Analyst Note | Apr 22 2020
We have recently updated all our European banking models to incorporate the effects of the coronavirus. We now forecast that the median bank that we cover will book a 38% earnings decline in fiscal 2020 compared with 2019. We then, however, forecast a solid recovery in 2021 and 2022, with median earnings growth of 31% and 17% respectively. A sharp median increase of 125% in loan-loss provisions is the main driver of the expected decline in earnings, but we also anticipate that median revenue will decline by 4%. Importantly, we do not believe that any of the European banks we cover, with the exception of Deutsche Bank, will record a loss for 2020.
The median European bank that we cover has lost half of its market value year to date and now trades at 0.45 times its fiscal 2019 tangible book value. A number of the banks we cover are now trading at their lowest valuation multiples ever. The market is clearly pricing in either deeply discounted capital calls or material structural declines in profitability. If our view, that banks will remain profitable, proves to be correct it is unlikely there will be a broad recapitalisation of the industry as we witnessed during the aftermath of the global financial crisis in 2009. Our midcycle estimates of profitability did not change materially for most banks that we cover. European banks are also better capitalised and more liquid than during the previous downturn. We believe there are good value opportunities among European banks. Wide-moat Julius Baer is trading at 0.7 times our fair value estimate and is offering a rare opportunity to get exposure to its high-quality franchise at such a discount. Narrow-moat Credit Suisse remains attractively priced at 0.5 times our fair value estimate, while narrow moat Santander also looks attractive at 0.5 times our fair value estimate--although we do caution investors that they will be exposed to currency risk as Santander generates the bulk of its earnings in emerging markets.
The outlook for 2020 remains very murky, however. Evidence of this is that the standard deviation of consensus earnings for 2020, collected by Visible Alpha, are wider than the standard deviation of 2022 earnings. Apart from the uncertainty over the extent of provisions that the banks will have to book for loan losses, we are concerned about the potential for mark-to-market losses on the security portfolios held by the banks. For now, liquidity in most asset classes seems to be holding up, thanks to substantial liquidity injections by central banks, but if liquidity becomes a problem banks can once again find themselves holding illiquid, hard-to-value securities on their balance sheets and having to book material impairments. We do not forecast earnings on a quarterly basis and we caution investors not to extrapolate the trend in first-quarter earnings into the full year. The vagaries of the new provisioning accounting standard, IFRS 9, may see a disproportionate amount of losses booked in the first quarter while trading revenues, on the other hand, will benefit from the volatility in the first quarter, which may not continue into the rest of the year.
Loan-loss provisions will increase materially. While we do anticipate a sharp increase in nonperforming loans over the coming quarters, the initial increase in loan-loss provisions will be driven by an increase in provisions for an increase in the expected losses that banks will suffer, more so than an increase in actual nonperforming loans. This will be the first deterioration in the credit cycle for which banks will apply the new accounting standard for impairments of financial assets, IFRS 9. When it comes to recognizing credit losses, IFRS 9 takes a much more forward-looking approach than the previous standard, IAS 39. IFRS 9 will therefore lead to credit losses hitting banks' profit and loss statements much earlier than in the past. We believe the initial hit will be the most pronounced in the first two quarters when banks have to book the provision for loans rolling from stage 1 (performing loans with a 12-month expected credit loss recognized) to stage 2 (loans still not defaulted but with a significant increase in credit risk, with a lifetime expected, credit loss recognized). We would be cautious to simply annualize the loan-loss provisions booked in the first quarter. We expect that IFRS 9 will lead to great divergence across the banks over how they account for expected losses, which will make comparability difficult.
Even before regulators recently urged banks to use their discretion on how to apply IFRS 9, the standard allowed substantial leeway for interpretation. We will be especially interested to see how banks will account for the impact of the extensive government support packages on asset quality as well as banks' own forbearance on loan repayments. We believe that, initially, more pressure will be felt on SME, commercial and corporate loans with the asset quality of household debt holding up, although some early pressure is likely on the unsecured portfolios. We are most concerned about banks with large oil and gas lending books, with Natixis, Credit Agricole, ING and ABN Amro all having material exposure. Although it is too early to tell, we have yet to see material house price declines and we believe governments and banks will bend over backwards to keep people in their homes. The continental European banks tend to have limited exposure to credit card debt, Barclays and Lloyds however have major card businesses. Santander and BNP Paribas have large consumer finance businesses, dominated by auto loans, while volumes would have dried up; and it will be interesting to see how asset quality has held up.
