Random Walk of Security Prices:

Empirical Evidence from KSE, LSE, and ISE

Yasir Kamal and Dr. Kashif-Ur-Rehman*

SZABIST
Islamabad, Pakistan

Abstract: Previously security market research had been Investors in weak-form efficient markets cannot expect to
focused mainly on developed economies with no attention find any patterns in the historical sequence of stock prices
paid to the security markets of developing countries of that would provide insight into future price movements
South East Asia. In an attempt to fill this gap in the and allow them to earn an abnormal rate of return. The
literature, this paper conducts an empirical investigation security prices fluctuate randomly if a market is efficient,
of the random walk of security prices in Pakistani stock and the degree of randomness of security prices increases
markets. The Augmented Dickey fuller test, Ljung Box Q with an increase in market efficiency. The most efficient
test, Variance ratio test and a non parametric Run test has market of all is one in which price changes are completely
been used for analysis of Random walk of security prices. random and unpredictable. For these reasons, the Random
Results indicate the presence of some predictable Walk Hypothesis and its close relative, the Efficient
elements, which contradict with previous studies on Markets Hypothesis, have become icons of modern
Karachi stock market. This is because of the difference in financial economics that continue to fuel the imagination
number of observation used in previous studies and this of academics and investment professionals alike.
particular study. To conclude, the Karachi stock exchange
and Islamabad stock exchange does show a weak random Augmented Dickey Fuller Unit Root Test and Serially Un-
walk of security prices, while Lahore stock exchange show Correlation Method have been used for the detection of
strong random walk of security prices. random walk in time series data, see, among others [4],
[5], [6] and [7]. Reference [8] shows that common stock
Keywords: Random Walk of Security Prices, Efficient prices have properties analogous to the movement of
Market hypothesis, ADF, Run Test, Variance ratio test, molecules. Osborne applies the methods of statistical
Ljung Box Q-statistics. mechanics to the stock market, with a detailed analysis of
stock price fluctuations from the point of view of a
1. INTRODUCTION physicist.

The behavior of security prices is one of the affluently Many research works on stock behavior suggest that the
documented works in empirical finance; the most enduring expected value of speculative strategies should be zero. In
model used for this purpose is the random walk an efficient market the stock prices would reflect all the
hypothesis. The Random Walk Hypothesis has an available information and as a result of different favorable
illustrious history, with remarkable intellectual forbears and unfavorable news the stock prices varies differently.
such as Bachelier, Einstein, L'evy, Kolmogorov, and These different variations are termed as random movement
Wiener. Reference [1] gives first contribution to literature and in econometric terminology it is called random walk
by using random walk hypothesis for financial markets. It of security prices.
was firmly believed amongst financial empiricists that
stock market price should reflect the intrinsic value of The purpose of this paper is to investigate evidence in
underlying assets. Over the past two decades ample support of random walk hypothesis in the Pakistani stock
research work has been undertaken to test the efficient markets (LSE, ISE and KSE) using three techniques: the
market hypothesis – the claim that a market in which run test, Augmented Dickey fuller test for unit root,
prices fully reflects available information is an efficient Variance ratio test and Ljung–Box Q test. Of the four tests
one. It was strongly argued that there were no used, three examine inter temporal structure of stock
opportunities for investors to make abnormal profits by returns and one examines unit root stationarity in the price
exploiting information contained in the history of revelation process. As [9] points out, if the evidence fails
fundamental market data. Reference [2] categorizes three to support weak-form tests there is no reason to examine
forms of market efficiency: weak, semi-strong and strong. semi-strong (and/or strong) forms before declaring the
These three forms differ in terms of the types of market is inefficient on the evidence.
information which are used in developing investment
strategies. As [3] points out, a sufficient condition for The significance of random walk hypothesis: The
weak-form efficiency is that the stock prices fluctuate significance of the random walk theory can be explained
randomly. As a result; a market is efficient in the weak both practically and theoretically. From practical point of
form if stock prices follow a random walk process. view canvasser of the hypothesis deals with the absence or
presence of systematic elements in their empirical studies,
and the detection of some nonrandom components is not
* Bahria University, Islamabad evidence against the RWH unless it can be shown that

Journal of Independent Studies and Research (JISR)

Volume 4, Number 1, January 2006 17
systematic tendencies in past price behavior alone present to out perform a buy-and-hold strategy in managing a
unexploited opportunities to make above normal profits. portfolio”.

