In Focus This Quarter

Introduction

The banking industry continued to perform strongly through fourth quarter 2000, despite a noticeable slowdown in earnings growth resulting from narrowing net interest margins (NIMs), increased loan loss provisions, and declining noninterest income. Industry performance during first quarter 2001 appears unlikely to improve in light of the recent slowdown in the U.S. economy. Furthermore, certain adverse trends in portfolio characteristics, including risk selection, loan concentrations, and funding and liquidity concerns, could pose risk management problems for certain groups of insured institutions. Overall, however, the financial strength of the industry positions it to continue to meet loan demand from creditworthy borrowers during a slowing economy. Using consensus economic scenarios, this article examines the potential effects of the current economic slowdown on banking industry earnings and capitalization.

The Banking Industry Remains Strong

Call report data continue to reflect the financial strength of the banking industry. According to a variety of financial measures, the industry is currently much stronger and better prepared for an economic slowdown than in 1990, when the last recession began. At that time, the industry was weakened and recovering from the fallout of excessive risk taking and problems relating to the agricultural sector, regional recessions, and sharply deteriorating conditions in local residential and commercial real estate markets.¹ Year-end 2000 data for capital, earnings, nonperforming loans, and exam ratings suggest that the financial position of the industry is historically strong, and the industry is better able to withstand similar adverse economic conditions.

¹ For further information regarding the state of the banking industry in the late 1980s and early 1990s, see Federal Deposit Insurance Corporation, History of the Eighties-Lessons for the Future. Vol. I: An Examination of the Banking Crises of the 1980s and Early 1990s, Washington, DC: FDIC, 1997. www.fdic.gov/bank/historical/history/contents.html.

Insured institutions with weak capital ratios have declined both in number and as a percentage of all insured institutions (see Chart 1). In the mid-1980s, about 17 percent of the banking industry, or 3,000 institutions, reported equity and reserves less than 6 percent of assets. At that time, nearly 2.5 percent of all institutions were insolvent (reporting equity and reserves less than zero percent of assets). As the economy entered a recession in the early 1990s, capital positions had improved somewhat for most of the industry. By the end of 2000, however, the percentage of thinly capitalized insured institutions had fallen dramatically: only about 0.5 percent of the industry, or 45 institutions, held equity and reserves less than 6 percent of assets.

Chart 1

[D]

Higher levels of profitability contributed significantly to the improvement in capital levels at insured institutions in the late 1980s and early 1990s. During 1986 and 1987, more than 20 percent of the banking industry, or roughly 3,500 institutions, was unprofitable, operating with return on assets of less than zero (see Chart 2). By the mid-1990s, the industry had reorganized, the credit quality of loans had improved, cost-cutting measures had led to improved efficiency, and fewer than 5 percent of all insured institutions were unprofitable. The number of unprofitable institutions has risen to more than 700 (7.3 percent of insured institutions) as of year-end 2000. More than half of those unprofitable institutions are new banks that have been chartered since 1997. While new banks are often unprofitable during their start-up period, the onset of adverse economic conditions will pose additional challenges for these institutions.

Chart 2

[D]

The percentage of institutions with high problem loan levels and the percentage of institutions with a composite CAMELS rating of 3, 4, or 5 show a similar trend.² As the real estate crisis and regional recessions hit the banking industry in the mid- to late 1980s, the percentage of institutions with noncurrent loans greater than 6 percent of total loans rose past 10 percent. As the national and regional economies began to improve in the early 1990s, the performance of loan portfolios also improved. Today only a handful of institutions have high levels of nonperforming loans. Similarly, insured institutions rated 3 or lower rose to more than 25 percent at the beginning of the past decade, fell to less than 5 percent by 1997, and rose to 6.4 percent by the end of 2000.

² The acronym CAMELS refers to the supervisory ratings system applied to insured depository institutions by federal and state regulators. Its component letters stand for: capital, asset quality, management, earnings, liquidity, and sensitivity to market risk.

Although the banking industry's financial performance remains solid overall, some signs of deterioration are emerging. Aggregate commercial bank and thrift earnings declined modestly in 2000, largely as a result of significantly higher loan loss provisions at a few large institutions.³ Commercial banks also continued to experience downward pressure on NIMs; the industry NIM fell to 3.90 percent during fourth quarter 2000, the lowest level since 1990. Earnings of insured savings institutions also fell slightly in 2000 in large part because of declining NIMs. However, compared to ten years ago, when many banks were saddled with impaired capital and negative earnings, banks and thrifts are currently well positioned to meet loan demand from creditworthy borrowers and help support economic recovery.

