Stock Market Efficiency in

Nepal: A Variance Ratio Test

Bitaran Jang Maden

Abstract

This paper focuses on “Assessing the market efficiency in Nepalese market through
variance ratio test” on daily and monthly return. Aligned with the statement, this study has
analyzed historical data for NEPSE along with 16 haphazardly selected individual companies
listed in Nepalese stock market. The tests for variance ratio is then performed for assessing
predictability of the stock return series.

The test for variance ratio has shown that the Nepalese stock market is weak from
inefficient and has mean reverting characteristics. The returns are negatively correlated; thus, the
market is prone to bubble effects and is affected by microstructure components representing
trading effects like bid ask spreads and infrequent trading.

Key Words: Market efficiency, Random-walk, Variance ratio

1
 I. Introduction
The term “Efficiency” in the field of economics is a general term for value describing a
situation by measures designed to capture the amount of waste or “friction” or other undesirable
economic features present. An efficient market is characterized by rapid dissemination of
information and its reflection in prices. The EMH has been the central proposition of finance since
the early 1970s and is one of the most controversial and well-studied propositions in all the social
sciences.

The patterns and behavior of the capital markets has always been a subject of interest in the
academic as well as the applied research. The behavior of stock returns has been extensively
debated over the years (Hiremath, 2014). In course of time, Cowles (1933), Cowles and Jones
(1937), Kendall and Hill (1953) tried to analyze the pattern and behavior of stock return trying to
describe the nature of stock returns. But the most conclusive work on return predictability was the
theory of efficient stock market was introduced by Fama (1970).

According to Fama (1970, 1991), in an efficient stock market, share prices reflect all
information available to market participants and that, by implication, share prices cannot be
predicted, thus precluding any abnormal profit returns. From view-point of market participants,
the stock price behaviour is very important to determine future abnormal returns. For weak form
tests, information can include only past history of security prices. Tests for weak-form market
efficiency are, more generally, referred to as test of return predictability (Fama 1991).

The weak form efficiency test using “Fair Game” model or the “Random Walk” model test is
used to understand whether present or future price movements are dependent only on the
information set of the historical prices. This study is based on the test for the presence of random
walk in the Nepalese security market.

Fama (1970) was the pioneer of the efficient market hypothesis. It argued that the market price
“fully reflect” all available information. Fama classified the market efficiency into three levels
based on the reflection of information on the markets. The first stage of efficiency is the weak
form efficiency which states that the market price fully reflects the informations of past price and
returns. Kendall and Hill (1953) obtained that stocks are random from time series analysis using
serial correlation test. Fama(1970) proposed the theory of random walk hypothesis and Fama and
French (1988) used serial covariance and first order autocorrelation to show that stock returns are
random and there is no dependence of future prices on the past prices. Guidi and Gupta (2011)
also found similar results for the stock exchange of Singapore and Thailand, their test claiming
presence of weak form efficiency in these stock markets. However, the test did not support weak
form efficiency in Indonesia, Malaysia and Philippines.

Lo and MacKinlay (1988) was one of the first study to claim stock prices do not follow random
walk through volatility based specification test. Shaker (2013), Ali, Naseem and Sultana (2013)
also insisted that stock markets don’t follow random walk. In Nepalese context Baral and Shrestha
(2006), Pradhan and Upadhyay (2006), Dangol (2010), Dangol (2011), Pradhan and KC (2010)
and Maharjan (2012) studied presence of random walk in Nepalese market. Only Pradhan and KC
(2010) among those studies, predicted presence of random walk in Nepalese stock market.

2
II Research Methods

Lo and MacKinlay (1988) studied the variance ratio statistic as the test for the weak form
efficiency. It is derived from the assumption that observation interval regarding the variance of
increments has linear relations. The other assumption underlying the VR tests is that stock returns
are at least identically, if not normally, distributed and that the variance of the random walk
increments in a finite sample is linear in the sampling interval (Al‐Khazali, Ding, & Pyun, 2007).
One of properties of random walk hypothesis is that the variance of increments is directly
proportional to the length of time interval. It can be stated as: Var [Xt│X0]=σ2t. This explains that
variance of weekly increments should be 7 times the daily increments and that of monthly
increments (with 30 days in a month) should be 30 times the daily increments. This property is the
principle for variance ratio test. It is conducted by comparing the variances of the increments from
different lengths of time interval. In the random walk hypothesis with IID normal errors, the null
hypothesis states that variance from q -period increment should be q times larger than that of one-
period increment. Lo & MacKinlay (1988) suggest the Variance-Ratio test statistic, denoted as
𝑉𝑅̅̅̅̅̅̅̅̅(q), is the ratio between two variances from different time intervals:

𝜎2(𝑞)
𝑉𝑅(𝑞)=𝜎2(1)

