Selected Issues

Abstract

Selected Issues

Interconnections within Panama’s Regional Banking Sector and with Foreign Banks1

This paper assesses the potential magnitude of the risks related to interbank exposures within the Panamanian banking sector and Panama’s connections with foreign banking systems through both credit and funding channels are potential sources of systemic risk to Panama. On balance, the results suggest that systemic risks from interconnections within Panama’s regional banking sector are moderate. While banks maintain relatively important interbank connections, primarily through deposits, banks’ existing capital buffers appear broadly sufficient to absorb contagion from the failure of individual banks in the system without the need for sizeable capital injections. However, caution is warranted as economic or financial shocks that simultaneously affect all banks are beyond the scope of this analysis and would heighten systemic risks associated with banks’ interconnections. Panama maintains significant upstream and downstream exposures to foreign banking systems and credit provision in Panama would be materially affected by foreign banks’ deleveraging.

A. Introduction

1. Exposures between individual banks and across banking systems can have important implications for financial stability. Disruptions to such financial linkages, even at the level of individual banks, can affect the stability of an economy’s entire banking system. Similarly, challenges faced by foreign banks operating abroad can spillover to other countries directly through both funding and credit channels. This paper assesses the potential magnitude of these risks for Panama. First, potential risks to Panama’s regional banking center from banks’ connections with other banks operating in Panama as well as foreign banks are assessed. Second, potential risks to Panama from its exposure to foreign banks through both credit and funding channels are considered.

B. Interbank Exposures in the Panamanian Banking System2

Stylized Facts

2. Interbank connections in the Panamanian banking system are relatively important. Banks are connected through their interbank deposits and their investment holdings (both through bonds and cross-equity holdings). Banks’ interbank deposits are a relatively small share of their total assets: of banks operating in the onshore banking sector, interbank deposits account for only about 4 percent of Panamanian-owned banks’ assets or about 9 percent of foreign-owned banks’ assets as of March 2018. As a share of total assets, the offshore banking sector is much more exposed, with about 30 percent of assets held in deposits in other banks. As a share of banks’ equity, banks’ interbank exposures are much more important: in the onshore banking sector, interbank deposits account for about 115 percent of Panamanian-owned banks’ total equity and 76 percent of foreign banks’ equity.3 Again, offshore banks’ exposure to their interbank deposits is also more important as a proportion of their equity at about 165 percent. Thus, although banks have relatively little of their assets concentrated in their interbank holdings, interbank deposits may still be an important transmission channel for bank stress given their importance as a percent of banks’ equity. By contrast, banks’ investment holdings in the bonds, stocks, other financial instruments of other banks are relatively small both as a share of banks’ equity and of their total assets (Table 2).

Table 1.

Panama: Interbank Deposits by Type of Bank License1/

article image
Source: SBP and author calculations.

GL stands for “General license, local”, GF for “General license, foreign”, IL for “International license”. Data is as of March 31, 2018.

Table 2.

Panama: Investment Holdings of Banks by Type of Bank License1/

article image
Source: SBP and author calculations.

GL stands for “General license, local”, GF for “General license, foreign”, IL for “International license”. Data is as of March 31, 2018.

3. Interbank deposits are relatively concentrated in foreign banks. On aggregate, over 80 percent of banks’ interbank deposits are held in foreign banks. While this ratio is boosted by the offshore banks, which are restricted in their ability to interact with the onshore banking system, Panamanian-owned onshore banks hold over 70 percent of their interbank deposits in foreign banks and foreign-owned offshore banks hold about 80 percent of their interbank deposits in foreign banks. Banks’ interbank deposits within the domestic onshore banking system are more limited. The concentration of bank’s interbank deposits within a few origin banks or destination banks also provides a preliminary indication of banks’ vulnerability to losses on their interbank deposits. Annex Tables 1, 2, and 3 show the composition of banks’ interbank deposits by origin bank and by destination bank as a percent of total interbank deposits in the Panamanian banking system (including both the onshore and offshore banks). By origin banks, while most banks account for a limited share of banks’ total interbank deposits, 4 banks hold shares in excess of 5 percent of total interbank deposits, suggesting higher vulnerability to losses on these deposits. However, for at least two of these banks, these deposits may be in their parent banks abroad. By destination bank, interbank deposits are even more diverse. Exposures within the Panamanian banking system are relatively small as a share of total interbank deposits, except for those to the Banco Nacional, which plays an important role in the operation of the payments system. On aggregate, the Panamanian banking system holds deposits in 128 banks abroad, but the bulk of these deposits are concentrated in the ten banks with the largest exposures: these banks hold about over 50 percent of Panamanian banks’ deposits abroad.

