“Modeling Indian Bank Nifty volatility using univariate GARCH models”

Nikhil M. N.

Suman Chakraborty

Lithin B. M.
AUTHORS

Sanket Ledwani

Satyakam

Nikhil M. N., Suman Chakraborty, Lithin B. M., Sanket Ledwani and Satyakam
ARTICLE INFO (2023). Modeling Indian Bank Nifty volatility using univariate GARCH models.
Banks and Bank Systems, 18(1), 127-138. doi:10.21511/bbs.18(1).2023.11

DOI http://dx.doi.org/10.21511/bbs.18(1).2023.11

Friday, 17 March 2023

RELEASED ON

RECEIVED ON Tuesday, 11 October 2022

ACCEPTED ON Monday, 06 February 2023

LICENSE This work is licensed under a Creative Commons Attribution 4.0 International
License

JOURNAL "Banks and Bank Systems"

ISSN PRINT 1816-7403

ISSN ONLINE 1991-7074

PUBLISHER LLC “Consulting Publishing Company “Business Perspectives”

FOUNDER LLC “Consulting Publishing Company “Business Perspectives”

NUMBER OF REFERENCES NUMBER OF FIGURES NUMBER OF TABLES

67 3 5

© The author(s) 2023. This publication is an open access article.

businessperspectives.org
 Banks and Bank Systems, Volume 18, Issue 1, 2023

Nikhil M. N. (India), Suman Chakraborty (India), Lithin B. M. (India),

Sanket Ledwani (India), Satyakam (India)

Modeling Indian Bank Nifty

BUSINESS PERSPECTIVES
LLC “СPС “Business Perspectives” volatility using univariate
Hryhorii Skovoroda lane, 10,
Sumy, 40022, Ukraine
www.businessperspectives.org
GARCH models
Abstract
The crumble of financial markets due to the recent crises has wobbled precarious-
ness in the stock market and intensified the returns vulnerability of banking indices.
Against this backdrop, this study intends to model the volatility of the Indian Bank
Nifty returns using a battery of GARCH specifications. The finding of the present re-
search contributes to the literature in three ways. First, volatility during the sample
period, which corresponds to a time of stress (a bear market), is more persistent, with
an estimated coefficient of 0.995695. Moreover, when volatility rises, it persists for a
long time before returning to the mean in an average of 16 days. Second, for a positive
γ, the results insinuate the possibility of an “anti-leverage effect” with a coefficient of
0.139638. Thus, the volatility of the Bank Nifty returns tends to rise in response to posi-
tive shocks relative to negative shocks of equal magnitude in India. Finally, the findings
Received on: 11th of October, 2022 demonstrate that EGARCH with Student’s t-distribution offers lower forecast errors in
Accepted on: 6th of February, 2023 modeling conditional volatility.
Published on: 17th of March, 2023

© Nikhil M. N., Suman Chakraborty, Keywords asymmetry, anti-leverage, leverage, return volatility,
Lithin B. M., Sanket Ledwani, bank nifty, GARCH, index returns, Indian stock
Satyakam, 2023.
JEL Classification C22, C52, G10, G17
Nikhil M. N., Doctoral Scholar,
Department of Commerce, Manipal
Academy of Higher Education, India. INTRODUCTION
Suman Chakraborty, Dr., Associate
Professor, Department of Commerce,
Manipal Academy of Higher Education, Following the consistent underperformance of the benchmark Nifty 50
India. (Corresponding author) over the last three years, Bank Nifty has taken the brunt of the wrath.
Lithin B. M., Doctoral Scholar, Notably, the benchmark index recorded an increase of 38%, while the
Department of Commerce, Manipal
Academy of Higher Education, India. Banking index plunged by 13% during the same period. However, in-
Sanket Ledwani, Doctoral Scholar, vestors frequently explore a strategy to expand exposure to high beta
Department of Commerce, Manipal indices with an appetite to outperform the benchmark. In particu-
Academy of Higher Education, India.
lar, an investor with a high-risk potential prefers to buy a high-beta
Satyakam, Assistant Professor (Senior
Scale), Department of Electrical and stock (Christoffersen & Simutin, 2017). In this context, several studies
Electronics Engineering, Manipal in mainstream finance showed that Bank Nifty is a high beta index,
Academy of Higher Education, India.
reiterating its ability to generate large returns in relation to the bench-
mark when the market is bullish. On the contrary, academic research
has been inconclusive in modeling the volatility of Bank Nifty during
periods under stress (bear market). In fact, the 1987 crisis, the ma-
jor global financial crisis of Asia in 1997, and the catastrophe in 2008,
which walloped the economy worldwide, have demonstrated the cut-
ting-edge volatility of the Bank Nifty index with substantial variations
in the returns (Bhattacharyay, 2013).
This is an Open Access article,
distributed under the terms of the Recently, COVID-19 had the most devastating effects on the finan-
Creative Commons Attribution 4.0
International license, which permits cial markets, both in terms of scale and intensity (Rout & Mallick,
unrestricted re-use, distribution, and 2022). As the crises intensified, the banking sector returns in India
reproduction in any medium, provided
the original work is properly cited. metamorphosed into acute vulnerability. Experts are of the opinion
Conflict of interest statement: that “Vulnerabilities in credit markets, emerging countries, and banks
Author(s) reported no conflict of interest could even cause a new financial crisis.” Nevertheless, there is a dearth

http://dx.doi.org/10.21511/bbs.18(1).2023.11 127
Banks and Bank Systems, Volume 18, Issue 1, 2023

of academic research on banking sector volatility explicitly in emerging markets like India. Thus, the
current study, motivated by this problem, employs a battery of GARCH specifications to empirically as-
sess the asymmetric behavior of the Indian Bank Nifty returns.

