ByDon Schlagenhauf , Lowell R. Ricketts
This is the inaugural edition of the full Quarterly Debt Monitor, a detailed report on consumer debt nationally compared to the four largest metropolitan statistical areas (MSAs) in the Eighth Federal Reserve District, which has headquarters in St. Louis.1 Each quarter—for this issue, the first quarter of 2016—this report will examine changes in total consumer debt and in specific types of liabilities: mortgages, home equity lines of credit (HELOC), automobile and student loans, and credit card balances. Our analysis also will offer a look at trends in liabilities from an age group perspective. This demographic lens is important, as different age groups account for different borrowing rates and delinquency rates within debt types.
The analysis is organized into three sections: 1) a comparison of debt trends across the nation and in the largest MSAs in the District; 2) measures of delinquency for the same regions; and 3) a special section offering more in-depth analysis of trends in auto lending. The figures in the first two sections will be presented each quarter, offering readers the ability to monitor developments in these indicators over time. The topic for the final section of this report will change each quarter; in this issue, we highlight auto loan originations by Equifax Risk Scores, the estimated creditworthiness of borrowers.
To monitor developments in consumer debt, it is imperative to examine microeconomic data rather than traditional aggregate data. Thus, this project uses the Consumer Credit Panel dataset assembled by the Federal Reserve Bank of New York. These data are constructed from raw credit report data provided by Equifax. See the methodology section in the appendix for definitions and technical details regarding our analysis.
The table offers a quick summary of important indicators and trends in debt levels and delinquency rates within the past year across geographic areas.2
This issue of the Quarterly Debt Monitor examines data for the first quarter of 2016. Highlights include:
Following a lengthy period of deleveraging, total debt is on the rise across the nation and the District’s four largest MSAs. (See Figure 1.)4 Aside from growth in mortgage lending, borrowing for automobile purchases and for higher education has been a substantial source of strength for lending activity. (See Figure 4.) Louisville had a faster growth rate in total debt than the nation and other MSAs did, almost entirely due to strong mortgage borrowing.
Mortgage debt growth has been modest, but slight increases have had a meaningful impact on overall debt growth. (See Figures 2A and 4.) Across age cohorts, 31- to 40-year-olds and 66- to 75-year-olds acquired the bulk of new mortgage debt. (See Figure 5B.) In contrast, 41- to 55-year-olds lowered their mortgage balances considerably.
Among the MSAs, Louisville had overall mortgage borrowing that outpaced that of the nation and stood in stark contrast to the other MSAs, which had either declining or unchanged mortgage debt. Local contacts in Louisville said that the combination of high demand and low supply in the housing market leads to several bids on homes as soon as they reach the market. This robust activity has led to rising house prices; the CoreLogic Home Price Index in Louisville increased at a rate not seen since before the recession. Mortgage borrowing in Louisville derived unusual strength from millennials (21- to 30-year-olds) and 56- to 65-year-olds, two groups that provided little or even negative growth for the nation and other District MSAs.
The average mortgage balance (see Figure 3.) across the nation was close to $30,000 higher than that of Memphis, the highest balance among MSAs.5 The lower mortgage balances for the MSAs reflect in large part the lower house prices found within the region in contrast to some of the more expensive markets in the country.
Across all areas, levels of auto debt exceed their pre-recession levels. (See Figure 2B.) Individuals aged 31 to 55 took on more than half of the new auto debt nationally. (See Figure 5C.) However, there was also significant borrowing activity at the youngest (1.5 percent, 21 to 30) and oldest (1.2 percent, over 66) age groups. Aside from St. Louis, the 41 to 55 and 66 to 75 age groups in the MSAs outpaced the growth rate nationally. The 31- to 40-year-old group accounted for nearly a third of the auto debt growth within Little Rock, the MSA with the highest rate. Increasing auto sales, rising auto prices, and a greater reliance on financing boosted auto debt. The average balance (see Figure 3) for a borrower was similar across geographic areas, ranging from $12,000 in Louisville and St. Louis to $16,000 in Little Rock.
