Journal of Accounting and Economics

Available online May 2022, 101515

The Impact of Stock market shocks on Pakistan’s macroeconomic variables: A

structural vector auto regression approach.

By

Mir Alam1, Imtiaz Hussain2, Faiza Kiran3

1. University of Baltistan, Skardu, Pakistan

2. Pakistan Institute of Development Economic Islamabad
3. Centre for Economic Research in Pakistan (CERP), Lahore.

Abstract:

This study uses monthly data from 1991-1 to 2019-12 to analyse the stock market shocks
on Pakistan’s macroeconomic variables. The study contains some macroeconomic
variables, including stock price (SP), industrial manufacturing production index (IP),
exchange rate (ER), interest rate (IR), money supply (M2), and consumer price index
(CPI). The structure vector autoregressive (SVAR) model has been applied to impose
long–term restrictions on macroeconomic variables. Our results found that there is a unit
root at level, but their first difference indicates that there is no seasonal unit root at zero
frequency, and also found that there is no autocorrelation in the model. The stock price
shocks are positive but have little effect on money supply, industrial production, and
consumer price. The shock did not transmit into them. On the other hand, the stock market’s
impact is negative and significant for interest rate and interest rate, which shows that in the
case of Pakistan, shocks are translated mainly into interest rate and exchange rates.

Keywords: Stock market shocks, Impulse response, variance decomposition, SVAR,

Pakistan.

Introduction:

The importance of stock and financial markets are increasing around the world and both
are considered aa important elements for economic growth. Especially, the stock market
provides different ways to the growth of industries and commerce, which further leads to
economic development of the country in a large scale (Ahmed and Development 1999).
The stock market has a significant role in the economy because it makes domestic resources
valuable into the utilization in economic activities, which further leads to increase
investment in the country. Therefore, the stock market has found as a significant indicator
of the economy and it has relation with aggregate demand in the economy, especially
through consumption and investment.

1
There are some channels, which link share prices to macroeconomic variables. First,
suggested by(Tobin and banking 1969), he reported in detail that share prices affect cost
of capital, which leads to increase investment in the economy. Second channel has
presented by (Modigliani and linkages 1971) studied that how the impact of share prices
transmits into gross domestic product. In this connection, he pointed out that higher share
price leads to increase wealth in an economy and then households are consuming this
wealth to maximize their utility by increasing consumption patterns in an economy. When
consumption is increasing then ultimately aggregate demand will also increase and
henceforth, production will increase to enhance the economic growth. This idea is also
supported by permanent income hypothesis. Third link has presented by (Gertler and
Bernanke 1989) in which they reported that share prices affects firm’s balance sheet due
to prevailing asymmetric information in credit market. Where borrowing of firms depends
significantly on the collateral they initiate.

Portfolio Balance Models of exchange rate determination also present the theoretical link
between share prices and exchange rate. Where, in any economy, individuals are trading
domestic assets and foreign asset as well as currencies of both. If domestic share price
increases in a country, then domestic individuals start to selling foreign assets. In this case,
domestic money demand will be increased again this will lead to appreciate local currency,
but here the appreciation of exchange rate is one unit of foreign currency in terms of
domestic currency (Sohail, Hussain et al. 2009). The life cycle hypothesis which has
developed by (Ando and Modigliani 1963) postulates that as for as share prices change, for
example, increase in share price leads to increase wealth holding of individual.

Moreover, the share prices have a positive impact on consumer price index as if
investors are net debtors in stock market, this will be a result of the higher value of firm’s
equity which leads to increase expected inflation (Chakravarty, Mitra et al. 2013). And the
share price has a negative impact on interest rate Because when share prices decrease in an
economy then investors will not invest in market and same time if Banks increases interest
rate to their depositor then the increased wealth due to fall in share prices will switch to the
Banks. This will lead to decrease the demand for share in stock market and will decrease
share prices and both will treat an inverse relationship (Alam, Uddin et al. 2009, Giri, Joshi
et al. 2017).

