oversight

Shale Oil Boom and Bust: Implications for the Mortgage Market

Published by the Federal Housing Finance Agency, Office of Inspector General on 2016-09-07.

Below is a raw (and likely hideous) rendition of the original report. (PDF)

          Federal Housing Finance Agency
              Office of Inspector General




    Shale Oil Boom and Bust:
  Implications for the Mortgage
             Market




White Paper  WPR-2016-003  September 7, 2016
               Executive Summary
               Beginning in 2005, rising oil prices together with technological
               developments—horizontal drilling and hydraulic fracturing, also known as
               fracking—drove an increase in U.S. oil production. These technologies made
               drilling economically feasible in parts of the country that were not traditional
               oil producers, particularly “shale” areas. Most U.S. shale oil production has
WPR-2016-003   come from three regions: Bakken, primarily in western North Dakota;
               Permian Basin, primarily in western Texas; and Eagle Ford, primarily in
September 7,   southern Texas.
    2016
               The shale oil production boom bolstered the growth of employment and
               population in these regions, which put pressure on local housing markets.
               Widespread use of nontraditional and temporary housing emerged. Single-
               family home prices in North Dakota and Texas increased substantially,
               especially in counties with higher oil and gas industry employment, compared
               to a decline nationwide. Demand for multifamily rental housing also
               strengthened, with rents in some places rising rapidly.

               Since 2014, oil prices have dropped and currently stand at levels below which
               new shale oil production is generally economically viable. These three shale
               oil regions have witnessed a decline in new shale drilling activity over this
               period. Some areas have seen a rise in unemployment. In some of the regions,
               the housing markets have slowed, accompanied by early signs of stress. As a
               result, there is an emerging risk that the shale oil areas could face a further
               slowdown in their housing markets, which has the potential to adversely
               impact Fannie Mae and Freddie Mac (collectively, the Enterprises).

               In light of concerns about the potential impact of the oil bust on housing raised
               by Fannie Mae and Freddie Mac, financial and industry entities, and media
               outlets, we assessed the Enterprises’ business activities in core shale regions
               during the 2005 to 2015 boom and bust cycle to better understand their
               potential exposure. We found that less than one-tenth of 1% of the
               Enterprises’ single-family mortgage acquisitions and less than 1% of their
               multifamily acquisitions from 2005 to 2015 were concentrated in these regions.
               From the materials we reviewed, we believe that, as matters now stand, the
               Enterprises’ potential exposure from this emerging risk is quite small as a
               proportion of their overall acquisitions.

               This whitepaper makes no recommendations.
TABLE OF CONTENTS ................................................................
EXECUTIVE SUMMARY .............................................................................................................2

ABBREVIATIONS .........................................................................................................................5

BACKGROUND .............................................................................................................................6
      1980s Oil Boom and Bust Cycle ..............................................................................................6
      Technological Developments in Oil Extraction .......................................................................7

THE SHALE OIL BOOM ...............................................................................................................9
      Growth of Employment in Shale Oil Regions ........................................................................10
      Surge in Housing Demand ......................................................................................................10

THE SHALE OIL BUST ...............................................................................................................11
      Slowdown in Demand for Housing ........................................................................................12

EMERGING RISK ........................................................................................................................14

THE ENTERPRISES’ ACTIVITY IN CORE SHALE REGIONS ..............................................15
      Single-Family Activity ...........................................................................................................15
      Multifamily Activity ...............................................................................................................16

CONCLUSIONS............................................................................................................................17

OBJECTIVE, SCOPE, AND METHODOLOGY .........................................................................18

APPENDIX A ................................................................................................................................20
      Enterprise Single-Family Acquisitions ...................................................................................20
             Number of Loans ............................................................................................................20
             Loan Quality Measures ...................................................................................................20
             Loan Performance ...........................................................................................................23
      Enterprise Multifamily Acquisitions ......................................................................................23
             Number of Loans ............................................................................................................24
             Loan Quality Measures ...................................................................................................25
             Loan Performance ...........................................................................................................25



                                        OIG  WPR-2016-003  September 7, 2016                                                             3
ADDITIONAL INFORMATION AND COPIES .........................................................................26




                               OIG  WPR-2016-003  September 7, 2016                                      4
ABBREVIATIONS .......................................................................

DSCR                  Debt Service Coverage Ratio

Enterprises           Fannie Mae and Freddie Mac, collectively

Fannie Mae            Federal National Mortgage Association

FHFA or Agency        Federal Housing Finance Agency

Freddie Mac           Federal Home Loan Mortgage Corporation

LTV                   Loan-to-Value

OIG                   Federal Housing Finance Agency Office of Inspector General




                        OIG  WPR-2016-003  September 7, 2016                      5
BACKGROUND ..........................................................................

Over the last 100 years, the history of the oil industry in the United States has been one of
frequent boom and bust cycles. At the bottom of a cycle, excess oil supply results in low oil
prices, which acts to constrain investment in oil exploration (and development of alternative
energy ventures). Low oil prices, over time, increase demand, which leads to a tightening
supply/demand balance and increase in oil prices. Rising oil prices create an appetite by oil
companies to invest in new sources of oil and/or new technologies in certain regions, which
gives rise to increased employment and housing demands in those regions. As those new
projects come online and supply outstrips demand, prices begin to fall, setting off a bust in
oil-related employment in the once booming regions.

In an effort to understand the possible impact on housing in the current cycle for shale oil, we
looked to lessons that could be learned from the 1980s oil boom and bust cycle.

