UNIVERSITY OF CUENCA

FACULTY OF ECONOMIC AND ADMINISTRATIVE SCIENCES

ECONOMY

SIMPLE KEYNESIAN MODEL APPLIED TO THE ECONOMY

ECUADOREAN
PERIOD

1970–2015
SYSTEM OF SIMULTANEOUS EQUATIONS
 SIMPLE KEYNESIAN MODEL: SIMULTANEOUS EQUATIONS ECUADORIAN ECONOMY

LITERATURE REVIEW

International Literature

SIMULTANEOUS EQUATIONS

One of the documents analyzed for this work is a research related to the
COMPUTATIONAL ASPECTS OF THE RESOLUTION AND OBTAINING OF EQUATION MODELS
Simultaneous" (Lopéz Espín, 2009). This research observes that the authors intend
analyze logarithms for the construction of simultaneous equation models and also
pretend to compare them both in sequential and parallel. This also shows the decomposition
matrix to accelerate basic algorithms in sequential taking them to the version
paralela. Su principal objetivo es solucionar modelos de ecuación simultáneas de manera eficaz.
As well as developing algorithms that run on high-performance systems, with the
the purpose of algorithms being as portable as possible and being able to be executed on
various systems. This research is composed of 9 chapters in which it develops the
different parts that make up the modeling of these equations. The author arrives at different
Conclusions highlight that the algorithm based on MCI continues to have
lower cost than this new algorithm based on inverse decomposition so that it
Propose a general resolution scheme, where MCI is used in the equations exactly.
identified and the MC2E algorithm based on the inverse decomposition in the super
identified. In cases where this algorithm cannot be used, the basic algorithm will be used.
of MC2E based on its expression. In addition to the decomposition of the inverse described, it has
used QR decomposition to propose new algorithms based on MC2E. They have
developed two new schemes, one based solely on QR decomposition by the
Householder's reflection method and another one that, in addition to this method, uses rotations
from Givens to re-triangulate the matrix in each equation.

MACROECONOMIC CONSISTENCY MODEL FOR ECUADOR

Another document analyzed for this work is a research article related to

"MACROECONOMIC CONSISTENCY MODEL FOR ECUADOR" (Pérez & Samaniego). The
researchers from the economic research and development department are proposing a
theoretical economic consistency model applied to programming needs
economic, whose main objective is to carry out the annual projection of a set of variables that
they allow to determine and know the performance of the economy as well as to establish as a core
the core of the model is a set of equations that describe the behavior of the sector
private and external sector, particularly investment decisions, and the evolution of the
 SIMPLE KEYNESIAN MODEL: SIMULTANEOUS EQUATIONS ECUADORIAN ECONOMY

imports, exports, and services. Regarding the equation, the model allows
choose four adjustment variables from a set of eight possible ones. Once selected,
endogenous variables and then situate the remaining ones as exogenous, it is necessary to solve a
linear system of equations with four unknowns. In which two projections are presented
calculated based on a closed rule and two scenarios for the evolution of the exogenous variables.
And the conclusions they reach are that this work serves as a tool for knowledge.
about equations and how this type of equations could be formulated to be able to understand
How could the country's economy be modeled.

THEORETICAL FRAMEWORK

Next, we will see some concepts.

The simultaneous equations model is a statistical model given by a set of

linear equations. The modeling of this type of variables must consider more structures.
complex that collect the influence of the endogenous variable on the predetermined and the
relationships that may exist among these. This type of models, formed by more than one
equation, is named systems of simultaneous equations or multiequational models.
They specify several equations in which variables that are explained in one of the
equations may appear as explanatory in one or more others. (Bernardo Pena Trapero, 1999)

The most important method for estimating simultaneous equation models is the method of variables.
instrumentals. Therefore, the solution to the problem of simultaneity in essence is the same.
that the VI solutions to omitted variable problems and measurement error problems.
(Wolldridge)

Model Identification
The identification of a simultaneous equations model consists of knowing, based on a
set of sample observations, which allows the estimation of the reduced form, can be
estimate the parameters of the structural form of the model. If this is possible, it is considered that the
The equation is identified; otherwise, the equation is not identified.

