The European Banking Authority has announced the launch of stress tests of the banking sector in the European Union. The exercise will involve 48 banks, including PKO BP and Pekao SA from Poland.
The results will be announced on November 2nd, 2018. The distant date is probably the result of great emphasis placed on the quality of the results. The previous editions of the tests were carried out at a much faster pace.
The adverse scenario provides for a sudden shock on the asset market and the return of a deep recession. These will be the first stress tests after a two-year break, in accordance with the previous announcements by the European Banking Authority (EBA). Since last year, every second year the stress tests are alternately supplemented by a transparency exercise. Last year’s exercise included 132 banks from 25 European countries, so the coverage was much wider than in the case of stress tests. The goal of the transparency exercise, implemented since 2011, is to strengthen market discipline by supplementing the disclosure requirements under Pillar 3. In this way, the stress tests and the transparency exercise are supposed to complement each other every year.
The number of banks covered by the tests is very similar to two years ago, when 51 European institutions participated. The fact that some are missing this year is the result of the tests carried out two years ago. The Banca Monte dei Paschi di Siena is not included because it was recapitalized a year ago by the Italian government in a so-called precautionary recapitalization.
Another important absentee is Banco Popular Espanol, just a year ago ranked as the 5th largest bank in Spain. It is the first institution in the EU subjected to an orderly liquidation (resolution) under the BRRD directive, i.e. pursuant to pan-European legislation and conducted by a European authority. Its assets were ultimately bought by Santander.
Meanwhile, the Polish bank, Pekao SA, has joined the ranks of the participants. The stress tests are carried out at the highest consolidated level, so Pekao SA previously participated in them, but only as a part of the UniCredit group. This year, after UniCredit sold a 40 per cent of stake, Pekao SA is participating on its own.
The results of the previous tests showed that since 2011 the banks have strengthened their CET1 core capital by approx. 400 bps, up to an average level of 13.2 per cent at the end of 2015. The two-year recession assumed in the extreme adverse scenario would reduce the average value of that ratio by 380 bps, to 9.4 per cent, and credit risk would have the strongest impact here. In this year’s edition – according to the structure of the scenario – credit risk may be a very important, but secondary factor.
The previous tests revealed a large number of weak banks. They all became a starting point for the subsequent SREP supervisory activities. Besides Banco Popular and Banca Monte Paschi di Siena, whose capital would shrink under extreme circumstances by 1,451 bps in three years to minus 2.44 per cent, the tests revealed several other sensitive institutions, whose capital would suffer a lot of damage in adverse conditions.
These included, among others, Allied Irish Banks (minus 847 bps), Royal Bank of Scotland (minus 746 bps), the German state development bank of North Rhine-Westphalia NRW Bank (minus 742 bps), Bayerische Landesbank (minus 690 bps), Commerzbank (minus 636 bps) and Deutsche Bank (minus 540 bps), as well as Raiffeisen, UniCredit and Barclays.
In the meantime, some of them have increased their capital, issued equity instruments, or sold part of the business (like UniCredit). But will this be enough to exhibit resilience in extreme conditions? Have the activities of the Single Supervisory Mechanism (SSM) in the Eurozone, and other national supervisory authorities increased the resilience of these institutions?
In total, 48 banks will participate in the tests, including 33 banks from the Eurozone, 14 banks from remaining EU member states, and one institution from Norway. Taken together, they represent 70 per cent of the European Union’s banking sector assets. Based on the test results, the supervisory bodies will reassess the resilience of the banks to shocks, identify the areas of uncertainty and will take pre-emptive activities. The tests are also intended to strengthen the market discipline due to the fact that the cohesive and detailed data on individual banks will be made available to the public.
“The stress tests will be a moment of truth for the banks,” said Daniele Nouy, the Chair of the Single Supervisory Mechanism (the body operating at the European Central Bank) a few days before the announcement of this year’s scenarios.
Methodology and new challenges
The tests are carried out throughout the European Union based on a common methodology developed by the EBA. This is intended to allow an assessment of the banks’ resilience in a comparable manner. The methodology of the current tests was announced well in advance. There is no established capital threshold – as in the previous edition – which would serve as a limit indicating whether the bank has passed the tests or not. Such a criterion was still used as a benchmark in the tests four years ago.
As part of the tests, the banks impose a method on their models and “samples” of assets, checking their resilience to the assumed external conditions. By extrapolating the results of the “samples” on the entirety of their assets, the banks calculate their hypothetical losses, and these are translated into the deterioration of capital ratios. These results are then carefully checked by local supervisors in coordination and in a dialogue with the EBA. Compliance with the methodology and the quality of results is ensured by the local supervision authorities – for the Eurozone banks this is the SSM, and for the Polish banks this is the Polish Financial Supervision Authority (KNF). After verification of the banks’ calculations, supervisors may order the calculations to be repeated or request the provision of additional data.
