UniCredit Global Themes Series

Transcription

1 7 September 16 Economics Research UniCredit Global Themes Series Economics & FI/FX Research No. 37 Credit Research 7 September 16 Equity Research Cross Asset Research Forecasting non-performing loans in Europe Outlook by on basis of macroeconomic fundamentals A proprietary model for forecasting non-performing loan ratios in EU countries is presented. Real GDP growth, unemployment and real interest rates impact non-performing loans with substantial time lags. In some countries, better macro data have therefore not affected loan books yet. Our model signals a decline in the Italian non-performing loan ratio from 18% (in % of gross loans) last year to 16½% in 17 and 14% by. Significant improvements can also be expected for Spain and some peripheral eurozone countries such as Portugal, Ireland and Greece. The development in Germany and France will remain rock-solid. Especially sharp rises in real long-term interest rates would endanger the recovery in loan books. Author: Dr. Andreas Rees, Chief German Economist (UniCredit Bank) Editor: Erik F. Nielsen, Global Head of CIB Research (UniCredit Bank London)

2 Contents 3 Executive Summary 4 I. Introduction 5 II. Status quo 6 III. Our macro model 1 IV. Scenario analysis 1 V. Conclusions 11 Technical Appendix 13 Charts Appendix 15 References 16 List UniCredit Research page See last pages for disclaimer.

3 Executive Summary Status quo Agenda Our model GDP growth Unemployment Real interest rates Time lags Input data Results Risk scenario The issue of non-performing loans continues to be high on the agenda of banks, financial investors and policymakers. Especially in some eurozone peripheral countries and Italy, nonperforming loan ratios (in % of gross loans) are still at elevated levels: 35% in Greece, 18% in Italy, 15% in Ireland, and 1% in Portugal (all 15 World Bank data). In the following, forecasts are derived by estimating a proprietary model based on macroeconomic variables. With this exercise, we aim at providing the fundamental setting of non-performing loan ratios by. It goes without saying that the legal and regulatory framework plus the creation of bad banks in some countries also play an important role. Especially three macro variables proved to be useful in forecasting non-performing loans: real GDP growth, unemployment and real interest rates. The following rules of thumb can be established: An increase in the real GDP growth rate by 1pp leads to a total decrease in the nonperforming loan ratio of.5pp over time. A decline in the unemployment rate of 1pp triggers a decrease in the non-performing loan ratio of.5pp. A reduction in the real interest rate (1Y) by 1pp leads to a lower non-performing loan ratio of 1.pp in total over time (real 3M:.pp). These macro variables impact loan books with (substantial) time lags. There are both fundamental macroeconomic and legal reasons for this pattern. Besides historical macro data, consensus and UniCredit forecasts for real GDP growth, unemployment and real interest rates are used as inputs to derive non-performing loan ratios by. Our assumptions are deliberately conservative in order to err on the side of caution. In Italy, the return to growth territory since 15 will start affecting markedly loan books next year. For 17, the forecast for the non-performing loan ratio is 16½% after 18% in 15. By, our model signals 14%, which would be the lowest level since 1. In Spain and Ireland, the already discernible downward trend in bad loans triggered both by macro fundamentals and the creation of bad banks will continue. The non-performing loan ratios in Portugal will decline to 4½% by from nearly 1% last year. In Greece, there will also be an improvement, although non-performing loans will still be at elevated levels (: %). In Germany and France, the development will remain rock-solid. Italian non-performing loans account for about one third of total non-performing loans in the eurozone. We therefore identify macroeconomic risk factors and run a scenario analysis for Italy. Especially sharp rises in real long-term interest rates would endanger the recovery in loan books in Italy and the periphery. UniCredit Research page 3 See last pages for disclaimer.

4 I. Introduction Non-performing loans at center stage Fundamental macro setting Structural framework Structural framework The issue of non-performing loans in some European countries continues to be high on the agenda of banks, financial investors and policymakers. From a macroeconomic perspective, bad loans are an obstacle on the way to further recovery. By tying up bank capital and reducing profitability, they weigh on bank lending and, hence, future economic activity. After all, more than 8% of debt financing in the non-financial corporate sector in EMU comes from bank lending compared to less than % in the US. The million-euro question is whether and to what extent bad loans will decline in the next few years. In the following, we present forecasts by for non-performing loan ratios in some European countries. Special attention will be paid to the development in Italy. The forecasts are derived by estimating a proprietary model based on macroeconomic variables. With this exercise, we aim at providing the fundamental macro setting of non-performing loans in the next few years. It goes without saying that bad loans are not only influenced by macroeconomic variables but also by structural factors. Examples are the legal and regulatory framework, the existence of well-functioning distressed debt markets, the tax regime, etc. 1 However, it is likely that macro variables will still play an important role in influencing bad loan volumes. Before presenting our model and results, the status quo of non-performing loans in the eurozone and its member states is shortly described to provide a starting point. 1 Shekhar, A. et al., (15), A Strategy for Resolving Europe s Problem Loans, IMF Staff Discussion Note, September UniCredit Research page 4 See last pages for disclaimer.

