## wpiea2020081-print-pdf

## Source details

**Canonical URL:** [wpiea2020081-print-pdf](https://www.imf.org/-/media/files/publications/wp/2020/english/wpiea2020081-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2020/english/wpiea2020081-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2020/english/wpiea2020081-print-pdf.pdf.json)

---

### I. Introduction — scope and focus
- FMIs facilitate clearing and settlement and must comply with the Principles for Financial Market Infrastructures (PFMI) (CPSS‑IOSCO, 2012).
- FMIs’ malfunction can significantly disrupt the financial system; illustrative incidents cited include outages in the United Kingdom in 2014, defaulting participants at Nasdaq Clearing in 2018, and a DDos attack on Norges Bank in 2014.
- Novelty and focus of the paper:
  - Includes FMIs explicitly as participants in LVPS network analysis (contrast with prior analyses that often omitted FMIs).
  - Measures the LVPS network as a “flow” per unit time (contrast with “stock” exposure networks).
  - Develops a composite risk indicator capturing both liquidity risk and systemic impact from the failure of a financial institution in an LVPS network.
  - Defines “failing” as inability to make any payments due to operational problems, cyber‑attack, severe power outage, operational fraud, or bankruptcy.
  - Corrects for FMIs that settle net positions (e.g., TARGET2 settlement hides gross underlying flows owing to netting efficiencies up to 98 percent).

### II. TARGET2 — system description and key statistics (2018)
- System characteristics:
  - TARGET2 is the LVPS and RTGS system for euro denominated payments owned and operated by the Eurosystem.
  - All payments settled gross and in real time in central bank money with immediate finality; no upper or lower limit on payment value.
  - Settles monetary policy and money market operations.
  - Every five days, TARGET2 processes a value close to the entire euro area GDP.
- Participation and ancillary systems (end-2018):
  - TARGET2 had 1,056 direct participants, 659 indirect Participants, and 4,091 correspondents.
  - TARGET2 settled the cash positions of 85 ancillary systems: 31 retail payment systems, 21 securities settlement systems and 23 clearing houses (including 4 CCPs), and 10 others.
  - Authors’ identified ancillary systems: 17 retail payment systems, 20 CSDs, 12 CCPs and 3 others.
- Daily activity and payment characteristics:
  - Daily averages: TARGET2 processed a daily average of 346,843 payments representing a daily average value of EUR 1.7 trillion.
  - Average transaction value: EUR 4.8 million.
  - Payment values distribution:
    - < 50,000 (70 percent of all TARGET2 payments)
    - > 1 million (10 percent of all TARGET2 payments)
  - Share large value payment traffic in euro: 90 percent in value and 63 percent in volume.
  - Payments processed in less than five minutes: 100 percent.
    - < 1 minute (99.9 percent of all enquiries)
    - 1 to 3 minutes (0.10 percent)

### III. Data coverage, aggregation, and netting considerations
- Dataset:
  - Covers all individual transactions settled in TARGET2 from 2007 to December 2018.
  - Aggregated to daily gross bilateral flows for each participant pair per payment subcategory.
  - Database levels available: i) account level; ii) institutions (aggregation of accounts belonging to one institution); iii) country level (aggregation by responsible central bank).
  - Analysis focuses on institutional level (level ii).
- Payment categories included in analysis:
  - Only payment types 1 and 3 are included: i) transactions between commercial banks; iii) payments with an ancillary system (AS) involved.
  - Category 1 generally reflects underlying economic value; Category 3 often reflects netted final positions in TARGET2 (underlying economic activity can be much larger).
- Indicative netting factors per FMI type (from Table 2)
  - ACH
    - Netting percentage: 98
    - Underlying economic impact factor: 50
    - Source: Expert judgement.
  - CCP
    - Netting percentage: 66
    - Underlying economic impact factor: 3
    - Source: Rosati and Secola, 2005
  - CSD
    - Netting percentage: 0
    - Underlying economic impact factor: 1
    - Source: Assumption that CCP settle securities on a net basis and CSDs settle securities on a gross basis.
  - Other (CLS, EURO1, EBA Step2)
    - Netting percentage: 96
    - Underlying economic impact factor: 25
    - Source: Stated by CLS and also used for EURO1 and EBA Step2.
  - Interbank payment
    - Netting percentage: 0
    - Underlying economic impact factor: 1
    - Source: By definition, no netting.
  - Note: Netting factors are indicative and may vary across CCPs and ACHs and from day to day; a netting factor may be close to 99 percent one day and not even 20 percent another day.