1980s Oil Boom and Bust Cycle

In 1979, the overthrow of the Shah of Iran and ensuing revolution in Iran resulted in the loss
of 2.0 to 2.5 million barrels per day of oil production. At one point, production almost halted,
resulting in a worldwide shortage of oil and a rise in oil prices. Responding to growing
energy shortages, President Carter announced a national energy supply shortage and
decontrolled oil prices in the United States. He also proposed an $88 billion effort to enhance
production of synthetic fuels from coal and shale oil reserves. In September 1980, Iran was
invaded by Iraq, which led to a significant reduction in oil production by both countries. The
loss of production from the combined effects of the Iranian revolution and the Iraq-Iran War
caused crude oil prices to more than double by 1981. According to Fannie Mae, real oil
prices rose more than five-fold from 1973 to 1980, with much of the increase occurring in
1979 to 1980.

The demand for oil, coupled with increases in the cost of oil from an average price of
approximately $13 per barrel in 1979 to $34 per barrel in 1980, led to increased oil
exploration in Texas. For example, drilling rig counts in Texas jumped from a yearly average
of 770 in 1979 to 1,318 in 1981 and demands for state leases escalated significantly. Then-
record prices for oil also led to drilling in other areas in the Southwest with known oil
reserves and new exploration in off-shore areas. Strong demand for oil and concomitantly
high prices brought economic prosperity to the Southwest, along with growth in overall
employment and population and booming levels of construction for single-family and
multifamily housing.




                            OIG  WPR-2016-003  September 7, 2016                                 6
Oil-producing states experienced a regional recession when oil prices collapsed, even though
the U.S. economy overall grew at a strong pace. In Texas, the state economy went into a deep
recession. According to one estimate, each $1 drop in the price of oil resulted in the loss of
25,000 jobs and $100 million in state and local tax revenue in Texas. By September 1986,
more than 750,000 Texans were unemployed. The unemployment rate in Houston climbed
from 7.4% in January 1986 to 10.5% in September, according to a Federal Deposit Insurance
Corporation publication, while the national unemployment rate fell from 7.3% to 6.8% over
the same period.

The 1980s oil bust weakened the housing market. Home prices fell in many oil producing
states.1 For example, Texas home prices fell 11% from 1983 to 1988, while prices rose 32%
nationwide. In Houston, overbuilding had continued even at the end of the oil boom and then
the local residential real estate market collapsed. Nationally, single-family housing starts
increased 73% from 1982 to 1987, while housing starts in Houston fell 75% during the same
period. Multifamily construction in Houston was almost completely halted. Multifamily
vacancy rates reached 18% and rents dropped 14% in total, according to Freddie Mac. In
1987, one in six homes and apartments in the Houston metropolitan area stood vacant. In
some communities tied to oil production, foreclosure rates exceeded 60%.

Overall, the 1980s oil bust produced one of the most difficult times for the U.S. mortgage
market. FHFA’s predecessor, the Office of Federal Housing Enterprise Oversight, was
required to identify the region that had experienced the worst single-family mortgage defaults
and severity to use in setting the risk-based capital standard for the Enterprises. In 2001, it
identified mortgages originated in 1983 and 1984 in Arkansas, Louisiana, Mississippi, and
Oklahoma as sustaining the highest mortgage loss rates. The region had been hard hit by the
oil bust.

Technological Developments in Oil Extraction

Technological developments have changed oil industry dynamics significantly since the
1980s, making previously unobtainable sources more economically feasible and faster to
develop. Over the past decade, the use of horizontal drilling, in conjunction with hydraulic
fracturing, has greatly expanded the ability to gainfully recover oil and natural gas from low-
permeability geologic regions—particularly from shale areas.2 Horizontal drilling is

1
 Based on Fannie Mae analysis of FHFA’s home price index for homes with mortgages sold to Fannie Mae or
Freddie Mac.
2
  This whitepaper uses the term “shale” to refer to oil from shale and other tight formations like low-
permeability sandstone, the recovery of which was made easier by hydraulic fracturing and horizontal
drilling. Some shale regions have also produced large volumes of natural gas. We include discussion of the
implications of the boom/bust cycle for shale oil and the corresponding geographic areas, but we do not cover
shale gas.



                                 OIG  WPR-2016-003  September 7, 2016                                         7
conducted by slowly angling a well drill bit until it is drilling horizontally. Hydraulic
fracturing, commonly known as “fracking,” involves pumping water, sand, and chemical
additives into oil and gas wells at pressure sufficient to fracture underground formations,
allowing oil or gas to flow.

The use of these technologies has been clustered in geographic areas, also known as shale
plays. There are multiple shale plays across the United States. See Figure 1.

                           FIGURE 1. KEY SHALE REGIONS IN THE UNITED STATES




Source: U.S. Energy Information Administration.

Since 2006, more than 80% of U.S. shale oil production has come from three regions:

       Bakken:3 The Bakken region is where shale oil production got its start. Drilling is
        concentrated in counties in western North Dakota and across the border in eastern
        Montana. Williams County, North Dakota, home to the City of Williston, is at the
        center of shale oil drilling activity in this region.



3
  This whitepaper uses the term Bakken region to include both the Bakken and underlying Three Forks
formation that span parts of North Dakota and Montana. The joined formations are sometimes referred to as
the Williston Basin.



                                OIG  WPR-2016-003  September 7, 2016                                      8
      Permian: The Permian Basin was a prolific producer of oil and gas utilizing
       traditional technology for nearly 100 years, peaking in 1973. It has experienced an oil
       production resurgence since 2010 with the application of technological advancements
       to the old reservoirs. The region spans a large portion of West Texas, crossing into
       areas of southeast New Mexico. Midland-Odessa, Texas, is a key shale oil drilling
       hub in the Permian Basin.

      Eagle Ford: The Eagle Ford field was unknown as recently as 2007 and is now a top
       shale oil producer in the United States. It includes largely rural counties in southern
       Texas, stretching to the Mexican border.