5 II. Status quo Definition and source In the following, we focus on the non-performing loan ratio as a share of gross loans (in %) provided by the World Bank. Hence, this measure includes non-performing loans before the deduction of loan-loss provisions by banks. The World Bank data have the advantage of being available since 1997/98 for the majority of countries. Comparisons over time and the estimation of a (panel) model are therefore feasible. However, it is important to note that the definition of non-performing loans is not internationally harmonized and may differ by country. Furthermore, there are even differences depending on the institution which publishes the data. For example, the ECB uses a narrower definition of non-performing loans than the European Banking Authority. Recently, the BIS made proposals to harmonize the measurement of bad loans going forward. Eurozone versus US EMU country comparison When comparing the aggregate eurozone figures with the US over time, a clear-cut pattern emerges (see chart 1). While increasing in sync to about 5% each by 9 from roughly 1% in 6 before the outburst of the financial crisis, a significant divergence started. Last year, the non-performing loan ratio in the US was 1½% compared to 5½% in EMU. The good news is that, at least on an aggregate level, some progress has been made in the eurozone. Non-performing loans peaked in 13 at nearly 8%. When looking at national patterns in the eurozone, the differences by country and over time are striking. In Cyprus, the non-performing loan ratio was nearly 46% last year (7: 3½%) followed by Greece with 35% (7: 4½%) and Italy with 18% (7: 6¾%). Although still at 15% in 15, Ireland has made substantial progress in reducing its bad loans after reaching nearly 6%. Another success story is Spain where the non-performing loan ratio declined to 6¼% last year from 9¼% in 13. However, both in Ireland and Spain, the creation of bad banks provided relief and contributed to the reduction. In Ireland, the National Asset Management Agency took over a whopping EUR 74bn of property-related loans. This equaled about one fifth of total loans to the private sector. In Spain, its equivalent Sareb absorbed ½% of total loans (or EUR 4bn). In Germany (15: ¼%) and the UK (15: 1½%), non-performing loans remained low during the crisis years or declined swiftly again. 3 Chart 1 Non-performing loan ratios, in % of gross loans Chart Non-performing loan ratios, in % of gross loans US EMU CY* GR IT IE PT SP FR AT NL* DE** UK *8 data; **14 data Source: World Bank, UniCredit Research See for more details, Mesnard, B. et al. (16), Non-performing loans in the Banking Union: stocktaking and challenges, Briefing, European Parliament, March; D Hulster et al. (14), Loan Classification And Provisioning: Current Practices In 6 Countries, Overview paper, World Bank and Basel Committee on Banking Supervision (16), Prudential Treatment of problem assets definitions of non-performing exposures and forbearance, Consultative Document, April 3 See also our Charts Appendix in which historical developments for major countries can be found. UniCredit Research page 5 See last pages for disclaimer.

6 III. Our macro model Country panel Macroeconomic fundamentals Real GDP Unemployment Long-term and short-term interest rates House price Time lags We developed a model for forecasting non-performing loan ratios based on macroeconomic fundamentals. The following fourteen EU countries were included in the panel estimate: Germany, France, Italy, Spain, the UK, the Netherlands, Austria, Ireland, Greece, Belgium, Portugal, Sweden, Denmark and Poland. A more detailed explanation of our model can be found in the Technical Appendix. A multitude of macro variables were tested as to whether they have any predictive power for forecasting non-performing loans. Four variables proved to be significant and had the expected sign: GDP growth, unemployment, interest rates and house prices. Economic activity. Higher growth translates into more profit for companies and income for consumers and improves the capability of borrowers to service their debts. According to our estimates, an increase in the real GDP growth rate by 1pp leads to a total decrease in the non-performing loan ratio of.5pp over time. Unemployment. A decline in the unemployment rate also lifts the debt service capability of private households. Higher purchasing power also means more private consumption and, hence, stronger domestic demand for companies. According to our estimates, a decrease of 1pp in the unemployment rate leads to a decrease in the non-performing loan ratio of.5pp. Interest rates. Lower interest rates lessen the difficulties of borrowers to repay their debt. A decline in the real interest rate (1Y) by 1pp reduces the non-performing loan ratio by 1.pp in total over time. A reduction in the real 3M interest rate by 1pp leads to a decrease in the non-performing loan ratio of.pp. 4 Real estate. Rising housing wealth eases access to credit by boosting the value of the underlying housing assets used as collateral. According to our estimates, an increase in real house prices by 1pp leads to a (minor) reduction of the non-performing loan ratio by.1pp. 5 It is important to note that these macro variables typically do not have any immediate impact. Instead, GDP growth, unemployment, etc., influence the non-performing loan ratio with (substantial) time lags. For instance, there is a lag of one year between changes in the unemployment rate and variations in non-performing loans. In the nexus between GDP growth and bad loans, there is even a time lag of -4 years. The recent development in Italy illustrates this effect well. Renewed GDP growth again in 15 (and 16 according to our forecast) did not lead to a decline in bad loans yet (see chart 3). Instead, the non-performing loan ratio increased further and stabilized at 18% last year. In other words, the recent positive macro development will only have an impact later on from 17 onwards. 4 Real interest rate defined as nominal interest rate minus the yearly change in consumer price inflation. 5 Real house prices defined as the yearly change in house prices minus the yearly change in consumer price inflation. UniCredit Research page 6 See last pages for disclaimer.

7 Chart 3 GDP growth and non-performing loans in Italy Chart 4 Time for resolving insolvency, in years Real GDP growth, yoy (%) Non-performing loans ratio (%; rs) * 13 IE UK NE AT DE SP CY IT FR PT GR US *UniCredit forecast Source: World Bank, World Bank Doing Business 16, UniCredit Research Economic reasons Legal reasons Using consensus forecasts as input The underlying rationale for these lags is that macro fundamentals need time to leave their marks on the loan book. In an economic downswing, companies and consumers will hold out for a while by working off backlog orders, selling assets, etc., to service their debts. Furthermore, companies might have accumulated financial means in better times which then act as some kind of buffer. In an upswing, the recovery has to be longer lasting before financial solvency improves. Furthermore, there are legal reasons such as insolvency laws and the efficiency of the legal system which additionally delay the impact from the macro world on bad loans. According to the Doing Business 16 Survey of the World Bank, it takes between one and two years in the majority of eurozone countries to legally resolve insolvency (see chart 4). In the Netherlands, Austria and Germany, the time span is about one year. In Italy, France and Portugal, it is close to two years. 6 In Greece, the time span is even 3½ years. Our model is used to derive fundamentally based forecasts for non-performing loan ratios by country. We focused on the four major eurozone member states and three peripheral countries: Germany, France, Italy, Spain, Portugal, Ireland and Greece. As input variables were used historical macro data (given the above-mentioned time lags) and projections for GDP growth, unemployment, etc. To take an example, unemployment rates in 15 can be used for forecasting the non-performing loan ratio in 16, given the time lag of one year. However, a projection for unemployment in 16 is needed to derive a forecast for 17 and so on. As input for these projections, we deliberately did not use our own forecasts but the consensus view as a kind of neutral yardstick. The latter equals the average forecast of economists with the extreme optimists and pessimists netting each other out. We used the consensus forecasts collected by Bloomberg. They are available for GDP growth, inflation (for calculating real interest rates) and the unemployment rate. The Bloomberg data have the advantage of being updated on a monthly basis to take new information and events swiftly into account. Forecasts of official institutions such as the IMF, the OECD and the EU Commission are typically only updated twice a year. Furthermore, the Bloomberg data do not only cover 16 and 17 but also include An excellent overview for Italy on legal issues is provided by Garrido, et al. (16), Cleaning-up Bank Balance Sheets: Economic, Legal, and Supervisory Measures for Italy, IMF Working Paper, July UniCredit Research page 7 See last pages for disclaimer.

