Identification and Prioritization of Banking Risks to Inform Liquidity and Credit Stress Testing: A Mixed-Methods Approach
Abstract
To identify and prioritize banking risks that may inform liquidity and credit stress-test development in banks listed on the Tehran Stock Exchange, and to examine their perceived interrelationships using expert judgments. A sequential exploratory mixed-methods design was adopted within a pragmatic paradigm. Semi-structured interviews with 20 banking experts informed qualitative coding in NVivo; Cohen’s kappa was 0.837. The quantitative phase covered 10 listed banks and 52 banking managers, while the direct-influence matrix represented mean ratings from 15 experts. DEMATEL-based analytic network process (DANP) calculations were performed in Excel. Eight risk categories were identified. Interest rate risk received the highest final weight (0.1634), followed by credit risk (0.1426) and legal and regulatory risk (0.1274). The remaining weights were 0.1151 for liquidity risk, 0.1146 for business risk, 0.1142 for financial risk, 0.1139 for market risk, and 0.1088 for foreign exchange risk. The reported network statistics and final weights describe expert assessments rather than estimated probabilities of bank failure. The analysis identifies interest rate, credit, and legal and regulatory risks as the three highest priorities within an expert-derived network. It provides an exploratory input to stress-test development, rather than a calibrated or validated stress-testing model. Losses, liquidity shortfalls, and resilience under adverse scenarios were not estimated.
Introduction
Banks mobilize savings and direct financial resources toward investment, production, and commercial exchange. Their ability to perform this intermediary role depends on managing exposures that arise from lending, funding, investment, and payment activities. Credit and liquidity risks are particularly important because repayment problems can weaken expected cash inflows, while funding pressures can constrain a bank’s capacity to meet obligations. Banking risk management therefore requires attention to the quality of individual exposures and to the interaction of risks across the balance sheet. The Basel Committee on Banking Supervision (BCBS, 2000) explicitly recognizes the importance of managing credit risk at both the transaction and portfolio levels.
Credit risk concerns the possibility that a borrower or counterparty will not meet obligations on agreed terms. Assessing this risk informs lending decisions, loan pricing, collateral requirements, exposure limits, and provisioning. Screening before credit approval and monitoring after disbursement are central to controlling information asymmetry. The quality of a bank’s loan portfolio is consequential for both earnings and continuity of operations. A sound lending policy therefore combines borrower assessment with portfolio diversification, asset classification, and procedures for identifying deteriorating exposures (BCBS, 2000, 2025).
Liquidity risk has a different immediate manifestation: an institution may be unable to meet payment obligations when they fall due without incurring unacceptable losses. Deposit withdrawals, mismatched asset and liability maturities, and difficulty converting assets into cash can generate funding pressures. The liquidity-risk framework developed by the BCBS (2008) emphasizes governance, measurement, contingency planning, and the maintenance of adequate liquidity resources. These considerations provide a rationale for examining liquidity alongside other banking risks rather than as an isolated cash-management problem. More recent analysis of the 2023 banking turmoil examines distressed-bank outflows and the adequacy of liquidity monitoring (BCBS, 2024b). Baudino et al. (2024) distinguish institution-specific, sector-wide, and bank–nonbank liquidity stress tests, highlighting the importance of outflow assumptions, asset-price changes, and contagion channels.
Stress testing provides a structured way to consider adverse conditions beyond routine operating assumptions. Its contribution depends on the purpose of the exercise, the relevance of the selected risk factors, and the quality of the underlying information. The BCBS (2018) sets out principles for integrating stress testing into governance and risk management. Risk prioritization is an input to that process: identifying influential exposures can inform the selection of factors and transmission channels that merit closer examination. It does not, by itself, quantify losses or demonstrate capital and liquidity adequacy under stress. Schuermann (2014) places scenario design, execution, and disclosure within a broader bank stress-testing framework. The European Banking Authority (EBA, 2024) published a common methodology and templates for its 2025 exercise, and the subsequent results report assesses bank resilience under the specified scenarios (EBA, 2025). These applications illustrate the distinction between selecting risk factors and estimating scenario-dependent financial outcomes.
The present study addresses banks listed on the Tehran Stock Exchange. It combines qualitative evidence from banking experts with a quantitative network-based prioritization of the identified risks. The objective is to identify the relevant categories, examine their perceived interdependencies, and derive relative priority weights for use in liquidity and credit stress-test planning. The analysis is organized around eight risk categories, while retaining the distinction between an expert-derived influence network and empirically established causal relationships.
Theoretical Background and Related Research
Banking risk management
Risk management in banks is complicated by the number and diversity of transactions, the limited capital available to absorb losses, and the need to protect depositors as well as shareholders. Borrowers differ in financial condition and repayment capacity, while banking employees make decisions involving substantial financial resources. These characteristics make reliable information, clear decision authority, and independent oversight important components of risk control. Figure 1 provides a conceptual taxonomy for this discussion; its categories are not treated as a mutually exclusive or empirically estimated hierarchy.