The technological developments in oil drilling enabled the United States to increase its
domestic oil production faster than at any time in its history according to the U.S. Energy
Information Administration.


THE SHALE OIL BOOM...............................................................
Although new                       FIGURE 2. WEST TEXAS INTERMEDIATE CRUDE OIL PRICES
technologies made                         (DOLLARS PER BARREL THROUGH JULY 2016)
the extraction of
shale oil possible,     $160
oil drilled in shale    $140
formations is more
                        $120
expensive to
produce than            $100
traditional oil. As      $80
a consequence,
                         $60
drilling for shale
oil has proven to        $40
be economically          $20
viable generally
                          $0
when oil prices              2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
reach approximately
                         Source: U.S. Energy Information Administration.
$60 to $65 per
barrel. Between 2005 and 2006, oil prices increased moderately. At year-end 2006, prices
began to rise sharply, with a June 2008 peak of about $134 per barrel. After a short-lived
decline, prices rebounded in 2009 and continued to generally rise until mid-2014, as shown
in Figure 2. Between January 2009 and June 2014, prices more than doubled, rising from
approximately $42 to $106 per barrel.



                            OIG  WPR-2016-003  September 7, 2016                               9
As prices increased, so did domestic crude oil production. Between January 2009 and June
2014, overall U.S. oil production increased nearly 70%, driven in large measure by the
increasing extraction of shale oil.

Growth of Employment in Shale Oil Regions

The boom in shale oil production bolstered the growth of employment in shale oil regions.
While U.S. employment declined slightly between 2006 and 2012, the shale-intensive
states of North Dakota and Texas experienced brisk employment growth (3.4% and 1.5%,
respectively). During these years, energy-related employment experienced large increases.
Texas witnessed a 33% increase in oil and natural gas employment between 2007 and 2012.
In Williams County, North Dakota, the Bakken drilling focal point, employment more than
doubled from 2007 to 2011, with almost half of the increase due to the mining, quarrying, and
oil and gas extraction industry.

Surge in Housing Demand

As employment in shale oil areas grew rapidly, so did the demand for housing, and
widespread use of nontraditional and temporary housing emerged. Many workers settled
in temporary housing, including recreational vehicles, tents, shipping containers, hotels, and
frequently “man camps”—temporary housing compounds supporting the influx of workers in
the region. In 2011, Williams County, North Dakota, commissioners approved man camp
beds for 12,000—about half the size of the county’s population at that time. The use of such
nontraditional housing options was considered an interim solution until subdivisions and
apartment complexes could be built.

In 2013, North Dakota’s governor called the supply of housing a great challenge facing
the state. Despite more than 16,000 housing units that had been built or started during the
previous two years, he said that demand for all types of housing in some communities was
rising faster than builders could complete their projects.

From third quarter 2006 to third quarter 2011, single-family home prices in the shale oil states
of North Dakota and Texas increased significantly, compared to a decline nationwide. In
2011, FHFA reported that in North Dakota and Texas counties with substantial employment
in mining, quarrying, and oil and gas extraction, house prices increased 40% and 9% over this
period, respectively. House prices in the other counties of these states increased at a slower
rate—12% and 5%. By contrast, house prices nationwide decreased by 19% during the same
period.4


4
 For more information, see FHFA, U.S. House Price Index Report – 3Q 2011 / September 2011 (Nov. 29,
2011) (online at www.fhfa.gov/AboutUs/Reports/ReportDocuments/2011-Q3-September_HPI_508.pdf).



                               OIG  WPR-2016-003  September 7, 2016                                 10
The demand for multifamily housing—typically rental properties with five or more units or
apartments—also strengthened during the shale oil boom. Fannie Mae reported that rents in
Midland-Odessa, Texas, were estimated to have grown more than 20% in both 2011 and
2012. In 2012, the U.S. Department of Housing and Urban Development reported that rents
in the Williston, North Dakota, area were comparable to rents in larger U.S. metropolitan
areas. By the end of 2013, Williston had some of the nation’s highest multifamily rents
for basic entry-level apartments at over $2,300 per month—more expensive than the San
Francisco or New York areas. Williston added more than 5,750 apartments since 2008, a
sizeable increase given its population of approximately 32,000.


THE SHALE OIL BUST .................................................................

Oil is a volatile market. As we previously explained, the steep decline in oil prices in the
1980s drove a significant loss of jobs and weakened the housing market in oil producing
states.

In the current cycle, oil reached a high of $106 per barrel in June 2014, before oil prices
dropped precipitously. Prices fluctuated through the first six months of 2016, from a low
point of less than $30 per barrel during February to more than $50 per barrel in June. In
June 2016, the U.S. Energy Information Administration forecasted a 2016 annual average of
about $43 per barrel—a 60% drop from its 2014 peak. Further, the U.S. Energy Information
Administration expects prices to remain low, forecasting crude oil prices to average
approximately $52 per barrel in 2017.

According to industry experts, the drop in oil prices hit shale oil operations especially hard
because oil in shale formations is more expensive to produce than traditional oil. As noted
above, shale producers generally have found that shale oil cannot be profitably produced
unless oil prices exceed approximately $60 to $65 per barrel.