8 Conservative extrapolation Interest rate assumptions Skipping house prices Results For 19 and, we did extrapolations for growth, inflation and unemployment. Our assumptions are deliberately conservative in order to err on the side of caution. All of the data can be found in table A.1 in the Technical Appendix. For instance, in the case of Germany, France and Italy, we simply left the 18 growth rate unchanged and used them for 19 and as well. For Germany, this means +1.4% p.a. in 19 and (France: +1.3%; Italy: +.9%). In the case of Spain and Ireland, it was assumed that they cannot sustain the comparatively high growth rates as expected by the consensus by 18 and that a gradual growth slowdown will kick in afterwards. In terms of unemployment, further moderate declines for all countries (apart from Germany due to recent immigration) are assumed. Interest rate consensus forecasts are only available to a limited extent. To be more precise, 16 and 17 (end of period) for 3M and 1Y could be used for Germany, France, Italy and Spain. From 18 onwards, extrapolations were done again. For the short end of the curve, it was assumed that nominal interest rates will remain at historically low levels. Given that the consensus expects inflation rates to return to 1½% to %, short-term real interest rates will decline. At the longer end of the curve, the consensus expects a gradual normalization in 17, which is assumed to continue from 18 onwards. In other words, 1Y real interest rates are assumed to rise somewhat over time. No interest rate forecasts are available for the periphery. As a rough approximation, we used current interest spreads to Germany to derive future data. Finally, there are no consensus forecasts for house prices available. We therefore excluded this variable from our equation. The information loss should be limited. Even in the unlikely case of a strong rebound or decline in house prices by, say, 5% p.a., the non-performing loan ratio would only decrease by.5pp each year. Our most important results are the following: Eurozone aggregate The non-performing loan ratio in EMU will decline from 5¾% in 15 to about 3% by (see chart 5). This would be the lowest level since 8. As pointed out above, our eurozone aggregate covers seven countries and more than 8% of bank loans in EMU. Major drivers behind the decrease are two countries: Italy and Spain. Italy Our model signals a significant decline in the non-performing loan ratio from 18% last year to 14% by (see chart 6). The moderate improvement in macro fundamentals which has taken place since 15 will start having a marked impact on loan books next year. For 17, the forecast for the non-performing loan ratio is 16½% after 18% in 15 and 18¼% in 16. On average, the reduction in the Italian non-performing loan ratio will be.8pp per year. Chart 5 Non-performing loan ratio in EMU* (%) Chart 6 Non-performing loan ratio in Italy (%) 9 8 Forecast 18 Forecast *Based on Germany, France, Italy, Spain, Portugal, Ireland and Greece Source: World Bank, UniCredit Research UniCredit Research page 8 See last pages for disclaimer.

9 Contributions by macro drivers Various quite long time lags and different coefficients between macro variables and bad loans make the underlying drivers of the decline not easy to spot. 7 Which macro factor contributes to the decrease in Italian non-performing loans over the forecasting horizon to what extent? In order to make our results 1% transparent for readers, we calculated the contributions from GDP growth, unemployment and real interest rates (see table 1). Our calculations also demonstrate where the biggest risks are. Between 16 and, growth contributes.7pp to the overall decline of 4pp in the non-performing loan ratio. The contribution from the lower unemployment rate is.6pp, from the 1Y real interest rate.6pp. The real 3M interest rate plays a comparatively small role (-.1pp). In other words, a sharp rise in the Italian real longterm interest rate could endanger the recovery in bad loans the most. TABLE 1: CONTRIBUTION TO CHANGE IN ITALIAN NON-PERFORMING LOANS RATIO, IN PP Total Real GDP Unemployment rate Real interest rate (1Y) Real interest rate (3M) Note: Some figures may not add up due to rounding issues Source: UniCredit Research Spain Peripheral countries Germany & France The already discernible downward trend in bad loans will continue in the next few years. Our model flags a decline from 6.3% last year to 1% by. The last time such historically low levels were seen was before the bursting of the real estate bubble about ten years ago. While the impact differs in each year, GDP growth, unemployment and real long-term interest rates contributed rather equally to the decline in non-performing loans on average. For all three peripheral countries, our model signals sharp declines in the non-performing loan ratios (see chart 7). For Portugal, it will decline to 4½% by from nearly 1% last year. For Ireland, the non-performing loan ratio will come down to 7½% by from nearly 15% in 15. In Greece, there will also be an improvement, although bad loans will remain at still elevated levels (: % after 35% in 15). In all three peripheral countries, the (lagged) impact from lower real interest rates after the crisis years (when interest rates were at painfully high levels) plays an important role. This is especially true for Greece. The development in Germany and France will remain rock-solid (see chart 8). Our model flags a slight decline to 1¾% by from roughly % plus some yearly volatility. In France, the non-performing loan ratio falls to 3% after 4% in 15. Chart 7 Non-performing loan ratio in periphery (%) Chart 8 Non-performing loan ratio in Germany & France (%) 4 35 Portugal Ireland Greece Forecast 7 6 Germany France Source: World Bank, UniCredit Research 7 There are also overlapping effects, since GDP growth is lagged by two and four years and the real 1Y rate by one and three years. See also the Technical Appendix for the concise lag structure of the explanatory variables in the panel regression. UniCredit Research page 9 See last pages for disclaimer.

