Professional Documents
Culture Documents
Multivariate models
• All the models we have looked at thus far have been single equations models of the
form y = X + u
• All of the variables contained in the X matrix are assumed to be EXOGENOUS.
• y is an ENDOGENOUS variable.
An example from economics to illustrate - the demand and supply of a good:
(1) Qdt Pt St ut
(2)
Qst Pt Tt vt
(3)
Qdt Qst
where = quantity
Qdt of the good demanded
= quantity of the good supplied
Qst
St = price of a substitute good
Tt = some variable embodying the state of technology
• Assuming that the market always clears, and dropping the time subscripts for
simplicity
(4) Q P S u
(5)
Q P T v
• Solving for Q,
P S u P T v (6)
• Solving for P,
Q S u Q T v (7)
• Rearranging (6),
P P T S v u
( ) P ( ) T S (v u)
vu (8)
P T S
u v
Q T S
(9)
• (8) and (9) are the reduced form equations for P and Q.
• But what would happen if we had estimated equations (4) and (5), i.e. the
structural form equations, separately using OLS?
• Both equations depend on P. One of the CLRM assumptions was that
E(Xu) = 0, where X is a matrix containing all the variables on the RHS of
the equation.
• It is clear from (8) that P is related to the errors in (4) and (5) - i.e. it is
stochastic.
• What would be the consequences for the OLS estimator, , if we ignore
the simultaneity?
• If the X’s are non-stochastic, E(Xu) = 0, which would be the case in a single
equation system, so that ) is the condition for unbiasedness.
E,(which
Solution
G = 3;
If # excluded variables = 2, the eqn is just identified
If # excluded variables > 2, the eqn is over-identified
If # excluded variables < 2, the eqn is not identified
Equation 14: Not identified
Equation 15: Just identified
Equation 16: Over-identified
2. Run the regression corresponding to equation (14).
3. Run the regression (14) again, but now also including the fitted values
Y2 , Y3 as additional regressors:
Y1 0 1Y2 3Y3 4 X 1 5 X 2 2Yˆ2 3Yˆ3 u1
1 1
(20)
4. Use an F-test to test the joint restriction that 2 = 0, and 3 = 0. If the
null hypothesis is rejected, Y2 and Y3 should be treated as endogenous.
• In fact, we can use this technique for just-identified and over-identified systems.
• Two stage least squares (2SLS or TSLS) is done in two stages:
Stage 1:
• Obtain and estimate the reduced form equations using OLS. Save the fitted
values for the dependent variables.
Stage 2:
• Estimate the structural equations, but replace any RHS endogenous variables
with their stage 1 fitted values.
• Recall that the reason we cannot use OLS directly on the structural equations is that the
endogenous variables are correlated with the errors.
• One solution to this would be not to use Y2 or Y3 , but rather to use some other variables
instead.
• We want these other variables to be (highly) correlated with Y2 and Y3, but not correlated
with the errors - they are called INSTRUMENTS.
• Say we found suitable instruments for Y2 and Y3, z2 and z3 respectively. We do not use the
instruments directly, but run regressions of the form
(27) & (28)
Y2 1 2 z2 1
Y3 3 4 z3 2
‘Introductory Econometrics for Finance’ © Chris Brooks 2013 23
Instrumental Variables (cont’d)
• Obtain the fitted values from (27) & (28), Y2 and Y3 , and replace Y2 and
Y3 with these in the structural equation.
• The S&P 100 Index has been traded on the CBOE since 1983 on a
continuous open-outcry auction basis.
• Equations (1) & (2) and then separately (3) & (4) are estimated using 2SLS.
C a ll B id -A s k S p r e a d a n d T r a d in g V o lu m e R e g r e s s io n
CBA i 0 1 CDUM i 2 C i 3 CL i 4 T i 5 CR i e i ( 6 .5 5 )
2 2
CL i 0 1 CBA i 2T i 3T i 4M i v i ( 6 .5 6 )
2
0 1 2 3 4 5 A d j. R
0 .0 8 3 6 2 0 .0 6 1 1 4 0 .0 1 6 7 9 0 .0 0 9 0 2 - 0 .0 0 2 2 8 - 0 .1 5 3 7 8 0 .6 8 8
( 1 6 .8 0 ) ( 8 .6 3 ) ( 1 5 .4 9 ) ( 1 4 .0 1 ) ( - 1 2 .3 1 ) ( - 1 2 .5 2 )
2
0 1 2 3 4 A d j. R
- 3 .8 5 4 2 4 6 .5 9 2 - 0 .1 2 4 1 2 0 .0 0 4 0 6 0 .0 0 8 6 6 0 .6 1 8
( - 1 0 .5 0 ) ( 3 0 .4 9 ) ( - 6 .0 1 ) ( 1 4 .4 3 ) ( 4 .7 6 )
N o te : t-ra tio s in p a re n th e s e s . S o u rc e : G e o rg e a n d L o n g s ta ff (1 9 9 3 ). R e p r in te d w ith th e p e rm is s io n o f
th e S c h o o l o f B u s in e s s A d m in is tra tio n , U n iv e rs ity o f W a s h in g to n .