³ Federal Deposit Insurance Corporation, Division of Research and Statistics, Quarterly Banking Profile, fourth quarter 2000, www.fdic.gov/bank/analytical/index.html.

Risk Management Issues Raise Concerns

A number of risk management concerns related to changes in portfolio characteristics should be weighed against this overall financial strength. These concerns include loan concentrations, risk selection, funding and interest rate risk, off-balance-sheet exposures, and operational risks.

Loan Concentrations

Of particular importance at this stage of the business cycle are the rising loan concentrations in traditionally higher risk loan categories. Chart 3 shows the percentage of insured institutions with concentrations of commercial and industrial (C&I) loans, commercial real estate (CRE) loans, and construction and development (C&D) loans expressed as a percentage of equity capital. During the mid-1980s, more than 20 percent of all insured institutions held total concentrations in these "commercial and construction" loans greater than 400 percent of equity capital. By the early 1990s, that percentage had fallen by more than half. Since 1995 the percentage of institutions with concentrations in these loan types has begun to rise sharply and again reached 20 percent by the end of 2000.

Chart 3

[D]

Loan concentrations such as these traditionally have been associated with a higher frequency of failure. Chart 4 shows the percentage of insured institutions that failed within three years, according to concentrations of commercial and construction loans-to-equity capital. Institutions with elevated concentrations in these loan categories have consistently failed in higher proportions than other insured institutions. From 1984 to 1998, institutions with commercial and construction loans in excess of 400 percent of capital were two to three times more likely to fail within three years.

Chart 4

[D]

Similar concerns are noted for institutions with concentrations specifically in CRE or C&D loans. Some 18.6 percent of insured institutions showed CRE loan concentrations greater than twice their capital in December 2000, compared with 12.5 percent in 1989. For C&D loans, slightly more than 10 percent of insured institutions show concentrations greater than twice their capital-about the same percentage as in 1989. Again, history shows that the greater the level of concentration in these loan types, the greater the risk of failure within three years. However, the number of institutions with the very highest concentrations in CRE and C&D loans is lower than it was in the late 1980s.

Risk Selection

Some insured institutions are also holding a higher-risk asset mix, including subprime consumer loans, high loan-to-value (LTV) mortgages, and syndicated loans to highly leveraged companies, than they did a decade ago. Subprime lending refers to loans extended to borrowers with weak credit histories or questionable repayment capacity. In 2000, 13 percent of mortgage originations were subprime mortgage loans, up from 5 percent in 1994. Insured institutions increasingly have become involved in subprime lending as this market has expanded. Nearly half of the top 25 subprime mortgage lenders are bank or thrift holding companies or their affiliates.⁴ Even the best grade of subprime mortgages (A-) is nearly four times as likely to be seriously delinquent as prime mortgages. Given the high delinquency rate on subprime loans even in a strong economy, insured institutions that hold a significant portion of these loans in their portfolios may need to prepare for greater credit quality challenges should the economy slow further.⁵

⁴ Inside B&C Lending, Bethesda, MD: Inside Mortgage Finance Publications, Inc., March 5, 2001, Volume 6, Issue 5, p. 2.

⁵ According to Mortgage Information Corporation data from September 2000, 0.5 percent of all prime mortgages were seriously delinquent (at least 90 days past due or in foreclosure), while 3.4 percent of all A- subprime loans were seriously delinquent.

Insured institutions have participated in an expanding market for high-LTV mortgage lending. More than 20 percent of all mortgages originated during 2000 were for more than 90 percent of the value of the house, almost three times the rate in 1990.⁶ Like subprime lending, high-LTV lending represents a new line of business that remains untested in a slower-growing economy. Should the economy enter a prolonged downturn, mortgage losses could be higher than in past recessions.⁷

⁶ Federal Housing Finance Board data.

