Against the Wind

As the U.S. banking sector closed the books on 2015, some market observers' expectations for the coming year's results shifted plainly to the negative.  Duncan Marvin's Bloomberg column on April, 2016 was titled "There's Bad, Really Bad, and This Round of Bank Earnings" and the columnist opened by describing his expectations for first quarter 2016 results thusly, "Disappointing. Dismal. Dreadful." Another article published by an independent research firm was titled "US Banks: 1Q16 Preview, How Bad Will It Be?"  While these two examples are not the only downbeat outlook commentaries observed over the past several weeks, they are among the most pessimistic regarding the domestic banking system's prospects for the coming year. Frankly, we feel the titles lean toward the melodramatic and our opinion stands in the face of the prevailing wind.

The following chart distills the banking system's income statement into its four broad component parts, with total revenue represented by the purple line.  Operating expenses (the red bars) and loan loss provisions (green bars) are subtracted from revenue to arrive at pretax earnings.  The upward slope of the revenue line from 2014 to 2015 graphically illustrates that the system increased total revenue by nearly $17 billion.  Operating expenses were refreshingly lacking the one-off "headline noise" charges that have littered income statements since 2008, and the system was able to realize approximately $5 billion more in bottom line results through lower operating expenses in 2015 vs. 2014 as a result.  The combination of higher revenue and lower operating expenses was sufficient to offset a modest increase in loan loss provisions to drive a 6% increase in pretax net income to a record $235 billion.

FDIC Income Statement

Chart 1 - Graphic depiction of the domestic banking system's income statement. Source data: Federal Deposit Insurance Corporation (FDIC).

Why So Glum? 

Broadly, the common threads woven through market observers' negative expectations for the coming year are related to three headwinds: (1) the persistence of the low-growth, low interest rate environment, (2) capital markets volatility at the end of 2015 that carried over into 2016 and pounded related noninterest revenue, and (3) loan asset quality deterioration related to banking industry exposures to the energy sector. All three headwinds are real, although we don't expect them to materially constrain the banking system's results given its capacity to reduce operating expenses to offset slow revenue growth. Moreover, the system is better positioned to preemptively ameliorate exposures to the currently stressed energy sector than it was to the residential mortgage business during the last down credit cycle and it has plenty of capacity to build loan loss reserves against further loan losses should they emerge.

Fed Policy Giveth and Taketh Away: Net Interest Revenue

The first concern focuses on the pressure being applied to the system's net interest margin, and undeniably the trend appears concerning. A fundamental measure of profitability in the banking sector, the net interest margin is the difference between the interest revenue that a bank earns on assets (i.e. loans) and the interest expense that it must pay out in funding costs, expressed as a yield on average interest-earning assets. As illustrated in the following chart, by reducing the banking system's funding cost, the Federal Reserve's zero interest rate policy begun in late 2008 was initially a powerful stimulant in reversing the six-year slide in the measure. The positive impact was ephemeral, however, and after touching 3.76% in 2010, the system's net interest margin narrowed every subsequent year through 2015 when it was 3.07%.

FDIC Insured Banks Net Interest Margin

Chart 2 - Collective net interest margin for the domestic banking system. Source data: FDIC.

The reason for the decline is illustrated in the following chart, in which asset yields (blue line) and funding costs (grey line) are isolated. The average cost of funding across the system reacted positively and immediately to the Federal Reserve's policy and the grey line tracking funding costs turned sharply lower (positive) as banks lowered the rates being paid on deposits and took advantage of federally sponsored low-cost funding programs.  Funding costs eventually bottomed out between 0.30% and 0.35% across 2014 and 2015.  Asset yields meanwhile did not fall as rapidly and effectively plateaued in 2010 before commencing a slow decline culminating in a level at the end of 2015 that was actually 0.07% lower than the system's average funding cost at the end of 2007 when the five-year economic expansion ended (red line in Chart 3).

FDIC Yield on Earning Assets vs. Funding Cost

Chart 3 - System-wide yield on interest-bearing assets vs. cost of funding. Note the red line, which shows 2015 yield on assets is actually lower than the interest banks had to pay on funding costs in 2007. Source data: FDIC.

Margin compression notwithstanding, the three year negative growth trend in net interest revenue reversed in 2014 as the year-over-year loan growth rate of 5.28% approached the average rate across the system since 1993 (5.37%).  Net interest revenue increased 2.2% in 2015 ($9.4 billion) as outstanding loan balances increased 6.38% in 2015, with particular strength in the real estate-secured and commercial & industrial lines. Stated simply, the system was able to make up for weak lending margins with higher loan volume.

 FDIC Insured Banks, Change in Gross Loans Outstanding

Chart 4 - System-wide change in gross loans outstanding. Net interest margin is being pressured (see Chart 2), but growth in net revenue is being fueled by more loans on the balance sheet. Source data: FDIC.