There will be two scenarios – a baseline scenario and an adverse scenario. The tests will examine the impact of credit risk, including securitization, market risk, counterparty risk and credit value adjustment risk (CVA), as well as operational risk, including conduct risk. The effects of the scenarios are to be reflected not only in the changes in capital ratios, but also in the net interest income and in the profit and loss account.
The exercise will cover a period from the Q1’18 to the end of 2020. Because of this – apart from the scenarios – there are two big changes.
The effect of Brexit and IFRS 9
The effects of Brexit have already been included in the variables forming the baseline scenario. In the years 2018-2020 the British economy is expected to grow at a rate of 1.6-1.7 per cent. Among the EU countries a lower rate is only predicted for Italy and Belgium. While unemployment in the European Union is forecasted to decline, it is expected to increase slightly in the United Kingdom. Inflation is expected to be significantly higher. Long-term interest rates are expected to grow slightly faster. In the adverse scenario, these phenomena are strengthened, but the United Kingdom is not among the countries most heavily affected.
On the other hand, it is not entirely certain whether, and in what way, the two scenarios take into account the effects of Brexit (for the European Union and for the United Kingdom) in the event of a so-called “hard Brexit”, i.e. if the United Kingdom leaves the European Union without any trade agreement or transitional periods. That is, if the WTO rates suddenly become applicable in trade. Such a risk exists and the scenarios do not indicate that it has been fully taken into account.
In connection with the IFRS 9 entering into force starting from this year, the banks are supposed to take into account its impact on the profit and loss account and on capital already in the output data, as well as throughout the forecast horizon. The European Banking Authority’s July exercise showed that as a result of the application of IFRS 9, and especially the Expected Credit Loss, on average the CET 1 core capital decreases by 45 bps. The banks announcing the results for the last year are already publishing such estimates.
“The total negative impact should not exceed 35 bps (…) Our capital ratio will not be immediately burdened with 100 per cent of the requirements. The burden will be spread out over 5 years. In the first year it will be 5-10 bps,” the vice-president of ING BŚK Bożena Graczyk said at a press conference.
At the same time, with respect to many issues the standard requires interpretations which are not yet available. The Polish Financial Supervision Authority has just started working on Recommendation R to ensure that the banks will be able to assess impairment losses more consistently. This recommendation could result in a revision of the opening balance. The latter could also further change due to tax interpretations.
The effects of Brexit, the risks associated with increased protectionism and the entry into force of IFRS 9 could prove to be the biggest challenges for ensuring the comparability and consistency of the results of this year’s tests.
A market shock will come first
The baseline scenario is based on the latest macroeconomic forecast of the ECB. The adverse scenario has been prepared by the European Systemic Risk Board (ERSB). The baseline scenario forecasts GDP growth in the EU and in the euro area at 5.9 per cent over three years (9.8 per cent in Poland), a decrease in unemployment, respectively, to 7.3 per cent (Eurozone) and 6.4 per cent (EU) in 2020 (to 4.1 per cent in Poland), annual inflation (HICP) at the level of 1.4-1.7 per cent in the Eurozone, and at 1.7-1.9 per cent in the EU (in Poland it should reach the middle of the band of the Polish Monetary Policy Council’s target), and an increase in long-term interest rates by 50 to 60 bps in the years 2017-2020 (by 50 bps in Poland). In an economic improvement and the accompanying mild increase in prices and financing costs, the banks will have the ability to make more money than in previous years.
The adverse scenario reflects the four systemic risks identified by the ERSB as the most material threats to the stability of the European Union’s financial sector. The first one is the most significant, while the others are derived from it. These threats include:
- an abrupt and sizeable repricing of risk premia in global financial markets, triggered, for example, by a policy expectation shock, leading to a tightening of financial conditions;
- adverse feedback loop between a weak bank profitability and low nominal growth;
- public and private debt sustainability concerns amid a potential repricing of risk premia and increased political fragmentation;
- liquidity risks in the non-bank financial sector with potential spillovers to the broader financial system.
The ESRB states that financial markets are susceptible to abrupt and sizeable revaluations, as stock and bond prices are very high, which is combined with minimal levels of volatility. The repricing of the risk premia could be caused by political uncertainty or geopolitical tensions, or a change in expectations regarding economic policy in the major economies of the world. According to the ESRB’s message this relates to the United States. The stock and bond markets there could start a fire that would spread all over the world.
Asset prices in global financial markets would decrease, and the yields of securities would rise throughout the curve. This would be more pronounced for countries with higher public debt stability risks. The volatility in the markets would result in the unwinding of short positions, and the high leverage in the business sector would amplify the shock.
