|
Does Net Interest Margin Matter to Banks?
On the surface, the answer to this question seems obvious. Net interest margin (NIM) is the difference between income generated by earning assets, such as loans and securities, and expenses incurred on interest-bearing liabilities, such as deposits and borrowings. Because banks are in the business of intermediation—taking funds from depositors and other sources and investing in interest-bearing assets—of course NIM matters. But, for some banks, it does not matter as much as it used to.
Over the past 25 years, deregulation, technology, and market forces have contributed to increased competition and significant changes in revenue sources for insured institutions. These trends have resulted in a secular decline in NIM and a concurrent increase in other revenue sources, particularly at very large institutions. As a result, the significance of NIM as a performance yardstick is not what it used to be. This article will focus on trends in NIMs and analyze institutions of similar asset size to identify reasons for differences in NIM performance.
For the purposes of this article, we have divided the commercial banking industry into three segments:1
- Megabanks—commercial banks with assets over $100 billion.
- Large and midsized banks—commercial banks with assets of $1 billion to $100 billion.
- Community banks—commercial banks with assets under $1 billion.
The article also explores whether alternative metrics exist that measure earnings performance more effectively than NIM.
Secular and Cyclical Factors Have Affected NIMs
The average NIM for the industry had fallen from 4.69 percent in 1992 to 4.10 percent by year-end 2003 (see Chart 1). NIMs for the banking industry have been under secular pressure for some time, partly as a result of increased price competition within the banking industry and from nonbanking firms that offer bank-like products. Improvements in technology and other marketplace innovations contribute to price competition. For example, in the past, loan and deposit pricing were largely set in local markets; they still are in some areas, but technology is leading to a convergence in pricing. One result is that depositors can now easily use the Internet to locate and move money to the offering with the highest yield.
Chart 1
[D]
Cyclicity has also played a role in the decline of NIMs, particularly in recent years. Yields on loans have fallen as a result of nominal interest rates at a level that has not been seen since the 1950s. Moreover, despite the current steepness of the yield curve, banks have not recently reaped much benefit in lower costs, as the persistent nature of low interest rates has caused depositors to resist further decreases in deposit rates, thus creating an effective "floor" for deposit costs.
|
Risk-Adjusted Margin
An alternative way to analyze NIM performance is by using a risk adjustment process for credit costs. Credit costs, namely net loan losses, are inherent in almost all forms of lending and should be considered when analyzing returns. The adjustment of loan yields and, ultimately, the margin for credit costs can facilitate comparisons among different types of institutions. For example, NIMs of credit card banks are usually much higher than those of community banks, because credit cards tend to be a riskier business line that yields higher revenues than the more traditional business mix of a typical community bank. However, adjusting for losses, margins between the two become much more comparable.
The NIM charts in this article show the unadjusted traditional NIM in a solid line and the risk-adjusted margin in a dashed line. In Chart 1, for the entire industry, the average risk-adjusted margin has fallen 52 basis points since 1992, while the average unadjusted NIM has fallen 59 basis points. The two metrics have been very close over the measurement period, compared with a much greater variance during the crises of the late 1980s and early 1990s. This closeness is due to improved credit performance over the measurement period. This improved performance has resulted from improved risk management and underwriting processes; enhanced regulatory requirements; and a shift in portfolio lending from large commercial real estate development and business lending to consumer-related lending (single-family mortgages and residential construction).
|
NIM Trends for the Three Asset Size Groups
As a group, megabanks have the lowest reported NIM in the industry. The median NIM for megabanks had fallen from 4.35 percent in 1994 to 3.84 percent by year-end 2003 (see Chart 2). Although NIMs for megabanks are lower than those for the other groups, megabank NIMs have not shown a greater decline over time; that is, historically, these banks have reported the lowest NIMs. As we discuss later in this article, significant differences in business strategies and availability of options between megabanks and other groups are major factors in lower megabank NIMs.
Chart 2
The NIMs of large and midsized banks have fallen much more than those of megabanks. The median NIM for large and midsized banks fell from 4.55 percent in 1994 to 3.91 percent by year-end 2003, a drop of 64 basis points. This drop was much greater than the corresponding 49 basis point decline at megabanks. Interestingly, the risk-adjusted NIMs for the same period for both groups fell about the same amount: For large and midsized banks, the risk-adjusted NIM fell 72 basis points; for megabanks, it fell 71 basis points. Risk-adjusted NIM at megabanks was adversely affected by credit losses on some very large corporate borrowers that experienced problems during the recent recession.
The NIM at community banks has experienced compression almost identical to that at megabanks. The median NIM for community banks had fallen from 4.59 percent in 1994 to 4.11 percent by year-end 2003. The 48-basis-point decline in the median NIM for community banks approximates the 51-basis-point decline for megabanks during the same period.
