Enter Search TextSubmitSearch Advanced Search     Home     Deposit Insurance     Consumer Protection     Industry Analysis     Regulations & Examinations     Asset Sales     News & Events     About FDIC

		Home > Industry Analysis > Research & Analysis > Atlanta Regional Outlook - Second Quarter 1998

Atlanta Regional Outlook - Second Quarter 1998

Regular Features Current Regional Banking Conditions Commercial banks performed well in the fourth quarter, despite margin compression and merger-related charges. Thrift performance, also affected by lower margins and higher overhead, was weaker in the quarter. Growth in net interest income (NII) in 1997 was driven by growth in earning assets, while interest rate movements actually constrained NII growth. Assessing the degree of interest rate risk at insured institutions becomes more important because of recent flattening in the yield curve. Insured institutions headquartered in areas where textile and apparel manufacturing employment is high have performed well since the last recession. Fourth-Quarter Regional Bank and Thrift Performance Is Good Atlanta Region banks performed well in the fourth quarter of 1997, although aggregate earnings were constrained by merger-related charges at some very large banks, as well as by margin compression that affected banks of all sizes. The Region reported an annualized fourth-quarter return on assets (ROA) of 1.01 percent, 31 basis points below the previous quarter and 25 basis points below the fourth-quarter 1996 return. Higher noninterest charges, particularly at four large banks in Virginia and Florida that were involved in pending mergers, contributed to a sharp increase in the overhead expense ratio during the quarter. The net interest margin fell 13 basis points to 4.29 percent in the period because of higher funding costs but remained above the national average. A modest increase in reserve provisions and a slight drop in noninterest income also affected quarterly earnings. The Region's aggregate leverage capital ratio fell slightly in the quarter, but overall capitalization remained strong. Delinquency, charge-off, and reserve coverage measures all indicated continued strong regional banking conditions. There was further deterioration in the consumer credit sector, however. The Region's credit card loss rate increased throughout the year and was above the national average at 5.54 percent during the fourth quarter. In general, large banks performed better than small ones. The weakness in small-bank earnings was largely seasonal, as overhead and provision expenses for this group tend to be highest in the fourth quarter. Atlanta Region thrifts underperformed their out-of-region peers during the quarter with an annualized ROA of 0.77 percent. Lower long-term interest rates reduced thrift net interest margins throughout the second half of 1997, and provision expenses and overhead spending were higher in the fourth quarter. The Region's thrifts remained very well capitalized, however, with a year-end aggregate leverage ratio above 9 percent. Delinquencies and charge-offs also remained low, and reserve coverage of nonperforming loans, already above the nationwide thrift average, improved during the quarter. Interest Rate Risk for Insured Institutions Could Increase Several forces currently at work in the general economy and the financial services sector could portend higher interest rate risk (IRR) for insured institutions. A flattened yield curve, record refinancing activity, potential devaluation of mortgage servicing assets, a shift in loan demand from adjustable-rate to low-fixed-rate products, and intense competition among financial institutions and nonbank lenders all could affect intermediated earnings. Change in these areas places increased emphasis on identifying the extent of IRR in insured institutions. IRR can be measured in several ways. Common methods include gap, duration, and simulation analysis, all of which are prospective measurement techniques designed to estimate net interest income (NII) variability or financial instrument price volatility in future periods. Measuring IRR is extremely difficult regardless of the technique used. Typically, most IRR models require estimates of, among other things, the direction of change for several key interest rates over time, the magnitude and timing of those changes, and the average volume and mix of earning assets and interest-bearing liabilities at the time of each change in rates. Embedded optionality in many financial instruments and the increased use of off-balance sheet derivatives and hedging programs add to the complexity of forecasting IRR. It is even more difficult to apply these forward-looking techniques off site, as the required inputs are not sufficiently detailed in Bank or Thrift Call Reports or other public filings. As discussed below, rate/volume analysis (see Table 1) is one alternative ex post method that can be applied off site to measure IRR exposure during a specified period. Table 1 Rate/Volume Analysis Methodology Variance Measures Formula Components Income Rate Variance (1997 earning asset yield - 1996 earning asset yield) x 1996 average earning assets Expense Rate Variance (1997 cost of funds - 1996 cost of funds) x 1996 average interest-bearing liabilities Income Volume Variance (1997 average earning assets - 1996 average earning assets) x 1996 earning asset yield Expense Volume Variance (1997 average interest-bearing