4. The degree to which banks are vulnerable to interbank contagion depends on their financial health in addition to their interbank exposures. Aggregate financial soundness indicators suggest that the Panamanian banking system, on aggregate, is well capitalized. Banks generally maintain relatively large capital buffers in excess of the minimum regulatory requirement of 8 percent of risk-weighted assets, but there is significant dispersion in the size of such buffers across individual banks. Non-performing loans remain low, but have been rising, and, while NPLs appear well provisioned on aggregate and has improved with the adoption of IFRS9, provisioning coverage varies significantly across banks. Similarly, while the banking system is profitable, larger banks by asset size tend to be more profitable than smaller banks, likely in part reflecting economies of scale. On aggregate, banks’ leverage appears contained and liquidity sufficient, but again with wide dispersion across banks. Annex I provides detailed heat-maps of individual banks’ financial health based on their capital adequacy, asset quality, profitability, liquidity, and leverage.

Methodology

5. The network-based approach of Espinosa-Vega and Sole (2010) is used to assess potential risks to the Panamanian banking system from banks’ interbank balance sheet exposures. The simulations examine the potential domino effects triggered by the failure of a bank on its interbank obligations. The approach sequentially simulates the failure of each bank in the Panamanian banking system, with each bank’s failure affecting those banks to whom it maintains interbank connections through both credit and funding channels. The failing bank affects its creditors through the credit channel as its creditor banks lose a fraction λ of their deposits and its borrowers through the funding channel as its borrowers can replace only a fraction (1-ρ) of the funding they were getting from the now failed bank and must restore their balance-sheet identify by selling assets at a (fire-sale) price of $1/(1+δ) on the dollar. Through these channels, the failure of a single bank can trigger a sequence of bank failures and defaults within the banking system through multiple contagion rounds, with the model continuing to simulate these interbank spillovers until there are no further bank failures.

6. The magnitude of interbank spillovers and the cascade of bank failures triggered by the failure of an individual bank depends on the depth of interbank linkages through credit and funding channels and the assumed severity of these shocks. Following the baseline simulations in Espinosa-Vega and Sole (2010), the simulations assume a loss given default (the parameterλ) of 1 for creditor banks. This is a severe scenario in which banks are unable to recover any of their loans, reflecting the substantial uncertainty banks face over recovery rates in the immediate aftermath of a credit event For Panama, it also reflects the potentially high uncertainty related to bankruptcy resolutions given existing gaps in the bank resolution framework. Borrower banks are assumed to be able to roll-over 65 percent of the funding previously received from the now failed bank (i.e. ρ = 0.35), which triggers a fire sale of assets by these banks at a 50 percent discount (i.e. δ = 1).4 In practice, the extent to which banks are able to replace an unforeseen withdrawal of interbank funding will depend on money market liquidity conditions. By design, the model excludes the possibility of banks raising new capital.

7. Banks may become insolvent even before their capital is fully depleted. As contagion spreads throughout the banking system, a bank may fail even before its capital is fully depleted if the severity of the shock forces the bank to file for bankruptcy or the supervisor, recognizing the severity of the shock, intervenes early to arrest further transmission of the shock throughout the system. Therefore, three scenarios are considered for the threshold of capital under which a bank would become insolvent. First, a high-sensitivity scenario that assumes banks default when their CAR falls below the regulatory threshold of 8 percent. Second, a medium-sensitivity scenario that assumes banks default when their CAR falls below 4 percent. Finally, a low-sensitivity scenario that assumes banks default only when their capital is fully depleted (or CAR<0). The various degrees of sensitivity of banks’ solvency to their capital is consistent with experience in past banking crises -the threshold depends on country- and market-specific factors including the prevailing regulatory environment and ex-ante stability of the banking system.

