Articles: Do Unexpected Events Cause Significant Market Volatility? - An Indian Perspective-By Debashis Kundu
Articles: Do Unexpected Events Cause Significant Market Volatility? - An Indian Perspective-By Debashis Kundu
A R T I C L E S
Do unexpected events cause significant market volatility? – An Indian Perspective—By Debashis Kundu 1
INTRODUCTION
Volatility is the child of uncertainty. Rational human beings favour either zero volatility or a quick return to certainty both in their
life and in their market investments. Unfortunately, volatility in stock prices is the norm rather than the exception. So investors
have learnt to live with it and also make occasional profit through speculation. But when volatility becomes extreme and persistent
over a period of time even seasoned investors prefer to move to the sidelines.
Extreme volatility is premised to be caused by news of unexpected events, among other things. But neither all events cause the
same level of volatility nor are they as persistent as others. The search for the type of events causing volatility and their statistical
measurement form the basis of this paper.
The paper starts with a brief introduction on volatility. The next section takes a look at the findings of other research papers on
volatility followed by the significance, objectives and sources of data for the present study. The methodology is outlined next fol-
lowed by the findings. The last section concludes the article.
LITERATURE REVIEW
This paper current paper is both an event study as well as a study of volatility in the Indian context. The past studies thus perused
comprise of both event-based as well as volatility tests conducted in and outside India. In the international context, Geske and Roll
(1983) conclude that stock returns forecast real activity (industry performance) and anticipated macroeconomic changes, causing
interest rates to rise. French and Roll (1986) report that private information and ‘mispricing’ were significant factors causing daily
volatility. Fama (1990) gives evidence of a positive relation between stock prices and announcements of real activities. Binder and
Merges (2001) identify four key determinants of market volatility that help to examine the past behaviour of stock market volatility
and in forecasting future volatility. Dolde, Saad, and Tirtiroglu (2002) reported that 63.5% of extreme volatility increases and
53.4% of decreases are associated with extraordinary reported news items. Brooks, Patel and Su (2003) found out that the response
time of stock prices to unanticipated news lasted about two hours after which they tended to reverse. Al-Khazali (2003) found out
that in the long term, there is a positive relationship among stock returns, consumer price index and industrial production. In the
Indian context, Kaur (2004) found that 1992 was the most volatile year and April was usually the most volatile month in the Indian
stock market. Sabnavis (2005) reports that economic events and natural disasters do not have much impact on Sensex movement
unlike political events and terrorist attacks. Padhi (2005) reports the same trend of volatility in the case of aggregate indices and
five selected sectors such as electrical, machinery, mining, non-metallic and power plant sector. Sarkar, Chakrabarti, and Sen
(2008) reports that shocks in Dow Jones, Jakarta stock index and BVSP has profound effect on the BSE Sensex. Capital goods and
consumer durables sectors were the most prominent contributors to the volatility of the Sensex.
SIGNIFICANCE OF STUDY
The paper looks at the effect of six major types of events on variance of returns of the proxy market portfolio, i.e., the NSE Nifty,
over a fourteen-year period from 1995 to 2009. The study is significant in the way it tries to statistically measure actual volatility
caused by six major types of events in the Indian market. The step-by-step analysis of volatility in Nifty using simple F-tests before
and after the events provides a clear picture to investors, speculators, researchers and the regulator alike about the possible ex-
tent of volatility from similar events in the future.
⇒ To empirically study the occurrence of statistically significant volatility from major events
⇒ To study the presence of statistically significant volatility up to nine trading days after the event day, if any
The chosen index Nifty comprises of fifty stocks that adequately represent the broad spectrum of the Indian industry. Nifty belongs
of National Stock Exchange (NSE), the largest Indian exchange in terms of turnover.
The events (126 of them) have been chosen on the basis of their effect on the market as a whole and not on any particular industry
1 The author is with Department of Commerce, Sivanath Sastri College, Kolkata
1
N S E N E W S L E T T E R 2
August 2009
or company. The events (Tables 6 to 12) have been regrouped into six broad types – budgets, macro-economic policy events, securi-
ties market matters, non-economic events, international matters (both economic and otherwise), and major disasters (man-made
and natural). In case of some events the market got the news after the event day. Hence calculation was performed on the closing
figure of the next trading day. They have been mostly sourced from Manorama, Penguin and Statesman’s yearbooks. The Nifty values
come from the official website of NSE.
METHODOLOGY
Volatility in returns has been measured through variance of returns over blocks of three (3) trading days. Six such blocks have been
considered, three before the event (X1, X2, X3) and three after (Y1, Y2, Y3), for each event. The blocks have been used to under-
stand changes in volatility as one move towards the event day, and later away from it. Only three such blocks were considered on
either side because consecutive events happening in quick succession tend to affect one another.