10 IV. Scenario analysis Some caveats Risk scenario It is important to note that our forecasts are, of course, conditional. Besides historical data as input variables (given time lags), the model also includes projections for future major macro variables. In other words, if the consensus and UniCredit forecasts are still too optimistic, the model overestimates the future decline in non-performing loans. In the following, we ran a scenario analysis in which we assumed hypothetical adverse macroeconomic conditions for Italy (see table ). Backdrop is the fact that Italian non-performing loans are of significant importance for EMU as a whole. They account for about one third of total non-performing loans in the eurozone. 8 In our risk scenario called no growth and higher interest rates, the Italian economy is assumed to persistently stagnate from 17 to (which is certainly a tough assumption). Unemployment starts climbing to more than 13% on a yearly basis, which would be higher than in the previous crisis years. While the ECB is expected to keep its key interest rates low and continue QE, inflation rates will be stuck in negative territory. As a result, real short-term and long-term interest rates are also assumed to increase. Furthermore, some financial investors may sell BTPs at the longer end of the curve at an early stage, thereby driving up the real 1Y rate further (but only to a certain significant extent, given QE). TABLE : SCENARIO ANALYSIS FOR ITALY Italy (extrapolated consensus baseline) Real GDP, yoy (%) Unemployment rate (%) Real interest rate (1Y; %) Real interest rate (3M; %) Italy (risk scenario) Real GDP, yoy (%) Unemployment rate (%) Real interest rate (1Y; %) Real interest rate (3M; %) Source: Bloomberg, UniCredit Research Results In the beginning of the forecast horizon, some historical and hence unchangeable macro variables impact the loan book with a time lag independent of the projections. For instance, the model still signals a decline to 17% in 17 (baseline: 16½%) despite harsher macro conditions. However, in the medium term, the non-performing loan ratio would be increasing again and reach 18% by. We think that the risks are tilted to the upside and that our model may paint a too optimistic picture in this tough scenario. Backdrop is that the return to growth territory only started last year after persistently shrinking economic activity between A renewed and long-lasting slide into stagnation could therefore put companies and consumers under stronger pressure than signaled by our model. V. Conclusions Some relief in loan books ahead Our model is able to forecast the fundamental trend in non-performing loans by country. Assuming no abrupt halt in the (moderate) recovery in Italy and the periphery, the longawaited turnaround in non-performing loans is within reach. Furthermore, some progress in the legal and regulatory framework has recently been made. In combination with the improvement in macro fundamentals, the recovery in loans books is well supported. 8 Garrido, et al. (16), Cleaning-up Bank Balance Sheets: Economic, Legal, and Supervisory Measures for Italy, IMF Working Paper, July UniCredit Research page 1 See last pages for disclaimer.

11 Technical Appendix Panel regression Sample Model The available data on non-performing loans are limited. Most of the historical time series only start in 1997 or 1998 on an annual basis (World Bank data). We therefore opted for a panel regression which allows for increasing the sample size, while gaining a cross-country perspective. Our sample covers annual data from 1998 to 15 for 14 EU countries: Germany, France, Italy, Spain, the UK, the Netherlands, Austria, Ireland, Greece, Belgium, Portugal, Sweden, Denmark and Poland. A panel regression of the following form was estimated: y it = β(l)x it + v i + ε it where y it is the non-performing loans ratio β(l) is the 1 k lag polynomial vector to be estimated X it is the k 1 vector of explanatory macro variables v i is a time-invariant unobserved country-specific effect ε it is the error term i = 1,, N is the cross-section indicator, i.e. the number of countries included in the panel t = 1,, T is the time dimension Unit root tests Fixed effects Lags Unit root tests were conducted to assess whether the used time series are stationary in level. This was the case. It is well known that there are differences in the classification of non-performing loans across countries, given various accounting approaches and regulations. Furthermore, there are different insolvency laws and other legal procedures by country. In order to take such effects into account, it was controlled for unobserved fixed effects in the panel regression. The following explanatory macro variables were used with the corresponding lags (years) in brackets: real GDP growth yoy () and (4) unemployment rate (1) real interest rate 1Y (1 and 3) real interest rate 3M () real house prices yoy (no lag; not included in the model for forecasting purposes) Adjusted R-squared =.9 Robustness tests In order to check for the robustness of the estimated results before and after the financial crisis, the sample was split into two sub-samples. The first one covers the period before the financial crisis from 1998 to 7; the second one the period from 8 to 15. No significant structural break could be detected. In general, there is a tendency that the coefficients become bigger after 7. In less technical terms, macro variables were getting more powerful in driving non-performing loans than before the financial crisis. UniCredit Research page 11 See last pages for disclaimer.

12 TABLE A.1: MACRO PROJECTIONS BASED ON BLOOMBERG CONSENSUS & UNICREDIT FORECASTS Germany Real GDP, yoy (%) Unemployment rate (%)* Real interest rate (1Y; %) Real interest rate (3M; %) France Real GDP, yoy (%) Unemployment rate (%)* Real interest rate (1Y; %) Real interest rate (3M; %) Italy Real GDP, yoy (%) Unemployment rate (%)* Real interest rate (1Y; %) Real interest rate (3M; %) Spain Real GDP, yoy (%) Unemployment rate (%)* Real interest rate (1Y; %) Real interest rate (3M; %) Portugal Real GDP, yoy (%) Unemployment rate (%)* Real interest rate (1Y; %) Real interest rate (3M; %) Ireland Real GDP, yoy (%) Unemployment rate (%)* Real interest rate (1Y; %) Real interest rate (3M; %) Greece Real GDP, yoy (%) Unemployment rate (%)* Real interest rate (1Y; %) Real interest rate (3M; %) *EU harmonized Source: Bloomberg, UniCredit Research UniCredit Research page 1 See last pages for disclaimer.

