present the macroprudential oversight process as described in the literature. Based upon the literature, we document through BPM a high-level representation of the macroprudential oversight process (i.e., without specifying responsible entities). Finally, we show the increase in complexity of a move to a low-level description of the process (i.e., specifying responsible entities) by providing a description of the European system of financial supervisors. Rather than providing adequate means for the analysis and improvement of current processes, the exercise functions as an illustration of PCPs' complexity, particularly in the case of financial stability.
3.1
Macroprudential oversight
Paraphrasing Milton Friedman's statement about Keynesians, Borio (2011) stated “We are all macroprudentialists now.” Since the date when the still ongoing global financial crisis broke out, the notion of a macroprudential approach to safeguarding financial stability has grown consensus among the academic and policymaking communities alike. Yet, it is no new concept. The Bank for International Settlements (BIS) applied the term to describe a system-wide orientation of regulatory frameworks already in the 1970s (see, e.g., Borio, 2011). The series of recently established macroprudential supervisory bodies obviously also motivates understanding and disentangling their specific tasks and functions. A macroprudential supervisory body is an institution tasked with macroprudential oversight of the financial system and the mandate of safeguarding financial stability. Examples are the European Systemic Risk Board in Europe, the Financial Policy Committee in the UK, and the Financial Stability Oversight Council in the US. A comprehensive macroprudential approach to safeguarding financial stability obviously starts from a thorough understanding of the inner (dys)functioning of the financial system. While a wide range of topics in the literature remain to be disputed, one notion that few oppose is that a key aim is to have a resilient and well-functioning financial system. One characterization of such a financial system is through the following three pillars (Fell and Schinasi, 2005): well-managed financial institutions, efficiently functioning financial markets and a strong and robust financial infrastructure. That said, the frequent incidences of costly financial crises do, however, indicate that the three pillars of have
5
defects. While each recurrence of financial instability may have sources of its own kind, market imperfections like asymmetric and incomplete information, externalities and public-good characteristics and incomplete markets are threaten the functioning of the financial system. These imperfections, when being related to a financial sector, may lead to significant fragility of not only individual institutions, but also the entire system (see, e.g., Carletti, 2008). De Bandt and Hartmann (2002) relate fragilities in financial systems to three causes: (i) the structure of banks, (ii) the interconnection of financial intermediaries, and (iii) the information intensity of financial contracts. The material risks of these fragilities support the role of governments and other supervisory authorities in addressing and monitoring financial instability, which also makes financial stability into a common good and systemic risk into an externality. To concretize the notion of systemic risk, we follow the definition of three forms by de Bandt et al. (2009): (i) endogenous build-up and unraveling of widespread imbalances; (ii) exogenous aggregate shocks; and (iii) contagion and spillover. The first form of systemic risk focuses on the unraveling of widespread imbalances and is illustrated by a thorough literature on the presence of risks, vulnerabilities and imbalances in banking systems and the overall macro-financial environment prior to historical financial crises. Early and later literature alike have identified common patterns in underlying vulnerabilities preceding financial crises (see, e.g., Minsky, 1982; Reinhart and Rogoff, 2008). The second type of systemic risk, exogenous aggregate shocks, have been shown to co-occur with financial instabilities (see, e.g., Gorton, 1988; Demirgüç-Kunt and Detragiache, 1998). One example is the collapse of banks during recessions due to the vulnerability to economic downturns. The contagion literature provides evidence on the final, third form of systemic risk, which involves the cross-sectional transmission of financial instability (see, e.g., Upper and Worms, 2004; van Lelyveld and Liedorp, 2006). Here, episodes of financial instabilities have been shown to relate to the failure of one financial intermediary causing the failure of another. For macroprudential oversight, policymakers and supervisors need to have access to a broad toolbox of approaches to measure and analyze system-wide threats to financial stability. Broadly speaking, tools and models can be divided into those for early identification and assessment of systemic risks. ECB (2010) provides a mapping of tools to the above listed three forms of systemic risk: (i) earlywarning models, (ii) macro stress-testing models, and (iii) contagion models. First, by focusing on the presence of vulnerabilities and imbalances in an economy, early-warning models can be used to derive probabilities of the occurrence of systemic financial crises in the future (see, e.g., Alessi and Detken, 2011; Betz et al., 2013). Second, macro stress-testing models provide means to assess the resilience of the financial system to a wide variety of aggregate shocks, such as economic downturns (see, e.g., Castrén et al., 2009; Hirtle et al., 2009). Third, contagion and spillover models can be employed to assess how resilient