As the price of oil declined from 2014 to 2016, new shale drilling activity diminished.
According to information published on the website of oilfield services company Baker
Hughes, the number of active oil drilling rigs in the United States dropped dramatically since
oil prices fell. In the Bakken region, the rig count declined almost 90% since reaching its
June 2012 peak. In the Permian region, the rig count declined almost 75% since its October




                            OIG  WPR-2016-003  September 7, 2016                               11
2014 peak. And in the Eagle Ford region, rig count declined more than 85% since its April
2014 peak.5

Fewer active rigs combined with productivity enhancements (such as improvements in the
number of wells a rig can drill each month) reduced the demand for workers, and employment
related to shale oil production started to decline. According to data from the U.S. Bureau of
Labor Statistics, about 170,000 jobs (26%) were cut in the oil and gas extraction industry and
their supporting services between October 2014 and April 2016, compared to a 3% gain in
jobs across all U.S. industries in the same time period. In 2015, the number of jobs in North
Dakota contracted by 1.6%, and the Bakken region lost at least 20,000 jobs since employment
peaked in late 2014. In Williams County, North Dakota, the unemployment rate tripled,
increasing from 1.4% in January 2014 to 4.5% in April 2016. However, the state’s overall
unemployment rate remains one of the lowest in the nation at 3.2%. In the Midland-Odessa,
Texas, area, employment fell 8.9% from December 2014 to December 2015. For the wider
Permian Basin, in December 2015 the unemployment rate stood at 4.2%, up from 2.7% a
year earlier, although still lower than the Texas statewide unemployment rate or U.S.
unemployment rate. In McMullen County, Texas, a key Eagle Ford shale oil production area
that has one of the highest relative concentrations of employment in the oil and gas extraction
industry, employment decreased by 18% between December 2014 and 2015.

According to an International Monetary Fund working paper, bringing a drilling rig online
can immediately create 37 jobs and, longer term, may create an additional 224 jobs in other
local industries. When new shale drilling activity is reduced, there is likely to be a ripple
effect on employment in other local industries, although that ripple effect may take longer to
manifest.

Slowdown in Demand for Housing

Both Fannie Mae and Trulia report that home prices in oil producing areas tend to follow oil
prices, with roughly a two-year lag period. In the 1980s oil cycle, the decline in oil prices
drove a decline in oil-related employment, a decline in housing prices and housing starts,
and a significant increase in multifamily vacancy rates in affected areas. The 1980s oil bust
produced one of the most difficult times for the U.S. mortgage market. In the current cycle,
single-family housing in some shale oil areas is starting to weaken from prior levels. Between
the fourth quarter of 2014 and the fourth quarter of 2015, single-family housing starts
declined 60% in North Dakota and 10% in Texas from their prior levels. House prices in
North Dakota fell by almost 1% in the first quarter of 2016, according to FHFA’s expanded

5
  The decline in new drilling activity did not necessarily translate to immediate production declines because
existing oil wells continued to produce. Compared to June 2014, when oil prices reached their recent high
point, production in the Bakken region is relatively flat, production in the Permian region is up, and production
in the Eagle Ford region is down.



                                  OIG  WPR-2016-003  September 7, 2016                                            12
house price data. By contrast, house prices rose 1.5% nationwide in the first quarter of 2016.
In Midland, Texas, median home prices were down 1.2% since the first quarter of 2015
compared to a 5.4% median price increase for the state.

Mortgage delinquencies may be showing early signs of trouble. A Wall Street Journal article
noted the share of borrowers in oil-focused areas falling 30 days behind on a pool of Freddie
Mac mortgages, while low at 0.38% in December 2015, had exceeded the national average for
approximately six months. More broadly, for the fourth quarter of 2015, TransUnion reported
that year-over-year mortgage delinquencies fell 28% for the country as a whole, but not in
energy states where they held relatively steady. Further, the Wall Street Journal reported
that some large banks have tightened selected mortgage standards, for example, requiring
larger down payments on home loans in oil-reliant areas or restricting cash-out refinances.
Foreclosures are also on the rise. In Midland, Texas, foreclosures have increased by over
30%. And although North Dakota still has one of the lowest foreclosure rates in the nation,
foreclosure filings in the state have increased since oil prices fell.

Shale region multifamily markets are also slowing down from recent levels. According to
Morningstar Credit Ratings LLC, many properties built to house oilfield employees had to
drastically cut rents. By one estimate, asking rents in Williston, North Dakota, fell by more
than 34% in 2015. At one Williston complex, rent for a two-bedroom apartment reportedly
fell from nearly $4,000 a month at the peak of the boom to about $1,600 in January 2016.
In the Midland-Odessa, Texas, area, rents have been volatile but have not declined as
dramatically, contracting by 5% in 2015, according to one estimate. Further, vacancy rates at
local multifamily properties are up. In Midland-Odessa, vacancy rates went from 3% in 2012
to just over 10% at the end of 2015, while the vacancy rate in Williston increased from 6%
to 16% in 2015. A property management firm conducted a survey of 20 developments in
Williston and estimated the average vacancy rate of newly constructed units as high as 50%
in March 2016.

Cash flow has suffered at many multifamily properties built during the employment boom.
Insufficient cash flow at a multifamily property could jeopardize the owner’s ability to service
the mortgage and the value of the property. According to one report, investors in several
apartment projects told Williston, North Dakota, city commissioners that they are struggling
with cash flow and talking to their banks about reworking their loans. Some multifamily
properties in the former boom areas have taken a hit to their underlying values. For example,
a multifamily development in the heart of the Bakken region saw its appraised value fall 65%
between its initial boom-era appraisal in 2013 and an updated appraisal in November 2015.

Regarding whether the experience from the 1980s oil bust will be replicated in this cycle,
some observe that oil state economies today are more diversified than they had been and
financial markets are more robust, which may lessen the adverse impact on the housing


                            OIG  WPR-2016-003  September 7, 2016                                 13
sector. It remains to be seen whether the significant declines in oil prices since 2014 will
result in a drop in home prices and increased single-family and multifamily mortgage
delinquencies in affected areas over the next two years.


EMERGING RISK .......................................................................