13 Charts Appendix Chart A.1 NPL ratio in Germany, in % of total loans 6 5 Chart A. NPL ratio in France, in % of total loans Long-term average Long-term average Chart A.3 NPL ratio in Italy, in % of total loans Chart A.4 NPL ratio in Spain, in % of total loans Long-term average Long-term average Source: World Bank, Bloomberg, UniCredit Research UniCredit Research page 13 See last pages for disclaimer.

14 Chart A.5 NPL ratio in Portugal, in % of total loans Chart A.6 NPL ratio in Ireland, in % of total loans Long-term average 15 1 Long-term average Chart A.7 NPL ratio in Greece, in % of total loans Long-term average Source: World Bank, Bloomberg, UniCredit Research UniCredit Research page 14 See last pages for disclaimer.

15 References Basel Committee on Banking Supervision (16), Prudential Treatment of problem assets definitions of non-performing exposures and forbearance, Consultative Document, April Beck, R. et al. (13), Non-Performing Loans. What Matters In Addition To The Economic Cycle, ECB Working Paper, February D Hulster, et al. (14), Loan Classification And Provisioning: Current Practices In 6 Countries, World Bank, Overview paper, August Garrido, et al. (16), Cleaning-up Bank Balance Sheets: Economic, Legal, and Supervisory Measures for Italy, IMF Working Paper, July Klein, N. (13), Non-performing Loans in CESEE: Determinants and Macroeconomic Performance, IMF Working Paper, March Louzis, et al. (11), Macroeconomic and bank-specific determinants of non-performing loans in Greece: A comparative study of mortgage, business and consumer loan portfolios, Journal of Banking and Finance Makri, V. et al. (14), Determinants of Non-Performing Loans: The Case of Eurozone, Panoeconomicus Mesnard, B. et al. (16), Non-performing loans in the Banking Union: stocktaking and challenges, Briefing, European Parliament, March Nkusu, M. (11), Nonperforming Loans and Macrofinancial Vulnerabilities in Advanced Economies, IMF Working Paper, July UniCredit Research page 15 See last pages for disclaimer.

16 List No Date Author(s) Title Aug 16 Dr. Harm Bandholz The rise of the machines: Economic and social consequences of robotization 35 1 Jul 16 Fadi Hassan, Marco Valli Business Opportunities in Emerging Markets: Which Country and Industry for Exporters? Jun 16 Roberto Mialich Central Banks (Mis)Communication & Exchange Rate Volatility 33 8 Jun 16 Harm Bandholz, Andreas Rees Thomas Strobel Internationalization of companies by 3 3 7Jun 16 Tobias Rühl European Football Championship 16 in France 31 3 Mar 16 Vasileios Gkionakis, Kathrin Goretzki How much ECB QE is in the price of EUR-USD? 3 19 Jan 16 Fadi Hassan Global Challenges and Prospects for Emerging Markets: Commodity, China, and Capitals 9 4 Nov 15 Tobias Rühl Introducing The EMU Financial Conditions Index by UniCredit 8 4 Jun 15 Roberto Mialich The lingering menace of FX wars 7 1 May 15 Marco Valli, Edoardo Campanella 6 19 Nov 14 Erik F. Nielsen, Harm Bandholz Andreas Rees 5 1 Jul 14 Roberto Mialich The EUR-USD resilience Growth uncertainties in China What it means for the eurozone Public investment: If not now, when? Kick-starting the economy and taking care of future generations 4 8 May 14 Andreas Rees Football World Cup 14 in Brazil Forecasting national football success 3 19 May 14 Harm Bandholz Deleveraging in Europe and the US: Not a brake on growth 13 May 14 Gillian Edgeworth, The newer EU states: Maximizing integration Carlos Ortiz, Dan Bucsa, Marcin Mrowiec, Pavel Sobisek, Kristofor Pavlov, Hrvoje Dolenec, Lubomir Korsnak, Mihai Patrulescu. 1 6 Mar 14 Daniel Vernazza, The damaging bias of sovereign ratings Erik F. Nielsen, Vasileios Gkionakis 4 Oct 13 Andreas Rees Introducing the Global Leading Indicator by UniCredit 19 1 Sep 13 Michael Rottmann The return of the macro-yield correlation & its implication for active duration management 18 3 Sep 13 Vasileios Gkionakis, Introducing BEER by UniCredit; Our new framework for modeling equilibrium exchange rates Daniel Vernazza 17 5 Jul 13 Gillian Edgeworth, CEE: Stress testing external financing shortfalls Dan Bucşa, Carlos Ortiz Jun 13 Roberto Mialich Too big to fall soon! Why the USD still remains the world's reserve currency 15 6 Jun 13 Gillian Edgeworth CEE: The 'normalisation' challenge 14 1 May 13 Marco Valli Inflating away the debt overhang? Not an option 13 7 May 13 Harm Bandholz, The quest for competitiveness in the eurozone Tullia Bucco, Loredana Federico, Alexander Koch 1 4 Jan 13 Luca Cazzulani, Short and long-term impact of the introduction of CACs in the EMU Elia Lattuga 11 Oct 1 Harm Bandholz US Fiscal Policies at a Crossroad: consolidation through the fiscal cliff? UniCredit Research page 16 See last pages for disclaimer.

17 List (Cont d) No Date Author(s) Title 1 18 Sep 1 Marco Valli The eurozone five years into the crisis: lessons from the past and the way forward 9 3 Jul 1 Erik F. Nielsen Europe in the second half of 1: Moving closer together or further apart? 8 18 Jul 1 Harm Bandholz, Andreas Rees 7 16 Jul 1 Luca Cazzulani, Chiara Cremonesi Reach out for the medal(s) EMU bond correlation & portfolio decisions 6 4 Jun 1 Andreas Rees Money scoring goals. Forecasting the European Football Championship Apr 1 Harm Bandholz How the Great Recession changed the Fed 4 16 Apr 1 Erik F. Nielsen Safeguarding the common eurozone capital market 3 1 Apr 1 Andreas Rees The hidden issue of long-term fiscal sustainability in the eurozone 3 Mar 1 Alexander Koch European housing: fundamentals and policy implications 1 1 Mar 1 Gillian Edgeworth, Vladimir Zlacký, Dmitry Veselov The EU: Managing capital flows in reverse UniCredit Research page 17 See last pages for disclaimer.