liabilities - 1996 average interest-bearing liabilities) x 1996 cost of funds Income Mix Variance (1997 average earning assets - 1996 average earning assets) x (1997 yield - 1996 yield) Expense Mix Variance (1997 average interest-bearing liabilities - 1996 average interest-bearing liabilities) x (1997 cost - 1996 cost) The "net" position for each of the variance measures is the difference between the income and expense variances. For example, the net rate variance is equal to income rate variance - expense rate variance. The sum of the three net variance measures (rate, volume, and mix) should equal the total change in net interest income during the period. Source: FDIC Interest Rate Risk Model Table 1 details how rate/volume analysis decomposes a financial intermediary's period change in NII. NII changes can be attributed to three factors: changes in yields and costs (rate variance), fluctuations in earning asset and interest-bearing liability volume (volume variance), and residual interest income and expense arising from the combination of rate and volume changes (mix variance). Conceptually, rate/volume analysis simply breaks down NII into its component parts and measures the contribution of each component during the period. This method is particularly suitable for off-site monitoring, since the required data are readily obtainable from Call Reports. Note that Table 1 assumes 1997 as the measurement period, but a rate/volume analysis can be applied to any two consecutive time periods. Subjecting Atlanta Region commercial banks and FDIC-supervised savings banks1 to a rate/volume analysis for 1997, we found that the increase in aggregate NII was driven entirely by strong growth in earning assets relative to interest-bearing liabilities, while rate and mix variations actually constrained NII growth during the year. Table 2 summarizes the effects of rate, volume, and mix variance on 1997 aggregate NII. The negative income rate variance and positive expense rate variance reflect the flattening of the yield curve (lower long-term rates and higher short-term rates) during the period. The negative net mix variance also reflects this flattening, as new earning assets yielded progressively less (long-term rates were falling), while marginal funding costs were higher (short-term rates were rising). 1 Credit card banks are excluded because of their nontraditional balance sheet structure and institutions involved in a merger in 1996 or 1997 because merger accounting can skew average yield and cost data. The universe included 1,055 institutions in 1997. Table 2 1997 Rate/Volume Analysis of Atlanta Region Institutions* Variable Year Total Change Change Due to 1997 1996 Rate Volume Mix Interest Income         (tax equivalent) 14,091,951 12,822,187 1,269,764 (76,970) 1,351,553 (4,819) Interest Expense   6,296,444   5,628,857     667,587     30,441       619,003   18,143   Net Interest Income 7,795,507 7,193,330 602,177 (107,411) 732,550 (22,962) Average Earning         Assets 169,604,372 153,414,712 16,189,660 Average Interest-         Bearing Liabilities 138,755,506 125,158,551 13,596,955 * Insured institutions include all commercial banks and FDIC-supervised savings banks.Institutions involved in mergers in 1996 or 1997 are excluded. Credit card institutions are excluded. Source: Bank and Thrift Call Reports For comparison, we conducted a rate/volume analysis on a similar universe2 of institutions using 1994 data, the latest year in which there was notable volatility in market interest rates. As in 1997, net earning asset growth accounted for the increase in NII in 1994. The net rate variance (NRV) again was negative, although the rate effect was much larger in 1994 than in 1997, implying higher IRR exposure. Mix variance was slightly negative in 1994. 2 The 1994 data set included 1,040 commercial banks and FDIC-supervised savings banks. Rate/volume analysis indicates that, on average, commercial banks and savings banks exhibited greater sensitivity to interest rate changes in 1994 than in 1997 (measured by NRV as a percentage of prior period NII). The interest rate environments that existed during each of those periods explain this finding. Specifically, in 1994 the Federal Reserve began pushing short-term (federal funds) rates higher very early in the year, so the increased funding costs affected NII throughout most of the measurement period. In 1997, however, the bulk of the decline in long-term rates occurred later in the year, so the detrimental impact on asset yields was felt for a shorter period. Although NRV was negative in both periods, the differing manner in which the yield curve flattened in each period affected how the negative NRV occurred. For instance, 72 percent of the negative NRV in 1997 was due to lower asset yields (declining long-term rates), whereas nearly all (94 percent) of the negative NRV in 1994 resulted from higher funding costs (rising short-term rates). If we use capitalization, fee income contribution, and the net interest spread as indicators of IRR tolerance, it appears that overall risk-bearing capacity was higher in 1997 than in 1994. The aggregate leverage capital ratio for the previously defined institution universe was 9.26 percent in 1997, up from 8.72 percent in 1994. In addition, institutions had a more diverse income stream in 1997, which reduced reliance on NII to maintain profit stability. For this analysis, reliance on NII was measured by the ratio of