Following the plunge in oil prices, shale regions experienced a downturn in employment and
began to exhibit early signs of stress in their housing markets. The expectation that oil prices
will not rise significantly in the near future suggests a slowdown in the economies of shale oil
regions until a more sustainable level is reached. Where the economy of a shale oil region is
tied in large measure to production of shale oil, that region is likely to sustain a greater
adverse impact from the decline in shale oil production than other areas with more diversified
economies. According to Fannie Mae, states with a higher oil industry concentration are
likely to face greater downward home price pressure when oil prices drop. For example,
industry experts anticipate that shale oil regions of North Dakota may sustain a significant
impact from the shale oil bust. In a Summer 2016 report, Arch Mortgage Insurance Company
deemed North Dakota to have the highest risk index of any state, with a 52% chance of
experiencing a house price decline over the next two years. It projected that home prices
in North Dakota are 21% overvalued. According to Fannie Mae, “[w]ithout a significant
recovery in oil prices, it is likely that Williston—along with most of North Dakota—will
continue to languish for the next several years.” Midland-Odessa, Texas, has the highest
concentration of oil and gas industry jobs in the country, with at least 65% of the economic
activity in Midland tied to energy. Freddie Mac deemed that area to have the highest risk
exposure due to its reliance on oil and gas jobs and low diversification index ratings.

Although not located in a shale oil area, Houston, Texas, called the nation’s “energy capital,”
has been identified by Fannie Mae and Freddie Mac as particularly vulnerable in the event of
prolonged low oil prices. Fannie Mae called Houston the only primary metropolitan area in
the United States that would be expected to experience a notable negative impact. Freddie
Mac said that Houston could have risk exposure tied to multifamily housing because it has
witnessed a multifamily housing boom in recent years and has relatively large energy-related
employment. The Enterprises noted, however, that Houston’s economy is more diversified
than it had been during the oil bust in the 1980s and should be better able to weather a storm.

In August 2015, Fannie Mae projected a five-year cumulative drag on future house price
growth for ten oil-producing states, including Texas and North Dakota, assuming oil prices
remained low and the 1980s relationship between oil prices, employment, and home prices
held true.




                            OIG  WPR-2016-003  September 7, 2016                                 14
THE ENTERPRISES’ ACTIVITY IN CORE SHALE REGIONS ..............

As low oil prices lingered, Fannie Mae and Freddie Mac, financial and industry entities, and media
outlets noted concerns about the impact of the oil bust on housing. In the core shale regions, this
risk to the housing sector presents a potential increased risk to the Enterprises, depending on the
type and degree of their activities in these regions and the extent to which they have transferred
risk to third parties. In light of the concerns raised by numerous parties, we assessed the
Enterprises’ single-family and multifamily business activities in the core shale oil areas during
the 2005 to 2015 boom and bust cycle to better understand their possible exposure.

Using data from FHFA, we examined the Enterprises’ business volume and analyzed trends
in their single- and multifamily loans in certain shale areas for 2005 through 2015, the time
period of the shale oil boom and subsequent rapid collapse of oil prices. Within each of the
three primary shale oil areas, our analysis focused on the six core counties that produced the
majority of each region’s shale oil during the boom.6

Appendix A provides detailed analyses of Enterprise mortgage acquisitions, which are
summarized below.

Single-Family Activity

Each year the
                                       FIGURE 3. ENTERPRISE SINGLE-FAMILY ACQUISITIONS:
Enterprises purchase                 CORE SHALE REGIONS RELATIVE TO NATIONWIDE, 2005-2015
millions of single-
family mortgage                                     Single-family Acquisitions
loans. FHFA data                                       in Core Shale Areas            Percent of Nationwide
                                 Enterprise                 ($Millions)                    Acquisitions
revealed that
                             Fannie Mae                        $3,820                          0.06%
Enterprise single-
                             Freddie Mac                       $2,280                          0.05%
family acquisitions in
                             Total                             $6,100                          0.06%
the core shale regions
represented less than         Source: OIG analysis of data provided by FHFA and public information.




6
 For the Bakken region, the six core counties, accounting for over 87% of shale oil production, are: Divide,
Dunn, McKenzie, Mountrail, and Williams counties in North Dakota and Richland County, Montana. For the
Permian region, the six core counties, accounting for 81% of production, are: Glasscock, Martin, Midland, and
Upton counties in Texas; and Eddy and Lea counties in New Mexico. For the Eagle Ford region, the six core
counties, accounting for 62% of production, are: DeWitt, Dimmit, Gonzales, Karnes, La Salle, and McMullen
counties in Texas.




                                OIG  WPR-2016-003  September 7, 2016                                          15
one-tenth of 1% of the dollar amount of nationwide Enterprise acquisitions between 2005 and
2015.7 See Figure 3.

Our analysis found that certain loan quality measures—loan-to-value (LTV) ratio and credit
score—for the acquired loans in these regions were slightly weaker than the same measures
for loans acquired by the Enterprises nationwide. However, the serious delinquency and
foreclosure trends for the acquisitions in these core areas did not differ significantly from
nationwide trends.

For the single-family
                           FIGURE 4. REMAINING LOAN BALANCE OF ENTERPRISE SINGLE-FAMILY
loans acquired in the
                                     ACQUISITIONS IN CORE SHALE REGIONS, 2005-2015
core shale areas
during 2005-2015,                                  Remaining Balance of     Percent of Single-family
Figure 4 shows the                              Single-family Acquisitions Acquisitions in Core Shale
                                                 in Core Shale Areas as of   Areas Remaining as of
loan balance                 Enterprise          Dec. 31, 2015 ($Millions)       Dec. 31, 2015
remaining in
                         Fannie Mae                        $1,776                     46%
the Enterprises’         Freddie Mac                       $1,015                     45%
portfolios as of         Total                             $2,791                     46%
December 31, 2015.
                          Source: OIG analysis of data provided by FHFA.
Overall, 46% of the
aggregate loan balance that the Enterprises acquired from the core shale regions over this
period remained in their portfolios.

