Chapter17 Econometrics SimultaneousEquationsModels
Chapter17 Econometrics SimultaneousEquationsModels
Similar to the classification of variables as explanatory variable and study variable in linear regression
model, the variables in simultaneous equation models are classified as endogenous variables and exogenous
variables.
Exogenous variables help is explaining the variations in endogenous variables. It is customary to include past
values of endogenous variables in the predetermined group. Since exogenous variables are predetermined, so
they are independent of disturbance term in the model. They satisfy those assumptions which explanatory
variables satisfy in the usual regression model. Exogenous variables influence the endogenous variables but
are not themselves influenced by them. One variable which is endogenous for one model can be exogenous
variable for the other model.
Econometrics | Chapter 17 | Simultaneous Equations Models | Shalabh, IIT Kanpur
1
Note that in the linear regression model, the explanatory variables influence the study variable but not vice
versa. So relationship is one sided.
The classification of variables as endogenous and exogenous is important because a necessary condition for
uniquely estimating all the parameters is that the number of endogenous variables is equal to the number of
independent equations in the system. Moreover, the main distinction of predetermined variable in estimation
of parameters is that they are uncorrelated with disturbance term in the equations in which they appear.
Example 1:
Now we consider the following example in detail and introduce various concepts and terminologies used in
describing the simultaneous equations models.
Consider a situation of an ideal market where transaction of only one commodity, say wheat, takes place.
Assume that the number of buyers and sellers is large so that the market is a perfectly competitive market. It
is also assumed that the amount of wheat that comes into the market in a day is completely sold out on the
same day. No seller takes it back. Now we develop a model for such mechanism.
Let
d t denotes the demand of the commodity, say wheat, at time t ,
By economic theory about the ideal market, we have the following condition:
dt = st , t = 1, 2,..., n .
Our aim is to study the behaviour of st , pt and rt which are determined by the simultaneous equation
model.
Since endogenous variables are influenced by exogenous variables but not vice versa, so
• st , pt and rt are endogenous variables.
Now consider an additional variable for the model as lagged value of price pt , denoted as pt −1 . In a market,
generally the price of the commodity depends on the price of the commodity on previous day. If the price of
commodity today is less than the previous day, then buyer would like to buy more. For seller also, today’s
price of commodity depends on previous day’s price and based on which he decides the quantity of
commodity (wheat) to be brought in the market.
So the lagged price affects the demand and supply equations both. Updating both the models, we can now
write that
• demand depends on pt , it and pt −1.
• Exogenous variables: pt −1 , it , rt .
The mechanism of the market is now described by the following set of equations.
• demand dt = 1 + 1 pt + 1t
• supply st = 2 + 2 pt + 2t
• equilibrium condition dt = st = qt
where ' s denote the intercept terms, ' s denote the regression coefficients and ' s denote the
disturbance terms.
These equations are called structural equations. The error terms 1 and 2 are called structural
qt = 1 + 1 pt + 1t (I)
qt = 2 + 2 pt + 2t (II)
So there are only two structural relationships. The price is determined by the mechanism of market and not
by the buyer or supplier. Thus qt and pt are the endogenous variables. Without loss of generality, we can
assume that the variables associated with 1 and 2 are X 1 and X 2 respectively such that
variables.
Each endogenous variable is expressed as a function of the exogenous variable. Note that the exogenous
variable 1 (from X 1 = 1, or X 2 = 1) is not clearly identifiable.
The equations (III) and (IV) are called the reduced form relationships and in general, called the reduced
form of the model.
The coefficients 11 and 21 are called reduced form coefficients and errors v1t and v2t are called the
reduced form disturbances. The reduced from essentially express every endogenous variable as a function
of exogenous variable. This presents a clear relationship between reduced form coefficients and structural
coefficients as well as between structural disturbances and reduced form disturbances. The reduced form is
ready for the application of OLS technique. The reduced form of the model satisfies all the assumptions
needed for the application of OLS.
Suppose we apply OLS technique to equations (III) and (IV) and obtained the OLS estimates of 11 and 12
1 − 2
ˆ11 =
2 − 1
−
ˆ 21 = 2 1 2 1 .
