Retail loans in the Polish Banking System Luigi Ruggerone – Intesasanpaolo Risk Management Dept. 1
Summary 3. Executive Summary 4. House price evolution in Poland 5. Main risk drivers in FX denominated housing loans 6. Impact of Zloty depreciation: assumptions 7. Analysis of the impact of Zloty depreciation on the Polish banking system 8. Estimated functional relationship between FX and NPLR 9. Estimated impact of a depreciation scenario on NPLR 10.Possible impact of Zloty depreciation on Ec. Capital 11.Summary of the results of the functional relationship 12. Impact of macroeconomic scenario 13.Vintage analysis on systemic figures 14. Macroeconomic stress test 15. Stress test on different vintages (1/2) 16. Stress test on different vintages (2/2) 17. Impact of unemployment on impaired rate 18. Conclusions 2
Executive Summary The main risk driver of the polish banking system is the retail credit portfolio, in particular the mortgage loans portfolio denominated in CHF; Despite the freshness of the portfolio and the limited time history, our analysis highlights a non linear relationship between the FX and the NPLR and allows us to draw some conclusions on additional provisioning that may become necessary in case of depreciation of the local currency. A 30% depreciation of the PLN determines an increase of more than 6% in the non performing loan rate; We also estimate the impact of an alternative scenario characterized, over a longer time span (until the end of 2014), by an appreciation of the PLN (15%) and higher CHF rates (to the level reached in 2007) and find out that the housing loans extended in 2007 until Q2 2008 are the riskiest. The reason being a residual capital well above the original amount extended and, above all, a sharp increase in the installments; We also run, with the help of Moody’s Credit Cycle model , an exercise to estimate the impact of unemployment on retail credit delinquency. A 350 bps increase in unemployment over the next 2 years, determines, on average, a 100 bps increase in the delinquency rate; 3
House price evolution in Poland We report below the house price evolution according to AMRON-SARFiN Report and we concentrate on the major cities; House price evolution price per square meter 10000 9000 8000 7000 6000 5000 4000 3000 2000 1000 0 1Q 2006 2Q 2006 3Q 2006 4Q 2006 1Q 2007 2Q 2007 3Q 2007 4Q 2007 1Q 2008 2Q 2008 3Q 2008 4Q 2008 1Q 2009 2Q 2009 3Q 2009 4Q 2009 1Q 2010 2Q 2010 3Q 2010 4Q 2010 1Q 2011 Warsaw Katowicw Wroclaw Gdansk Krakow Poznan From 2007 on, the value of the houses does not change significantly, so it is reasonable to believe that the re-assessment of house values impacts mainly on 2006 vintages; For vintages from 2Q 2007 up to 2Q 2008 the re-assessment could negatively impact on the LTV level; 4
Main risk drivers in FX denominated housing loans The majority of the loans has been granted to customers during a credit boom period , with low interest rates, that probably determined excessively benign assessment of the creditworthiness of clients; The impact of the rules set by the National Bank, within Recommendation S on good practices with regard to management of credit exposures that finance property and are mortgage-secured , could be relevant: On the one hand it is stated that the repayment burden on retail borrower’s net income should not exceed 42% On the other hand, starting from June 2012, the risk weight to be applied for foreign currency denominated retail exposure will increase to 100%; This last change, considering the standard methodology for calculating Capital Requirement, will determine a consistent increase of capital requirement for FX loans , and a consequent reduction of the solvency ratio; considering only the FX denominated portfolio, capital absorption will consequently increase by 33%. Economic downturn, especially in terms of unemployment rate increase, can trigger an increase in the non performing loan ratio; this is also confirmed in the Financial Stability report , where it is stated that “ The deterioration in the condition of some household related, inter alia, to unemployment growth during the recent economic slowdown, may continue to exert a negative effect on loan quality ”. 5