Megabanks Have Diversified Their Income Streams, Resulting in Less Reliance on NIM
The narrowing NIM at megabanks is part of a gradual alteration of the income stream over the past two decades, in which these banks have purposefully reduced their dependence on spread income. On average, megabanks' net interest income as a share of total revenue fell from 90 percent in 1984 to 65 percent in 2003 (see Chart 3).
Chart 3>
[D]
Several factors have driven the change in earnings composition for megabanks. Traditional key lending areas—such as a large corporate loans—have diminished, in both volume and yield, as a result of competition, technology, and other market forces, including the expansion of the capital markets. To replace this lost revenue, megabanks have expanded into new business lines—such as investment banking, asset management, and insurance—to generate fee income and grow revenue. Megabank managers have sought to diversify their revenue streams to lessen their dependence on any one source and reduce the volatility in revenue over time.
Community banks, on the other hand, still rely heavily on net interest income. For these banks, net interest income as a share of total revenue fell slightly, from roughly 95 percent in 1984 to 89 percent in 2003. Additionally, while there has been an increase in fee income at community banks, for the most part, this income is closely related to net interest income rather than a result of expansion into new business lines. These fees are mainly associated with deposit accounts: monthly service charges, check-cashing fees, and insufficient funds charges. Price increases for these items are the main reason for the small increase in the proportion of fee income to total revenue.
The size of the megabanks (over $100 billion) suggests institutions with the scale and management to operate multiple business lines over multiple geographies and with the largest array of strategic and funding options. Conversely, the size of the community banks (under $1 billion) suggests institutions that operate a traditional local banking business. Income streams at the large and midsized bank group—$1 billion to $100 billion—are more complicated to analyze because of the mixed composition of this group. Some banks in the group tilt more toward the traditional, while the larger banks may emulate the megabanks in terms of business lines and strategies. Still others may be niche players.
Because of the varied nature of the banks in the large and midsized group, the change in revenue stream falls at the midway point between that of megabanks and that of community banks. Net interest income as a share of total revenue fell from 90 percent in 1984 to 80 percent in 2003. The rate of decline has picked up somewhat in the past three years as declines in interest rates have prompted these banks to seek alternative revenue sources and technology improvements have made it easier to access these sources.
Diversifying Income Streams Has Been Beneficial for Megabanks
Historically, earnings have deteriorated during economic downturns as credit costs generally rise. However, record levels of income were generated during the recent recession, partly because of increased revenue from business lines not tied to net interest income. Earnings problems were also muted at megabanks by the relatively mild nature of the recession and a shift in portfolio lending from commercial (where most credit problems were concentrated) to consumer lending. In addition, advances in active credit portfolio management techniques and the development of secondary markets have created new options for managing and transferring credit risk.
Risk management processes have become more sophisticated at megabanks, so that optimization of the NIM ratio often is not a primary goal. A principal focus among managers in this group is the maximization of revenue and total shareholder return. Business lines are scrutinized to determine whether they are exceeding a specified hurdle rate of return on a risk-adjusted basis. Simply stated, business lines with risk-adjusted returns above the hurdle rate are adding economic profit while those below are not.
This hurdle rate methodology is also used to determine the overall profitability of individual customer relationships at megabanks. While one product or service in a relationship may not exceed the hurdle rate of return, other products or services sold to the same customer could raise the overall profitability of the relationship above the prescribed minimum. For example, commercial credit facilities offered by megabanks may be priced below the business line return hurdle. This loss-leader strategy is used to cross-sell other, higher margin, products that lead to profitable relationships in the long run.
A number of banks in the large to midsized group are able to diversify income streams and employ some of the same revenue optimization techniques as megabanks. However, among community banks, the lack of diversification in the income stream has magnified the current cyclical pressures on NIM. Because of the heavy reliance on NIM and the prolonged and very low level of nominal interest rates, which has created a floor on deposit costs, 40 percent of community banks experienced a decline in net interest income in 2003. In 2002, only 18 percent of community banks saw net interest income fall. At many community banks, the year-over-year drop in net interest income occurred even though earning assets rose as a share of total assets. Community banks in the bottom 10th percentile had a 2003 NIM of only 3.14 percent, the lowest level in 20 years (see Chart 4).
Chart 4
[D]
As NIM's Usefulness as a Performance Benchmark Diminishes for the Largest Banks, Are There Alternatives?
Because several factors have changed the revenue stream of the industry over the past two decades, NIM may no longer be the most effective tool for measuring performance. The change is most evident at the megabanks that now control the majority of assets in the industry. Because the megabank group disproportionately influences the NIM aggregate trend, NIM has become less useful as a tool for measuring industry performance.
Megabanks themselves focus much more on net interest income than NIM in their earnings discussions with equity analysts and investors. One metric that could be used is risk-adjusting margins for credit costs, as described in the text box above. Another metric could be the calculation of the ratio of pretax net income to gross revenues. In this calculation, gross revenues are defined as interest income plus fee income. Essentially, this measure determines the percentage of total revenue that flows to the bottom line; it allows greater comparability across different business models, balance sheet and off-balance sheet structures, and tax status. In Chart 5 the time-series pro forma graphic for pretax net income to gross income shows the very strong earnings performance of the banking industry over the past decade—during the past few years in particular.