The credit-risk management process includes identifying exposures embedded in existing and proposed activities, defining lending standards, classifying assets, monitoring recoverability, and maintaining adequate provisions. Banks also need procedures for approving, renewing, restructuring, and refinancing credit. Portfolio-level information should reveal concentration, changes in exposure quality, and deviations from approved limits. Independent review and reporting to senior management help connect these activities to governance (BCBS, 2000, 2025).
The risk-management framework considered here distinguishes identification, measurement, control, and monitoring. Identification specifies exposures; measurement translates them into indicators; control concerns avoidance, reduction, mitigation, and the balance between return and exposure; and monitoring detects adverse changes through information and reporting systems. These functions are complementary and provide the organizational context for risk prioritization. Effective implementation also depends on accurate, comprehensive, and timely risk information. The risk-data aggregation principles and the subsequent implementation update emphasize data governance and traceability as foundations for sound risk management (BCBS, 2013, 2026).
Relationships between credit and liquidity risks
Credit and liquidity risks can interact through expected repayments, the value and marketability of assets, and the stability of funding. An institution with deteriorating credit quality may face pressure on funding, while liquidity constraints may limit its ability to maintain or adjust lending activities. The direction and magnitude of these relationships depend on the setting and should not be presumed to be uniform across banks. Abdelaziz et al. (2022), using conventional-bank data from the Middle East and North Africa, examined credit risk, liquidity risk, and profitability jointly. Their results associated higher credit and liquidity risks with lower profitability and highlighted the relevance of institutional quality. Imbierowicz and Rauch (2014) find no reliable direct relationship between credit and liquidity risks across their US bank sample, while showing that their interaction can influence default risk differently at different overall risk levels. Their findings support investigating joint exposure without presuming a uniform transmission pattern.
The financial-institution taxonomy in Figure 2 groups market, credit, liquidity, and operational risks and shows their constituent categories. Its purpose is to make the distinctions used in the qualitative discussion explicit. The eight categories ultimately included in DANP are listed separately in Table 2; the background taxonomy and the analytical criteria are not interchangeable.
Stress testing and risk prioritization
Patra and Padhi (2022) developed a macroeconomic credit-risk stress-testing model for Indian banks using panel-data methods and baseline, medium, and severe scenarios. They examined how macroeconomic changes affected non-performing loans and regulatory capital across bank groups. This application illustrates the additional modelling needed to move from identifying important risks to assessing the financial consequences of adverse scenarios. The present study instead focuses on the earlier stage of identifying risks and assigning relative influence weights. In a more recent regional application, Alter et al. (2024) use granular corporate and banking data to assess MENA vulnerabilities in a higher-for-longer interest rate environment, including corporate stress and liquidity pressures. This offers a relevant regional comparison, although it does not establish equivalent effects for the Iranian banks examined here.
Stress-testing results also require careful interpretation. Borio et al. (2014) discuss the strengths and limitations of macro stress-testing frameworks and caution against expectations that exceed their analytical capabilities. In the present context, a DANP weight indicates relative priority within a specified expert-assessment network. It is not an estimate of default probability, a capital requirement, a liquidity shortfall, or a measured contribution to systemic losses.
The DANP approach combines DEMATEL-based influence assessment with network-derived priorities. A published methodological application by Kuo et al. (2015) demonstrates how DEMATEL relationships can be integrated with an analytic network process when criteria are interdependent. That work is used here as methodological context rather than as evidence about banking risk. The contribution of the present application lies in its banking setting and the combination of expert-derived categories with a network prioritization of those categories.
Methods
Research design and setting
The study adopted a pragmatic paradigm and a sequential exploratory mixed-methods design. Qualitative work identified concepts and categories that informed expert-based network analysis. The setting comprised 10 banks listed on the Tehran Stock Exchange. Financial information was obtained from Codal disclosures and Tehran Stock Exchange databases. The influence matrices were constructed from expert ratings; bank-level financial variables were not used to estimate stress losses. The analysis therefore concerns qualitative risk identification and expert-based prioritization.
Qualitative data collection and analysis
Semi-structured interviews were conducted with 20 banking experts. Recordings were transcribed, and the transcripts were read repeatedly before coding in NVivo. The coding sequence drew on Corbin and Strauss (2015): open coding identified concepts, axial coding examined relationships among categories, and selective coding integrated the framework. The qualitative evidence supported the identification of risk criteria; it did not establish a formally validated theory of banking crises.
Familiarization and iterative coding also drew on Braun and Clarke (2006). These activities supported the examination of patterns in the material, while open, axial, and selective coding described its conceptual organization. Researcher triangulation supported interpretation. Cohen’s kappa was 0.837, reflecting agreement beyond chance in the coding comparison. This coefficient was considered an agreement measure rather than evidence of conceptual validity or theoretical saturation.
Risk identification combined literature review with interview-based concept development. A questionnaire based on the initial factors was subsequently distributed to experts, and their feedback informed revisions. This stage provided expert feedback for organizing the framework. Table 1 presents 46 coding entries with their recorded identifiers; the descriptions are translated coding material rather than verbatim interview quotations.
Quantitative participants and influence ratings
The quantitative phase involved 52 senior, middle, and first-line banking managers across 10 listed banks. The direct-influence matrix aggregated ratings from 15 experts. The network calculations and weights presented below pertain specifically to that 15-expert assessment. The sample counts refer to different components of the study and are not combined into a single analytical sample.