⁷ For additional information on high-LTV lending, see Kathy R. Kalser and Deborah L. Novak, "Subprime Lending: A Time for Caution," Regional Outlook, third quarter 1997; Diane Ellis, "High Loan-to-Value Lending: A New Frontier in Home Equity Lending," Regional Outlook, first quarter 1999; and Alan Deaton, "Rising Home Values and New Lending Programs Are Reshaping the Outlook for Residential Real Estate," Regional Outlook, third quarter 2000.

Another sign of rising credit risk on bank balance sheets is an increase in the proportion of syndicated loan originations extended to leveraged borrowers.⁸ Originations of leveraged loans as a percentage of total syndicated loans rose from 17 percent in 1996 to 31 percent in 1999.⁹ As the level of problem loans rose and economic conditions weakened in 2000, banks scaled back lending to leveraged borrowers to 26 percent of total syndicated loans. However, total syndicated loan originations rose to a historic high of nearly $1.2 trillion in 2000.

⁸ Ronald L. Spieker, "An Overview of Trends in Syndicated and Leveraged Lending," Bank Trends: Analysis of Emerging Risks in Banking, Number 98-04, February 1998; and Ronald L. Spieker and Steve E. Cunningham, "Recent Trends in Syndicated Lending," Regional Outlook, first quarter 1999.

⁹ Loan Pricing Corporation.

Funding and Interest Rate Risk

A scarcity of core deposits, particularly at community banks, during a period of sustained loan growth has resulted in an increased reliance on noncore sources of funds. This trend has contributed to an ongoing decline in NIMs and, in some cases, to increasing sensitivity to rising interest rates.¹⁰ The recent decline in short-term interest rates could alleviate pressure on NIMs for many institutions. However, a general decline in interest rates also could spur significant growth in mortgage refinancing activity, which could lead to greater interest-rate sensitivity for institutions that invest heavily in fixed-rate assets, such as mortgage-backed securities (MBS) and long-term fixed-rate mortgages. The Mortgage Bankers Association Refinancing Index has risen sharply since fourth quarter 2000, when interest rates began to fall.¹¹ Call report data indicate that the interest-rate sensitivity of insured institutions may be rising. The number of banks with long-term assets greater than 40 percent of assets and noncore liabilities greater than 20 percent of assets has more than doubled since year-end 1997. At the same time, capital market developments-such as the emergence of deep, liquid markets for interest-rate swaps and floating-rate MBS-have expanded the range of tools available to bankers to manage interest-rate risk.

¹⁰ For additional information on the effects of increasing levels of noncore funding on interest rate risk, see Allen Puwalski, "Increasing Interest Rate Risk at Community Banks and Thrifts," Bank Trends: Analysis of Emerging Risks in Banking, Number 00-01, March 2000; and Allen Puwalski and Brian Kenner, "Shifting Funding Trends Pose Challenges for Community Banks," Regional Outlook, third quarter 1999.

¹¹ The Refinancing Index rose from 761.5 at year-end 2000 to 2287.0 as of mid-April 2001. Mortgage Bankers Association of America/Haver Analytics.

Off-Balance-Sheet Exposures

Innovations in risk management practices and the desire for enhanced fee income have led to a significant increase in derivatives activity over the past decade. Since 1990, the notional volume of derivatives contracts held by commercial banks has risen from $6 trillion to $40 trillion.¹² The growth in volume highlights the importance of maintaining risk management practices that are appropriate to this complex business activity. Risk-based capital requirements also have created incentives for lenders to remove risk from their balance sheets through securitization activities and credit derivatives. Significant growth in these off-balance-sheet risks poses a variety of issues, such as accurately valuing derivatives exposures and residual assets.

¹² Bank call report data.

Since 1992, unused commercial loan commitments and letters of credit held by banks have increased rapidly- from $776 billion to $1.889 trillion.¹³ Unused commercial loan commitments present banks with adverse selection risks when financially stressed companies draw on their credit lines. For example, in recent months several large corporations drew on commercial paper backup credit lines following downgrades in their commercial paper ratings. Also, many households that refinanced mortgages as a means of consolidating other debts between 1998 and 1999 now may have more unused credit available on credit cards and other revolving lines that these households could draw upon immediately if they experience financial distress.

¹³ Bank call report items used to calculate unused commercial loan commitments include commercial and similar letters of credit, performance standby letters of credit and foreign office guarantees, and "other unused commitments," which include commitments to extend credit through overdraft facilities or commercial lines of credit and retail check credit and related plans.