8. The release of dynamic provisions may also affect the extent of contagion from banks’ interbank exposures. Panama introduced dynamic provisioning requirements in 2013 (with application starting in 2014).5 Each bank is required to maintain additional capital or dynamic provisions, over and above the regulatory requirement of 8 percent of risk-weighted assets, based on its outstanding loans (on a risk-weighted basis) and the quarterly change in the amount of its risk-weighted loan exposures and quarterly variation in specific provisions.6 Effectively, the dynamic provisioning requirements in Panama act as an additional capital buffer that is included in its capital, but cannot be used to meet the minimum 8 percent regulatory requirement. Draw-down of this buffer is restricted and subject to the discretion of the SBP.7 Therefore, banks’ dynamic provisions may reduce the extent of contagion from banks’ interbank exposures, but, in practice, the effectiveness of dynamic provisions will depend on the ability of the SBP to determine the appropriate commencement of the draw-down phase. To examine the potential impact of dynamic provisions on stemming interbank contagion, the three sensitivity scenarios are considered both including and excluding the impact of banks’ dynamic provisions on their capital.

9. Bank capital losses are classified into three categories: buffer losses, injection required to restore CAR, and excess losses. Following the default of an individual bank, the resulting capital losses of all other banks in the system can be decomposed based on whether banks’ capital remains above the solvency threshold. The portion of each banks’ total capital loss that reflects a reduction in its capital buffer (i.e. above the regulatory minimum of 8 percent) is classified as a buffer loss. Injection required to restore CAR is the amount of capital required to restore CAR to the 8 percent regulatory minimum (or the difference between 8 percent of risk-weighted assets and actual capital). Finally, any additional losses for banks with capital below the solvency thresholds assumed in the three scenarios is classified as excess loss.

Data

10. The interbank contagion analysis is based on data as of March 2018:

  • Banks’ interbank exposures are based on banks’ interbank deposits as reported to the SBP in banks’ weekly liquidity report. Banks report on a weekly basis their deposits in other depository institutions, both in Panama and abroad. As only banks with Panamanian operations are required to report, the data do not include information on foreign banks’ deposits in Panama. In total, the interbank deposit matrix includes data from 75 banks operating in Panama, of which 49 have general banking licenses and operate in the onshore banking system, and 26 have international banking and are part of the offshore banking system. Of the general license banks, 19 are owned by residents and 30 are foreign-owned banks licensed to operate in Panama.

  • Banks’ capital adequacy is calculated using banks’ reported regulatory capital and risk-weighted assets. However, regulatory capital and risk-weighted assets data are only available for these banks for which the SBP is the supervisor of origin. For the banks for which the SBP is the host supervisor, assumptions were made as to their capital adequacy for the purposes of the interbank contagion simulations. Risk-weighted assets for these banks (including both onshore and offshore) were estimated by multiplying their unweighted assets by the average ratio of risk-weighted assets to the unweighted assets of the banks under the SBP’s supervision (based on their balance sheet information provided to the SBP). A similar approach was taken with regulatory capital. For those banks for which these assumptions would leave their CAR below 8 percent, their capital was set at the median CAR of the other banks.8

Results

11. The failure of an individual bank cascades through the banking system through capital losses and defaults of other banks. The capital losses and number of defaults resulting from the failure of each bank in the system are plotted in Figure 2.9 For each of the three sensitivity scenarios, the results show that system-wide capital losses from the failure of an individual bank are only weakly related to the number of additional bank failures caused by the failure of that bank as it defaults on its interbank liabilities. But, both the capital loss of the banking system as well as the number of contagious defaults are important indicators of the banking system’s resilience to interbank contagion. The information on potential capital losses is critical to assess potential capital needs and sovereign exposure to contingent liabilities, while that on the number of defaults is important to gauge the risk of a banking crisis. In the high-sensitivity scenario, 6 simulations generate losses greater than US$1.5 billion (or about 1 percent of total consolidated banking system assets and 3.7 percent of 2017 GDP). While a maximum of 7 contagious defaults takes place in the worst of these simulations, the failure of 5 of these banks results in zero contagious defaults. As suggested in Section B, the failure of banks that receive large deposits from comparatively few banks tend to be the riskiest in term of fueling contagion, causing the failure of other banks and increasing the threat of a systemic banking crisis. By contrast, those banks that receive deposits from a large number of banks may be less systemic in the event that these deposits are relatively small compared to the capital of the depositor bank. Results from the medium- and low-sensitivity scenarios suggest a lower degree of contagion within the banking sector from the failure of any individual banks. In each case, while 6 simulations still generate losses in excess of US$1.5 billion, the maximum number of contagious defaults amongst all scenarios is 3 relative to the maximum of 7 in the high-sensitivity scenario.