Each pair of criteria was assessed on a five-point influence scale: 0, no influence; 1, very low influence; 2, low influence; 3, high influence; and 4, very high influence. Ratings represented the perceived influence of one criterion on another. The arithmetic mean across the 15 experts formed the direct-influence matrix. Excel was used for normalization, total-influence calculations, the influence map, and the limiting supermatrix.
DEMATEL and network weighting
The analytical sequence comprised averaging expert ratings, normalizing the direct-influence matrix A, calculating the total-influence matrix T, and deriving prominence and net influence. The normalization divisor is the larger of the maximum row sum and maximum column sum of A, yielding N. The total-influence matrix is N multiplied by the inverse of I minus N, where I is the identity matrix. This calculation aggregates direct and indirect influence in the specified network; it does not estimate causal effects from bank financial observations.
For each criterion, D is the row sum of T and represents total outgoing influence; R is the column sum and represents total incoming influence. D + R describes prominence in the network, while D − R distinguishes net influence. A positive D − R identifies a net influencing criterion, and a negative value identifies a net receiving criterion within the expert-assessment model. The conventional DEMATEL terms “cause” and “effect” refer to this classification and do not establish causal identification from observational or experimental data.
The eight-criterion total-influence matrix was normalized by row and transposed to obtain a column-stochastic network matrix. Successive matrix powers were used to derive the final weights. The calculation used the fifth power to obtain the displayed limiting-supermatrix values. The network was analyzed at the level of the eight criteria. Full-precision iteration statistics and an explicit convergence tolerance are not available, limiting independent assessment of the stopping criterion.
Results
Qualitative coding and selected criteria
The coding material emphasized credit, interest rate, and liquidity risks and addressed repayment, funding stability, maturity mismatch, credit assessment, portfolio concentration, information systems, capital adequacy, and stress-testing procedures (Table 1). References to procurement, wages, and product quality provide general corporate examples; they are not treated as direct bank-specific observations.
No. | Original identifier | Translated coding extract | Open code |
|---|---|---|---|
1 | C1-001 | Banks and financial institutions face multiple risks from the outset because of the nature of their activities. The breadth and diversity of banking operations have prevented agreement on a single classification of banking risks. | Breadth and diversity of activities; operational risk |
2 | C2-001 | Credit, interest rate, and liquidity risks are among the principal risks in banking. | Credit risk; interest rate risk; liquidity risk |
3 | C3-001 | Among the risks facing banks and financial institutions, credit risk is particularly important because of its centrality, transaction volume, and sensitivity. | Credit risk |
4 | C4-001 | Liquidity risk is the risk of insufficient cash to meet obligations. Firms may encounter liquidity problems when they cannot sell products, collect sales proceeds, or contain excessive cost increases, leading to lower efficiency. | Liquidity risk |
5 | C5-001 | Liquidity problems weaken a firm’s financial structure and make it difficult to procure raw materials and meet current obligations, including wages. The firm may consequently use cheaper, lower-quality inputs. | Liquidity problems; payment of current obligations |
6 | C6-001 | Weaknesses in the operating structure reduce the quantity and quality of activities and performance and, in turn, affect the firm’s financial structure. | Weakness in the operating structure |
7 | C7-001 | Deposit withdrawals at maturity instead of renewal, early withdrawals, or negative deposit growth can increase liquidity risk. | Deposit withdrawals at maturity; liquidity risk |
8 | C8-001 | Because idle cash does not generate returns, banks hold assets as loans or other earning assets. The convertibility of these assets into cash is important for reducing liquidity risk. | Non-earning cash; assets held as loans |
9 | C9-001 | The relationship between the average maturity of loans and the maturity of deposits can be used to assess the maturity gap between assets and liabilities. | Asset–liability maturity gap |
10 | C10-001 | A lower net present value of the balance sheet indicates less favorable conditions and greater liquidity risk. | Liquidity risk |
11 | C11-001 | Credit risk concerns losses caused by nonpayment or delayed payment of loan principal or interest by customers. | Credit risk |
12 | C12-001 | Before a loan is granted, the recipient’s creditworthiness and ability to repay principal and interest must be assessed. | Credit risk |
13 | C13-001 | Internal credit rating by banks is described in the interview extract as a practice introduced less than 10 years earlier. Although detailed information about the institutions’ methodologies and processes is unavailable, the criteria suggest limited differences in the methods used. | Credit risk |
14 | C14-001 | Higher credit risk increases the marginal cost of debt and equity, which in turn raises a bank’s cost of funding. | Credit risk; marginal cost of debt |
15 | C15-001 | Credit risk includes non-repayment of loan principal and interest and potential transaction losses arising from product quality, the items exchanged, and the reliability and creditworthiness of counterparties. | Non-repayment of loan principal and interest |
16 | C16-001 | Banks should identify and manage credit risk in all products and activities. Before new products or activities are introduced or committed to, their risks should be assessed with appropriate controls and approved by the board or another suitable committee. | Credit risk; risks of new products and activities |
17 | C17-001 | Credit should be granted under appropriate and clearly defined criteria. | Bank lending |
18 | C18-001 | Banks should establish explicit and transparent procedures for approving new credit and for modifying, renewing, and refinancing existing credit. | Credit approval |