18 Legal Notices Glossary A comprehensive glossary for many of the terms used in the report is available on our website: Link Disclaimer Our recommendations are based on information obtained from, or are based upon public information sources that we consider to be reliable but for the completeness and accuracy of which we assume no liability. All estimates and opinions included in the report represent the independent judgment of the analysts as of the date of the issue. This report may contain links to websites of third parties, the content of which is not controlled by UniCredit Bank. No liability is assumed for the content of these third-party websites. We reserve the right to modify the views expressed herein at any time without notice. Moreover, we reserve the right not to update this information or to discontinue it altogether without notice. This analysis is for information purposes only and (i) does not constitute or form part of any offer for sale or subscription of or solicitation of any offer to buy or subscribe for any financial, money market or investment instrument or any security, (ii) is neither intended as such an offer for sale or subscription of or solicitation of an offer to buy or subscribe for any financial, money market or investment instrument or any security nor (iii) as an advertisement thereof. The investment possibilities discussed in this report may not be suitable for certain investors depending on their specific investment objectives and time horizon or in the context of their overall financial situation. The investments discussed may fluctuate in price or value. Investors may get back less than they invested. Changes in rates of exchange may have an adverse effect on the value of investments. Furthermore, past performance is not necessarily indicative of future results. In particular, the risks associated with an investment in the financial, money market or investment instrument or security under discussion are not explained in their entirety. This information is given without any warranty on an "as is" basis and should not be regarded as a substitute for obtaining individual advice. Investors must make their own determination of the appropriateness of an investment in any instruments referred to herein based on the merits and risks involved, their own investment strategy and their legal, fiscal and financial position. As this document does not qualify as an investment recommendation or as a direct investment recommendation, neither this document nor any part of it shall form the basis of, or be relied on in connection with or act as an inducement to enter into, any contract or commitment whatsoever. Investors are urged to contact their bank's investment advisor for individual explanations and advice. Neither UniCredit Bank, UniCredit Bank London, UniCredit Bank Milan, UniCredit Bulbank, Zagrebačka banka, UniCredit Bank Czech Republic and Slovakia, Bank Pekao, UniCredit Russia, UniCredit Bank Romania nor any of their respective directors, officers or employees nor any other person accepts any liability whatsoever (in negligence or otherwise) for any loss howsoever arising from any use of this document or its contents or otherwise arising in connection therewith. This analysis is being distributed by electronic and ordinary mail to investors, who are expected to make their own investment decisions without undue reliance on this publication, and may not be redistributed, reproduced or published in whole or in part for any purpose. Responsibility for the content of this publication lies with: UniCredit Group and its subsidiaries are subject to regulation by the European Central Bank a) UniCredit Bank AG (UniCredit Bank), Am Tucherpark 16, 8538 Munich, Germany, (also responsible for the distribution pursuant to 34b WpHG). The company belongs to UniCredit Group. Regulatory authority: BaFin Bundesanstalt für Finanzdienstleistungsaufsicht, Lurgiallee 1, 6439 Frankfurt, Germany. b) UniCredit Bank AG London Branch (UniCredit Bank London), Moor House, 1 London Wall, London ECY 5ET, United Kingdom. Regulatory authority: BaFin Bundesanstalt für Finanzdienstleistungsaufsicht, Lurgiallee 1, 6439 Frankfurt, Germany and subject to limited regulation by the Financial Conduct Authority, 5 The North Colonnade, Canary Wharf, London E14 5HS, United Kingdom and Prudential Regulation Authority Moorgate, London, ECR 6DA, United Kingdom. Further details regarding our regulatory status are available on request. c) UniCredit Bank AG Milan Branch (UniCredit Bank Milan), Piazza Gae Aulenti, 4 - Torre C, 154 Milan, Italy, duly authorized by the Bank of Italy to provide investment services. Regulatory authority: Bank of Italy, Via Nazionale 91, 184 Roma, Italy and Bundesanstalt für Finanzdienstleistungsaufsicht, Lurgiallee 1, 6439 Frankfurt, Germany. d) UniCredit Bulbank, Sveta Nedelya Sq. 7, BG-1 Sofia, Bulgaria Regulatory