Simultaneous Equation Models

Lekshmi krishna M.R(100609) M.Tech. |Technology Management Department of Futures Studies University of Kerala Lekshminair.k@gmail.com

Introduction
The recognition in economics everything depends on everything else Extensive research Simultaneous Equation Models (SEM)

System of simultaneous equations

System describing the joint dependence of variables If we use ordinary least squares in an equation of the simultaneous system the estimate will be biased Simultaneous equation bias

Variables in SEM
Simultaneous Equation Models

Endogenous Variables (Interaction with other Variables)

Exogenous Variables (Independent)

Lagged Variables (Similar- exogeneous)

Predetermined Variables

Example of an SEM
D (t) = a1 + a2 P (t) + a3y(t) + u1(t) S (t) = b1 + b2 P (t 1) + b3 P*(t) + b4 R (t) + b5 r (t) + u2(t) D (t) = S (t) D (t) S (t) P (t) P*(t) y (t) R (t) r (t) u1 & u2 Domestic Demand Domestic Supply Domestic Price International Price Real National GDP Rainfall index Cost of rural credit Disturbances

Agricultural commodity

 Endogenous Variables : D (t) ,S (t) ,P (t) Exogenous Variables : R(t) ,y(t) ,P*(t) ,r(t) Predetermined :R(t),y(t),P*(t),r(t) & P(t-1)

Notations
Suppose we have T observation M endogenous variable T observation - K predetermined variables Assumption : Number of equation is SEM=Number of endogenous variables

 This enables us to write Y +XB = U Where X =[ x1.xk] Y= [y1..ym] ij = Coefficient of the ith endogenous variable in the jth equation ij = Coefficient of the ith predetermined variable in the jth equation

Identification : Basic Concepts

A parameter of the structural form of an SEM is said to be identified if its value can be deduced from knowledge of the likelihood function of the model An equation of the structural form is identified if all its parameters are identified The structural form as a system is identified if every equation of the system is identified

Reduced form
Endogenous variables are expressed as a function of predetermined variables only. yi = i1x1 +i2x2+i3x3+..+ikxk+vi i =Reduced form coefficients

 If: M = number of endogenous variable K = number of variables missing from the equation (both endogenous and exogenous) IDENTIFIED K = M 1 OVERIDENTIFIED K > M 1 UNDERIDENTIFIED K < M 1

IMPLICATIONS OF IDENTIFICATION
If an equation (model) is under-identified it is impossible to estimate all its parameters with any econometric technique.

If an equation (model) is identified in general its coefficients can be estimated. The appropriate estimation technique will depend upon whether it is exactly identified or over-identified.

Check for model estimate

Order Condition : Necessary but not sufficient condition for identification Rank Condition : Necessary and sufficient condition for identification

Order Condition
For an equation to be identified the total number of variables excluded from it must be equal to or greater than the number of endogenous variables in the model less one.

G = total number of equations (total number of endogenous variables. K= total number of variables in the model (endogenous and pre-determined). M = number of variables, endogenous and pre-determined, in a particular equation.

K M G-1

Rank Condition
G* = number of endogenous variables contained in a particular equation. An equation containing G* endogenous variables is identified if it is possible to construct at least one non-zero determinant of order (G*-1) from the reduced form coefficients of the exogenous variables excluded from that particular equation.

Solutions for the Simultaneous Equations

Indirect least squares (ILS) Method of instrumental variable (IV) Limited information maximum likelihood (LIML) Three stage least square (3SLS) Full information maximum likelihood (FIML)

Thank you!!!