FDIC Home - Federal Deposit Insurance Corporation
FDIC - 75 years
FDIC Home - Federal Deposit Insurance Corporation

 
Skip Site Summary Navigation   Home     Deposit Insurance     Consumer Protection     Industry Analysis     Regulations & Examinations     Asset Sales     News & Events     About FDIC  


Home > Industry Analysis > Research & Analysis > FDIC Outlook




FDIC Outlook

Regional Perspectives

Atlanta Regional Perspectives

Manufacturing in the Atlanta Region: A Help or Hindrance to a Recovery in Employment Growth?

Although a rebound in employment growth late in 2003 fueled optimism about the nation's economy in 2004, the manufacturing sector has continued to shed jobs, prompting concern about whether this sector will be a net asset or liability. Manufacturing is a critical economic component in several Atlanta Region states; shares of employment in the manufacturing sector in Alabama, Georgia, and North and South Carolina exceed the national average of 12 percent. The manufacturing sector's overall contribution to economic growth in the Region likely will be a function of the type of industry and the extent of geographic concentration. This article identifies critical manufacturing industries in the Region, examines their recent performance and prospects for future growth, and assesses implications for local economies and the banking industry.

Perspectives on Industry Performance: Structural and Cyclical Considerations

Manufacturing industry performance is influenced by long-term (structural) and short-term (cyclical) trends. In the short term, manufacturing closely follows the economic cycle, although performance tends to be more volatile than in other sectors of the economy. If economic growth picks up, the performance of the manufacturing industry typically improves. In the long term, structural factors can constrain industry perfor-mance and potential gains in employment.1 Nationwide, during the past few years, increasing competition (and industry relocation) from overseas firms and the slow-paced economic recovery have contributed to significant job losses in the manufacturing sector. Locally, the net effect of structural and cyclical forces can be magnified by specific industry concentrations. Areas with large shares of employment in particular industries—those that continue to shed jobs because of structural forces—are less likely to participate in an economic recovery once cyclical effects moderate.

The Atlanta Region is home to several critical manufacturing industries.
Although the Atlanta Region is home to a variety of manufacturers, our analysis is limited to the relationship between structural and cyclical effects in two types of industries—those classified as "traditional" or "emerging" based on location quotient analysis.2 Traditional industries are defined as those that are relatively more concentrated in the Region than in the nation in terms of shares of total employment or, statistically, those with a location quotient greater than one in 2002.3 Emerging industries are classified as those with a location quotient that increased between 1992 and 2002, indicating that these industries have become increasingly concentrated in the Atlanta Region. Seven industries satisfied the definition of a traditional industry (see Table 1), and they can be described in a number of ways.4 First, these industries continued to represent a large share of the manufacturing sector, accounting for just over 33 percent of all employment, compared with 21 percent nationally. The textiles subsector, despite decades of erosion, remains the Region's largest industrial employer, with more than 300,000 workers, and is characterized by a location quotient that exceeds four. Second, the location quotient in each traditional industry has declined during the past ten years. Third, during the past year, national employment in traditional industries has declined at a more rapid rate than employment in emerging industries. National employment in the apparel manufacturing subsector alone fell by more than 16 percent. Comparatively weaker performance in traditional industries presumably can be attributed in part to greater ongoing structural erosion during the cyclical downturn. Given the greater number of traditional compared with emerging industry exposures in the Atlanta Region, manufacturing employment losses have continued to outpace those at the national level.

Table 1

The Atlanta Region’s Industrial Base Reflects Its Diversity
  Atlanta Region Nation
Industry
(Ranked by 2002 location quotient)
Location Quotients Industry Employment Year-Ago Job Growth
  1992 2002 2002 2003 Q3
Manufacturing 1.03 0.94 2,746,501 -4.3%
Traditional Industries
Textiles 4.42 4.11 316,279 -9.8%
Beverage and Tobacco Products 1.68 1.39 47,945 -5.4%
Furniture 1.70 1.38 118,925 -5.0%
Wood Products 1.38 1.32 177,086 -2.9%
Apparel 1.72 1.16 104,330 -16.1%
Paper Products 1.13 1.08 117,662 -4.2%
Nonmetallic Mineral Products 1.11 1.07 104,407 -3.3%
Emerging Industries
Food Manufacturing 0.91 0.94 248,947 0.5%
Primary Metal Manufacturing 0.79 0.81 83,735 -6.6%
Printing and Publishing 0.78 0.80 196,797 -2.1%
Transportation Equipment 0.66 0.77 224,696 -3.3%
Computer and Electronic Manufacturing 0.65 0.67 208,234 -7.5%
Petroleum and Coal Products 0.23 0.37 8,024 -1.7%
Sources: Bureau of Labor Statistics/Haver Analytics; Global Insight, Inc.

In contrast, six critical industries in the Atlanta Region are characterized by location quotients that increased between 1992 and 2000 and therefore are classified as emerging industries.5 Although all industries except food processing lost employment during the past year, the decline was almost half that experienced by traditional industries. Transportation equipment, namely automobile manufacturing, is a good example of this type of industry. Manufacturers have relocated or built new facilities in recent years to take advantage of the Region's low cost of doing business. Similarly, during the high-tech boom, the Region benefited from growth in computer and electronics equipment manufacturing. If the economic recovery continues to strengthen in 2004 or cyclical forces moderate, these industries may contribute more toward job growth, as they, unlike traditional industries, may not be vulnerable to structural employment losses.

The strength of the economic recovery in certain areas of the Atlanta Region may be a function of industry exposure.
The degree of economic diversity varies widely across the Atlanta Region, and geographic exposure to either traditional or emerging industries may affect an area's economic growth prospects significantly. Some areas stand out in our analysis as a result of high exposures to a number of traditional industries. The Greensboro and rural North Carolina areas, for example, have location quotients greater than one for every traditional industry identified in our study. Moreover, employment in traditional industries is high, accounting for 11 percent of all jobs in the Greensboro metropolitan statistical area (MSA). Although the Greensboro MSA is characterized by a concentration in printing and publishing, employment in traditional industries is nearly three times as high as employment in this emerging industry. The Danville, VA; Florence, SC; and Hickory, NC MSAs not only have relatively high concentrations in multiple traditional industries but also are characterized by greater levels of aggregate employment in traditional compared with emerging industries (see Table 2); rural areas of Alabama, Georgia, and South Carolina can be described similarly. Continued erosion in traditional industries, combined with increasing overseas competition and greater use of automation, likely will constrain any recovery in economic growth in areas where exposures to traditional industries remain high.

