




































 

 

 
16 

© 2021 Conscientia Beam. All Rights Reserved. 

TESTING THE VALIDITY OF ARBITRAGE PRICING THEORY: A STUDY ON DHAKA 
STOCK EXCHANGE BANGLADESH   

 

 

 Syed Mohammad  
Khaled Rahman1+ 

 Priyanka   
Mazumder2 

 

1Associate Professor, Department of Business Administration, Shahjalal 
University of Science & Technology, Bangladesh. 

 
2Ex-student of MBA Program, Department of Business Administration 
Shahjalal University of Science & Technology, Bangladesh. 

 

 
(+ Corresponding author) 

 ABSTRACT 
 
Article History 
Received: 25 January 2021 
Revised: 23 February 2021 
Accepted: 19 March 2021 
Published: 14 April 2021 
 

Keywords 
Arbitrage  
Pricing theory index interest 
exchange. 

 
JEL Classification: 
G12. 

 
The purpose of the study was to test the validity of Arbitrage Pricing Theory (APT) in 
Dhaka Stock Exchange (DSE) of Bangladesh. Secondary data has been used which was 
composed of observable macroeconomic and stock market variables. Study period was 
from January 2013 to October 2018, making a total of 70 monthly observations. Study 
found that interest rate and exchange rate has significant influence but market 
capitalization and tax rate have insignificant impact on return of DS-30 index. Except 
exchange rate, other three variables were negatively related with DS-30 index return. 
1% increases in exchange rate results 0. 993% increase in stock prices while 1% 
increases in interest rate results 0. 486% decrease in stock prices and vice-versa. Strong 
negative correlation was seen between interest rate and stock index return. APT have 
failed to fully explain the change of DS-30 index return due to presence of two 
insignificant explanatory variables. This research has practical implications on stock 
market participants as investors’ optimal strategy largely influenced by precision of 
asset pricing models. This research has also policy implications for Securities & 
Exchange Commission, government, and other regulators as findings of the study will 
assist them to develop more efficient capital market.  
 

Contribution/Originality: This study contributes to the existing literature of asset pricing model by judging 

its reliability in Bangladeshi capital market. This study is one of very few studies which have investigated the 

validity of Arbitrage Pricing Theory in Dhaka Stock Exchange with the help of index of blue chip companies. 

 

1. BACKGROUND OF THE STUDY 

The capital market is a channel of direct finance. It enables mobilization of funds for investment purposes from 

purchaser of capital market securities or investor to issuer of securities or borrowers (Imam, 2001). It is a market 

where savings are made available to investors as firms get capital through issuing financial assets such as bonds and 

equities. In other words, through capital market funds are transferred from investors to different economic entities 

in exchange of transfer of financial instruments (Chowdhury, 2005). Like other markets it has a set of rules to 

ensure protection of the buyers i.e. investors. 

Capital market lies at the heart of the economic development of a country as it circulates the life blood of 

industrialization, i.e. the capital, in an economy; like a heart circulates blood in a human body. But it requires 

rigorous, conscious, and continuous human efforts to result in a fully and efficiently functioning organ of an 

economy. These efforts encompass some fundamentals related to the very nature of the market without establishing 

which the true benefits of the market cannot be harnessed. Several theories and models have been developed for 

Financial Risk and Management Reviews 
2021 Vol. 7, No. 1, pp. 16-25. 
ISSN(e): 2411-6408 
ISSN(p): 2412-3404 
DOI: 10.18488/journal.89.2021.71.16.25 
© 2021 Conscientia Beam. All Rights Reserved. 

 
 
 

 
 
 

 

 
 
 
 

https://orcid.org/0000-0002-8713-6933
https://orcid.org/0000-0002-9708-0507
https://www.doi.org/10.18488/journal.89.2021.71.16.25


Financial Risk and Management Reviews, 2021, 7(1): 16-25 

 

 
17 

© 2021 Conscientia Beam. All Rights Reserved. 

estimating prices or values of financial instruments. Arbitrage Pricing Theory (APT) is one of them. Determination 

of asset prices through well-established models or theories is an indicator of a well-functioning capital market. 