19 | C19-001 | Banks should maintain information systems and analytical procedures that enable management to measure credit risk across balance-sheet activities. Management information should describe the credit portfolio and identify concentration risk. | Information systems; balance-sheet activities; credit portfolio composition |
20 | C20-001 | Banks should establish a system to monitor the overall composition and quality of the credit portfolio. | Portfolio composition |
21 | C21-001 | When evaluating individual credits and the overall portfolio, banks should consider possible changes in economic conditions and assess credit exposures under stress. | Credit risk; stress conditions |
22 | C22-001 | Banks should maintain an independent, continuing review of credit-risk management, with findings reported directly to the board and senior management. | Credit risk |
23 | C23-001 | Stress testing is an important risk-management tool used by banks as part of internal risk management. | Stress testing; risk management; internal risk management |
24 | C24-001 | Stress testing measures the vulnerability of a portfolio or financial system to hypothetical scenarios. | Measurement of portfolio vulnerability |
25 | C25-001 | Stress testing is a computer-based simulation technique for examining bank and investment-portfolio performance under severe economic scenarios. | Bank performance; investment portfolio |
26 | C26-001 | Stress testing helps assess investment risk, asset adequacy, and internal processes and controls. | Investment risk; asset adequacy |
27 | C27-001 | Regulation requires banks and funds to conduct different stress scenarios and report on internal capital- and risk-management practices. | Capital management; risk management |
28 | C28-001 | Stress testing may examine changes in one risk factor through sensitivity analysis or simultaneous changes in several factors through scenario analysis. | Risk factor; sensitivity analysis |
29 | C29-001 | Financial markets mobilize and allocate investment resources to economic activity; bank credit is part of this market. | Bank credit |
30 | C30-001 | Credit risk is among the oldest and largest risks in exchange, including the possibility that credit is lost when borrowers default on debt payments. | Credit risk; credit extended |
31 | C31-001 | Diversifying counterparties and the forms of activities are approaches to credit-risk management. | Counterparty diversification; diversity of activities |
32 | C2-001 | Loans for which principal and interest are not repaid in full, or are repaid late, generate credit risk for a bank. | Loan non-repayment |
33 | C2-002 | Under dynamic and unpredictable economic conditions, the complex interaction between liquidity and credit risk creates important banking challenges. Banks are also major providers of finance to industry, agriculture, and services. | Liquidity risk; credit risk |
34 | C2-003 | Problems in evaluating the effect of risk on bank performance arise when the risks of different loan types are not distinguished. | Failure to distinguish risks across loan types |
35 | C2-004 | The assessment of loans to individual and corporate borrowers differs, as does their relationship with bank performance. Despite the volume of outstanding claims, this distinction has received insufficient practical attention. | Differentiation of bank loans |
36 | C2-005 | Credit scoring uses applicants’ historical and current information to evaluate the probability of loan non-repayment and assign a score. | Credit scoring |
37 | C2-006 | Measuring credit risk by forecasting non-repayment losses and establishing a relationship between risk and return supports portfolio optimization, asset pricing, and the determination of economic capital. This can reduce capital costs and help preserve competitiveness. | Credit risk |
38 | C2-007 | Every transaction and service in banking carries particular inherent risks. | Risk inherent in banking services |
39 | C2-008 | Insufficient bank capital to withstand financial crises is a principal constraint and can amplify a crisis. | Insufficient bank capital |
40 | C2-009 | Market fluctuations and loss of depositor confidence can lead to deposit withdrawals and liquidity risk. | Market fluctuations; depositor distrust |
41 | C2-010 | Adverse economic conditions intensify banks’ credit risks. | Adverse economic conditions |
42 | C2-011 | Credit risk has a complex reciprocal role in the emergence of liquidity risk. | Credit risk; liquidity risk |
43 | C2-012 | Loans constitute a large share of commercial-bank assets and expose banks to default. Identifying borrowers’ default probability is therefore important for credit-risk management. | Identification of borrower default probability |
44 | C2-013 | Simulating default probabilities is particularly important for effective credit-risk management in Iran. | Simulation of default probabilities |
45 | C2-014 | Liquidity is generally defined as a financial institution’s ability to meet debt obligations without incurring large, unacceptable losses. Funding liquidity risk also concerns its ability to meet expected and unexpected cash-flow and collateral requirements without materially disrupting operations or its financial position. | Liquidity; debt servicing; inability to estimate current and future requirements |
46 | C2-015 | Banks’ liquidity risk reflects shortages and uncertainty in available liquidity. Adequate cash balances help ensure that depositor obligations and liquidity needs can be met on time. | Liquidity risk; liquidity shortage and uncertainty |
Note. The 46 entries retain their recorded coding identifiers. Repeated identifiers do not identify unique participants. Descriptions are translated coding material rather than verbatim quotations.
Eight criteria entered the network: liquidity, credit, market, interest rate, foreign exchange, financial, legal and regulatory, and business risks (Table 2). These labels describe the criteria evaluated in the rating matrices. They are not necessarily mutually exclusive: financial risk is a broad label, while market risk may overlap with interest rate and foreign exchange risks. Potential overlap is considered when interpreting the priorities.