As a rule, we are more bearish than consensus on the outlook for loan losses. We do not incorporate a material change in the outlook for net interest margins into our models--except for banks with material United States dollar deposits. If anything, European yield curves have steepened and this could prove positive over the longer term. Especially large commercial clients have drawn down on committed lines of credit at banks: the higher volumes will be beneficial to net interest income. An interesting trend that we witnessed in the U.S. banks' reporting is a robust increase in deposits. This could be reflection of the drawdown on credit lines--that were then merely deposited again or just merely a reflection on the slowdown of economic activity, with clients depositing liquidity released from working capital. The extension of payment terms offered to support clients may lead to some margin compression as repayments are spread over a longer term, reducing effective interest rates. Lower asset values will affect assets under management and consequently fees for banks with large asset- and wealth management franchises.
It is important to highlight though that equity investments typically make up less than 40% of assets under management for banks' asset- and wealth management operations so the impact of lower market values may be less pronounced than some investors believe. The wealth management businesses of, especially, the Swiss banks will also see the benefit of higher brokerage as client activity increased materially in the first quarter. Primary market investment banking fees are under pressure as, especially, merger and acquisition advisory work and equity underwriting declines, dent underwriting may prove to be more resilient as corporate clients hunt for liquidity. Transactional-based income will be under pressure from the general economic slowdown. Clients have simply transacted much less, which will reduce fees. Lower credit card volumes will also lead to lower card fees, important for the United Kingdom banks, less so for continental banks. Widening spreads and much higher volumes will have a positive impact on trading income.
The U.S. investment banks that recently reported results, all reported strong increases in their trading businesses. What stood out was the strong performance from equity derivatives, which could be a positive sign for BNP Paribas and Societe Generale, both with strong equity derivative franchises. The trading line may, however, be distorted by mark-to-market and counterparty credit risk adjustments, but this is impossible to forecast. The banks have some flexibility in their cost bases, although it is unlikely we will see a good indication of this in the first quarter as management would probably have only turned off the taps in the March. We do believe that banks' variable costs as a portion of overall costs have reduced since the previous downturn. Variable compensation has been reduced as a portion of overall compensation to comply with regulations and the move toward digitalisation has also increased the overall fixed cost base. We will be very surprised to see any announcements of new redundancies and, if anything, we could expect to see a moratorium on previously planned redundancies. It will be a public relations nightmare for a bank to announce large-scale layoffs in the current environment.
Business Strategy and Outlook | 03 Jan 2020
Swedbank is one of the largest retail banks in Sweden and the largest in the Baltic states. While its foundation is based in Swedish savings banks, a position that still benefits Swedbank today and is part of the rationale for our narrow moat rating for the lender, its dominant presence in the Baltics is owed to the acquisition of Hansabank in 2005. Since then, Swedbank has transformed both its Swedish and Baltic lending arms into highly efficient banking operations, taking advantage of economic growth in the Baltics, a booming real estate market in Sweden, and digitalisation initiatives to improve efficiency. As a result, Swedbank has transformed itself into the best-performing Swedish bank recently.
All is not good, however. Reports of suspicious money flows as part of about EUR 135 billion in gross transactions from nonresident customers over the last decade, and more importantly the handling of said transactions, weigh heavily on the bank's short- to medium-term outlook. Not only does the bank face potentially substantial fines from regulators, the scandal takes focus away from Swedbank's core business and investments into the future.
Next to the more direct impact of fines, Swedbank's operations in the Baltics are in limbo. Although management is adamant that the Baltics will remain core, this may not be Swedbank's decision alone. Danske, after winding down most of its Estonian operation amid its money laundering scandal, was required to close shop entirely by Estonian regulators. A similar fate could overcome Swedbank. We do not include such a scenario in our forecasts, given that it is unlikely for Estonia to push out all international banks as the country lacks national banks to support its economic development. But, uncertainty about Swedbank's place in the Baltics as well as its strategy going forward are high as a result.
Fair Value & Profit Drivers | 08 Apr 2020
Our fair value estimate is SEK 183, which corresponds to 1.4 times book value and 28 times earnings based on 2020 estimates. Given the impact of the coronavirus, we think the 11 times 2021 earnings is a more meaningful gauge, however. In the medium term, we anticipate an average return on equity of 12%, which is above our cost of equity assumption of 9%.
