EL ACCESO A LA FINANCIACIÓN DE LOS MERCADOS ......Mar Delgado y Javier J. Pérez Banco de España...

DIRECCIÓN GENERAL DE ECONOMÍA Y ESTADÍSTICA

EL ACCESO A LA FINANCIACIÓN DE LOS MERCADOS POR PARTE DE LAS COMUNIDADES AUTÓNOMAS

Mar Delgado y Javier J. PérezBanco de España

Seminario Banco de España – MEDEBanco de España, Madrid, 28 de abril de 2017

Las opiniones y los análisis contenidos en esta presentación son responsabilidad exclusiva de los autores y, por tanto, no coinciden necesariamente con los del Banco de España o el Eurosistema.

MONITORING THE SPANISH ECONOMY THROUGH THE LENSES OF STRUCTURAL BAYESIAN VARs

Alberto UrtasunBanco de España, Jefe de la División de Análisis Coyuntural y Previsión Económica

Seminario CEMLA: Desarrollos recientes y herramientas para el análisis de la actividad económica

Banco de España, 5-7 de junio de 2019

DIRECCIÓN GENERAL DE ECONOMÍA Y ESTADÍSTICA 2

Introduction

MONITORING THE SPANISH ECONOMY THROUGH THE LENSES OF STRUCTURAL BAYESIAN VARS

Danilo Leiva-Leon

Banco de España

Documentos Ocasionales

N.º 1706

2017


Introduction

• Goal: Provide a suite of empirical models for:

1. Assessing the structural shocks that mainly affect the Spanish economy over time

2. Producing forecasts of GDP and inflation conditional on assumptions about key macroeconomic variables

• Tool: Suite of Structural Vector Autoregressions (SVAR) , estimated with Bayesian methods, which allows us to:

1. Impose restrictions that help to identify in a relatively accurate way key macroeconomic structural shocks

2. Easily provide robust and detailed inference about conditional forecasts obtained from the suite of models


Outline

1. Suite of Models

2. Results

2.1. Structural Analysis

2.2. Forecasts


Outline

1. Suite of Models

2. Results


2.2. Forecasts


1. BENCHMARK

REAL GDPCORE INFLATION

OIL PRICEINTEREST RATE

Suite of Bayesian VAR Models


2. LABORBENCHMARK VARIABLES

+UNEMPLOYMENT RATE

WAGES

3. FISCALBENCHMARK VARIABLES

+GOVERNMENT EXPENDITURE

GOVERNMENT REVENUE

5. INTERNATIONALBENCHMARK VARIABLES

+GLOBAL DEMANDEXCHANGE RATE

1. BENCHMARK


OIL PRICEINTEREST RATE 4. FINANCIAL

BENCHMARK VARIABLES+

STOCK RETURNSUNCERTAINTY INDEX




+UNEMPLOYMENT RATE

WAGES



GOVERNMENT REVENUE



1. BENCHMARK






ESTRUCTURAL ANALYSIS:Narrative of the contribution of shocks



+UNEMPLOYMENT RATE

WAGES



GOVERNMENT REVENUE



1. BENCHMARK






6. ONE-FITS-ALL



+

UNEMPLOYMENT RATEWAGES

+

GOVERNMENT EXPENDITUREGOVERNMENT REVENUE

+


+

GLOBAL DEMANDEXCHANGE RATE




+UNEMPLOYMENT RATE

WAGES



GOVERNMENT REVENUE



1. BENCHMARK






6. ONE-FITS-ALL



+

UNEMPLOYMENT RATEWAGES

+

GOVERNMENT EXPENDITUREGOVERNMENT REVENUE

+


+

GLOBAL DEMANDEXCHANGE RATE


FORECASTS: conditionalon assumptions


Identification of Shocks: Benchmark Model

11DG ECONOMICS, RESEARCH AND STATISTICS

• SBVAR mode for the Spanish economy

• Sign and Exclusion Restrictions: The reduced form innovations are a function of the structural innovations

��,��,��,��,�

=

+ − − ∗

+ + − ∗

+ + + ∗

0 0 0 ∗

��,�

��,��,�

��,�

�� !�"�

= # +Φ(")

��'��'�� '!�"��'

+

��,��,��,��,�






��,��,��,��,�

=

+ − − ∗

+ + − ∗

+ + + ∗

0 0 0 ∗

��,�

��,��,�

��,�

�� !�"�

= # +Φ(")

��'��'�� '!�"��'