### IV. Composite indicator — definitions, measures, and construction
- Objective: measure the impact on the payment system when a bank or FMI cannot make payments in TARGET2.
- Liquidity risk measures:
  - Macro-level: Net Multilateral Flow (NMF)
    - NMF is the total amount of liquidity sent minus the liquidity received by the failing participant.
    - NMF formula (notation preserved): 푁푀퐹
      ௜
      = ෍푎
      ௜௝
      − ෍푎
      ௝௜
      for ௝ ≠ ௜
    - Interpretation: positive NMF implies the failing bank was supposed to pay more that day than it was about to receive; negative NMF implies incoming payments exceed outgoing.
  - Micro-level: Net Bilateral Flows (NBF)
    - NBF reflects remaining net liquidity flow between two participants after bilateral netting on a daily basis.
    - NBF can be positive (NBFpos) or negative (NBFneg) for a participant pair.
    - High NBF indicates “circular” payments not netted out bilaterally and reflects dependency of participants on each other’s liquidity.
- Systemic impact measure:
  - Degree measures (Dorogovtsev and Mendes (2003)): Out-Degree, In-Degree, Total Degree.
  - If failing participant has positive outflow (NBFpos) use out-degree; if negative (NBFneg) use in-degree.
  - Degree calculated for every participant for every day in three varieties.
- Composite (total) risk indicator:
  - Rationale: combine liquidity risk (NMF or NBF) and systemic impact (normalized degree) with equal weighting.
  - Total risk definition (as presented): 푇표푡푎푙 푟푖푠푘 ൌ√ሺሺ푙푖𝑞𝑢𝑖𝑑𝑖𝑡𝑦 𝑟𝑖𝑠𝑘ሻ ଶ ൅ሺ푠𝑦𝑠𝑡𝑒𝑚𝑖𝑐 𝑟𝑖𝑠𝑘ሻ ଶ ሻ ( 4 )
  - Normalization: largest euro value and largest degree observed in the data in a given quarter set to 1; all other values scaled against these largest observed values (range 0 to 1).
  - Graphical risk regions: circles centered at origin with radii 1/3, 2/3, and 1 mark low, medium, high risk regions; colors from green (near origin) to red (outside radius 1) map to risk levels.
- Interpretation links:
  - NBFs / NMF interpreted as “size”.
  - Degree interpreted as “interconnectedness”.
  - Comparable to systemic‑importance frameworks that consider size and interconnectedness (BCBS, 2013).

### V. Illustrative example — Matrix C and payment flow schematic
- Matrix C example where both NMF and NBF are unequal to zero for participant C1:
  - NMF = 55 (100 + 25 - 30 - 40 = 55).
  - NBFpos for this participant = 70 (with respect to participant C2).
  - NBFneg for this participant = 15 (with respect to participant C3).
- Implication: in failure, the balance of inflows and outflows is lost and external liquidity may be needed to settle payments.

### VI. Application to TARGET2 data — banks
- Visualization approach: NMF (or NBF) on x-axis, normalized degree on y-axis; each dot represents a bank group on a business day.
- Bank group findings (2017–2018 and 2018:Q4 subsets):
  - Light blue banks: near horizontal axis → overall low criticality (low liquidity risk and low systemic impact).
  - Red group: high degree (highly connected) and large variation in liquidity positions; often mid to high negative NMF values and some large positive NMFs — regularly highly critical.
  - Green group: intermediate connectivity and slightly higher liquidity risk than light blue.
- NMF vs NBF distinctions:
  - NMF measures system‑wide (multilateral) impact; NBF measures bilateral impacts on specific counterparts.
  - More dots with positive NMF and NBF than with negative values → more participants potentially face liquidity shortages than liquidity sinks.
- Temporal variation:
  - Behaviour not constant over time; December shows majority of outliers (potentially due to end‑of‑year balance sheet cleaning and higher payment activity).
  - Differences observed in 2018:Q4 compared with full-sample patterns (e.g., absence of some red outliers).

### VII. Application to TARGET2 data — FMIs (ACHs, CCPs, CSDs, other)
- Normalization uses same levels as banks to allow direct comparison.
- FMI observations:
  - CSDs vary widely in degree.
  - ACHs, CCPs, and other FMIs typically have relatively low degrees (under 0.33-radius).
  - CCPs: NMF and NBF sizes are very low (small payment instruction sizes); participation concentrated in largest banks.
  - ACHs: often paid in by fewer (large) participants than they pay out to.
- Netting-adjusted analysis:
  - FMIs apply multilateral netting; failing to adjust for netting can misrepresent FMI importance.
  - Figures adjusted for netting factors (Table 2) reveal ACHs and “other FMIs” (e.g., CLS and EURO1) as most critical in Q4 2018 after scaling for netting.
  - ACHs and “other FMIs” show high positive NMF and NBF and very large NBFpos and NBFneg — reflecting concentrated payers vs payees and large underlying economic impact when netting is undone.
  - CCPs and CSDs are less critical for the payment system in TARGET2: payment instruction sizes are relatively low and CCP participation is limited.
- Comparative conclusions:
  - ACHs and “other FMIs” can be more critical than banks due to centralization and payment concentration.
  - FMIs’ criticality is driven mainly by size of payment flows (after netting adjustment); banks’ criticality typically driven by degree (interconnectedness).
  - TARGET2 data limitations: difficulty in ascertaining whether TARGET2 participants represent broader economic actors (clients of banks).