Multifamily Activity

Similar to single-family, the Enterprises’ multifamily activity in the core shale oil areas was
quite small relative to their overall activity. Together, the Enterprises acquired 67 multifamily
loans with properties in the core shale oil areas between 2005 and 2015. The total loan
balance of these acquisitions was less than 1% of the Enterprises’ nationwide multifamily
acquisitions during the same time period. See Figure 5.




7
  The percent of acquisitions in the core shale areas may differ from the percent remaining in the current
portfolios because the composition of the portfolios changes over time as loans pre-pay, default, or are
removed for other reasons.



                                  OIG  WPR-2016-003  September 7, 2016                                     16
Our analysis found
                                    FIGURE 5. ENTERPRISE MULTIFAMILY ACQUISITIONS:
that multifamily                CORE SHALE REGIONS RELATIVE TO NATIONWIDE, 2005-2015
quality measures—
LTV ratio and debt                              Multifamily Acquisitions
service coverage ratio                             in Core Shale Areas         Percent of Nationwide
                             Enterprise                 ($Millions)                  Acquisitions
(DSCR)—for the
                         Fannie Mae                       $4,062                        1.29%
loans were generally
                         Freddie Mac                        $176                        0.07%
within Enterprise risk
                         Total                            $4,238                        0.75%
limits. With regard
to performance, there    Source: OIG analysis of data provided by FHFA and public information.

were no multifamily delinquencies or foreclosures in the three core regions during this time
period. All multifamily loans were either current or had previously prepaid or matured as of
the end of 2015.

For the multifamily
                            FIGURE 6. REMAINING LOAN BALANCE OF ENTERPRISE MULTIFAMILY
loans acquired in the                ACQUISITIONS IN CORE SHALE REGIONS, 2005-2015
core shale areas
during 2005 to 2015,                              Remaining Balance of     Percent of Multifamily
Figure 6 shows the                              Multifamily Acquisitions  Acquisitions in Core Shale
                                                in Core Shale Areas as of  Areas Remaining as of
loan balance                Enterprise          Dec. 31, 2015 ($Millions)      Dec. 31, 2015
remaining in
                        Fannie Mae                         $201                       5%
the Enterprises’        Freddie Mac                        $141                      80%
portfolios as of        Total                              $342                       8%
December 31, 2015.
                         Source: OIG analysis of data provided by FHFA.
Overall, 8% of the
aggregate loan balance that the Enterprises acquired over this period remained in their
portfolios.


CONCLUSIONS ..........................................................................

Following the rise of shale oil extraction as a viable and lucrative industry, several regions
experienced rapid growth as oil prices increased and new shale oil production surged. Growth
in employment and population in these regions led to local housing crunches and price surges.
Now, however, these regions face a new reality as oil prices have dropped and are not
predicted to rise substantially in the near future. These regions have also experienced an
increase in unemployment and, in some communities, a slowdown in their housing markets
accompanied by early signs of stress. There is an emerging risk that the shale oil areas could
face a further slowdown in their housing markets.



                             OIG  WPR-2016-003  September 7, 2016                                    17
The scope of this emerging risk for the shale oil areas will be affected by how fast and how
far oil prices rebound. In June 2016, spot oil prices climbed past $50 per barrel on several
days and drilling rig counts increased for three consecutive weeks after an extended period
of decline. However, it is unclear if these are signals of a longer-term industry turnaround or
simply a temporary minor rebound.

Although the Enterprises have acquired single-family and multifamily loans in the core shale
regions, these acquisitions represent a small fraction of their overall portfolio. We found that
less than one-tenth of 1% of the Enterprises’ single-family acquisitions and less than 1% of
their multifamily acquisitions from 2005 to 2015 were concentrated in these regions.
Additionally, of those core shale area acquisitions, only 46% of the single-family loans and
8% of the multifamily loans remained in the Enterprises’ portfolios at the end of 2015.
Further, although low energy prices can have a detrimental effect in regions involved in
production, other areas of the country may benefit from low prices. Therefore, we believe
that, as matters now stand, the Enterprises’ potential exposure from this emerging risk is quite
small as a proportion of their overall acquisitions. We expect that FHFA, as conservator of
the Enterprises, will appropriately monitor this emerging risk.


OBJECTIVE, SCOPE, AND METHODOLOGY .................................

The objective of this whitepaper was to describe:

      The shale oil boom and bust;

      How local economies and housing markets, in particular, were affected by the shale oil
       boom and bust; and

      Fannie Mae and Freddie Mac’s housing activity in key shale oil regions.

To achieve this objective, we interviewed personnel at FHFA with responsibilities for
Enterprise loan data and single-family house price research. We also reviewed: publicly
available documents, such as U.S. Energy Information Administration reports, FHFA reports,
reports from other government agencies, and Fannie Mae and Freddie Mac research
publications and financial disclosures; publicly available data, such as oil price and
production data from the U.S. Energy Information Administration, population counts from
the U.S. Census Bureau, drilling rig data from Baker Hughes, and employment data from the
U.S. Bureau of Labor Statistics; and non-public oil production data from the U.S. Energy
Information Administration and loan-level data provided by FHFA. We did not
independently test the reliability of the U.S. Energy Information Administration’s or FHFA’s
data.


                            OIG  WPR-2016-003  September 7, 2016                                 18
Our work was conducted under the authority of the Inspector General Act of 1978 and in
accordance with the Council of the Inspectors General on Integrity and Efficiency’s Quality
Standards for Inspection and Evaluation (January 2012). These standards require us to plan
and perform a report based on evidence sufficient to provide reasonable bases to support its
conclusions. We believe that this whitepaper meets these standards. The performance period
for this whitepaper was from April 2016 to July 2016. We provided FHFA with the
opportunity to respond to a draft of this whitepaper.