Operational Risks

Well-run financial institutions use a variety of risk management and audit-related controls to safeguard against excessive risk taking, fraud, noncompliance with regulatory directives, and errors in financial reporting. The need for such systems and controls only increases as an institution expands its scope of operations or enters new lines of business. However, operational controls do not directly generate revenues, and the temptation to cut expenses in this area may be great, particularly when an institution is faced with prospects for reduced profitability. The role that fraud has played in some of the more recent bank failures highlights the importance of adequate internal control and audit procedures. To a great extent, operational risk, as well as off-balance-sheet exposures, can be evaluated only by on-site supervisory analysis of risk management practices and modeling techniques.

Current Economic Conditions: Are We Still in a "New Economy"?

The economic environment for the banking industry is now marked by considerable uncertainty.

As recently as nine months ago, virtually all the vital signs for the U.S. economy were positive. In February 2000, the economy entered its 108th month of expansion, making this the longest expansion in U.S. history. Sustained high rates of productivity growth were making it possible for the economy to grow rapidly (above 4.0 percent annually for four consecutive years) with low to moderate inflation. Equity market valuations reflected the optimism of investors and analysts, who began to conclude that much had indeed changed in the so-called "New Economy."¹⁴

¹⁴ See Maureen E. Sweeney et al., "Banking Risk in the New Economy," Regional Outlook, second quarter 2000.

However, over the past year, three factors have significantly slowed the pace of economic activity. First, the cost of borrowing by households and businesses rose in 2000, in part a result of a cumulative 175 basis point increase in the Federal Reserve federal funds target rate between June 30, 1999, and May 16, 2000. Second, the price of oil in 2000 averaged more than twice that of two years earlier, increasing costs for consumers and businesses. As winter approached, the energy situation worsened as spot prices for natural gas soared and electricity production fell short of demand in California and other western states. Finally, high rates of capital investment in certain industry sectors led to overcapacity that was pressuring some companies' ability to service debt. Examples of these companies could be found in the entertainment, telecom, and dot-com industries, among others.¹⁵

¹⁵ See Steven Burton, "Credit Problems for U.S. Businesses Continue to Rise," Regional Outlook, first quarter 2001.

By fall 2000, signs of deterioration in economic activity were clearly visible. Equity analysts began cutting estimates of corporate earnings growth for the first half of 2001. Equity valuations continued to retreat from the high levels of early 2000, led by rapidly falling stock prices in the high-tech sector. Perhaps more troubling, credit spreads on risky corporate debt rose to higher levels than had been experienced during the financial crisis of late 1998.¹⁶ Capital market conditions became particularly impaired in December 2000 as issuance of high-yield corporate debt virtually stopped.¹⁷

¹⁶ Burton, p. 14.

¹⁷ Board of Governors of the Federal Reserve System, "Monetary Policy Report to Congress," February 13, 2001, p. 10.

Signs of the economic slowdown were also visible on Main Street. Corporate layoff announcements in December 2000 tripled from year-ago levels to 133,000, and remained above the 100,000 mark through April 2001 (see Chart 5). The Conference Board's Index of Consumer Confidence fell for five consecutive months beginning in September 2000. Data released in early 2001 showed that capital investment by U.S. companies declined in the fourth quarter of 2000 for the first time since the 1990 to 1991 recession, while growth in consumer spending was only half what it had been the previous year. Weakening in these "twin pillars" of U.S. economic activity slowed the growth rate of the economy to a low 1.1 percent in fourth quarter 2000 before rebounding slightly to 1.3 percent in first quarter 2001.

Chart 5

[D]

Despite the negative economic news, it is by no means clear that the economy will slip into recession. Monetary and fiscal policies appear to be aimed at stimulating economic activity. Between January and mid-May 2001, the Federal Reserve reduced the federal funds target rate five times by a total of 250 basis points. Current budget proposals under consideration by Congress call for tax cuts in excess of $1 trillion over the next ten years. Moreover, despite earnings difficulties in the corporate sector, the consumer sector appears to be in better financial shape-at least for the present. Unemployment continues to be low (4.5 percent as of April 2001), while declining interest rates have contributed to a strong housing market and a renewed wave of mortgage refinancing.