the financial system is to cross-sectional transmission of financial instability (see, e.g., IMF, 2009). In addition, the literature has also provided a large set of coincident indicators to measure the contemporaneous level of systemic risk (see, e.g., Holló et al., 2012). While coincident measures may be used to identify, signal and report on heightened stress, they are not designed for early identification and assessment of risk. The above described market imperfections, and thereby caused systemic risks, are a premise for macroprudential oversight. Accordingly, the above three analytical approaches aim at signaling these systemic risks at an early stage. In terms of a process, Figure 1 puts forward the steps in the process that a macroprudential supervisory body follows. As described in ECB (2010) and author (reference omitted for blind review), macropudential oversight can be related to three steps: (i) risk identification, (ii) risk assessment, and (iii) policy assessment and implementation, as well as giving risk warnings and policy recommendations. The process in Figure 1 deviates from ECB (2010) by disentangling the final step into two separate feedback loops, as proposed by author (reference omitted for blind review). In the figure, red components represent risks and vulnerabilities, green components represent the need for risk identification and assessment, blue components represent risk warnings and policy
6
recommendations, and gray components represent policy assessment, implementation and follow-up. With no detailed treatment, we present herein the key tasks and tools used in each step. Impacts of internal and ext ernal ris k co mmunica tion Risk communicatio n Risk identification
Po tentia l sources o f systemic risk
Yes Vulnera -bility
Yes Risk ass ess ment
No
Risk warnings
Mat erial r is k
No P olicy assessment
Market imp erfe ction s behind systemic risk
Policy recommendation
Policy implementation
Impacts of policy and follo w-up reco mmenda tions
Notes: The figure illustrates the macroprudential oversight process. The red components represent risks and vulnerabilities and the green components represent the need for risk identification and assessment, blue components represent risk warnings and policy recommendations, and gray components represent policy assessment, implementation and follow-up. The figure is an adapted version of that in ECB (2010) and author (reference omitted for blind review).
Figure 1.
The macroprudential oversight process.
In the first step of the supervisory process, the key focus is on identifying risks to stability and potential sources of vulnerability. The vulnerabilities and risks could exist in any of the three pillars of the financial system: financial intermediaries, financial markets and financial infrastructure. The necessary analytical tools to identify possible risks, vulnerabilities and triggers come from the set of early-warning models and indicators, combined with the use of market intelligence, and expert judgment and experience. This involves ranking risks and vulnerabilities as per intensity, as well as for assigning probabilities to specific shocks or future systemic events. In the second step of the process, the rankings and probabilities may be used to assess the identified risks. Beyond market intelligence, as well as expert judgment and experience, risk assessment makes use of analytical tools mainly from the set of macro stress-testing models and contagion models. In macro stress-testing, simulations of most plausible risk scenarios show the degree of impact severity on the overall financial system, as well as its components. Contagion models, on the other hand, might be used through counterfactual simulations to assess the impact of specific failures on the entire financial system and individual institutions. The first and the second step of the process should not only provide a list of risks ordered according to possible severity, but also contain their materialization probabilities, losses given their materialization, and real losses in output and welfare, as well as their possible systemic impact. Hence, these two initial steps in the process aim at early risk identification and assessment and steer subsequent actions for safeguarding financial stability. The third step of the process involves the assessment, recommendation and implementation of policy actions as early preventive measures, as well as the communication of risks and vulnerabilities. Based upon the identified and assessed risks, a macroprudential supervisory body can consider giving a wide variety of risk warnings and recommendations for other parties to use policy instruments, as well as implementations of policies given the instruments at hand. To steer their decisions, the policy assessment step can make use of the same analytical tools used for risk identification and assessment. Likewise, risk warnings and policy recommendations can make use of the analytical tools for better
7
communication. While the use of policy tools is oftentimes beyond the mandate of a macroprudential supervisory body, actions tailored to the needs of a system-wide orientation are obviously a key part of macroprudential regulation and supervision. As illustrated in Figure 1, policies have an impact, not only on the assessment of policy and identification and assessment of risks, but obviously also directly on market imperfections and the accumulation of systemic risks. The feedback of risk warnings and policy recommendation can, likewise, be divided into those that affect the risk identification and assessment and those affecting the financial market at large, where the former can be targeted with internal communication and the latter with external communication.