X1 X2 X3 Y1 Y2 Y3
Event day
The study has used the statistical technique of hypothesis testing with the help of F-test. The test statistic F is calculated as follows:
-
F = s(Xi)2 / s(Yi)2
[where s(Xi)2 = å(RXi – RXi)2 / (n1-1) and s(Yi)2 = å( RYi – RYi)2 / (n2-1)]
Here, Xi and Yi are the two sample time periods, s(Xi)2 and s(Yi)2 being the sample return variances, and n1 and n2 being their respec-
tive numbers of observations.
In all the cases, variance of return of the succeeding period (say, X3) has been compared with the preceding period (say, X2). The
null hypotheses in all the tests assume no change in variance, i.e., the variances are equal. The alternative hypotheses H1 want to
prove that variance during the succeeding period is more than the preceding period. Hence the comparisons are: s2X1 < s2X2, s2X2 < s2X3,
s2X3 < s2Y1, s2Y1 < s2Y2 and s2Y2 < s2Y3.
INTERPRETATION OF FINDINGS
The next few paragraphs analyse the F-test results based on variances of returns at the standard 5% level of significance. As already
mentioned, five F-tests have been conducted for each event. Some of the events have occurred previous to the day when the market
got the news. This happens if the event occurred in the after-market hours or on a trading holiday.
Table 1 shows the F-test values comparing the variances of returns around sixteen budgets. The results clearly show that in only five
out of sixteen instances, the budgets were found to cause statistically significant volatility. In fact, two significant instances oc-
curred even before the budget day. This may be due to the pre-budget press leaks and expectations that are already built into the
stocks. Instant volatility (statistically significant volatility caused immediately after the event; Y1 on X3) is present only in two cases
2
N S E N E W S L E T T E R 3
August 2009
Table 2 shows the F-test values that compare the variances of return around thirty-six macro-economic events. The results show
a curious pattern of the maximum number of statistically significant cases occurring before the actual announcement day. A
closer look reveals that most of these events were the announcement of rate changes by RBI. This indicates that the market was
expecting such an announcement despite RBI's attempt to keep these secret. Most other major policy decisions seem to be al-
ready either discounted or deemed to be not so significant by the market. Instant volatility is present only in three cases.
Table 3 exhibits the results from F-tests comparing the variances of returns around seventeen stock market-related events. The
events that are supposed to directly affect the stock market cause hardly any statistically significant volatility. Only the matters
affecting foreign investment cause some volatility and that too well after the event day. Instant volatility is present only in one
case.
Table 4 shows the F-test values of twenty-four non-economic events. These are events (mostly political ones) that are widely
speculated upon by the public. Hence these events hardly cause any significant volatility. Instant volatility is present only in two
cases.
Table 5 shows the F-test values comparing the variances of returns around twenty-one international (both economic and non-
economic) events. International events are said to influence Indian markets much more than ever. However, the results depict a
completely different picture. They hardly cause any volatility in Indian markets. Instant volatility is present only in one case.
Table 6 shows the F-test values comparing the variances of returns around twelve man-made and natural disasters. Terrorist at-
tacks and natural disasters also hardly cause any statistically significant volatility. Instant volatility is present only in one case.
CONCLUSION
The paper helps to empirically verify the widely held notion that major events (economic and non-economic) cause substantially
changes in returns. A total of 126 events, grouped into six types, have been studied over a fourteen-year period.
The 630 (126 x 5) F-tests report only thirty-eight (38) statistically significant cases of volatility. In fact, these cases are evenly
distributed both before and after the events, indicating that the selected events cannot be held responsible for causing such vola-
tility.
In the light of the triple objectives outlined at the outset, the study conclusively proves the absence of significant volatility
caused by major events. Even in instances where such volatility occurs, they do not linger beyond the third trading day post
event. Out of the six major types of events, only budgets and macro-economic announcements cause a few cases of volatility
more than others.
All-in-all, the investors and the Regulator can rest assured that major domestic and international events do not cause too much
uncertainty in stock prices, at least in the Indian context.
REFERENCES
Al-Khazali, Osamah M. (2003): “Stock prices, Inflation and Output: Evidence from the Emerging Markets”, Journal of Emerging
Finance, pp. 287-314.
Binder, John J. and Merges, Matthias J. (2001): “Stock Market Volatility and Economic Factors”, Review of Quantitative Finance
and Accounting, Volume 17(1).
Brooks, R. M., and Patel, Ajay and Su, Tie (2003): “How the Equity Market Responds to Unanticipated Events”, Journal of Busi-
ness, pp. 109-133.