+

��,��,��,��,�


Identification of Shocks: Labor Model


• The (i) unemployment rate and (ii) wages are incorporated to the Benchmark model

• The assumed restrictions follow the line of Peersman and Straub (2009), and Foroni et al. (2015)

��,��,��,��,��(��,��,�

=

+ − − + ∗ ∗

+ + − − ∗ ∗

+ + + ∗ ∗ ∗

∗ ∗ ∗ ∗ ∗ ∗

∗ ∗ ∗ − ∗ ∗

0 0 0 0 0 ∗

��,�

��,��,�

)*+,�-./,0�(��,��,�


Identification of Shocks: Fiscal Model


• The (i) government revenues and (ii) government expenditures are incorporated to the Benchmark model

• The assumed restrictions follow the line of Mountford y Uhlig (2009)

��,��,��,��1�,��2�3,��,�

=

+ − − + − ∗

+ + − ∗ ∗ ∗

+ + + ∗ ∗ ∗

∗ ∗ ∗ + ∗ ∗

∗ ∗ ∗ ∗ + ∗

0 0 0 0 0 ∗

��,�

��,��,�

)45,�67/,0)458�968,0��,�


Identification of Shocks: Financial Model


• The (i) stock returns and (ii) a financial uncertainty index are incorporated to the Benchmark model

• Gil, Pérez and Urtasun (2016) have documented significantly negative effects of uncertainty on real activity

��,��,��,��,��,��,�

=

+ − − ∗ − ∗

+ + − ∗ ∗ ∗

+ + + ∗ ∗ ∗

∗ ∗ ∗ ∗ − ∗

∗ ∗ ∗ ∗ + ∗

0 0 0 0 0 ∗

��,�

��,��,�

��2��,�).:;690,0��,�


Identification of Shocks: International Model


• A measure of global demand and the Dollar-Euro exchange rate are incorporated to the Benchmark model

• Sign restriction of the Benchmark model are kept, and exclusion restrictions are added

��,��,��,��,��12,��,�

=

+ − − ∗ ∗ ∗

+ + − ∗ + ∗

+ + + ∗ ∗ ∗

0 0 0 ∗ ∗ ∗

0 0 ∗ 0 + ∗

0 0 0 0 0 ∗

��,�

��,��,�

)4*5�<6=,0

)679�9+06,0��,�


Esquema

1. Suite of Models

2. Results


2.2. Forecasts


GDP Historical Decomposition: Benchmark Model


-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

Demand Shocks Cost-push Shocks

Monetary Policy Shocks Oil Price Shocks




-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8



Labor MarketReforms




-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8



Monetary Union




-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8



Housing boom,Great Recession




-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8



Period of“Lowflation”




-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8



Fall in Oil Pirces




-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8



Monetary Union

Housing boom,Great Recession

Labor MarketReforms

Fall in Oil Pirces


-0.8

-0.3

0.2

0.7

GDP Historical Decomposition: Labor Model


Labor Supply Shocks

Cost-push Shocks (Labor vs. Benchmark)

Monetary Policy Shocks (Labor vs. Benchmark) Oil Price Shocks (Labor vs. Benchmark)

-0.8

-0.3

0.2

0.7

-0.8

-0.3

0.2

0.7

-0.8

-0.3

0.2

0.7

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

Labor ReformLaw 10/ 1994 High Unemployment

Levels

Demand Shock (Labor vs. Benchmark)


GDP Historical Decomposition: Fiscal Model


Government Expenditure and Revenue Shocks

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8Exp-Gov-Shock

Inc-Gov-Shock

-0.8

-0.3

0.2

0.7

-0.8

-0.3

0.2

0.7

-0.8

-0.3

0.2

0.7

-0.8

-0.3

0.2

0.7

Sovereign DebtCrisis

Cost-push Shocks (Fiscal vs. Benchmark)

Monetary Policy Shocks (Fiscal vs. Benchmark) Oil Price Shocks (Fiscal vs. Benchmark)

Demand Shock (Fiscal vs. Benchmark)


GDP Historical Decomposition: International Model


International Shocks

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

-0.8

-0.3

0.2

0.7

-0.8

-0.3

0.2

0.7

-0.8

-0.3

0.2

0.7

-0.8

-0.3

0.2

0.7

Great Recession in Advanced Economies

Cost-push Shocks (International vs. Benchmark)

Monetary Policy Shocks (International vs. Benchmark) Oil Price Shocks (International vs. Benchmark)