3.2
Documenting the macroprudential oversight process
While there are many standards for documenting process (see, e.g., Ko et al., 2009 for an overview), we have chosen to use the Event-driven Process Chain (EPC) method. For our purposes, a graphical standard is essential. Any process documentation method requires some insights to the semantics and rules for the method in question, but graphical standards such as EPC are typically easier to read and thus most natural for humans (as opposed to standards that can, for example, be interpreted by a computer). EPC is an effective method to convey a general overview of a process and contains a minimal amount of different symbols that need to be conveyed to the reader. It should be noted, however, that research has shown that the most effective way to depict a process is through the use of both a graphical representation, and a text-based case description (Ottensooser et al., 2012). While this is relevant for communicating the details of a process, the aim of the illustrative application in this paper is not limited by only relying on a graphical representation. In its simplest form, as originally proposed by Keller et al. (1992), an EPC contains three elements: functions, events, and connectors. A function corresponds to an activity, whereas an event depicts the conditions before and after the function. As such, a function is typically preceded and followed by an event. The crucial differences between these lie in their fundamental nature. Functions consume time and resources, and can make decisions, whereas events have neither of these attributes. Events represent a status, and can be divided to four types (Rosemann, 2003): (i) an event to depict the starting point or final status of a process, (ii) an event to depict an update to an object (that has been processed in a function), (iii) an event to depict a certain point in time, or (iv) an event to depict a change in status that triggers a function. Connectors allow for alternative courses of actions to be selected in a function. AND-connectors model parallel execution. An incoming arc is split to two or more outgoing arcs that are all performed simultaneously. XOR-connectors depict alternative outgoing arcs, and OR-connectors model one or more alternative parallel arcs. Criticism regarding the semantics of these different connectors has been put forward, mainly related to the flow of events and functions after a connector. A connector, much like an event, does not have decision power. While an EPC can neither depict the timing of the flows after the connector nor the timing of activities (Gruhn and Laue, 2007), we do not see this as a concern in providing a high-level representation of the macroprudential oversight process.
8
Figure 2.
A documented macroprudential oversight process.
9
The sequel of this subsection documents the above described macroprudential oversight process. We make use of the EPC method for documenting the process, and entirely rely on the description in the literature. Thus, the documentation in Figure 2 aims at describing the macroprudential oversight process, with the common goal of financial stability. While key elements of Figure 1 can be observed in the EPC documentation, we do not focus on impacts, feedback and learning, which does not involve explicit forms of functions and events. The description in Figure 2 makes one crucial assumption: as if safeguarding of financial stability would be performed by one single organization. This neglects the fact that it, as a PCP, contains multiple organizations with the same objective, where the actors take a collaborative approach to reaching the goal. In the case of financial stability, the organizations are oftentimes dispersed and cooperate under unstructured forms. Next, we will view the case of the European system of financial supervisors, taking safeguarding financial stability to a low-level representation.
3.3
European system of financial supervisors
To understand the complexity of macroprudential oversight in the European case, we need to focus on the entire system of financial supervisors in Europe. In this section, we briefly discuss the actors and their roles in the European system of financial supervisors. Without including strictly political institutions, the actors involved in financial supervision, micro- and macroprudential, are the following: European Central Bank (ECB), European Systemic Risk Board (ESRB), national central banks (NCBs), European Commission (EC), European Banking Authority (EBA), European Insurance & Occupational Pensions Authority (EIOPA), European Securities & Market Authority (ESMA), national banking supervisors, national insurance supervisors and national securities supervisors. While their common aim is safeguarding financial stability at various levels, it is needless to say that disentangling activities and functions, as well as their connectors, at the level of responsible entities substantially increases complexity. European System of Financial Supervisors
Microprudential supervisors
European Supervisory Authorities
European Banking Authority
European Insurance & Occupational Pensions Authority
Macroprudential supervisors
National supervisory authorities
European Securities & Market Authority
National banking supervisors
National insurance supervisors
ECB President and Vice President
European Systemic Risk Board
European Central Bank
National securities supervisors
European Supervisory Authorities
Chair of Economic & Financial Committee
29 NCB Governors
European Commission
3 European Supervisory Authorities chairs
National central banks
Advisory Technical Committee Chair
Advisory Scientific Committee Chair and 2 Vice Chairs
Notes: The figure is an adapted version of that in Hartmann (2013) and describes the structure depicted by the EU Commission (2009a; 2009b).