 

2. PROBLEM STATEMENT 

An investor’s appropriate investment decision largely depends on accuracy level of forecasting stock prices or 

values. In general, market prices of firms’ shares in Dhaka Stock Exchange (DSE) fluctuates every now and then 

and price variation enhances as investors’ holding periods lengthen (Mollik & Bepari, 2010). DSE stock return 

behave abnormally for example stock market price crush of 1996 and 2010. Many investors gained and many faced 

tremendous losses from the unexpected price abrupt DSE trade stock. As the shares of stocks of different industries’ 

firms are traded in DSE and the economy consists of different industries so certain macro-economic variables 

should have significant impact on stock prices. Some macro-economic variables are treated as explanatory variable 

in APT and in an efficient and well developed capital market true asset prices can be forecasted through APT which 

guides rational investment behavior. Individual and institutional investors’ investment decisions influence the 

degree of abnormal return and price instability.   

 

3. THEORETICAL FRAMEWORK OF ARBITRAGE PRICING THEORY (APT) 

The APT is a multi-index model which along with market index describes both macroeconomic and industrial 

factors that changes most of the security prices to a specific direction. Influence of non-market factors are 

incorporated. Initial equation for APT test is:  

Rit= 𝜶 + 𝜷1 X1+ 𝜷2X2 + 𝜷3X3 + ….+ 𝜷iXi + 𝜺  (Dimson & Mussavian, 1999) 

Here, Rit the return of ith stock at t period. 𝜶 is the intercept, 𝜷1, 𝜷2,……i  are relative slopes of X1, X2,….Xi 

macro variables. 𝜺 is the error term.  

The Fama-French three factor model: 

Fama and French (1993); Fama and French (1995) proposed a three factor model to evaluate stock market 

return; 

E(Ri )= Rf + β1 [E(Rm) – Rf] + β2.SMB+ β3.HML 

Here, HML is the difference in return between portfolios of assets with high book to market and low book to 

market ratio. SMB is the difference in return between portfolios of assets with small market capitalization and big 

market capitalization.  

 

4. LITERATURE REVIEW AND RESEARCH GAP 

Lehmann and Modest (2005) published a study examining the different strategies for constructing basis 

portfolios that are highly correlated with the factors affecting security rates of return. Study found that performance 

of portfolios with large number of securities was better than portfolios with small number of securities. They found 

factor analysis as superior technique to principal components procedures. Ingersoll (1984) is one of the supporters 

of the APT model, stating in his study that the APT does not incorporate CAPM’s dubious assumptions in 

describing simple linear relationship between security prices and various factors those have impact on rates of 

return. According to opinion of some researchers, although equally weighted stock market index does not truly 

represent market portfolio, its movement of returns should be highly associated with the true market portfolio 

(Kandel & Stambaugh, 1987; Shanken, 1985).  

Chen, Roll, and Ross (1986) in their study verified the soundness of the APT in US stock markets. They found 

several macroeconomic factors such as industrial production, variations in risk premiums, and the shifts in the yield 

curve have profound influence on stock’s return. The relation between risk and return for agricultural assets was 

examined by Arthur, Carter, and Abizadeh (1988) who concluded that the APT was better at explaining returns for 

these assets than the CAPM. Handa and Linn (1993) in their study depicted a linear relationship between expected 



Financial Risk and Management Reviews, 2021, 7(1): 16-25 

 

 
18 

© 2021 Conscientia Beam. All Rights Reserved. 

asset returns and their factor betas; however, they also concluded that when there is more information available, 

predicted prices were higher while factor betas were relatively lower. On the other hand, less available information 

underestimated prices and overestimated factor betas.  

Sarver and Philippatos (1993) evaluated the nature of spot foreign exchange risk premiums using the APT. 