Where, σ2(q) is 1/q the variance of the q-differences and σ2(1) is the variance of the first
differences. According to Lo and MacKinlay (1988), formulas for the calculation of σ2(q) and
σ2(1) are as follows: 𝜎2(𝑞)=11/(𝑞(𝑛𝑞 − 𝑞 + 1)(1 − 𝑞/𝑛𝑞) ) ∑2𝑛 𝑡=1(Yt − Yt − 1 − μ̂ )
2

Where,
1
𝜇̂=𝑛𝑞 (𝑌𝑛𝑞−𝑌0)

Y0 and Ynq are the first and last observations of the time series. The test is performed
under both homoskedastic and heteroskedastic specifications. Under homoskedasticity, the
asymptotic variance of the variance ratio is expressed as follows

Under heteroskedasticity, the asymptotic variance can be expressed as:

Where,

3
 The homoskedasticity and heteroskedasticity consistent Z-statistics are denoted by Z(q)
and Z*(q) and expressed as follows:

And

Under a single variance-ratio test, the null hypothesis is that VR(q) = 1 or that the chosen
index follows a random-walk. If the null hypothesis is rejected and VR(q) > 1, then the computed
Z(q) and Z*(q) are positive and returns are positively serially correlated. If the null hypothesis is
rejected and VR(q) < 1, then the computed Z(q) and Z*(q) are negative and returns are negatively
serially correlated, i.e., mean reverting.

III Empirical Test Results

Test for Hypothesis : Variance ratio in stock return series.

This study use Lo and Mackanlay variance ratio test to examine the predictability of the
stock return series. The method involves comparing variances of the difference of the returns over
various intervals. The variance of q period difference should be q times the variance of one period
difference, if the data follows random walk. For the analysis, the Hypotheses were formulated as
under:

H0: “Variance ratio in the stock return series is unity”

H1: “Variance ratio in the stock return series is not unity.”
The table 1 and 2 shows test result of NEPSE along with returns for 16 other individual
companies for monthly and daily data. Table 4.11 depicts that for monthly data ACE, AHPC,
LINC, PLIC, OHL and SUPREME supports the null hypothesis for lag 16 of homoscedastic
increments and lag 8 and 16 of heteroskedastic increments. ADBL supports the null hypothesis
only for lag 16 of homoscedastic increments. SCB, NABIL, CHCL, UNL and SLICL support the
nulls for lag 16 of both homoscedastic and heteroskedastic increments. BPCL supports the null for
the lags 8 and 16 for homoscedastic and all lags for heteroskedastic increments. BNL supports the
null for the homoscedastic increments but not the heteroskedastic increments. NTC support nulls
for lag 16 in homoscedastic increments and lag 4, 8 and 16 for heteroskedastic increments. The
RMDC supports null for all lags for both homoscedastic and heteroskedastic increments.

Table 2 shows that the variance ratio test result for daily return series does not support the
random walk in majority of the stock returns. Only, BNL supports the null for all lag lengths for

4
both homoscedastic and heteroskedastic increments and SCB and NABIL support the null
hypothesis for random walk for the 2nd and 4th lag for heteroskedastic increments. This shows
that Nepalese stock market in general is weak form inefficient, with only few series showing signs
of random behavior in price formation process. The Z-statistics is negative for all series for all q
cases. The variance ratio for all return series is below one and they decrease as the value of q
increases. This indicate negative serial correlation in the return with potential mean reverting price
formulation process. The result is like Fama and French (1988) which indicated negative serial
correlation implying mean reversion for long horizon security returns. The result is like Dangol,
(2011) who found mean reversion in Nepalese stock market and strongly rejected the Random
Walk Hypothesis. The test result contradicts the results of Lo and MacKanlay, (1988) and Urrutia,
(1995) who document mean aversion for US stock indices. To address this difference, Lo and
MacKinlay (1988) show that individual stock returns show more of a tendency towards mean
reversion. They also demonstrate that small stocks show comparatively greater degree of mean
aversion than large stocks. Hence, mean aversion for index returns can be justified by presence of
small stocks. Dangol (2012) describes this type of behaviour to be prominent in emerging financial
markets prone to bubble effects.
Lo and MacKinlay (1988) also hypothesize that the returns for individual securities to be
affected by three components: a systematic component representing general market risk factors
and is positively auto-correlated. The idiosyncratic component representing firm-specific risk and
is a white noise process, which implies random walk. The market microstructure component
representing trading effects such as bid ask spreads and infrequent trading, and is negatively
autocorrelated (as cited in Grieb and Reyens, 1999)
Hence based on the observations, it can be concluded that Neplese stock market is not weak form
efficiency since Random Walk Hypothesis is strongly rejected. Summers, (1986) pointed out that
when the returns are negatively correlated, it can also be said that the market is prone to bubble
effects (as cited in Dangol, 2012) and is affected by microstructure components representing
trading effects like bid ask spreads and infrequent trading (Lo and MacKinlay, 1988).