Table 2

Several Areas of the Atlanta Region Have Multiple Exposures to Traditional Industries
  Traditional Industries1  
Area Number with High2 Concentration Employment, 2002 Traditional/Emerging Employment Ratio
Greensboro, NC 7 71,838 2.9
Rural North Carolina 7 106,270 2.1
Rural Alabama 6 75,088 2.1
Danville, VA 6 9,365 3.8
Florence, SC 6 3,615 2.2
Rural Georgia 6 32,026 2.1
Hickory, NC 6 48,947 4.4
Rural South Carolina 5 41,766 2.1
1 Textiles, beverage and tobacco products, furniture, wood products, apparel, paper, and nonmetallic minerals.
2 Location quotient greater than one.
Source: Global Insight, Inc.

In contrast to the potential downside risks associated with a concentration of traditional industries, relatively high exposures to emerging industries may contribute to a rebound in economic growth if cyclical pressures moderate. Although these industries exist in several areas in the Region, our analysis determined that their presence frequently was eclipsed by the scale of traditional industry employment. Greenville, SC, for example, boasts a healthy transportation equipment industry (automobiles), which is an emerging industry. However, this area also is home to high employment concentrations in textiles, apparel, furniture, and nonmetallic mineral products (four traditional industries); employment in the aggregate for these traditional industries was 50 percent higher than in emerging industries. The situation is reversed in other areas, such as the Tuscaloosa, Birmingham, Dothan, and Huntsville, AL; and Charlottesville, VA, MSAs, which have relatively high exposures to emerging industries and aggregate employment in emerging industries that exceeds that for traditional industries. A number of metropolitan areas in Florida have employment exposures to single emerging industries, such as computer or transportation equipment manufacturing, and no traditional industries; however, employment in emerging industries remains low.

Although equity performance points toward a moderation in cyclical constraints in several traditional and emerging industries, total manufacturing job growth has yet to recover. Many economists are forecasting a pick-up in the manufacturing sector because of a decline in the relative trade value of the dollar over the past year. Exports showed strength in late 2003, and a December survey of purchasing managers indicated the strongest foreign demand for U.S. goods in 14 years.6 While accelerating export activity is positive news, it is uncertain which industries will benefit most.

What Are the Implications for Banking?

During the past 12 months, industry employment exposures in the Atlanta Region may have played a role in community bank loan performance.7 Overall, credit quality has improved; the average noncurrent loan level reported by community banks in the Region declined during the year ending September 30, 2003. However, community banks based in states characterized by significant exposure to traditional industries reported an increase. For example, insured institutions based in Virginia, North and South Carolina, and Georgia reported an average increase in noncurrent loans of 11 basis points between September 30, 2002, and September 30, 2003. Community banks based in other states in the Region, or in states characterized by multiple exposures to emerging industries and where employment in emerging industries is greater than in traditional sectors, reported an average decline in noncurrent loan levels of 15 basis points during the same period (see Chart 1).

Chart 1

The Region's Large Banks Report Solid Financial Performance


Profitability remained solid among large banks headquartered in the Atlanta Region. Median return on assets (ROA) grew for a second straight year and finished September 30, 2003, at 1.22 percent, up 6 basis points from a year earlier. A drop in net interest income was offset by lower provision expenses and securities gains, which led to the higher ROA figure. After rising in third quarter 2002, the median net interest margin (NIM) slid during the most recent 12-month period, falling 36 basis points to 3.60 percent by September 30, 2003. Nevertheless, a steeper yield curve during third quarter 2003 may result in an easing of margin pressures going forward, but a substantial improvement in the NIM is not likely to occur until commercial loan growth improves.

However, since third quarter 1999, commercial loans as a percentage of total loans have fallen 11.5 percentage points to 19 percent at September 30, 2003. Large banks have shifted portfolio emphasis from this traditionally higher-yielding asset class into lower-yielding one-to-four family mortgages, which have increased 8 percentage points during the same period to 33 percent of total loans. A healthy housing market and an unprecedented level of refinancing activity in 2003 combined with strong consumer demand helped offset weakness in the commercial sector.

Asset quality has continued to improve among the Region's large banks. The ratio of median past-due and nonaccrual loans as a percentage of total loans fell for a second straight 12-month period to finish September 30, 2003, at 1.44 percent, down 42 basis points from September 30, 2002. However, strong loan growth, especially in the one-to-four family mortgage portfolio, may be amplifying the level of improvement.

Obviously, a variety of factors affect the performance of institutions in these states, and many of these factors take time to work through the financial statements of individual institutions. It is reasonable to expect that weaknesses in traditional industries are among the factors leading to differences in noncurrent ratios among institutions in states with concentrations in traditional industries versus emerging industries. These differences could widen if weakness in the manufacturing sector continue.

Jack Phelps, CFA, Regional Manager
Scott Hughes, Regional Economist
Ron Sims, CFA, Senior Financial Analyst
Pam Stallings, Senior Financial Analyst

1Long-term industry trends often include greater automation, cost competitiveness (both domestically and internationally), use of part-time or contract workers, and niche or specialized manufacturing.
2 Approximately 30 percent of manufacturing employment in the Atlanta Region was not classified as either traditional or emerging because these jobs did not meet our criteria. Chemicals, although a large employer in the Region, with more than 150,000 workers and considered a critical industry in certain local areas, is characterized by a location quotient less than one in 2002, down 0.17 points from a decade earlier.
3 A location quotient measures an industry's share of local employment relative to the corresponding national share. Algebraically, the calculation is shown as
Industry Location Quotient (LQ) =
[Local Industry Employment/Total Local Employment]/
[National Industry Employment/Total National Employment].
If the calculated ratio is greater than one, an industry is more concentrated locally than nationally. Refer to the Atlanta Regional Outlook, Spring 2003, for more information.
4 Textiles, beverage and tobacco products, furniture, wood products, apparel, paper products, and nonmetallic mineral products.
5 Food processing, primary metal manufacturing, print and publishing, transportation equipment, computers and electronic manufacturing, and petroleum and coal products.
6 James C. Cooper and Kathleen Madigan, "The Economy's Big Mo Is Beating Expectations," Business Week, January 26, 2004.
7 Community banks are defined as commercial banks that hold assets less than $1 billion and exclude specialty institutions.