Criterion | Code |
|---|---|
Liquidity risk | C1 |
Credit risk | C2 |
Market risk | C3 |
Interest rate risk | C4 |
Foreign exchange risk | C5 |
Financial risk | C6 |
Legal and regulatory risk | C7 |
Business risk | C8 |
Direct and total influence matrices
The direct-influence matrix in Table 3 represents the arithmetic mean of the 15 experts’ ratings. Diagonal entries are zero by construction. Table 4 presents the normalized matrix, and Table 5 gives the total-influence matrix. The latter incorporates direct and indirect relationships among the eight criteria.
C1 | C2 | C3 | C4 | C5 | C6 | C7 | C8 | |
|---|---|---|---|---|---|---|---|---|
C1 | 0 | 3 | 2.467 | 3 | 2.067 | 2.2 | 2.267 | 2.2 |
C2 | 2.6 | 0 | 1.667 | 3.067 | 1.533 | 1.867 | 2.333 | 1.4 |
C3 | 2 | 2.933 | 0 | 3.066667 | 2 | 2.333 | 2.533 | 2.867 |
C4 | 0.933 | 0 | 2.267 | 0 | 2 | 2 | 2.133 | 1.8 |
C5 | 2.133 | 2.933 | 1.867 | 2.8 | 0 | 1.6 | 1.733 | 1.467 |
C6 | 1.933 | 3.333 | 1.733 | 2.933 | 1.867 | 0 | 2.467 | 2.467 |
C7 | 2.6 | 3.667 | 2.267 | 3.4 | 1.933 | 1.933 | 0 | 2 |
C8 | 2.133 | 3.467 | 1.4 | 3.067 | 1.8 | 1.933 | 2.2 | 0 |
Note. C1–C8 denote the criteria listed in Table 2.
C1 | C2 | C3 | C4 | C5 | C6 | C7 | C8 | |
|---|---|---|---|---|---|---|---|---|
C1 | 0 | 0.141 | 0.116 | 0.141 | 0.097 | 0.103 | 0.106 | 0.103 |
C2 | 0.122 | 0 | 0.078 | 0.144 | 0.072 | 0.088 | 0.109 | 0.066 |
C3 | 0.094 | 0.138 | 0 | 0.144 | 0.094 | 0.109 | 0.119 | 0.134 |
C4 | 0.044 | 0 | 0.106 | 0 | 0.094 | 0.094 | 0.1 | 0.084 |
C5 | 0.1 | 0.138 | 0.088 | 0.131 | 0 | 0.075 | 0.081 | 0.069 |
C6 | 0.091 | 0.156 | 0.081 | 0.138 | 0.088 | 0 | 0.116 | 0.116 |
C7 | 0.122 | 0.172 | 0.106 | 0.159 | 0.091 | 0.091 | 0 | 0.094 |
C8 | 0.1 | 0.163 | 0.066 | 0.144 | 0.084 | 0.091 | 0.103 | 0 |
Note. C1–C8 denote the criteria listed in Table 2.
C1 | C2 | C3 | C4 | C5 | C6 | C7 | C8 | |
|---|---|---|---|---|---|---|---|---|
C1 | 0.251 | 0.441 | 0.349 | 0.487 | 0.324 | 0.340 | 0.372 | 0.342 |
C2 | 0.322 | 0.269 | 0.283 | 0.437 | 0.269 | 0.292 | 0.334 | 0.275 |
C3 | 0.344 | 0.448 | 0.251 | 0.500 | 0.328 | 0.352 | 0.390 | 0.375 |
C4 | 0.214 | 0.229 | 0.264 | 0.249 | 0.248 | 0.255 | 0.280 | 0.251 |
C5 | 0.304 | 0.390 | 0.290 | 0.426 | 0.200 | 0.280 | 0.310 | 0.276 |
C6 | 0.327 | 0.445 | 0.313 | 0.475 | 0.309 | 0.240 | 0.372 | 0.345 |
C7 | 0.366 | 0.472 | 0.348 | 0.512 | 0.324 | 0.337 | 0.283 | 0.340 |
C8 | 0.324 | 0.435 | 0.290 | 0.464 | 0.296 | 0.312 | 0.350 | 0.229 |
Note. C1–C8 denote the criteria listed in Table 2.
Prominence and net influence
Table 6 and Figure 3 distinguish prominence from net influence. With rows representing outgoing influence, liquidity risk (D − R = 0.455), market risk (0.601), foreign exchange risk (0.178), financial risk (0.418), legal and regulatory risk (0.291), and business risk (0.266) are net influencing criteria. Credit risk (−0.649) and interest rate risk (−1.560) are net receiving criteria. This classification follows the signs of D − R under the matrix convention defined in Section 3.4.