### VIII. Policy implications and uses of the composite indicator
- Identification of critical participants:
  - Supports payment system operators in integrated risk view (PFMI Principle 3).
  - Enables monitoring and information sharing with bank and FMI supervisors.
  - Operational risk management: critical participants may require business continuity, incident management, and security measures aligned with PFMI Principle 17.
  - Financial risk management: informs precautionary measures (e.g., intraday credit limit policies tailored to participant criticality).
- Informing central bank liquidity provision to non-bank FMIs:
  - Composite indicator can feed cost‑benefit analysis for granting intraday/overnight liquidity access and determining appropriate liquidity types.
  - Considerations include monetary policy implementation, financial stability, market efficiency, neutrality, and central bank costs.
- Crisis management and contingency planning:
  - Indicator informs scenarios for business continuity drills and default‑management exercises.
  - Critical participants can be invited to exercises; findings can inform resolution planning and tailoring of resolution tools/strategies.
- Systemwide financial stability analysis and oversight:
  - Central bank oversight can use composite indicator data for monitoring and assessing FMIs and broader payment arrangements.
  - Helps understand links between FMIs and other financial institutions/markets to apply oversight objectives at system and individual FMI levels.

### IX. Conclusions and contribution
- The composite risk indicator:
  - Combines liquidity risk (NMF/NBF) and systemic impact (degree) to determine participant criticality in a payment system.
  - Is an informative and relatively straightforward data‑driven measure for banks and FMIs using payment transaction data.
  - Identifies critical participants whose default could cause significant problems for other market infrastructure participants.
- Practical value and contribution:
  - Provides regulators with an a‑priori indication of potential system‑wide impact from highly critical participants, supporting targeted interventions or procedures.
  - Adds to network analysis literature by incorporating FMIs and using transaction (flow) data rather than exposure (stock) data, enabling cross-sectional comparisons across FMI types.

*Source: wpiea2020081-print-pdf*

### References .............................................................................................................

### References

### Tables
- 1. TARGET2 Facts in 2018.
- 2. Indicative Netting Factor per FMI Type

### Figures
- 1. Schematic Example of Payment Flows in a Three-Participant System
- 2. Schematic Illustration of Measuring Level of Criticality
- 3. Composite Risk Indicator for Banks (2017–2018).
- 4. Composite Risk Indicator for Banks (2018:Q4)
- 5. Composite Risk Indicator for FMIs (2018)
- 6. Composite Risk Indicator for FMIs (2018:Q4)
- 7. Composite Risk indicator for FMIs, Adjusted for Netting Factor (2018:Q4)

### I. INTRODUCTION
- FMIs facilitate the clearing and settlement of financial obligations of financial institutions and their customers and must comply with the Principles for Financial Market Infrastructures (PFMI) (CPSS‑IOSCO, 2012).
- FMIs’ malfunction can significantly disrupt the financial system and be a source of contagion; examples of incidents:
  - Outages in the United Kingdom in 2014 (Bank of England, 2015).
  - Defaulting participants at Nasdaq Clearing in 2018.
  - DDos attack on Norges Bank in 2014 resulting in suspended services and subsequent liquidity exposures to other participants (Bouveret, 2018).
- Previous network analyses often omitted FMIs, assuming perpetual proper functioning; this paper includes FMIs as network participants.
- Focus of the paper:
  - The LVPS network measured as a “flow” per unit time (contrast with “stock” exposure networks).
  - Development of a composite risk indicator capturing both liquidity risk and systemic impact from the failure of a financial institution in an LVPS network.
  - Financial institutions include banks and FMIs that settle in the LVPS; nodes are participants, links are payment flows.
  - “Failing” defined as inability to make any payments due to operational problems, cyber‑attack, severe power outage, operational fraud, or bankruptcy.
- Measurement approaches:
  - Liquidity risk measured as the difference between incoming and outgoing payments at macro level (difference between all incoming and outgoing payments) or micro level (bilateral flows between two participants).
  - Systemic impact measured by the degree of a participant (its level of interconnectedness in the payment system).
  - The proposed indicator aims to give insight into the impact on participants, not their shock‑absorbing capacity.
- Correction for FMIs that settle net positions in TARGET2:
  - Adjust for the actual underlying gross flow of their settlement instructions to measure actual economic impact of failure.
  - Note: Some FMIs have a netting efficiency of 98 percent leading to an actual potential impact that is 50 times larger than the flows seen in the LVPS.
- Data and scope:
  - Transaction data from TARGET2 (the European real time gross settlement system (RTGS) for euro denominated payments) are used.
  - Data include transaction information of commercial banks and different types of FMIs settling their positions in TARGET2.
- Contributions and novelty:
  - Identifies TARGET2 participants (FMIs and commercial banks) that are critical and whose failure may cause substantial liquidity risk and/or systemic impact.
  - Novelty lies in inclusion of FMIs as participants and differentiation between flows observed in an LVPS and the actual underlying gross flows.
- Relation to prior literature and methods:
  - Builds on Berndsen and Heijmans (2020), Heijmans and Zhou (2019), Li and Perez‑Saiz (2018), Squartini, van Lelyveld, and Garlaschelli (2013).
  - Compares and contrasts with measures and studies such as SinkRank (Soramäki and Cook, 2013), Baek, Soramäki, and Yoon (2014), Arjani, Li, and Sabetti (2020), Pröpper, van Lelyveld, and Heijmans (2013), Heijmans, Heuver, Lelyveld, and Levallois (2016), León and Pérez (2014), Bardoscia, Bianconi, and Ferrara (2018).
  - Differs by combining liquidity risk and systemic impact and explicitly identifying the role of FMIs in an LVPS.
- Policy and practitioner relevance:
  - Method useful for payment system operators, overseers, financial stability experts, central banks, ESMA, CFTC, FSB studies, IMF, World Bank, and other institutions for network analysis, stress testing, and surveillance.
  - Examples of related supervisory exercises: ESMA and CFTC coordinated stress tests for CCPs (ESMA, 2018; CFTC, 2019); FSB Study Group on Central Clearing Interdependencies (FSB, 2017).
- Organization:
  - Section II provides an overview of the TARGET2 system and its data.
  - Section III describes the composite indicator for systemic and liquidity risk.