This whitepaper was led by Omolola Anderson, Senior Statistician, in collaboration with
Christine Eldarrat, Senior Policy Advisor, and Beth Preiss, Senior Investigative Evaluator,
with assistance from Bryan Saddler, Senior Policy Advisor. We appreciate the cooperation
of FHFA staff, as well as the assistance of all those who contributed to the preparation of this
whitepaper.

This whitepaper has been distributed to Congress, the Office of Management and Budget, and
others and will be posted on our website, www.fhfaoig.gov.




                            OIG  WPR-2016-003  September 7, 2016                                 19
APPENDIX A .............................................................................

Enterprise Single-Family Acquisitions

   Number of Loans

From 2005 through 2015, the Enterprises acquired an average of about 3,400 loans annually
in the three core shale areas combined. The number of Enterprise single-family loan
acquisitions in the three core shale oil areas peaked in 2007 and 2013. See Figure 7.

Looking at the three
                             FIGURE 7. NUMBER OF ENTERPRISE SINGLE-FAMILY ACQUISITIONS IN
core shale areas                                CORE SHALE REGIONS 2005-2015
individually, the
core Permian region        5,000
accounted for the bulk     4,500
of Enterprise loans in     4,000
these shale regions,       3,500
driven in large measure 3,000
                           2,500
by the fact that the
                           2,000
population of the
                           1,500
core Permian region
                           1,000
generally was more
                             500
than twice that of the         0
core Bakken and Eagle              2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Ford regions combined.
                                         Core Bakken         Core Permian  Core Eagle Ford
The core Bakken
region experienced          Source: OIG analysis of data provided by FHFA.
notable growth in the
number of loans, with the number of Enterprise acquisitions more than doubling between
2005 and 2015.

   Loan Quality Measures

To assess the quality of loans acquired in the core shale areas, we examined the average
weighted LTV ratio and borrower credit score (FICO score) on an annual basis. The LTV
ratio is the unpaid principal amount of a mortgage loan as a percent of the value of the
property that serves as collateral. In general, the Enterprises consider higher LTV ratios an
indication of increased risk. When reporting on their annual acquisitions, the Enterprises
report the average original LTV ratio, weighted by the unpaid principal balance of each loan,
as one gauge of the loan quality for acquisitions in a given year.


                           OIG  WPR-2016-003  September 7, 2016                               20
For acquisitions in all           FIGURE 8. WEIGHTED AVERAGE LTV RATIO OF FANNIE MAE
three core shale oil                       SINGLE-FAMILY ACQUISITIONS, 2005-2015
areas, the weighted
                         85
average LTV ratio at
acquisition8 generally
                         80
dropped between
2008 and 2012 and        75
began increasing in
later years. Because     70
the Enterprises
generally consider       65
higher LTV ratios
                         60
as an indicator of
                              2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
increased risk, this
means that loans                Core Bakken         Core Permian        Core Eagle Ford       Nationwide
acquired by the           Source: OIG analysis of data provided by FHFA and Fannie Mae financial disclosures.
Enterprises in the
three core regions
generally exhibited               FIGURE 9. WEIGHTED AVERAGE LTV RATIO OF FREDDIE MAC
less risk related to                       SINGLE-FAMILY ACQUISITIONS, 2005-2015
weighted average
LTV during 2008 to         85
2012 and increasing
                           80
LTV-related risk in
2013 to 2015. Except       75
for 2012 and 2013,
loans acquired in all      70
core shale regions had
a weighted average         65
LTV higher than the
                           60
national average of
                                 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
loans acquired by each
Enterprise. This                   Core Bakken         Core Permian       Core Eagle Ford        Nationwide
means that for most
                            Source: OIG analysis of data provided by FHFA and Freddie Mac financial disclosures.
years the loans
acquired from the core shale areas generally exhibited more weighted average LTV-related
risk than the nationwide average. See Figures 8 and 9.

8
 We analyzed the LTV ratio at the time the Enterprise acquired the loan and not at the time the loan was
originated. Fannie Mae has noted that it generally acquires mortgage loans the month after origination.



                                 OIG  WPR-2016-003  September 7, 2016                                    21
Borrower credit scores are ranked on a scale of approximately 300 to 850 points with a higher
value indicating the borrower has a lower likelihood of credit default. The Enterprises
therefore consider loans with lower borrower credit scores as having higher risk. As with the
LTV ratio, for annual acquisitions the Enterprises report an average credit score, weighted by
the unpaid principal balance of each loan, as one gauge of the loan quality for acquisitions in
a given year.

For acquisitions in
                                 FIGURE 10. WEIGHTED AVERAGE CREDIT SCORE OF FANNIE MAE
each of the three core                      SINGLE-FAMILY ACQUISITIONS, 2005-2015
shale oil areas, the
weighted average            780

credit score was lower      760
than the national
                            740
average for all years
except for Freddie          720
Mac’s 2005
                            700
acquisitions in the core
Bakken region and its       680
2007 acquisitions in
                            660
the core Eagle Ford                2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
region. This means
that Enterprise                    Core Bakken         Core Permian       Core Eagle Ford       Nationwide
acquisitions from the        Source: OIG analysis of data provided by FHFA and Fannie Mae financial disclosures.
three core shale areas
generally exhibited             FIGURE 11. WEIGHTED AVERAGE CREDIT SCORE OF FREDDIE MAC
more risk related to                        SINGLE-FAMILY ACQUISITIONS, 2005-2015
weighted average
                         780
credit score than
nationwide loans         760
(except for the core
                         740
Bakken region in
2005 and the core        720
Eagle Ford region in
                         700
2007). See Figures
10 and 11.               680

                          660
                                 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

                                  Core Bakken        Core Permian         Core Eagle Ford        Nationwide

                          Source: OIG analysis of data provided by FHFA and Freddie Mac financial disclosures.