Capital Markets Drag Is Limited

Results for the larger domestic banks with capital markets and / or asset management businesses are further expected to be pressured by capital markets volatility - and subsequently lower activity - that began to emerge at the tail end of 2015. Typically the strongest quarter of the year for security underwriting, the first quarter of 2016 featured sharply lower issuance in both debt and equity markets that was exacerbated by low trading volumes as investors were sidelined by tumbling values. Merger and acquisition advisory fees additionally are anticipated to be sharply lower as completed transactions plunged sequentially (63%) and year-over-year (76%).

Ominous as the figures above appear, trading account gains and fees together with investment banking fees are a relatively small revenue component for the domestic banking sector, as only a handful of the larger institutions operate with any scale in the capital markets.  In aggregate, trading and investment banking revenues have accounted for approximately 10% of noninterest revenue and less than 4% of total net revenue on average since 1992.  In 2015, their aggregate contribution to total net revenue was $33.3 billion, or 4.8% of annual net revenue, which is on par with $34.6 billion in revenue from service charges on deposits and $32.9 billion earned from trust and other fiduciary services. 

Not Another Storm

Potential asset quality deterioration represents the final top-line headwind confronting the banking system entering 2016 and while some concern is warranted, some perspective is also in order.  Since 2010, system earnings have been supported by loan loss reserve releases as the economy emerged from the 2009 recession and loan asset quality improved. As illustrated in the following chart, when loan loss provisions (dollars set aside for delinquent loans, represented by the blue line) exceed loan losses (a.k.a. net charge-offs, represented by the grey line), the banking system builds loan loss reserves.  The opposite relationship holds true when loss provisions are lower than recognized losses. The greater the divergence between the two lines, the larger the loss reserve build (expense) or release (expense reduction).

FDIC Insured Banks Loss Provisions vs. Net Charge-offs

Chart 5 - The domestic banking system's loan loss provisions vs. net charge-offs. The inflection point in 2015 may indicate the system moving in to a reserve building phase to offset loans in the energy sector, but we feel the magnitude will be more similar to that seen in 2000.  Source data: FDIC.

How this impacts earnings is illustrated by the green segments in Chart 1. The better the asset quality experience, the less loan loss provisioning is required (and the smaller the green bars) to build reserves against expected future losses.  The loss provision bar hit its peak in 2009 (a $249.5 billion expense) and it shrunk in every subsequent year - until 2015.  As illustrated in Chart 5 above, loan loss provisioning turned visibly higher in 2015 and what had been a windfall for sector earnings trends since 2010 was over. Over that period, approximately $122 billion (or 11% of pretax earnings) fell to the bottom line in the form of reserve releases.

Going into 2016, energy lending exposure has become a top-of-mind concern after the sharp decline in oil prices that began in 2014 looks to persist over the next 18 months (see our oil analyst's views here). Loan losses are a fact of life for the banking sector however, and associated provisioning represents an unavoidable cost of doing business.  Although the industry has been reasonably aggressive in reserving for those risks, there is no obvious offsetting lever that can be pulled (i.e. increased lending volumes or increasing fee revenue) to diffuse the negative impact of increased provisioning and the earnings headwind will persist until the down-cycle currently pressuring the oil and gas sectors runs its course. 

Fortunately, emerging asset quality deterioration is broadly limited to the energy sector and to those geographies where oil and natural gas extraction is concentrated, in contrast with the last down credit cycle that was driven by the unprecedented national collapse of the residential real estate sector.

In addition, commercial and industrial loans (of which the energy sector is but a single component) accounted for roughly 21% of total outstanding gross balances at the end of 2015 compared with the 36% concentration in first- and second-lien mortgages at the end of 2007. Consequently, individual exposures to the energy sector should be materially lower than those that existed to residential real estate prior to the last down credit cycle and the financial sector should not face something like the unadulterated catastrophe of 2008 - 2009. 

Moreover, commercial loans generally exhibit sufficient visibility of deterioration as it develops to allow banks to take appropriate mitigating actions.  For example, among our coverage universe the highest concentration of energy-related loans is roughly 6% of that institution's total gross outstanding balances. The bank reported a 4% impairment rate at the beginning of 2Q15 against which it was holding loss reserves equivalent to 114% of nonperforming balances.  The first cohort of U.S. banks reporting results for the first quarter of 2016 are signaling only incremental loss provision growth for the current year, rather than something like the sharp spike experienced between 2007 and 2010 due to the mortgage crisis.

In addition to being able to build loss reserves in anticipation of worsening asset quality, commercial credit facilities afford banks flexibility to amend terms at regular intervals as the economic winds shift.  Terms for energy sector loans specifically are reset twice a year, with one "redetermination" process occurring in the spring and the other in the fall. In addition to demanding increased collateral levels, banks have begun to incorporate "anti-cash hoarding" covenants during the current redetermination process that require borrowers to repay existing loan balances with any surplus cash they may have and - in the more extreme circumstances - they have the discretion to cut credit lines outright. 