Chart 5
[D]
Several megabanks are now publicly disclosing their internal performance metrics, such as risk-adjusted return on capital, economic profit, and shareholder value added. These disclosures not only contribute to more complete information on earnings performance but also indicate the risks that are being taken to achieve those earnings.
While NIM may be losing some of its relevance, as long as banks serve as intermediaries, it provides a useful, though not exclusive, measure of earnings performance. For the more than 7,300 community banks and many of the banks in the large and mid-sized group, NIM is still a very important performance measure.
At the FDIC and other bank regulatory agencies, examiners are trained to analyze the quality and composition of earnings sources rather than focusing solely on ratio analysis and peer group comparisons. This is particularly true at large financial institutions, where the complexity and uniqueness of each institution require a more comprehensive and idiosyncratic evaluation.
Jack Phelps, CFA, Regional Manager, Atlanta
Scott Hughes, Regional Economist, Atlanta
Ron Sims, CFA, Senior Financial Analyst, Atlanta
Robert L. Burns, CFA, CPA, Senior Examiner, Atlanta, Division of Supervision and Consumer Protection
1 The article focuses on commercial banks because of the historical importance of net interest income as a revenue source. Before deregulation and the development of secondary markets, thrifts almost exclusively underwrote residential mortgages, with fee income earned in origination and servicing. Funding sources were mainly savings and time deposits. The cost structure of thrifts was significantly below that of commercial banks because of lower underwriting and deposit-servicing costs. Therefore, NIMs have been historically much narrower for thrifts.
Bank Investment Portfolios: Strong Gains since 2000—Will They Continue?
Investment portfolios have traditionally served multiple purposes at banks. A well-managed securities portfolio can reduce a bank's credit risk profile and the volatility of the income stream. Additionally, securities portfolios can provide an alternative investment opportunity in times of sluggish loan demand and can serve as a source of liquidity to fund bank operations. Over time, the increased availability of other liquidity options and the search for higher yields has led to reductions in the relative size of bank securities portfolios. Moreover, yield pressures and supply constraints have changed the composition of the securities portfolio, resulting in far less reliance on U.S. Treasury securities.
Recently, through a period of historically low interest rates, banks benefited
from the inverse relationship between bond values and interest rates and
realized significant securities gains. As reported in the FDIC
Quarterly Banking Profile, gains on securities sold supported
strong aggregate bank profitability throughout the recession and subsequent
recovery.1 However, as banks sold securities and rebuilt their
portfolios, replacement securities often had low yields because of changes
in prevailing interest rates.
This article examines the reasons behind the changes in the size and composition
of bank securities portfolios. It also discusses how the inverse relationship
of bond prices to interest rates may affect bank earnings and capital
cushions going forward, particularly in a rising interest rate scenario.
The implications of rising interest rates on debt security yields and
valuations will be influenced in part by shifts in the composition of
securities portfolios, the magnitude of interest rate changes, and the
shape of the yield curve.2
Funding Alternatives and Yield Pressures Influence Banking Investment Strategies
Securities portfolios have represented a shrinking proportion of most bank balance sheets over the past decade. The median securities-to-total-asset ratio dropped from 30 percent at year-end 1994 to 23 percent by year-end 2003. Even the smallest banks now have numerous liquidity alternatives, a situation that contributes to this decline. In the past, banks maintained securities portfolios to provide a source of liquidity and to help meet deposit outflows or loan demand. Now, banks of all sizes commonly have an array of liquidity options, such as interbank or Federal Home Loan Bank borrowing lines and access to brokered or Internet deposits.
In addition, in the trade-offs among returns, liquidity, and credit risk, securities generally pay lower yields than loans. Thus, when lending opportunities are abundant, as has been the case over the past decade, the incentive is strong to maintain higher loan balances and lower security balances.
Until the early 1990s, insured institutions traditionally invested heavily in federal, state, and local government bonds, with some exposure to U.S. Agency (agency) debt and mortgage-backed securities (MBS).3 However, banks have increasingly shifted securities investments away from default-free U.S. Treasury (UST) instruments and toward agency-issued notes, MBS, and various other debt issues.4 For instance, between 1994 and 2003, the share of institutions with at least 25 percent of securities invested in UST instruments plummeted, while banks investing above this threshold in agency, pass-through MBS, municipal securities, and other debt classes increased (see Chart 1).5
Chart 1
[D]
Yield considerations as well as developments in the capital markets contributed to the shift in investment portfolio mix. In general, MBS, agency, and other debt securities offer more attractive yields than UST instruments, reflecting in part the potentially higher credit and interest rate risk of these investments. In addition, secondary market developments throughout the 1990s and considerable mortgage refinancing activity over the past three years accelerated both MBS and agency debt issuance. According to the Bond Market Association, agency and MBS debt issuance outpaced UST issuance over the past several years.6 By 1999, the volume of outstanding MBS instruments exceeded outstanding UST obligations (see Chart 2).