2. REVIEW OF LITERATURE According to [21] , [15] and [22] most of the research has
conducted on stock market of developed countries and less
Believer of the efficient-markets concept also tend to attention has been paid to emerging stock markets, so this
espouse the concept of the random walk that the market study is a step toward research on emerging stock markets.
behaves in discernible way. The advocates of random walk
holds that it is impossible to predict the prices of a security Researchers have consensus that the power of variance
from the past performance because changes in economic ratio test is superfluous than any other test used for
condition, securities valuations, corporate profits and random walk [10]. Reference [23] derived the non-
market as a whole all occur in a myriad of different ways. overlapping VR (NVR) statistic, which follows a Beta
In random walk process, successive stock returns must be distribution. As argued by [10], the OVR test is expected
identically distributed and independent so that the to have higher power than the NVR test. The advantages
correlation between one period’s return and the immediate of the VR test are summarized by [24]. The application of
following period is zero, [9], [10], and [11]. the Variance Ratio test to measure the time series data was
observed by [25], [26] and [27].
So the technical analyst would exploit any non-random
fluctuation or speculators would buy before an expected For Karachi stock market [28] observed the random walk
rise in price or sell short before an expected fall in price. of security prices traded in KSE 100 index. [29] took the
In random walk the flow of information is random, and if event of 9/11 and check out the presence of predictable
security prices adjusted with that information so the new element in stock return, his finding suggested that before
security prices would also be randomly attuned, and hence 9/11 the predictable element exists in stock returns (e.g.
each day securities have different prices depending on the week day effect) but the security prices also followed a
flow of information. And no body can predict about the random walk. The study about the market efficiency of
future security prices. Wall street journal has different Karachi stock market [30] suggested that after 9/11,
thinking on the subject. It believes that past stock prices do security prices followed a random walk and no calendar
show foreseeable trends and that it is possible to forecast related anomalies exist in KSE 100 index.
the market based on past performance.
3. DATA & METHODOLOGY
So by taking this issue several studies have uncovered
empirical evidence, which suggests that stock returns 3.1 Data
contained predictable components (e.g. see [12], [13] and
[14]. Despite these staggering findings their results of Daily stock index data was taken for Karachi stock
failing to reject the random walk applies only to the stock exchange, Lahore stock exchange, and Islamabad stock
markets of industrialized nations. It is of great interest to exchange from Business recorder, which is a well known
explore if the similar patterns can be identified in the and reliable source of business indicators in Pakistan. The
Asian stock markets. Reference [15] reports that stock daily index for all three markets consists of five year data
returns of emerging countries are highly predictable and and included almost 1200 observations for each stock
have low correlations with stock returns of developed market.
countries. He suggested that emerging markets are less
efficient than developed markets and that higher return 3.2 Measure of Daily Return
and lower risk can be obtained by incorporating emerging
market stocks in investor’s portfolios. Reference [16] used The daily closing value of KSE, LSE and ISE were used
variance ratio test to find out random walk for Asian stock for calculating the daily returns. The continuously
market and rejected random walk for all the observed compounded annual rate of return is a well-accepted
Asian stock markets. Reference [17] findings suggested approach to measuring the daily returns. The natural log of
that random walk hypothesis for London stock market is daily closing index value is, thus the measure of daily
rejected but the weak form of market efficiency can not be return used for this study.
rejected. Reference [18] findings suggest that the monthly
stock price for the Taiwan stock market exhibits weak- Following is the formula used to find the return:
form efficiency. Reference [19] used variance ratio test  I 
and reject the random walk hypothesis for Latin American Rt = Ln  t 
stock markets, however run test indicate that Latin  I t −1 
American equity markets are weak-form efficient. where:
According to [20] “Thus an accurate statement of the Rt = return on day ‘t’
narrow form of the random-walk hypothesis goes as It = index closing value on day ‘t’
follows: the history of stock price movements contains no It-1 =index closing value on day ‘t-1’
useful information that will enable an investor consistently Ln= natural log.