*Source: wpiea2020081-print-pdf - References .............................................................................................................*

### Section IV applies this composite indicator to TARGET2 data. Section V provides policy

### wpiea2020081-print-pdf - Section IV applies this composite indicator to TARGET2 data. Section V provides policy

### II. TARGET2 — The System and Its Participants
- TARGET2 is the LVPS and RTGS system for euro denominated payments owned and operated by the Eurosystem.
- All payments are settled one by one (gross) and on a continuous basis (real time) in central bank money with immediate finality.
- There is no upper or lower limit on the value of payments.
- TARGET2 also settles monetary policy and money market operations.
- Every five days, TARGET2 processes a value close to the entire euro area GDP.
- TARGET2 had over 1,000 direct participants, of which most are commercial banks; other participants include central banks of the Eurosystem, some central banks in the European Union not having the euro, and national treasuries.
- FMIs can participate in TARGET2 as ancillary systems (ASs) to settle (net) cash positions resulting from their clearing and settlement business. Ancillary systems that participate in TARGET2 include:
  - automated clearing houses (ACH);
  - CCPs;
  - the continuous linked settlement (CLS) system;
  - Euro1;
  - securities settlement systems (SSS) and central securities depositories (CSDs).
- At the end of 2018, TARGET2 settled the cash positions of 85 ancillary systems, including 31 retail payment systems, 21 securities settlement systems, and 23 clearing houses (including four CCPs).
- Notes on data labeling and classification differences:
  - ECB reports 31 retail payment systems; the authors’ data finds fewer due to not labeling different ancillary system interfaces (ASIs) as different payment systems.
  - ECB reports 4 CCPs; the authors’ data finds more because some CCPs have a bank license and are labeled by activity.
- Definitions and roles:
  - ACH: electronic clearing system for high-volume, low-value retail payments; typically settles on a deferred net basis with one or multiple settlement batches per day. Retail payments can also be settled through instant payment arrangements with direct settlement in TARGET2.
  - CCP: interposes itself between counterparties, becoming buyer to every seller and seller to every buyer; guarantees performance of open contracts; nets positions multilateral and results in net payment transfers in the RTGS system; settles cash collateral transfers (e.g., margin payments) through the payment system.
  - CLS: provides global multi-currency settlement services settling FX contracts on a multilateral net basis using a payment versus payment (PvP) mechanism; has access to central bank money in each eligible currency; euro funding delays can affect multi-currency settlement and other currency areas, in particular the Asia-Pacific region.
  - EURO1: LVPS for cross-border and domestic transactions in euro between banks operating in the EU; settles its positions at the end of the day on a net basis in TARGET2; uses a guarantee account mechanism if a settlement bank fails to meet its obligation in EURO1 end-of-day settlement.
  - CSD: operates an SSS enabling securities transactions to be processed and settled by book entry; provides custodial services; plays an active role in ensuring integrity of securities issues. In this report, CSD is defined as operator of an SSS. CSDs typically settle on a gross basis.

### II. TARGET2 — Key Facts and Statistics (2018)
- Number of participants: TARGET2 had 1,056 direct participants, 659 indirect Participants, and 4,091 correspondents.
- Number of ancillary systems: TARGET2 settled the cash positions of 85 ancillary systems: 31 retail payment systems, 21 security settlement systems and 23 clearing houses (including 4 CCPs), and 10 others.
- Number of ancillary systems identified in the authors’ data: 17 retail payment systems, 20 CSDs, 12 CCPs and 3 others.
- Daily averages: TARGET2 processed a daily average of 346,843 payments representing a daily average value of EUR 1.7 trillion.
- Average transaction value: EUR 4.8 million.
- Payment values distribution:
  - < 50,000 (70 percent of all TARGET2 payments)
  - > 1 million (10 percent of all TARGET2 payments)
- Share large value payment traffic in euro: 90 percent in value and 63 percent in volume.
- Payments processed in less than five minutes: 100 percent.
  - < 1 minute (99.9 percent of all enquiries)
  - 1 to 3 minutes (0.10 percent)
- Source for table: ECB.