The trend in productivity growth is also unclear. Output per hour at nonfarm businesses rose by a strong 4.3 percent in 2000 and at an annualized rate of 2.0 percent in fourth quarter 2000. But productivity essentially stood still in first quarter 2001, raising the question of whether the high levels of productivity growth that have been associated with the New Economy will persist. Strong productivity gains in recent years can be attributed, at least in part, to investment in new technologies and efficiencies generated in corporate restructuring. It would follow that maintaining productivity growth at or near these levels in the future will require a recovery in corporate investment in new technologies and continued access to sufficient financial capital to carry out corporate mergers and acquisitions. Therefore, it appears that a prompt recovery in financial market performance and economic activity will be essential to maintain the productivity growth that has been a hallmark of the New Economy.

Three Factors Will Shape the Economic Outlook

In this time of economic uncertainty, three factors will largely determine whether the United States will fall into a recession and how long such a slump could last.

Consumer and Business Confidence

The degree of optimism on the part of consumers and business managers is an intangible factor that is difficult to predict-or even explain after the fact-but is nonetheless crucial to the economic outlook. Although the index of consumer confidence declined from a cyclical high of 143 in September 2000 to a low of 109 in April 2001, consumer spending has continued to increase.¹⁸ Retail sales increased 2.1 percent in the first quarter of 2001 from year-ago levels. Concern exists, however, that high levels of consumer indebtedness and a negative household savings rate could result in consumers reducing their spending significantly if the economy enters a recession. Part of this concern centers on the so-called wealth effect, whereby consumers are thought to increase their spending in proportion to stock market gains when equity prices are booming. The corollary is that consumers may reduce their expenditures and increase their savings if equity market losses persist.

¹⁸ The Conference Board, www.conference-board.org.

The confidence of business managers is also tied to events in the capital markets, which to a large extent determine the cost of capital. Willingness to invest in new equipment is tied not only to the cost of capital but also to expectations for future earnings. Investment spending fell in the fourth quarter of 2000 as profit expectations plummeted (see "Corporate Profits," below). Business investment likely will begin growing again at the double-digit rates of recent years only after managers see a sufficient decline in the cost of capital and a corresponding increase in profits to justify additional spending on equipment and software.

Availability of Capital

One of the keys to the New Economy has been an increasing reliance on market-based sources of financing. In 1997, for the first time, corporate sector borrowing from market-based sources surpassed its borrowing from traditional financial institutions.¹⁹ Continual access to the capital markets has been an integral part of many New Economy business plans. Equity initial public offerings (IPOs) on the New York Stock Exchange and the NASDAQ totaled 445 in 1999 and 408 in 2000.²⁰ However, only 17 new companies issued shares on the two major exchanges during first quarter 2001. A resumption of capital market funding from these depressed levels appears necessary for a continuation of the New Economy and its high levels of new business formation and productivity growth.

¹⁹ See Sweeney et al.

²⁰ See "Gap Between NYSE and NASDAQ IPOs Gets Smaller," The Investment Dealers' Digest, April 9, 2001, and "A Quarter to Forget for Initial Offerings," EBN, April 9, 2001.

Banks and other traditional intermediaries have proved to be more reliable sources of business finance than the capital markets in recent years. Holdings of C&I and CRE loans by FDIC-insured institutions rose 9.4 percent in 2000, following increases of 11.5 percent in 1998 and 9.8 percent in 1999. Underwriting surveys conducted by the federal banking agencies show that banks recently have tightened pricing and terms on commercial loans as evidence of corporate weakness and slowing economic activity has emerged. However, there is little evidence to suggest a significant reduction in the availability of bank credit for businesses, particularly in comparison with the recent volatility in the capital markets. Indeed, compared with the condition of the banking industry heading into the last recession, today's banking industry is composed of far fewer institutions with impaired capital positions and troubled loan portfolios-two factors that can interfere with the ability of insured institutions to make loans to creditworthy borrowers. On the whole, the current strong financial condition of the banking industry, and its ability to make loans to creditworthy borrowers, appears to be one of the brighter spots in the near-term economic outlook.

Corporate Profits

Prospects for a rapid and sustained economic recovery likely depend on the health of the corporate sector. One of the most important and most tangible determinants of the near-term economic outlook will be the corporate profit equation.

