The U.S. Shale Oil Boom, the Oil Export Ban, and the Economy: A - - PowerPoint PPT Presentation

▶
the u s shale oil boom the oil export ban and the economy
SMART_READER_LITE
LIVE PREVIEW

The U.S. Shale Oil Boom, the Oil Export Ban, and the Economy: A - - PowerPoint PPT Presentation

Nov 15, 2023 117 likes •502 views

The U.S. Shale Oil Boom, the Oil Export Ban, and the Economy: A General Equilibrium Analysis Nida akr Melek Federal Reserve Bank of Kansas City Michael Plante Federal Reserve Bank of Dallas Mine K. Ycel Federal Reserve Bank of Dallas

slide-1
SLIDE 1

The U.S. Shale Oil Boom, the Oil Export Ban, and the Economy: A General Equilibrium Analysis

Nida Çakır Melek Federal Reserve Bank of Kansas City Michael Plante Federal Reserve Bank of Dallas Mine K. Yücel Federal Reserve Bank of Dallas

September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 1 / 36

slide-2
SLIDE 2

The views expressed herein are solely those of the authors and do not necessarily reflect the views of the Federal Reserve Bank of Dallas, the Federal Reserve Bank of Kansas City or the Federal Reserve System.

September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 2 / 36

slide-3
SLIDE 3

Motivation

◮ Shale oil boom dramatically increased U.S. oil production ◮ U.S. oil imports declined ◮ Key feature of the boom: production of primarily “light” oil ◮ Refining sectors specialize to some degree across oil types ◮ U.S. crude oil export ban in place through 2015

Production Refining Export ban background September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 3 / 36

slide-4
SLIDE 4

Questions we address

◮ What are the implications of the shale oil boom?

◮ for oil prices? ◮ for the refining sector? ◮ for the broader economy?

◮ What are the implications of an export ban on crude oil?

◮ Was the ban actually binding at any point in time? ◮ Are the effects limited to the energy sector? September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 4 / 36

slide-5
SLIDE 5

What we do in this paper

◮ Develop a two-country DSGE model with three sectors (oil, refining,

non-oil)

◮ What’s new: different types of oil, refining sector, and

the U.S. crude oil export ban

◮ Calibrate the model using oil market and macro data (pre-shale boom) ◮ Use the model to examine the effects of the U.S. shale oil boom

◮ Start with case of no ban ◮ Consider implications of the export ban September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 5 / 36

slide-6
SLIDE 6

Literature

◮ DSGE literature with oil

◮ Backus and Crucini (2000), Leduc and Sill (2007), Bodenstein et. al. (2011),

Nakov and Nuno (2013), Plante (2014)

◮ International business cycles and trade

◮ Backus, Kehoe, and Kydland (1992, 1994), Crucini and Kahn (1996), Kose

and Yi (2001, 2006), Farrokhi (2016)

◮ U.S. shale oil boom

◮ Manescu and Nuno (2015), Mohaddes and Raissi (2016), Walls and Zheng

(2016), Kang et. al. (2016), Kilian (2016, 2017)

◮ Studies on the export ban (mostly non-academic)

◮ CRS (2014), ICF (2014), IHS (2014), Bordoff and Hauser (2015), Medlock

(2015), Brown et. al. (2014), etc.

September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 6 / 36

slide-7
SLIDE 7

Introduction to crude oil quality

◮ Oil generally viewed as a homogenous commodity ◮ But can vary across a number of dimensions, such as

◮ Density (light, medium or heavy) (API gravity) ◮ Sulfur content (sweet or sour) Characteristics of several crudes

◮ We split oil into three broad types

◮ Light (API gravity ≥ 35◦) ◮ Heavy (API gravity < 26◦) ◮ Medium (everything in between) September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 7 / 36

slide-8
SLIDE 8

U.S. light oil production almost tripled

◮ Total oil production increased by 72%

1 2 3 4 5 6 7 8 9 10 2010 2011 2012 2013 2014 2015 Light Medium Heavy

U.S. Crude Oil Production by Type

Million barrels per day SOURCE: Eni. September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 8 / 36

slide-9
SLIDE 9

U.S. light oil imports down substantially

◮ Shifts in the quantity and types of oil imported ◮ Total oil imports down by 20%

Table: U.S. imports of crude oil, mb/d U.S. crude imports Light Medium Heavy 2000 2.2 4.6 2.3 2005 2.3 4.3 3.5 2010 2.1 3.3 3.8 2011 1.7 3.3 4.0 2012 1.4 3.1 4.0 2013 0.9 3.0 3.9 2014 0.6 2.7 4.1 2015 0.6 2.6 4.2