Our fair value estimate is primarily driven by net interest margin, provision ratio, and operating cost assumptions. We pencil in 1.3% as our midcycle net interest margin assumption, which is about 6 basis points above 2019. We anticipate a progressive increase in credit costs toward a midcycle level of about 20 basis points, up from just 9 basis points in 2019. Finally, we believe Swedbank will keep its low-cost-leading position relative to peers, although higher anti-money-laundering costs in the form of staff and compliance systems will lower efficiency to some degree. We forecast a midcycle efficiency ratio of 45% compared with 44% in 2019. We included the SEK 4 billion administrative penalty imposed by Swedish supervisors for inadequate AML processes in our 2020 estimates.
For 2020, we have adjusted our estimates based on currently available information and the potential impact the coronavirus pandemic may have on Swedbank. We assume a 10-basis-point drop in the group's net interest margin and more importantly, a spike in the provision ratio from 9 basis points in 2019 to 26 basis points in 2020. Taking into account a first-quarter 2020 loss, as guided by management, we don't expect a loss on a full-year basis. As a result, we currently don't see a material risk to Swedbank's capitalisation.
Scenario Analysis | 08 Apr 2020
In addition to our base-case scenario, we create upside and downside fair value estimates of SEK 226 and SEK 139 per share, respectively.
In our upside scenario, we assume a net interest margin of 1.4% in the medium term while provisions for loan losses stay subdued at about 12 basis points. In this scenario, Swedbank can offset higher anti-money-laundering staffing and systems expenses and maintain its currently low fixed-cost base. As a result, the efficiency ratio declines to 42% from 44% in 2019. In this scenario, we do not include any regulatory fines beyond the SEK 4 billion administrative penalty by Swedish watchdogs. The fair value estimate in this scenario is SEK 226 per share.
In our downside scenario, we expect a stronger impact of higher funding costs on net interest margins in the short term and model for generally lower midcycle net interest margins. Our net interest margin assumption is about 1.2% in 2024. Provisions for loan losses increase to 32 basis points and operating costs increase meaningfully to a cost/income ratio of 48% in the medium term. In this scenario, we forecast another SEK 4 billion fine imposed by U.S. regulators on top of the administrative penalty in Sweden. Our fair value estimate in this scenario is SEK 139 per share. Swedbank remains profitable in our bear-case scenario, protecting its sound capital base.
Economic Moat™
We assign Swedbank a narrow moat rating based on cost advantages and customer switching costs. Swedbank's narrow moat stems from its strong position in the stable Swedish market, where it operates a highly efficient bank while maintaining access to loyal customers through its collaboration with regional savings banks. Nearly every second household in Sweden has an account at Swedbank, and although not all of these are primary accounts, it generates competitive advantages for Swedbank in the form of economies of scale and scope. Underlying our moat assessment for Swedbank is our view of the Swedish banking system, which we rate as good. Swedbank also has a high market share in the Baltics, a segment that has been profitable for Swedbank, but that also carries risks.
Following our banking moat framework, we analyzed cost advantages based on operating, credit, and funding costs and are confident that Swedbank enjoys advantages in operating costs, as evidenced by its low cost/income ratio.
Sweden has one of the most efficient banking sectors in Europe. In Sweden itself, Swedbank has the lowest cost/income ratio among the four largest banks, the result, we believe, of Swedbank's digitalisation efforts amplifying its large customer base. Swedbank has 4.3 million customers in Sweden and shows stable market shares in both deposits and lending. While every bank aims to achieve economies of scale, and operating leverage is inherent in every bank's business model, Swedbank has a competitive advantage beyond its large scale. The cooperation with 58 savings banks in Sweden, which use Swedbank's IT system and offer Swedbank's products to its clients, allows Swedbank to increase its operating leverage while gaining access to loyal regional customers with low risk profiles. The "independent" savings banks contribute about 30% of product sales for Swedbank in Sweden. This additional distribution channel is difficult to replicate, especially since the agreement allows Swedbank to push its products to clients without binding capital in branches and staff. In fact, it allows Swedbank to offload some of the IT development spending to the savings banks. Although the cooperation agreements are renegotiated about once every decade, the close relationship between Swedbank and the savings banks makes it unlikely that a competitor could replace Swedbank in this arrangement. The savings banks collectively own 14% of Swedbank, creating a symbiotic relationship.
