You are on page 1of 3

( Nath, Golaka C.

and Dalvi, Manoj, Day of the Week Effect and Market Efficiency Evidence from Indian Equity Market Using High Frequency Data of National Stock
Exchange (December 2004). Available at SSRN: http://ssrn.com/abstract=1092765
or http://dx.doi.org/10.2139/ssrn.1092765 )

DAY OF THE WEEK EFFECT AND MARKET EFFICIENCY


EVIDENCE FROM INDIAN EQUITY MARKET USING HIGH
FREQUENCY DATA OF NATIONAL STOCK EXCHANGE

Nath, Golaka C. and Dalvi, Manoj (December, 2004):The research was conducted to
find out the day of the week effect anomaly in the Indian equity market for the period
from 1999 to 2003 using both high frequency and end of day data for the benchmark
Indian equity market index S&P CNX NIFTY in National Stock Exchange. In recent
times the testing for market anomalies has become an active field of research in finance.
Earlier studies have found the presence of the day of the week effect in many countries
like USA, Hong Kong, Malaysia. If an anomaly exists in the market, the investors can
take advantage of the same and adjust their buying and selling strategies accordingly to
increase their returns with timing the market. The present study aims to find the day of
the week effect on India equity market using high frequency data. This study is different
in two aspects: (1) it uses the high frequency data to study the day of the week effect; (2)
the study also does a comparative analysis using the closing values to understand if any
additional valuable information can be obtained from high frequency data.

High frequency data for the index S&P CNX NIFTY from January 1999 to December
2003 has been used. S&P CNX Nifty is a benchmark stock index based on the selected
stocks traded at National Stock Exchange (NSE). It was noticed that there were 1228
days of data running into millions of tick level index values. Normally a day has 335
minutes of trade. More than 410652 data points of 1 minute index values were used to
compute the logarithmic returns. The same has been taken out from millions of S&P
CNX NIFTY values from daily data provided by NSE. It is also to be taken into account
that 30 days of data is missing. The mean has to be found out to be statistically zero as

expected from the high frequency return series.


The research was now moved to see if there is any day of the week effect in both high
frequency as well as the close to close return series using dummy variables. Also day
wise returns were segregated to study the behaviour of returns of individual days to
understand if the returns are statistically significant or not . Robust regression method
was used as it assigns weight to each observation.
As the compulsory rolling system was introduced on the first day of January, 2002 , the
entire data was divded into two sets January 1999 to December 2001 having about 731
observations and the other period was from January 2002 to December 2003 having 497
observations . The high frequency as well as the end of day data, the results for the subsample period 1999-2001 shows that the coefficient of mean return is significant for
Monday and Wednesday. The sign is negative for both days. The coefficients for none of
the days except Friday are significant. The sign has been found to be positive indicating
the effect of Wednesday has vanished after introduction of rolling settlement. This clearly
indicates the Friday being the last day of the week; traders would like to close their
positions before the weekends. This clearly indicates that markets have become more
efficient after introduction of rolling settlement. The results have also pointed out the
Mondays have the highest standard deviation for high frequency data while for close to
close data Fridays have the highest standard deviation. For Wednesdays, the returns are
not only positive and significantly different from zero but also the risk (standard
deviation) is less compared to other days. The robustness check dividing the data into two
separate sets it was found that for the period from January 1999 to December 2001, the
coefficients for Wednesdays and Monday were found to be significant and negative for
both the datasets. The period from January 2001 to December 2003 showed the
coefficient of Friday being significant and positive.
Wednesday has been a significant historic day in NSE as it used to follow a trading cycle
from Wednesday to Tuesday. Wednesday being the first day of the trading cycle reflected
the action of market participants who used to roll over their positions from Tuesdays. But
in recent times after the introduction of rolling settlement, Fridays, being the last day of

the week, have become significant. The inefficiency in the market still exists and market
is yet to price the risk appropriately.

Bibliography
Gokla C Nath, M. D. (n.d.). Social Science Research Network. Retrieved from Nath,
Golaka C. and Dalvi, Manoj, Day of the Week Effect and Market Efficiency - Evidence
from Indian Equity Market Using High Frequency Data of National Stock Exchange
(December 2004). Available at SSRN: http://ssrn.com/abstract=1092765 or
http://dx.doi.org/10.2139/ssrn.1092765
Gokla Nath, M. D. (2004). Day of the Week Effect and Market Efficiency - Evidence
from Indian Equity Market Using High Frequency Data of National Stock Exchange.

By:
Anuj Kumar Jalan
1311545
4 BBA D

You might also like