### II. TARGET2 — Data Coverage and Aggregation
- Dataset covers all individual transactions settled in TARGET2 from 2007 to December 2018.
- TARGET2 payment categories (four main categories; main subcategories summarized):
  1. Transactions between commercial banks.
     - Subdivided into interbank payments on behalf of the bank’s own business or that of its client.
  2. Payments with a central bank involved.
     - Payment types include cash operations (bank notes) and payments between two central banks of the Eurosystem.
  3. Payments with an ancillary system (AS) involved.
     - Differentiates payments with i) an SSS and CCP; ii) an ACH; iii) EURO1; iv) CLS; v) EBA Step2.
  4. Liquidity transfers.
     - Includes subtypes such as internal and commercial transfers between different accounts of the same participant.
- For euro settlement, CLS holds an account with the ECB and receives and sends euro payments via TARGET2.
- Data aggregation:
  - Aggregated to daily gross bilateral flows for each participant pair per payment subcategory.
  - Database levels available: i) account level; ii) institutions (aggregation of accounts belonging to one institution); iii) country level (aggregation of accounts of all participants belonging to one responsible central bank in TARGET2).
  - This paper focuses on the institutional level (level ii).
- For analysis purposes, only payment types 1 and 3 are included (commercial banks and FMIs).
  - Category 1 (commercial bank flows) generally reflects the underlying economic value.
  - Category 3 (FMI flows) often reflect netting processes; only final positions after netting are settled in TARGET2, so actual underlying economic activity can be much larger than the settled flow.
  - Example: one euro not sent to CLS potentially means an actual missing value of EUR 100 due to netting efficiency.

### Table of Indicative Netting Factors per FMI Type (from Table 2)
- ACH
  - Netting percentage: 98
  - Underlying economic impact factor: 50
  - Source: Expert judgement.
- CCP
  - Netting percentage: 66
  - Underlying economic impact factor: 3
  - Source: Rosati and Secola, 2005
- CSD
  - Netting percentage: 0
  - Underlying economic impact factor: 1
  - Source: Assumption that CCP settle securities on a net basis and CSDs settle securities on a gross basis.
- Other (CLS, EURO1, EBA Step2)
  - Netting percentage: 96
  - Underlying economic impact factor: 25
  - Source: Stated by CLS and also used for EURO1 and EBA Step2.
- Interbank payment
  - Netting percentage: 0
  - Underlying economic impact factor: 1
  - Source: By definition, no netting.
- Note: The netting factors are indicative and may vary across CCPs and ACHs and from day to day; depending on business cleared, netting factor may be close to 99 percent one day and not even 20 percent another day.

### III. Composite Indicator for Liquidity Risk and Systemic Impact — Overview
- Purpose: measure the impact of a bank or FMI that is not able to make payments in TARGET2.
- Structure:
  - Section A: Liquidity risk measure.
  - Section B: Measure for systemic impact.
  - Section C: Composite risk indicator combining liquidity risk and systemic impact to determine a participant’s criticality to the payment system.
  - Section D: Connection to the common economic approach for defining systemically important financial institutions.

### III.A Liquidity Risk — Definitions and Measures
- Liquidity risk definition: risk that a failing participant in a payment system poses to other participants when it is not able to settle its payment obligations in full when due, although it may be able to do so in the future.
- Macro-level liquidity risk measure: Net Multilateral Flow (NMF).
  - NMF is the total amount of liquidity sent minus the liquidity received by the failing participant.
  - NMF formula (notation preserved):
    - 푁푀퐹
      ௜
      = ෍푎
      ௜௝
      − ෍푎
      ௝௜
      for ௝ ≠ ௜ (Equation (1) notation in source)
  - Interpretation:
    - If NMF is positive, the failing bank was supposed to pay more that day than it was about to receive — this is liquidity the system will be lacking even when other banks do not send payments to the failing bank.
    - If NMF is negative, the failing participant received more that day than it paid out to other participants.
    - NMF gives insight into the impact of a failing participant on system-wide liquidity.
- Micro-level (pairwise) measure: Net Bilateral Flows (NBF).
  - NBF of one participant pair reflects remaining net liquidity flow between these two participants if all gross payment flows are netted on a daily basis.
  - NBF can be positive or negative:
    - Positive NBF (NBFpos): on a bilateral net basis the failing participant has to pay out more than it receives.
    - Negative NBF (NBFneg): incoming payments for the failing participant are higher than the outgoing.
  - Definitions preserved from source:
    - NBFpos for bank i defined by Equation (2) (notation preserved in source).
    - NBFneg for bank i defined by Equation (3) (notation preserved in source).
  - High NBF (positive and/or negative) indicates “circular” payments in the system which cannot be netted out at a bilateral level — reflecting dependency of participants on each other’s liquidity.
- Behavioral effect: A participant unable to meet obligations causes liquidity shortages at its counterparts; liquidity sent to the failing participant temporarily disappears into a liquidity “sink” (Clarke and Hancock, 2012).