 

Chicago Regional Perspectives

Signs of Economic Improvement Are Uneven among Industry Sectors and States in the Chicago Region

Certain key developments indicate that the Region's economy is performing better than at any time since before the 2001 recession. For example, the Midwest Manufacturing Index (MMI) rose in third quarter 2003, the first gain in four quarters and the largest since early 2000. This upturn accompanied improvement in the Region's labor market, as third-quarter job losses slowed to less than an annual rate of 0.5 percent (see Chart 1). The October MMI reading suggests that, even should no additional advance occur in November and December, this gauge of manufacturing activity in the Chicago Region will post an annual rate of increase of at least 2.5 percent in fourth quarter 2003.

Chart 1

To date, however, growth in output has not led to net gains in employment in the Region. Several sectors, such as leisure and hospitality, education and health, and professional and business services, hired additional workers even as large layoffs occurred in manufacturing and government during third quarter 2003. Employment conditions among states also are uneven. Wisconsin for example, is the only state in the Region to report an increase in employment during each of the first three quarters of 2003, while employment in Michigan fell by 1.3 percent during the same time frame.

Nationally, the length of the workweek and number of overtime hours have been rising among factory workers, according to the Bureau of Labor Statistics. Longer workweeks and additional overtime help boost wages and salaries for this group, but manufacturers are not likely to hire additional workers until they use their existing workforce more fully. Gains in manufacturers' orders in recent quarters suggest that factory output should continue increasing as 2004 unfolds, boosting capacity utilization rates, reducing job layoffs, and perhaps triggering hiring.

However, until total employment in the Region shows sustained and broad-based gains, financial strains among some households and repercussions such as high personal bankruptcy and mortgage foreclosure rates likely will persist. In addition, some retired workers—such as those who worked for steel companies that filed for bankruptcy—are experiencing financial setbacks because of dramatic reductions in their pension and health benefits. When the Pension Benefit Guaranty Corporation assumes the defined-benefit pension obligations of firms that file for bankruptcy, pensioners' monthly benefits are subject to a maximum amount that may be far less than they had been receiving or anticipated.1

Not unexpectedly, signs of consumer repayment problems have emerged, and loan performance has deteriorated fairly quickly among one-to-four family mortgages. On September 30, 2003, the percentage of past-due or nonaccrual (PDNA) residential mortgages held by community institutions in the Region was relatively high, at 2.45 percent, and matched the rate for consumer loans.2 Putting this figure into perspective, during 1997 through 2001, third-quarter PDNA rates for one-to-four family mortgages ranged 45 to 60 basis points lower than for consumer loans; in 1992 through 1996, the difference was at least 90 basis points. Compared with other segments of the loan portfolio, the September 2003 PDNA rate for mortgages was exceeded only by commercial and industrial (C&I) loans (3.24 percent) and construction and development loans (2.73 percent). To date, however, the charge-off rate for mortgage loans remains relatively low.

In contrast, the third-quarter PDNA ratio on one-to-four family mortgages held by the Region's largest institutions (those holding assets of at least $20 billion) was 3.32 percent, higher than for C&I loans (3.12 percent) and for other major loan groups. This relatively high PDNA ratio for mortgages could reflect a number of factors, including these institutions' strategic policies and greater risk tolerance; geographic exposure beyond their local area; greater exposure to subprime, jumbo, and nonconforming loans; and use of third-party brokers or appraisers. As past-due mortgage rates rose, so did the average charge-off rate for mortgage loans among large institutions based in the Chicago Region. Third-quarter charge-off rates for one-to-four family mortgages have been 0.30 percent or higher since 2001, about triple the rate in the previous few years.

In recent years, many homeowners refinanced their debt and locked in fixed-rate mortgages at low rates, an act that should help shelter them from rising debt burdens as interest rates rise. However, refinancing activity slumped in recent months as mortgage rates rose and the pace of home appreciation in the Region slowed. As a result, some households may be less able to support spending by taking equity out of their homes and lowering debt payment burdens.

Rising Interest Rates Are Likely to Affect Earnings in a Variety of Ways

Economic growth is improving and becoming more broad based across the nation; as a result, interest rates are expected to rise. Indeed, although the Federal Open Market Committee maintained the target fed funds rate at 1 percent at its December 2003 meeting, market forces have pushed up yields on intermediate- and longer-term Treasury securities since midyear. For example, the yield on the five-year constant-maturity Treasury note in December was 100 basis points above the June low of 2.27 percent. With the exception of a brief interval from late 2001 into early 2002, the rise in rates during the second half of 2003 for securities with a maturity of one year or more reversed the three-year trend of falling rates that began early in 2000.

Looking ahead, the Blue Chip Economic Indicators consensus forecast calls for the yield curve to show a parallel upward shift during 2004, as yields on three-month Treasury bills and ten-year Treasury notes are expected to rise by 80 basis points.3 A subgroup of this forecast's participants expects not only greater increases in interest rates but also a flattening of the yield curve; they forecast a 140-basis-point increase during 2004 in the three-month Treasury bill rate and a 110-basis-point increase for the ten-year note.

Improving economic conditions and a shift from a sustained period of low interest rates to one of rising rates are expected to affect insured institutions based in the Chicago Region in a variety of ways. Some of the impact to date is illustrated by the following interest-sensitive components of the return on assets (ROA) ratio, which posted a modest decline in third quarter 2003 relative to a year earlier (see Table 1).

Table 1

Net Income of Chicago Region’s Institutions Changed Little in Past Year
Income statement contribution (as a percentage of average assets)
  Three months ended Sept. 30 Basis point change
  2002 2003  
Net interest income 3.33 3.15 -0.18
Total noninterest income 1.87 1.93 0.06
Noninterest expense -2.97 -2.88 0.09
Provision expense -0.55 -0.43 0.12
Security gains (or losses) 0.23 0.03 -0.20
Income taxes -0.62 -0.56 0.06
Net income (return on assets) 1.29 1.24 -0.05
Source: Bank and Thrift Call Reports for all institutions in the Chicago Region.

Securities gains: Unrealized gains on securities held for sale peaked in third quarter 2002. The decline since then reflects the fact that insured institutions not only realized some gains by selling securities but also lowered securities portfolio valuations as interest rates rose after midyear (see Chart 2). In third quarter 2003, realized gains from securities sales boosted ROA by 3 basis points among insured institutions in the Chicago Region, noticeably less than the 23-basis-point contribution a year earlier. Given the general expectation that interest rates will rise over the next year, the contribution to ROA from unrealized securities gains likely will shrink or turn negative in coming quarters.

