A cautious outlook on EU banks
admin | October 5, 2018
This document is intended for institutional investors and is not subject to all of the independence and disclosure standards applicable to debt research reports prepared for retail investors.
The post-crisis regulatory cleansing has tightened the correlation of credit spread to ratings category among banks and insurers. Yankee banks continue to exhibit more spread volatility than their US counterparts, in part due to more challenging markets in Europe and, in some cases, ongoing regulatory investigations. We evaluate opportunities to selectively reduce holdings among EU banks.
Correlation between US bank spreads and ratings has tightened
For the last 18+months, domestic bank spreads have been more closely correlated than in the prior decade. This should seem intuitive as the banking system benefits from the “cleansing” process brought about by the financial crisis of 2008-09. As seen in the regression graph below, the domestic bank 5-year senior holding company (HC) bond index follows a very close correlation to the mean spread regression. In select cases there are some reference bonds that are wide of the regression line, but for seemingly logical reasons. For example, the KEY 5.10% ’21 senior HC notes trade tight as they are the only senior HC bond anywhere close to 5-year maturity, and the front-end of the curve tends to trade rich. We think it is the scarcity value of the KEY notes that drives the spreads so much tighter than for peers’ bonds.
Exhibit 1: Domestic bank holdco 5-year comps (g-spread)
This tight correlation is also evident in the insurance space, but to a lesser extent. The REIT issuers tend to trade with the least correlation to the mean among the three primary financial sectors, which could be due in large part to the low trading volume of the bonds and the small issuance sizes. This reinforces the idea that diminished liquidity in a bond or issuer will elevate the risk premium needed by investors. Until the domestic banking or insurance space experiences capital losses or other material challenges, we don’t expect this correlation to break-down.
Yankee banks still facing more volatility than US banks
European banks continue to trade wide of the domestic issuers, and with less correlation to the mean, however with less variation than in recent years. As seen in the second graph, the variance for bank spreads from the mean regression is slightly more than for the US banks (lower r-squared value vs US banks). This will remain true for European banks as long as they continue to face more challenging markets in Italy, the UK, and peripheral markets where they operate. Additionally, banks from a number of countries have faced recurring operational problems, key among them recently have been money laundering violations.
Exhibit 2: EU bank holdco 5-year comps (g-spread)
The money laundering probe at Danske is escalating, with the incremental risk widening spreads. Danske shares are down 8% in EU markets as warnings for the size of the fine are increasing to €1.8 billion ($2.07 billion) or more. This seems manageable from an earnings standpoint given the last twelve months net income of $3.07 billion. Nevertheless, the bank had cited higher operating costs and impacts from IFRS9 as eating into profits at the 1H18 presentation (marginal profits low with ROA of 56 bp).
Danske could temporarily be shut out of the debt funding markets until primary law enforcement and regulatory agencies finish their investigation, and levy fines and/or penalties. If the process drags on for weeks or months, it could erode a bit of liquidity. This is not an immediate concern given their liquidity coverage ratio was 142% at 1H18. What’s more, while the maturity curve for Danske is very front-end heavy, most of the company’s wholesale funding is in Danish krone, which should ease funding pressures when they come back to the market.
In June this year DANBNK 3.875% ’23 senior notes were issued at +120 bp, and most recently traded only slightly wider at +128 bp (8 bp tighter w/w too). We look to reduce exposures at current levels given what could still turn up in the course of the investigations. We remain cautious on the ’23 issuance in June based on the poor relative value to peers, which preceded any hint of legal problems and reinforces our level of caution. For example, the BNP 3.50% ’23 SNP was issued at +90 bp in mid-February and now trades +115 bp, is rated one-notch higher than DANBNK by both S&P and Fitch, and does not face the legal issues of Danske.
While most of the European banks have materially improved capital, liquidity and asset quality levels, which has reduced inherent risk in the system, there remains a problem with valuation. We had been bullish on EU banks for a number of years coming out of the banking crisis of 2012. The approach proved rewarding until 1Q16 when the broader market sold off, but has been a good play for investors again into 2018. However, given the relatively low spreads among EU banks, and the elevated risk in the system, it may be a good time to reduce selective holdings among the names.