noninterest income to total income, where total income was equal to noninterest income plus net interest income. A higher ratio indicates greater diversification of the income stream and, consequently, a higher capacity to withstand negative fluctuations in NII. Noninterest income accounted for 26.3 percent of aggregate total income in 1997, a modest increase from the 1994 level of 25.4 percent. Finally, a comparison of aggregate net interest spreads showed a slight (2 basis points) contraction from 1994 to 1997-- not enough to offset the favorable effects of higher capital and greater income diversification. The IRR exposure of savings banks was higher, on average, than that of commercial banks in both 1994 and 1997 as measured by the ratio of NRV to the prior year's NII, and their overall risk-bearing capacity was comparatively lower. Savings banks' risk tolerance benefited from higher capital levels in both periods; however, these institutions exhibited significantly lower net interest spreads and had very little income source diversification. In 1997, the savings banks in the data set had an aggregate leverage capital ratio of 13.47 percent, which was well above the commercial bank ratio of 9.17 percent, but noninterest income accounted for only 9.34 percent of savings banks' total income during the year, versus 26.6 percent for commercial banks. Also, savings banks' 1997 net interest spread of 3.02 percent was 77 basis points below that of commercial banks. Given lower net interest spreads and considerably higher mortgage loan concentrations, savings banks may be more vulnerable to an increase in refinancing. As of year-end 1997, residential mortgage loans comprised nearly 84 percent of total loans for savings banks in the data set, compared with 32 percent for commercial banks. Similar results were evident in the 1994 data. The referenced universe of insured institutions did not include traditional thrifts, as average balance and tax-equivalency data were not readily available. It is reasonable to conclude, however, that they would exhibit an IRR posture similar to that of the savings banks in the data set. Although aggregate IRR exposure as measured by rate/volume analysis was lower in 1997 than in 1994 for the Region's commercial and savings banks, examiners and institution managers should be aware that the ongoing yield curve flattening that began in the second half of 1997 could result in higher IRR in 1998. Insured Institutions Have Performed Well despite Declines in Textile and Apparel Manufacturing The secular decline in textile and apparel manufacturing has a disproportionate impact on the Atlanta Region (see Regional Economy in this issue). While the amount of direct lending by insured institutions to textile manufacturers is not extensive, many local economies in the Region are dependent on this industry sector. Hence, further plant closings or reductions in employment could adversely affect some insured institutions. Currently, there are 29 rural counties in the Region where textile and apparel employment represents 20 percent or more of total employment. Headquartered in these counties are 53 insured institutions (see Chart 1), which are concentrated mostly in three states: Georgia (20), Alabama (13), and North Carolina (10). Most are commercial banks, but six are traditional thrifts and three are FDIC-supervised savings banks. The performance of these insured institutions has improved since the 1991 recession, as shown in Table 3. Significant improvements have occurred in the weighted-average capital ratio and in noncurrent loans and charge-offs. However, the improvement in noncurrent loans and charge-offs appears to have reached a cyclical low in the past year. In general, the performance of commercial and industrial (C&I) loans has been weaker than the overall loan portfolio. Extensions of credit to textile and apparel manufacturers are reported as C&I loans. Over the past two years, C&I loan growth has been extremely robust, leading to an increase in this type of lending relative to capital and a notable increase in the loan-to-deposit ratio. Because this lending segment performed poorly at these insured institutions during the last economic downturn, attention to the ramifications of a future downturn on loan quality is warranted. Chart 1 Table 3 Performance Trends at Institutions Headquartered in Rural Counties with High Textile and Apparel Manufacturing Employment* Year Number of Institutions Average Size ($ Millions) Capital Ratio C&I Loans to Capital (x) Non- current Loans Noncurrent C&I Loans Loan Charge- Offs C&I Loan Charge- Offs Loans to Deposits C&I Loan Growth Return on Assets 1991 49 78.0 9.17 0.98 1.39 2.52 0.43 1.25 67.9 -- 0.96 1992 49 82.8 9.52 0.89 1.09 1.76 0.44 1.18 64.7 0.74 1.27 1993 49 86.8 10.01 0.77 0.86 1.92 0.14 0.44 64.4 (4.69) 1.31 1994 50 89.0 9.85 0.80 0.75 1.60 0.14 0.20 67.8 6.90 1.17 1995 50 96.5 10.76 0.72 0.57 0.85 0.14 0.44 69.4 3.61 1.22 1996 53 98.4 10.93 0.76 0.60 0.68 0.23 0.39 74.7 13.22 1.17 1997 53 115.1 10.57 0.95 0.59 0.85 0.19 0.38 77.8 28.57 1.29 * Defined as counties where textile and apparel manufacturing employment is 20 percent or more of total employment. Source: Bank and Thrift Call Reports; WEFA, Inc. Jack M.W. Phelps, Regional Manager W. Brian Bowling, Financial Analyst Scott C. Hughes, Regional Economist Pamela R. Stallings, Financial Analyst Regional Outlook Information Return to Regional Outlook main page