Chart 2

Net interest income: From September 30, 2002, to September 30, 2003, insured institutions headquartered in the Chicago Region reported an 18-basis-point decline in net interest income as a percentage of average assets. During this period, the yield on earning assets fell 83 basis points, while the cost of funding earning assets declined to a lesser degree.

Whether a rising yield curve will enhance net interest income in coming quarters depends on insured institutions' interest-rate risk management strategies. A few institutions seemingly were caught by surprise when interest rates started rising recently, as several took charges to unwind funding vehicles with option features that started incurring losses as rates rose.

Banks and thrifts that rely heavily on deposits as a source of funding may benefit if the spread between the yield on earning assets and deposit rates widens. Compression of this spread in recent years likely reflected, at least in part, the reluctance of institutions to lower rates on deposits in tandem with rates on assets, especially after deposit rates fell below the psychologically sensitive level of 1 percent.

Fee income: The upturn in interest rates also is expected to dampen fee income, notably among insured institutions with significant mortgage origination and refinancing activity. Refinancing of residential mortgages has plunged since midyear, and growth in home purchase applications has slowed considerably. Indeed, such large national mortgage lenders as Washington Mutual recently announced planned layoffs of thousands of employees in response to the drop in mortgage underwriting activity.4 Other institutions are taking similar actions, and nationwide employment by credit intermediaries, which includes mortgage banking, fell by 22,000 in fourth quarter 2003, following gains of about a quarter-million over the prior three years that largely reflected increased mortgage refinancing activity.

The Brighter Side of Rising Interest Rates

As insured institutions adjust to some short-term or adverse impacts from rising interest rates, rising rates may help widen the spread between deposit rates paid and yields on earning assets. In addition, other aspects of insured institutions' operations would be expected to benefit from improving economic conditions.

Lower provision expenses already contributed positively to ROA among the Region's community banks in the third quarter. Even though the past-due rate on one-to-four family mortgages is relatively high compared with other loan types, the 30- to 89-day past-due rate on September 30, 2003, for all loans held by community institutions was 51 basis points lower than two years earlier. The improvement in the 30- to 89-day past-due rate for all loans suggests that the percentage of loans seriously delinquent (i.e., past due by 90 or more days or on nonaccrual basis) may ease in coming quarters. Although the Region's economic recovery has not been vibrant and areas of concern remain, the fact that general economic conditions are stabilizing suggests that loan quality may not deteriorate further.

In addition, growth in demand for loans typically lags upturns in economic growth. Consequently, in coming quarters banks and thrifts may be able to expand loan portfolios without easing underwriting standards. The Federal Reserve's recent survey of senior loan officers indicated that demand for consumer loans strengthened in the third and fourth quarters, although demand for home mortgage loans declined. In the same period, a smaller net percentage of banks reported weaker demand for C&I loans. In contrast, demand for commercial real estate loans weakened at about the same pace as in the third quarter.5

Looking Ahead

Although economic growth across the Region at year-end 2003 was neither vigorous nor widespread, certain conditions and leading indicators suggested that momentum was building that could sustain more robust and balanced growth in future quarters. In this environment, insured institutions will continue to face challenges, such as pressure on net interest margins and credit quality concerns. Meanwhile, business and household demand for nonmortgage loans may strengthen, but fee income from mortgage origination activity and the contribution to income from securities gains may fade quickly. A shift to a sustained period of rising interest rates also will contrast with conditions of recent years, reinforcing the need for insured institutions to monitor and manage interest rate risk continually.

Chicago Staff

1 Details about the Pension Benefit Guaranty Corporation (PBGC) and maximum monthly guarantee levels can be found at www.pbgc.gov. For information on defined-benefit pension plans and the PBGC, refer to "Could a Bull Market Be a Panacea for Defined Benefit Pension Plans?" FYI: An Update on Emerging Issues in Banking, January 13, 2004, at www.fdic.gov/bank/analytical/fyi/2004/011304fyi.html.
2 Community institutions are nonspecialty banks and thrifts that have been in existence at least three years and hold assets less than $1 billion.
3 Aspen Publishers, Inc., Blue Chip Economic Indicators, Vol. 28, No. 12, December 10, 2003.
4 Bradley Meacham, "WAMU Cuts Jobs, Profit Outlook as Mortgage Business Slows," The Seattle Times, December 10, 2003.
5 Federal Reserve Board, Senior Loan Officer Opinion Survey on Bank Lending Practices, October 2003.


 

Dallas Regional Perspectives

Banking Industry Consolidation May Mask Competitive Effects of Increased Branching Activity

Consolidation in the banking industry has been dramatic, with the total number of Federal Deposit Insurance Corporation (FDIC)-insured institutions declining 29 percent during the past decade, from nearly 13,000 to approximately 9,200. Over the same period, however, the number of physical branch offices increased 15 percent nationwide.1 The growth in the number of physical branches is all the more striking in that it occurred during a period of rapid technological advances, including the rise of the Internet and increasing broadband capacity, which enabled customers to bank on-line.

Consumers have been the engine of economic growth through the recent recession and period of gradual recovery. The branch has become the most prominent delivery channel in the competition for consumer business; as a result, the number of de novo branches has increased. However, some observers now believe that banks may have overplayed branch expansion, particularly if the consumer sector cools.2

Banking industry consolidation in the Dallas Region has been on a par with that of the nation, but growth in the number of branches, at 42 percent, is nearly triple that of the nation, although it varies significantly among states. This article discusses trends in consolidation and branching in the Dallas Region. It also examines differences in overall performance and risk profiles based on the nature of branching activities to determine the effects of certain branching strategies.

Economic and Demographic Conditions Are Driving New Branch Activity

Branching activity has varied among states in the Region (see Table 1). Colorado leads the group, as the number of branches has more than doubled in that state. Branch growth rates in Texas, Oklahoma, and Arkansas also significantly outpaced those of the nation, while Mississippi lagged the nation with only 8 percent growth.

Table 1

Despite Steady State-Level Declines in the Number of Institutions,
the Number of Branches Has Increased
State 2Q03 Insured Institutions Change in Insured from 2Q94 National % Rank 2Q03 Branch Count Change of Branches from 2Q94 National Rank % Change
Arkansas 174 -37% 7 1,128 55% 6
Colorado 178 –41% 4 1,167 111% 2
Louisiana 171 –32% 16 1,340 17% 24
Mississippi 105 –22% 37 1,008 8% 31
New Mexico 60 –35% 10 426 20% 23
Oklahoma 278 –25% 32 945 52% 7
Tennessee 209 –26% 29 1,820 20% 22
Texas 707 –33% 13 4,438 60% 4
Region 1,882 –32%   12,272 42%  
Nation 9,232 –29%   77,712 15%  
Source: Bureau of Labor Statistics, Summary of Deposits.

