S&P Global Ratings thinks the extraordinary moves announced by the Federal Reserve in the past several days--most notably a return to quantitative easing (QE), a fed funds rates near zero, and CP funding and primary dealer credit facilities--will support financial stability, as well as banks' liquidity, bolstering their ability and willingness to meet client demands for funding. At the same time, we expect ultra-low interest rates to cause banks' net interest income and earnings to decline meaningfully this year even before considering potential pressures from higher loan-loss provisions and lower fee income.

Even before the Fed's actions, we viewed the liquidity of the U.S. banking system (in aggregate) as fairly robust thanks to stricter regulations enacted after the financial crisis. The Fed's plan to purchase at least $700 billion of securities will add to that liquidity by increasing the roughly $14.5 trillion of deposits held by U.S. banks by roughly about 5% (all else being equal). On top of that, the Fed has also launched a CP funding facility, a primary dealer credit facility, and encouraged banks to borrow from the discount window. The eight U.S. globally systemically important banks (GSIBs) responded by accessing the discount window to remove the stigma associated with it during times of strain. We think these moves, collectively, will lower the odds that the effects of the coronavirus will test banks' liquidity.

We view the benefits of the Fed's actions to bank creditors for liquidity and financial stability outweighing the pressure that low rates exert on bank earnings. Still, compared with 2019, we think bank earnings will fall by a double-digit percentage--and perhaps by nearly 25% depending on how quickly banks can reduce their funding costs--solely based on lower rates. Given the coronavirus's major shock to aggregate demand, a rise in loan loss provisions and a fall in fee income will only compound those effects. Banks with concentration in certain sectors (for example, energy and consumer discretionary) in particular are likely to experience an amplified hit to revenues and earnings.

U.S. banks in aggregate have entered this period with robust liquidity and capital, as well as low levels of nonperforming assets--we believe this will help them manage the stress. That strength has been an important factor in our Banking Industry Country Risk Assessment for the U.S., our methodology for assessing the economic and industry risks of a country's banking system. We currently have stable trends on both of those risk assessments.

Where Does Liquidity Currently Stand For The U.S. Banking System

Following the financial crisis, liquidity in the U.S. banking system improved sharply from precrisis levels partially due to three rounds of QE (when the Fed purchased securities). Since then, the system's overall liquidity has been in far better shape than it was in the run-up to the financial crisis. Large banks, which hold most of the assets and are responsible for providing the lion's share of liquidity to borrowers, appear to be in good shape.

Chart 1

The strength of large banks' liquidity stems in part from their need to meet the liquidity coverage ratio (LCR) requirement. The LCR aims to measure a bank's ability to meet a cash outflow assumed to occur during a 30-day stress period based on the nature of its funding and other cash obligations. High-quality liquid assets (HQLA)--essentially unrestricted cash and very liquid securities--make up the numerator of LCR. Net cash outflows--the cash outflows netted for cash inflows assumed to occur during the stress period--make up the denominator. For example, a bank with a 100% LCR would theoretically meet all net cash outflows during the 30-day stress period and wind up depleting all of its HQLA.

Before passage of regulatory reform legislation (The Economic Growth, Regulatory Relief, and Consumer Protection Act) in 2018, all banks under enhanced supervision (those with at least $50 billion in assets) were subject to an LCR. Banks with $50 billion-$250 billion had to meet a "modified" LCR by multiplying the net cash outflow (the denominator in the equation) by 70%. Shortly after the regulatory reform legislation, banks with $50 billion-$100 billion in assets were released from enhanced supervision and LCR.

Banks with assets of $100 billion-$250 billion remained under enhanced supervision. But in late 2019, they were released from the modified LCR requirement, as part of the Fed's "tailoring" rule. That rule, which adapts regulation on large banking organizations based on their risk profiles, also reduced the LCR requirement to 70%-85% (rather than 100%) for nonglobally systemically important banks with assets of $250 billion-$700 billion. (The Fed also considers other risk factors, but asset size is the most prominent driver for determining the LCR requirement.)

The end to LCR for banks with less than $250 billion in assets only occurred within months of the current pandemic, meaning it is unlikely that those banks significantly reduced their liquidity between late 2019 and the start of the crisis.