Chart 2
[D]
The Shift Away from USTs May Have Increased the Risk Profile of Bank Securities Portfolios
MBS and agency bonds pose unique interest rate risks compared with UST securities, because most MBS and many agency securities contain embedded options. MBS have option risk, as mortgage borrowers have the right to prepay their loans. During periods of refinancing activity induced by low interest rates, MBS holders tend to receive cash flows earlier than originally expected, forcing them to reinvest proceeds at the prevailing lower interest rates. Conversely, MBS holders may face extension risk when prepayments fall because of rising rates and the expected life of the investment increases. In other words, with many MBS investments, the investor receives money faster when reinvestment options are less desirable and more slowly when similar but higher yielding securities are available.
Similarly, agency bonds can have option risk, because they often are callable or "structured." For Call Report purposes, structured notes include "debt securities whose cash flow characteristics (coupon rate, redemption amount, or stated maturity) depend upon one or more indices and/or that have embedded forwards or options."7 Structured notes have never been a large portion of community bank securities portfolios; however, they have come in and out of favor as investment options as interest rates and yield-curve steepness have changed (see Chart 3).8
Chart 3
[D]
Structured notes can be appropriate investment vehicles, and not all structured notes carry the same degree of risk. Many banks were drawn to structured notes in the early 1990s, because they were issued by government-sponsored enterprises and had attractive yields compared with those of other agency bonds or notes. Typically, initial yields were high, but the embedded options were very difficult to price. Some institutions found themselves with highly depreciated, low-paying investments when interest rates moved higher. In fact, on a median basis, banks that held structured notes reported net unrealized losses on these instruments between 1995 and 2000 and registered only mild appreciation throughout 2001, 2002, and 2003.9
While some structured notes have very little credit risk, virtually all have characteristics that require prepurchase scrutiny and ongoing assessment of their sensitivity to interest rate movements. Data on structured note issuance as well as anecdotal reports suggest that these instruments currently tend to be relatively straightforward step-up bonds; in the early 1990s, they were more likely to feature exotic derivative aspects.10 Managers of insured institutions should understand the unique characteristics of these instruments and how they might fit into the bank's overall strategies for investment and interest rate risk management. Because of the embedded options, structured notes need to be monitored closely during the holding period.11
Low Interest Rates Hampered Bond Yields and Boosted Market Values
Declining interest rates pushed the median 2003 year-end yield on securities among insured commercial banks to 3.73 percent, down steadily from 6.13 percent in 2000 (see Chart 4). Given their positive correlation with changes in interest rates, investment portfolio yields likely will improve should rates rise prospectively, as cash flows are invested at progressively higher rates. This correlation was demonstrated when rising interest rates in the 1994 to 1995 and 1999 to 2000 periods provided a temporary lift to bond yields.
Chart 4
[D]
During a period of increasing interest rates, however, rising securities yields may not offset declines in bond values, especially as there is typically a lag in the repricing dates for securities. For instance, indices compiled by Merrill Lynch on total bond returns, which include recurring yield income as well as price changes, suggest that interest rate increases during 1994 to 1995 and 1999 to 2000 created negative year-over-year total returns in many bond classes, because declines in bond values outweighed yield increases.
In addition to boosting earnings, securities appreciation can serve as a cushion to capital and liquidity in the form of unrealized gains. Financial statements prepared in accordance with generally accepted accounting principles (GAAP) adjust asset and capital balances for unrealized gains and losses on available-for-sale debt securities.12 Although regulatory capital standards do not include such adjustments, the agencies recognize that large unrealized losses may impair earnings in the event securities have to be sold to meet liquidity needs. Similarly, unrealized losses may limit liquidity options, as bank management may decide not to sell investments because of the potential effect on earnings and capital.
Unrealized gains and losses on bank balance sheets have fluctuated over time, depending on the level and direction of interest rates. Rising interest rates during periods in the early and late 1990s drove down bond values and caused net portfolio depreciation. After periods of rising interest rates at the end of 1994 and 1999, the median net unrealized loss-to-amortized-cost ratios among insured banks were negative 3.14 percent and negative 1.90 percent, respectively. However, declining interest rates over the past few years enabled many insured banks to augment earnings by selling higher yielding securities for gains. Whereas only 17 percent of insured banks reported gains on the sale of securities during 2000 (a period of rising interest rates—see Chart 5), nearly half of insured banks realized securities gains during 2001, 2002, and 2003.
Chart 5
[D]
How Will Increasing Interest Rates Affect Securities Portfolios?
Changes in portfolio mix have contributed to a general lengthening of investment maturities, which implies that investment portfolios may have greater price risk. As of year-end 2003, debt securities with a next-earliest repricing, maturity, or estimated average life of more than three years typically accounted for 59 percent of securities held by insured banks, up from 41 percent in 1997.13 Most of the lengthening occurred in the pass-through MBS segment, which often accounts for a large share of total securities (see Chart 6).