September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 9 / 36

slide-10
SLIDE 10

Oil inputs to refineries up

◮ U.S. light oil input to refineries up by 52% ◮ U.S. refineries are geared toward processing heavy crude relative to the ROW

Table: Refiner inputs by type, U.S. and ROW, mb/d

U.S. refiner inputs ROW refiner inputs Light Medium Heavy Light Medium Heavy 2000 4.3 7.5 3.1 17.8 30.4 5.3 2005 4.0 7.1 4.2 17.5 36.0 5.9 2010 4.2 6.1 4.4 18.6 35.9 6.0 2011 4.2 5.7 4.6 18.0 37.6 5.8 2012 4.9 5.5 4.5 18.7 38.3 5.7 2013 5.3 5.3 4.5 18.9 37.9 5.8 2014 5.9 5.1 4.7 18.8 38.8 5.9 2015 6.4 5.1 4.8 19.0 39.5 6.6

Details September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 10 / 36

slide-11
SLIDE 11

Model - Households

Countries denoted i = 1, 2 (U.S., ROW)

◮ A typical household in country i maximizes utility subject to budget, time

and capital accumulation constraints, where E0

∞

  • t=0

βt (cµ

i,tL1−µ i,t

)γ γ ci,t is aggregate consumption and Li,t is leisure.

◮ Aggregate consumption given by

ci,t =

  • ωi
  • ca

i,t

−ρ + (1 − ωi)

  • cf

i,t

−ρ− 1

ρ

◮ Key feature: ρ allows for inelastic demand of petroleum products

September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 11 / 36

slide-12
SLIDE 12

Model - Firms - Oil sector

Representative firms maximize profits subject to production technologies

◮ Introduce a cost function to model production of oil, as in Balke,

Plante, and Yucel (2015)

◮ Assume it costs Ck i units of the non-oil good to produce a unit of oil

type k (k = L, M, H) in country i

◮ Production cost is a convex function of the level of output

Ck

i,t =

  • yok

i,t

zok

i,t

1+ 1

ηk i

1 + 1

ηk

i

◮ Key feature: ηk allows us to make oil supply price-inelastic

September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 12 / 36

slide-13
SLIDE 13

Model - Firms - Refining sector

◮ Refined petroleum products are produced using labor, capital and oil ◮ Production function, yf i,t, is a nested CES of value-added and oil

  • wf

i

  • zf

i (nf i,t)χf

i (K f

i,t)1−χf

i

−ρf

i + (1 − wf

i )G(of Li,t, of Mi,t, of Hi,t) −ρf

i

  • 1

−ρf i

allowing us to model differences between U.S. and ROW refining sectors in

◮ labor-share of value-added ◮ value-added share of gross output

ρf allows us to model the fact that it is hard to substitute between oil and other inputs when it comes to producing fuel

September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 13 / 36

slide-14
SLIDE 14

Model - Firms - Refining sector

◮ Different types of oil are imperfect substitutes as inputs into the

refining process

◮ The aggregator G(of Li,t, of Mi,t, of Hi,t) is given by

 w o

i (of Hi,t )−ρoil

i

+ (1 − w o

i )

  • ωo

i (of Li,t )−ηoil

i

+ (1 − ωo

i )(of Mi,t )−ηoil

i

ρoil

i ηoil i

 

1 −ρoil i

allowing us to

◮ capture differences in use of the oil types across countries ◮ emphasize the fact that substitution is easier between light and

medium oils

September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 14 / 36

slide-15
SLIDE 15

Model - Firms - Non-oil sector

◮ Production function is a nested CES of value-added and fuel

ya

i,t =

  • wa

i

  • za

i,t(na i,t)χa

i (K a

i,t)1−χa

i

−ρa

i + (1 − wa

i )(mf i,t)−ρa

i

  • 1

−ρa i

◮ Key feature: ρa allows for inelastic demand of petroleum products ◮ Key feature: χa and wa let us match cost-shares

September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 15 / 36

slide-16
SLIDE 16

Model - Trade

◮ Free trade in all goods in baseline case ◮ Free trade in all goods except crude oil in export ban case

◮ The export ban in country 1 is introduced by three inequality

constraints

  • f

k1,t − y ok 1,t ≥ 0

◮ LOOP and PPP hold for all goods when the ban does not bind ◮ If the ban binds for type k, oil prices for type k can diverge in U.S.,