Figure 3.
The European system of financial supervisors.
Following the relatively vague description by the EU Commission (2009a; 2009b), as well as the illustration in Hartmann (2013), we can move towards an understanding of the overall division of aims and focus areas. First, Figure 3 shows that supervision is divided into micro and macroprudential. Second, within each of these two branches, the actors consist of both national and European institutions. Third, there exists a large share of interaction and collaboration both between and within these two groups of supervisors. For instance, while microprudential supervisory bodies do not have macroprudential oversight as their key aim, they are oftentimes the ones implementing policies according to recommendations by macroprudential supervisory bodies. Likewise, microprudential information is an important input to analysis of system-wide risks by macroprudential supervisors.
10
Yet, while this subsection functions as an illustration of the complexity of macroprudential oversight, it is worth to note that this provides no detailed description of tasks, interaction and collaboration. For a further understanding of the low-level process, we should make use of operative and governing experts from a broad range of organizations.
4
Discussion and future work
Stepping back from the case of safeguarding financial stability, we can generally review what PCPs are, what their implications are on BPM and how they should be managed. PCPs are processes occurring in the public domain, which involve multiple organizations with one common objective and where the actors take a collaborative approach to reaching the final goal. Hence, PCPs mostly involve dispersed organizations cooperating under unstructured forms. While being of different nature than most processes managed through BPM, PCPs can still make use of concepts and approaches common in BPM, such as graphical representations provided by EPC. Without delving into the details of performing each step in BPM, we propose PCPs to be managed through the elements of the matrix in Table 1, where rows represent the abstraction level of the process model and columns represent the source of the process model and its temporal perspective (current and the desired future state). The elements of the matrix include the person that provides an input to the process model and the order in which the elements are to be performed. The first step involves a process analyst describing the current state of the high-level process as described in the literature. Second, in collaboration with governing experts, the literature-based high-level process model is to be adjusted. Third, the same group of governing experts provides input on the desired future state of the high-level process. Fourth, when moving to documenting a low-level process, a group of operative experts provide additional input to document the responsible actors on an organizational level. It should be noted that the literature-based low-level process is lacking due to the fact that the literature seldom provides information on tasks at a detailed level. Finally, the fifth step involves the same group of operative experts to provide input on the desired future state of the lowlevel process. With this five-step procedure at hand, the EPC notation is highly suitable for the analysis of PCPs in that it provides a graphical representation that enables views on multiple abstraction levels.
High-level Low-level
Literature (as-is) 1. Process analyst
Current (as-is) 2. Governing experts 4. Operative experts
Proposed (to-be) 3. Governing experts 5. Operative experts
Notes: Rows represent the abstraction level of the process model, whereas columns represent the source of the process model and its temporal perspective (status quo and the desired future state).
Table 1.
A matrix of BPM in the context of PCPs.