Figure 1:
Figure 1:

Panama: Panamanian Bank Performance by Bank License Type

Citation: IMF Staff Country Reports 2019, 012; 10.5089/9781484394069.002.A004

Figure 2.
Figure 2.

Panama: Total Capital Losses and Number of Defaults

(based on the three scenarios related to the sensitivity of bank failures to CAR)

Citation: IMF Staff Country Reports 2019, 012; 10.5089/9781484394069.002.A004

12. The failure of foreign banks is more likely to result in contagion to banks in the Panamanian banking system. Figure 3 shows, for each scenario, total capital losses and the number of defaults caused by the 30 banks that inflict the largest total losses on the system. The license type of the original defaulting bank is displayed in lieu of its name for confidentiality reasons. Of these banks, it is the default of foreign banks that results in the largest number of contagious defaults with the exception of the default of one local Panamanian bank, whose default in the high-sensitivity scenario results in the failure of 7 additional institutions. In contrast, it is the failure of banks operating in the Panamanian banking system (3 Panamanian-owned, one offshore, and two foreign-owned banks) that account for the largest capital losses for the banking system. The largest capital loss is caused by the failure of a Panamanian-owned bank, causing losses of US$2.8 billion or 7 percent of GDP in the high-sensitivity scenario. However, with the relatively strong ex-ante capital position of banks, the losses generated by the failure of banks are largely absorbed by banks’ existing capital buffers. This is true even in the scenarios in which banks’ additional capital buffers from their dynamic provisions are not counted toward their capital (see Figures 4 and 5). Consistent with the concentration of banks’ deposits in foreign banks, the largest capital injections required result from the failure of foreign banks.10

Figure 3.
Figure 3.

Panama: Capital Losses as a Function of Original Defaulting Bank 1/2/

(US$ millions, 30 most systemically important institutions)

Citation: IMF Staff Country Reports 2019, 012; 10.5089/9781484394069.002.A004

1/ Each column displays total capital losses in the banking system (excluding those of the original defaulting bank) following the exogenous default of one bank (the license type of the original defaulting bank is shown in lieu of its name). Buffer losses are those over and above 8 percent of RWA, whereas injection to restore CAR represents the difference between 8 percent of RWA and actual capital after contagion. For defaulting banks (i.e. those whose capital is below 8, 4, and 0 percent of CAR in the high-, medium-, and low-sensitivity scenarios, respectively), if remaining capital is positive then that amount is classified as excess loss.2/ Total number of failures excludes original defaulting institution.
Figure 4.
Figure 4.

Panama: Total Capital Losses and Number of Defaults Excluding Capital from Dynamic Provisions

(based on the three scenarios related to the sensitivity of bank failures to CAR)

Citation: IMF Staff Country Reports 2019, 012; 10.5089/9781484394069.002.A004

Figure 5:
Figure 5:

Panama: Capital Losses as a Function of Original Defaulting Bank Excluding Capital from Dynamic Provisions 1/2/

(US$ millions, 30 most systemically important institutions)

Citation: IMF Staff Country Reports 2019, 012; 10.5089/9781484394069.002.A004

1/ Each column displays total capital losses in the banking system (excluding those of the original defaulting bank) following the exogenous default of one bank (the license type of the original defaulting bank is shown in lieu of its name). Buffer losses are those over and above 8 percent of RWA, whereas injection to restore CAR represents the difference between 8 percent of RWA and actual capital after contagion. For defaulting banks (i.e. those whose capital is below 8, 4, and 0 percent of CAR in the high-, medium-, and low-sensitivity scenarios, respectively), if remaining capital is positive then that amount is classified as excess loss.2/ Total number of failures excludes original defaulting institution.

13. Total capital losses are similar regardless of the sensitivity threshold considered. Total capital losses are US$24.9 billion in the high-sensitivity scenario, US$ 24.6 billion in the medium-sensitivity scenario, and US$24.5 billion in the low-sensitivity scenario. These capital losses are broadly comparable in the scenarios when banks’ capital buffers from dynamic provisions are not available. The similarity of the estimated capital losses reflects the fact that the most vulnerable banks (those which fail as a result of the failure of a high number of other banks) tend to be the least contagious (banks whose failure would cause a high number of other banks to fail) and, as a result, the banks affected in the three scenarios are broadly comparable.