Recently, the news has been mostly negative. Equity analysts repeatedly have cut their 2001 estimates for earnings growth by S&P 500 companies, to the point that earnings for the year may not grow at all. Companies in a wide range of sectors have been hit by higher energy, labor, and debt service costs at a time when a slowing economy has reduced sales growth.

Some of the stress in the corporate sector is evidenced by rising corporate bond defaults and increasing losses on C&I loan portfolios. For example, net charge-offs of C&I loans by FDIC-insured institutions rose to more than $8 billion in 2000 from less than $5.5 billion in 1999. On the basis of the slowdown in economic activity in late 2000, commercial loan performance could deteriorate further before it improves.

Economic Scenarios and Bank Performance

Regardless of whether the U.S. economy goes into recession, a transition from the rapid growth of the 1990s to a period of slower growth will pose challenges to insured institutions. Some insight into insured institution performance in a slower-growing economy can be gained from an analysis of historical relationships in past business cycles.²¹ Our analysis emphasizes the relationships of the major components of banking industry profitability to aggregate economic factors, acknowledging that these relationships can differ for individual banks. Although history need not repeat itself, the relationships discussed below have been evident in many U.S. business cycles since 1960.

²¹ Economic factors such as the growth of aggregate demand, gross domestic product (GDP) growth, equity market performance, and interest rate levels and term structures are expected to be related to banking industry performance.

In recognition of the current uncertainty in the economic outlook, a consensus among business and financial analysts has developed since late 2000 outlining three possible directions an economic slowdown might take. These scenarios have been labeled in the business press as the V, U, and L scenarios because of the shape each suggests for the depth and duration of the slowdown and the rate of recovery in overall economic activity.²²

²² Another way to view the degree of uncertainty is to recall that, as late as November 2000, most analysts were discussing the timing of the "soft landing," interpreted as GDP growth slowing to a sustainable rate of 3 percent. By mid- to late-December 2000, analysts had shifted to the V, U, or L scenario characterization of the slowdown, with the V-shape being most prominent. By late March 2001, the U-shaped scenario had gained more support as the accepted way to view the current slowdown in economic activity.

The V, U, and L scenarios are described in the inset box. Chart 6 characterizes, from a historical perspective, the V and U scenarios that have characterized U.S. business cycle downturns since 1960. The chart shows that gross domestic product (GDP) growth typically accelerates four quarters before the beginning of a recession (period 0) and reaches a negative value (a decline in GDP) about three quarters after the beginning of the recession. The U scenario characterizes the 1990 to 1991 recession and differs from the typical post-1960 recessions in that GDP growth did not accelerate just before the recession and recovered at a much slower pace after the trough of the recession. For example, GDP growth did not reach 3 percent (near the long-run potential growth rate of 3.3 percent) until seven quarters after the start of the recession.

Chart 6

[D]

Net Interest Margin and the Term Structure of Interest Rates

A steeper yield curve generally results in improvements in banks' NIMs. A consistent, positive relationship between banks' NIMs and the yield spread between short-term and long-term interest rates has been well established for some time.²³ As shown in Chart 7, an increase in the yield spread is associated with an increase in the NIM after one or two quarters. Since the early 1990s, moreover, the lag in this relationship has shortened.

²³ Gerald A. Hanweck and Thomas E. Kilcollin, "Bank Profitability and Interest Rate Risk," Journal of Economics and Business, Spring 1982, Volume 36, Number 1, p. 77.

Chart 7

[D]

Over the past 25 years, an inversion of the yield curve (a negative spread) has led most recessions by several quarters, although inverted yield curves have not always been followed by a recession.²⁴ An inverted yield curve suggests that banks are paying higher rates to borrow short term and earning lower rates of return on long-term investments. The yield curve became inverted in the final two quarters of 2000, and, as anticipated, net interest income declined (see Chart 7). The Fed's aggressive easing of the federal funds rate during the first quarter of 2001 has already reversed the negative spread in the yield curve that was in place at the beginning of the year.

²⁴ Christopher J. Neely, "What Is the Slope of the Yield Curve Telling Us?" Monetary Trends, Federal Reserve Bank of St. Louis, MO, August 2000, p. 1.

Chart 8 indicates a tendency for the yield curve to become steeper as economic downturns run their course. Other things constant, this steepening would serve to buffer the NIM, offsetting the effects of declining credit quality as the recession progresses. Under a V scenario, the yield curve would be expected to invert before the recession and become positive within 12 months (see Chart 8).