Chart 6
[D]
While Chart 6 suggests that the estimated average life of non-pass-through MBS declined over a six-year period, this may not indicate declining interest rate risk in this category. For instance, estimated average lives among non-pass-through MBS appear to have lengthened during 1999—a period of rising interest rates—but shortened as interest rates declined during subsequent years. Although the specific types of MBS held in this category cannot be known with certainty on the basis of Call Report data alone, this alternating pattern of extension and contraction in estimated average life may be symptomatic of heightened interest rate and prepayment risk.14 Thus, the estimated lives of these instruments could lengthen quickly should interest rates rise sharply. Understanding the maturity or duration profile of an investment product is important, because—given an equal change in short-term and long-term interest rates (that is, a parallel yield curve shift)—bonds with longer maturities may exhibit larger percentage-point price swings.
Not surprisingly, as long-term rates rose in the second half of 2003, aggregate securities gains and portfolio appreciation decelerated among the nation's banks (see Chart 7).15 In the future, the effects of rising interest rates on bond valuations will ultimately depend on how quickly rates increase and the shape of a post-shift yield curve (for example, flat versus steep).
Chart 7
[D]
If both short- and long-term rates change by equal amounts, prices for longer duration bonds likely would decline by a greater magnitude than prices for shorter duration investments. However, if the yield curve flattens because short-term rates rise faster than long-term rates, prices for bonds with shorter durations could suffer disproportionately.
For instance, interest rates increased sharply and in a parallel fashion during 1994, maintaining a very steep yield curve. These conditions triggered relatively large amounts of unrealized losses in bank bond portfolios, in particular for agency and MBS securities (see Chart 8). However, when both long- and short-term rates increased during 1999, the rate rise was less severe and the yield curve was flatter than that in 1994. As a result, although the bonds depreciated in value, the correction was less pronounced.
Chart 8
[D]
Should interest rates move up sharply, institutions with portfolios of longer-term assets might be faced with the prospect of holding many securities to maturity at below-market interest rates or realizing losses to reinvest in new assets with higher yields. Most institutions may be able to hold depressed securities until maturity and technically never realize a loss, but the trade-off is lower yields over time.
Banks Are in the Business of Managing Risk
Banks have been investing in securities for many years, and individual bank managements adopt policies and strategies to manage their portfolios under various scenarios, balancing loan demand, liquidity needs, and the effects of interest rates on yields and values. Some banks use the securities portfolio as a hedge to reduce credit risk or interest rate risk elsewhere on their balance sheets. Bank investment policies and strategies are typically dynamic and unique to individual institutions and, as a result, are a key area of review for bank directors, auditors, and examiners.
Innovation in the capital markets results in the introduction of new types of investment vehicles, which can make investment selection more difficult and challenge efforts to model the potential effects of changing interest rates on investment holdings. In this environment, it is incumbent upon bank managers and directors to understand fully the unique characteristics of each investment they own. As noted in the Interagency Supervisory Policy Statement on Investment Securities and End-User Derivatives Activities, risk limits associated with capital market activities should be consistent with a bank's strategic plans and overall asset/liability management objectives. Policies should seek to manage market, credit, liquidity, and interest rate risks. In addition, insured institutions that have increased their holdings of interest-sensitive investments and have purchased investments with extended maturities or repricing intervals must ensure that these holdings fit into their overall asset and liability management strategies.
Judy Plock, Senior Financial Analyst,
San Francisco Region
Mike Anas, Senior Financial Analyst, Chicago Region
David Van Vickle, Regional Manager, Chicago Region
1 To view the fourth quarter 2003 and other editions of the FDIC Quarterly Banking Profile, go to http://www2.fdic.gov/qbp/qbpSelect.asp?menuItem=QBP.
2 Data used for this article came from Call Reports filed by commercial banks. Because of differences in data availability, information from Thrift Financial Report filers was not used.
3 U.S. Agency securities include direct debt issued by government-sponsored enterprises (GSEs) such as Fannie Mae, Freddie Mac, and Ginnie Mae; MBS include bonds backed by single-family mortgages issued through private parties or GSEs.
4 Other debt instruments include asset-backed securities, trust preferred securities, and foreign government bonds.
5 In a pass-through MBS structure, principal, interest, and prepayments made on the underlying pool of mortgages are passed through to the ultimate certificate holders. In contrast, investors in non-pass-through MBS such as collateralized mortgage obligations and real estate mortgage investment conduits receive cash flows structured differently from the payments on the underlying mortgages.
6 Data on the composition of bond markets are from the Bond Market Association website at http://www.bondmarkets.com/Research/ osdebt.shtml.
7 For additional information, see the instructions for the preparation of schedule RC-B of the Consolidated Reports of Condition and Income (http://www.ffiec.gov/ffiec_report_forms.htm), which define structured notes to include step-up bonds, index amortizing notes, dual index notes, deleveraged bonds, range bonds, and inverse floaters.