They tested whether the differences in the pure returns on currencies depend on systematic risk. They came to the 

result that single-factor APT model can describe expected exchange returns. Due to non-stationarity problem of 

security returns, Koutmos and Theodossiou (1993) examined the impact of conditional heteroskedasticity in the 

APT with observed variables. Study revealed that forming portfolios can’t remove the existence of conditional 

heteroskedasticity which leads to inefficient estimates of factor betas, and if ignored leads to erroneous pricing of 

factors. Nshom (2007) analyzed 18 stocks listed in London Stock Exchange (LSE) to explore the association 

between return of stock and currency exchange rates. Study revealed that in case of some firms, return of stock was 

significantly affected by exchange rate changes. Cauchie, Hoesli, and Isako (2004) conducted a study on the 

determinants of stock returns using an APT framework in the Swiss stock market which is particular because it 

includes a large number of firms that are susceptible to foreign economic conditions. They used statistical and 

macroeconomic implementations of the APT on 19 industrial sector indices. They concluded that statistically 

determined factors were better at explaining determinants of stock returns than macroeconomic variables.  

Altay (2003) analyzed German and Turkish economies and used the factor analysis technique on key economic 

indicators for identifying significant factors that affect security returns in an APT framework. Altay used the same 

economic indicators for both countries and found four factors affecting the German market while he found only 3 

for the Turkish market. The German market had two factors relating to unexpected economic changes. Ramadan 

(2012) conducted a research on 12 industrial portfolios which consisted of listed stocks traded in Amman Stock 

Exchange of Jordan over the period 2001-2011. Regressing six variables on stock return study found that four 

variables explained 84% changes in stock return. Study also revealed that the impact changes across industries. 

Zubairi and Farooq (2012) in their research tested both CAPM and APT using stock returns of fertilizer and power 

industries listed in Karachi Stock Exchange, Pakistan. Study found that macroeconomic variables were not the 

significant determinant of stock returns and hence APT was not valid. For testing efficiency of APT, (Basu & 

Chawla, 2012) analyzed 10 portfolios which consists of 50 stocks traded in Indian stock exchange over the period 

2003 to 2008. Study found that APT is valid since macroeconomic variables of APT were significantly associated 

with portfolios’ return. Iqbal, Khattak, Khattak, and Ullah (2012) tested the validity of APT in Karachi Stock 

Exchange of Pakistan using four macroeconomic factors and 26 listed firms’ stocks during 2004-2008. Study found 

that APT was valid since actual return did not significantly deviate from projected return.  

Depending on the previous studies mentioned above, it can be concluded that some of these studies supported 

the APT while others don’t. All studies used different methods in testing the validity of the models, some used 

unconditional APT while others used conditional APT, and thus the results are inconclusive. So, here some space of 

investigating the validity of model on Bangladesh Stock Market is found. Although few research works were done 

on Capital Asset Pricing Model (CAPM) but no research work was done on validity of APT from the context of 

Bangladeshi Stock Market. Aim of the present study is to fill this research gap.  

 

5. RESEARCH OBJECTIVES: 

The main purpose of the study is to test the validity of Arbitrage Pricing Theory (APT) model in Dhaka Stock 

Exchange (DSE). The specific objectives are as follows: 

a. To know about some descriptive statistics of DS-30 index return and macro-economic variables used in 

APT model. 

b.  To identify the strength and direction of co-movement between DS-30 index return and macro-economic 

variables used in APT model. 



Financial Risk and Management Reviews, 2021, 7(1): 16-25 

 

 
19 

© 2021 Conscientia Beam. All Rights Reserved. 

c. To explore the significance of impact of macro-economic variables on DS-30 index return. 

 

6. HYPOTHESES 

The hypotheses of the study are as follows: 

            H0-1: Exchange rate has significant effect on return of DS-30 index. 

            H0-2: Interest rate has significant influence on return of DS-30 index. 

            H0-3: Market capitalization has significant impact on return of DS-30 index. 

            H0-4: Tax rate has significant influence on return of DS-30 index. 

 

7. RESEARCH METHODOLOGY 

7.1. Data and Sample 

The data in this particular study consist of secondary data composed of observable variables that has been 

obtained from the DSE library, Bangladesh Bureau of Statistics, and Bangladesh Institute of Bank Management. 

