Decoding Stress in the Banking Sector 

The details: 

• We view Silicon Valley Bank (SVB), Silvergate Capital and Signature Bank as outliers due to outsized business concentration risks, lack of diversity in loans and deposits, and poorly managed asset-liability duration mismatches. They were not subject to the regulatory stress tests or liquidity coverage ratio requirements that larger banks face, and oversight was generally looser because they were not viewed as systemically important.
• We do not expect the deposit runs and liquidity problems at the failed banks to spread to other regional banks. Some regional banks that share individual characteristics with the failed banks (size, geography, asset investments) have seen their stocks pressured. However, in our assessment, regional banks are generally more diversified, have a greater amount of retail deposits and fewer uninsured deposits, perform internal stress tests, and appear more experienced with duration and interest rate exposure control. In addition, regulatory oversight of small banks could increase as result of the crisis.
• Banks are currently seeing deposit costs increases, which is slowing the widening of net interest margins. With this and the greater dependence on lending at the regional banks, our credit research has generally favored large US banks over regional banks in recent months.
• We expect risk premiums to remain elevated as markets work through the discovery phase of this situation.

The details: 

• The Fed remains on course despite market turmoil. The Federal Open Market Committee (FOMC) announced its decision to raise the federal funds rate by 25 basis points on 22 March 2023, in line with market expectations. The FOMC has been careful not to surprise markets during this hiking cycle, and this meeting was no different. In our view, the move signals confidence in the US banking system and reaffirms the FOMC’s commitment to bringing inflation down.
• Emergency intervention. We view the SVB situation as a liquidity crisis, not a solvency crisis. The Fed implemented the Bank Term Funding Program (BTFP) to address liquidity and the duration mismatch. The Fed should not be at risk of loss due to high-quality collateral and $25 billion in funding backed by the Treasury. But the Fed is taking duration risk in the name of avoiding fire sales that could further transmit bank stress through the system. 
  

The details: 

• With significant financial assistance from the Swiss government, UBS agreed on 19 March to acquire Credit Suisse for 3 billion Swiss francs in an all-stock deal. In the process, the Swiss government determined that Credit Suisse had reached the "point of non-viability," which triggered a full writedown of the group's deeply subordinated additional tier 1 (AT1) bonds. (AT1 bonds, a form of debt that emerged after the 2008 global financial crisis, are designed to be converted to equity or written down to improve a bank’s capital when capital levels fall below required thresholds.)
• In our view, the UBS-Credit Suisse deal should help prevent contagion and will have little direct fundamental impact on other global banks. Global banks have exposure to each other through covered bonds (very senior debt backed by mortgages and held for liquidity purposes) and as counterparties, but these should not be affected because of this acquisition in our view.
• Though we think contagion risk is limited, we expect market volatility to persist for global banks. This should generally put upward pressure on bank funding costs, both wholesale funding and retail deposits, which may lead to tighter lending standards that exacerbate already weakening economic projections.
• Near term, we anticipate greater dispersion between stronger issuers and weaker issuers, and between more senior and more subordinated debt in the capital structure. Banks still need to address their regular funding needs and will have to return to the new issue market at some point; issuing into weakness could weigh on spreads.

The details: 

• We believe the Credit Suisse AT1 writedown illuminated two key takeaways for investors. First, losses at this level of the capital structure are possible even for a global systematically important financial institution. Second, the AT1 level of subordinated debt is highly susceptible to decisions made by local regulators, who often have a number of different stakeholders in mind.
• Importantly, the structure of Swiss AT1s is different from that of AT1 bonds issued by United Kingdom and eurozone banks. UK and eurozone AT1 bonds typically stipulate that they will be converted to equity or temporarily written down. By contrast and per local regulatory framework, the Swiss regulator was obligated to permanently write down the AT1 debt in full. Conversion and temporary writedown options would avoid treating equity holders as senior to AT1 holders, a detail of the Credit Suisse deal that created substantial backlash. However, these measures would not reduce the likelihood of the AT1 bonds being bailed-in (the process through which private creditors take losses). The Bank of England and European Banking Authority emphasized these differences, which appears to have reassured investors and helped calm markets for now.
• AT1 bonds are callable instruments. At this time, almost all calls are likely uneconomical due to the steep price declines the market experienced after the deal announcement. Though much of the market has rebounded, we expect continued choppiness. Technical factors could also play a role as funds dedicated to investing in subordinated debt may need to sell securities in order to meet redemptions.
• Markets penalized T2 bonds along with AT1s, but in our view there is an important difference between deeply subordinated AT1 bonds and more traditional subordinated T2 bonds. AT1 bonds are considered "going concern capital," meaning they are specifically designed to absorb losses during times of distress so that a bank can recapitalize itself and continue operations. T2 bonds, though still subordinated debt, are considered "gone concern capital." This means they are only bailed-in once AT1s have been exhausted and there are no options other than more extensive government intervention and potential breakup of the bank. T2 bonds, based on the regulatory framework, cannot as easily be bailed-in.
• Senior holdco bonds operate similarly to T2 subordinated debt, but they are more senior in the capital structure and cannot be bailed-in until T2 bondholders have been fully wiped out. (In countries where banks can operate without a holding company, senior non-preferred bonds were created to function similarly to holdco bonds.)
• In a liquidity event, as was the case for Credit Suisse, T2 and senior holdco bonds are less likely to be affected. However, if a bank experiences more severe capital shortages, they are more likely to be bailed-in. In our view, spillover volatility in the T2 and senior holdco bond markets may not be warranted short term, but this debt could underperform longer term if banking system stress leads to economic recession.