14. The onshore banking system remains relatively isolated from the offshore banking system. With little interbank deposits between the onshore and offshore banking systems, the impact of the failure of an offshore bank has little impact on banks in the onshore financial system (see Annex III, Figures 1, 2 and 3). While some offshore banks do feature amongst the most systemic in the system (defined by size of capital losses caused by their default) the impact of the default of these banks is almost entirely on other offshore banks (or potentially foreign banks abroad), with no banks in the onshore system defaulting due to the default of an offshore bank.

C. Spillovers from Stress in International Banks11

Methodology

15. The IMF Bank Contagion Model is used to assess the exposure of Panama’s regional banking center to foreign banks. As a first step, the model is used to measure the downstream and upstream exposures of Panama’s banking sector. Downstream exposures capture Panama’s vulnerability to crises in countries that borrow from its banks, including potential losses on direct cross-border lending, off-balance sheet accounts, and affiliates’ claims. Upstream exposures capture rollover risks from Panamanian residents’ borrowing from international banks and the proportion of lending by foreign affiliates that were funded by their parent bank. The analysis is based on BIS banking statistics and bank-level data as of 2017Q4. This data provides an important complement to the analysis based on banks’ interbank deposits in Section C, which cannot fully assess foreign funding risks as the dataset does not include foreign banks’ deposits in or lending to Panamanian banks.

16. The IMF Bank Contagion Model is also used to assess the potential propagation of financial shocks from foreign banks to Panama through bank losses and deleveraging.12 Based on Panamanian banks’ identified upstream exposures, the model simulates several rounds of asset and funding shocks. The first round considers foreign bank losses on asset that partially or fully deplete their capital. These losses are calculated based on an assumed 10 percent loss in the value of banks’ private and public sector assets in selected BIS-reporting countries. In the second round, if banks do not have sufficient capital buffers to cover the losses, they restore their capital ratios by uniformly deleveraging across both domestic and external claims, thus, in the third round, banks reduce their lending to other banks including those in Panama and other countries, causing funding shocks to these banks and further deleveraging. Final convergence is achieved when no further deleveraging needs to occur. This analysis provides an indication of the potential impact on credit availability in Panama from foreign banks’ deleveraging versus the impact on banks’ capital positions considered in Section C.

Results

17. Panamanian banks maintain sizeable downstream and upstream exposures to foreign banking systems (Table 3). Total downstream exposures amount to over 30 percent of GDP. These high bilateral downstream exposures imply significant potential credit losses from Panamanian banks’ lending to foreign clients. Downstream exposures are concentrated primarily to the United States and to neighboring economies in Central and South America. Upstream exposures are even more significant at over 50 percent of GDP, implying that potential funding risks from Panamanians’ foreign borrowing are larger than credit risks from Panamanian banks’ lending to foreign clients.13 These exposures capture the upper bound of rollover risks to Panama from the loss of credit by BIS-reporting banks to Panamanian borrowers. Panama’s largest upstream exposures are to a handful of Asian countries, the most important of which is Japan, the United States and Canada and several European economies.

Table 3.

Panama: Panamanian Bank’s Downstream and Upstream Exposures to Foreign Banking Systems

article image

18. Foreign credit availability to Panama would be materially affected by severe shocks to foreign banks’ balance sheets (Table 4). For example, a combined 10 percent loss on assets of BIS-reporting banks in Canada and the United States would reduce credit in Panama by about 11 percent of GDP. Based on the identified upstream exposures, the most sizeable impact on foreign credit availability for Panamanian borrowers would stem from losses on Japanese assets, which would reduce credit availability in Panama by over 13 percent of GDP. These calculations do not take into account the amount of local stable funding for foreign banks from deposits in Panama, which would provide some cushion against banks’ need to deleverage in Panama.

Table 4.

Panama: Spillovers to Panama from Upstream Exposures to Foreign Banks

article image
Source: IMF, Research Department Macro – Financial Division Bank Contagion Module based on BIS, ECB, IFS, and Fitchconnect data.

Percent of on – balance sheet claims (all borrowing sectors) that default

Reduction in foreign banks’ redit due to impact of the shock on their balance sheet, assuming uniform deleveraging across domestic and external claims. All simulations are based on 2017Q4 data.