Criterion | D | R | D + R | D − R |
|---|---|---|---|---|
C1 | 2.907 | 2.452 | 5.358 | 0.455 |
C2 | 2.480 | 3.129 | 5.610 | −0.649 |
C3 | 2.989 | 2.388 | 5.377 | 0.601 |
C4 | 1.989 | 3.549 | 5.539 | −1.560 |
C5 | 2.475 | 2.298 | 4.773 | 0.178 |
C6 | 2.825 | 2.407 | 5.232 | 0.418 |
C7 | 2.983 | 2.691 | 5.674 | 0.291 |
C8 | 2.698 | 2.433 | 5.131 | 0.266 |
Note. D is the row sum of total influence and R is its column sum. Positive D − R indicates net influence; negative D − R indicates net reception. Minor discrepancies between displayed totals may reflect rounding.
Network weights and priorities
Table 7 presents the transposed, row-normalized total-influence matrix. Table 8 gives the final supermatrix weights, repeated across columns. Figure 4 displays the weights in criterion-code order. The weights sum to 1.0000 and express relative priority within the eight-criterion expert-assessment network.
C1 | C2 | C3 | C4 | C5 | C6 | C7 | C8 | |
|---|---|---|---|---|---|---|---|---|
C1 | 0.086 | 0.130 | 0.115 | 0.108 | 0.123 | 0.116 | 0.123 | 0.120 |
C2 | 0.152 | 0.108 | 0.150 | 0.115 | 0.157 | 0.158 | 0.158 | 0.161 |
C3 | 0.120 | 0.114 | 0.084 | 0.133 | 0.117 | 0.111 | 0.117 | 0.107 |
C4 | 0.168 | 0.176 | 0.167 | 0.125 | 0.172 | 0.168 | 0.172 | 0.172 |
C5 | 0.111 | 0.108 | 0.110 | 0.125 | 0.081 | 0.109 | 0.109 | 0.110 |
C6 | 0.117 | 0.118 | 0.118 | 0.128 | 0.113 | 0.085 | 0.113 | 0.116 |
C7 | 0.128 | 0.135 | 0.131 | 0.141 | 0.125 | 0.132 | 0.095 | 0.130 |
C8 | 0.118 | 0.111 | 0.125 | 0.126 | 0.111 | 0.122 | 0.114 | 0.085 |
C1 | C2 | C3 | C4 | C5 | C6 | C7 | C8 | |
|---|---|---|---|---|---|---|---|---|
C1 | 0.1151 | 0.1151 | 0.1151 | 0.1151 | 0.1151 | 0.1151 | 0.1151 | 0.1151 |
C2 | 0.1426 | 0.1426 | 0.1426 | 0.1426 | 0.1426 | 0.1426 | 0.1426 | 0.1426 |
C3 | 0.1139 | 0.1139 | 0.1139 | 0.1139 | 0.1139 | 0.1139 | 0.1139 | 0.1139 |
C4 | 0.1634 | 0.1634 | 0.1634 | 0.1634 | 0.1634 | 0.1634 | 0.1634 | 0.1634 |
C5 | 0.1088 | 0.1088 | 0.1088 | 0.1088 | 0.1088 | 0.1088 | 0.1088 | 0.1088 |
C6 | 0.1142 | 0.1142 | 0.1142 | 0.1142 | 0.1142 | 0.1142 | 0.1142 | 0.1142 |
C7 | 0.1274 | 0.1274 | 0.1274 | 0.1274 | 0.1274 | 0.1274 | 0.1274 | 0.1274 |
C8 | 0.1146 | 0.1146 | 0.1146 | 0.1146 | 0.1146 | 0.1146 | 0.1146 | 0.1146 |
Note. Each row’s common value is the corresponding criterion’s final weight. C1–C8 are defined in Table 2.
Interest rate risk ranked first with a weight of 0.1634, credit risk ranked second at 0.1426, and legal and regulatory risk ranked third at 0.1274. The remaining criteria, in descending order, were liquidity risk (0.1151), business risk (0.1146), financial risk (0.1142), market risk (0.1139), and foreign exchange risk (0.1088). These priorities reproduce the reported limiting-supermatrix weights. They should be distinguished from the sign of D − R: a high priority weight does not necessarily identify a net influencing factor.
Discussion
This study combines qualitative identification of banking risks with network-based prioritization for liquidity and credit stress-test planning. The most prominent final priorities were interest rate, credit, and legal and regulatory risks. The qualitative discussion also emphasized liquidity and the reciprocal relationship between funding and credit conditions. Together, the findings support examining several interconnected exposures when deciding which risks to include in a stress-testing framework. They do not demonstrate that any single risk causes a quantified change in bank losses or resilience.
The first-ranked weight for interest rate risk indicates that experts assigned it substantial relative importance in the specified network. This ranking is distinct from its negative net-influence value in Table 6. The latter means that, within the reported matrix, interest rate risk receives more total influence than it transmits. There is no logical requirement that the highest-weighted criterion also be a net cause. Maintaining that distinction prevents a priority ranking from being interpreted as evidence of an upstream causal mechanism. The broader importance of this exposure is recognized in the banking-book interest rate risk standard and its subsequent shock recalibration (BCBS, 2016, 2024a). Jiang et al. (2024) examine how monetary tightening, asset valuation losses, and uninsured deposit funding can interact in bank fragility. This literature supports joint assessment of asset and funding exposures; it does not validate the magnitude of the present expert-derived weight.