The details: 

• European bank loan books are generally well diversified across retail and corporate clients, and corporate clients tend to be active across a range of economic sectors. Deposit bases are usually diversified across demand, savings and time deposits from retail and corporate clients. Deposits tend to be stickier primarily due to fewer alternatives (i.e., money market funds), less competitive markets dominated by large national-champion banks, lower loan growth and lower absolute rates.
• The deposit beta (the percentage of rate increases passed on to depositors) is generally much lower in Europe. While deposit betas naturally increase as rates rise, Europe has substantially more sight deposits (deposits earnings no interest) than term deposits with higher rates.
• Regulations implemented in Europe since the global financial crisis (GFC) have strengthened liquidity profiles (e.g., the liquidity coverage ratio, which requires banks to hold enough high-quality liquid assets to meet estimated stressed cash outflows for a 30-day period). This has helped promote consumer confidence, in our view a critical factor in bank stability. In the event a bank was required to sell bonds for liquidity purposes (which we consider unlikely), many of these bonds are classified as available for sale. These bonds are marked-to-market through shareholders equity for most European banks, meaning unrealized losses are already factored in to current bank capital ratios, and a sale would not trigger additional losses from a regulatory capital perspective. This was not the case for SVB in the US, which received an exemption based on its size.
• SVB had a significant amount of excess deposits invested in bonds. While European banks hold some government bonds for liquidity purposes, they tend to rely more on wholesale funding (i.e., financing through financial markets or heavy reliance on covered bonds) than US banks. The quantum of excess deposits in Europe is less; however, it is worth noting that loan-to-deposit ratios for most European banks are currently near their lowest/strongest level since the GFC. During the COVID-19 pandemic, deposit volumes increased but bond holdings remained fairly stable as these banks chose to hold more liquidity in cash than securities. This has helped create a solid starting point for the sector in our view. Japan is the global exception as a country that has significant excess deposits invested in government bonds. At this time, Bank of Japan monetary policy remains fairly loose, with lower rates compared to other developed markets. In our view, this is an area to watch closely when the country eventually begins interest rate increases.
• Going forward, bank investors who had been focused on a potential earnings tailwind from higher interest rates may shift focus to banks’ deposit base, the risk of potential deposit outflows and how excess deposits are being deployed.
  

The details: 

• After each economic boom, we look for the hidden rot in the system. Some of it has finally surfaced with the collapse of SVB, Silvergate Capital and Signature Bank. The economy has held on to the late cycle despite some typical hallmarks of a recession, such as the inverted yield curve, the magnitude of Fed tightening, and the slow-motion decline in corporate earnings (S&P 500 profits turned negative in 4Q22 and a few sectors have entered an earnings recession). We see the SVB crisis as a symptom of more acute stress in the economy.
• While the SVB collapse was a failure of risk management rather than a solvency issue, it further tightens credit conditions. As the lenders and key drivers of the credit cycle, financial institutions can be a catalyst for the next crisis. Credit rating agencies are on alert, with Moody’s downgrading its outlook for the US banking system to negative from stable on 14 March. We’ll be watching for indications of a broader collapse in corporate earnings.

Will Banks Take Advantage of the Fed's New Funding Facility?

The Fed launched its Bank Term Funding Program (BTFP) to help the nation’s banks meet deposit needs. However, it’s unclear how much uptake the program will get given the traditional stigma associated with borrowing from Fed facilities. Regional banks tend to use the Federal Home Loan Bank (FHLB) system for short-term funding needs.

The Fed’s “Factors Affecting Reserve Balances” report for the week ending 15 March 2023, otherwise known as the H.4.1 report, gave us our first glimpse into banks’ funding preferences in the immediate aftermath of the Silicon Valley Bank collapse. Here’s what we learned:

• The H.4.1 report[i] showed net borrowing of $307.6 billion split across three lending facilities:
  • The Fed’s discount window, its main direct lending facility, lent $152.9 billion, a notable number. However, filings show that First Republic Bank accounted for the vast majority of that number.[ii] It’s worth noting that the BTFP facility did not launch until the fourth day of the reporting period, so we’d expect some of this activity to shift to the BTFP.
  • “Other Credit Extensions” showed borrowing to the tune of $142.8 billion, representing a loan given to the FDIC for the takeover of SVB and Signature Bank.
  • The BTFP saw a paltry $11.9 billion. In our view, the limited uptake may be partly due to trepidation about using a facility that had not yet been vetted from an operational perspective. It may also indicate the traditional stigma associated with borrowing from the Fed remains.
• FHLB borrowing reached $264.4 billion for the week ending 15 March.[iii] This number represents the largest amount of borrowing among the individual lending facilities. In our view, this indicates the FHLB will likely remain the lender of first resort for banking institutions.

[i] https://www.federalreserve.gov/releases/h41/20230316/
[ii] Source: First Republic Bank 8-K Filing released on 15 March 2022.
[iii] FHLB data compiled by Loomis Sayles in conjunction with primary dealers for the week ending 15 March 2022. Number reflects both term and FRN issues (floating rate notes).