5
6
7
8
9
IV Conclusions

The variance ratio test shows that the Nepalese market is weak from inefficient and has
mean reverting. The returns are negatively correlated; thus, the market is prone to bubble effects
and is affected by microstructure components representing trading effects like bid ask spreads and
infrequent trading.
This proves Nepalese Stock market is not weak form efficient and there is an opportunity
of earning excess returns by beating the market. The rejection of the random walk and weak form
efficiency may not represent possibility of abnormal returns from profitable trading strategies. It
may be the effect of the thin trading in the security market because of the low volume of trade in
developing and undeveloped markets. Studies such as Al‐Khazali et al. (2007) and Moustafa
(2004) considered problems of infrequent trading as a subject to correction. Moustafa (2004)
insisted in using individual prices of the stock instead of the stock index returns for better analysis.
Contrary to Urrutia (1995), Al‐Khazali et al., (2007) and Moustafa (2004), Ma and Barnes (2001)
insisted that the correlated return pattern signifying return predictibility are results from a “thin”
market and followed the work of Butler and Malaikah (1992), claiming the “thin” market as a
characteristic of marekt inefficiency. This study uses the concept of Moustafa (2004) and used
individual prices of the company for testing the Efficient Market Hypothesis. However, the VR
tests shows that even after taking the individual prices of the companies, the returns are not free
from low trade volume.

10
Reference:

Al‐Khazali, O. M., Ding, D. K., & Pyun, C. S. (2007). A new variance ratio test of random walk in
emerging markets: A revisit. The Financial review, 42(2), 303-317.
Baral, K. J., & Shrestha, S. K. (2006). Daily stock price behavior of commercial banks in Nepal.
Journal of Nepalese Business Studies, 3(1), 100-110.
Butler, K. C., & Malaikah, S. J. (1992). Efficiency and inefficiency in thinly traded stock markets:
Kuwait and Saudi Arabia. Journal of Banking & Finance, 16(1), 197-210.
Cowles, A. (1933). Can stock market forecasters forecast? Econometrica, 1(3), 309-324.
Dangol, J. (2010). Testing random-walk behaviour in Nepalese stock market. PYC Nepal Journal of
Management, 3, 28-36
Dangol, J. (2011). Test of random-walk and stock market efficiency in Nepal and India. MIMT
Journal of IT and Management, 73-85.

Dangol, J. (2012). Efficient market hypothesis and adjustment of stock prices in Nepal. (Doctoral
Thesis, Tribhuvan University) available at central library of TU.

Fama, E. F. (1965). The behavior of stock-market prices. Journal of business, 38(1), 34-105.

Fama, E. F. (1970). Efficient capital markets: A review of theory and empirical work. The Journal of
Finance, 25(2), 383-417.

Fama, E. F. (1991). Efficient capital markets: II. The Journal of Finance, 46(5), 1575-1617.

Fama, E. F. (1998). Market efficiency, long-term returns, and behavioral finance. Journal of
Financial Economics, 49(3), 283-306.

Fama, E. F., & French, K. R. (1988). Permanent and temporary components of stock prices. The
Journal of Political Economy, 96(2), 246-273.
Grieb, T., & Reyes, M. G. (1999). Random walk tests for Latin American equity indexes and
individual firms. Journal of Financial Research, 22(4), 371-383.
Guidi, F., & Gupta, R. (2011). Are ASEAN stock market efficient? Evidence from univariate and
multivariate variance ratio tests. Discussion Papers in Finance No, 13. Retrieved from
https://www120.secure.griffith.edu.au/research/file/f5c91485-927f-a354-4c61-fcafd164531b/1/2011-
13-are-asean-stock-market-efficient-evidence-from-univariate-multivariate-variance-radio-tests.pdf
Hiremath, (2014) Random walk characteristics of stock returns. Indian Stock Market. Springer India,
2014. 19-39.
Kendall, M. G., & Hill, A. B. (1953). The analysis of economic time-series-part I: Prices. Journal of
the Royal Statistical Society. Series A (General), 116(1), 11-34.

11
Lo, A. W., & MacKinlay, A. C. (1988). Stock market prices do not follow random walks: Evidence
from a simple specification test. Review of financial studies, 1(1), 41-66.
Maharjan R. K. (2012). Development of Capital Market in Nepal. (Master’s thesis, London School of
Business and Finance) Retrieved from
http://www.academia.edu/3629775/Development_of_Capital_Market_in_Nepal
Moustafa, M. A. (2004). Testing the weak-form efficiency of the United Arab Emirates stock market.
International Journal of Business, 9(3). Retrieved from http://ssrn.com/abstract=551102
Pradhan, R. S., & K. C., Saraswari (2010). Efficient market hypothesis and behaviour of share prices:
the Nepalese evidence. SEBON Journal, 4, 104-117.

Pradhan, R. S., & Upadhyay, B. D. (2006). The efficient market hypothesis and the behaviour of
share prices in Nepal. Doi:http://dx.doi.org/10.2139/ssrn.1403745

12