8 Generally, the federal agencies that issue most structured notes (Federal Home Loan Bank, Fannie Mae, and Freddie Mac) do so when interest rates are low and the yield curve is steep.
9 Because Call Report information on structured notes was not collected until 1995, data are not available for 1994 and earlier.
10 Mauro, Martin J., and Michele Chesnicka, Merrill Lynch Fixed Income Digest, January 8, 2004, p. 9.
11 For additional information on risks posed by structured notes, refer to the 1998 Interagency Supervisory Policy Statement on Investment Securities and End-User Derivatives Activities, http://www.fdic.gov/ regulations/laws/rules/5000-4400.html#5000supervisoryps.
12 For additional details on GAAP treatment of held-to-maturity, available-for-sale, and trading securities, refer to Financial Accounting Standard Number 115. As of December 31, 2003, banks classified most securities as available-for-sale and reported associated unrealized gains in GAAP capital accounts.
13 Per Call Report instructions, banks report non-pass-through MBS according to estimated average life. For other categories of debt securities, banks report by earliest repricing or maturity date. The estimated average life calculation considers expected prepayments and is dollar- and time-weighted. As a result, it is not equivalent to contractual maturity or expected final maturity.
14 Because non-pass-through MBS are issued in tranches that differ in terms of priority for receiving principal and interest payments on the underlying pool of mortgages, some classes may have relatively more or less exposure to prepayment risk. Continuous declines in mortgage interest rates during 2001, 2002, and parts of 2003 triggered so much refinancing activity that some tranches that were contractually last in line to receive principal cash flows ended up receiving payments earlier than expected.
15 For a discussion of the elements that contributed to interest rate volatility during 2003, see "Causes and Implications of Recent Interest Rate Volatility," FDIC Outlook, Winter 2003 (http://www.fdic.gov/bank/ analytical/regional/ro20034q/na/infocus.html).
Implications of Rural Depopulation in the Great Plains for Community Banks
The United States is currently undergoing a major demographic event: the depopulation of a significant number of rural counties. This subject has been under review by the FDIC for a number of years. For example, an FDIC report examined rural depopulation trends in the Kansas City Region and concluded that while depopulation is a slow-moving event, it does have an effect on the economic viability of counties experiencing out-migration and on the banks operating in those counties.1 In particular, that report found that lack of growth was the most prominent negative factor affecting community banks in counties with declining populations. These banks reported lower growth rates for assets, loans, deposits, and core deposits than banks in growing rural counties.
This article again examines performance trends of community banks located in depopulating counties.2 However, it also attempts to identify strategies that some banks have employed to remain successful, despite the unfavorable demographic trends unfolding around them. The article is an excerpt of an expansive analysis of rural depopulation trends in the United States and rural bank performance that was released on May 18, 2004, as part of the FDIC's Future of Banking in America (www.fdic.gov/bank/analytical/future).
Depopulation Trends Are Most Pronounced in the Great Plains
The analysis in this report employs a method developed by the FDIC in which counties are divided into categories depending on their rurality and on population trends between 1970 and 2000 (see Map 1). Rural counties that added population over the 30-year span are called "growing counties"; rural counties that lost population at a relatively constant rate are called "declining counties"; and rural counties that not only lost population but saw the rate of loss increase in the 1990s are called "accelerated declining counties." Metropolitan counties, which almost universally added population, were not analyzed in this report.
Map 1
[D]
As Map 1 shows, rural depopulation is most prevalent in the middle of the country but can also be seen in the South and Northeast. Our analysis in this article focuses on the Great Plains region (outlined in Map 1), because the problem of rural depopulation is far more advanced there than anywhere else in the country. In fact, of the 424 rural counties in the Great Plains region, 304 (72 percent) are either declining or accelerated declining counties.
The Great Plains region is also striking from a banking perspective. As of year-end 2003, either declining or accelerated declining counties were home to more than 500 community banks—more than half of all community banks in the Great Plains. In addition, banks in the Great Plains tend to be much smaller than banks located elsewhere. The median size of a bank in the Great Plains is just $56 million, and only about $39 million in rural counties with declining populations. Institutions in other depopulating areas are significantly larger—even the Corn Belt's median bank has $89 million in assets—reflecting the fact that although other regions may also be experiencing depopulation, their financial institutions have much larger beginning customer bases.
Rurality Affects Growth Rates, but Not Performance Measures
Despite the demographic challenges that face the Great Plains, rural community banks headquartered there report performance measures that are in line with community banks located elsewhere. As Table 1 indicates, measures related to earnings and asset quality are very similar, and Great Plains community banks have considerably higher levels of equity capital. The most significant difference between the groups of institutions is the level of farm loans. Not surprisingly, community banks in the rural Great Plains have a far higher concentration of farm loans than do community banks in other rural areas. This leaves Great Plains' financial institutions much more dependent on federal farm policy and vulnerable to swings in net farm income caused by commodity price fluctuations, persistent drought conditions, and unexpected impacts, such as the "mad cow" discovery that led to a steep drop in cattle prices in early 2004. A challenge rural bank managers continuously confront is that many rural farm banks have few local options with which to diversify their loan portfolios.