2 − 1
Econometrics | Chapter 17 | Simultaneous Equations Models | Shalabh, IIT Kanpur
5
Note that ˆ11 and ˆ 21 are the numerical values of the estimates. So now there are two equations and four
unknown parameters 1 , 2 , 1 and 2 . So it is not possible to derive the unique estimates of parameters of
the model by applying OLS technique to reduced form. This is known as problem of identifications.
dt = st = qt .
So we have two structural equations model in two endogenous variables ( qt and pt ) and one exogenous
variable (value is 1 given by X 1 = 1, X 2 = 1) . The set of three equations is reduced to a set of two equations as
follows:
Demand : qt = 1 + 1 pt + 1t (1)
Supply: qt = 2 + 2 pt + 2t (2)
Before analysis, we would like to check whether it is possible to estimate the parameters 1 , 2 , 1 and 2 or
not.
where = 1 + (1 − ) 2 , = 1 + (1 − ) 2 , t = 1t + (1 − ) 2t and is any scalar lying between 0 and
1.
Comparing equation (3) with equations (1) and (2), we notice that they have same form. So it is difficult to
say that which is supply equation and which is demand equation. To find this, let equation (3) be demand
equation. Then there is no way to identify the true demand equation (1) and pretended demand equation (3).
A similar exercise can be done for the supply equation, and we find that there is no way to identify the true
supply equation (2) and pretended supply equation (3).
Suppose we apply OLS technique to these models. Applying OLS to equation (1) yields
n
( p − p )( q t t −q)
ˆ1 = t =1
n
= 0.6, say
( p − p)
2
t
t =1
1 n 1 n
where p = t
n t =1
p , q = qt .
n t =1
( p − p )( q − q )
t t
ˆ = t =1
n
= 0.6.
( p − p)
2
t
t =1
Note that ˆ1 and ˆ have same analytical expressions, so they will also have same numerical values, say 0.6.
Looking at the value 0.6, it is difficult to say that the value 0.6 determines equation (1) or (3).
Applying OLS to equation (3) fields
n
( p − p )( q − q ) t t
ˆ2 = t =1
n
= 0.6
( p − p)
2
t
t =1
the value 0.6. Increasing the number of observations also does not helps in the identification of these
equations. So we are not able to identify the parameters. So we take the help of economic theory to identify
the parameters.
The economic theory suggests that when price increases then supply increases but demand decreases. So the
plot will look like
and this implies 1 0 and 2 0. Thus since 0.6 0, so we can say that the value 0.6 represents ˆ2 0
and so ˆ2 = 0.6. But one can always choose a value of such that pretended equation does not violate the
sign of coefficients, say 0. So it again becomes difficult to see whether equation (3) represents supply
equation (2) or not. So none of the parameters is identifiable.
1 − 2
ˆ11 =
2 − 1
− 2 1
ˆ 21 = 1 2 .
2 − 1
There are two equations and four unknowns. So the unique estimates of the parameters 1 , 2 , 1 and 2
cannot be obtained. Thus the equations (1) and (2) can not be identified. Thus the model is not identifiable
and estimation of parameters is not possible.
Suppose a new exogenous variable income it is introduced in the model which represents the income of
buyer at time t . Note that the demand of commodity is influenced by income. On the other hand, the supply
of commodity is not influenced by the income, so this variable is introduced only in the demand equation.
The structural equations (1) and (2) now become
Demand : qt = 1 + 1 pt + 1it + 1t (6)
Supply: qt = 2 + 2 pt + 2t (7)
where 1 is the structural coefficient associate with income. The pretended equation is obtained by
multiplying the equations (6) and (7) by and (1 − ) , respectively, and then adding them together. This is
obtained as follows:
qt + (1 − )qt = 1 + (1 − ) 2 + 1 + (1 − ) 2 pt + 1it + 1t + (1 − ) 2t
or qt = + pt + it + t (8)
Suppose now if we claim that equation (8) is true demand equation because it contains pt and it which
influence the demand. But we note that it is difficult to decide that between the two equations (6) or (8),
which one is the true demand equation.
Suppose now if we claim that equation (8) is the true supply equation. This claim is wrong because income
does not affect the supply. So equation (6) is the supply equation.
Thus the supply equation is now identifiable but demand equation is not identifiable. Such situation is termed
as partial identification.