We think Swedbank's moat is supported by implicit switching costs based on its ability to cross-sell various products to its large share of customers, connection to loyal customers through its regional savings banks agreement, and high mortgage lending market share. Typically, banks in Sweden do not charge customers when closing accounts or terminating products. Some products, such as fixed-rate mortgages, could require a fee to be terminated. In the larger scheme of things, this is not enough to create a significant switching cost argument, however. We believe the low explicit cost to switch banks is largely owed to the good availability of loan approval, default and income data on banking customers reducing the information asymmetry between former and new bank. This reduces risk premiums incurred by the bank in assessing new customers’ creditworthiness, which if significant would be passed on to customers in the form of explicit switching costs. We do not believe that Swedbank customers are exposed to explicit switching costs that are material enough for a moat rating.
A bank can also create implicit switching costs in the form of customer relationships, trust, and products. We believe that offerings are largely commoditised in Sweden, but economies of scope allow banks with a large product offering to cross-sell to and lock in its clients. Prerequisite for achieving such switching costs is a large customer base, which Swedbank has. Next to Swedbank's high market share in mortgage lending, which tend to be rather sticky products, Swedbank has been expanding its other products. The most prominent example is Robur, Swedbank's asset-management arm, which manages about SEK 1.3 trillion in assets. Swedbank also offers pension, savings, credit card, life and nonlife insurance, and payment products. While none of these are particularly moaty in their own right, a one-stop-shop solution like Swedbank offers can lead to implicit switching costs if combined with strong customer trust and satisfaction.
Strong customer relationships are another form of implicit switching cost, even though this can be costly in the form of staff and fixed costs for branches and therefore is often believed to be suboptimal in the age of digitalisation. Although digitalisation has its perks, such as lower costs through leveraging fixed IT costs on a large customer base, we don't think a fully digital offering is the be all and end all. Swedish banks are front-runners when it comes to online and mobile banking solutions, leaving little room for banks to carve out a competitive advantage on the digital side. Customers seem to be largely agnostic to the minor digital differences in banks' offerings. On the other hand, Swedish customers continue to rank regional banks with local branches highest in satisfactory metrics. As a result, we believe that banks able to connect its digital offerings with a regional branch network can enjoy implicit switching costs through closer client relationships to customers who value such services. Swedbank has the largest customer base in Sweden, with nearly half of every household having an account at the bank. Not all these accounts are primary accounts, but we believe the fact that half of Swedish banking customers choose at least one of Swedbank's offerings indicates a strong trust in Swedbank and allows for easier cross-selling of products. Additionally, Swedbank's collaboration with regional savings banks gives Swedbank access to more loyal customers. In sum, Swedbank can combine its digital offerings with a local branch network, which leads to a high share of customers who value consistency and trust higher than price savings from a potential switch of banks. This is evidenced in Swedbank’s ability to maintain high market shares in Sweden despite challenger banks undercutting the four largest banks in Sweden on mortgage pricing over the last couple of years now.
Although Swedbank itself does not rank high in customer satisfaction surveys in Sweden, the overall customer relationships Swedbank has with clients is likely better than peers’, except for Handelsbanken. The independent savings banks that offer Swedbank’s products to its customers rank in the top of customer satisfaction surveys, allowing for more loyal customers, ultimately benefiting Swedbank.
Swedish banks depend to a large extent on wholesale funding, which leads to stable net interest margins as interest expenses on debt vehicles have more room to decline than deposit funding. This is advantageous in a scenario such as low or negative interest rates; however, it does not generate a competitive advantage through a cycle. Swedbank is no different in that respect. The bank finances about half of its asset base with wholesale funds. Following our moat framework, we would prefer higher levels of deposit funding, in particular non-interest-bearing current accounts, which provide a tailwind in higher interest-rate environments.
From an operating environment point of view, we rate the Swedish banking system as good under our framework, thanks to its stable economic environment and good regulatory setup (which has been on an increasing trend in recent years). Finansinspektionen (the Swedish FSA) and the Riksbank (Sweden’s central bank) are primarily responsible for monitoring compliance and maintaining financial stability in the country. These supervisory bodies believe strongly in tougher capital and liquidity requirements for the system, and they monitor banks closely.