This has covered the period from January 2013 to October 2018, making a total of 70 monthly observations. Data 

has comprised of interest on commercial bank loans & advances; interest on savings of commercial banks; exchange 

rate of the U.S dollar, Value of market capitalization, Taxes paid to government by investor and rate of return of 

DS 30 index. 

 

7.2. Model Specification 

The analysis has involved performing test of relative pricing model to know how much of the movement in 

stock returns is picked up by the several explanatory variables. From the full sample of 30 stocks, the firms selected 

spanned the full spectrum of stocks contained in the sample. The model for APT test is: 

RRDt = 𝜶 + 𝛽1 EXPt + 𝜷2INTt + 𝜷𝟑CAPt + 𝜷4 TAXt + 𝜺t       (Sharpe & Cooper, 1972)    

Here, RRDt representing Rate of Return on DS 30 companies at time period t is the dependent variable. Returns 

are articulated in percentage forms for study periods. 

Independent variables: Exchange Rate (EXPt), Interest Rate (INTt), Market Capitalization (CAPt), and Taxes 

paid by investors (TAXt) at time period t. 𝛽1, 𝛽2, 𝛽3, 𝛽4 are  coefficients of exchange rate , interest rates , market 

capitalization  & taxes respectively while 𝛼 is constant and 𝜀t is the error of analysis 

 Stock return –Return of a stock (RRD): It depends on stock prices which is largely influenced by value creation.  

 Exchange Rate- Ratio of two currencies (EXP): An exchange rate estimates the estimation of one cash in units of 

another money.  

 Interest rate (INT): It is the general market interest rate which indicates the fee paid for obtaining the fund. It 

is the cost of borrowing for the fund receiving entity. 

 Market capitalization (CAP)-Aggregate market stock value: It alludes to the all-out dollar market value of an 

organization's outstanding shares. It is determined by duplicating an organization's offers remarkable by the 

present market price of one share. 

 Taxes - Mandatory charges paid to government (TAX): A charge which is an obligatory budgetary charge or 

some other kind of toll forced upon a citizen (an individual or other lawful element) by a legislative association 

so as to support different open expenditures. 

 

7.3. Data Analysis Technique 

Here in the study, data have been analyzed using Ordinary Least Squares (OLS) or linear least squares method. 

It is a method for estimating the unknown parameters in a linear regression model. For APT test a regression 

model for establishing an equation is formed. This equation describes the relationship of RRD with other 

independent variables such as exchange rates, interest rates, market capitalization and taxes. 



Financial Risk and Management Reviews, 2021, 7(1): 16-25 

 

 
20 

© 2021 Conscientia Beam. All Rights Reserved. 

a) Data were dissected to information utilizing E-Views 10. Some specific tests were performed as followings:  

    1) Descriptive measurements. 

    2) Correlations between factors. 

   3) Estimation of regression coefficients. 

   4) Diagnostic tests: Normality, heteroskedasticity, serial auto correlation, model stability, and multicollinearity. 

b) Hypothesis test was performed through multiple linear regression.  

 

8. RESULTS AND DISCUSSIONS 

8.1. Descriptive Statistics of Variables 

It is necessary to comment on some preliminary features of the data for analyzing the result of the study. Table 

1 shows descriptive statistics for the Return of DS30 and other independent variables. 

 
Table-1. Descriptive statistics of dependent variable and independent variables. 

Statistics RRD EXP INT CAP TAX 

Mean 3.249404 1.897652 0.808625 3.753781 5.075939 

Median 3.245375 1.894316 0.814229 3.778838 5.076294 

Maximum 3.358549 1.922985 0.938520 3.834866 5.311029 

Minimum 3.108089 1.888741 0.684845 3.443106 4.436799 

Std. Dev. 0.059084 0.009821 0.089872 0.103793 0.137199 
Skewness -0.129869 1.371836 0.082612 -2.399796 -

1.381987 Kurtosis 2.486674 3.632694 1.488254 7.121232 8.240003 
Probability 0.634397 0.000018 0.041589 0.000000 0.000000 

Sum Sq. Dev. 0.226910 0.006269 0.525002 0.700242 1.223524 
Observations 70 70 70 70 70 

Note: Here all variables are in the percentage form. Data have been compiled by the researchers and calculated in EViews10. 