ROW

◮ To solve the model with inequality constraints, we use the Guerrieri

and Iacoviello (2015) OccBin toolkit

September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 16 / 36

slide-17
SLIDE 17

Calibration: Overview

◮ Use data on the oil market, refining sectors and macroeconomy to

guide our calibration

◮ annual frequency, match data from 2010

◮

List of data

◮

U.S. calibration

◮

ROW calibration

◮

Model parameters

◮

Productivity shocks September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 17 / 36

slide-18
SLIDE 18

Calibration: Refining sector

Table: Calibration of refining sector parameters U.S. ROW Data source Labor share .16 .30 WIOD Capital share .84 .70 WIOD Share of light oil .29 .31 Calculations Share of medium oil .41 .59 Calculations Share of heavy oil .30 .1 Calculations

September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 18 / 36

slide-19
SLIDE 19

Calibration: Refining sector elasticities

No long time series on refiner inputs by type of oil or No previous works that calibrate this type of elasticity

◮ Long time series available on oil prices of different oil types

◮ use price data on LLS (light), Dubai (medium) and Maya (heavy) ◮ match two moments to calibrate elasticities: correlation between light

and medium oil prices, and correlation between light and heavy oil prices

◮ Use total oil input to refineries data for substitution between oil and value-added

◮ match volatility of U.S. total oil input to refineries

◮ Our calibration gives us

◮ elasticity between oil and value-added (ρf ): 0.285 ◮ elasticity between light oil and medium oil (ηoil): 3.65 ◮ elasticity between heavy and composite (ρoil): 2.13 September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 19 / 36

slide-20
SLIDE 20

Policy experiment

◮ Start from initial steady state ◮ Replicate the path of U.S. light oil production from 2011 to 2015

◮ First consider effects under free trade scenario ◮ Then consider effects under export ban case

◮ Consider a persistent increase in light oil production

◮ Replicate the EIA’s 2015 projections for light oil production through

2020

◮ effects under free trade and export ban September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 20 / 36

slide-21
SLIDE 21

Responses to positive U.S. light oil shock w/ export ban

2010 2015 2020 100 200 2010 2015 2020 50 100 Ignoring ban With export ban 2010 2015 2020

  • 3
  • 2
  • 1

2010 2015 2020

  • 30
  • 20
  • 10

2010 2015 2020

  • 20
  • 10

2010 2015 2020

  • 15
  • 10
  • 5

2010 2015 2020

  • 200
  • 100

2010 2015 2020 20 40 60 2010 2015 2020

  • 5

5 10

Figure shows percent deviations from steady state

September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 21 / 36

slide-22
SLIDE 22

Responses to positive U.S. light oil shock w/ export ban

2010 2015 2020 2 4 2010 2015 2020

  • 0.03
  • 0.02
  • 0.01

2010 2015 2020

  • 4
  • 2

2 10-3 2010 2015 2020 2 4 2010 2015 2020 0.5 1 1.5 2010 2015 2020 0.5 1 1.5

Figure shows percent deviations from steady state

September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 22 / 36

slide-23
SLIDE 23

Responses to positive persistent U.S. light oil shock w/ export ban

2010 2015 2020 100 200 300 2010 2015 2020 50 100 Ignoring ban With export ban 2010 2015 2020

  • 3
  • 2
  • 1

2010 2015 2020

  • 40
  • 20

2010 2015 2020

  • 30
  • 20
  • 10

2010 2015 2020

  • 20
  • 10

2010 2015 2020

  • 300
  • 200
  • 100

2010 2015 2020 50 100 2010 2015 2020 5 10 15

Figure shows percent deviations from steady state

September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 23 / 36

slide-24
SLIDE 24

Responses to positive persistent U.S. light oil shock w/ export ban

2010 2015 2020 2 4 6 2010 2015 2020

  • 0.03
  • 0.02
  • 0.01

2010 2015 2020

  • 4
  • 2

2 10-3 2010 2015 2020 2 4 6 2010 2015 2020 1 2 2010 2015 2020 0.5 1 1.5

Figure shows percent deviations from steady state

September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 24 / 36

slide-25
SLIDE 25

Conclusion

◮ We introduce a DSGE model with three types of oil and a refining

sector to explore the effects of the shale oil boom and the U.S. oil export ban

◮ We find that the oil boom causes light oil prices and fuel prices to fall ◮ It also leads higher processing of light oil and dramatic declines in

U.S. imports of light oil

◮ GDP increases and petroleum trade balance improves ◮ The export ban primarily distorts the price of light oil and the refining

sectors with slight impact on GDP and negligible impact on fuel prices

September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 25 / 36

slide-26
SLIDE 26

Refiner input calculations

  • 1. Assume U.S. crude exports are “light” oil
  • 2. Construct an estimate for U.S. refiner inputs of type j for each year