*Source: https://www.imf.org/-/media/files/publications/wp/2020/english/wpiea2020081-print-pdf.pdf*

### 90. Matrix C (and most realistic one) shows a situation in which both the NMF and the NBF

### wpiea2020081-print-pdf - 90. Matrix C (and most realistic one) shows a situation in which both the NMF and the NBF

### Matrix C example and payment flow schematic
- Matrix C example: both the NMF and the NBF are unequal to zero for a participant.
  - Example for participant C1: NMF = 55 (100 + 25 - 30 - 40 = 55).
  - NBFpos for this participant = 70 (with respect to participant C2).
  - NBFneg for this participant = 15 (with respect to participant C3).
- Normal circumstances: all in-flows and out-flows are “balanced,” or can be absorbed by available liquidity of participants.
- In failure: this balance is lost and external liquidity may be needed to get all payments settled.
- Figure 1: schematic example of payment flows in a three-participant system (matrices A, B, C shown).

### Systemic impact (degree measures)
- Systemic impact measured by number of participants affected by a failing participant (degree).
- Degree concepts (as defined by Dorogovtsev and Mendes (2003)):
  - Out-Degree: number of connections for outgoing payments.
  - In-Degree: number of connections for incoming payments.
  - Total Degree: number of connections for both incoming and outgoing payments.
- Use of degree in composite indicator:
  - If failing participant has positive outflow (NBFpos) use out-degree; if negative (NBFneg) use in-degree.
- Interpretation:
  - High degree implies many links and greater systemic impact from a participant’s failure.
  - Degree calculated for every participant for every day in three varieties: Out-Degree, In-Degree, Total Degree.

### Combined liquidity risk and systemic impact — composite risk indicator
- Objective: measure liquidity risk and systemic impact simultaneously to determine criticality using one composite risk indicator.
- Rationale: failing participants with many links and high NBFs/NMF are more critical.
- Equal weighting: liquidity risk and systemic risk are deemed equally important in the total risk indicator (footnote: weights can be adjusted).
- Total risk definition (as presented):
  - 푇표푡푎푙 푟푖푠푘 ൌ√ሺሺ푙푖푞푢푖푑푖푡푦 푟푖푠푘ሻ ଶ ൅ሺ푠푦푠푡푒푚푖푐 푟푖푠푘ሻ ଶ ሻ ( 4 )
- Normalization:
  - Largest euro value and largest degree observed in the data in a given quarter set to 1.
  - All other euro values and degrees scaled against these largest observed values, falling in the range between 0 (zero) and 1 (one).
  - Values close to 1 indicate relatively high liquidity risk or systemic impact.
- Graphical use (Figure 2):
  - Liquidity risk (NMF or NBF) on x-axis; systemic risk (normalized degree) on y-axis.
  - Circles centered at origin with radii 1/3, 2/3, and 1 mark low, medium, high risk regions.
  - Colors from green (near origin) to red (outside radius 1) correspond to low to high risk.
  - Radii thresholds can be set depending on expert preferences.

### Composite indicator interpretation and links to typical systemic-importance indicators
- Composite combines:
  - NBFs / NMF → interpreted as “size”.
  - Degree → interpreted as “interconnectedness”.
- Comparison to typical systemic indicators (BCBS, 2013): size, interconnectedness, cross-border activity, substitutability, complexity.
- TARGET2 context:
  - Size translates to liquidity (value of transactions) processed by participants.
  - Interconnectedness translates to number of counterparts.
  - Cross-border activity and substitutability are difficult to translate in TARGET2’s closed-loop euro payment environment.

### Application to TARGET2 — Interbank (banks) results
- Figure 3a (NMF vs normalized degree) and Figure 3b (NBF vs normalized degree) for three bank groups (light blue, green, red) across 2017–2018 business days; each dot is a group on a specific business day.
- Risk-region interpretation via half-circles with radius 1/3, 2/3, and 1:
  - Observations outside radius 1 considered outliers or potentially high-risk situations.
- Bank group findings:
  - Light blue banks: near horizontal axis → overall low criticality (low liquidity risk and low systemic impact).
  - Red group: high degree (highly connected) and large variation in liquidity positions; often mid to high negative NMF values and some large positive NMFs — regularly highly critical.
  - Green group: intermediate connectivity and slightly higher liquidity risk than light blue.
- NMF vs NBF differences:
  - NBF analysis shows red and green groups with no dots on vertical axis → positive outflow with part of the market and positive inflow with rest; failure would cause some counterparts to face liquidity shortages.
  - Dots outside outer circle with positive NMF and NBF: substantial liquidity risks if entities fail on those days.
  - Dots outside outer circle with negative NMF and NBF: risk of “liquidity sink” where incoming payments accumulate and are not forwarded.
- Aggregate observations:
  - Number of dots with positive NMF and NBF > number with negative NMF/NBF → more participants potentially face liquidity shortages than liquidity sinks.
  - Light blue group is balanced between shortages and sinks but systemic impact remains low.
- Temporal variation (Figure 4; 2018:Q4 subset):
  - Results largely similar to full sample but some differences (e.g., absence of a tuft of red outliers in Q4 2018).
  - Behaviour is not constant over time; temporary changes can indicate network shifts with larger liquidity risk/systemic impact.
  - Majority of outliers occurred in December — potentially due to end-of-year balance sheet cleaning and higher payment system activity.
- Distinction between NMF and NBF:
  - NMF measures impact to whole system (multilateral); NBF measures impact to individual counterparts (bilateral).