authority: Financial Supervision Commission (FSC), 33 Shar Planina str.,133 Sofia, Bulgaria e) Zagrebačka banka d.d., Trg bana Jelačića 1, HR-1 Zagreb, Croatia Regulatory authority: Croatian Agency for Supervision of Financial Services, Miramarska 4B, 1 Zagreb, Croatia f) UniCredit Bank Czech Republic and Slovakia, Na Príkope 858/, CZ-1111 Prague, Czech Republic Regulatory authority: CNB Czech National Bank, Na Příkopě 8, Praha 1, Czech Republic g) Bank Pekao, ul. Grzybowska 53/57, PL--95 Warsaw, Poland Regulatory authority: Polish Financial Supervision Authority, Plac Powstańców Warszawy 1, -95 Warsaw, Poland h) ZAO UniCredit Bank Russia (UniCredit Russia), Prechistenskaya emb. 9, RF-1934 Moscow, Russia Regulatory authority: Federal Service on Financial Markets, 9 Leninsky prospekt, Moscow , Russia i) UniCredit Bank Czech Republic and Slovakia, Slovakia Branch, Šancova 1/A, SK Bratislava, Slovakia Regulatory authority: CNB Czech National Bank, Na Příkopě 8, Praha 1, Czech Republic and subject to limited regulation by the National Bank of Slovakia, Imricha Karvaša 1, Bratislava, Slovakia. Regulatory authority: National Bank of Slovakia, Imricha Karvaša 1, Bratislava, Slovakia j) UniCredit Bank Romania, Bucharest 1F Expozitiei Boulevard, RO-111 Bucharest 1, Romania Regulatory authority: National Bank of Romania, 5 Lipscani Street, RO-331, 3rd District, Bucharest, Romania k) UniCredit Bank AG Hong Kong Branch (UniCredit Bank Hong Kong), 5/F Man Yee Building, 68 Des Voeux Road Central, Hong Kong. Regulatory authority: Hong Kong Monetary Authority, 55 th Floor, Two International Financial Centre, 8 Finance Street, Central, Hong Kong l) UniCredit Bank AG Singapore Branch (UniCredit Bank Singapore), Prudential Tower, 3 Cecil Street, #5-1, Singapore 4971 Regulatory authority: Monetary Authority of Singapore, 1 Shenton Way MAS Building, Singapore m) UniCredit Bank AG Tokyo Branch (UniCredit Tokyo), Otemachi 1st Square East Tower 18/F, Otemachi, Chiyoda-ku, 1-4 Tokyo, Japan Regulatory authority: Financial Services Agency, The Japanese Government, 3--1 Kasumigaseki Chiyoda-ku Tokyo, Japan, The Central Common Government Offices No. 7. n) UniCredit Bank New York (UniCredit Bank NY), 15 East 4nd Street, New York, NY 117 Regulatory authority: BaFin Bundesanstalt für Finanzdienstleistungsaufsicht, Lurgiallee 1, 6439 Frankfurt, Germany and New York State Department of Financial Services, One State Street, New York, NY POTENTIAL CONFLICTS OF INTEREST UniCredit Bank AG acts as a Specialist or Primary Dealer in government bonds issued by the Italian, Portuguese and Greek Treasury. Main tasks of the Specialist are to participate with continuity and efficiency to the governments' securities auctions, to contribute to the efficiency of the secondary market through market making activity and quoting requirements and to contribute to the management of public debt and to the debt issuance policy choices, also through advisory and research activities. ANALYST DECLARATION The author s remuneration has not been, and will not be, geared to the recommendations or views expressed in this study, neither directly nor indirectly. ORGANIZATIONAL AND ADMINISTRATIVE ARRANGEMENTS TO AVOID AND PREVENT CONFLICTS OF INTEREST To prevent or remedy conflicts of interest, UniCredit Bank, UniCredit Bank London, UniCredit Bank Milan, UniCredit Bulbank, Zagrebačka banka, UniCredit Bank Czech Republic and Slovakia, Bank Pekao, UniCredit Russia, and UniCredit Bank Romania have established the organizational arrangements required from a legal and supervisory aspect, adherence to which is monitored by its compliance department. Conflicts of interest arising are managed by legal and physical and non-physical barriers (collectively referred to as Chinese Walls ) designed to restrict the flow of information between one area/department of UniCredit Bank, UniCredit Bank London, UniCredit Bank Milan, UniCredit Bulbank, Zagrebačka banka, UniCredit Bank Czech Republic and Slovakia, Bank Pekao, UniCredit Russia, UniCredit Bank Romania, and another. In particular, Investment Banking units, including corporate finance, capital market activities, financial advisory and other capital raising activities, are segregated by physical and non-physical boundaries from Markets Units, as well as the research department. In the case of equities execution by UniCredit Bank AG Milan Branch, other than as a matter of client facilitation or delta hedging of OTC and listed derivative positions, there is no proprietary trading. Disclosure of publicly available conflicts of interest and other material interests is made in the research. Analysts are supervised and managed on a day-to-day basis by line managers who do not have responsibility for Investment Banking activities, including corporate finance activities, or other activities other than the sale of securities to clients. UniCredit Research page 18

