modeling the random walk hypothesis for select industries listed in bse
DESCRIPTION
The objective of any professional investor is to carefully watch the volatility return of measure industries and companies. Among the various techniques of the volatility, random walk hypothesis tests the return of industries pure random or non random. This study made an attempt to study on random walk hypothesis with top Indian industries. The study was conducted for log return of industries from October 2011 to June 2014. various factors were carefully analyzed and interpreted. The finding and suggestions were made on the basis of giving a suggestion to the investors.TRANSCRIPT
-
International Journal of Management (IJM), ISSN 0976 6502(Print), ISSN 0976 - 6510(Online),
Volume 6, Issue 1, January (2015), pp. 527-535 IAEME
527 G.S. David Sam Jayakumar & A.Sulthan, Modeling the Random Walk Hypothesis for Select Industries
Listed in BSE (ICAM 2015)
MODELING THE RANDOM WALK HYPOTHESIS FOR
SELECT INDUSTRIES LISTED IN BSE
Dr. G.S. David Sam Jayakumar
Assistant Professor, Jamal Institute of Management, Tiruchirappalli
A. Sulthan
Research scholar, Jamal Institute of Management, Tiruchirappalli
ABSTRACT
The objectives of any professional investor is to carefully watch the volatility return of
measure industries and companies. Among the various techniques of the volatility, random walk
hypothesis tests the return of industries pure random or non random. This study made an attempt to
study on random walk hypothesis with top Indian industries. The study was conducted for log return
of industries from October 2011 to June 2014. various factors were carefully analyzed and
interpreted. The finding and suggestions were made on the basis of giving a suggestion to the
investors.
INTRODUCTION AND RELATED WORK
The empirical evidence in the random- walk literature existed before the theory was
established. That is so say empirical results were discovered first and then an attempt was made to
develop a theory that could possibly explain the results. After these initial occurrences, more results
and more theory were uncovered. This has led then to a diversity of theories. Which are generically
called the random walk theory. Worthington, Andrew C. & Higgs, Helen (2004)tests for random
walks and weak-form market efficiency in European equity markets. Daily returns for sixteen
developed markets (Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy,
Netherlands, Norway, Portugal, Spain, Sweden, Switzerland and the United Kingdom) and four
INTERNATIONAL JOURNAL OF MANAGEMENT (IJM)
ISSN 0976-6502 (Print)
ISSN 0976-6510 (Online)
Volume 6, Issue 1, January (2015), pp. 527-535
IAEME: http://www.iaeme.com/IJM.asp
Journal Impact Factor (2014): 7.2230 (Calculated by GISI)
www.jifactor.com
IJM
I A E M E
-
International Journal of Management (IJM), ISSN 0976 6502(Print), ISSN 0976 - 6510(Online),
Volume 6, Issue 1, January (2015), pp. 527-535 IAEME
528 G.S. David Sam Jayakumar & A.Sulthan, Modeling the Random Walk Hypothesis for Select Industries
Listed in BSE (ICAM 2015)
emerging markets (Czech Republic, Hungary, Poland and Russia) are examined for random walks
using a Augmented Dickey-Fuller (ADF). Ali F. Darrat and Maosen Zhong (2000) investigate, with
new daily data, whether prices in the two Chinese stock exchanges (Shanghai and Shenzhen) follow
a random-walk process as required by market efficiency. Srinivasan (2010) examined the random
walk hypothesis to determine the validity of weak-form efficiency for two major stock markets in
India. Similarly, Kabir Hassan, Al-Sultan (2003) explained unlike previous studies in stock market
efficiency literature on KSE, examines the weak-form efficiency by taking into consideration the
institution - al features of the KSE. On the other hand,Faiq Mahmood et.al (2006)endeavored to
examine the efficiency of Chinese stock market and how the global financial crisis influences the
efficiency of Chinese stock market. Hin Yu Chung (2006) also examined the random walk
hypothesis to determine the validity of weak-form efficiency for two major stock markets in China.
