6 posts from April 2013

Driving to the bank: Auto loans rebounding in Ninth District

Maybe people like their cars more than their homes. While the housing market appears to be finally finding its legs, auto sales have been on a tear—especially in the Ninth District.

During and subsequent to the Great Recession, total auto loan debt declined precipitously to its trough around the end of 2011 (see first set of charts). Since then, however, auto loans have been squealing the tires. In the district, inflation-adjusted auto loan debt rose to 98 percent of prerecession levels, while national auto loan balances are at 88 percent, according to data from the Federal Reserve Bank of New York Credit Panel/Equifax.

Auto loan balances CH1

The lending rebound has been shared among various financing options, but banks are in the lead car, particularly in the Ninth District. Federal Reserve Bank of New York Credit Panel/Equifax data show that banks (which include credit unions and savings and loans) hold the majority of vehicle loan debt in the district, and debt balances didn’t dip as far during the recession. Debt balances have since rebounded above prerecession levels, to $9.5 billion (see charts below). At the same time, vehicle loan debt held by finance companies (dealers and auto or sales finance companies, like car makers) plunged during the recession and remains considerably below the prerecession peak.

Auto loan balances -- Banks v. Finance CH2

Nationally, the share is flipped. Finance companies still account for a majority of loans, but the margin has narrowed, in part because finance company loans saw a steeper drop through 2010 compared with banks, and their subsequent growth since 2011 has been very modest.

Performance among Ninth District states has varied—in fact, some states saw inflation-adjusted auto loan balances decline well before the recession. Northwestern Wisconsin has experienced little recovery since the end of the recession; by the end of 2012, real auto loan balances stand at only 90 percent of 2003 levels. At the other end of the spectrum, North Dakota auto loan balances are 40 percent higher over the same period. Other district states lie somewhere in the middle, though Montana did have a notable runup in debt levels prior to the onset of the recession.

Creditworthiness and delinquency also play a big role in the rebound. Vehicle loan balances generally dropped less during the recession and rose more afterward as borrower Equifax Risk Scores rose. So-called super-prime borrowers are responsible for a large percentage of vehicle debt, and they have been taking advantage of their access to credit to take out more vehicle loans given today’s low interest rates. And, again, this trend has been more prevalent in the Ninth District. Ninth District loan delinquency rates in the district also have been well above national rates before, during and after the recession.

Revamping the district stock index

The composition and methodology of the Ninth District Mid-Cap Stock Index changed at the beginning of April 2013.

In terms of composition, four companies were removed from the index as they relocated their headquarters outside the district or their market capitalization exceeded the maximum threshold for the index. Five new companies were added to the index, expanding the diversity of industries represented as well as geographical coverage. Details on these changes and a current list of companies in the index are available here: Ninth District Mid-Cap Stock Index.

In terms of methodology, the index is now more closely aligned with calculations used in the benchmark S&P MidCap 400 index. Conceptually the district index, like other market-capitalization-based indexes, such as the S&P MidCap 400 and S&P 500, represents the total value of underlying companies normalized to a base year. The rate of change in the index therefore represents the rate of change in the total value of companies included. Under the previous methodology, price changes had a stronger weight and tended to overestimate growth.

As shown in the chart below, the Ninth District Mid-Cap Stock Index moves closely with the benchmark S&P Midcap 400 index. District mid-cap companies appear to have weathered the financial crisis better than the national average; the cumulative growth rate since February 2008 was about 8 percentage points higher than the S&P 400 index. For more details on the Ninth District Mid-Cap Stock Index, see the index methodology.

Stock index -- 4-17-13

The economic impact of closing Minnesota's achievement gap: A theoretical construct

An education achievement gap by race and income has long persisted in the nation and in Minnesota. While there is a clear moral argument for closing the gap, there are some compelling economic ones as well.

Differences in high school graduation rates and achievement scores between white students and Native American, black and Hispanic students in Minnesota are some of the largest in the country. The chart below shows a substantial difference in average math scores of white and black eighth grade students since 2003.

Achievement gap charts -- 4-12-13

If test scores of black and Latino students and low-income students could be raised to those of white and higher-income students, presumably graduation rates would increase, as would the overall skills of the workforce, leading to productivity gains and stronger economic growth. But by how much, and what net effect would it have for closing these gaps in Minnesota?