In this paper, we have concluded that the macroprudential oversight process follows our outlined definition of a PCP. By definition, it involves a dispersed set of actors, and in its current form relies on implicit and explicit agreements regarding working procedures and roles and responsibilities. At the same time, there are indications that there is room for improvement (e.g., De Grauwe, 2013). As such, we examine the macroprudential oversight process with techniques familiar from BPM. In doing this, our goal is to achieve a process that follows Hammer’s (2010) characterization of a mature, wellfunctioning process (see Section 3). Our ultimate aim, which goes beyond the focus in this paper, is to go through the whole BPM cycle (as depicted by, e.g., Segatto et al. (2013)). From the viewpoint of macroprudential oversight, our overall objective is hence twofold: (i) to validate the macroprudential oversight process with all relevant stakeholders (as-is), (ii) and to suggest possible improvements to it (to-be). Although the main aim of this paper is the notion of PCPs, we take a number of steps in order to fulfill the two above proposed objectives. Hence, this paper only
11
represents the first steps in reaching the overall objectives. We provide a review of BPM practices, the nature of PCPs, and the macroprudential oversight process in this setting. Based upon the literature, we make use of EPC to document a high-level representation of the macroprudential oversight process, as well as describe the complexity in moving to a low-level description of the process. Following this, our future intention is to review the process together with experts in the field, including policymakers and analysts at relevant regulatory bodies, to gain insights to potential gaps and possible solutions. Hence, as is depicted in Table 1, the process documentation focuses on three different perspectives: (i) one based upon the literature, (ii) one based upon experts' opinions, and (iii) one based upon experts' suggestions for future improvements. Beyond the representation herein, we adopt a stance where the process is documented on a lower level. While the high-level process description focuses on the tasks and goals in macroprudential oversight without specifying responsible actors, the low-level process description attempts to document the responsible actors on an organizational level. Following the complexities inherent to PCPs, we expect this task to be challenging. Relating to this, many of the process enablers outlined by Hammer (2010) are yet to be clearly defined. Our priority is on a more detailed design of the process, including the identification of all the involved entities (performers). This will serve as a starting point to put together a board of experts (similar to Bertolini et al., 2011) to define the metrics of the process (if at all possible) and assign the process owner or owners (again, if possible). Any form of process simulation and discussion of infrastructure (e.g., IT systems and data) is also a subject of future research. More broadly, this paper sets merely a starting point not only for the case of modeling the macroprudential oversight process, but also for a broad range of applications to PCPs in general.
5
Conclusions
This paper has introduced the notion of PCPs (public collaborative processes). These processes occur in the public domain, involve multiple organizations with a common objective, and involve a number of dispersed organizations collaborating under various unstructured forms to reach a final goal. While the nature of PCPs differs from more standard processes, PCPs can still make use of concepts and approaches common in BPM. The analysis of PCPs is in this paper exemplified with the case of safeguarding financial stability, particularly macroprudential oversight in Europe. Following the literature on BPM, this paper models the macroprudential oversight process as an event-driven process chain, including activities, responsible entities, as well as inputs and outputs. We view the process from two directions. As an initial step, we document a high-level representation of the process (i.e., without specifying responsible entities) as depicted in the literature. Moreover, we show the increase in complexity when moving to a low-level description of the process (i.e., specifying responsible entities). Along these lines, we have put forward a matrix view consisting of five necessary steps to managing PCPs. Thus, in terms of managing the macroprudential oversight process, this motivates further work in collaboration with experts to document, analyze and improve the macroprudential oversight process at a more detailed level.
References ABPMP (2009). Business process management BPM common body of knowledge (BPM CBOK), version 2.0., 2nd release, Association of Business Process Management Professionals, Chicago, IL. Alessi, L. and Detken, C. (2011). Quasi real time early warning indicators for costly asset price boom/bust cycles: A role for global liquidity. European Journal of Political Economy, 27(3), 520533. Bertolini, M., Bevilacqua, M., Ciarapica, F. and Giacchetta, G. (2011). Business process reengineering in healthcare management: a case study. Business Process Management Journal, 17(1), 42-66.