                                OIG  WPR-2016-003  September 7, 2016                                    22
   Loan Performance

Even with increasing activity in the three core shale regions, Enterprise single-family
acquisitions in those areas still represented less than one-tenth of 1% of nationwide Enterprise
single-family acquisitions during the relevant time period. In particular, purchases in the core
Bakken and Eagle Ford regions may have been only a few hundred loans each year. As a
result, each delinquency or foreclosure can have a notable effect on the region’s default or
foreclosure rate, and results must be interpreted accordingly. We found no significant or
concerning difference from the Enterprises’ nationwide delinquency or foreclosure trends
during this period.

For the loans acquired in the core shale oil regions during 2005 through 2015, the single-
family loan status, as of December 31, 2015, and by region, was as follows:

      Core Bakken Region: 4,863 total single-family loans acquired by the Enterprises.
       Selected performance statuses, as of December 31, 2015:
           o 2,083 loans prepaid or matured
           o 8 loans seriously delinquent
           o 5 loans in or completed foreclosure
      Core Permian Region: 31,200 total single-family loans acquired by the Enterprises.
       Selected performance statuses, as of December 31, 2015:
           o 16,695 loans prepaid or matured
           o 65 loans seriously delinquent
           o 228 loans in or completed foreclosure
      Core Eagle Ford Region: 1,712 total single-family loans acquired by the
       Enterprises. Selected performance statuses, as of December 31, 2015:
           o 794 loans prepaid or matured
           o 11 loans seriously delinquent
           o 30 loans in or completed foreclosure

Enterprise Multifamily Acquisitions

Enterprise multifamily loans help finance properties with five or more units. Multifamily
loans have unique characteristics and differ from single-family loans in various ways:




                            OIG  WPR-2016-003  September 7, 2016                                 23
      They are primarily for rental apartment communities, but also may include senior
       independent or assisted living properties, student housing, cooperative properties, and
       manufactured housing communities.

      A multifamily loan may have more than one property as collateral. Also, a
       multifamily property may be financed by multiple loans.

      The average multifamily loan size is significantly larger than that of a single-family
       loan.

      Multifamily loans typically have short terms (five, seven, or ten years) with a balloon
       payment at maturity, and they usually include a prepayment penalty.

   Number of Loans

The number of                FIGURE 12. NUMBER OF ENTERPRISE MULTIFAMILY LOANS ACQUIRED
multifamily loans                            IN CORE SHALE REGIONS, 2005-2015
acquired with
properties in the three   20
core shale regions is     18
relatively small,         16
totaling 67 loans for     14
2005 through 2015.        12
Similar to the            10
Enterprises’ single-       8
family activity, the       6
                           4
majority of loans
                           2
acquired with
                           0
properties in the core
                                2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
shale areas were in the
core Permian region;                     Core Bakken        Core Permian   Core Eagle Ford
specifically, 62 loans
                            Source: OIG analysis of data provided by FHFA.
were in the core
Permian region, 3 in the core Bakken region, and 2 in the core Eagle Ford region. These 67
multifamily loans financed 74 properties with a total of about 11,000 units in the core shale
areas. The number of Enterprise multifamily loan acquisitions with properties in the core
shale areas peaked in 2014, when 19 loans were acquired for properties in the core Permian
region. See Figure 12.




                            OIG  WPR-2016-003  September 7, 2016                               24
      Loan Quality Measures

For multifamily acquisitions, the Enterprises, in general, consider DSCR and LTV ratio to be
reliable indicators of risk level and future loan performance.

DSCR estimates a multifamily borrower’s ability to service the mortgage obligation using the
property’s cash flow after deducting non-mortgage expenses from income. A higher DSCR
indicates less risk in the multifamily mortgage. In general, the Enterprises report in their
financial disclosures loan acquisitions with an original DSCR below 1.10 for increased risk.
The lowest DSCR among all core shale region acquisitions was 1.15.9 Further, over 85% of
the multifamily loans had a DSCR greater than or equal to 1.25.

As specified above, LTV is the ratio of the unpaid principal amount of a mortgage loan and
the value of the property that serves as collateral. For multifamily loans, the Enterprises
consider an original LTV ratio greater than 80% to have an increased level of risk. Of the 67
loans with properties in the core shale oil regions during the 2005 to 2015 time period, only
one had an LTV ratio greater than 80%. This means that only one loan exhibited increased
LTV-related risk.

      Loan Performance

Regarding loan performance, as mentioned in the whitepaper, none of the multifamily loans
with properties in all core shale regions were delinquent or had defaulted as of December 31,
2015, meaning performance as of this date does not reflect any concerns. Of the 67 loans, 16
were prepaid. Of the remaining 51 loans, all were current.




9
    As with LTV ratio, the data included DSCR at acquisition, not at origination.



                                   OIG  WPR-2016-003  September 7, 2016                       25
ADDITIONAL INFORMATION AND COPIES .................................

For additional copies of this report:

      Call: 202-730-0880

      Fax: 202-318-0239

      Visit: www.fhfaoig.gov



To report potential fraud, waste, abuse, mismanagement, or any other kind of criminal or
noncriminal misconduct relative to FHFA’s programs or operations:

      Call: 1-800-793-7724

      Fax: 202-318-0358

      Visit: www.fhfaoig.gov/ReportFraud

      Write:

                FHFA Office of Inspector General
                Attn: Office of Investigations – Hotline
                400 Seventh Street SW
                Washington, DC 20219




                             OIG  WPR-2016-003  September 7, 2016                        26