Understanding the Operating Expense Lever

The other side of the income statement is expense control and pulling it makes the system more efficient. One option banks have available to ameliorate the low revenue growth and rising loss provision headwind is tighter control of operating expenses. The system's operating overhead ratio (noninterest expenses / total net revenue) has been stubbornly wrapped around 60% since 1992, with the weakest level (68.2%) reported in 2012 as legal expenses peaked, while the strongest ratio was struck in 2002 (55.9%).  The overhead ratio reported by the system in 2015 was 60.6%, which in plain English means that the system as a whole spent $0.61 for every $1.00 of revenue that was earned. Wringing operating leverage from their respective franchises to improve the ratio emerged as a near universal strategic priority across our coverage universe in 2015, as management teams grappled with maintaining returns against a backdrop of higher required levels of equity capital and low-return liquid asset requirements imposed by regulators.    

The single largest noninterest expense line item is employee compensation, which accounted for roughly 43% of total expenses on average since 1992 and 26% of total net revenue over the same time frame.  Although employee compensation has increased in every year since 1992, except 2008, it remains one of the more variable industry expenses and one over which management teams have the most discretionary flexibility. Despite required additions to risk management and oversight positions, headcount is down roughly 9% from 2007, helping to slow expense growth - and we believe there will be additional reductions.

FDIC Insured Banks Operating Expenses

Chart 6 - The domestic banking's system operating expenses. Personnel represents salary expense, Occupancy equals facilities rental and Other represents all other expenses (unrelated to payroll or facilities rental).

Concurrent with taming growth rates in variable expenses and - in conjunction with optimizing to account for electronic banking and other technological change -banks have begun to turn increasing attention to reducing their fixed cost bases. In addition to re-aligning and / or exiting businesses outright to optimize their use of available balance sheet capital, for the first time since the mid-1930's the system has shrunk its physical footprint over a sustained period.  Since peaking in 2012, over 1,050 branches were closed through the end of 2014 (the last available annual data), plus an additional 289 branches in the first quarter of 2016 alone. In addition, we expect continued industry consolidation to provide opportunity to squeeze excess capacity and cost out of the system, continuing the two-decade decline in bank charters which includes the surrender of over 2,350 charters since the end of 2008.

Technological innovation has allowed for the reduction in physical plant, while still providing adequate services to banking customers. Bank of America is a case in point. The bank reported in its 1Q16 earnings release that it has 19.6 million customers who conduct the majority of their business with the bank on mobile devices, vs. 17.1 million at the end of the same period last year. The bank further reported that 16% of its total deposit account transactions in the most recent reporting period were done on mobile devices.  Growth in consumption of banking products and services through non-traditional channels allowed Bank of America to close 146 branches over the trailing 12 months and reduce its total employee headcount by 6,475 (3%), even while bringing 100 more automatic teller machines on-line. We expect these types of efficiency adjustments to be made across the system in coming years.

FDIC Insured Commercial Bank Branches

Chart 7 - Open bank branches, system-wide. The decrease in number of bank branches beginning in 2009 represents a sea change and represents one way the banking sector is working to control expenses. Source data: FDIC.

Shelter Against the Wind

Although we acknowledge that the concerns described above represent real headwinds against which the banking system will strive in coming reporting periods, we do not view them as insurmountable individually nor in aggregate.  Rising impairments in loan portfolios reflect the natural ebb and flow in the credit cycle, with the sector well positioned to manage through exposures to the energy sector.  Improving demand for loans has offset historically narrow lending margins and pressure from capital markets volatility and lower client activity impacts a relatively thin cohort of institutions across our coverage universe.  Finally, the sector has numerous levers available to lean on to improve operating leverage and wring cost efficiencies from the existing base, including traditional methods of shrinking the physical branch footprint as well as through the application of emerging and evolving technologies, both of which allow for reduction in the single-largest operating expense line.  Ultimately we believe the "doom and gloom" is a bit excessive.  We view the domestic banking sector as healthy and are accordingly comfortable with our current sector positioning.  

Sources:

Marvin, Duncan, "There's Bad, Really Bad, and This Round of Bank Earnings," Bloomberg, April 11, 2016.

CreditSights, "US Banks: 1Q16 Preview, How Bad Will It Be?" April 11, 2016.

Schoen, John W., "7 Years on From Crisis, $150 Billion in Bank Fines and Penalties," CNBC, April 30, 2016.

Federal Deposit Insurance Corporation, "Statistics on Banking," quarterly publication.

Bloomberg

Bank of America, "1Q16 Financial Results," April 14, 2016.


 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Funding cost is interest a bank pays on deposits, coupled with the interest it pays to bondholders on debt instruments it issues.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Loan loss provisions are capital a bank sets aside in anticipation of loans becoming stressed and not paying. While setting provisions aside against potential bad loans is a good thing, it also means that the money cannot be put to work for other corporate initiatives. On the income statement, loan loss provisions are considered an expense. Further, lower loan loss provisions are symptomatic of a healthy loan portfolio.