A 2009 McKinsey report, using a methodology developed by Eric Hanushek in a 2008 study in the Journal of Economic Literature, projects that national GDP in 2008 could have been 2 percent to 4 percent higher had the United States bridged the racial achievement gap by raising the performance of black and Latino students to that of white students by 1998 after a successful 15-year reform period. The report estimates that GDP could have been 3 percent to 5 percent higher had the United States closed the income achievement gap by raising the performance of students with household incomes below $25,000 to that of students with higher household incomes.

The same framework discussed in the McKinsey report was applied to Minnesota using National Assessment of Educational Progress (NAEP) data for the last five survey years. Closing the racial achievement gap for eighth grade students in Minnesota would improve the state’s overall average math scores by about 2 percent; closing the income achievement gap would improve average math scores by about 3 percent.

Using Hanushek’s estimate – that long-run GDP growth rate increases by 1.3 percentage points per standard deviation improvement in test scores (about 0.6 percentage points per 10 percent increase in average test scores) – closing the achievement gap in Minnesota would translate into a 0.1 to 0.3 percentage point increase in the long-run economic growth rate.

Even a small change in a growth rate over time adds up. For example, if a hypothetical 15-year reform plan could close the achievement gaps, the level of Minnesota’s GDP would diverge from trend, raising the GDP level by 1 percent or more after 30 years and by more than 3 percent to 6 percent after 50 years (see table below).

Achievement gap table -- 4-10-13

In terms of dollars, these increases translate to a few hundred million dollars per annum after 15 years from the start of the reform period to a couple of billion dollars after 30 years to more than $10 billion after 50 years. In 2011, Minnesota’s real GDP was $282 billion. However, caution should be used with these projections because it’s unclear whether Hanushek’s estimate applies at the state level.

The calculated economic impact of closing the achievement gap in Minnesota is smaller than the national estimates by McKinsey. One explanation is the lower percentage of black and Latino students in Minnesota (22 percent) relative to the national average (45 percent). Likewise, low-income students also comprise a smaller percent of population in Minnesota than in the nation.

Another explanation could be different assumptions used in McKinsey’s and Hanushek’s models. Although details are not clear, the McKinsey report seems to assume that after the 15-year reform period, the entire workforce achieves the projected gains in cognitive skills commensurate with the closing of the achievement gap in test scores. Hanushek’s paper assumes a more gradual displacement of the existing workforce with higher-quality graduates. Correspondingly, estimates for Minnesota using this assumption yield a smaller impact of bridging the gap.

Even if the economic impact of closing the gap is estimated to be smaller in Minnesota than nationally, it is by no means a trivial one. As anyone planning a retirement learns, small changes in growth rates can have a big impact on the future value of investments, more so for longer-term investments.

Furthermore, this analysis doesn’t take into account benefits to government from closing the achievement gap, such as reductions in remedial education and crime costs, and eventually higher tax revenue, nor does this analysis estimate the cost of a 15-year education reform. Both of these data points are needed to assess whether the government would achieve a positive rate of return from investing in education reform. An analysis by Henry Levin and colleagues suggests that investments in early childhood education and some reforms for school-age children do just that.

And, finally, this is not the only achievement gap whose closure would likely lead to faster economic growth. Nationally, for example, Asian students have the highest average test scores. If, hypothetically, educational reform could boost the performance of white students to the level of Asian students, overall average math scores would increase by about 2 percent, with about a 0.2 percentage point increase in economic growth. Furthermore, if test scores of all non-Asian students were raised to the average of Asian students, average math scores would increase by over 6 percent, with about a 0.6 percentage point increase in economic growth, almost 70 percent larger than the effect of closing the black-Hispanic and white achievement gap. This particular analysis, however, isn’t relevant to Minnesota, where average test scores for Asian students are lower than both Asian students nationwide and Minnesota white students.

The other Bakken pipelines: Water for fracking

Much pipeline development in the Bakken region of North Dakota and Montana is focused on transporting crude oil. But pipe is also being laid to carry a humble commodity essential to oil production in the region: water.

Copious amounts of water are required to extract oil from the Bakken’s shale beds. The fracking process—injecting a mixture of water and chemicals into shale rock to release oil—consumes up to 8 million gallons of freshwater per well. And along with oil and natural gas, wells produce even larger amounts of subterranean saltwater over their operating life. “The first thing that’s produced out of a well is water; the last thing ever produced out of a well is water,” said Rodney Wren, president of New Frontier Midstream, a Texas-based developer of oil and gas infrastructure.