As shown in Table 2, the variations in branching activity by state are generally well correlated with economic and demographic trends. Colorado's heavy branching activity occurred at a time when the state led the Dallas Region in level of and growth in per capita personal income; the state also experienced relatively high population and employment growth rates. Robust economic and demographic factors over the past decade also explain the relatively high level of branching activity in Texas. Conversely, states with branching activity close to or less than the national average (Louisiana, New Mexico, Tennessee, and Mississippi) have been characterized by less favorable economic or demographic factors during the past decade.

Table 2

The Variation in State-Level Economic and Demographic Conditions Helps to Explain Differences in Branching Levels
  Per Capita Personal Income Population Employment Top 5 Institutions’ Control of Market Share
State Average Annual Level, 1993–2002 National Level Rank Average Annual Change, 1993–2002 National Change Rank Average Annual Change, 1993–2002 National Change Rank Average Annual Growth, 1993–2002 National Growth Rank 2Q 1994 2Q 2003
Arkansas 20,065 48 3.7% 45 3.4% 2 1.8% 26 18% 30%
Colorado 28,095 8 4.6% 4 2.6% 3 3.2% 4 40% 44%
Louisiana 21,352 44 4.2% 13 0.4% 45 1.6% 35 37% 57%
Mississippi 18,976 50 4.3% 7 0.9% 23 1.6% 34 48% 50%
New Mexico 20,187 47 3.9% 31 1.5% 12 2.5% 9 39% 54%
Oklahoma 21,478 42 3.7% 43 0.8% 29 2.0% 17 26% 34%
Tennessee 23,584 34 3.9% 34 1.4% 14 1.7% 28 41% 51%
Texas 24,362 29 4.1% 19 2.1% 7 2.6% 8 42% 44%
Region 24,546   4.1%   1.6%   2.3%   17% 26%
Nation 26,259   3.9%   1.2%   1.8%      
Source: Bureau of Labor Statistics, Summary of Deposits.

Economic and demographic factors are not the only explanations for the level of branching activity. Less concentrated markets also have experienced growth as competitors opened branches to gain market share. For example, as shown in the last column of Table 2, the Arkansas market was highly fragmented in 1994, with the top five institutions controlling only 18 percent of the deposit market. Despite poor economic fundamentals, including low levels of per capita personal income and weak employment growth, the number of branches in the state increased 55 percent, with the top five institutions controlling 30 percent of the market at the end of the decade.

Some of the increase in branching activity can be attributed to changes in state and federal laws, which eased restrictions on branching within and across state lines. Empirical studies have analyzed the effects of an easing in branching restrictions; the results suggest that deregulation contributes to greater profit efficiency (during a time when costs have increased and spreads have declined) and an increase in the number of offices per capita.3 As a result, during the past ten years, it is reasonable to assume that branching would have been greater in states with previously restrictive laws, such as Oklahoma, Texas, and Colorado.

It is instructive to review branching activity below the state level, because branching decisions are typically market specific—often at the county level, or in urban areas at the ZIP code level or below. For the purposes of this article, it is not practical to review the conditions and trends for all 738 counties and 3,509 ZIP codes that are home to branches in the Dallas Region.

The banking analyses conducted for this article are limited to the county level. Analysts researching branching issues and trends in banking industry competition can access the FDIC’s website at www.fdic.gov for data at the ZIP code level, as well as state, metro, and county levels. The Industry Analysis section provides links to bank data (Institution Directory, Summary of Deposits) and statistics (Regional Economic Conditions) with helpful user guides.

However, a comparison of trends in a sample of counties that exhibited the most rapid and slowest rates of branching activity provides helpful insights. The rapid growth group excludes counties that were home to fewer than ten branches as of June 30, 2003, and comprises 76 counties that ranked in the top decile for growth in the number of branches. The slow growth group consists of 76 counties that ranked in the bottom decile for growth in the number of branches. It is important to note that the number of branches actually declined in 68 counties in the latter group during the past ten years.

Not surprisingly, the rapid growth counties overwhelmingly are in metropolitan areas that experienced generally favorable economic and demographic trends during the past decade. Indeed, Austin, Dallas, Denver, Fort Worth, Houston, and Oklahoma City each added more than 100 branches and together accounted for more than a third of all new branches in the Region. In contrast, the vast majority of the slow growth counties are in rural areas that have been characterized by decidedly less favorable trends (see Map 1). The median per capita personal income level in the rapid growth counties was almost 116 percent of that in the slow growth counties during the past decade. In addition, median annual population and employment growth levels in the rapid growth counties were 5.2 and 4.2 times greater, respectively, than in the slow growth counties.

Map 1

Performance Varies Markedly Depending on Branching Strategy

There are significant differences in performance and risk characteristics based on the existence and nature of branching activities among the 1,943 banks operating in the Dallas Region as of June 30, 2003.4 For analytical purposes, insured institutions were categorized in four groups:

  • Group 1 operated branches exclusively in metropolitan statistical areas (MSAs).
  • Group 2 operated branches exclusively outside MSAs.
  • Group 3 operated a combination of MSA- and non-MSA-based branches.
  • Group 4 had no branches.

Our analysis also identified Subgroup A, which consists of banks with headquarters in non-MSAs that have attempted to improve performance by branching into MSAs. Banks in Subgroup A also fall into Group 1 or Group 3.

Overall, insured institutions that operate branches displayed significantly stronger growth rates, higher rates of lending, and higher operating profits than those without branches (see Table 3). Banks that operate branches also reported lower average ratios of Tier 1 risk-based capital to risk-weighted assets, indicating that they have greater opportunities to leverage risk.