Journal of Independent Studies and Research (JISR)

Volume 4, Number 1, January 2006 18
3.3 Method and Procedure If the null hypothesis of randomness is sustainable,
following the properties of the normal distribution, we
The ADF unit root, Variance ratio test, Ljung Box Q should expect that
statistics and a non parametric Run test has been used for
the analysis of the data. The details of method Prob [E ( R) − 1.96σ R ≤ R ≤ E ( R) + 1.96σ R ] = 0.95 2
implemented are given below:
3.6 Variance Ratio Test
3.4 Augmented Dickey Fuller (ADF) Unit root test1
Lo and MacKinlay variance ratio test is derived on the
The Augmented Dickey Fuller (ADF) unit root test is assumption that if the natural logarithm of a time series Pt
given as: is a pure random walk, the variance of its qth difference
grows proportionally with the difference q, that is the
m
∆Yt = β 1 + β 2 t + δYt −1 + α i ∑ ∆Yt −1 + ε t ------ (A) variance of its qth difference variable would be q times the
i =1 variance of its first difference. So if we obtain n+1
observations Po , P1 , P 2 ,…..,Pn at equally spaced
Where εt is a pure white noise error term and where
intervals, 1/q of the variance of Pt , Pt-q is expected to be
∆Yt −1 = (Yt −1 − Yt − 2 ) , the same as the variance of Pt –Pt-1 , for a time series
characterized by random walks.
∆Yt − 2 = (Yt − 2 − Yt −3 ) , etc.
The variance ratio, VR (q), is defined as
The number of lagged difference terms to include is often
determined empirically, the idea being to include enough
σ q2
VR (q) = 2
terms so that the error term in equation is serially σ1
uncorrelated. Since Eq: “A” is a first-order difference or
equation, so the stability condition requires that P > 1
1 Var ( Pt − Pt − q )
3.5 Run Test =1
q Var ( Pt − Pt −1 )
If a series of positive and negative residual occurred, then
the randomness of these residual can be checked through Where σ q2 is the unbiased estimator of 1/q of the variance
runs test which is also known as Geary test, a
of qth difference of the logged security return (Pt-Pt-q) and
nonparametric test. The run test determined whether the
successive price changes are independent. Unlike to any σ 12 is an unbiased estimator of the variance of the logged
parametric test the assumption of normal distribution is return (Pt-Pt-1). Reference [10] demonstrates that the
not applied on run test. If the return series exhibits greater
σ q2 and σ 1
2
estimators can be computed by:
tendency of change in one direction, the average run will
be longer and the number of runs fewer than that 1 n
generated by a random process. σ q2 = ∑
m t =q
( Pt − Pt − q − qu ) 2

Under the null hypothesis that the successive outcomes are

independent and assuming that N1>10 and N2>0, the 1 n
number of runs is (asymptotically) normally distributed
with
σ 12 = ∑
n − 1 t =1
( Pt − Pt −1 −u ) 2

2 N1 N 2 Where
Mean: E(R) = +1 q
N m = q (nq − q + 1)(1 − )
nq
2 N1 N 2 (2 N1 N 2 − N ) and
Variance: σ R2 =
( N ) 2 ( N − 1) 1
u= ( Pnq − Po )
Where N=N1+N2
nq

2
Decision rule: Do not reject the null hypothesis of randomness with
Decision rule: the computed τ value should be more negative than
1
95% confidence if R the number of runs, lies in the preceding confidence
the critical τ value, and the large negative τ value is generally an interval; reject the null hypothesis if the estimated R lies outside these
indication of stationarity. limits.

Journal of Independent Studies and Research (JISR)