Table 1
|
Great Plains Rural Community Banks Perform Similarly to Those in the Rest of the Nation (%)
|
| |
2003 |
2002 |
2001 |
2000 |
1999 |
| GP—Pretax ROA |
1.44 |
1.49 |
1.42 |
1.59 |
1.55 |
| Nation—Pretax ROA |
1.44 |
1.51 |
1.39 |
1.50 |
1.54 |
| GP—Net Interest Margin |
4.12 |
4.25 |
4.17 |
4.34 |
4.24 |
| Nation—Net Interest Margin |
4.05 |
4.24 |
4.08 |
4.24 |
4.23 |
| GP—Loans-to-Assets Ratio |
58.51 |
59.59 |
58.92 |
59.25 |
57.45 |
| Nation—Loans-to-Assets Ratio |
61.94 |
62.39 |
63.02 |
64.52 |
63.04 |
| GP—Total PD Loan Ratio |
2.59 |
2.89 |
2.86 |
2.53 |
2.50 |
| Nation—Total PD Loan Ratio |
2.59 |
2.82 |
2.92 |
2.62 |
2.29 |
| GP—Net Charged-Off Loans |
0.31 |
0.34 |
0.46 |
0.30 |
0.30 |
| Nation—Net Charged-Off Loans |
0.30 |
0.33 |
0.31 |
0.23 |
0.22 |
| GP—Equity Capital |
10.97 |
11.19 |
10.95 |
10.81 |
10.16 |
| Nation—Equity Capital |
10.52 |
10.59 |
10.25 |
10.34 |
10.05 |
| GP—Ag Loans/Total Loans |
40.33 |
40.68 |
40.84 |
40.35 |
40.81 |
| Nation—Ag Loans/Total Loans |
13.76 |
13.68 |
13.27 |
13.22 |
13.42 |
| GP—Ag Inst./Total Inst. |
79.97 |
80.08 |
80.44 |
81.22 |
82.21 |
| Nation—Ag Inst./Total Inst. |
28.46 |
28.55 |
28.07 |
28.62 |
29.03 |
Notes: "GP" refers to banks and thrifts with less than $250 million in assets in rural counties in the Great Plains. "Nation" refers to banks and thrifts with less than $250 million in assets in ruralcounties in the United States, excluding the Great Plains.
Source: Bank and Thrift Call Reports. |
In comparing community bank performance in the rural Great Plains, it is interesting to note that institutions in growing, declining, and accelerated declining counties perform similarly. Earnings measures are generally satisfactory regardless of the institution's location, although institutions in growing counties have earned a bit more pretax revenue, largely through higher sources of noninterest income. Net interest margins (NIMs) are similar, as declining and accelerated declining county banks have offset lower loan yields with lower funding costs. Loan quality measures tend to favor growing county institutions moderately, but institutions in declining and accelerated declining counties offset this difference with higher levels of equity capital.
Similarly to national comparisons, differences in farm loan levels exist within the Great Plains region. Growing county community banks have about 30 percent of all loans invested in farm loans, while community banks in depopulating areas average just under 50 percent. Growing counties, which likely are adding to their populations through nonagricultural job growth, tend to offer community banks more diversified lending opportunities.
Growth rates clearly show that depopulation trends have adversely affected community banks. Because declining populations translate into dwindling bases of potential borrowers and depositors, growth rates for total assets, loans, and deposits for community banks in declining and accelerated declining counties have been lower than the corresponding growth rates in growing counties' financial institutions. Table 2 shows a ten-year trend of annualized growth rates for balance sheet accounts. The most striking point in the table is the difference in the Great Plains region between metropolitan community banks and those in rural areas. Across the board, the economic vibrancy of metropolitan areas has translated into higher growth rates. In rural areas, community banks in growing counties were able to increase assets, loans, and deposit accounts at a significantly higher rate than declining or accelerated declining institutions.
Table 2
|
Great Plains Metro Community Banks Have Grown Balance Sheets Far More Quickly than Rural Banks)
|
| Annualized Growth Rate (%) between Year-End 1993 and Year-End 2003 |
| Great Plains County Type |
Total Assets |
Total Loans |
Total Deposits |
Core Deposits |
| Metropolitan |
8.87 |
11.16 |
8.61 |
7.87 |
| Rural |
4.37 |
6.77 |
3.84 |
3.04 |
| Rural County Breakdown: |
| Growing |
4.78 |
6.96 |
4.28 |
3.47 |
| Declining |
4.04 |
6.32 |
3.45 |
2.64 |
| Accelerated Declining |
4.10 |
7.16 |
3.61 |
2.84 |
Demographic Problems Have Not Yet Accelerated Consolidation Trends
The number of insured banks and thrifts has been declining in the United States for more than two decades.3 Because of the large number of depopulating rural counties in the Great Plains region, one might expect that bank consolidation would have been more severe in that region. However, reductions in the number of banks in the Great Plains are similar to those in rural areas in the rest of the nation (see Chart 1). Perhaps surprisingly, the reduction in insured institutions is consistent among all three types (growing, declining, and accelerated declining) of Great Plains rural counties.