Credit card debt is the most ubiquitous of all types of debt within our data. In the first quarter of 2016, over half of our sample had a credit card balance. Given that credit card balances are the lowest among debt types (see Figure 3), it isn’t surprising that these liabilities are the most common. Consumers across the nation and MSAs aggressively lowered their credit card debt following the recession. (See Figure 2C.) This deleveraging involved paying down balances and closing accounts. Memphis had the most extensive deleveraging; total credit card debt in the first quarter of 2016 was 63 percent of its 2003 level. (See Figure 5D.) However, within the past year, Memphis credit card debt grew by 6 percent, almost double the rate for the next highest MSA. Interestingly, a third of that growth came from the 41 to 55 age group, the same age group that saw balances decline across all other areas.
Total student debt (Figure 2D) increased for the nation and for the MSAs over the past year, by as much as 4.9 percent in Memphis. It’s important to note that consumers didn’t deleverage their student debt during and after the recession. In fact, student debt is considered to be countercyclical: Consumers may choose to take on more debt to invest in additional education during a recession when the job market is particularly weak. In general, college enrollment rates have been on an upward trend since the mid-1980s. This trend is partly reflected by the share of the population that has student loans. In the nation, the rate jumped from 12 percent to 15 percent between 2007 and 2010. However, this understates the increased investment in higher education during this time, as some students don’t rely on loans. The increase does highlight a greater reliance on student loans to finance higher education. Across the MSAs, an additional 3.6 percent to 5 percent of the population borrowed for education during the same period. The average balance (See Figure 3.) for a student loan borrower ranged from about $27,000 in Louisville to $31,000 in Memphis. Interestingly, most of the growth in overall student debt (see Figure 5E) was concentrated in the 41 to 55 age group. This growth may represent family members co-signing the loans of young borrowers.
Prior to the housing crash, HELOC debt was carried by a small share of individuals. Of the District MSAs, Louisville had the largest share (10 percent), while Little Rock had the smallest (4 percent). Those shares are much lower as of the first quarter; only 6 percent of consumers in Louisville have HELOC loans, and in Little Rock the rate has fallen to 3 percent. The deleveraging period following the recession hasn’t concluded, either. (See Figure 2E.) The four MSAs and the nation registered declines, including a 6 percent decline for both St. Louis and Memphis. (See Figure 5F.) Across age groups and areas, the majority of the decline was concentrated among the 41 to 55 age group. There were some positive outliers among those over 56, but not enough to keep the total amount steady. It is unlikely that HELOC debt will rebound until people in the 41 to 65 age range begin borrowing again, since they carry two-thirds of all HELOC debt nationally.
We define the serious delinquency rate as the share of outstanding debt for which payment is at least 90 days past due.6 Looking at the rate of debt over 90 days delinquent focuses on those instances where borrowers are having significant difficulty in meeting their payment obligations.
Serious delinquency rates for mortgages reflect the severity of the housing crash. (See Figure 6.) The rate for the nation presents a sharp contrast to the MSAs’, as the national rate peaked at 7.7 percent in the first quarter of 2010. Among the District MSAs, Memphis was the hardest hit by the crash, with a peak of 5.8 percent in the same period. The legacy of the crash continues to wane: Serious delinquency rates declined across all areas over the past year. In fact, Memphis and Louisville have reached their pre-recession level. The majority of seriously delinquent mortgage debt remains on the balance sheet of the 41 to 55 age group. (See Figure 7A.) This group represents an even greater share of delinquent mortgage debt in Memphis.
Serious delinquency rates for auto debt are under greater scrutiny given some signs of looser underwriting standards within the market. (See our special section.) While many of the rates remain higher than pre-recession levels, only a few of the areas shown here have sustained increases. Louisville and Little Rock saw their rates increase by 0.5 and 0.6 percentage points, respectively. Despite a decline over the past year, Memphis had a far higher rate of delinquency than any other area featured in this report. In addition to a higher rate of delinquency for everyone under 55, Memphis also had an elevated rate of delinquency among the 56 to 65 age group. (See Figure 7B.) This is unusual, as people in that age group typically have stronger balance sheets, which allow them to better weather financial setbacks.
After borrowers’ concerted deleveraging, the serious delinquency rate for credit card loans has reached historic lows for the MSAs and the nation. The number of credit card accounts has substantially declined, and the remaining card users are apparently better able to manage their repayment obligations. The country and the MSAs—except Little Rock—saw a continued decline in delinquency rates within the past year. (See Figure 6).