 

From the table, we can see that average return of DS30 is 3.25%.  The mean of TAX is the largest (5.08) and 

the lowest one is INT (.81) over the period. In term of standard deviation, highest value is TAX with 0.137 

followed by CAP, INT, RRD and EXP with 0.104, 0.0899, 0.059 and 0.0098 respectively. All of the variables except 

EXP and INT have negative skewness and positive kurtosis values.  

 

8.2. Correlation between Macroeconomic variables and DS-30 Index 

 
Table-2. Correlation among variables. 

Variables RRD EXP INT CAP TAX 

RRD 1 0.65223686 -0.8295003 -0.4721129 0.45743023 

EXP 0.65223686 1 -0.6582619 -0.8305811 0.41359074 

INT -0.8295003 -0.6582619 1 0.45274716 -0.5795148 

CAP -0.4721129 -0.8305811 0.45274716 1 -0.5046696 

TAX 0.45743073 0.41359074 -0.5795148 -0.5046696 1 
 

 

From Table 2 it is seen that relationship between RRD and INT is strong and negative as the value of 

correlation coefficient is -0.829. Moderate and positive correlation is seen between RRD and EXP (0.65). Negative 

and weak correlation is seen between RRD and CAP (-0.472) while TAX is positively related with RRD.  

 

8.3. Impact of Macroeconomic Variables on Return of DS-30 Index 

Regression model paves the way of constituting the mentioned equation. The independent variable which has 

high coefficient value at low probability has impact RRD more. 

From the Table 3, it is clear that constant value is 1.932. Coefficient values of EXP, INT, CAP and TAX are 

0.993, -0.486, -0.015 and -0.023 respectively. The probabilities are less than 5% for EXP and INT which means that 



Financial Risk and Management Reviews, 2021, 7(1): 16-25 

 

 
21 

© 2021 Conscientia Beam. All Rights Reserved. 

null hypothesis is rejected and these two macroeconomic variables has significant impact on return of DS-30 index. 

The other two variables CAP and TAX are insignificant as probabilities are greater than 5% and thus null 

hypothesis is accepted. From the regression model the actual equation established such as- 

             RRDt =   1.93 + 0.993 EXPt + (-0.486) INTt + (-0.015) CAPt + (-0.023) TAXt 

From the empirical study it is obvious that exchange rate significantly and positively affects stock return of 

DS30 companies. 1% increases in exchange rate results in 0. 993% increase in stock prices at 5% significant level in 

the safe range having 0.097% standard error. Another variable that has significant impact is INT but it is 

negatively related with return of DS-30 index. 1% increases in interest rate results in 0. 486% decrease in stock 

prices at 5% significant level having standard error 0.075%. The other two variables TAX and CAP have negative 

impact on stock prices but have insignificant impact. This supports the findings of Dimitrova (2005). 1% increase in 

market capitalization results in 0.015% decrease in Return of DS30 having standard error 0.187%. It is noted that 

1% increase in TAX causes 0.023% decrease in stock price and vice versa. So, it is relatively more worthy in 

considering exchange rate and interest rate effect for investing in stock market although market capitalization and 

tax rate has insignificant effect. So, APT is not fully valid for determining stock prices of DS-30 companies of 

Dhaka Stock Exchange. 

 
Table-3. Coefficient table of regression analysis. 

Variable Coefficient Std. Error t-Statistic Prob.(p-value) Decision on H0 

C 1.932142 2.245506 0.860448 0.3929  
EXP 0.992655 0.097534 10.17746 0.0003*** H0-1 accepted 

INT -0.486033 0.074792 -6.498472 0.0004** H0-2 accepted 

CAP -0.015400 0.082530 -0.186598 0.8526 H0-3 rejected 
TAX -0.022779 0.041635 -0.547117 0.5863 H0-4 rejected 

     

8.4. Model Fitness and Diagnosis Tests 

Tests for model fitness are quite positive in the empirical analysis. The Table 4 shows the fitness of the overall 

model. 