2
Data and Econometric Technique

Data

The variables used in the analysis are Industrial Manufacturing Production Index, exchange
rate, interest rate, Share Prices, Money supply and Consumer Price Index. For this purpose,
the period average monthly data is taken from 1991:1 to 2019:12. The reason for choosing
this period for analysis is that there was huge volatility in stock market due to political
instability, earth quack and other factors in this time period. Data for all these variables are
taken from International Financial Statistics (IFS). The nature of data has presented in
below figure.

CPI EX GDP
160 200 140

120
160
120
100
120
80 80
80
60
40
40
40

0 0 20
92 94 96 98 00 02 04 06 08 10 12 14 92 94 96 98 00 02 04 06 08 10 12 14 92 94 96 98 00 02 04 06 08 10 12 14

INTR M2 SPR
25 12,000,000 400

10,000,000
20
300
8,000,000
15
6,000,000 200
10
4,000,000
100
5
2,000,000

0 0 0
92 94 96 98 00 02 04 06 08 10 12 14 92 94 96 98 00 02 04 06 08 10 12 14 92 94 96 98 00 02 04 06 08 10 12 14

Econometric Methodology

The identification problem deals with imposing restrictions on some of the

structural parameters. On the other hand, Cholesky decomposition deals little with
economic interpretation or theory. In response to this criticism, the Structural Vector Auto
regression (SVAR) approach was developed by Sims (1986), Bernanke (1986) and
Blanchard and Watson (1986). The SVAR approach allows for imposing restrictions on

3
the basis of economic theory for the identification of structural parameters. An extension
to the SVAR of Sims (1980) and others was made by (Blanchard 1989) by imposing long-
run and short run restrictions on structural parameters. And this methodology assumes that
the structural shocks are orthogonal and these shocks are normalized to have unit variance.
The main purpose of using SVAR methodology by considering economic theories to
overlap the structural shock from the residuals 𝑒1𝑡 𝑎𝑛𝑑 𝑒2𝑡 . If the correlation between
𝑒1𝑡 𝑎𝑛𝑑 𝑒2𝑡 is low, then there is no need to ordering. Moreover, in VAR model with some
variables it is not possible that all correlation between errors or variables is also low and
when the errors of VAR model are correlated then other all practical usage of ordering is
useless (Applied time series econometrics by Walter Andres).
To understand SVAR model below model has considered as first order n number of
variables by using algebra matrix.
𝐵𝑍𝑡 = 𝛾𝑜 + 𝛾1 𝑍𝑡−1 + 𝜀𝑡 ………………………………………………………..…1
Where B is element of diagonal box and all elements are unity.
Now by Multiplying 𝐵 −1 on both sides we obtain
𝑍𝑡 = 𝐴𝑜 + 𝐴1 𝑍𝑡−1 + 𝜀𝑡 ……………………………………………………….2
Where:
𝐴𝑜 = 𝐵 −1 𝛾𝑜 , 𝐴1 = 𝐵 −1 𝛾1 and 𝑒𝑡 = 𝐵 −1 𝜀𝑡 …………………………………3
Here B elements of diagonal box are all unity and also B consist of (n2 – n) unknown values
furthermore there are n unknown values and var (𝜀𝑡 ) for total (n2 ) unknown values in the
structural model that is n2 – n values of B plus n values var (𝜀𝑡 ) now the identification
problem is simple. To identify n2 unknown from the known (n2 + n)/2 independent elements
of sum, it is important to impose additional restrictions that is n2 – (n2 + n)/2 = (n2-n)/2
restriction on the system. To estimate structural model from VAR model we need to impose
(n2 –n) / 2 restrictions on SVAR model. Hence, if we have (n2 – n)/2 number of restrictions
then it is just identified.
We have used SVAR modeling for six variables that was basis on Blanchard and
Quah (1986) Also Kim and Roubani (2000), to measure stock market prices shocks on
macroeconomic variables and reverse effect of macroeconomic variables to create stock
market prices bubbles for Pakistani economy. So, we have the system of structural
equations are as follow.