Inputj

t = Productionj t + Importsj t − Exportsj t

  • 3. ROW estimate for type j is given by

ROW inputj

t = World productionj t − U.S. inputj t

  • 4. Data limitations prevent us from adjusting for inventories

◮ Annual changes for U.S. crude inventories in mb/d relatively small

(2010: +.02, 2011: -.01, 2012: +.1, 2013: -.02, 2014: +.1, 2015: +.15)

back September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 26 / 36

slide-27
SLIDE 27

U.S. crude oil production by type

1 2 3 4 5 6 7 8 9 10 2010 2011 2012 2013 2014 2015 Light Medium Heavy

U.S. Crude Oil Production by Type

Million barrels per day SOURCE: Eni.

back September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 27 / 36

slide-28
SLIDE 28

U.S. processed more than 40 percent of world’s heavy crude in 2010

10 20 30 40 50 60 70 80 90 Light Medium Heavy U.S. ROW Percent Source: Authors' calculations.

back September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 28 / 36

slide-29
SLIDE 29

Crude by type

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 20 25 30 35 40 Sulfur content (percentage) API gravity (a measure of crude oil density) sour sweet heavy light Maya Mars Brent LLS WTI

Characteristics of Various Crude Oils

SOURCES: Bloomberg; Platts.

back September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 29 / 36

slide-30
SLIDE 30

Overview of the U.S. crude oil export ban policy

◮ Policy in place since the 1970s, lifted end of 2015 ◮ Prohibited exports of crude oil, no ban on refined products ◮ Exports possible under several exemptions

◮ Exports of Alaskan crude oil (negligible since 2000) ◮ Exports to Canada for use there-in ◮ Few other exemptions Back September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 30 / 36

slide-31
SLIDE 31

U.S. crude exports to Canada

100 200 300 400 500 600 Total exports Exports to Canada Thousand barrels per day SOURCE: Energy Information Administration. Export ban lifted December 2015 Back September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 31 / 36

slide-32
SLIDE 32

Calibration: List of select data and sources

Sources: Bloomberg, BEA, BLS, EIA, Eni, IEA, IMF, Oil and Gas Journal, UN, WIOD

◮ Light, medium and heavy oil production ◮ Total crude oil production ◮ Refined products production ◮ Energy consumption by sector and source ◮ Crude input to refineries and refinery processing gains ◮ U.S. share in world population ◮ U.S. share in global GDP ◮ Prices of light and heavy oil ◮ Others

back September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 32 / 36

slide-33
SLIDE 33

Calibration: Moments used for the U.S.

Data Model Value Source U.S. GDP 1 Normalization Price of fuel pf

1

1 Normalization Total time ¯ L1 1 Normalization Total time allocated to leisure L1 2/3 Literature Household fuel consumption pf

1cf 1

0.022 BEA, EIA Firm fuel use pf

1mf 1

0.022 BEA, EIA Fuel production yf

1

0.965(cf

1 + mf 1)

EIA Total crude oil production yol + yom + yoh 0.35yf

1

EIA Total crude oil input to refineries

  • f

l + of m + of h

2.675(yol + yom + yoh) EIA back September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 33 / 36

slide-34
SLIDE 34

Calibration: Moments used for ROW

Moment Model Value Source ROW share of global oil production 0.927 Eni ROW share of global GDP 0.83 IMF ROW share of world population 0.955 UN Time devoted to leisure L2 (2/3)¯ L2 Literature ROW Refinery gains 1.017 IEA, EIA Ratio of household to firm fuel use cf

2 /mf 2

0.50 EIA, Exxon, WIOD back September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 34 / 36

slide-35
SLIDE 35

Calibration: Parameters

Description Parameter Value Discount factor β 0.96 Substitution parameter ρ, ρa 4 Intertemporal substitution γ −1 Depreciation rate δ 0.10

back September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 35 / 36

slide-36
SLIDE 36

Calibration: Productivity shocks and dynamics

◮ We calibrate the parameters using Simulated Method of Moments ◮ Our data: U.S. and ROW GDP, U.S. and ROW crude oil production ◮ We detrend annual data using an HP filter ◮ We calculate volatilities and first-order autocorrelations for the

detrended data

◮ We treat model deviations from the steady state as de-trended data,

match their moments to the actual data

back September 6, 2017 IAEE Conference Presented by Mine K. Yücel - Dallas Fed 36 / 36