Moat Trend
Swedbank’s moat trend is stable. Although new challenger banks, including online and mobile banks, have aggressively entered the mortgage market, Swedbank can counteract this with its own digital offerings. As competition for customers increases and the financial landscape threatens to become more fragmented as a result of open banking initiatives, we believe Swedbank’s collaboration agreement with regional savings banks will gain in weight as a competitive advantage. In sum, Swedbank can compete on price-sensitive customers while maintaining loyal customers who closer and steadier banking relationships, for which they are willing to pay a premium.
Risk & Uncertainty
We rate Swedbank's uncertainty as very high, given the significant allegations regarding money laundering in the company's Baltic operations, the investigation into potential insider information violations, and inquiries by U.S. authorities about withholding information on the involvement of Swedbank's clients in the tax scheme revealed in the Panama Papers.
As of now, EUR 135 billion in gross transactions is believed to have flowed through Swedbank's nonresident accounts in its Baltic operations. We do not believe that this figure is a good indication of potential fines Swedbank may face. Rather, we think that the potential economic impact can fall in a wide range of possible scenarios, hence our very high uncertainty rating. The ultimate fines, if any, will depend on the culpability of Swedbank, what was known when, and which illegal activities were reported adequately. Bank privacy laws and investigation-related requirements to withhold information from the public make a narrow assessment of the potential impact fruitless.
Swedbank's exposure to the Swedish mortgage market could pose a downside risk in the event of a sudden decline in growth but also asset prices. Higher amortization requirements have cooled real estate markets somewhat in Sweden, but a structural imbalance between affordable housing supply and demand are likely to persist in the medium term.
Financial Strength
Swedbank depends highly on wholesale funding, especially covered bonds, which are correlated to money markets and will affect funding costs when a bond requires rolling. The bank plans to concentrate on covered bonds rather than senior bonds for funding in the near term to circumvent higher credit spreads in the midst of its money laundering scandal. However, we are not convinced that this will shield Swedbank from higher funding costs entirely. Yes, covered bond costs are a function of the creditworthiness of the underlying assets (mortgages), more so than the issuer's credit rating, but a prolonged increased credit spread will eventually affect the bank's funding costs.
Swedbank's core equity Tier 1 ratio at the end of the fourth quarter of 2019 stood at 17% versus a requirement of 15.1%.
Stewardship | 03 Jan 2020
Our Poor stewardship rating is based on management’s unsatisfactory handling of the money laundering case. We had thought that the pattern seemed strangely familiar to Danske’s conduct when faced with similar accusations. Indeed, involvement was denied by both banks until made public by whistleblowers or third parties gaining access to internal documents to prove otherwise, upon which both lenders acknowledged what was revealed but hid behind bank privacy laws when asked about the full extent of the matter. This resulted in a strange attempt to display transparency by way of citing official press releases and reports. The latest similarity between Danske's and Swedbank’s cases has been the release of Swedbank's former CEO, Birgitte Bonnesen, who, like Danske’s former chief executive, used to oversee the Baltic operations. We would have preferred a quick and comprehensive acknowledgement of involvement, getting ahead of the news cycle and showing full transparency in the matter. A bank’s business relies on trust by clients, regulators, and even competitors, and we believe that management’s actions have shown poor stewardship with regard to its most valuable asset.
When it comes to expansion to the Baltics, Swedbank followed a similar merger trend to many of its peers. In 1996, the bank expanded its operations to Estonia, Latvia, and Lithuania when it acquired a 12.5% stake in Eesti Hoiupank, a bank that merged with Hansabank in 1998. In 1999, it acquired a more than 50% stake in Hansabank, and in 2005, it acquired the remaining outstanding shares. It changed its name to Swedbank in 2006. While the bank expanded to Russia and Ukraine in 2007, it closed its Russian and Ukrainian operation in 2013. However, its expansion in the Baltics became volatile when market conditions for financial institutions deteriorated in 2008, as they did for all Swedish banks with exposure to the Baltics. In December 2008, the bank made a cash call from shareholders for SEK 12.4 billion. In 2009, following the financial crisis, its Baltic operations balance sheet suffered from bad debt, and the bank was forced to raise SEK 15 billion in capital via a share issue. While no major Swedish bank was forced to obtain government help during the crisis, Swedbank and SEB were the only two large players who participated in the Swedish government’s loan guarantee program.