The scale of declines
How severe would the market shock be? It would vary in the different EU countries and between the different asset classes. Capital would flow out of many countries perceived in various market segments as more risky. And so, the greatest volatility in interest rate swaps (IRS) would be recorded in Poland and Hungary, much more at the long end of the curve than at the short end. According to these assumptions, the GBP interest rate swaps would behave in the same way as in the Eurozone. However, real earth shattering changes would occur in the case of interest rate contracts for the USD.
In the case of government bond yields, the securities of Greece, Italy, Spain and Portugal would be the most susceptible to shocks – their yields would increase by over 100 basis points. On the other hand, Polish government securities would be very resilient, and the potential for growth in their yields would be lower than in the case of many Eurozone countries. On the other hand, the yields of United States 10-year bonds would skyrocket by 235 bps in 2018, and then fall to a difference of 192 bps compared to the starting point. On average long-term interest rates in the European Union would increase by 83 bps in 2018.
The shock would start on the stock and ETF markets and they would be subjected to the strongest fluctuations, especially in 2018. The adverse scenario assumes that the value of the S&P500 would drop by 23.6 per cent by 2020; however, in 2018 the stock prices in the United States would fall by 41.4 per cent. The widest FTSE index would lose 27.5 per cent by 2020, the CAC 40 would lose 33.2 per cent, and DAX 30 would lose 31.4 per cent, but in 2018 the share prices in the European Union would fall by an average of 30.1 per cent. Against this background, the broadest index of the Warsaw Stock Exchange (WIG) would drop by only 24.1 per cent. A smaller revaluation would take place on the commodity market – the BCOM index would lose 14.8 per cent, and the Brent crude oil three-month futures would only lose 3.8 per cent.
The EUR would gain 9.6 per cent against the GBP, and would appreciate against all currencies outside the Eurozone, including an 11.2 per cent appreciation against the PLN. However, it would weaken by 5.4 per cent against the USD and by 8.2 per cent against the CHF. Prices of residential real estate in the EU member states would fall by 27.7 per cent in relation to the baseline price forecast for 2020, and the cumulative decrease in real estate prices in the European Union in the time horizon of the scenario would amount to 19 per cent. In the tests the banks must also take into account the effects of these shocks on the economy.
A fall in global demand would lead to a decline in economic activity in the EU, which would translate into GDP being 8.3 per cent lower than in the baseline scenario in 2020. The result of two years of recession would bring a drop in GDP in the Eurozone and in the European Union by 2.4 and 2.7 per cent, respectively, at the end of 2020. In relation to the baseline scenario, unemployment in the European Union would increase by 3.3 percentage points by 2020, and consumer price growth (HICP index) would be 1.9 percentage points lower in 2020. As for non-European economies, their GDP would be between 2.5 to 7.4 per cent lower.
Additional internal shocks were assumed in the tests in the case of 11 countries, including Poland, although the ESRB points out, that they do not reflect any significant and particular imbalance. In the case of Poland in the adverse scenario, in 2018 GDP growth declines to 1.1 per cent, in 2019 there is a recession, and in 2020 there is stagnation. The three-year balance is a 0.2 per cent drop in GDP compared to 2017. In the adverse scenario, after three years the GDP would be 9.4 per cent lower than in the baseline GDP forecast.
The economic shock is the most abrupt in 2018, after which it gradually softens, but not everywhere, as some economies experience it with a certain delay. It has a similar course as in the 2016 tests, except that the decline in GDP is deeper – in the three-year horizon by more than one percentage point.
The shock comes from the outside, and therefore in the EU the countries with more open economies are the most vulnerable to shocks. This also applies to countries characterized by high levels of overvaluation in their property markets. The assumed drop in residential real estate prices in Sweden is as high as 56.4 per cent, and in the case of commercial real estate it amounts to 50 per cent in relation to the baseline scenario. The third category of the most susceptible countries are the ones where the stability of the public debt is in doubt.
An increase in unemployment and an increase in long-term interest rates would weaken the creditworthiness of households and their ability to service their debts, including mortgage loans. A decrease in production combined with an increase in corporate bond yields would increase the credit risk of enterprises.
An increase in credit risk means an increase in non-performing loans (NPL), which further adds to the mountains of bad debts that haven’t been sorted out after the previous crisis. They currently amount to approx. EUR800bn. A drop in demand would put pressure on the price of credit, and on the other hand, an increase in the risk of banks increases their financing costs. The banks’ ability to generate income falls, and their losses increase. After the tests, as part of the SREP supervisory activities, the SSM will decide individually whether a given bank that has demonstrated a capital shortfall in the adverse scenario will have to be recapitalized. “Banks with a capital deficit in the baseline scenario will of course have to cover it,” said Daniele Nouy.