Greece, Ireland, Portugal, Italy, Spain, France, Germany, Netherlands, and the UK

These lender countries stopped disclosing these bilateral positions

D. Conclusions

Both interbank exposures within the Panamanian banking sector and Panama’s connections with foreign banking systems through both credit and funding channels are important potential sources of systemic risk to Panama. On balance, results from the network analysis of interconnections within Panama’s regional banking sector and with foreign banks suggest that systemic risks from these interconnections are moderate. While banks maintain relatively important interbank connections, primarily through deposits, banks’ existing capital buffers appear broadly sufficient to absorb contagion from the failure of individual banks in the system without the need for sizeable capital injections. However, caution is warranted as economic or financial shocks that simultaneously affect all banks are beyond the scope of this analysis and would heighten systemic risks associated with banks’ interconnections. Results from the assessment of Panama’s exposures to foreign banking system suggest that Panama maintains significant upstream and downstream exposures to foreign banking systems and credit provision in Panama would be materially affected by foreign banks’ deleveraging. Results of both exercises highlight the importance of strong supervision and regulation to mitigating associated risks, including by continuing to build banks’ ex-ante capital buffers through macroprudential policy.

References

  • Cerutti, E., S. Claessens, and P. McGuire (2012), “Systemic Risks in Global Banking: What can Available Data tell us and What More Data are Needed?BIS Working Paper 376, Bank for International Settlements.

    • Search Google Scholar
    • Export Citation
  • Espinosa-Vega, M. and J. Sole (2010), “Cross-Border Financial Surveillance: A Network Perspective”, IMF Working Paper WP/10/105.

  • Ong, L. L., P. Jeasakul, and S. Kwoh (2013), “HEAT! A Bank Health Assessment Tool”, IMF Working Paper WP/13/177.

Annex I. Bank’s Interbank Exposures by Origin and Destination Bank

Annex Table 1:

Interbank Deposits by Origin Bank to Destination Bank by Bank Type1/

(in percent of total interbank deposits)

article image

GL stands for “General license, local”, GF for “General license, foreign”, IL for “International license”. Data is as of March 31, 2018. Destination banks are classified as follows: 1=GL, 2=GF, 3=IL, 4=abroad.

Annex Table 2.

Interbank Deposits by Destination Bank within Panamanian Banking System from Origin Bank by Bank Type1/

(in percent of total interbank deposits)

article image

GL stands for “General license, local”, GF for “General license, foreign”, IL for “International license”. Data is as of March 31, 2018. Origin banks are classified as follows: 1=GL, 2=GF, 3=IL, 4=abroad.

Annex Table 3.

Interbank Deposits by Destination Bank Abroad System from Origin Bank by Bank Type 1/

(in percent of total interbank deposits)

article image

Excluding 98 destination banks that hold less than 0.5 percent of total interbank deposits. Origin banks are classified as follows: 1=GL, 2=GF, 3=IL, 4=abroad where GL stands for “General license, local”, GF for “General license, foreign”, IL for “International license”.

Annex II. Heatmaps of Banks’ Financial Performance14

Annex Table 1:

Capital Adequacy 1/

(capital in percent of risk-weighted assets)

article image

High vulnerability banks (in red) are defined as those with CAR of less than the 8 percent of risk-weighted assets minimum regulatory requirement, medium-vulnerability banks (in yellow) are those with CAR of less than the consolidated CAR for the national banking system, low vulnerability banks (in green) are those with CAR greater than the consolidated CAR for the banking system. Heat map is shown only for general license banks that report CAR to the Superintendent of Banks and excludes international (offshore) license banks. GL stands for “General license, local” and GF for “General license, foreign”. Bank ordering is based on banks’ asset size.

Annex Table 2:

Asset Quality 1/

(non-performing loans in percent of total loans)

article image

High vulnerability banks (in red) are defined as those with NPL ratios in the top 10th percentile of the pooled distribution of banks, medium-vulnerability banks (in yellow) are those with NPL ratios in the top 50th percentile of the pooled distribution (excluding the top 10th percentile), low vulnerability banks (in green) are those with NPL ratios in the bottom 50th percentile of the pooled distribution. Heat map is shown only for general license banks and excludes international (offshore) license banks. GL stands for “General license, local” and GF for “General license, foreign”. Bank ordering is based on banks’ asset size.

Annex Table 3:

Provisioning Coverage of NPLs 1/

(NPL Ratio Less Provisions to Total Loans)

article image

High vulnerability banks (in red) are defined as those with coverage ratios in the top 10th percentile of the pooled distribution of banks, medium-vulnerability banks (in yellow) are those with ratios between zero and the top 10th percentile, and low vulnerability banks (in green) are those with ratios less than zero. Heat map is shown only for general license banks and excludes international (offshore) license banks. GL stands for “General license, local” and GF for “General license, foreign”. Bank ordering is based on banks’ asset size.