This study also provides evidence on the existence of the day-of-the-week effect on Chinese stock
market. Bizhan Abedini (2009) explored some evidence whether Kuala Lumpur Stock Exchange
(KLSE) is efficient in the weak form or not for the period January 2006 to June 2008 using daily
General Index. The methods used for the study are Autocorrelation Function test (ACF), Runs tests,
Variance ratio test and Unit root test (Augmented Dickey-Fuller Test (ADF)). Ankit Agarwal (2006)
tells us that the efficient market hypothesis has been and continues to be one of the most contentious
issues in finance..Nikolay Angelov (2009) explained that the development of capital markets in
transition and emerging market economies is a very important prerequisite for fostering their growth
as it allows the available funds to be channeled easily to their best investment uses. Kian-Ping Lim
(2000)provided a systematic review of the weak-form market efficiency literature that examines
return predictability from past price changes, with an exclusive focus on the stock markets. Korkmaz
(2001) explained in his study, that the efficiency concept was emphasized. Weak Form Market
Efficiency of Efficient Market Hypothesis was tested in Istanbul Stock Exchange (ISE). Ibrahim
Awad (2009) examined the efficiency of the Palestine Security Exchange (PSE) at the weak-level for
35 stocks listed in the market by using daily observations of the PSE indices: Alquds index, general
index, and sector indices. Parametric and nonparametric tests for examining the randomness of the
PSE stock prices were utilized. The parametric tests include serial correlation test, and Augmented
Dickey-Fuller (ADF) unit root tests.
METHODOLOGY AND TECHNIQUES OF DATA ANALYSIS
The objectives of any professional investor is to carefully watch the volatility return of
measure industries and companies. Among the various techniques in measuring the volatility.
Random walk hypothesis tests the return industries pure random or non random. This is study is
attempt to study on random walk hypothesis of top Indian industries. The study was conducted for
log return of industries from October 2011to June 2014.The monthly data of following industries
Automobile, Health Care, PSU, Capital Goods, Bank, Consumer Durable, FMCG, IT, Power, Metal,
Oil & Gas are considered. The analysis was conducted at different stages In Stage-1Multivariate test
of Normality for security returns was promoted and in Second stage, summary statistics for industry
returns and selection of optimal lag length are formulated, Stage-3 Augmented dickey-filler test is
used on industrial returns for understanding the random walk of the industrial return with ADF
regression equation.
-
International Journal of Management (IJM), ISSN 0976 6502(Print), ISSN 0976 - 6510(Online),
Volume 6, Issue 1, January (2015), pp. 527-535 IAEME
529 G.S. David Sam Jayakumar & A.Sulthan, Modeling the Random Walk Hypothesis for Select Industries
Listed in BSE (ICAM 2015)
MULTIVARIATE NORMALITY TESTS
Multivariate normality tests check a given set of data for similarity to the multivariate
normal distribution. The null hypothesis is that the data set is similar to the normal distribution,