P u t B id -A s k S p r e a d a n d T r a d in g V o lu m e R e g r e s s io n
PBA i 0 1 PDUM i 2 P i 3 PL i 4 T i 5 PR i u i ( 6 .5 7 )
2 2
PL i 0 1 PBA i 2T i 3T i 4M i w i ( 6 .5 8 )
2
0 1 2 3 4 5 A d j. R
0 .0 5 7 07 0 .0 3 2 5 8 0 .0 1 7 2 6 0 .0 0 8 39 - 0 .0 0 1 2 0 - 0 .0 8 6 6 2 0 .6 7 5
( 1 5 .1 9) ( 5 .3 5 ) ( 1 5 .9 0 ) ( 1 2 .5 6) ( - 7 .1 3 ) ( - 7 .1 5 )
2
0 1 2 3 4 A d j. R
-2 .8 9 3 2 4 6 .4 6 0 - 0 .1 5 1 5 1 0 .0 0 3 39 0 .0 1 3 4 7 0 .5 1 7
( - 8 .4 2 ) ( 3 4 .0 6 ) ( - 7 .7 4 ) ( 1 2 .9 0) ( 1 0 .8 6 )
N o te : t-ra tio s in p a re n th e s e s . S o u rc e : G e o rg e a n d L o n g s ta ff (1 9 9 3 ). R e p rin te d w ith th e p e rm is s io n o f
th e S c h o o l o f B u s in e s s A d m in is tra tio n , U n iv e rs ity o f W a s h in g to n .
Adjusted R2 60%
1 and 1 measure the tick size constraint on the spread
2 and 2 measure the effect of the option price on the spread
3 and 3 measure the effect of trading activity on the spread
4 and 4 measure the effect of time to maturity on the spread
5 and 5 measure the effect of risk on the spread
1 and 1 measure the effect of the spread size on trading activity etc.
• The paper argues that calls and puts might be viewed as substitutes
since they are all written on the same underlying.
• So call trading activity might depend on the put spread and put trading
activity might depend on the call spread.
• The results for the other variables are little changed.
• Bid - Ask spread variations between options can be explained by reference to the
level of trading activity, deltas, time to maturity etc. There is a 2 way relationship
between volume and the spread.
• The authors argue that in the second part of the paper, they did indeed find
evidence of substitutability between calls & puts.
Comments
- No diagnostics.
- Why do the CL and PL equations not contain the CR and PR variables?
- The authors could have tested for endogeneity of CBA and CL.
- Why are the squared terms in maturity and moneyness only in the
liquidity regressions?
- Wrong sign on the squared deltas.
uˆuˆ ̂
LR T log ˆ r log ˆ u
‘Introductory Econometrics for Finance’ © Chris Brooks 2013 42
Choosing the Optimal Lag Length for a VAR
(cont’d)
where ̂ r is the variance-covariance matrix of the residuals for the restricted
model (with 4 lags), ̂ u is the variance-covariance matrix of residuals for the
unrestricted VAR (with 8 lags), and T is the sample size.
•The test statistic is asymptotically distributed as a 2 with degrees of freedom
equal to the total number of restrictions. In the VAR case above, we are
restricting 4 lags of two variables in each of the two equations = a total of 4 *
2 * 2 = 16 restrictions.
•In the general case where we have a VAR with p equations, and we want to
impose the restriction that the last q lags have zero coefficients, there would
be p2q restrictions altogether
•Disadvantages: Conducting the LR test is cumbersome and requires a
normality assumption for the disturbances.
y1t 10 11 y1t 1 12 y2t 1 11 y1t 2 12 y2t 2 11 y1t 3 12 y2t 3 u1t
y2t 20 21 y1t 1 22 y2t 1 21 y1t 2 22 y2t 2 21 y1t 3 22 y2t 3 u2t
• This is done by determining how much of the s-step ahead forecast error
variance for each variable is explained innovations to each explanatory
variable (s = 1,2,…).
Lags of Variable
Dependent variable SIR DIVY SPREAD UNEM UNINFL PROPRES
SIR 0.0000 0.0091 0.0242 0.0327 0.2126 0.0000
DIVY 0.5025 0.0000 0.6212 0.4217 0.5654 0.4033
SPREAD 0.2779 0.1328 0.0000 0.4372 0.6563 0.0007
UNEM 0.3410 0.3026 0.1151 0.0000 0.0758 0.2765
UNINFL 0.3057 0.5146 0.3420 0.4793 0.0004 0.3885
PROPRES 0.5537 0.1614 0.5537 0.8922 0.7222 0.0000
Months ahead I II I II I II I II I II I II
1 0.0 0.8 0.0 38.2 0.0 9.1 0.0 0.7 0.0 0.2 100.0 51.0
2 0.2 0.8 0.2 35.1 0.2 12.3 0.4 1.4 1.6 2.9 97.5 47.5
3 3.8 2.5 0.4 29.4 0.2 17.8 1.0 1.5 2.3 3.0 92.3 45.8
4 3.7 2.1 5.3 22.3 1.4 18.5 1.6 1.1 4.8 4.4 83.3 51.5
12 2.8 3.1 15.5 8.7 15.3 19.5 3.3 5.1 17.0 13.5 46.1 50.0
24 8.2 6.3 6.8 3.9 38.0 36.2 5.5 14.7 18.1 16.9 23.4 22.0
0.06
0.04
0.02
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
-0.02
-0.04