Demand Shock (International vs. Benchmark)


GDP Historical Decomposition: Financial Model


Uncertainty Shocks

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

-0.8

-0.3

0.2

0.7

-0.8

-0.3

0.2

0.7

-0.8

-0.3

0.2

0.7

-0.8

-0.3

0.2

0.7

High levels of uncertainty“Double-dip recession”

Cost-push Shocks (Financial vs. Benchmark)

Monetary Policy Shocks (Financial vs. Benchmark) Oil Price Shocks (Financial vs. Benchmark)

Demand Shock (Financial vs. Benchmark)


GDP Historical Decomposition: Models by Block


Uncertainty Shocks

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

High levels of uncertainty“Double-dip recession”

Labor Supply Shocks

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

Labor ReformLaw 10/ 1994

High UnemploymentLevels

Government Expenditure and Revenue Shocks

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8Exp-Gov-Shock Inc-Gov-Shock

Sovereign DebtCrisis

International Shocks

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

Great Recession in Advanced Economies


Esquema

1. Suite of Models

2. Estimation Algorithms

3. Results


3.2. Forecasts


Forecasts: conditional and unconditional variables


BENCHMARK FISCAL LABOR INTERNATIONAL FINANCIAL ONE-FITS-ALL

REAL GDP * * * * * *CORE INFLATION * * * * * *INTEREST RATE X X X X X X

GOVERNMENT EXPEDITURE

X X

GOVERNMENT REVENUE

X X

UNEMPLOYMENT RATE * *WAGE * *GLOBAL DEMAND

X X

EXCHANGE RATE X XSTOCK RETURNS X X

UNCERTAINTY * *OIL PRICE X X X X X X

Nota. * denotes unconditional variables and X denotes conditional variables


Forecasts: Benchmark Model


Growth Rate of Real GDP

2000 2005 2010 2015 2020

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

Proyección BdEPredicción modelo

• The forecasts of GDP from the Benchmark model tend to be slightly lower than the BdE projections


Forecasts: Benchmark Model



2000 2005 2010 2015 2020

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

2000 2005 2010 2015 2020

-0.2

0

0.2

0.4

0.6

0.8

1

1.2

Core Inflation Rate



• The forecasts of GDP from the Benchmark model tend to be slightly lower than the BdE projections

• The forecasts of inflation from the Benchmark model tend to be significantly lower thanthe BdE projections


Forecasts: Models by Block



• The forecasts of GDP from most of the models by blocks are aligned with the BdEprojections. Although, there is certain heterogeneity.

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

1.00

Benchmark Finance International

Labor Fiscal BdE


Forecasts: Models by Block


Growth Rate of Real GDP Core Inflation Rate

• The forecasts of GDP from most of the models by blocks are aligned with the BdEprojections. Although, there is certain heterogeneity.

• For the case of inflation, the forecasts of all the models by blocks point to a significantly lower inflation than the BdE projections.

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

1.00


Labor Fiscal BdE

0.00

0.10

0.20

0.30

0.40

0.50

0.60


Labor Fiscal BdE


Forecasts: One-Fits-All Model (“Agnostic” priors)




• Once it is conditioned on all the assumptions, the One-Fits-All model produces forecasts that are more optimistic than the projections of the BdE



Forecasts: One-Fits-All Model (“Agnostic” priors)





• Once it is conditioned on all the assumptions, the One-Fits-All model produces forecasts that are more optimistic than the projections of the BdE

• The forecats of inflation from the One-Fits-All model follow a similar path to the BdEprojections




Forecasts: One-Fits-All Model (“Mimic” priors)


2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020

-0.2

0

0.2

0.4

0.6

0.8

1

1.2




• Once the model uses “mimic” priors and conditions on all the assumptions, theforecasts of GDP from the One-Fits-All model are very similar to the projections of theBdE

• The forecats of inflation from the One-Fits-All model also follow a similar path to theBdE projections


Forecast Uncertainty: Benchmark Model


-3 -2 -1 0 1 2 3

0

0.1

0.2

0.3

0.42017:Q2

0.63

0.87

-3 -2 -1 0 1 2 3

0

0.1

0.2

0.3

0.42017:Q3

0.56

0.74

-3 -2 -1 0 1 2 3

0

0.1

0.2

0.3

0.42017:Q4

0.56

0.68

-3 -2 -1 0 1 2 3

0

0.1

0.2

0.3

0.42018:Q1

0.50

0.56


Nota. Numbers in blue indicate the forecasts from the Benchmark model, whilenumbers in black indicate the BdE projections