Chart 8

[D]

Noninterest Income and Stock Prices

Noninterest income has become an increasingly important source of revenue for banks and thrifts. As a proportion of earning assets, this source of income has grown steadily since the mid-1980s, and its growth has accelerated since 1992.²⁵ Furthermore, since 1992 this income source has become more market sensitive, particularly for large banks, and has shown a positive relationship with the rate of change in the S&P 500 stock index (see Chart 9). Changes in this index (year-over-year) have preceded changes in the noninterest income ratio by one or two quarters, and since 1992, this time lag has shortened. For example, the stock market declines of the latter half of 2000 were accompanied by a decline in the ratio of noninterest income to earning assets reported over the same period.

²⁵ Bank call report data.

Chart 9

[D]

The rate of change of the S&P 500 index has shown a consistent relationship with the business cycle over the past 40 years (see Chart 10). Declines in the S&P 500 returns have preceded recessions by 12 months and have recovered to positive returns within at least 12 months after the start of the recession. Through March 27, 2001, the S&P 500 index had dropped 20 percent from its high in September 2000; this decline is considerably greater than during past recessions (see Chart 10). Regardless of whether the economy enters recession, if the correlation between stock prices and noninterest income is maintained, as suggested in Chart 9, the effect could be a decline in noninterest income, with the duration and magnitude depending on the length of the economic slowdown.

Chart 10

[D]

Noninterest Expense

Wages, salaries, and other operating expenses drive changes in bank noninterest expenses. The ratio of these expenses to earning assets has remained practically constant since 1992. Historically, the ratio of noninterest expense to earning assets shows no discernible cyclical relationship, possibly because of offsetting factors making up noninterest expense. For example, wages and salaries for some banking functions tend to slow during a recession, whereas expenses for loan workouts and related costs rise.

Loan Losses

Over a number of business cycles, bank loans have tended to be procyclical, with total loan growth falling for three quarters before a recession, continuing to decline through the recession, and beginning to rise four to five quarters after the beginning of the recession (see Chart 11). During the 1990 to 1991 recession, which exhibited characteristics of a U-shaped scenario, loan growth was negative for five quarters and returned to increasing, positive growth some eight quarters after the recession had begun. These patterns are similar for growth in mortgage, CRE, consumer, and C&I loans.

Chart 11

[D]

This cyclical pattern suggests that the rapid loan growth during the upswing of the business cycle may be associated with loans whose credit quality will become weaker in the event of an economic slowdown. Recent loans, however, typically will not be the ones that initially become nonperforming. Consequently, growth in nonperforming loans, loan loss provisions, and loan charge-offs would be expected to lag total loan growth and the GDP growth cycle.

During the 1990 to 1991 recession, loan loss provisions began to rise six quarters before the beginning of the recession. As loan growth slowed dramatically, loan loss provisions increased, peaking about five quarters after the recession had begun as loan growth was nearing its trough (see Chart 12). In many ways, however, the 1990 to 1991 recession was unique. Coming into that recession, commercial real estate was experiencing an unrelated downturn, causing loan loss provisions to rise. Loan loss provisions may behave differently during other business cycles.

Chart 12

[D]

Summary of the Scenario Analysis

At this point we can summarize our general observations about how banks in the aggregate might be expected to perform in a slower-growing economy. Noninterest income and loan loss provisions would typically worsen in the early part of an economic slowdown. At some point, the return to an upward sloping yield curve and a stock market rebound would boost the NIM and noninterest income, respectively. This boost could occur even as loan loss provisions, a more lagging indicator, continue to rise.

These observations are distilled from the experience of the past. However, at least two caveats are needed. Typically, each business cycle has its own unique features; the present economic slowdown need not follow the course of any previous economic downturn. Even more important, the results of the scenario analysis applies to banks in the aggregate, whereas the outcome for an individual bank depends on the risks that bank assumes and the way it manages those risks.

Richard A. Brown, Chief, Economic and Market Trends Section
Alan Deaton, Financial Economist
Gerald Hanweck, Visiting Scholar
Kim Lowry, Writer/Editor
Norman Williams, Regional Economist

The authors would like to thank Paul Bishop and Lisa Ryu for their contributions to the analysis of economic scenarios in this article.