Econometrics | Chapter 17 | Simultaneous Equations Models | Shalabh, IIT Kanpur
9
Now we find the reduced form of structural equations (6) and (7). This is achieved by first solving equation
(6) for pt and substituting it in equation (7) to obtain an equation in qt . Such an exercise yields the reduced
Applying OLS to equations (9) and (10), we get OLSEs ˆ11 , ˆ 22 , ˆ12 , ˆ 21 . Now we have four equations
((6),(7),(9),(10)) and there are five unknowns (1 , 2 , 1 , 2 , 1 ). So the parameters are not determined
However, here
22
2 =
11
11
2 = 21 − 22 .
22
If ˆ11 , ˆ 22 , ˆ12 and ˆ 21 are available, then ˆ 2 and ˆ2 can be obtained by substituting ˆ ' s in place of ' s. So
2 and 2 are determined uniquely which are the parameters of supply equation. So supply equation is
identified but demand equation is still not identifiable.
Now, as done earlier, we introduce another exogenous variable- rainfall, denoted as rt which denotes the
amount of rainfall at time t . The rainfall influences the supply because better rainfall produces better yield of
wheat. On the other hand, the demand of wheat is not influenced by the rainfall. So the updated set of
structural equations is
Demand : qt = 1 + 1 pt + 1it + 1t (11)
Supply: qt = 2 + 2 pt + 2 rt + 2t . (12)
The pretended equation is obtained by adding together the equations obtained after multiplying equation (11)
by and equation (12) by (1 − ) as follows:
scalar.
Econometrics | Chapter 17 | Simultaneous Equations Models | Shalabh, IIT Kanpur
10
Now we claim that equation (13) is a demand equation. The demand does not depend on rainfall. So unless
= 1 so that rt is absent in the model, the equation (13) cannot be a demand equation. Thus equation (11) is
a demand equation. So demand equation is identified.
Now we claim that equation (13) is the supply equation. The supply is not influenced by the income of the
buyer, so (13) cannot be a supply equation. Thus equation (12) is the supply equation. So now the supply
equation is also identified.
The reduced form model from structural equations (11) and (12) can be obtained which have the following
forms:
pt = 11 + 12it + 13rt + v1t (14)
qt = 21 + 22it + 23rt + v2t . (15)
Application of OLS technique to equations (14) and (15) yields the OLSEs ˆ11 , ˆ12 , ˆ13 , ˆ 21 , ˆ 22 and ˆ 23 . So
now there are six such equations and six unknowns 1 , 2 , 1 , 2 , 1 and 2 . So all the estimates are uniquely
determined. Thus the equations (11) and (12) are exactly identifiable.
Finally, we introduce a lagged endogenous variable pt −1 which denotes the price of the commodity on the
previous day. Since only the supply of wheat is affected by the price on the previous day, so it is introduced
in the supply equation only as
Demand : qt = 1 + 1 pt + 1it + 1t (16)
Supply: qt = 2 + 2 pt + 2 rt + 2 pt −1 + 2t (17)
The pretended equation is obtained by first multiplying equations (16) by and (17) by (1 − ) and then
adding them together as follows:
qt + (1 − )qt = + pt + it + rt + (1 − ) 2 pt −1 + t
or qt = + pt + it + rt + pt −1 + t . (18)
Now we claim that equation (18) represents the demand equation. Since rainfall and lagged price do not
affect the demand, so equation (18) cannot be demand equation. Thus equation (16) is a demand equation
and the demand equation is identified.
The reduced form equations from equations (16) and (17) can be obtained as of the following form:
pt = 11 + 12it + 13rt + 14 pt −1 + v1t (19)
qt = 21 + 22it + 23rt + r24 pt −1 + v2t . (20)
Applying the OLS technique to equations (19) and (20) gives the OLSEs as
ˆ11 , ˆ12 , ˆ13 , ˆ14 , ˆ 21 , ˆ 22 , ˆ 23 and ˆ 24 . So there are eight equations in seven parameters
1 , 2 , 1 , 2 , 1 , 2 and 2 . So unique estimates of all the parameters are not available. In fact, in this case,
the supply equation (17) is identifiable and demand equation (16) is overly identified (in terms of multiple
solutions).
The whole analysis in this example can be classified into three categories –
(exogenous) variables x1 , x2 ,..., xK . Let there are n observations available on each of the variable and there
are G structural equations connecting both the variables which describe the complete model as follows:
11 y1t + 12 y2t + ... + 1G yGt + 11 x1t + 12 x2t + ... + 1K xKt = 1t
21 y1t + 22 y2t + ... + 2G yGt + 21 x2t + 22 x2t + ... + 2 k xK 2 = 2t
This is the reduced form equation of the model where = B −1 is the matrix of reduced form coefficients
and vt = B −1 t is the reduced form disturbance vectors.