Credit risk was the second-ranked criterion and was central to many coding entries, including borrower assessment, loan recovery, information quality, and portfolio composition. Its importance is consistent with the attention given to credit administration and monitoring in banking supervisory guidance (BCBS, 2000, 2025). However, consistency with that guidance does not validate the numerical weight or establish generalizability beyond the participating experts and banks. The present weights are specific to the criteria and judgments included in the analysis.
The third-ranked legal and regulatory category places governance and institutional conditions alongside financial exposures. Abdelaziz et al. (2022) also report relationships involving institutional quality in their analysis of banking risk and profitability, although their econometric results concern a different sample and method. The comparison provides a relevant contextual link rather than a replication of the present findings. It does not establish that a particular legal reform would produce a measurable reduction in credit or liquidity risk in the studied banks.
Transparent financial reporting and the implementation of International Financial Reporting Standards (IFRS) are possible considerations for risk-management practice. Inflation, cash-flow forecasting difficulties, and legal and institutional infrastructure may affect implementation. However, the present study did not compare IFRS adopters and non-adopters or estimate the effects of reporting reform. IFRS, factoring, credit guarantees, and organizational changes are therefore discussed as possible applications for separate evaluation, not as interventions shown by this study to reduce risk.
An important boundary concerns the meaning of a stress-testing model. The present application identifies factors and their relative weights but does not report calibrated baseline, moderate, or severe shocks or corresponding losses, capital ratios, or liquidity gaps. Patra and Padhi (2022) provide an example of the additional macroeconomic modelling needed to evaluate bank resilience under specified scenarios. The BCBS (2018) similarly places risk-factor selection within a wider stress-testing process. The current framework can inform that process, while its predictive and operational performance remains untested.
Practical implications
The priority weights can inform discussion of which exposures merit closer examination during stress-test development. Strengthening risk-management functions and considering factoring or credit guarantees are possible management responses, but their effectiveness does not follow from the rankings. No cost, feasibility, intervention, or comparative analysis of these responses was conducted.
Financial-reporting education, alignment of legal and reporting frameworks, economic stability, and implementation capacity provide context for possible institutional applications. They were not outcomes evaluated in the network analysis. Accordingly, no claim is made that changing tax rules, adopting IFRS, providing parallel financial statements, or expanding international engagement would improve resilience in the studied banks. Any such proposal requires independent legal, institutional, and empirical assessment.
Limitations
The study is exploratory and based on expert judgments. Its weights do not estimate population-level risk, default probabilities, causal effects, or stress losses. Selection effects and temporal relevance cannot be fully assessed because participant recruitment, overlap among the interview and questionnaire groups, collection dates, and the financial observation period were not comprehensively recorded. The network findings pertain to the 15-expert matrix and should not be generalized to all managers or listed banks.
The qualitative component combines literature review, coding, and expert feedback. The available coding records do not provide a complete audit trail from every entry to the eight final criteria. Limited information about independent coding, saturation, and the provenance of individual extracts constrains assessment of qualitative rigor. General corporate examples and repeated coding identifiers also limit the specificity of the evidence. Cohen’s kappa of 0.837 should be interpreted within these reporting limitations.
Potential category overlap, rounded matrix entries, and the absence of an explicit convergence tolerance limit reproducibility. The arrow-selection rule for the directional diagram is not available, so additional conclusions are not drawn from individual arrows. Sensitivity to expert composition and plausible rating changes was not evaluated. The closely spaced lower-ranked weights consequently cannot be assumed to represent stable or substantively distinct priorities.
Finally, the study does not test the proposed framework against realized crisis outcomes or calibrated stress scenarios. The effectiveness of the suggested reporting, training, and liquidity-management measures therefore remains outside the empirical scope. Further applications could assess stability across expert panels and examine the framework’s contribution to quantified credit and liquidity stress tests; these are future research directions rather than completed analyses.
Conclusion
The study identifies eight banking-risk categories and reports an expert-based prioritization to inform liquidity and credit stress-test development. Interest rate risk has the highest weight (0.1634), followed by credit risk (0.1426) and legal and regulatory risk (0.1274). Under the stated matrix convention, interest rate and credit risks are net receiving criteria, while liquidity risk is a net influencing criterion. Priority and net influence describe different network properties. The framework has not been validated against stress losses, liquidity shortfalls, or realized crisis outcomes, and its rankings remain subject to the documented sampling, measurement, and reproducibility limitations.
Declarations
Ethical considerations
The study concerns non-clinical research involving banking experts and managers. Respect for participant autonomy, voluntary participation, and confidentiality provides the ethical framework for interview and questionnaire research. The findings are presented as aggregate assessments and coded extracts; participant names and individual bank identities are not disclosed in the reported results.
Author contributions
Sadegh Adelnia: Conceptualization, methodology, investigation, data collection, data curation, formal analysis, and writing—original draft; Seyyed Abbas Borhani: Conceptualization, methodology, supervision, validation, and writing—review and editing; Seyed Kazem Chavoshi: Methodology, validation, interpretation of findings, and writing—review and editing; Mojgan Safa: Methodology, validation, interpretation of findings, and writing—review and editing. All authors reviewed and approved the final manuscript and agree to be accountable for the accuracy and integrity of the work.