therefore a sufficiently small p-value indicates non-normal data. Multivariate normality tests include
the Cox-Small test and Smith and Jain's adaptation of the Friedman-Rafsky test. Mardia's test is
based on multivariate extensions of skewness and kurtosis measures. Under the null hypothesis of
multivariate normality, the statistic A will have approximately a chi-squared
distribution with (1/6)*k(k + 1)(k + 2) degrees of freedom, and B will be approximately standard
normal N(0,1).Mardia's kurtosis statistic is skewed and converges very slowly to the limiting normal
distribution. For medium size samples (50 n
-
International Journal of Management (IJM), ISSN 0976 6502(Print), ISSN 0976 - 6510(Online),
Volume 6, Issue 1, January (2015), pp. 527-535 IAEME
530 G.S. David Sam Jayakumar & A.Sulthan, Modeling the Random Walk Hypothesis for Select Industries
Listed in BSE (ICAM 2015)
TABLE-2: SUMMARY STATISTICS FOR INDUSTRY RETURNS Industry
Name Mean
Standard
Deviation C.V Skewness Ex.kurtosis
AUTOMOBILE 0.0565792 1.29719 22.9270 0.175903 1.08326
HEALTHCARE 0.0723178 0.861176 11.9082 -0.185413 0.794022
PSU -0.0423310 1.14071 26.9473 -0.0875869 0.691177
CAPITALGOODS -0.0205306 1.55488 75.7348 -0.0348614 0.846336
BANK 0.0380047 1.60244 42.1642 0.196942 1.71437
CONSUMER
DURABLE 0.0529387 1.52398 28.7876 -0.351868 2.73797
FMCG 0.0902287 1.06559 11.8099 0.0585206 1.76327
IT 0.0654255 1.41274 21.5932 -0.461118 8.40083
POWER -0.0555093 1.25673 22.6401 -0.239719 0.930730
METAL -0.0432414 1.69574 39.2158 0.217300 1.18240
OIL &GAS 0.00843248 1.29422 153.481 0.0292861 0.276246
TABLE-3: SELECTION OF OPTIMAL LAG LENGTH
INDUSTRY
NAME
LOG
LIKELYHOOD MINIMUM AIC NO. OF LAGS
AUTOMOBILE -1649.60702 3.357578 3
HEALTHCARE -1235.14655 2.516034 3
PSU -1502.91225 3.075964 11
CAPITALGOODS -1823.76358 3.707134 1
BANK -1858.58515 3.777838 1
CONSUMER DURABLE -1813.54191 3.686380 1
FMCG -1462.33178 2.973262 1
IT -1726.96985 3.516690 4
POWER -1619.64059 3.294702 2
METAL -1907.98357 3.894383 9
OIL GAS -1641.41990 3.353137 9
TABLE- 4:AUGMENTED DICKEY-FULLER TEST FOR INDUSTRIAL RETURNS
INDUSTRY NAME NO. OF LAGS TAU- STATISTIC P-VALUE
AUTOMOBILE 3 -16.2014
-
International Journal of Management (IJM), ISSN 0976 6502(Print), ISSN 0976 - 6510(Online),
Volume 6, Issue 1, January (2015), pp. 527-535 IAEME
531 G.S. David Sam Jayakumar & A.Sulthan, Modeling the Random Walk Hypothesis for Select Industries
Listed in BSE (ICAM 2015)
TABLE-5: AUGEMENTED DICKEY- FULLER REGRESSION OF AUTOMOBILE
INDUSTRY
DEPENDENT VARIABLE (Yt) INDEPENDENT
VARIABLE CO-EFFICIENT
STANDARD
ERROR T-RATIO P- VALUE
Yt-1 -0.954715 0.0589279 -16.20 0.000
Yt-1 0.0525487 0.0507943 1.035 0.3011
Yt-2 0.0840064 0.0426575 1.969 0.0492
Yt-3 0.0132796 0.0317385 0.4184 0.6757
TABLE-6: AUGEMENTED DICKEY- FULLER REGRESSION OF HEALTH CARE
INDUSTRY
DEPENDENT VARIABLE (Yt) INDEPENDENT
VARIABLE CO-EFFICIENT
STANDARD
ERROR T-RATIO P- VALUE
Yt-1 -0.932412 0.0590844 -15.78 0.000
Yt-1 0.0449442 0.0506426 0.8875 0.3750
Yt-2 0.0572486 0.0422931 1.354 0.1762
Yt-3 -0.0293119 0.0316401 -0.9264 0.3545
TABLE-7: AUGEMENTED DICKEY- FULLER REGRESSION OF PSU INDUSTRY
DEPENDENT VARIABLE (Yt) INDEPENDENT
VARIABLE CO-EFFICIENT STANDARDERROR T-RATIO P- VALUE
Yt-1 -0.927421 0.100902 -9.191 0.000
Yt-1 0.0548600 0.0963860 0.5692 0.5694
Yt-2 0.111401 0.0924200 1.205 0.2283