Chart 2

What The LCR Tells Us About Banks' Ability To Meet Draws On Credit Facilities

In their calculation of LCR, banks estimate what liquidity outflows they would experience as a result of commercial customers drawing on their credit and liquidity facilities. For instance, based on the LCR rule, banks assume commercial customers that are not financial institutions would draw 10% of the undrawn portion of their committed credit facilities.

Disclosure on the LCR calculation for the banks currently subject to that requirement (and those subjected to it through late 2019) suggest that it would take massive draws on their revolving facilities--probably significantly larger than those draws that occurred during the financial crisis for the industry as whole--to jeopardize the liquidity of most of those banks. The expected outflows on commitments in LCR equated to a median 18% of HQLA for banks subject to LCR, implying that outflows would need to be far larger than expected to drain HQLA to dangerous levels.

During the 2008-2009 financial crisis, data from the FDIC show that utilization rates on commercial facilities for the industry rose to the low- to mid-40%s from the mid-30%s. In aggregate, these draws totaled about 10% of the undrawn amount. In some cases, we believe borrowers could draw funds in a defensive move without immediately deploying the cash. That cash might sit in the bank that lent it, thereby leaving its HQLA unchanged.

Still, substantial draws on revolving facilities could significantly reduce or eliminate any buffer banks have over their minimum LCR requirements. If other liquidity outflows compound draws on revolving facilities, LCRs clearly would fall much more substantially.

Furthermore, draws on revolving facilities could be materially higher than in the financial crisis. The prevalence of closed or partially closed businesses for social distancing may cause many to seek liquidity wherever they can, particularly if the economic impact of the pandemic is felt over a long period.

Banks not subject to LCR are also exposed to this risk. They report commitments, but not what portion they expect to flow out in a stress scenario. Based on their commitment disclosure and using assumptions similar to those used in LCR, we believe rated banks would have outflows on commitments similar to the 18% median reported for LCR banks, though some banks are materially above that level. This indicates that most rated banks are well positioned to absorb significant draws on their revolvers. That assumes no other major liquidity stresses and that draws on revolvers are not multiples higher than in the financial crisis.

Chart 3

Some of the banks we rate have weaker liquidity ratios than peers, which factors holistically into our ratings. But some of these banks have offsetting factors that these ratios do not capture. For instance, some have good core deposit funding and are at lower risk of having asset quality problems. Those characteristics may make them less likely to experience a run on their liquidity, all else equal. In contrast, some of the banks that appear to have higher liquidity are more reliant on brokered or online deposits and take greater credit risk.

What Will Happen To Earnings?

Net interest income

We previously estimated that a cut in the fed funds rate back to 0.00%-0.25% could cause a 23% drop in bank earnings over time (see "COVID-19 And Falling Rates Cloud The Outlook For U.S. Financial Institutions," published March 10, 2020, on RatingsDirect). This assumed 3% earning asset growth, a 50% earning asset beta, and a 25% liability beta. (The betas measure how far asset yields and liability costs, respectively, would fall in proportion to the change in the Fed funds rate; for example, a 50% earning asset beta implies the average yield banks earn on earning assets would fall 25 bps for every 50-bps cut in the fed funds rate.)

While this could be a realistic estimate, we have updated that rough estimate for the Fed's most recent actions. We think the purchase of at least $700 billion of securities will ease funding competition to some degree and perhaps allow banks to reduce their funding costs more significantly. It also may result in some increase in asset growth even if this is in low-risk, low-yielding securities. Therefore, for our rough estimate, we have raised asset growth to 5% and the liability beta to 30%. This would result in a 16% drop in earnings--just on a net interest income decline--versus the earnings of all FDIC-insured banks in fourth-quarter 2019.

These results could be notably better or worse depending on those variables, however. For instance, a lower liability beta will result in greater earnings deterioration, as shown in our sensitivity table (see below); performance of individual banks also could vary substantially.

Table 1

Table 2

Table 3

Fee income

An abrupt slowdown in economic activity in the first half of 2020 also would constrain fee income, though to varying degrees for banks depending on their areas of concentration. The areas of noninterest income most likely be most pressured are shown in table 4.

Expenses

As their revenues comes under pressure, we expect banks to accelerate cost-cutting. Variable compensation will also naturally decline on lower revenue. Not all banks have the same ability to cut costs based on differences in fixed-cost structure (branches and technology and regulatory spending). Still, we expect expenses to decline for all banks in general, although efficiency ratios could rise.