Volume 4, Number 1, January 2006 19
And Po and Pnq are the first and last observations of the k r j2
time series and n is the sample size. The first test statistic, QLB = T (T + 2)∑
Z (q) is derived under the assumption of homoscedasticity, j =1 T−j
with the asymptotic variance of the VR statistic φ (q )
defined as: Where rj is the jth autocorrelation and T is the number of
observations. The Q statistic is often used as a test of
2(2q − 1)(q − 1) whether the series is white noise. Q is asymptotically
φ (q) = distributed as a chi square distribution with degree of
3q (nq ) freedom equal to the number of autocorrelations. This test
was developed by Ljung & Box, in 1979. To compute auto
And the associated standard Z test statistic, Z (q) as: correlation and Ljung-Box Q statistics, we compute
correlogram, which graphs the value of auto correlation at
successive lags against the length of the lag and the last
VR(q ) − 1 two columns reported in correlogram are the Ljung-Box
Z (q ) = → N (0,1) Q-statistics and their P values.4
φ (q)
4. DISCUSSION AND ANALYSIS
There is a growing consensus amongst finance empiricist
that volatility is time varying as documented by most of
4.1 ADF unit root test
the researchers that variances of most stock returns are
conditionally heteroscedastic with respect to time. As a
An Augmented Dickey-Fuller (ADF) test is used for a unit
result, a linear relation does not exist over the observation
root in time series sample; it is the advance version of the
intervals. To overcome this difficulty, Lo and McKinley
Dickey-Fuller test. The greater the negative value of the
derive the heteroscedasticity-consistent variance
ADF test statistic the greater the chance of the rejection of
estimator φ *(q) is given by: the hypothesis of having a unit root.

q
2(q − j ) 2 ^ From table 1, the ADF test result shows that for all the
φ *(q) = ∑ [ ] δ ( j) three stock markets returns, we can reject the null
j =1 q hypothesis of unit root. It was recommended by many
financial researchers that existence of random walk
Where component does not mean that future stock return is
q −1 ^ unpredictable. If stock returns are characterized by a white
∑ (P − P t t −1 − u ) 2 ( Pt − j − Pt − j −1 − u ) 2 noise process, the correspondence price indices are said to
t = j +1 follow the random walk.
δ ^
(J) = nq ^
[∑ ( Pt − Pt −1 − u ) 2 ] 2 In that case that stock returns are considered to be
t =1 unpredictable. If the stock returns do not follow white
noise or they are integrated of order one or I (1), there
The VR statistic can be standardized asymptotically to a exists some predictable components. The ADF is used to
standard normal test-statistic, Z*(q) which as reported by find out only the stochastic trend components while the
[10] Lo and McKinley (1988), is computed as: purpose variance ratio approach is to detect if the short-
term fluctuations dominate the stochastic trend
VR (q ) − 1 components or not. According to [30] the Karachi Stock
Z*(q) = .... → N (0,1) 3 market follows a random walk for its daily returns, the
φ * (q) same results was also determined by [28].

3.7 Ljung Box-Q Statistics Table 1: ADF Test Results

MacKinnon critical values
ADF Test R-
The first approach we use here is to consider the Statistic squared
for rejection of hypothesis
autocorrelation structure of stock returns and test the joint of a unit root.
significance of the autocorrelation using the Ljung–Box KSE
-34.3 0.491 1% Critical Value -2.56
portmanteau statistic (Q). The Ljung–Box Q-statistics are LSE -35.5 0.507 5% Critical Value -1.93
given by: ISE -27.8 0.389 10% Critical Value -1.66

3 4
Decision Rule: if Z (q) and Z*(q) calculated values is less than Z Decision rule: Reject null hypothesis under LB statistics if calculated
tabulated values then we reject Ho: VR (q) =1, Z value of LB is greater than tabulated value of chi square distribution table
tabulated =1.96 for 5% significance level. values with ‘m’ d.f.

Journal of Independent Studies and Research (JISR)