If B is singular, then one or more structural relations would be a linear combination of other structural
relations. If B is non-singular, such identities are eliminated.
Assume that t ' s are identically and independently distributed following N (0, ) and Vt ' s are identically
p ( yt | xt ) = p (vt )
t
= p ( t )
vt
= p ( t ) det( B )
t
where is the related Jacobian of transformation and det( B) is the absolute value of the determinant of
vt
B.
Applying a nonsingular linear transformation on structural equation S with a nonsingular matrix D, we get
DByt + Dxt = D t
or S *: B * yt + * xt = t* , t = 1, 2,..., n
where B* = DB, * = D, t* = D t and structural model S * describes the functioning of this model at time
t*
p ( yt | xt ) = p ( t* )
vt
= p ( t* ) det ( B *) .
p ( )
n
L* = det( B*)
n *
t
t =1
n
t
= det( D) det( B) p ( )
n n
t *
t =1 t
n
= det( D) det( B) p ( ) det( D −1
n n
t )
t =1
= L.
Thus both the structural forms S and S * have the same likelihood functions. Since the likelihood functions
form the basis of statistical analysis, so both S and S * have same implications. Moreover, it is difficult to
identify whether the likelihood function corresponds to S and S * . Any attempt to estimate the parameters
will result into failure in the sense that we cannot know whether we are estimating S and S * . Thus S and
S * are observationally equivalent. So the model is not identifiable.
A parameter is said to be identifiable within the model if the parameter has the same value for all equivalent
structures contained in the model.
If all the parameters in the structural equation are identifiable, then the structural equation is said to be
identifiable.
Given a structure, we can thus find many observationally equivalent structures by non-singular linear
transformation.
Econometrics | Chapter 17 | Simultaneous Equations Models | Shalabh, IIT Kanpur
15
The apriori restrictions on B and may help in the identification of parameters. The derived structures may
not satisfy these restrictions and may therefore not be admissible.
The presence and/or absence of certain variables helps in the identifiability. So we use and apply some
apriori restrictions. These apriori restrictions may arise from various sources like economic theory, e.g. it is
known that the price and income affect the demand of wheat but rainfall does not affect it. Similarly, supply
of wheat depends on income, price and rainfall. There are many types of restrictions available which can
solve the problem of identification. We consider zero-one type restrictions.
When the zero-one type restrictions are incorporated in the model, suppose there are G jointly dependent
and K* predetermined variables in S having nonzero coefficients. Rest ( G − G ) jointly dependent and
Without loss of generality, let and * be the row vectors formed by the nonzero elements in the first row
of B and respectively. Thus the first row of B can be expressed as ( 0 ) . So B has G coefficients
Similarly, the first row of can be written as ( * 0 ) . So in , there are K* elements present ( those take
or ( 0 ) yt + ( * 0 ) xt = 1t , t = 1, 2,..., n.
Assume every equation describes the behaviour of a particular variable, so that we can take 11 = 1.
If 11 1, then divide the whole equation by 11 so that the coefficient of y1t is one.
Partition
**
= *
* **
where the orders of * is ( G K* ) , ** is ( G K** ) , * is ( G K* ) and ** is ( G K** )
We can re-express
( 0 ) = − ( * 0 )
( 0 ) = − ( * 0** )
**
or ( 0 ) * = − ( * 0** )
* **
* = − * (i )
** = 0**. (ii )
Assume is known. Then (i) gives a unique solution for * if is uniquely found from (ii). Thus
identifiability of S lies upon the unique determination of . Out of G elements in , one has coefficient
As
** = 0**
or (1 ) ** = 0** .
rank ( ** ) = G − 1.
rank ( ** ) = G − 1.
This is known as a rank condition for the identifiability of parameters in S . This condition is necessary
and sufficient.
Another condition, known as order condition is only necessary and not sufficient. The order condition is
derived as follows:
We now use the result that for any matrix A of order m n , the rank ( A) = Min(m, n). For identifiability, if
rank ( A) = m then obviously n m.