12
Betz, F., Oprica, S., Peltonen, T. and Sarlin, P. (2013). Predicting distress in European banks. ECB Working Paper No. 1597.. Borio, C. (2011). Implementing a macroprudential framework: Blending boldness and realism. Capitalism and Society, 6(1). Carletti, E. (2008). Competition and regulation in banking. In Handbook in Financial Intermediation (Boot, A. and Thakor, A. Eds.). Elsevier, North Holland, pp. 449-482. Castren, O., Fitzpatrick, T. and Sydow, M. (2009). Assessing portfolio credit risk changes in a sample of EU large and complex banking groups in reaction to macroeconomic shocks, ECB Working Paper No. 1002. de Bandt, O. and Hartmann, P. (2002). Systemic risk in banking: A survey. In Financial crisis, contagion and the lender of last resort: A book of readings (Goodhart, C. and Illing, G. Eds.). Oxford University Press, Oxford. de Bandt, O., Hartmann, P. and Peydro, J. (2009). Systemic risk in banking: An update. In Oxford Handbook of Banking (Berger, A., Molyneux, P. and Wilson, J. Eds.). Oxford University Press, Oxford. de Grauwe, P. (2013). Design failures in the eurozone: Can they be fixed?, EQS Paper No. 57/2013. Demirgüc-Kunt, A. and Detragiache, E. (1998). The determinants of banking crises in developing and developed countries. IMF Staff Papers, 45(1), 81-109. ECB (2010). Analytical models and tools for the identification and assessment of systemic risks. In Financial Stability Review (June 2010). European Central Bank, Frankfurt, Germany. Eikebrokk, T., Iden, J., Olsen, D. and Opdahl, A. (2011). Understanding the determinants of business process modelling in organisations. Business Process Management Journal, 17(4), 639-662. EU Commission (2009a). Proposal for a regulation of the European Parliament and the Council on Community macro-prudential oversight of the financial system and establishing a European Systemic Risk Board, September 23, 2009. EU Commission (2009b). Proposal for a Council Decision entrusting the European Central Bank with specific tasks concerning the functioning of the European Systemic Risk Board. September 23, 2009. Fell, J. and Schinasi, G. (2005). Assessing financial stability: Exploring the boundaries of analysis. National Institute Economic Review, 192, 102-117. Gorton, G. (1988). Banking panics and business cycles. Oxford Economic Papers, 40(4), 751-781. Gruhn, V. and Laue, R. (2007). What business process modelers can learn from programmers. Science of Computer Programming, 65(1), 4-13. Hammer, M. (2010). What is business process management? In Handbook on business process management 1 (vom Brocke, J. and Rosemann, M. Eds.). Springer-Verlag, Heidelberg. Hartmann, P. (2013). Fitting Macroprudential into the Picture: Models, Institutions and Instruments. Presentation at the Banco de España/World Bank policy conference on “Central Bank (R)Evolutions” in Madrid, Spain on June 17th, 2013. Hirtle, B., Schuermann, T. and Stiroh, K. (2009). Macroprudential supervision of financial institutions: Lessons from the scap, Staff Report No. 409, Federal Reserve Bank of New York. Holló, D., Kremer, M. and Lo Duca, M. (2012). CISS - a composite indicator of systemic stress in the financial system. ECB Working Paper No. 1426. IMF (2009). Assessing the systemic implications of financial linkages. Global Financial Stability Report April, International Monetary Fund, Washington, 73-110. Islam, S. and Ahmed, M. (2012). Business process improvement of credit card department: case study of a multinational bank. Business Process Management Journal, 18(2), 284-303. Keller, G., Nüttgens, M. and Scheer, A. (1992). Semantische processmodellierung auf der grundlage ereignisgesteuerter processketten (epk), veröffentlichungen des Instituts für Wirtschaftsinformatik, Heft 89 (in German), University of Saarland, Saarbrücken. Küng, P. and Hagen, C. (2007). The fruits of business process management: an experience report from a swiss bank. Business Process Management Journal, 13(4), 477-487.
13
Ko, R., Lee, S. and Lee, E. (2009). Business process management (bpm) standards: a survey. Business Process Management Journal, 15(5), 744-791. Minsky, H. (1982). Can "it" Happen Again?: Essays on Instability and Finance. M.E. Sharpe, Armonk, N.Y. Ottensooser, A., Fekete, A., Reijers, H., Mendling, J. and Menictas, C. (2012). Making sense of business process descriptions: An experimental comparison of graphical and textual notations. Journal of Systems and Software, 85(3), 596-606. Reinhart, C., Rogoff, K. (2008). Is the 2007 us sub-prime financial crisis so different? An international historical comparison. American Economic Review, 98(2), 339-344. Rosemann, M. (2003). Preparation of process modeling. In Process management: a guide for the design of business processes (Becker, J., Kugeler, M. and Rosemann, M. Eds.). Springer-Verlag, Heidelberg. Segatto, M., de P_adua, S. and Martinelli, D. (2013). Business process management: a systemic approach? Business Process Management Journal, 19(4), 698-714. Shim, S. and Kumar, A. (2010). Simulation for emergency care process reengineering in hospitals. Business Process Management Journal, 16(5), 795-805. Upper, C. and Worms, A. (2004). Estimating bilateral exposures in the German interbank market: Is there a danger of contagion? European Economic Review, 48(4), 827-849. van Lelyveld, I. and Liedorp, F. (2006). Interbank contagion in the Dutch banking sector: A sensitivity analysis. International Journal of Central Banking, 2(2).
14