Volume 4, Number 1, January 2006 20
4.2 Non-Parametric Runs Test Table 2
Runs Test
A non-parametric runs test has been commonly used to
examine the random walk hypotheses which further KSE LSE ISE
implies the existence of weak-form efficiency in stock Test Value 0.001 0.002 0.001
markets return (e.g. see, [19] and [32] ). Cases < Test Value 584 643 558
Cases >= Test Value 638 582 666
A non-parametric runs test is applicable here as a test of
Total Cases 1222 1225 1224
randomness for the sequence of the return. In other words,
it tests whether Pakistani stock markets returns are Number of Runs 571 571 541
predictable. If the successive change in the returns is Z -2.282 -2.349 -3.87
random, it is an indication of weak-form efficiency. A test Asymp. Sig. (2-tailed) 0.022 0.019 0.007
for a run up and down requires dividing the return series
into two different types, increasing and decreasing. A homoscedastic and heteroscedastic consistent Z (q) and
positive sign (+) is assigned if the return in the sequence is Z*(q) test statistics.
higher than the preceding return, whereas a negative sign
(-) is assigned if the return in the sequence is smaller than The table no 3 reported that except LSE (under the
the preceding return. A zero (0) is assigned if there is no assumption of heteroscedasticity) no other stock market
change in the returns. For example, a series of- - - - - + + + returns follows a random walk hypothesis and hence the
+ - - - - -00000 + + + + would indicate five runs-two market efficiency concept was void in this prediction of
positive runs of length four, two negative runs of length variance ratio test. As concluded by [30] and [28] the
five and one zero run of length five. presence of random walk in security price for the KSE 100
index has no more validation under these circumstances.
The run-up and run-down test identifies the existence of Now to remove these discrepancies in results further
persistence but does not indicate the direction and degree theoretical supports are needed.
of persistence. Rejection of the null hypothesis of
randomness indicates the persistence in returns does exist, As for the [28] study is concerned he used the standard
which violates the weak-form of market efficiency. tests for autocorrelations, like Ljung Box Q statistics, ADF
test for unit root and correlogram method to find the
As suggested in the table 2 that, the successive return for random walk of security prices. So one reason of this
all the markets do not show independence and hence upon discrepancy is that [28] did not use a higher power test like
the above results we can reject the null hypothesis of Variance ratio test, the other reason for inconsistency is
randomness which is inconsistent with [30] and [28] study that [28] used a weekly data for his analysis while in this
for KSE 100 index. research daily observation has is used. Now if we see
Ljung Box Q statistics of our analysis we can easily
4.3 Variance ratio test conclude that it was consistent with our variance ratio
statistic results.
Variance ratio test is developed by [10] and it is more
powerful than any standard autocorrelations tools. Under The LB-Q statistics indicate auto correlation for both ISE
the null hypothesis of variance ratio test the stock return and KSE but no autocorrelation for LSE at any observe
follows a random walk and the variance ratio are expected lags. The validity of Z (q) depends to some extent on its
to be equal to one. Table no.3 reports the estimated asymptotic distribution towards normality and one can
variance ratios for lags q = 2, 4, 8, 12, 16, and 20 for both
K=4 K=8 K=12 K=16 K=20 K=24 K=28 K=32 K=36 K=40
KSE 0.256 0.120 0.086 0.059 0.051 0.045 0.037 0.031 0.029 0.026

(-13.9)* (-10.4) * (-8.52) * (-7.47) * (-6.67) * (-6.09) * (-5.66) * (-5.31) * (-5) * (-4.75) *

(-7.38)* (-6.27) * (-5.49) * (-5) * (-4.58) * (-4.27) * (-4.05) * (-3.87) * (-3.71) * (-3.58) *

LSE 0.243 0.123 0.083 0.063 0.042 0.036 0.032 0.029 0.026 0.243

(-14.2)* (-10.37)* (-8.56)* (-7.45)* (-6.12)* (-5.67)* (-5.31)* (-5.01)* (-4.75)* (-14.16)*

(-1.162) (-1.153) (-1.15) (-1.15) (-1.148) (-1.147) (-1.145) (-1.143) (-1.141) (-1.162)

ISE 0.257 0.124 0.090 0.064 0.054 0.048 0.039 0.036 0.034 0.030

(-13.9)* (-10.36)* (-8.48)* (-7.44)* (-6.66)* (-6.07)* (-5.65)* (-5.29)* (-4.98)* (-4.73)*

(-4.17)* (-3.99)* (-3.79)* (-3.67)* (-3.54)* (-3.44)* (-3.36)* (-3.28)* (-3.21)* (-3.15)*