Dolde, W., Saad, R. and Tirtiroglu, Dogan (2002); “Evidence that Extreme Volatility in Stock Prices is Associated with Extraordi-
nary Reported News Items”, http://ssrn.com/abstract=334602
Fama, Eugene F. (1990): “Stock returns, expected returns and real activity”, Journal of Finance, pp. 1089-1108.
French, Kenneth R. and Roll, Richard (1986): “Stock Return Variances: The Arrival of Information and the Reaction of Traders”,
Journal of Financial Economics, pp. 5-26.
Geske, Robert and Roll, Richard (1983): “The fiscal and monetary linkage between stock returns and inflation”, Journal of Fi-
nance, pp. 1-33.
3
N S E N E W S L E T T E R 4
August 2009
Gujarati, Damodar (1999): "Essentials of Econometrics", Irwin Mcgraw Hill, p. 274, 281.
Kaur, Harvinder (2004): “Stock Market Volatility in India”, The Indian Journal of Commerce, Vol. 57(4), pp. 55-70.
Kothari, C. R. (2004): "Research Methodology: Methods and Techniques", New Age International (P) Ltd., New Delhi, p. 95.
Padhi, Puja (2005): “Stock Market Volatility in India: A Case of Select Scripts”, Indian Institute of Capital Markets 9th Capital
Markets Conference Paper.
Sabnavis, Madan (2005): “Irrational, and remarkably resilient”, Business Standard, Kolkata, p. 8.
Sarkar, A., Chakrabarti, G. and Sen, C. (2008): “Indian Stock Market Volatility in Recent Years: Transmission from Global and
Regional Contagion and Traditional Domestic Sectors”, http://ssrn.com/abstract=1089763.
Statesman’s Yearbook (various issues), The Macmillan Press Ltd., London and Basingstoke.
d.f. = 2 and 2
T.V. (5%) = Table value at 5% level of significance, d.f. = degrees of freedom, figures with * = statistically significant results]
4
N S E N E W S L E T T E R 5
August 2009
d.f. = 2 and 2
5
N S E N E W S L E T T E R 6
August 2009
Table 3: F-test values from Securities market events
7
N S E N E W S L E T T E R 8
August 2009
Table 7: Budgets
Sl. Date Event
No.
11. 08/07/04 UPA government’s first budget presented by P. Chidambaram. Securities Transaction Tax
introduced.
12. 28/02/05 Second budget presented by P. Chidambaram.
13. 28/02/06 Chidambaram presents the budget for 2006-07.
14. 28/02/07 Chidambaram presents the budget for 2007-08.
8
N S E N E W S L E T T E R 9
August 2009
3. 22/10/01 RBI cuts bank rate to 6.5%, lowest since May 1973.
7. 29/10/02 RBI cuts Bank rate from 6.5 to 6.25 and CRR from 5 to 4.75.
9. 15/06/04 Petrol and diesel prices hiked; kerosene price left untouched.
15. 31/03/06 Foreign exchange reserves touch new high. Current account deficit narrows.
21. 17/02/07 RBI directs banks to tighten procedures on loans for properties.
22. 30/03/07 RBI hikes repo rate to 7.75% and CRR by 50 bps.
9
N S E N E W S L E T T E R 10
August 2009
11. 07/10/99 BJP-led NDA Govt. leads towards absolute majority in elections.
12. 11/03/01 BJP President Bangaru Laxman caught on the camera of Tehelka.com taking bribes.
13. 04/12/01 Madras H.C. acquits Jayalalitha, a key supporter of the NDA Government, in TANSI land deal case.
15. 22/05/02 Vajpayee asks troops to prepare for decisive battle against Pakistan.
17. 18/05/04 Manmohan Singh, former Finance Minister, tipped to become next Prime Minister.
18. 27/05/04 UPA government releases its Common Minimum Programme (CMP).
19. 07/03/05 President’s Rule imposed in Bihar ending 15-year RJD rule. RJD is an important ally of the Central
Government.
20. 23/05/05 CP(I)M assures UPA govt of continuing support.
21. 15/11/05 Left warns Govt of serious consequence if India votes against Iran in UNSC.
23. 23/12/07 BJP wins in Gujarat. Modi becomes CM for 3rd time.
24. 18/06/08 UPA and Left seem to be parting ways on Indo-US nuclear deal.
25. 22/07/08 UPA Govt wins trust vote in Lok Sabha by 19 votes.
10
N S E N E W S L E T T E R 11
August 2009
2. 12/05/98 US impose economic sanctions after Pokhran nuclear blasts takes place.
5. 14/04/00 Crash in the NASDAQ composite index. Beginning of the dotcom bust.
7. 11/09/01 On Sept. 11 terrorists attack WTC, New York. Stock markets crash worldwide.
14. 21/07/05 China appreciates yuan by 2% and removes its dollar peg.