Chart 1
[D]
Although consolidation trends in Great Plains rural community banks have been consistent with national figures, two trends suggest that consolidation in the Great Plains may increase in the future. First, the elderly population in depopulating counties is very large. At some point in the relatively near future, these people are going to begin to pass away, taking a disproportionate amount of banking business with them. Second, many rural community banks in the Great Plains may lack adequate succession plans. In many cases, when the owner/operators of these institutions retire, no family members are ready to take their places, because the younger relatives have long since migrated to counties with more economic opportunities. And because of the shortage of young professionals in rural areas, no qualified nonfamily members may be available to take over the operations. In such cases, owner/operators may simply continue working well into their retirement-age years. When these bankers finally do retire, their institutions may be sold, which could increase the pace of rural bank consolidation.
What Strategies Can Help Rural Community Banks Remain Successful?
While many counties in the Great Plains face similar economic issues, community banks in the region have responded differently and reported disparate operating results. Two metrics—profitability and growth—were used to try to identify common strategies employed by the more successful rural banks. Most analysts would agree that profitability is an appropriate measure of success. For the purposes of this analysis, profitability is measured by the five-year pretax return-on-assets (ROA) ratio.4 Asset growth is also often used as an indicator of success, although some banks can achieve success in other metrics (such as profitability) without significant growth. For this analysis, growth is measured by the five-year annualized asset growth rate, adjusted to negate the effects of mergers.
Among the 483 banks studied, profitability and growth performance differed significantly from bank to bank. Annualized profitability ranged from a low of -1.07 percent to a high of 3.53 percent, with the middle 80 percent of banks in the range of 0.62 percent to 2.10 percent. Annualized asset growth ranged from -11.71 percent to 79.65 percent, with the middle 80 percent of banks falling between -0.51 percent and 9.04 percent.
To analyze relatively high- and low-performing institutions, each of the two metrics was divided into thirds, creating a nine-cell matrix (see Table 3). The corner cells contain summary analyses for the 203 banks at the upper and lower levels of both performance metrics. The other 280 institutions, which fall into the middle range, are lumped into a single unit, the Middle Cross, and used as a control group for the analysis.
Table 3
[D]
Notes:
1. Asset growth figures are merger-adjusted, asset-weighted, annualized five-year growth rates.
2. Pretax ROA figures are for merger-adjusted, asset-weighted, annualized five-year pretax return-on-asset performance.
3. A total of 483 institutions analyzed above met the following descriptives:
a. as of December 31, 2003, had total assets of $250 million or less;
b. were established on December 31, 1993, or earlier;
c. were headquartered in rural counties within the Great Plains region with either a declining population or an accelerated declining population.
Source: Bank and Thrift Call Reports.
|
In looking at the data for the corner banks, one might ask, for example, about the successful business strategies of the 49 community banks in the upper right-hand corner (those that reported high asset growth and high profitability). By contrast, why do the 61 institutions in the lower left-hand corner report both low growth and low profitability? The other corners indicate institutions that were able to achieve high profits despite low growth and those that reported high growth but low profits.
Briefly, here are the reasons for the widely disparate performances of community banks in the corners of the matrix:
Low-growth/high-earnings banks. These institutions have maintained a high level of profitability in the absence of asset growth by controlling operating costs extremely well. Seventy percent of these banks operate a single (albeit likely large) banking office, which helps keep costs down. Lending activity and capital levels also suggest effective management strategies.
High-growth/high-earnings banks. These banks tend to be larger (as Table 3 shows, the median bank in this group is the largest in the matrix), allowing them to control operating costs through scale efficiencies. They achieved the highest NIMs by maintaining reasonable funding costs and maximizing loan volumes in relation to total assets. Many banks in this group have branches in metropolitan areas or growing rural counties, which helps explain why they have been able to achieve higher than average asset growth.
Low-growth/low-earnings banks. These institutions are the smallest, with a median asset size of just $21.5 million. They report by far the lowest NIMs of any group, have not controlled costs well, and have significantly higher levels of past-due loans than the other groups. Nearly two-thirds of these institutions operate a single banking office.
High-growth/low-earnings banks. Like high-growth/high-earnings banks, these institutions have aggressively pursued growth through branching activities but without the earnings success. These banks report lower than average NIMs coupled with the highest operating costs of any group.
One strategy bankers are pursuing is branching activity. Because community banks in depopulating counties have declining customer bases, many institutions have turned to branching in more economically vibrant areas to attract new loan and deposit customers. Many institutions that have achieved high asset growth have adopted this strategy (see Table 4).
Table 4
|