4
 𝐴1𝑌𝑡 + 𝐴2𝑍𝑡 + 𝑉𝑡 ………………………………. 4
𝐵𝑌𝑡 = 𝐴𝑌𝑡−1 + 𝐶𝑍𝑡 + 𝑉𝑡 ………………………… 5
𝑌𝑡 = 𝐵 −1 𝐴𝑌𝑡−1 + 𝐵 −1 𝐶𝑍𝑡 + 𝛽 −1 𝑉𝑡 ….......................... 6
Where A1, A2, B, C, are coefficients of matrix. And 𝑌𝑡 is n x 1 vector of
endogenous variables in the system as well as Zt is k x1 vector of exogenous variable (stock
market prices shock) and finally vt is n x 1 vector of structural parameter with zero mean
and variance (vt) = m where m is a diagonal matrix. All The diagonal values are
representing variance of structural disturbances so we can assume that not all structural
disturbances are correlated with one another. We have reduced above equation 3 into
reduced form model for structural equation as
𝐴(𝐿)𝑌𝑡 = 𝐵(𝐿)𝑍𝑡 + 𝜇𝑡 …….. 7
A(L) and B(L) are representing matrix polynomial of the lag operator and 𝜇𝑡 shows
error term of VAR model with identical independent distribution with zero mean and
variance constant ii~ ND (0,1). Because of reduced form model, we would estimate
structural parameter in numerous ways. In this connection, we need to impose some
restriction on elements of matrix and estimating structural VAR model based on economic
theories. Different group of studies have done on the imposition of restrictions as Cholaski
decomposition, these methods only accept recursive method. In such a case mostly results
depending on the ordering of variables which are under considered. Now applying SVAR
method which supports non-recursive approach in which we impose restrictions on
structural parameters see Sim, (1986) Blanchard and Quah (1989), Blanchard and Watson
(1986). By applying ordinary least square OLS method we have estimated 𝜇𝑡 from equation
(2) and the coefficient matrix D have not contemporaneous coefficient in the structural
form equation and this relationship has showed through below model.
𝐴(𝐿) = 𝐼 + 𝐵 −1 𝐴(𝐿)…………………………………… 8
The correlation between reduced form equation and structural form equation has shown as
below.
𝐵(𝐿) = 𝐵 −1 𝐶 9
Now in addition association between residual of VAR model and disturbances of structural
equations given by following models.
𝑈𝑡 = 𝐵 −1 𝑉𝑡 10

5
 𝐸(𝑢𝑡 , 𝑢′ 𝑡 ) = 𝐵 −1 𝐸(𝑉𝑡 , 𝑉 ′ 𝑡 )𝐵 −1 11
𝐸(𝑢𝑡 , 𝑢′ 𝑡 ) = 𝐵 −1 𝐷. 𝐵 −1 12
Σ = 𝐵 −1 𝐷𝐵 −1 13
𝐵 −1 Is constant estimate and D is coefficient matrix and both are estimated by
sample estimate of Σ through maximum likelihood function. We have estimated n (n1)
free parameters in equation (8) and contains only n (n1)/2 parameters so we also need to
impose additional restriction on the system. So finally, we need to impose restriction by n
(n-1)/2 for the element of B to be unity.
Non-Recursive Method.
The structural vector autoregressive (SVAR) model is reduced from VAR model in which
economic interpretation is being considered. In this model we restrict variables in a matrix
according to their theoretical perspective on the basis of non-recursive method which is as
follow.

The left Colum show the shocks of all variables and the right Colum shows all reduced
form residual. In addition, first equation in above matrix is share prices equation, which
represents stock market, and we assume that exchange rate, interest rate and consumer price
index have contemporaneous impact on share prices. 2nd and 3rd equations present money
market equilibrium and theoretically we assumed that money supply is contemporaneously
effected by industrial production (IP) and interest rate (IR).