Table 3

Financial Trends Vary among Those with Branches and Contrast Significantly with Those without Branches
Group Name Number of Insured Institutions Deposit Growth (median, %) Core Funding to Total Assets (median, %) Time Deposits to Total Assets (median, %) Loan-to-Asset Ratio (median, %) Past-Due Ratio (median, %) Return on Assets Pretax (median,%) Quarterly Net Interest Margin (median, %) Tier 1 Risk-Based Capital to Risk Weighted Assets
Group 1 558 9.3 36.4 33.5 62.8 1.9 1.45 4.36 12.85
Group 2 494 5.0 27.8 41.0 57.4 2.8 1.50 4.25 15.27
Group 3 353 7.3 35.4 35.6 64.5 2.0 1.63 4.32 12.22
Group 4 538 3.7 27.8 40.4 51.9 2.7 1.38 4.05 19.04
Subgroup A 196 7.0 31.7 40.1 64.1 2.3 1.54 4.39 12.85
1 Subgroup A banks also appear in Groups 1 and 3.
2 Core funding includes demand deposit accounts, money market deposit accounts, and savings accounts.
3 Time deposits include certificates of deposits and time open accounts held in domestic offices.
Source: Bank and Thrift Call Reports, Summary of Deposits.

Among those with branches, the groups operating at least one branch in an MSA reported the highest median asset and deposit growth rates, roughly 2.5 times those of institutions without branches. A similar observation applies to median pretax return on assets, with the banks in Group 1 and Group 3 realizing an advantage of more than 25 basis points compared with banks without branches. The ability of banks operating in MSAs to invest significantly greater shares of assets in loans likely explains much of their edge in earnings performance. Earnings also may benefit from greater levels of core funding (demand, savings, and money market deposit accounts) and the lower costs typically associated with these funding sources. Finally, banks that operate branches in MSAs have reported significantly lower median past-due ratios than those without branches or those that branch only in rural areas. These performance data seem to suggest that banks in Subgroup A (banks with headquarters outside MSAs) have benefited from branching into more robust markets.

Looking Ahead—Will the Pace of Branch Growth Continue?

The decision to open or acquire a branch or maintain an existing branch is based on a determination that doing so will provide a net benefit/profit. Only bank management can make this determination, as it is specific to markets, branch types, and the institution's strategy and business mix. Although growth undoubtedly will continue in various markets, one simple measure of feasibility—the number of people per branch—suggests that overall branch growth may moderate. In fact, the number of customers available to support a branch in the Dallas Region declined by approximately 20 percent during the past decade, falling to an average of 2,422 in non-MSAs and 4,562 in MSAs.

Other trends in retail business conditions also have implications for a particular bank's branching strategy. Nationwide, deposit growth varied during the ten years ending June 30, 2003 (averaging 5.5 percent), with the strongest gains coming after 2000, a trend attributable at least in part to the decline in the equity markets. However, with the recent rebound in the stock market, the third quarter 2003 FDIC Quarterly Banking Profile reported the first quarter-over-quarter decline in deposits since first quarter 1999. Moreover, while the consumer sector has remained strong, management must ask whether retail banking will retain its attraction when other types of businesses rebound or when interest rates rise.

Clearly, bank management must consider a number of factors related to current business conditions when making branching decisions—the increased competition arising from a greater number of branches, higher land and building costs, the decline in the number of people per branch, and challenges facing the retail banking business. All these factors together could indicate that the time required for a new branch to become profitable may increase, if it has not done so already in some markets—a key calculation that must be factored into an overall branching strategy.

Memphis Staff

1 The Federal Deposit Insurance Corporation collects deposit data at the branch level as of June 30 every year. Data from June 1994 through June 2003 were used for this article.
2 Greta Sundaramoorthy, "Deposit Drop Looks Like More Than a Blip: Some See Effect on Industry's Branch-building Binge," The American Banker, December 22, 2003.
3 A.A. Dick, Nationwide Branching and Its Impact on Market Structure, Quality and Bank Performance, Finance and Economics Discussion Series, Washington, DC: Federal Reserve Board, 2003, and R.B. Avery et al., "Changes in the Distribution of Banking Offices," Federal Reserve Bulletin, September 1997, pp. 707-725.
4 Included in the 1,943 are 61 banks that operate branches in the Dallas Region but are headquartered outside the Region.


Kansas City Regional Perspectives

Hydrological Drought Conditions Are Expected to Affect Farmers and Their Lenders

In the Winter 2003 FDIC Outlook, the Kansas City Regional Perspectives article described how drought conditions have existed in the Kansas City Region since 2000. Nebraska, western Kansas, and southern South Dakota have been the hardest hit, experiencing at least moderate levels of "agricultural" drought during three of the past four years. This article discusses another type of drought that is affecting much of the Region and is being aggravated by agricultural drought conditions: "hydrological" drought.

Hydrological Drought Conditions Are Significant and Increasing

Agricultural drought refers to topsoil moisture levels that are important for proper crop development. Hydrological drought focuses on the longer-term availability of water for all uses, including farming, urban uses, manufacturing, and recreation. Specifically, hydrological drought refers to shortages in surface or subsurface water supplies, such as reservoirs, rivers, and aquifers. According to the Drought Mitigation Center, a research institute at the University of Nebraska, precipitation shortfalls typically contribute the most to hydrological drought conditions, but factors such as increased land development, landscape, and construction of dams may also have a significant effect. Precipitation deficiencies can cause agricultural drought to manifest very quickly, but they take longer to cause hydrological drought. Hydrological drought can be observed in declining lake and reservoir levels, reduced stream and river flows, and depleted aquifer levels.

In the Kansas City Region, the effects of hydrological drought on surface water levels have increased in severity as a result of lower than normal rainfall and snowfall levels during the past few years. As shown in Map 1, Kansas and Nebraska are experiencing the most severe drought. In Kansas, the river system is running quite low; the flows of the Arkansas, Cimarron, Republican, and North and South Platte Rivers all have declined during the past decade.1 The greatest decrease in flow has been in the Arkansas River because of drought and upstream water diversion for irrigation and recreational purposes. In Nebraska, reservoir levels show the greatest impact of the drought. The water level in the state's two largest reservoirs, Lake McConaughy and Lake Harlan, declined 24 percent and 29 percent, respectively, between October 30, 2002, and September 30, 2003. As of September 30, 2003, these lakes stood at 25 percent and 36 percent of their normal capacities, the lowest levels since they were originally filled.2

Map 1

As disturbing as surface water levels are, the worst may not be over. Climatologists such as Al Dutcher with the University of Nebraska predict that it will take several years of much higher than normal precipitation, typically in the form of snowfall, to recharge these water levels.3 However, a multifederal agency study that combines various climatological models predicts that the Kansas City Region will continue to see abnormally dry to moderate drought conditions over the next five years, which does not bode well for replenishment of water supplies.4

Although the low reservoir and river levels are troubling, they are only a readily apparent, visual indication of a much larger problem. The hydrological drought has had a profound effect on the Region's underground water system, the largest part of which is the Ogallala Aquifer, a vast geologic formation that sprawls below eight states from South Dakota to Texas (see Map 2).5 Nebraska, Kansas, and South Dakota are positioned over 77 percent of this massive aquifer's available water. Under Nebraska alone, the aquifer contains approximately 2,130 million acre-feet of water, and under Kansas it contains 320 million acre-feet. For comparison, the cumulative level of the top 17 reservoirs in Kansas, even if filled to capacity, is just 6.7 million acre-feet of water. The agricultural drought has affected the Ogallala Aquifer in two ways: less precipitation has caused the replenishment rate to be far below average, and it has also caused farmers to draw more water from the system for crop irrigation. In some of the most severely affected areas, the water table levels have declined by as much as ten feet per year. As a result, farmers have incurred higher costs to drill deeper wells and have had to pay more in extraction costs to bring water up from lower pumping levels.