Order Condition
To apply this condition, the number of variables, both endogenous, is compared to an equation.
as predetermined, excluded in the equation with the number of equations of the system minus
a (G-1). (Bernardo Pena Trapero, 1999)

If G∆∆ +K**G-1, the equation is not identified.

If G∆∆ +K**=G-1, the equation is exactly identified.
If G∆∆ +K**=G>1, the equation is over-identified.

Being:
 SIMPLE KEYNESIAN MODEL: SIMULTANEOUS EQUATIONS ECUADORIAN ECONOMY

G∆∆The number of excluded endogenous variables.

K**The number of excluded defaults
G: The number of equations in the system.

Range Condition
The application of this condition requires obtaining the coefficient matrix of the structural form.
what is defined by the following expression:

= ( | ´)

From it, the rank of the submatrix of the excluded variables is calculated (A∆∆ A**in the equation
what needs to be identified. This submatrix is formed by the coefficients that in the other
equations take the mentioned excluded variables. In the range of this submatrix it must
compare it with the number of equations in the system, minus one.

If p(A∆∆ A**)<G∆-1, the equation is not identified

If p(A∆∆ A**=G∆-1, the equation is exactly identified
If p(A∆∆ A**G∆-1, the equation is over-identified.

Estimation of the System of Equations

Estimation methods in multiequation models are classified into two main groups:

a) Most commonly used methods with limited information:

Direct Least Squares (DLS)
Indirect least squares (ILS)
Two-Stage Least Squares (2SLS)
Maximum likelihood with limited information
b) Most commonly used methods with complete information
Three-stage least squares (TSLS)
Maximum likelihood with complete information

The most important thing, from the perspective of economic relations, is the estimation of the
parameters of the structural form. However, as a previous step of this estimation, it
they calculate the estimators of the reduced form. For this, and since the regressors of the form
reduced are independent of the disturbance variables, the least squares method is used
ordinary squares.

Types of simulation:
Static simulation.
Dynamic simulation.

There are two types of simulation, but we will focus on dynamics due to analyzing this type.
Simulation is our study objective.

Dynamic Simulation.
 SIMPLE KEYNESIAN MODEL: SIMULTANEOUS EQUATIONS ECUADOREAN ECONOMY

Characteristics:

Uses observed values for exogenous variables and for the included endogenous ones with
backlogs.
After the initial year, the values of the current and lagged endogenous variables are
calculated through simulation.

Conditions or properties of dynamic models.

It depends on the quality of the estimation.
In behavioral equations, it depends on the structure of the model itself, of the
correct application of statistical techniques for estimation, and transformation
correct algebraic.
Finally, the correct specification of the variables in the models.
 SIMPLE KEYNESIAN MODEL: SIMULTANEOUS EQUATIONS ECUADORIAN ECONOMY

MODEL OF SIMULTANEOUS EQUATIONS

SIMPLE KEYNESIAN MACROECONOMIC MODEL

METHODOLOGY

SYSTEM OF SIMULTANEOUS EQUATIONS:

= + +
= + + +
= + +
Where:
Ct: Aggregate Consumption
Total Investment
Gross Domestic Product
Total Government Spending

The first equation models a consumption function, the second equation models a function of
investment and the last equation is the condition for product equilibrium in the market

CONSIDERATIONS OF THE VARIABLES:

Ct
It Endogenous Variables
Yt

Variables Exógenas = Variables Predeterminadas

Gt
Gt-1 Exogenous current: Gt
Delayed Exogenous: Gt-1

1. Substituting the identity equation into the equations we have that:

(1) = + + +
(2) = + + + − +
 SIMPLE KEYNESIAN MODEL: SIMULTANEOUS EQUATIONS ECUADORIAN ECONOMY

FORMULATION OF THE STRUCTURAL FORM OF THE SIMPLE KEYNESIAN MODEL

To propose the structural form of the model, considering N endogenous variables and K
predetermined, we have:

+ + =0

WHERE:

Vector corresponding to the t-th observation of N

: Vector of the values of k predetermined variables corresponding also to the
t-th observation
: Vector of N perturbation terms
Being the matrices that contain the coefficients of the endogenous variables and
predetermined, respectively.