Yt-3 0.0332749 0.0883061 0.3768 0.7064
Yt-4 0.0228120 0.0828515 0.2753 0.7831
Yt-5 -0.0134660 0.0774038 -0.1740 0.8619
Yt-6 -0.0365102 0.0709886 -0.5143 0.6072
Yt-7 0.00462035 0.0642714 0.07189 0.9427
Yt-8 -0.0274312 0.0573316 -0.4785 0.6324
Yt-9 0.0678314 0.0496639 1.366 0.1723
Yt-10 0.0753571 0.0422990 1.782 0.0751
Yt-11 -0.00269448 0.0320442 -0.08409 0.9330
TABLE-8: AUGEMENTED DICKEY- FULLER REGRESSION OF CAPITAL GOODS
INDUSTRY
DEPENDENT VARIABLE (Yt) INDEPENDENT
VARIABLE CO-EFFICIENT STANDARDERROR T-RATIO P- VALUE
Yt-1 -0.865620 0.0417141 -20.75 0.000
Yt-1 -8.55909e-05 0.0317015 -0.002700 0.9978
-
International Journal of Management (IJM), ISSN 0976 6502(Print), ISSN 0976 - 6510(Online),
Volume 6, Issue 1, January (2015), pp. 527-535 IAEME
532 G.S. David Sam Jayakumar & A.Sulthan, Modeling the Random Walk Hypothesis for Select Industries
Listed in BSE (ICAM 2015)
TABLE-9: AUGEMENTED DICKEY- FULLER REGRESSION OF BANKING INDUSTRY
DEPENDENT VARIABLE (Yt)
INDEPENDENT
VARIABLE CO-EFFICIENT STANDARDERROR T-RATIO P- VALUE
Yt-1 -0.920905 0.0421715 -21.84 0.000
Yt-1 0.0389763 0.0316758 1.230 0.2188
TABLE-10: AUGEMENTED DICKEY- FULLER REGRESSION OF CONSUMER GOODS
INDUSTRY
DEPENDENT VARIABLE (Yt) INDEPENDENT
VARIABLE CO-EFFICIENT STANDARDERROR T-RATIO P- VALUE
Yt-1 -0.963386 0.0431395 -22.33 0.000
Yt-1 0.0386268 0.0316752 1.219 0.2230
TABLE-11: AUGEMENTED DICKEY- FULLER REGRESSION OF FMCG INDUSTRY
DEPENDENT VARIABLE (Yt) INDEPENDENT
VARIABLE CO-EFFICIENT STANDARDERROR T-RATIO P- VALUE
Yt-1 -0.949519 -0.0438541 21.65 0.000
Yt-1 -0.00859111 0.0316886 -0.2711 0.7864
TABLE-12: AUGEMENTED DICKEY- FULLER REGRESSION OF IT INDUSTRY
DEPENDENT VARIABLE (Yt) INDEPENDENT
VARIABLE CO-EFFICIENT STANDARDERROR T-RATIO P- VALUE
Yt-1 -0.933471 0.0696107 -13.41 0.000
Yt-1 -0.0123527 0.0626152 -0.1973 0.8436
Yt-2 -0.0569377 0.0536147 -1.062 0.2885
Yt-3 -0.0863471 0.0435286 -1.984 0.0476
Yt-4 -0.0232004 0.0317325 -0.7311 0.4649
TABLE-13: AUGEMENTED DICKEY- FULLER REGRESSION OF POWER INDUSTRY
DEPENDENT VARIABLE (Yt) INDEPENDENT
VARIABLE CO-EFFICIENT STANDARD ERROR T-RATIO P- VALUE
Yt-1 -0.897502 0.0509804 -17.60 0.000
Yt-1 -0.0223555 0.0430953 -0.5187 0.6041
Yt-2 0.0333378 0.0317127 1.051 0.2934
-
International Journal of Management (IJM), ISSN 0976 6502(Print), ISSN 0976 - 6510(Online),
Volume 6, Issue 1, January (2015), pp. 527-535 IAEME
533 G.S. David Sam Jayakumar & A.Sulthan, Modeling the Random Walk Hypothesis for Select Industries
Listed in BSE (ICAM 2015)
TABLE-14: AUGEMENTED DICKEY- FULLER REGRESSION OF METAL INDUSTRY
DEPENDENT VARIABLE (Yt)
INDEPENDENT
VARIABLE CO-EFFICIENT STANDARDERROR T-RATIO P- VALUE
Yt-1 -0.837903 0.0958598 -8.741 0.000
Yt-1 -0.133425 0.0919120 -1.452 0.1469
Yt-2 -0.0580228 0.0868697 -0.6679 0.5043
Yt-3 -0.106771 0.0813799 -1.312 0.1898
Yt-4 -0.153140 0.0748676 -2.045 0.0411
Yt-5 -0.142044 0.0681419 -2.085 0.0374
Yt-6 -0.164812 0.0605182 -2.723 0.0066
Yt-7 -0.114785 0.0526939 -2.178 0.0296
Yt-8 -0.110444 0.0443702 -2.489 0.0130
Yt-9 -0.0291774 0.0318417 -0.9163 0.3597
TABLE-15: AUGEMENTED DICKEY- FULLER REGRESSION OF OIL & GAS
INDUSTRY
DEPENDENT VARIABLE (Yt) INDEPENDENT
VARIABLE CO EFFICIENT STANDARDERROR T-RATIO P- VALUE
Yt-1 -1.04475 0.107867 -9.686 0.000
Yt-1 0.0463984 0.102453 0.4529 0.6507