Since ** = 0 and has only (G − 1) elements which are identifiable when
rank ( ** ) = G − 1
K − K* G − 1.
This is known as the order condition for identifiability.
There are various ways in which these conditions can be represented to have meaningful interpretations. We
discuss them as follows:
1. ( G − G ) + ( K − K* ) ( G − G ) + ( G − 1)
or ( G − G ) + ( K − K* ) G − 1
y1t + 12 y2t + ... + 1G yG + 11 x1t + ... + 1K xKt = 1t
Thus if the total number of variables excluded in this equation exceeds (G − 1), then the model is
identifiable.
2. K K* + G -1
Here
K* : The number of predetermined variables present in the equation.
3. Define L = K − K* − (G − 1)
L measures the degree of overidentification.
Note
We have illustrated the various aspects of estimation and conditions for identification in the simultaneous
equation system using the first equation of the system. It may be noted that this can be done for any equation
of the system and it is not necessary to consider only the first equation. Now onwards, we do not restrict to
first equation but we consider any, say ih equation (i = 1, 2, …,G).
Suppose
0 * 0**
B= , = ,
B B * **
where , * , 0 and 0** are row vectors consisting of G , K* , G and K** elements respectively in them.
((G − 1) K* ) and ** is ((G − 1) K** ). Note that B and ** are the matrices of structural coefficients for
the variable omitted from the i th equation (i = 1,2,…,G) but included in other structural equations.
B −1 = ( B −1 B B −1 )
= (I − )
I 0 − ,* − ,**
= .
0 I , − ,* − ,**
If
0 0**
* =
B **
then clearly the rank of * is same as the rank of since the rank of a matrix is not affected by enlarging
0 − ,**
rank ( B −1* ) = rank
I , − ,**
= rank ( I , ) + rank ( ,** )
= ( G − G ) + ( G − 1)
= G −1
rank ( B −1* ) = rank ( * ) = rank ( B ** ) .
So
rank ( B ** ) = G − 1
Note that ( B ** ) is a matrix constructed from the coefficients of variables excluded from that particular
equation but included in other equations of the model. If rank ( ** ) = G − 1, then the equation is
identifiable and this is a necessary and sufficient condition. An advantage of this term is that it avoids the
inversion of matrix. A working rule is proposed like following.
Working rule:
1. Write the model in tabular form by putting ‘X’ if the variable is present and ‘0’ if the variable is
absent.
2. For the equation under study, mark the 0’s (zeros) and pick up the corresponding columns
suppressing that row.
3. If we can choose ( G − 1) rows and (G − 1) columns that are not all zero, then it can be identified.
1 X 0 X X 0 X 2–1=1 2 3–3=0
2 X 0 0 X 0 X 1–1=0 2 3–2=1
3 0 X X X X 0 2–1=1 2 3–3=0
• G = Number of equations = 3.
• `X’ denotes the presence and '0 ' denotes the absence of variables in an equation.
• G = Numbers of ‘X’ in ( y1 y2 y3 ) .
• K* = Number of ‘X’ in ( x1 x2 x3 ) .
Equation (2) 0 0
Equation (3) X X
❖ Check if any row/column is present with all elements ‘0’. If we can pick up a block of order (G − 1)
is which all rows/columns are all ‘0’ , then the equation is identifiable.
0 0
Here G − 1 = 3 − 1 = 2, B = , ** = . So we need ( 2 2 ) matrix in which no row and no
X X
column has all elements ‘0’ . In case of equation (1), the first row has all ‘0’, so it is not identifiable.
Notice that from order condition, we have L = 0 which indicates that the equation is identified and
this conclusion is misleading. This is happening because order condition is just necessary but not
sufficient.
Equation (1) 0X 0
Equation (3) XX X
0 X 0
❖ B = , ** = .
X X X
❖ G − 1 = 3 − 1 = 2.
❖ We see that there is atleast one block in which no row is ‘0’ and no column is ‘0’. So the equation (2)
is identified.
❖ Also, L = 1 0 Equation (2) is over identified.
Equation (1) X X
Equation (2) X X
X X
❖ So B = , ** = .
X X
❖ We see that there is no ‘0’ present in the block. So equation (3) is identified.
❖ Also, L = 0 Equation (3) is exactly identified.
y1t = 12 y2t − ... − 1G yG − 11 x1t − 12 x2t − ... − 1K xK*t + 1t , t = 1, 2,.., n.