 
Table-4. Fitness of regression model. 

R-squared 0.709438 

Adjusted R-squared 0.690385*** 

S.E of regression 0.032876* 

F-stat 37.234449** 

No of observations 66 
 Note: (***), (**) and (*) indicates strongly, moderate and poor results. 

 

Here R-squared is 0.709 meaning that about 71% of independent variables explain change of dependent variable 

RRD. Summary results of model diagnostic tests are given below: 

 
Table-5. Overall results of diagnosis tests. 

Name of Tests Specific Model Remarks 
Normality Test Jarque-Bera Yes 
Model specification test Ramsey-Reset Test Support 

Heteroskedastiticity Breusch–Pagan–Godfrey(BPG) test. No 
Serial  auto correlation Durbin-Watson test/ Breusch- Godfrey Serial Correlation 

LM Test 
No 

Testing for multicollinearity Variance Inflation Factors (VIFs) No 
 

 

From Table 5 it is seen that the data used in the study are normally distributed. It is found that from Jarque-

Bera normality test. The estimated equation is stable and specified having 73% probabilities (Ramsey Reset test). 

The equation is free from Heteroskedasticity and serial auto correlation. So, the model has passed all of the standard 

tests. 



Financial Risk and Management Reviews, 2021, 7(1): 16-25 

 

 
22 

© 2021 Conscientia Beam. All Rights Reserved. 

9. CONCLUSION 

In this study, four explanatory variables namely interest rate, exchange rate, market capitalization, and tax rate 

were used in APT model to assess the significance of impact of these variables on return of DS-30 index. It is found 

that two explanatory variables namely interest rate and exchange rate has significant influence but market 

capitalization and tax rate have insignificant impact on return of DS-30 index. It is also found that rise of interest 

rate adversely affects the return of DS-30 index. So, monetary policy has immense importance in this regard. 

Explanatory variables used in APT model have failed to fully explain the change of return of DS-30 index due to 

presence of two insignificant variables. Further studies can be undertaken to assess the validity of APT model by 

incorporating other macroeconomic variables such as inflation rate, per capital income, foreign direct investment 

etc.  

 

Funding: This study received no specific financial support.    
Competing Interests: The authors declare that they have no competing interests.  
Acknowledgement: Authors are indebted to Dhaka Stock Exchange (DSE) and Bangladesh 
Bank authority for providing necessary data and reports. They also express their gratitude to 
different finance scholars, academicians, and officials of DSE who assist them through their 
invaluable suggestions, insight thoughts and constructive criticisms. 

 

REFERENCES 

Altay, E. (2003). The effect of macroeconomic factors on asset returns. A comparative analysis of the German and the Turkish 

stock markets in an APT foreign market (pp. 217-237). Germany: Financ: University Library of Munich. 

Arthur, L. M., Carter, C. A., & Abizadeh, F. (1988). Arbitrage pricing, capital asset pricing, and agricultural assets. American 

Journal of Agricultural Economics, 70(2), 359-365. Available at: https://doi.org/10.2307/1242076. 

Basu, D., & Chawla, D. (2012). An empirical test of the arbitrage pricing theory—the case of Indian stock market. Global 

Business Review. Global Business Review, 13(3), 421–432. Available at: https://doi.org/10.1177/097215091201300305. 

Cauchie, S., Hoesli, M., & Isako, D. (2004). The determinants of stock returns in a small open economy. International Review of 

Economics & Finance, 13(2), 167-185. Available at: 10.1016/j.iref.2003.07.001. 

Chen, N.-F., Roll, R., & Ross, S. A. (1986). Economic forces and the stock market. Journal of Business, 59(3), 383-403. 

Chowdhury, T. A. (2005). An overview of Bangladesh stock market (Vol. 45). Portfolio: Chittagong Stock Exchange Ltd. 