4th and 6th equations are industrial production and consumer price index which
representation goods market equilibrium theory and we assume that M2, IR and ER have
not contemporaneous impact on Industrial production. On the other hand, industrial
production is contemporaneous effect by share price and consumer price index. 5th equation
is exchange rate equation and theoretically we assumed that it is function of all variables.
In this connection we will solve the problem of identification and estimation by imposing
n2-n/2 restrictions.

6
 RESULTS AND DISCUSSION

Table 1: Results of Seasonal Unit Root Using Beaulieu and Miron approach

8 SP M2 IR IP EX CPI ∆SP ∆M2 ∆IR ∆IP ∆EX ∆CPI

t : 1  0 -2.02 -2.76 -1.84 -0.48 -1.93 -0.60 -4.31** -3.84** -6.20** -7.21** -4.10** -3.14**

(-3.19) (-3.19) (-2.76) (-2.76) (-2.79) (-2.76) (-1.91) (-2.76) (-2.76) (-2.76) (-2.79) (-2.76)

t : 2  0 -4.96** -5.68** -5.79** -5.66* -4.83** -5.34** -4.70** -4.53** -5.27** -4.58** -3.92** -5.46**

(-2.76) (-2.76) (-2.76) (-2.76) (-1.88) (-2.76) (-1.88) (-2.76) (-2.76) (-2.76) (-1.88) (-2.76)

F :3   4  0 24.94** 30.27** 26.81** 17.83** 21.86** 22.58** 22.42** 16.32** 26.07** 14.16** 15.88** 31.14**

(6.24) (6.24) (6.27) (6.27) (3.03) (6.27) 3.05 (6.27) (6.27) (6.27) (3.05) (6.27)

F :5   6  0 34.74** 34.86** 17.91** 18.44** 35.17** 39.98** 24.27** 21.69** 15.28** 15.11** 22.9** 22.89**

(6.26) (6.26) (6.28) (6.28) (2.99) (6.28) (3.01) (6.28) (6.28) (6.28) (2.99) (6.28)

F : 7  8  0 26.22** 33.24** 23.53** 31.96** 25.09** 47.47** 21.75** 22.26** 21.72 17.11** 17.62** 38.40**

(6.18) (6.18) (6.21) (6.21) (3.02) (6.21) (3.05) (6.21) (6.21) (6.21) (3.02) (6.21)

F :  9   10  0 30.91** 25.03** 45.74** 47.27** 48.16** 33.88** 22.43** 13.15** 36.23** 22.53** 33.73** 27.42

(6.24) (6.24) (6.22) (6.22) (3.04) (6.22)** (3.06) (6.22) (6.22) (6.22) (3.04) (6.22)

F :  11   12  0 42.55** 33.19** 22.43** 11.27** 14.78** 17.15** 28.16** 18.23** 20.46** 10.14** 10.69** 32.10**

(6.24) (6.24) (6.21) (6.21) (3.06) (6.21) (3.09) (6.21) (6.21) (6.21) (3.06) (6.21)

Auxiliary C, D, T C, D, T C, NT, C, NT, C, NT, C, NT, NC, NT C,D,NT C,D,NT C,D,NT C, ND, C ,D,

Regression D D ND D ND NT NT

7
Critical values given by (Franses and Hobijn 1997) are in parentheses and ** shows 5%
level of significance

Table-1 shows the results of the (Beaulieu and Miron 1993) seasonal unit root test both at
level and at first difference. We consider 5 percent significance level using (Franses and
Hobijn 1997) critical values for detection of seasonal unit root. The results showed that at
level the calculated values of the t-statistics of  1 are -2.02 for Share prices, -2.76 for
money supply, -1.84 interest rate, -0.48 for industrial production, -1.93 for exchange rate
and -0.60 for consumer price index. These calculated values at zero frequency unit root are
greater than their critical values, so null hypothesis cannot be rejected which implies the
presence of unit root at zero frequency i.e. series of SP, M2, IR, IP, EX and CPI rates are
non-stationary at level. Therefore, we have transformed the variables by using first
difference filter at zero frequency that is (1  B) yt  yt  yt 1 . After transforming, the