Annex Table 4:

Earnings 1/

(Return on Assets (ROA))

article image

High vulnerability banks (in red) are defined as those with ROA<0, medium-vulnerability banks (in yellow) are those with ROA ratios in the bottom 50th percentile of the pooled distribution (excluding banks with ROA<0), low vulnerability banks (in green) are those with ROA in the top 50th percentile of the pooled distribution. Heat map is shown only for general license banks and excludes international (offshore) license banks. GL stands for “General license, local” and GF for “General license, foreign”. Bank ordering is based on banks’ asset size.

Annex Table 5:

Liquidity 1/

(Liquid asset to Deposit Ratio)

article image

Legal liquidity index not available on a bank-by-bank basis. High vulnerability banks (in red) are defined as those with a liquid asset to deposit ratio in the bottom 10th percentile of the pooled distribution, medium-vulnerability banks (in yellow) are those with a liquid asset to deposit ratio in the bottom 50th percentile of the pooled distribution (excluding banks in the bottom 10th percentile), low vulnerability banks (in green) are those with a liquid asset to deposit ratio in the top 50th percentile of the pooled distribution. Heat map is shown only for general license banks and excludes international (offshore) license banks. GL stands for “General license, local” and GF for “General license, foreign”. Bank ordering is based on banks’ asset size.

Annex Table 6:

Leverage 1/

(Total Equity/Total Assets)

article image

High vulnerability banks (in red) are defined as those with a leverage ratio of less than 3 percent (the regulatory requirement), medium-vulnerability banks (in yellow) are those with a leverage ratio >3 and </=7 percent, low vulnerability banks (in green) are those with a leverage ratio >7 percent. Heat map is shown only for general license banks and excludes international (offshore) license banks. GL stands for “General license, local” and GF for “General license, foreign”. Bank ordering is based on banks’ asset size.

Annex III. Contagion Analysis – Capital Losses by License Type

The top-left panels in Figures 13 correspond to the results shown in Figure 3 in the main text, whereas the remaining panels present the breakdown by the license type of the affected institutions.

Annex Figure 1.
Annex Figure 1.

Capital Losses as Function of Bank Originating Cascade 1/2/ High-Sensitivity Scenario – Breakdown by Type of Affected Bank

(US$ millions, 30 most systemically important institutions)

Citation: IMF Staff Country Reports 2019, 012; 10.5089/9781484394069.002.A004

1/ Each column displays total capital losses in the banking system (excluding those of the original defaulting bank) following the exogenous default of one bank (the license type of the original defaulting bank is shown in lieu of its name). Buffer losses are those over and above 8 percent of RWA, whereas injection to restore CAR represents the difference between 8 percent of RWA and actual capital after contagion. For defaulting banks (i.e. those whose capital is below 8, 4, and 0 percent of CAR in the high-, medium-, and low-sensitivity scenarios, respectively), if remaining capital is positive then that amount is classified as excess loss.2/ Total number of failures excludes original defaulting institution.
Annex Figure 2.
Annex Figure 2.

Capital Losses as Function of Bank Originating Cascade 1/2/ Medium-Sensitivity Scenario – Breakdown by Type of Affected Bank

(US$ millions, 30 most systemically important institutions)

Citation: IMF Staff Country Reports 2019, 012; 10.5089/9781484394069.002.A004

1/ Each column displays total capital losses in the banking system (excluding those of the original defaulting bank) following the exogenous default of one bank (the license type of the original defaulting bank is shown in lieu of its name). Buffer losses are those over and above 8 percent of RWA, whereas injection to restore CAR represents the difference between 8 percent of RWA and actual capital after contagion. For defaulting banks (i.e. those whose capital is below 8, 4, and 0 percent of CAR in the high-, medium-, and low-sensitivity scenarios, respectively), if remaining capital is positive then that amount is classified as excess loss.2/ Total number of failures excludes original defaulting institution.
Annex Figure 3.
Annex Figure 3.