Yt-2 0.0713329 0.0955918 0.7462 0.4557
Yt-3 -0.0153071 0.0888546 -0.1723 0.8633
Yt-4 -0.0637455 0.0817805 -0.7795 0.4359
Yt-5 -0.0656524 0.0737200 -0.8906 0.3734
Yt-6 -0.0426841 0.0646261 -0.6605 0.5091
Yt-7 -0.0602768 0.0548418 -1.099 0.2720
Yt-8 -0.104721 0.0450953 -2.322 0.0204
Yt-9 -0.00959004 0.0320666 -0.2991 0.7650
DISCUSSION
Table-1 visualises the result of Mardia Multi-Variate test of normality such as Mardias
Skewness test, MardiasKutosis test and HenzeZirklertest. The test was applied for the returns of top
securities listed in BSE. The result of the test confirms that the security returns of securities are
departed from Multi-Variate normality and the returns are non-normally distributed. Hence, the
researcher assumed that the returns of securities are non-normally distributed. Table-2 exhibits the
descriptive statistics of security returns of Automobile Companies. Among the top 5 companies in
automobile industrythe Mean Returns of Bajaj Auto Ltd., and Tata Motors Ltd., are same followed
by Hero Motor Corp., and Maruti Suzuki Ltd., respectively. As for as Maruti Suzuki is concerned the
Standard Deviation of returns are less compared to the remaining companies. This shows the security
returns of Maruti Suzuki Ltd., is less volatile and highly consistent. Finally the Uni-Variate
Skewness, Kurtosis, Sharpe-Wilk Statistics and Anderson-Darling Statistics confirms that the returns
of automobile companies are departed from UniVariate normality and it follows the non-normal
distribution. Table-3Exhibit the selection of optimal lags length for the result of top industries listed
-
International Journal of Management (IJM), ISSN 0976 6502(Print), ISSN 0976 - 6510(Online),
Volume 6, Issue 1, January (2015), pp. 527-535 IAEME
534 G.S. David Sam Jayakumar & A.Sulthan, Modeling the Random Walk Hypothesis for Select Industries
Listed in BSE (ICAM 2015)
in BSE.The minimum AIC ( Akaike Information criteria). So the optimal lag length for the return of
automobile industries. The followed by the lag length for Health care, PSU, Capital goods are the 11,
1 respectively. Finally the return of PSU industry is have the lag length of 11 days and if is greater
combined to remember industries. Table-4 Exhibit the unit root test are Augmented Dickey-Filler
test which explains accepted or rejected are random walk hypothesis. The TAU statistics of the
Dickey-Filler test for all the industry statistical perform on present level. Hence reject the Null
hypothesis of random walk and accepted the Alternative hypothesis. Which reviews the all industry
do not follow a random walk is having a pattern. Table-5 to Table-15 Visualize the Augmented
Dickey-Filler test for selected industries. In order to perform ADF refers in researcher Yt is
dependent variable in independent variable Yt-1, Yt-2, Yt-3 and etc.., respectively. The shows
the first and endured result in the industries return. Yt-1, Yt-2 , Yt-3 in the return lags of the industry
returns. The result compare that Yt-1, Yt-2, Yt-3 is that of 1 % and 5 % significant level. The
comparison rejects random walk hypothesis for all selected return of industries. The following
suggestions were given to the investor:The return of auto mobile industry is high in all the years.