calculated values of t-statistics for all variables  1 and  2 are less than their critical values
for all variables. On the other hand, the calculated values of F-test statistics are greater than
critical values, which lead to the rejection of null hypothesis. Therefore, SP, M2, IR, IP,
EX and CPI are all become stationary at first difference. These calculated values are less
than the critical value at 5% significance level, which implies that all variables contains no
unit root at zero frequency. Also the calculated values of  2 for all variables are less than
their critical value and the F-statistics values are greater than their critical values which
leads to the conclusion that all variables contains no unit root at First difference.

8
 Lag Length Criteria
We need lags length criteria before estimating VAR model. For this purpose, we have
different econometric criteria’s including, AIC, SC, HQ, LR and others criteria we followed AIC
and LR for the selection of lags, so the results are presented in Table 4.

Table:2 Lags Selection Criteria

Lag Log-L LR FPE AIC SC HQ

0 3021.348 NA 4.03e-17 -20.72404 -20.64830* -20.69370

1 3098.491 150.5752 3.03e-17 -21.00681 -20.47664 -20.79442*

2 3147.197 93.06067 2.78e-17 -21.09414 -20.10954 -20.69970

3 3177.441 56.53844 2.90e-17 -21.05458 -19.61555 -20.47810

4 3217.811 73.80323 2.81e-17 -21.08461 -19.19115 -20.32608

5 3258.085 71.96736 2.74e-17 -21.11399 -18.76609 -20.17341

6 3316.528 102.0240 2.36e-17 -21.26823 -18.46591 -20.14561

7 3349.680 56.50647 2.41e-17 -21.24866 -17.99190 -19.94399

8 3393.822 73.41749* 2.30e-17* -21.30462* -17.59342 -19.81790

Note: * indicates Minimum Vales of different methods to select Maximum Lags.

By applying least square method from the above Table 2 we showed that three criteria’s suggesting
us to select 8 lags as AIC, FPE and LR. While SC suggests zero and HQ offers one lag to select.
Since, the minimum AIC or maximum value of LR we have selected 8 lags, but the model is
selected because of minimum value of AIC.

Variance Decomposition
Table 8 shows the estimated results of variance decomposition. This test is used to report variances
of selected variables are in more informative detail. The method of forecast error variance
decomposition is also used to decide variability proportion of errors for the sacking to forecast

9
variables at current time as well as in the long run. Because the variability proportion changes
when structural shocks occur.

Estimation of Structural VAR Model

This study uses non-recursive approach to estimate the SVAR model by imposing
theoretical restrictions on contemporaneous coefficient in the system, the imposition of restriction
based on theories is given in chapter 4 of Non-Recursive method, and estimated coefficients are
presented in Table 9.
Table 4 Estimated Coefficients Share price into Macroeconomic variables

Coefficient Std. Error z-Statistic Prob.

𝛼21 0.065312 0.085461 0.764231 0.4447
𝛼31 -0.96187 0.150124 -6.40714 0.000
𝛼41 0.043441 0.155535 0.279299 0.7853
𝛼51 -0.98593 0.429587 -2.29506 0.0217
𝛼61 0.030829 0.109687 0.281065 0.7787

In the above Table 4 we have estimated the coefficients of share prices to macroeconomic

variables like industrial production, exchange rate, interest rate, consumer price index and money

supply to determine whether shock of share prices into macroeconomic variables are significant

or not. In such a case 𝛼21 is coefficient of share prices in money supply equation and it is positive

but insignificant means that it increases money supply. On the other hand stock market responses

are more better when money supply increases see (Maskay 2007). Moreover, according to real

activity economists if share prices are increasing, is the result of higher cash flows in the economy

and higher cash flows are possible when there are higher economic activities. And further

economic activities is the result of increase in money demand and finally the increment in money

demand is the function of increment in money supply so the channel is positive from share prices

to money supply (Talla 2013).

