EU-Wide Stress Test Results Adding to Systemic Risk

As professional investors, we have been trained by the markets, the CFA Institute, and business school professors early on to avoid certain kinds of investments. Examples include scams such as Florida swampland (if it is too good to be true, it usually is) and obsolete businesses such as U.S. textile mills (by far Warren Buffett’s worst investment in his storied career) or newspapers & magazines whose revenues are still dependent on print advertising.

As I previously pointed out, Italian banks, along with systemically-important Deutsche Bank, have the distinction of fitting into both categories. Deutsche Bank’s core investment banking & trading businesses are obsolete in the post-crisis, capital-constrained world (think Basel III, Basel IV, Solvency II, and IFRS9), and yet Deutsche’s leadership remains in denial. Meanwhile–as I pointed out in two earlier blog posts (“Strengthening the Global Banking System’s Weakest Link” published on April 12, 2016 & “Italy’s NPLs Still the Global Banking System’s Weakest Link” published on July 7, 2016)–Italian policymakers have for years refused to reform or even acknowledge the growing NPL problem within the Italian banking system. It wasn’t until this year that investors finally forced Italian policymakers and banks to act; more than 7 months into the year, they have yet to provide a stronger or more comprehensive solution that will likely involve a combination of (hugely dilutive) capital raises, offloading NPLs to foreign investors, along with some kind of “bail-in” steps for the Italian bank with the worst balance sheet, i.e. Banca Monte dei Paschi di Seana (BMPS).

The publication of the 2016 EU-Wide Stress Test last Friday was intended to alleviate solvency concerns and to halt the vicious cycle of declining European bank stock prices that began early this year. 51 large European banks holding 70% of all European banking assets were covered. These banks were domiciled in the following countries: Austria, Belgium, Denmark, Finland, France, Germany, Hungary, Ireland, Italy, Netherlands, Norway, Poland, Spain, Sweden, and the UK. Banks in two of the Euro Zone’s most troubled economies, i.e. Greece and Portugal, were not included in the stress test.

Key macroeconomic assumptions/projections of the 2016 stress test (see Figure 1 below) were similar in severity relative to those in the 2014 stress test. E.g. the 2014 stress test assesses a 3-year cumulative real GDP “shock” of -7.0% in the Euro Area, vs. a 3-year cumulative -7.1% shock in the 2016 stress test, as shown in Figure 1 below.

Figure 1: Key Macroeconomic Assumptions/Projections of the 2016 EBA Stress Test
eba2016stresstestassumptions

The 2016 stress test, however, did refine its overall framework by adding two explicit assumptions relative to the 2014 stress test: 1) explicit bottom-up treatment of FX-related trading losses to entities in central and eastern Europe in the event of EM currency depreciation, and 2) cumulative losses of 71 billion euros driven by the concept of “conduct risk,” with 15 of the 51 banks tested having a conduct risk liability of more than one billion euros.

On the surface, the results of the 2016 stress test were relatively benign. The 51 banks in the sample had a weighted average fully loaded CET1 capital ratio of 12.6% as of year-end 2015. This starting point was 150 bps above that for the 2014 and 400 bps over that for the 2011 stress test. Since December 2013, the fully loaded CET1 capital for the 51 banks sampled has increased by 180 billion euros. Under the “adverse scenario,” the 2016 stress test estimates a decline of 340 bps in the weighted average CET1 ratio to 9.2% by the end of 2018, driven mostly by projected credit losses, conduct losses, and losses stemming from counter-party risks. More importantly–out of the 51 banks sampled–only 10 banks finished with a fully loaded CET1 capital ratio of less than 8.0% by year-end 2018 (post any mitigation measures enacted after year-end 2015). They are: BMPS (-2.44%), Allied Irish Banks (4.31%), Raiffeise-Landesbanken-Holding (6.12%), Bank of Ireland (6.15%), UniCredit (7.10%), Barclays (7.30%), Commerzbank (7.42%), SocGen (7.50%), Deutsche Bank (7.80%), and Criteria Caixa (7.81%). Figure 2 below shows the EBA’s projected fully loaded CET1 capital ratios as of year-end 2018 on a bank-by-bank basis:

Figure 2: Fully Loaded CET1 Ratios at Year-End 2018 Under the EBA’s Adverse Scenario
(source: Goldman Sachs, EBA)
CET1 Ratios

Unfortunately, as I have asserted in my weekly global macro newsletters and my last two blog entries on the Italian banking system, what investors are looking for is not a “rubber stamp” approval by bank regulators on the integrity of European banks’ balance sheets or a halfhearted fund-raising effort by BMPS, but a strong, decisive, and system-wide capital-raising effort with government backing, along with a multi-year reform plan to rebuild the sustainability of long-term profits currently being impeded by: 1) prevalence of negative interest rates across much of the Euro-wide yield curve, 2) slow structural growth rates, especially within countries such as Italy, Portugal, and Greece, and 3) obsolete business models where certain business operations (e.g. trading) are being quickly replaced by hedge funds and fintech companies.