The bold numeric represents the variance ratio while the values in parenthesis represents the Z (q) and Z (q)* respectively
and the value with a (*) representing the value is significant at 5%.
Journal of Independent Studies and Research (JISR)
Volume 4, Number 1, January 2006 21
also prefer a large sample such as in [30] study more than followed by [10] Variance ratio test and L-Jung Box Q
2000 daily observations has been taken for analysis for statistics. For the assumption of stationarity, the
KSE 100 index and only heteroscedastic consistent Z* (q) Augmented Dickey Fuller unit root test is also applied.
is free of autocorrelation and indicating randomness.
The analysis section gives us contradictory results from
Since [30] concluded that KSE 100 index is efficient in different statistical tools, since some of the tests are more
term of information, so rejection of random walk statistically powerful than others ([10] and [33] ). That is
hypothesis do not imply that market is not efficient. In why we analyzed and interpreted the results with the
variance ratio statistic when exceeds one leading to appropriate assumption and statistical power of the tool
rejection of null hypothesis, which is indication of positive used. The Augmented Dickey fuller test of unit root is
serial correlation that manifest itself in short term used only for the existence of unit root in time series data,
fluctuations. which suggest the randomness of the observed time series
for all the markets. The LB-Q statistic, Run Test and the
4.4 Ljung Box Q test variance ratio test results were consistent with each other,
but LB-Q and VR statistics contradict with RUN test in
The LB-Q5 statistic is often used as a test of white noise case of LSE. Our results of KSE also contradict with the
for a time series. LB-Q is asymptotically distributed as a general assumption of many researchers who suggested
chi square distribution with degree of freedom equal to the that KSE 100 index follows a Random walk hypothesis,
number of autocorrelations. for example [30] and [28].

LB-Q statistics obtained for Pakistani stock market returns The Pakistani stock markets follow a random walk
(shown in Table 4), suggested that there exist hypothesis if the assumption of variance ratio test and LB-
autocorrelation for KSE and ISE for few lags but not Q statistic is fulfilled, like large number of observation,
shows any autocorrelation at any lags for LSE. This result etc. One important finding of this result is that although
was consistent with not only to previous research work of this study contradicts [30], but it is consistent with one
[30], but also consistent with variance ratio analysis of the part of his study where he argued that in pre 9/11 era KSE
same stock markets. 100 index does not follow a random walk and Monday
gives a highest positive return which an indication of day
Table 4: LB-Q statistics of the week effect.
KSE LSE ISE
Lags Q-Stat Q-Stat Q-Stat REFERENCES
(P-value) (P-value) (P-value)
1 0.18 0.33 3.02 [1] Samuelson, Paul (1965). "Proof That Properly
(0.67) (0.56) (0.08) Anticipated Prices Fluctuate Randomly", Industrial
2 0.45 0.54 3.64 Management Review, 6, pp. 41-49.
(0.80) (0.76) (0.16) [2] Fama, E.F. (1970) “Efficient capital markets: A
review of theory and empirical work”, Journal of
3 3.65 1.88 13.88 Finance, 25, 383-417.
(0.30) (0.60) (0.003)
[3] McInish, H., and Pulisi, D.J. (1982) “The efficiency
4 3.79 2.21 16.02 of the international money market”, Journal of
(0.43) (0.69) (0.003) Business Finance and Accounting, 9(2), 167–177.
[4] Barnes, P. (1986), “Thin Trading and Stock Market
5 10.03 2.64 32.21
(0.07) (0.75) ( 0.00) Efficiency: The Case of the Kuala-Lumpur Stock
Exchange”, Journal of Business Finance &
Accounting, pp. 609 ,17.
[5] Cheung, Wong and Ho (1993), “The Pricing of Risky
4. CONCLUSION Assets in two Emerging Asian Markets, Korea and
Taiwan”, Applied Financial Economics, Vol. 3, No. 4,
In this paper we assessed the random behavior of the pp. 315, 24.
security prices for three major Pakistani stock markets i.e. [6] Dickinson, J.P. and Muragu, K. (1994) “Market
Karachi stock market, Lahore Stock market and Islamabad efficiency in developing countries: a case study of the
stock market. To put more validity and accuracy to our Nairobi Stock Exchange”, Journal of Business
analysis, we used a number of statistical models for Finance and Accounting, 21(1), 133–50.
random walk assessments. These statistical tools started [7] Ayadi, O. and C. Pyun (1994), “An Application of
from a non-parametric statistical test i.e. a Run Test, Variance Ratio Test to the Korean Securities Market”,
Journal of Banking and Finance, Vol. 18, pp. 643±
5 58.
The Ljung-Box Q statistic has an asymptotic chi square distribution
with p degrees of freedom equaling the first ρ lags. If the associated [8] Osborne, M. F. M. (1962) “Periodic structure in the
probabilities are less than 5%, the null hypothesis of no serial correlation Brownian motion of the stock market”, journal of
is rejected at 95% level of confidence. Operations Research, 10, 345–79.