To obtain the structural form we move all the variables to one side.

− + + + + =
− + + + + − + =

2. Rewrite in matrix form to obtain the structural form:

F0 H0
[ ] [ -1 h 1̄
f(a) =
] + [1
-1
Gg Gg- 1 Huh2H2
[ ] + [fan1 ]=0
fan2
0 H3
3. Identifying:

=[ ]

= [1 Go Go- 1 ]

= [Fan 1 is ]running, Fan 2 is running.

−1 left parenthesis 1
=[ ]
Fae1 -1
Hae0 Hae0
= [ Hae2 Hwag2
]
0 H3

4. Therefore, with this identification we obtain the structural form of the model.

+ + =0
 SIMPLE KEYNESIAN MODEL: SIMULTANEOUS EQUATIONS ECUADORIAN ECONOMY

FORMULATION OF THE REDUCED FORM OF THE SIMPLE KEYNESIAN MODEL

It consists of expressing the current endogenous variables in terms of the predetermined variables.

= + , with =− -1

1. Mathematically it is given:

f0 h0
-1 -β1
= − [ F2 H2] ∗ [ ]
-a1 -1
0 H3

α0 + α1β0α0β1 + β0
= [ α2 + α1β2a2b1 + b2 ]
Hae1Be3 F3

α0 + α1β0a0b1 + b0
[ ] = [1 Gg Gg- 1 ] [ 2 + H1B2 A2B1 + B2 ] + [Fen1 ] A
A Fen2
Fae1Fae3 H3

2. By rewriting the equations, we obtain the reduced form of the model:

In these equations, the endogenous variables are expressed as a function of the variables
predetermined.

=(α0 + α1β0 + ) F2+( 1 2 )+ 1 ( 3 )- 1 +Fan1

=( a0b1 + b0 +) F2B1+
( H2 )+ F3G0 - 1 + ε2G0
 SIMPLE KEYNESIAN MODEL: SIMULTANEOUS EQUATIONS ECUADORIAN ECONOMY

ANALYSIS OF IDENTIFICATION

IDENTIFICATION WITH NULLITY RESTRICTIONS

A general rule for identifying an equation would be:

− < − Overidentified equation

− = − EXACTLY IDENTIFIED EQUATION
− > − SUBIDENTIFIED EQUATION

1. Establish the matrix A

=[]

-1 H1
Fare 1 -1
=
0 0
Hae2 Hae2
[0 H3 ]

Where:
N=2
K=5

Therefore for the first equation:

N=2 N-1= 2-1= 1

K=5 k1 = 2 K- k1= 5-2= 3

− > −
5–2> 2–1
3 > 1 IDENTIFIED EQUATION

For the second equation:

N=2 N-1= 2-1= 1

K=5 k2 = 3 K- k1= 5-3= 2

− > −
5–3> 2–1
2 > 1 IDENTIFIED EQUATION
 SIMPLE KEYNESIAN MODEL: SIMULTANEOUS EQUATIONS ECUADORIAN ECONOMY

IDENTIFICATION UNDER ORDER CONDITION

= 0+ 1 + 1
= 0+ 1 + 2 -1 + 2
= + +

M = Endogenous variables in the model: 3 (Ct It Yt)

K = Predetermined variables in the model: 2 (Gt Gt-1)

For the first equation:

− > −
3-1> 2–0
1 > 0 OVERIDENTIFIED
For the second equation:

− = −
3–1= 2–1
1 = 1 EXACTLY IDENTIFIED

ELECTION OF THE ESTIMATION METHOD

CONCLUSION: with this information it can be deduced that the method to be used is the OLS model,
and MCO2S.