### Application to TARGET2 — FMIs results
- Figures 5–7 present FMIs (ACHs, CCPs, CSDs, other payment systems) for 2018; Figure 6 adjusts for netting factor (Table 2 referenced).
- Normalization uses same levels as banks to allow bank-FMI comparison.
- FMI-specific observations:
  - CSDs vary widely in degree.
  - ACHs, CCPs, and other FMIs typically have relatively low degrees (under 0.33-radius).
  - Participation in CCPs concentrated in largest banks (high access criteria).
  - NMF and NBF sizes for CCPs are very low (small payment instruction sizes).
  - ACHs often paid in by fewer (large) participants than they pay out to (Figure 5b: normalized degree higher when NBF is positive).
- Netting factor considerations:
  - FMIs apply multilateral netting, resulting in fewer and smaller settlement instructions in TARGET2.
  - If netting ratios not considered, FMIs’ importance may be misrepresented.
  - NMF and NBFs corrected for netting factor (Table 2) for Figure 7.
  - Alternative approach: change normalization to new largest values (principle remains unchanged).
- Figure 6 (2018:Q4) observation:
  - FMI business more stable month-to-month than banks; netting factor and relatively low account balances result in stability.
- Figure 7 (Adjusted for Netting Factor, 2018:Q4) observations:
  - Many outliers in NMF for ACHs beyond 1-radius — attributed to data noise (outgoing transactions incorrectly classified) and exaggeration when multiplied by factor 50.
  - Most critical FMIs in TARGET2 in Q4 2018: ACHs and “other FMIs” (e.g., wholesale systems CLS and EURO1).
    - High positive NMF and NBF reflect substantial payment flows and critical importance for system liquidity as individual participants.
    - Bottom graph of Figure 7: very large NBFpos and NBFneg for this group → participants who pay in are different from those who get paid by these FMIs.
  - CCPs and CSDs are less critical for the payment system in TARGET2:
    - Size of payment instructions following securities transactions is relatively low.
    - Derivatives payments often only reflect margin calls (a small percentage of exposure).
    - Number of CCP participants is relatively low due to concentration in large banks and strict access requirements.
- Comparative conclusions (banks vs FMIs):
  - ACHs and “other FMIs” are more critical than banks (expected due to centralization and payment concentration).
  - FMIs’ criticality driven mainly by size of payment flows rather than degree.
  - Banks’ criticality typically driven by degree (interconnectedness), e.g., correspondent banks.
  - TARGET2 data limitations: difficulty in determining whether TARGET2 participants represent broader economic actors that are clients of banks.

### Policy implications
- Composite risk indicator uses and benefits:
  1. Identification of critical participants in a payment system:
     - Supports payment system operator in integrated view of risks (PFMI Principle 3).
     - Enables monitoring, information sharing with bank and FMI supervisors.
     - Operational risk management: critical participants may need similar operational requirements as the payment system (business continuity, incident management, security) (PFMI Principle 17).
     - Financial risk management: identification informs precautionary measures (e.g., limits on intraday credit extensions tailored to participant criticality).
  2. Informing central bank liquidity provision to non-bank FMIs:
     - Access policies to intraday/overnight liquidity for FMIs are heterogeneous across central banks.
     - Composite indicator can feed cost-benefit analysis for granting liquidity access and determining appropriate liquidity type.
     - Considerations include impacts on monetary policy implementation, financial stability, market efficiency, neutrality, and central bank costs.
  3. Crisis management and contingency planning:
     - Indicator informs scenarios for business continuity drills and default-management exercises.
     - Critical participants can be invited to exercises; findings can inform resolution planning and tailoring of resolution tools/strategies.
  4. Systemwide financial stability analysis and oversight:
     - Central bank oversight of payment, clearing and settlement systems uses composite indicator data for monitoring and assessing FMIs and broader payment arrangements.
     - Helps understand links between FMIs and between FMIs and other financial institutions/markets to apply oversight objectives at system and individual FMI levels.

### Conclusions
- The composite risk indicator:
  - Combines liquidity risk and systemic impact through interconnections to determine individual participant criticality.
  - Is an informative and relatively easy data-driven measure for banks and FMIs in a payment system.
  - Identifies critical participants whose default could cause significant problems for other market infrastructure participants.
  - Is a useful tool for central banks in payment system operations, oversight, and financial stability analysis.
- Practical value:
  - Provides regulators an a-priori indication of potential system-wide impact from highly critical participants.
  - Detection of very critical participants may lead to different interventions or procedures to prevent large systemic damage.
- Contribution to literature:
  - New addition to network analysis for FMIs by using payment transaction data (rather than exposure data) to map interconnectedness and liquidity risks.
  - Enables comparison of systemic importance between banks and FMIs and across FMI types, yielding cross-sectional insights on interconnections in a payment system.