Journal of Independent Studies and Research (JISR)

Volume 4, Number 1, January 2006 22
[9] Wong, K.A. and Kwong, K.S. (1984) “The behavior [22] Claessnes, S., S.Dasgupta, and J.Glen, “Stock price
of Hong Kong stock Prices”, Applied Economics, 16, behavior in emerging stock markets”, working paper,
905–17 World Bank.
[10] D’ Ambrosio, C. (1980) “Random walk and the
stock exchange of Singapore”, The Financial Review, [23] Tian, G., Y. Zhang and W. Huang, (1999), “A Note
1-12. on the Exact Distributions of Variance Ratio
[11] Cooper, J.C. (1983) “The Korean stock exchange: Statistics”, Peking University, Mimeo.
a qualitative and quantitative assessment”, The [24] Cecchetti, S. G. and Lam, P. S. (1994) “Variance-ratio
Investment Analyst, 70, 5-12. tests: Small-sample properties with an application to
[12] Fama, E., and K. French, (1988) “permanent and international output data”, Journal of Business and
temporary component of stock prices,” Journal Economic Statistics 12, 177–86.
Political Economy, 96, pp. 246-73. [25] Campbell, J.Y. and N.G. Mankiw, (1987), “Are
[13] Lo, A.W. and A.C. MacKinlay, (1989), “The Size and Output Variations Transitory?” Quarterly Journal of
Power of the Variance Ratio Test in Finite Samples: Economics, 102, 857 - 880.
A Monte Carlo Investigation”, Journal of [26] Cogley, T., (1990), “International Evidence on the
Econometrics, 40, 203 — 238. Size of the Random Walk in Output”, Journal of
[14] Keim, D. and Stambough, R. (1986) “Predicting Political Economy, 98, 501 — 518.
returns in stock and bond markets”, Journal Of [27] Poterba, J.M. and L.H. Summers, (1988), “Mean
Financial Economics, 17, 357-90. Reversion in Stock Returns: Evidence and
[15] Harvey, C. (1993), “Portfolio enhancement using Implications”, Journal of Financial Economics, 22, 27
emerging markets and conditioning information, - 60.
working paper World Bank. [28] Syed, Ali, Abass (2004) “Does KSE 100 index
[16] Huang (1995), “Do Asian stock market prices follow follows a random walk: an empirical study” Social
random walks? Evidence form the variance ratio test” sciences Research Network.
Journal of Applied Financial Economics, 5,251-256. [29] Yasir Kamal, and Zafar Mueen Nasir (2005), “weak
[18] Fawson, Glover, and Chang (1996) “The weak-form day effect in stock return: evidence from Karachi
efficiency of the Taiwan share market” journal of stock exchange” Social Science Research Network.
Applied Economics Letters, 1996, 3, 663–667 [30] Yasir Kamal and Kashif Ur Rehman (2005), “weak
[19] Urrutia (1995),”Test of random walk and market form of market efficiency: evidence from Karachi
efficiency for Latin American emerging equity stock exchange” Social Science Research Network.
market” journal of financial research”, vol XVIII, [31] Ljung, G.M. and G.E.P. Box (1978),”On a measure of
no.3, page 299-309 Lack of Fit in Time series models” Biometrika,
[20] Malkiel, B, (1995) “Returns from Investing in Equity Vol.66, pp. 297-303
Mutual Funds 1971 to 1991,” Journal of Finance, pp. [32] Long, D. M., Payne, J. D. and Feng, C. (1999)
549-572. “Information transmission in the Shangai equity
[21] Errunza, V. and E.Losq, (1985), “The behavior of market”, The Journal of Financial Research, 22, 29–
stock prices on LDC markets”, Journal of Banking 45.
and Finance, 561-75. [33] Ye, Cai and Mototsugu. Shintoni (2004) “On the
Long-Run Variance Ratio Test for a Unit Root”
Vanderbilt University, USA

Journal of Independent Studies and Research (JISR)

Volume 4, Number 1, January 2006 23