For academic purposes, we will use our model to analyze the simultaneous equations.
through MCO, MCO2S, and MCO3S, a comparative table of the will be presented next
estimates.
 SIMPLE KEYNESIAN MODEL: SIMULTANEOUS EQUATIONS ECUADORIAN ECONOMY

PREDICTED VALUES

The following will be an analysis of the forecasts of the variables we used in order to.
to reduce uncertainty about the future of the Ecuadorian economy.
In the following graphs we can observe the behavior of the predicted values of the
macro variables versus the actual values of the variables of Ecuador.

GRAPH # 1

OWN PREPARATION.- EVIEWS

In graph #1, we will appreciate the behavior of the real values of the consumption log.
to the predicted values, which exhibit the same behavior and the same trend
adjusting the forecasted values to the actual values.

GRAFICO # 2

OWN ELABORATION.- EVIEWS

In graph #2, we observe the behavior of the actual values of the Investment against the
predicted values of the same variable, here it can be seen that they do not exhibit the same trend,
es decir la curva de la Inversion observada tiene decrecimientos fuertes ciclicos a lo largo de la
 SIMPLE KEYNESIAN MODEL: SIMULTANEOUS EQUATIONS ECUADORIAN ECONOMY

The history of study and the predicted inversion curve has constant growth during the
analysis period, here the predicted values do not resemble the actual values.

MODEL SIMULATION

To carry out the simulation of the model, we start from the reduced function obtained earlier,
Then, with the parameters obtained through the MC2E estimation, we proceed to obtain
our " dear ones, the same ones that will be used to obtain the simulated values.

ANALYSIS OF THE COEFFICIENTS OBTAINED FROM THE SIMULATION

To carry out this analysis, we have considered estimates by OLS, MC2E, and MC3E, for
find the coefficients of the reduced form of the Simple Keynesian model, as it is
show in the following tables.

COMPARATIVE TABLES OF ESTIMATES

OWN CREATION.- EVIEWS

 SIMPLE KEYNESIAN MODEL: SIMULTANEOUS EQUATIONS ECUADORIAN ECONOMY

GLOBAL ANALYSIS OF THE SIMULATED MODEL

Because our chosen estimation method is an MC2E, which allows parameters

consistent and efficient, we present the estimated reduced equation with the coefficients
calculated with the matrix .

α0 + α1β0f0b1 + b0
= [ α2 + α1β2A2B1 + B2 ]
Hwa1Hwa3 F3

MATRIX OF ESTIMATED VALUES

0.95839972 0.4172435
= [ 0.17176072 0.16125476 ]
0 0

FORMULATION OF THE EQUATIONS IN REDUCED FORM

C0 = 0.95839972 + 0.17176072G0 +F1

I₀ = 0.4172435 + 0.16125476G₀ + ε₂₀

NORMALITY TEST OF THE GLOBAL OLS MODEL

OWN PREPARATION.- EVIEWS

SIMPLE KEYNESIAN MODEL: SIMULTANEOUS EQUATIONS ECUADORIAN ECONOMY

NORMALITY TEST OF THE GLOBAL MODEL

MC2E

OWN PREPARATION.- EVIEWS

GLOBAL MODEL NORMALITY TEST

MC3E

OWN PREPARATION.- EVIEWS

 SIMPLE KEYNESIAN MODEL: SIMULTANEOUS EQUATIONS ECUADORIAN ECONOMY

Bibliography
Bernardo Pena Trapero, J. E. (1999).Cien ejercicios de Econometria .Piramide.

Lopéz Espín, J. J. (DECEMBER 2009). UNIVERSITY OF MURCIA. Obtained from

[Link]

Pérez, W., & Samaniego, P. (n.d.). CENTRAL BANK OF ECUADOR. Retrieved from
[Link]
[Link].

Wolldridge, J. M. (s.f.).Introducion ala Econometria un Enfoque [Link]: Cengage

Learning.