Forecast Uncertainty: One-Fits-All Model


-3 -2 -1 0 1 2 3

0

0.1

0.2

0.3

0.42017:Q2

0.91

0.87

-3 -2 -1 0 1 2 3

0

0.1

0.2

0.3

0.42017:Q3

0.77

0.74

-3 -2 -1 0 1 2 3

0

0.1

0.2

0.3

0.42017:Q4

0.67

0.68

-3 -2 -1 0 1 2 3

0

0.1

0.2

0.3

0.42018:Q1

0.67

0.56


Nota. Numbers in red indicate the forecasts from the One-Fits-All model, whilenumbers in black indicate the BdE projections


Forecast Uncertainty: Benchmark vs. One-Fits-All

Model



Nota. Numbers in blue (red) indicate the forecasts from the Benchmark (One-Fits-All)

model, while numbers in black indicate the BdE projections

-3 -2 -1 0 1 2 3

0

0.1

0.2

0.3

0.42017:Q2

0.63

0.91

0.87

-3 -2 -1 0 1 2 3

0

0.1

0.2

0.3

0.42017:Q3

0.56

0.77

0.74

-3 -2 -1 0 1 2 3

0

0.1

0.2

0.3

0.42017:Q4

0.56

0.67

0.68

-3 -2 -1 0 1 2 3

0

0.1

0.2

0.3

0.42018:Q1

0.50

0.67

0.56


Forecast Uncertainty: Benchmark Model


Core Inflation Rate

Nota. Numbers in blue indicate the forecasts from the Benchmark model, whilenumbers in black indicate the BdE projections

-3 -2 -1 0 1 2 3

0

0.1

0.2

0.3

0.42017:Q2

0.17

0.87

-3 -2 -1 0 1 2 3

0

0.1

0.2

0.3

0.42017:Q3

0.17

0.74

-3 -2 -1 0 1 2 3

0

0.1

0.2

0.3

0.42017:Q4

0.28

0.68

-3 -2 -1 0 1 2 3

0

0.1

0.2

0.3

0.42018:Q1

0.24

0.56


Forecast Uncertainty: One-Fits-All Model


Core Inflation Rate

Nota. Numbers in red indicate the forecasts from the One-Fits-All model, whilenumbers in black indicate the BdE projections

-3 -2 -1 0 1 2 3

0

0.1

0.2

0.3

0.42017:Q2

0.50

0.87

-3 -2 -1 0 1 2 3

0

0.1

0.2

0.3

0.42017:Q3

0.52

0.74

-3 -2 -1 0 1 2 3

0

0.1

0.2

0.3

0.42017:Q4

0.37

0.68

-3 -2 -1 0 1 2 3

0

0.1

0.2

0.3

0.42018:Q1

0.37

0.56


Forecast Uncertainty: Benchmark vs. One-Fits-All

Model


Core Inflation Rate

Nota. Numbers in blue (red) indicate the forecasts from the Benchmark (One-Fits-All)

model, while numbers in black indicate the BdE projections

-3 -2 -1 0 1 2 3

0

0.1

0.2

0.3

0.42017:Q2

0.17

0.50

0.87

-3 -2 -1 0 1 2 3

0

0.1

0.2

0.3

0.42017:Q3

0.17

0.52

0.74

-3 -2 -1 0 1 2 3

0

0.1

0.2

0.3

0.42017:Q4

0.28

0.37

0.68

-3 -2 -1 0 1 2 3

0

0.1

0.2

0.3

0.42018:Q1

0.24

0.37

0.56


Conclussions

• The proposed suite of models provide a narrative of the contribution of fundamental economic shocks that agrees with main historic events of the Spanish economy.

• The proposed SBVAR models are used to provide forecasts of output and inflation conditional on different scenarios about the development of key macroeconomic variables.

• The models provide conditional forecasts that are significantly aligned to the BdE projections

• Therefore, the suite could be incorporated to the toolkit of quantitative models that the Banco de España uses to perform forecasts.


THANK YOU.

EL ACCESO A LA FINANCIACIÓN DE LOS MERCADOS ......Mar Delgado y Javier J. Pérez Banco de España...

Documents

Transcript of EL ACCESO A LA FINANCIACIÓN DE LOS MERCADOS ......Mar Delgado y Javier J. Pérez Banco de España...