Map 2

Long-term factors also have affected the aquifer system adversely. Crop irrigation, which began in earnest in the 1940s, has gradually reduced the volume of water in the Ogallala Aquifer. According to the University of Nebraska Water Center, the aquifer lost 56 million acre-feet of water between 1987 and 2002. The greatest water level changes occurred in southwest Kansas and the southwestern part of the Texas Panhandle, where up to 50 percent of the water has been depleted, compared with pre-irrigation levels.6

The Ability to Irrigate Is the Key to Many Farmers' Fortunes

An estimated 95 percent of the water extracted from the Ogallala Aquifer each year is used to irrigate crops. In the Region's western half, some crops, such as corn, require more water to produce profitable crop yields than precipitation alone can provide. Crops that require less water, such as wheat and soybeans, are planted in areas where irrigation is not available or cannot be utilized fully. However, the returns to farmers are typically far less than if they grew irrigated corn, which produces much higher yields. During the growing season, the average corn crop requires 25 inches of water—from rainfall or irrigation—to reach maximum yield potential. In normal precipitation years, rainfall accounts for about 13 inches, and farm operators apply about 12 inches of irrigation water. By contrast, in severe drought years, such as the Region experienced in 2002 and 2003, many farm operators had to apply as much as 20 inches of water.

Water shortages have led many water districts in Nebraska, Kansas, and South Dakota to limit the amount of water that farmers can use for crop irrigation. In these areas, water meters have been installed on wells, and water allocations typically are provided over a five-year period. Because of the severe drought that has affected the western half of the Region, examiners note that some farmers used more than their yearly water allocations to grow irrigated corn in 2002 and 2003, effectively "borrowing" water from future years. If higher than normal rainfall does not occur in upcoming growing seasons, these farmers will be forced to make tough decisions. They could reduce water application rates, which will result in lower corn yields, or they could substitute lower-earning crops such as wheat or soybeans. Either way, farmers' cash flows are vulnerable in the short term. Even farmers who have adequate water allocations remaining could face higher pumping expenses to bring water up from declining water tables.

Even more significant than short-term considerations are the long-term effects of water shortages. Communities use water supplies not only for crop irrigation but also for related agricultural operations, hydroelectric power, recreation, and barge traffic. Usage is determined politically; urban population growth, a changing economic mix (less oil and gas extraction, more light industry), and increased environmental concern have contributed to a change in priorities in drought-affected states. Crop irrigation has represented the primary use of water supplies to date, but now priority has begun to be assigned to wildlife habitats, recreation, and water quality.7 Governors in Nebraska and Kansas have initiated task forces to study the effects of water shortages and recommend actions to prevent disruptions. Many foreseeable scenarios involve increased restrictions on crop irrigation; in fact, in Nebraska the recent settlement of a lawsuit with Kansas regarding use of the Republican River has resulted in the installation of water meters (to be completed by year-end 2004) and a moratorium on new irrigation wells.8 The next logical step will be water allocations where none had previously existed.

Banks in the Region May Feel the Effects

Hydrological drought could eventually have serious consequences for many of the Region's insured financial institutions. Approximately 22 percent of all counties in the Region are irrigated significantly and have been affected adversely by drought conditions (see Map 3).9 Most of these counties are in Nebraska and Kansas. If hydrological drought conditions result in irrigation problems, farmers will face the prospect of lower cash flows, as well as the potential for declining land values.10 Banks in these counties would be the most vulnerable to any resulting weakness in farm income. Eighty percent of the 299 banks headquartered in these counties are considered farm banks because of their relatively high agricultural lending concentrations.11 The current agricultural drought conditions already have stressed credit quality among these farm banks. At September 30, 2003, about one-quarter of the farm banks based in these counties reported past-due or nonaccrual loans that exceed 5 percent of total loans, up from 15 percent of banks a year ago. By contrast, only 10 percent of the Region's farm banks in areas that have not experienced multiple years of drought reported this level of problem loans.

Map 3

In conclusion, the hydrological drought could have significant adverse effects on farmers and their lenders. In the short term, farmers face cash flow difficulties from a variety of sources—from reduced crop yields for farmers who have used more than their annual water allocations to higher water-pumping costs. Over the long term, changes in water policy during the next few years likely will be incremental, barring the return of extreme agricultural drought conditions. However, any restrictions on the use of water beyond the status quo would hurt farmers and their lenders.