10
 𝛼31 is coefficient of share prices in interest rate equation which is significant but negative
meaning that transmission mechanism from share prices to interest rate is significant in Pakistan.
However, its negative relationship is concern if shock comes from share prices and share prices
goes down then this will lead to switch investors towards Banks and the same time if Banks
increases interest to their depositors. As we know theoretically if investment increase, it will also
decrease the rate interest. On the other hand, decrease in interest rate will also lead to decrease
lending interest rate, which further increases investment in the economy by increasing demand of
shares and decreasing prices of shares. While there is also possibility of above channel vice-versa
as if banks are paying good interest for their depositors, they will switch from stock market to
Banks. Resulting decrease in investment in stock market and lead to decrease aggregate demand
for shares and share prices. Finally, if there is permanent increment in interest rate by Banks to
their depositors will also increase lending interest rate and this will further decrease investment. in
such a case the negative relationship can exist between share prices to interest rate (Alam, Uddin
et al. 2009). 𝛼41 Is coefficient of share prices in industrial production equation is positive but
insignificant means shock of share price into industrial production is insignificant. On the other
hand, share price is leading indicator of economic growth as positive shock of share prices leads
to increase industrial production. According to Discounted Cash Flow Model share prices impact
cost of capital, which has captured by Tobin’s Q coefficient (1969). Which is equal to the ratio of
the market value of current capital to the cost of replacement capital.in this connection if share
prices are increasing it leads to increase cost of replacement capital. Which further leads to increase
investment and result to enhance production capacity in an economy. Aloof from this, permanent
income hypothesis increment in share prices increases wealth holding of individuals and they are
starting to increase their consumption smoothly to maximize their utility, it changes the patterns
of consumption in an economy resulting in increment in the ratio of investment and production
level. 𝛼51 Is coefficient of share prices in exchange rate equation that shows negative and
significant relationship and this means that shock of share price is transmitting into exchange rate.
There is no theoretical consensus between the relationship between share prices and exchange rate
either, on the other hand portfolio Balance Model of determination of exchange rate captured the
negative association, which runs from share prices to exchange rate. This model postulates that in
portfolios individuals keeping domestic as well as foreign assets including their currencies. Now
if the share price of domestic increases it leads to more demand of domestic assets by individuals,

11
to purchase domestic assets domestic investors starting to sell foreign asset this causes to
appreciate domestic currency. “But most importantly the currency appreciation means, it defines
exchange rate as a price of one-unit foreign currency in domestic currency terms. This means that
decrease in exchange rate. Hence, we can say that share prices have negative relationship with
exchange rate and this increased prices of share leads to increase wealth of investors and it also
raises money demand of investors, thus domestic interest rate will be increase. In such a case, once
again appreciate domestic currency which, has been also reported by (Muhammad, Rasheed et al.
2002). Finally, 𝛼61 is coefficient of share prices in consumer price index equation is positive and
insignificant, meaning that as share price’s shock does not transmits into CPI but it has positive
impact as share prices increases consumer price index also increases. There is possibility of both
positive and negative relationship between share prices and consumer price index. Positive
relationship runs from share prices to CPI as if firms are net debtors in stock market then their
equity value will also be increase. Therefore, this will lead to increase expected inflation. On the
other hand, according to (Fama 1981) high inflations predict downturns of economies over the
time. Based on this tendency in an economy firms are starting to sell their stock by increasing
supply of stock, result will be decrease in share prices.