Dimitrova, D. (2005). The relationship between exchange rates and stock prices: Studied in a multivariate model. Issues in 

political Economy, 14(1), 3-9. 

Dimson, E., & Mussavian, M. (1999). Three centuries of asset pricing. Journal of Banking & Finance, 23(12), 1745-1769. 

Fama, E. F., & French, K. R. (1993). Common risk factors in the returns on stocks and bonds. Journal of Financial Economics, 

33(1), 3-56. Available at: https://doi.org/10.1111/1540-6229.00717. 

Fama, E. F., & French, K. R. (1995). Size and book-to-market factors in earnings and returns. The Journal of Finance, 50(1), 131-

155. 

Handa, P., & Linn, S. C. (1993). Arbitrage pricing with estimation risk. Journal of Financial and Quantitative Analysis, 28(1), 81-

100. 

Imam, M. O. (2001). Capital market development in Bangladesh: Problems and prospects. Portfolio: Chittagong Stock Exchange Ltd. 

Ingersoll, J. E. (1984). Some results in the theory of arbitrage pricing. The Journal of Finance, 39(4), 1021-1039. Available at: 

https://doi.org/10.1111/j.1540-6261.1984.tb03890.x. 

Iqbal, N., Khattak, S. R., Khattak, M. A., & Ullah, I. (2012). Testing the arbitrage pricing theory on karachi stock exchange. 

Interdisciplinary Journal of Contemporary Research in Business, 4(8), 839-853. 

Kandel, S., & Stambaugh, R. F. (1987). On correlations and inferences about mean-variance efficiency. Journal of Financial 

Economics, 18(1), 61-90. 

Koutmos, G., & Theodossiou, P. (1993). APT with observed factors and conditional heteroskedasticity. Managerial Finance, 

19(3/4), 24-39. 



Financial Risk and Management Reviews, 2021, 7(1): 16-25 

 

 
23 

© 2021 Conscientia Beam. All Rights Reserved. 

Lehmann, B. N., & Modest, D. M. (2005). Diversification and the optimal construction of basis portfolios. Management Science, 

51(4), 581-598. 

Mollik, A., & Bepari, M. K. (2010). Instability of stock beta in Dhaka stock exchange, Bangladesh. Managerial Finance, 36(10), 

886-902. Available at: https://doi.org/10.1108/03074351011070251. 

Nshom, A. M. (2007). The association of exchange rates and stock returns. Master’s Thesis, UMEA School of Business, UMEA 

University, Sweden  

Ramadan, I. Z. (2012). The validity of the arbitrage pricing theory in the Jordanian Stock market. International Journal of 

Economics and Finance, 4(5), 177-185. Available at: http://dx.doi.org/10.5539/ijef.v4n5p177. 

Sarver, L., & Philippatos, G. C. (1993). The arbitrage pricing theory and foreign exchange risk premia. Managerial Finance, 

19(3/4), 40-67. 

Shanken, J. (1985). Multi-beta CAPM or equilibrium APT? A reply. Journal of Finance, 40(4), 1189-1196. 

Sharpe, W. F., & Cooper, G. M. (1972). Risk-return classes of New York Stock Exchange common stocks, 1931-1967. Financial 

Analysts Journal, 28(2), 46-54. 

Zubairi, H. J., & Farooq, S. (2012). Testing the validity of CAPM and APT in the oil, gas and fertilizer companies listed on the Karachi 

stock exchange. Paper presented at the Financial Markets & Corporate Governance Conference.  

 

 APPENDIX  

1. Jarque-Bera Normality Test: 

: Variables are normally distributed. 

: Variables are not normally distributed. 

0

2

4

6

8

10

12

14

-0.08 -0.06 -0.04 -0.02 0.00 0.02 0.04 0.06 0.08

Series: RESID

Sample 1 67

Observations 66

Mean       6.30e-16

Median   0.004404

Maximum  0.083677

Minimum -0.075474

Std. Dev.   0.031849

Skewness   0.025388

Kurtosis   2.894579

Jarque-Bera  0.037653

Probability  0.981350


 
Figure-A.1. Normality test. 