Today, most freshwater used for fracking is trucked to wellheads, and tanker trucks also haul away saltwater and used frac (“flow back”) water for disposal in deep wells. Trucking water raises costs for producers—fees in the Bakken range from 2 to 10 cents per gallon, depending on miles traveled—and contributes to wear and tear on rural roads.

It’s cheaper to pipe water to and from the wellhead, especially in areas where wells are close together. Brigham Exploration, an oil company acquired by Statoil of Norway in 2011, was a leader in laying water pipelines in the Bakken, installing them simultaneously with crude oil and gas lines. Incoming pipe delivers freshwater for fracking from municipal or rural water systems; outgoing pipelines carry away wastewater for disposal.

There are no public data on oil-industry water networks in the Bakken, but Statoil, New Frontier and other petroleum and energy transportation firms are laying new water pipe to wellheads. New Frontier has plans to build wastewater gathering systems and disposal wells near Dickinson, N.D., and Sidney, Mont., to serve oil and gas producers in those areas.

The next step in oilfield water management is recycling frac flowback water. Statoil has tested a fracking method that uses a 50:50 mixture of flow back water and freshwater. The company aims to raise the proportion of flow back water used to 80 percent—greatly reducing the volume of freshwater that must be transported to the wellhead.

For much more on pipelines and other energy transportation infrastructure in the Bakken, look for the upcoming April issue of fedgazette.

Some Ninth District MSAs gained or lost counties

Every few years, the White House’s Office of Management and Budget revises the definitions of MSAs (metropolitan and micropolitan statistical areas). This is important for data geeks because they analyze all sorts of information. In late February, the OMB published the new delineations. The Ninth District MSAs added nine counties and lost two (see map).

The nine counties that were added include three to the Minneapolis-St. Paul-Bloomington region, which now encompasses 16 counties in Minnesota and Wisconsin (see map; the Twin Cities metro is outlined). Two were also added to the fast-growing Bismarck, N.D., region. Other MSAs that added one county are Billings, Mont.; Rapid City, S.D.; and Rochester, Minn. Pierre, S.D., is the only micropolitan statistical area that added a county.

The two counties dropped are also from micropolitan designations—Billings, N.D. (estimated population 905) and Hamlin, S.D. (estimated population of 5,918), presumably because of small, rural populations.

  District MSA map -- 4-4-13

Banks paring back their branches

It has taken some time for the ball to start rolling the other way, but banks across the country and Ninth District are slowly pulling back on branches. Call it “too small to bail.”

The total number of Ninth District bank branches rose steadily from 2001 to 2006—increasing by nearly 25 percent—before plateauing during the recession (see Chart 1). Branches saw some gains and losses over the next several years, but still rose on net from 2006 to the fourth quarter of 2009, to 3,027 branches. But since then the Ninth District has officially lost about 70 branches (more on this in a bit; the actual number is likely higher).

The Upper Peninsula of Michigan has seen the biggest loss of branches, but the trend started well before the recession. Branches there peaked in 2004 at 172 and were down to 139 by the end of 2009. By the fourth quarter of last year, the U.P. had lost another six branches. Minnesota has shed 35 bank branches, but from a much larger base of more than 1,400 branches. Other district states (including northwestern Wisconsin, the only portion technically in the Ninth District) lost only a small handful of branches—even booming North Dakota saw branches drop by a half dozen over this period.

The outlier, with caveats, is Montana, which officially saw the number of branches rise by nine, or almost 3 percent since the end of 2009 (see Chart 2). However, at least part of this bump appears to come from full-fledged banks getting converted by a parent company to branch status. Last year, for example, Glacier Bancorp, one of the largest bank holding companies in that state, consolidated 11 bank subsidiaries—five of them in Montana—into a single commercial bank, effectively converting previously independent banks into branches; nothing else changed except the regulatory designation of the building.

At the same time, Montana is still something of an outlier in terms of total “banking service locations”—in essence, the number of banks plus branches. Montana saw a net-zero change from 2009 to 2012, while every other district state saw a decline of banking locations of between six (in the U.P.) and 72 (Minnesota).

For more information on the health of Ninth District banks, see the Minneapolis Fed’s Banking Conditions website, which is updated quarterly.

Bank branches -- 4-2-13

 Economist Jason Schmidt contributed to this post.