Writing this equation in vector and matrix notations by collecting all n observations, we can write
y1 = Y1 + X 1 + 1
where
variables where G denote the number of jointly dependent variables present on the right hand side of the
1
Assume E ( ) = 0, E ( ') = where is positive definite symmetric matrix.
n
The reduced form of the model is obtained from the structural equation model by post multiplying by B −1 as
YBB −1 + X B −1 = B −1
Y = X +V
where = −B −1 and V = B −1 are the matrices of reduced-form coefficients and reduced form disturbances
respectively.
where A = (Y1 , X 1 ) , = ( ) ' and = 1 . This model looks like a multiple linear regression model.
X 1 = XJ1
where J 1 is called as a select matrix and consists of two types of elements, viz., 0 and 1. If the
Step 1: Apply ordinary least squares to each of the reduced form equations and estimate the reduced form
coefficient matrix.
Step 2: Find algebraic relations between structural and reduced-form coefficients. Then find the structural
coefficients.
where yt = ( y1t , y2t ,..., yGt ) ', xt = ( x1t , x2t ,..., xKt ) '.
ˆ = ( X ' X ) X ' Y .
−1
This is the first step of ILS procedure and yields the set of estimated reduced form coefficients.
predetermined (exogenous) variables in the equation and is ( n 1) vector of structural disturbances. Write
this model as
1
(y1 Y1 X1 ) − =
−
or more general
1
−
( y1 Y1 Y2 X1 X 2 ) 0 =
−
0
where Y2 and X 2 are the matrices of observations on ( G − G + 1) endogenous and ( K − K ) predetermined
variables which are excluded from the model due to zero-one type restrictions.
Write
B = −
1
or − = .
0 0
Substitute as ˆ = ( X ' X ) X ' Y
−1
and solve for and . This gives indirect least squares estimators b
These equations (i) and (ii) are K equations is (G + K* − 1) unknowns. Solving the equations (i) and (ii)
2. Two stage least squares (2SLS) or generalized classical linear (GCL) method:
This is more widely used estimation procedure as it is applicable to exactly as well as overidentified
equations. The least-squares method is applied in two stages in this method.
Consider equation y1 = Y1 + X 1 + .
Stage 1: Apply least squares method to estimate the reduced form parameters in the reduced form model
Y1 = X 1 + V1
ˆ1 = ( X ' X ) −1 X ' Y1
Y1 = X ˆ1.
Stage 2: Replace Y1 is structural equation y1 = Y1 + X 1 + by Yˆ1 and apply OLS to thus obtained structural
equation as follows:
y1 = Yˆ1 + X 1 +
= X ( X ' X ) X ' Y1 + X 1 +
−1
= X ( X ' X ) X ' Yˆ1 X1 +
−1
= Aˆ + .
( )
−1
ˆ = Aˆ Aˆ ˆ
Ay1
−1
ˆ Y1' HY1 Y1' X 1 Y1' y1
or = '
ˆ X 1Y1 X 1' X 1 X 1' y1
−1
Y 'Y − Vˆ 'Vˆ Y1' X 1 Y1' − Vˆ1'
= 1 1 ' 1 1 y1.
X 1Y1 X 1' X 1 X 1
'
equations, we get ˆ and ˆ which are the two stage least squares estimators of and respectively.
( ) Ayˆ
−1
ˆ = Aˆ Aˆ 1
ˆ − = ( Aˆ ' Aˆ ) Aˆ '
−1
−1
( )
1
1
n
plim ˆ − = plim Aˆ ' A plim Aˆ ' .
n
( )
Thus plim ˆ- = 0 and so the 2SLS estimators are consistent.
( )
Asy Var ˆ = n −1 plim n ˆ − ˆ − '
( )( )
plim n ( Aˆ Aˆ ) Aˆ ' ' Aˆ ( Aˆ ' Aˆ )
−1 −1
= n −1
−1 −1
1 1 1
= n plim Aˆ ' Aˆ plim Aˆ ' ' Aˆ plim Aˆ ' Aˆ
−1
n n n
−1
1
= n −1 2 plim Aˆ ' Aˆ
n
where Var ( ) = 2 .
where
s2 =
( y − Y ˆ − X ˆ ) ' ( y − Y ˆ − X ˆ )
1 1 1 1
n−G − K