Capital Losses as Function of Bank Originating Cascade 1/2/ Low-Sensitivity Scenario – Breakdown by Type of Affected Bank

(US$ millions, 30 most systemically important institutions)

Citation: IMF Staff Country Reports 2019, 012; 10.5089/9781484394069.002.A004

1/ Each column displays total capital losses in the banking system (excluding those of the original defaulting bank) following the exogenous default of one bank (the license type of the original defaulting bank is shown in lieu of its name). Buffer losses are those over and above 8 percent of RWA, whereas injection to restore CAR represents the difference between 8 percent of RWA and actual capital after contagion. For defaulting banks (i.e. those whose capital is below 8, 4, and 0 percent of CAR in the high-, medium-, and low-sensitivity scenarios, respectively), if remaining capital is positive then that amount is classified as excess loss.2/ Total number of failures excludes original defaulting institution.
1

Prepared by Kimberly Beaton. This Selected Issues Paper updates and extends the analysis in IMF Country Report No. 15/238.

2

See chapter 2 on financial integrity issues for an overview of the structure of Panama’s financial system, including a detailed description of the banking system.

3

Based on banks’ reported consolidated capital.

4

Importantly, the scenarios do not capture the potential indirect effect of the resulting fire sales potentially caused by the forced sale of similar assets by multiple borrower banks at the same time, which can trigger further declines in the market value of banks’ portfolios and further rounds of forced asset sales.

5

Regulation No. 004–2013.

6

The amount of dynamic provisions required (DPR) by each bank in period (quarter) t is defined as DPR(t) = α L(t) + βmax{ΔL(t),0}-SP(t), where α=1.5 percent, β=5 percent, L(t) = risk-weighted assets for loans classified under the normal category and SP(t)= variation in the balance of specific reserves. The amount of DPR is capped at 2.5 percent of qualifying risk-weighted assets and is subject to a floor of 1.25 percent of risk-weighted assets.

7

Article 37 of Regulation No. 004–2013 stipulates that the amount of dynamic provisions “cannot be less than the amount established in the previous quarter, unless the decrease is the result of a conversion to specific provisions”, and the SBP “will establish the criteria” for the conversion.

8

This follows the approach in IMF Country Report No. 15/238.

9

BCT Bank, the new owner of the reorganized Balboa Bank previously intervened by the Superintendent of Banks is excluded from the analysis.

10

The systemic importance of foreign banks would be larger if the data included foreign exposures to Panama. The dataset used here does not include foreign banks’ deposits in Panama and is therefore insufficient to assess foreign funding risks.

11

Model estimates provided by Antoine Malfroy-Camine and Damien Puy.

12

See Cerutti and others (2012) for details on the methodology.

13

Based on the consolidated claims on Panama of BIS reporting banks – excluding domestic deposits of subsidiaries of these banks in Panama.

14

Heatmap methodology is adapted from that in Ong, Jeasakul, and Kwoh (2013). While Ong, Jeasakul, and Kwoh (2013) normalize banks’ financial ratios with a z-score approach to assess bank’s financial performance relative to their peers, the approach adopted here assesses banks’ absolute financial health, although some indicator-specific thresholds are defined based on the pooled distribution of banks’ financial ratios.

Panama: Selected Issues
Author: International Monetary Fund. Western Hemisphere Dept.
  • View in gallery

    Panama: Panamanian Bank Performance by Bank License Type

  • View in gallery

    Panama: Total Capital Losses and Number of Defaults

    (based on the three scenarios related to the sensitivity of bank failures to CAR)

  • View in gallery

    Panama: Capital Losses as a Function of Original Defaulting Bank 1/2/

    (US$ millions, 30 most systemically important institutions)

  • View in gallery

    Panama: Total Capital Losses and Number of Defaults Excluding Capital from Dynamic Provisions

    (based on the three scenarios related to the sensitivity of bank failures to CAR)

  • View in gallery

    Panama: Capital Losses as a Function of Original Defaulting Bank Excluding Capital from Dynamic Provisions 1/2/

    (US$ millions, 30 most systemically important institutions)

  • View in gallery

    Capital Losses as Function of Bank Originating Cascade 1/2/ High-Sensitivity Scenario – Breakdown by Type of Affected Bank

    (US$ millions, 30 most systemically important institutions)

  • View in gallery

    Capital Losses as Function of Bank Originating Cascade 1/2/ Medium-Sensitivity Scenario – Breakdown by Type of Affected Bank

    (US$ millions, 30 most systemically important institutions)

  • View in gallery

    Capital Losses as Function of Bank Originating Cascade 1/2/ Low-Sensitivity Scenario – Breakdown by Type of Affected Bank

    (US$ millions, 30 most systemically important institutions)