*Source: wpiea2020081-print-pdf (excerpt content provided).*

### REFERENCES

### REFERENCES

### Measurement and monitoring of payment and money markets
- Arciero, L., R. Heijmans, R. Heuver, M. Massarenti, C. Picillo, and F. Vacirca (2016). How to Measure the Unsecured Money Market? The Eurosystem’s Implementation and Validation using TARGET2 Data. International Journal of Central Banking March, 247–280.
- Heijmans, R., R. Heuver, and D. Walraven (2010). Monitoring the unsecured interbank money market using target2 data. De Nederlandsche Bank Working Paper 276.
- Heijmans, R. and R. Heuver (2014). Is this bank ill? The diagnosis of doctor TARGET2. Journal of Financial Market Infrastructures 2(3), 3–36.
- Heijmans, R., R. Heuver, I. Lelyveld, and C. Levallois (2016). Dynamic visualization of large financial networks. Journal of Network Theory in Finance 2(2), 1–23.
- Rosati, S. and S. Secola (2005). Explaining Cross-border Large-value Payment Flows - Evidence from TARGET and Euro 1 Data. ECB Working Paper Series 443.
- Soramäki, K., W. Beyeler, M. Bech, and R. Glass (2007). New Approaches for Payment System Simulation Research. In H. Leinonen (Ed.), Simulation studies of liquidity needs, risks and efficiency in payment networks., Proceedings from the Bank of Finland Payment and Settlement System Seminars 2005, pp. 15–40.
- Pröpper, M., I. van Lelyveld, and R. Heijmans (2013). Network Dynamics of TOP Payments. Journal of Financial Market Infrastructures 1(3), 3–29.

### Detection and analysis of operational outages and intraday liquidity risk
- Arjani, N. and R. Heijmans (2020). Detecting operational outages from lvts transaction data. Payments Canada Discussion Paper, February.
- Glowka, M., J. Paulick, and I. Schultze (2018). The absence of evidence and the evidence of absence: an algorithmic approach for identifying operational outages in target2. Journal of Financial Market Infrastructures 6(2/3), 63–91.
- Klee, E. (2010). Operational outages and aggregate uncertainty in the federal funds market. Journal of Banking and Finance 34(10), 2386–2402.
- Arjani, N., F. Li, and L. Sabetti (2020). Monitoring intraday liquidity risk in a real time gross settlement system. Payments Canada Discussion Paper, forthcoming.
- Clarke, A. and J. Hancock (2012). Payment system design and participant operational disruptions. Reserve Bank of Australia, Research Discussion Paper 2012-05.
- Heijmans, R. and C. Zhou (2019). Outlier detection in target2 risk indicators. DNB Working Paper (624).

### Network analysis, systemic importance, and early-warning signals
- Baek, S., K. Soramäki, and J. Yoon (2014). Network indicators for monitoring intraday liquidity in BoK-wire+. Journal of Financial Market Infrastructures 2(3), 37–66.
- Bardoscia, M., G. Bianconi, and G. Ferrara (2018). Multiplex network analysis of the UK OTC derivatives market. Bank of England Working Paper 726.
- León, C., R. Berndsen, and L. Renneboog (2014). Financial stability and interacting networks of financial institutions and market infrastructures. Borradores de Economia 848.
- León, C. and J. Pérez (2014). Assessing financial market infrastructures’ systemic importance with authority and hub centrality. Journal of Financial Market Infrastructures.
- Li, F. and H. Pérez-Saiz (2018). Measuring systemic risk across financial market infrastructures. Journal of Financial Stability 34, 1–11.
- Squartini, T., I. van Lelyveld, and D. Garlaschelli (2013). Early-warning signals of topological collapse in interbank networks. Scientific Reports 3(3357), 1–9.
- Sinkrank: Soramäki, K. and S. Cook (2013). Sinkrank: An algorithm for identifying systemically important banks in payment systems. Economics: The Open-Access, Open-Assessment E-Journal 28(7).

### Central counterparties, stress testing, and oversight frameworks
- CFTC (2019). CCP supervisory stress tests: Reverse stress test and liquidation stress test. CFTC Staff Report April.
- ESMA (2018). Eu-wide CCP stress test 2017. ESMA Report ESMA70-151-1154.
- Wendt, F. (2015). Central counterparties: Addressing their too important to fail nature. Journal of Financial Market Infrastructures 4(1), 59–84.
- FSB (2017). Analysis of central clearing interdependencies. With BCBS, CPMI and IOSCO.
- CPSS-IOSCO (2012). Principles for Financial Market Infrastructures. Bank for International Settlements and International Organization of Securities Commissions.
- CPSS (2003). The Role of Central Bank Money in Payment Systems. Bank for International Settlements.
- CPSS (2005). Central bank oversight of payment and settlement systems. Bank for International Settlements.
- Bank of England, B. (2015). Independent review of RTGS outage on 20 October 2014: Bank of England’s response.

### Cyber risk, quantitative frameworks, and country assessments
- Bouveret, A. (2018). Cyber risk for the financial sector: A framework for quantitative assessment. IMF Working Paper.
- IMF (2018), Jamaica Financial Sector Stability Assessment, IMF Country Report No. 18/347.

*Source: wpiea2020081-print-pdf - REFERENCES*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2020/english/wpiea2020081-print-pdf.pdf_