Impact of Zloty depreciation: assumptions Our analysis is focused on the impact of zloty depreciation vs Swiss franc as the main risk factor ; The short time series of exposures and Non Performing Loans require a very cautious approach; Even if an increase of CHF interest rates can impact banking system portfolios, we do not foresee this as the major risk over the short term , considering the statement of the Swiss National Bank, on September the 6 th, and the current and forecasted value of inflation in Switzerland: Reuters: “Using some of the strongest language from a central bank in the modern era, the SNB said it would no longer tolerate an exchange rate CHF/€ below 1.2 and would defend the target by buying other currencies in unlimited quantities ”. The current and the forecasted value of Switzerland inflation rate is reported below: Any interest rate increase, in the current economic environment, will affect the commitment of maintaining the FX at the current target level. 6
Analysis of the impact of Zloty depreciation on the Polish banking system This analysis assesses the impact of zloty depreciation vs Swiss franc on the Non Performing Loan Rate; Due to the lack of data, but above all due to the conversion into PLN of problematic housing loans, we both analyze CHF denominated housing loans and total housing loans (independently of the currency); Based on the available data, we estimate for the Polish banking system a functional relationship between the level of the CHF/PLN exchange rate and the level of the Non-Performing Loan Ratio (NPLR from now on); The two estimated relations between NPLR and FX rate are non linear and monotonic; i.e. larger depreciations have increasing effects on NPLR; In our analysis we use monthly average FX rate, calculated starting from the figures provided by the Polish National Bank (NBP) on a one year time horizon; We assume that the effect of a change in the FX rate will impact on NPLR with a delay of three months; so the relation between NPLR and FX has been estimated considering a time lag of 3 months; Furthermore we estimate the expected loss under three different FX stressed scenarios. In particular three different thresholds of depreciation of PLN vs. CHF have been applied: 10% - 20% - 30%. 7
Estimated functional relationship between FX and NPLR NPLR NPLR Total NPLR and CHF/PLN rate CHF NPLR and CHF/PLN rate 2.40% 2.00% 2.20% 1.80% 2.00% 1.60% 1.80% 1.40% 1.60% 1.20% 1.40% 1.00% 1.20% 0.80% 2.5 2.6 2.7 2.8 2.9 3 3.1 2.5 2.6 2.7 2.8 2.9 3 3.1 Fx NPL vs CHF Fx NPL vs CHF As already anticipated the functional relationship is monotonic and non linear. The NPLR grows faster than the PLN depreciates vs. CHF; Since the dynamics of total NPLR and NPLR of the CHF denominated housing loans are very similar over time, the regression function has a similar trend; this confirms the hypothesis that the housing loans denominated in CHF after the restructuring fall within Total Housing Loan Portfolio ; As already stated in the previous page, the issue of restructured housing loans should be carefully analyzed. 8
Estimated impact of a depreciation scenario on NPLR Using the non linear regression function presented above, we shocked the monthly average FX rate calculated on a one year time horizon by 10%, 20%, and 30% respectively. We provide below the results of this stress test: Total housing loans Δ NPLR Systemic NPLR June 11 Strees on FX Stressed FX* Tot NPLR 2.07% 10% 3.58 4.22% 2.15% 2.07% 20% 3.90 6.08% 4.01% 2.07% 30% 4.23 8.41% 6.34% * The fx means in Apr 11 and Oct 11 were 3.009 and 3.25 respectively CHF housing loans Systemic NPLR June 11 Strees on FX Stressed FX* CHF NPLR Δ CHF NPLR 1.53% 10% 3.58 3.51% 1.98% 1.53% 20% 3.90 5.20% 3.66% 1.53% 30% 4.23 7.29% 5.76% * The fx means in Apr 11 and Oct 11 were 3.009 and 3.25 respectively The impact, in term of additional provisioning, will be determined through the Cover ratio prevailing within each Polish bank; We have estimated also the impact on NPLR of interest rate level; however due to the short time series available, currently interest rates have a very limited explanatory power, so the only variable considered here has been the FX rate level; 9
Possible impact of Zloty depreciation on Ec. Capital Impact of FX level on Ec. Cap. NPLRC 2.40% atual nplrc values 1.90% Estimated function 95% 1.40% 99% 0.90% 0.40% 2.70 2.80 2.90 3.00 3.10 CHF/PLN Apart from an estimation of the level on NPLRC, our estimated function allows to calculate the economic capital (i.e. the unexpected loss) with different confidence intervals; This analysis could be very useful in case of a comprehensive ICAAP exercise. The economic capital is the difference between the unexpected loss (red or green lines in the graph above, depending upon the confidence interval) and the expected loss (black estimated function) for each given FX rate; 10
Recommend
More recommend