19 ADDITIONAL REQUIRED DISCLOSURES UNDER THE LAWS AND REGULATIONS OF JURISDICTIONS INDICATED You will find a list of further additional required disclosures under the laws and regulations of the jurisdictions indicated on our website Notice to Austrian investors: This analysis is only for distribution to professional clients (Professionelle Kunden) as defined in article 58 of the Securities Supervision Act. Notice to investors in Bosnia and Herzegovina: This report is intended only for clients of UniCredit in Bosnia and Herzegovina who are institutional investors (Institucionalni investitori) in accordance with Article of the Law on Securities Market of the Federation of Bosnia and Herzegovina and Article of the Law on Securities Markets of the Republic of Srpska, respectively, and may not be used by or distributed to any other person. This document does not constitute or form part of any offer for sale or subscription for or solicitation of any offer to buy or subscribe for any securities and neither this document nor any part of it shall form the basis of, or be relied on in connection with or act as an inducement to enter into, any contract or commitment whatsoever. Notice to Brazilian investors: The individual analyst(s) responsible for issuing this report represent(s) that: (a) the recommendations herein reflect exclusively the personal views of the analysts and have been prepared in an independent manner, including in relation to UniCredit Group; and (b) except for the potential conflicts of interest listed under the heading Potential Conflicts of Interest above, the analysts are not in a position that may impact on the impartiality of this report or that may constitute a conflict of interest, including but not limited to the following: (i) the analysts do not have a relationship of any nature with any person who works for any of the companies that are the object of this report; (ii) the analysts and their respective spouses or partners do not hold, either directly or indirectly, on their behalf or for the account of third parties, securities issued by any of the companies that are the object of this report; (iii) the analysts and their respective spouses or partners are not involved, directly or indirectly, in the acquisition, sale and/or trading in the market of the securities issued by any of the companies that are the object of this report; (iv) the analysts and their respective spouses or partners do not have any financial interest in the companies that are the object of this report; and (v) the compensation of the analysts is not, directly or indirectly, affected by UniCredit s revenues arising out of its businesses and financial transactions. UniCredit represents that: except for the potential conflicts of interest listed under the heading Potential Conflicts of Interest above, UniCredit, its controlled companies, controlling companies or companies under common control (the UniCredit Group ) are not in a condition that may impact on the impartiality of this report or that may constitute a conflict of interest, including but not limited to the following: (i) the UniCredit Group does not hold material equity interests in the companies that are the object of this report; (ii) the companies that are the object of this report do not hold material equity interests in the UniCredit Group; (iii) the UniCredit Group does not have material financial or commercial interests in the companies or the securities that are the object of this report; (iv) the UniCredit Group is not involved in the acquisition, sale and/or trading of the securities that are the object of this report; and (v) the UniCredit Group does not receive compensation for services rendered to the companies that are the object of this report or to any related parties of such companies. Notice to Canadian investors: This communication has been prepared by UniCredit Bank AG, which does not have a registered business presence in Canada. This communication is a general discussion of the merits and risks of a security or securities only, and is not in any way meant to be tailored to the needs and circumstances of any recipient. The contents of this communication are for information purposes only, therefore should not be construed as advice and do not constitute an offer to sell, nor a solicitation to buy any securities. Notice to Cyprus investors: This document is directed only at clients of UniCredit Bank who are persons falling within the Second Appendix (Section, Professional Clients) of the law for the Provision of Investment Services, the Exercise of Investment Activities, the Operation of Regulated Markets and other Related Matters, Law 144(I)/7 and persons to whom it may otherwise lawfully be communicated who possess the experience, knowledge and expertise to make their own investment decisions and properly assess the risks that they incur (all such persons together being referred to as relevant persons ). This document must not be acted on or relied on by persons who are not relevant persons or relevant persons who have requested to be treated as retail clients. Any investment or investment activity to which this communication related is available only to relevant persons and will be engaged in only with relevant persons. This document does not constitute an offer or solicitation to any person to whom it is unlawful to make such an offer or solicitation. Notice to Hong Kong investors: This report is for distribution only to professional investors within the meaning of Schedule 1 to the Securities and Futures Ordinance (Chapter 571, Laws of Hong Kong) and any rules made thereunder, and may not be reproduced, or used by or further distributed to any other person, in whole or in part, for any purpose. This report does not constitute or form part of an offer or solicitation of any offer to buy or sell any securities, nor should it or any part of it form the basis of, or be relied upon in connection with, any contract or commitment whatsoever. By accepting this report, the recipient represents and warrants that it is entitled to receive such report in accordance with, and on the basis of, the restrictions set out in this Disclaimer section, and agrees to be bound by those restrictions. Notice to investors in Ivory Coast: The information contained in the present report have been obtained by Unicredit Bank AG from sources believed to be reliable, however, no express or implied representation or warranty is made by Unicredit Bank AG or any other person as to the completeness or accuracy of such information. All opinions and estimates contained in the present report constitute a judgement of Unicredit Bank AG as of the date of the present report and are subject to change without notice. They are provided in good faith but without assuming legal responsibility. This report is not an offer to sell or solicitation of an offer to buy or invest in securities. Past performance is not an indicator of future performance and future returns cannot be guaranteed, and there is a risk of loss of the initial capital invested. No matter contained in this document may be reproduced or copied by any means without the prior consent of Unicredit Bank AG. Notice to New Zealand investors: This report is intended for distribution only to persons who are wholesale clients within the meaning of the Financial Advisers Act 8 ( FAA ) and by receiving this report you represent and agree that (i) you are a wholesale client under the FAA (ii) you will not distribute this report to any other person, including (in particular) any person who is not a wholesale client under the FAA. This report does not constitute or form part of, in relation to any of the securities or products covered by this report, either (i) an offer of securities for subscription or sale under the Securities Act 1978 or (ii) an offer of financial products for issue or sale under the Financial Markets Conduct Act 13. Notice to Omani investors: This communication has been prepared by UniCredit Bank AG. UniCredit Bank AG does not have a registered business presence in Oman and does not undertake banking business or provide financial services in Oman and no advice in relation to, or subscription for, any securities, products or financial services may or will be consummated within Oman. The contents of this communication are for the information purposes of sophisticated clients, who are aware of the risks associated with investments in foreign securities and neither constitutes an offer of securities in Oman as contemplated by the Commercial Companies Law of Oman (Royal Decree 4/74) or the Capital Market Law of Oman (Royal Decree 8/98), nor does it constitute an offer to sell, or the solicitation of any offer to buy non-omani securities in Oman as contemplated by Article 139 of the Executive Regulations to the Capital Market Law (issued vide CMA Decision 1/9). This communication has not been approved by and UniCredit Bank AG is not regulated by either the Central Bank of Oman or Oman s Capital Market Authority. Notice to Pakistani investors: Investment information, comments and recommendations stated herein are not within the scope of investment advisory activities as defined in sub-section I, Section of the Securities and Exchange Ordinance, 1969 of Pakistan. Investment advisory services are provided in accordance with a contract of engagement on investment advisory services concluded with brokerage houses, portfolio management companies, non-deposit banks and the clients. The distribution of this report is intended only for informational purposes for the use of professional investors and the information and opinions contained herein, or any part of it shall not form the basis of, or be relied on in connection with or act as an inducement to enter into, any contract or commitment whatsoever. Notice to Polish Investors: This document is intended solely for professional clients as defined in Art. 3.39b of the Trading in Financial Instruments Act of 9 July 5 (as amended). The publisher and distributor of the document certifies that it has acted with due care and diligence in preparing it, however, assumes no liability for its completeness and accuracy. This document is not an advertisement. It should not be used in substitution for the exercise of independent judgment. Notice to Serbian investors: This analysis is only for distribution to professional clients (profesionalni klijenti) as defined in article 17 of the Law on Capital Markets. Notice to UK investors: This communication is directed only at clients of UniCredit Bank who (i) have professional experience in matters relating to investments or (ii) are persons falling within Article 49()(a) to (d) ( high net worth companies, unincorporated associations, etc. ) of the United Kingdom Financial Services and Markets Act (Financial Promotion) Order 5 or (iii) to whom it may otherwise lawfully be communicated (all such persons together being referred to as relevant persons ). This communication must not be acted on or relied on by persons who are not relevant persons. Any investment or investment activity to which this communication relates is available only to relevant persons and will be engaged in only with relevant persons. ENP e 11 UniCredit Research page 19