    

From Figure A-1, it is revealed that p value of obtaining such a value from chi square is 0.9814(>0.05). 

Normality assumption of this study is not rejected. 

2.   Detection of heteroskedasticity:  

: The error variance is homoscedastic. 

:  The error variance is not homoscedastic. 

 
Table-A-1. Heteroskedasticity Test (Breusch-Pagan-Godfrey (BPG) Test). 

F-statistic 1.029456 Prob. F(4.61) 0.3994 

Obs *R-squared 4.173610 Prob. Chi-Square(4) 0.3830 

Scaled explained SS 3.377274 Prob. Chi-Square(4) 0.4968 
 

 

http://dx.doi.org/10.5539/ijef.v4n5p177


Financial Risk and Management Reviews, 2021, 7(1): 16-25 

 

 
24 

© 2021 Conscientia Beam. All Rights Reserved. 

From Table A-1, it is seen that the observed chi-square value 0.4968 has high p value of 0.40 which suggesting 

that it is failed to reject null hypothesis. So, the model does not suffer from heteroskedasticity. 

 

3. Model specification test (Ramsey-Reset Test): 

:  No specification error 

:  Having specification error 

 
Table-A-2. Ramsey-Reset Test (Omitted variables: Squares of fitted values). 

 Value df Probability 

t-statistic 0.462769 60 0.6452 

F-statistic 0.214155 (1,60) 0.6542 
Likelihood ratio 0.235151 1 0.6277 

 

 

Table A-2, showed that the P value of Ramsey’s Reset Test F-statistics value is quite high (0.65>0.05). So, we 

must accept null hypothesis and say that the equation is not mis-specified.  

 

4. Detection of serial auto correlation: 

:  There is no positive auto correlation               :  There is positive auto correlation 

 
Table-A-3. Brusch- Godfrey Serial Correlation LM test. 

F-statistic 40.72243 Prob. F(2,59) 0.083 

Obs *R-squared 38.27382 Prob. Chi-Square(2) 0.0012 
 

 
Table-A-4. Durbin watson test. 

R-squared 0.579906 Mean dependent var 6.30E-16 
Adjusted R-squared 0.537185 S.D. dependent var 0.031849 

S.E. of regression 0.021667 Akaike info criterion -4.726075 

Sum squared resid 0.027697 Schwarz criterion -4.493839 

Log likelihood 162.9605 Hannan-Quinn criter. -4.634308 
F-statistic 13.57414 Durbin-Watson stat 1.881036 
Prob (F-statistic) 0.000000   

 

 

Table A-4, showed that the estimated value of Durbin-Watson value is 1.88 which is near 2.0. It means that we 

may accept null hypothesis. So, there is no serial correlation in the study. For further testing, Breusch-Pagan LM 

test also has been used which is shown in Table A-3. Here, we see that p value .083 is above the level of significance 

(.05). So, null hypothesis is accepted.  

 

5. Multicollinearity Test: 

:  There is significant correlation among independent variables. 

:  There is no significant correlation among independent variables. 

 



Financial Risk and Management Reviews, 2021, 7(1): 16-25 

 

 
25 

© 2021 Conscientia Beam. All Rights Reserved. 

Table-A-5. Results of multicollinearity test. 

Variable Coefficient Variance Uncentered VIF Centered VIF 

C 5.042298 307900.5 NA 
EXP 0.951302 209191.9 5.517513 

INT 0.005594 226.0663 2.717109 

CAP 0.006811 5865.005 4.412733 

TAX 0.001734 2729.313 1.962352 
 

 

Table A-5 showed that no variable has centered VIF more than 10. So, we can say that variables do not suffer 

from strong relation within themselves or multicollinearity. 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Views and opinions expressed in this article are the views and opinions of the author(s), Financial Risk and Management Reviews shall not be responsible or 
answerable for any loss, damage or liability etc. caused in relation to/arising out of the use of the content. 

 