CONCLUSION

In this empirical investigation we studied some deriving shocks of the stock market into
macroeconomic variables for Pakistan, by using monthly data from 1991-1 to 2019-6. Applying a
structural vector autoregressive (SVAR) approach with restrictions identifying in the long-run.
These restrictions have been identified from closed-economy macroeconomic model, such as
Dividend Discount model (DDM), Arbitrage Pricing Theory (APT), Discounted Cash Flow Model
(DCM), Portfolio Balance Model of Exchange Rate Determination as factor deriving shocks of
stock prices. We applied Monthly Unit root proposed by (Beaulieu and Miron 1993) using seasonal
dummies and trends to capture seasonal unit root at zero frequency. the results showed that unit
root exist at level, and variables are stationary at their first differences. AIC has allowed to take
lag lengths to 8 to overs come time series problems, such as autocorrelation. In addition, Jarque-
Bera tests also indicated the residuals are normally distributed.
The estimated results of SVAR model postulates that share prices shock into money supply is
positive but statistically insignificant, meaning that shock does not transmits into money supply.
Such a results validate that an increases in share prices increases money supply for the case of
Pakistan. On the other hand, share prices shock into interest rate is statistically significant but
negative, meaning that shocks of share price are transmitting significantly into interest rate.

12
Moreover, the transmission mechanism of share prices into industrial production is positive but
statistically insignificant.
We also found negative relationship between share prices and exchange rate and
statistically significant meaning that shock of share price is significantly transmitting into
exchange rate. This is mainly because of increase in share prices domestic investors are tend to be
selling foreign currencies and assets. this process makes domestic currency to appreciate and tend
to decrease exchange rate as of one-unit foreign currency in terms of domestic currency. We found
that shock of share prices does not transmit into consumer price index and we showed that as share
prices increases consumer price index also increases. We cannot reject our study hypothesis that
share prices have significantly impact on macroeconomic variable. Finally, interest rate is the
variable which showed high variation in short run as well as in the long run in Pakistan from shocks
of share prices.

13
References:

Ahmed, M. F. J. S. and Development (1999). "Stock market, macroeconomic variables, and causality: The
Bangladesh case/Bourse des valeurs, variables macroeconomiques et causalite: Le cas du Bangladesh."
109-130.

Alam, M., et al. (2009). "Relationship between interest rate and stock price: empirical evidence from
developed and developing countries." 4(3): 43-51.

Ando, A. and F. J. T. A. e. r. Modigliani (1963). "The" life cycle" hypothesis of saving: Aggregate implications
and tests." 53(1): 55-84.

Beaulieu, J. J. and J. A. J. J. o. e. Miron (1993). "Seasonal unit roots in aggregate US data." 55(1-2): 305-
328.

Blanchard, O. J. J. T. A. E. R. (1989). "A traditional interpretation of macroeconomic fluctuations." 1146-

1164.

Chakravarty, S., et al. (2013). "Stock prices and inflation: relationship revisited." 9(3): 201-215.

Fama, E. F. J. T. A. e. r. (1981). "Stock returns, real activity, inflation, and money." 71(4): 545-565.

Franses, P. H. and B. J. J. o. A. S. Hobijn (1997). "Critical values for unit root tests in seasonal time series."
24(1): 25-48.

Gertler, M. and B. Bernanke (1989). Agency costs, net worth and business fluctuations. Business cycle
theory, Edward Elgar Publishing Ltd.

Giri, A., et al. (2017). "The impact of macroeconomic indicators on Indian stock prices: An empirical
analysis." 12(1): 61-78.

Maskay, B. (2007). "Analyzing the effect of change in Money supply on stock prices." The park place
economist 15(1): 72-79.

Modigliani, F. J. C. s. and m. p. t. linkages (1971). "Monetary policy and consumption." 9-84.

Muhammad, N., et al. (2002). "Stock prices and exchange rates: Are they related? evidence from south
asian countries [with comments]." 535-550.

14
Sohail, N., et al. (2009). "Long-run and short-run relationship between macroeconomic variables and stock
prices in Pakistan: The case of Lahore Stock Exchange." 183-198.

Talla, J. T. (2013). "Impact of Macroeconomic Variables on the Stock M arket Prices of the Stockholm Stock
Exchange (OMXS30)." Jonkoping International Business School.

Tobin, J. J. J. o. m., credit and banking (1969). "A general equilibrium approach to monetary theory." 1(1):
15-29.

15