Shelly M. Yeager, Financial Analyst
Allen E. McGregor, Supervisory Examiner

1 Kansas Geological Survey, Kansas Geological Survey Open File Report 2003-41, p. 12.
2 Nebraska Department of Natural Resources, Surface WaterInformation, 2003, http://waterdata.usgs.gov/ne/nwis/current.
3 Agweb.com, February 11, 2003, www.agweb.com/news_show_news_article.asp?file=AgNewsArticle_20032111447_5412&articleid=95259&newscat=GN.
4 Climate Prediction Center, National Climatic Data Center, and National Oceanic and Atmospheric Administration, December 13, 2003, www.cpc.ncep.noaa.gov/products/predictions/experimental/edb/lbfinal.gif.
5 Sometimes the terms "Ogallala Aquifer" and "High Plains Aquifer" are used interchangeably; while they are related, they are two separate water tables. The High Plains Aquifer is a large (approximately 33,500 square miles of surface area) body of sands, gravels, silts, and clays. In western Kansas it is generally identical with the Ogallala formation, and the aquifer system was originally known as the Ogallala Aquifer. However, the part of the aquifer extending into south-central Kansas (east of Ford County) is now recognized as a hydrologically similar but geologically different formation, and the combined aquifer system is referred to as the High Plains Aquifer. Kansas Geological Survey Open File Report 2000-29, Lawrence: University of Kansas.
6 Water-level Changes in the High Plains Aquifer, Predevelopment to 2001, 1999 to 2000, and 2000 to 2001, Lincoln: University of Nebraska Water Center, 2003, http://watercenter.unl.edu/whatsnew/water_levels.htm.
7 Managing Water: Policies and Problems, Lincoln: Drought Mitigation Center, University of Nebraska, 2003, www.drought.unl.edu/plan/managewater.htm.
8 Information regarding the lawsuit and the settlement can be found at www.accesskansas.org/kda/dwr/Interstate/Republican_River.htm.
9 Irrigated is defined as greater than 10 percent of cropland in the county is irrigated; the median for the Region is approximately 8 percent.
10 For example, in Kansas in 2003 an acre of irrigated farmland sold for an average of $1,100, while an acre of nonirrigated farmland sold for about $650. Agricultural Land Values, p. 1, Kansas Agricultural Statistical Service, 2003. Permanent reductions in water allocations would likely case farmland values to drop to somewhere within that range.
11 The Federal Deposit Insurance Corporation defines farm banks as institutions with at least 25 percent of loan portfolios in farm operating loans or loans secured by farm real estate.


New York Regional Perspectives

Housing in the Northeast

Against the backdrop of an overall weak national economy—at least until recently—housing has stood out as a principal source of economic strength. Whether evaluated by construction activity, rate of home price appreciation, or residential mortgage credit quality, the housing sector has performed strongly nationwide, including in the Northeast. Nonetheless, concerns about future performance of the housing industry and sustainability of current rates of home price appreciation have increased. This article examines housing market conditions in the Region and the implications of rising interest rates and a more tepid housing market on insured institutions.

The Region's Residential Construction Activity Has Increased in Recent Years, but Growth Is Less than the Nation's

While the number of housing permits in the nation has increased significantly in recent years, the increase in the Northeast has been more modest. This situation is due largely to factors such as lower birth rates, unfavorable migration patterns, and less land on which to build. However, economic developments also have played a part. For example, during the 1980s, the housing boom in the Northeast coincided with an economic revival in the Region. That boom and the accompanying revival did not last. The economic recovery began to falter by the end of the 1980s, and rising interest rates on the heels of speculative overbuilding of real estate sealed the fate of the housing sector. In recent years, new home construction in the Northeast has increased but has not reached previous peaks (see Chart 1). Consequently, housing prices in the Region have surged, far outstripping home price appreciation nationwide.

Chart 1

Many of the Nation's Top Housing Markets Are in the New York Region

The rate of home price appreciation had begun to ease in most of the Region's metropolitan statistical areas (MSAs) through third quarter 2003, although some markets continued to report strong price growth. Of the 50 housing markets with the highest rate of price appreciation as of third quarter 2003, 20 are in the New York Region.1 Areas that warrant monitoring because of rapid and potentially unsustainable rates of home price appreciation generally are clustered around the Region's larger, higher-priced housing markets, which include Providence, RI; New Bedford, MA; and Monmouth-Ocean, Atlantic-Cape May, and Jersey City, NJ. Housing markets in many parts of New Jersey have benefited from favorable employment and immigration trends and constraints on the supply of single-family housing. The Providence and New Bedford markets recently have become attractive as alternatives to the very expensive Boston market. The housing markets that have experienced more significant easing in the rate of home price appreciation in third quarter 2003 include Lowell, Lawrence, and Boston, MA; Nashua, NH; and New York, NY, reflecting, in part, localized softening in these economies following rapid increases in housing prices.2

Home prices have not appreciated to the same extent in all of the Region's markets. Communities throughout much of Pennsylvania, upstate New York, and parts of New England that have weaker, typically manufacturing-based economies have lower rates of home price appreciation than those of the nation. Unlike some of the Region's more vibrant housing markets, population in some of these areas has declined, constraining the demand for housing. Nonetheless, according to third quarter 2003 data from the Office of Federal Housing Enterprise Oversight, although rates of home price appreciation generally have eased, none of the Region's housing markets experienced a decline in the median home price during the past year.

Rates of Home Price Appreciation: Too Much Too Fast?

Strong home price appreciation in some of the Region's housing markets has prompted concern about sustainability. Professors Karl Case and Robert Schiller addressed this possibility in a paper prepared for the Brookings Institution.3 For the most part, the paper supports a soft, rather than hard, landing for housing prices. It cites favorable levels of affordability owing to historically low mortgage rates as a key positive factor.

Other data also support the potential for a soft landing. Unlike during the 1980s, new housing supply has moved with, not ahead of, rising demand. Inventories of unsold existing homes compared with sales during this past year were only slightly elevated, unlike the record lows of previous years. Also, the share of sales of completed houses compared with sales of houses not started or under construction has been near the lowest recorded level and well below levels in the 1970s and 1980s.4 This proxy for supply-demand for single-family housing suggests that speculative construction and lending activity remain in check.

The Region's Community Banks Report Favorable Credit Quality and Strong Growth in Housing-Related Assets

Overall, insured institutions in the New York Region and nationwide have reported favorable residential loan quality in recent years. While parts of the Region, predominantly metropolitan areas in upstate New York and western Pennsylvania, report residential delinquency rates above the national measure, the Region's median residential mortgage loan past-due rate declined steadily throughout the most recent recession, and at 1.0 percent is considerably below the 1.7 percent level in the rest of the nation (see Chart 2).

Chart 2

The Region's community banks are active in residential real estate lending; approximately one-third specialize in residential lending, compared with 10 percent for the rest of the nation.5 In addition, weak commercial and industrial (C&I) loan demand during this economic downturn likely has contributed to strong growth in mortgage-related assets among the Region's community banks (see box for detail on the Region's large banks). Since the beginning of the 2001 recession, community bank portfolios of residential mortgages, home equity lines of credit (HELOCs), and mortgage-backed securities (MBSs) have grown from $229 billion to $285 billion, a 24 percent increase, compared with a modest decline in C&I loans (see Chart 3). The comparatively high growth rate in MBSs is largely the result of the increased size of, and banks' participation in, the secondary mortgage market. The significant growth in HELOCs is primarily the result of the increased popularity of these loans among consumers, a trend that has been aided by appreciating home values. Contributing to the more modest growth rate of first mortgage loans is the large size of the portfolio of first mortgages (making high percentage growth rates more difficult to attain), banks' selling of first mortgage