Which reveals that accordingly to the augmented dickey-filler test, if the investors invest their funds
in the auto mobile industry. They can achieve a maximum possible returns while compared to the
other industries. Moreover the returns of PSU, metal , oil & gas industry are greater when compared
to other industries in over all period basis. The returns earned from oil &gas, metal, PSU industries
follows the above maximum yielding industries. After analyzing the top 10 companies data, the
following recommendations were suggested to investors to invest and cautions to invest the market
share respectively. We find that the following industries is less risk because the unsystematic risk or
marketable risk is less for securities. This shows the random walk hypothesis risk of the following
industries is less compared to non-marketable risk of the industry. The are automobile, health care,
PSU, capital goods, bank, consumer durable, FMCG, IT, power, metal, oil &gas. So investors will
made investment in this industries it avoid risk of loss of securities.
CONCLUSION
Based on the analysis, the researcher comes to a concrete conclusion. This study deals with
the Random walk hypothesis the top industries listed in BSE. At first the researcher observe the
returns of the industries are not normally distributed and its having a different pattern. More over the
researcher emphasis the investors to look into the average returns of the security and also the
industries involved before investing their funds. If the invest observe the industries they can use the
returns of auto mobile industry is play a vital role. Followed by metal, oil & gas, banking industries.
Finally the random walk hypothesis is the most important of an investors whether he/she may be a
individual institutional investor. According to result of the analysis the researcher recommends the
investors to invest their funds in auto mobile, metal, PSU, oil & gas, banking industries. Then only
they can earn a maximum return with the nominal risk.
REFERENCE
[1] Ali F.Darrat and Maosen Zhong (2000) On Testing the Random-Walk Hypothesis: A Model-Comparison Approach Volume 35, Issue 3, pages 105124.
-
International Journal of Management (IJM), ISSN 0976 6502(Print), ISSN 0976 - 6510(Online),
Volume 6, Issue 1, January (2015), pp. 527-535 IAEME
535 G.S. David Sam Jayakumar & A.Sulthan, Modeling the Random Walk Hypothesis for Select Industries
Listed in BSE (ICAM 2015)
[2] Ankit Agarwal (2006).Testing weak form of Market efficiency by application of simple technical trading rules on the Indian stock market, Dissertation.
[3] Kantesha Sanningammanavara and Kiran Kumar K V, Economic Indicators and Stock Market Performance - An Empirical Case of India, International Journal of Management
(IJM), Volume 5, Issue 8, 2014, pp. 107 - 114, ISSN Print: 0976-6502, ISSN Online:
0976-6510.
[4] Bizhan Abedini(2009).An Evaluation of Efficiency of Kuala Lumpur Stock Exchange, Interdisciplinary Journal of Contemporary Research in Business, August 2009, vol 1, no4.
[5] Darrat, A. F., and Zhong, M. (2000): On testing the random-walk hypothesis: a model-comparison approach. The Financial Review, vol. 35, no. 3, 105-124.
[6] Dickey, D. A., and Fuller, W. A. (1981): Likelihood ratio statistics for autoregressive time series with a unit root. Econometrica, vol. 49, 1057- 1072.
[7] Faiq Mahmood, Xia Xinping, Humera Shahid, Muhammad Usman(2006) Global Financial Crisis: Chinese Stock Market Efficiency, ISSN 2229 3795, Asian Journal of management
research.
[8] Kabir M. Hassan, Waleed S. Al-Sultan, and Jamal A. Al-Saleem (2003) Stock Market Efficiency in the Gulf Cooperation Council Countries (GCC): The Case of Kuwait Stock
Exchange. Scientific journal of administrative development, Vol 1 No 1, I.A.D.
[9] P. Srinivasan (2010), Testing Weak-Form Efficiency of Indian Stock Markets, APJRBM Volume 1, Issue 2, November ISSN 2229-4104.
[10] Worthington, Andrew C. & Higgs, Helen (2004) Random walks and market efficiency in European equity markets. Global Journal of Finance and Economics, 1(1), pp. 59-78.