Funding
No specific funding source is acknowledged for this study.
Competing interests
No competing interests are disclosed in this article.
Acknowledgments
The contribution of the banking experts and managers whose interview and questionnaire responses informed this research is gratefully acknowledged.
Data availability
The data supporting the findings of this study are available from the corresponding author upon reasonable request, subject to participant confidentiality and applicable institutional requirements.
References
Abdelaziz, H., Rim, B., & Helmi, H. (2022). The interactional relationships between credit risk, liquidity risk and bank profitability in MENA region. Global Business Review, 23(3), 561–583. https://doi.org/10.1177/0972150919879304
Alter, A., Hlayhel, B., Kroen, T., & Piontek, T. (2024). Financial stability in a higher-for-longer interest rate environment: The case of the Middle East and North Africa (IMF Working Paper No. 2024/080). International Monetary Fund. https://doi.org/10.5089/9798400270017.001
Basel Committee on Banking Supervision. (2000). Principles for the management of credit risk. Bank for International Settlements. https://www.bis.org/publ/bcbs75.htm
Basel Committee on Banking Supervision. (2008). Principles for sound liquidity risk management and supervision. Bank for International Settlements. https://www.bis.org/publ/bcbs144.htm
Basel Committee on Banking Supervision. (2013). Principles for effective risk data aggregation and risk reporting. Bank for International Settlements. https://www.bis.org/publications/201301-guidelines-principles-effective-risk-data-aggregation-and-risk-reporting
Basel Committee on Banking Supervision. (2016). Interest rate risk in the banking book. Bank for International Settlements. https://www.bis.org/publications/201604-standards-interest-rate-risk-banking-book
Basel Committee on Banking Supervision. (2018). Stress testing principles. Bank for International Settlements. https://www.bis.org/bcbs/publ/d450.htm
Basel Committee on Banking Supervision. (2024a). Recalibration of shocks for interest rate risk in the banking book. Bank for International Settlements. https://www.bis.org/publications/202407-standards-recalibration-shocks-interest-rate-risk-banking-book
Basel Committee on Banking Supervision. (2024b). The 2023 banking turmoil and liquidity risk: A progress report. Bank for International Settlements. https://www.bis.org/publications/2023-banking-turmoil-and-liquidity-risk-progress-report
Basel Committee on Banking Supervision. (2025). Principles for the management of credit risk. Bank for International Settlements. https://www.bis.org/publications/202504-guidelines-principles-management-credit-risk
Basel Committee on Banking Supervision. (2026). Implementation of the Principles for effective risk data aggregation and risk reporting (BCBS 239 Principles) (BCBS Newsletter No. 36). Bank for International Settlements. https://www.bis.org/publications/implementation-principles-effective-risk-data-aggregation-and-risk-reporting-bcbs-239-principles
Baudino, P., de Carvalho, P., & Svoronos, J.-P. (2024). Liquidity stress tests for banks: Range of practices and possible developments (FSI Insights No. 59). Bank for International Settlements. https://www.bis.org/publications/fsi-insight-59-liquidity-stress-tests-banks-range-practices-and-possible-developments
Borio, C., Drehmann, M., & Tsatsaronis, K. (2014). Stress-testing macro stress testing: Does it live up to expectations? Journal of Financial Stability, 12, 3–15. https://doi.org/10.1016/j.jfs.2013.06.001
Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology, 3(2), 77–101. https://doi.org/10.1191/1478088706qp063oa
Corbin, J., & Strauss, A. (2015). Basics of qualitative research: Techniques and procedures for developing grounded theory (4th ed.). SAGE.
European Banking Authority. (2024, November 12). The EBA publishes methodology, draft templates, and key milestones for its 2025 EU-wide stress test. https://www.eba.europa.eu/publications-and-media/press-releases/eba-publishes-methodology-draft-templates-and-key-milestones-its-2025-eu-wide-stress-test
European Banking Authority. (2025). 2025 EU-wide stress test: Results. https://www.eba.europa.eu/publications-and-media/publications/2025-eu-wide-stress-test-results
Imbierowicz, B., & Rauch, C. (2014). The relationship between liquidity risk and credit risk in banks. Journal of Banking & Finance, 40, 242–256. https://doi.org/10.1016/j.jbankfin.2013.11.030
Jiang, E. X., Matvos, G., Piskorski, T., & Seru, A. (2024). Monetary tightening and U.S. bank fragility in 2023: Mark-to-market losses and uninsured depositor runs? Journal of Financial Economics, 159, 103899. https://doi.org/10.1016/j.jfineco.2024.103899
Kuo, T. C., Hsu, C.-W., & Li, J.-Y. (2015). Developing a green supplier selection model by using the DANP with VIKOR. Sustainability, 7(2), 1661–1689. https://doi.org/10.3390/su7021661
Patra, B., & Padhi, P. (2022). Resilience of Indian banks: Macroeconomic stress test modeling for credit risk. Journal of Public Affairs, 22(1), e2350. https://doi.org/10.1002/pa.2350
Schuermann, T. (2014). Stress testing banks. International Journal of Forecasting, 30(3), 717–728. https://doi.org/10.1016/j.ijforecast.2013.10.003