With the exception of Memphis, the District MSAs outperformed the nation with lower overall delinquency rates. (See Figure 7C.) Much of the delinquent credit card debt is concentrated in the 41 to 55 age group. However, every age group contributes to the overall rate, even the relatively small group over the age of 75.
The serious delinquency rate for student loans has risen for years. In Little Rock, the rate has tripled, in effect, in the past 13 years. During that time, the serious delinquency rate nationally and in the MSAs has surpassed that of credit cards for the infamous title of worst serious delinquency rate among the debt types considered here. The younger age groups represent a larger share of the overall delinquent debt. (See Figure 7D.) This is especially true for Little Rock, where 21- to 30-year-olds had almost double the rate of delinquency than they had nationally. Making matters worse, the serious delinquency rate likely understates the severity of the problem because many students may be in deferment, forbearance or a grace period before repayment begins. Undoubtedly, some of those borrowers will fall behind on making payments.
During the recession, the national serious delinquency rate for HELOC loans rose like the delinquency rate for mortgages. However, delinquency rates for HELOC debt didn’t peak until close to 2012. Many of the District MSAs, given their relatively insulated housing markets, didn’t see a sustained rise. As conditions have improved, the national rate has shrunk a great deal. In contrast, the delinquency rate in St. Louis after 2010 has continued to increase. (See Figure 6.) The 66 to 75 age group accounts for almost a third of the delinquent debt in St. Louis, more than double the share seen nationally. (See Figure 7E.)
The Federal Reserve Bank of New York/ Equifax Consumer Credit Panel consists of Equifax credit report data for a longitudinal quarterly panel of individuals from 1999 to the present.8 Due to reporting irregularities prior to 2003, we restrict our sample so that it begins in the first quarter of 2003. The panel is a nationally representative 5 percent random sample of all individuals with a Social Security number and a credit report. Thus, our analysis excludes debt held by individuals without Social Security numbers. We use only data for individual consumers, but the sample also includes all other individuals living at the same address as the primary sample members. The resulting database includes a maximum sample of over 44 million individuals in each quarter.9
The Consumer Credit Panel offers a variety of variables for several important consumer debt categories:
To reduce computing requirements, we use a 2 percent subsample for all national estimates and figures. For all regional analysis, we use the entire 5 percent sample in order to maximize the sample size available for smaller geographic areas. While estimating individual total balances, we avoid double counting by attributing 50 percent of the balance associated with each joint account to each individual. Aggregate debt totals are calculated by multiplying estimated figures by 20 (1 /.05) for the regional analysis and by 1,000 (1/ (.05 *.02)) for the nation. All debt figures, with the exception of average debt balance, are adjusted for inflation using the personal consumption expenditures chain-type price index, with a base year of 2009. Analysis of regional MSAs uses the latest definitions provided by the Office of Management and Budget.10
Total Debt Balance: Total balance across all accounts, excluding those in bankruptcy.
Average Balance: All average balances are conditional on carrying that type of debt. For example, average student debt pertains to the average balance among individuals who have a student loan. Also, average debt balances are reported unadjusted for inflation. This was done to make those balances more relatable to a reader’s personal financial obligations.
Equifax Risk Score: Predicts the likelihood of a consumer becoming seriously delinquent (90-plus days past due on debt). The score ranges from 280 to 850, with someone with a higher score being viewed as a better risk than someone with a lower score.
Serious Delinquency Rate: The share of debt that has a payment at least 90 days past due, divided by the total outstanding balance for that type of debt.
Debt at least 90 days past due includes:
The total outstanding balance includes debt at least 90 days past due in addition to:
Not all creditors provide updated information on payment status, especially after accounts have been derogatory for a longer period of time.
Student Loan Reporting Issues:There are some reporting issues related to student loans in the raw Equifax data. This involves a temporary or sustained period of nonreporting of student loan accounts. We leave the data as they were provided by Equifax.
Auto Loan Originations: In calculating the balance of new auto loan originations, we use both the number of accounts and total balance on an individual credit report in consecutive quarters. New auto loan originations are then defined as an increase in the balance accompanied by an increase in the number of accounts reported.