Over the last two trading days, European banks’ stock prices (ETF ticker EUFN) have fallen by nearly 3%. In particular, BMPS is down by 13%, UniCredit by 15%, Barclays by 5%, and Deutsche Bank by 7%. Major criticisms of the framework/results of the EBA’s 2016 stress test have been as follows:

  1. Banks in the region’s two most NPL-challenged countries, Greece and Portugal, were not included in the stress test;
  2. Under the BMPS rescue plan conceived last Friday, the bank will securitize its entire bad loan portfolio (face value of 27.7 billion euros) at a price of 33 cents on the euro. A securitization at this price is below investors’ expectations; this means if other Italian banks are forced to take similar haircuts on their NPLs, their capital ratio position would be less than what the 2016 stress test currently implies, thus requiring a more substantial (and dilutive) capital raise on the part of Italian banks;
  3. The stress test did not include the impact of regulations that are likely to be implemented over the next several years. Known collectively as “Basel 4,” KPMG estimates that these would lead to an extra 350 billion euros of capital requirements for the world’s largest 100 banks.

Bottom line: The publication of the EBA’s 2016 stress test results have failed to quell investors’ concerns of a systemic meltdown of the Italian banking system. Moreover, a successful recapitalization of BMPS is far from assured, as its capital raise is contingent on the sale of its 27.7 billion euro NPL portfolio, the latter of which will likely result in a larger-than-expected haircut on the Italian-wide 360 billion euros of NPLs. Continue to underweight European financials, especially Deutsche Bank and Italian banks in general.

Italy’s NPLs Still the Global Banking System’s Weakest Link

I last discussed the vulnerabilities in the Italian banking system in our April 12, 2016 blog post (“Strengthening the Global Banking System’s Weakest Link“), where I asserted that–given its global inter-connectivity  today–the world’s financial system can only be as strong as its weakest link. Typically, a liquidity or solvency issue can linger on indefinitely, simply due to the absence of external shocks or because the overall global economy is doing well. In the case of the current NPL issues with the Italian banking system, an effort in January to encourage sales of NPLs by providing government-backed guarantees unfortunately attracted investors’ attention to Italian banks’ NPL issues. When we last covered this issue three months ago, Italian banking stocks were “only” down 20% YTD; today, they are collectively down by 55% YTD.

The vulnerability of the Italian banking system–and by extension, that of the Western European banking system–has come under increased scrutiny over the last several months, exacerbated by: 1) the unexpected, ongoing deflationary malaise in much of the developed world; the May 2016 Italian inflation reading was -0.3% year-over-year, worse than market expectations of -0.2%. June 2016 Italian inflation is expected to hit -0.4% year-over-year, resulting in six straight months of deflationary readings, 2) the dramatic flattening and downshift of the Western European yield curve; globally, nearly US$12 trillion of government bonds now have negative yields, and 3) an unexpected vote for “Brexit,” equivalent to a negative growth shock within the EU, as well as heightened political and economic uncertainty.

The Italian banking system in particular is saddled with 360 billion euros of NPLs, equivalent to about one-third of all of the Euro Zone’s NPLs. Moreover–as efforts since January have demonstrated–a concerted sales effort in NPLs in Europe is not a simple task. Firstly, EU rules explicitly ban the use of government-backed guarantees to cushion NPL losses. Secondly, the average restructuring period for Italian bad loans is an abnormally long 8 years; a quarter of cases take 12 years. Finally, the European market for NPLs is small and underdeveloped relative to the overall stock of NPLs in the banking system. In other words, the market for selling Italian NPLs is relatively small, and is almost non-existent without government-backed guarantees (e.g. A proposal by Apollo to purchase 3.5 billion euros of NPLs held by Italian bank Carige back in March made no progress). Italy’s NPL issues are especially concerning given the lack of core profitability of the Italian banking system (see Figure 1 below).

Figure 1: Return on Regulatory Capital of European Banks by Country – June 2015 (source: EBA, Goldman Sachs)

returnoncapitalEBA

It is generally agreed upon that an Italian government-led recapitalization of 40 billion euros into some of Italy’s largest banks (Unicredit, BMPS, and Intesa, for example) would be adequate to resolve the Italian NPL issue, as long as Euro area growth re-accelerates; at the very least, the immediate probability of a Euro-wide banking contagion would be reduced by an order of magnitude. There are two real obstacles to this “happy scenario,” however: 1) the EU, backed by Germany, is resistant to any Italian government-led efforts to recapitalize the banks at no cost to Italian bank debt holders, as this directly goes against EU rules. Any attempt to “bail-in” Italian banks would increase contagion risks among all of EU banks as both depositors and debt holders will likely take their capital and flee to either the U.S. or other safe haven asset classes, such as gold, and 2) Italian Prime Minster Matteo Renzi has promised to resign if he loses the constitutional referendum to be held in October. Recent opinion polls suggest Renzi’s campaign will fall short; this will likely lead to significant Italian and EU-wide instability given the surge of the populist Five Star Movement in recent opinion polls. Seen in this light, the fragility of the Italian banking system is an ongoing cause for concern.