1. LITERATURE REVIEW & Wu, 2009). Therefore, the susceptibility of

AND HYPOTHESES banking stock returns exerts a considerable effect
on the common stocks of financial institutions
DEVELOPMENT (Kasman et al., 2011). However, the sensitivity
changes over time, and the returns are countered
As the world becomes more fast-paced and dy- by an asymmetric response (Âhou & Gueyie,
namic, financial institutions have become in- 2001). As a result, analyzing and capturing bank
creasingly important; they help balance de- risk can be difficult. The rising complexity of the
mand and money supply. The performance of business models of the banks can make measur-
firms, specifically the growth of banking sectors ing and observing the underlying risks of banks
on account of diversified business activities, is more difficult (Begley et al., 2017).
of paramount importance (Anagnostopoulos,
2018). Investments in bank savings schemes Furthermore, because banks are critical to eco-
are often viewed as safe, assuring fixed returns. nomic development, a banking meltdown can
Additionally, due to divergence in various ser- considerably impact other sectors of the economy
vices, the banking sector’s stocks have become (Hoggarth et al., 2002). Consequently, strengthen-
a popular trading and investment option (De ing the banking system and preventing crises are
Jonghe, 2010). Since 2004, the banking sector of utmost importance for policymakers and regu-
stocks have shown tremendous growth in market lators (Gupta, 2010). Thus, the “New Basel Capital
capitalization and trade volume. Considering the Accord,” developed by the “Basel Committee
importance of banks for economic development, on Banking Supervision” in 2003, includes mar-
several empirical studies have been conducted by ket discipline as one of its three pillars in recog-
researchers and analysts hitherto. nition of the significance of market discipline in
safeguarding banks and financial institutions
Recent research indicates that banking crises are (Crockett, 2002). As a result of market discipline,
likely to have a spillover effect on other sectors stock returns, and prices are used to measure the
(Candelon et al., 2016), which necessitates prop- riskiness of banks (Flannery, 2001). In particular,
er evaluation and constructive measurement of unlike accounting-based returns, the information
bank risks to ensure the stability of the banking inherent in asset returns and prices is oftentimes
industry (Begley et al., 2017). The major upheaval a forward-looking metric, and such informa-
in the stock markets on account of the economic tion might serve as a useful benchmark for mar-
and financial crisis of 2008 induced government ket discipline (Acharya et al., 2012; Hasan et al.,
to adopt global banking rules to monitor and 2015; Stiroh, 2006). Besides, extreme movements
measure bank risk (Kumar & Prakash, 2019). For in stock prices and returns may indicate concern
decades, one of the most prevalent notions in the about the future economic situation. As a result,
banking literature has been that stock returns can volatility in the banking stock returns may signal
capture their potential long-run performance risk the banking sector’s performance stability of a
(Baele et al., 2007). country (Moshirian & Wu, 2009).

A significant contribution to banking litera- The asset returns dynamics have piqued the inter-
ture dates back to Diamond (1984), Brealey et al. est of several researchers. The empirical studies
(1977), and Diamond and Dybvig (1983) on stock on volatility modeling of stock returns have gar-
returns of banks and financial intermediaries. A nered the attention of academicians (Thiripalraju
fundamental assumption in this literature is that & Acharya, 2010). Numerous scholars have pro-
market returns transmit trustworthy informa- posed various methods for assessing and quan-
tion about profit prospects and risk (Moshirian tifying bank risks (Anginer et al., 2014; Baele et

128 http://dx.doi.org/10.21511/bbs.18(1).2023.11
 Banks and Bank Systems, Volume 18, Issue 1, 2023

al., 2007; Bennett et al., 2015; Demirer et al., 2018; ing sector stocks were highly volatile during the
Laeven et al., 2016; Stiroh, 2006; Fratzscher & Mortgage Crisis relative to COVID-19. Mahajan et
Rieth, 2015). On their part, Anginer et al. (2014) al. (2022) further confirmed the existence of lever-
demonstrate that systemic risk measured using age in the Indian stock market.
stock returns and bank competition are inverse-
ly related. Nonetheless, the study fails to consid- Despite the difficulty of quantifying the accu-
er banks’ time-varying stock return volatility. rate degree of the influence of market-based per-
However, stock returns often show time-varying formance metrics on bank stability in a rapidly
fluctuations (Fratzscher & Rieth, 2015). As a result, evolving world, it is evident that these metrics
the empirical findings show that the volatility of imply abrupt contractions of both financial per-
stock returns of leading global banks is dynami- formance and bank risk (Elnahass et al., 2021). To
cally connected across time (Demirer et al., 2018). gauge the financial stability of the banking sector,
modelling their return volatility could be a useful
Karmakar (2005) employs the GARCH (1,1) mod- benchmark (Suhadak et al., 2019). It is worth men-
el and estimates the parameters for BSE Sensex. tioning, however, that such quantifications aid in
Further, it claims that the GARCH model with forecasting future volatility and offer an extra tool
one lag order is the suitable model in the Indian for positioning adjustments.
stock market for forecasting and modelling vol-
atility. However, the study’s findings contend the Thus, the current study aims to disentangle the
relevance of using asymmetric models to examine asymmetric volatility puzzle by modeling Bank
the asymmetry, persistence of volatility, and its Nifty returns volatility in India. Based on theoreti-
clustering. Following this, Padhi (2006) applied cal and empirical frameworks developed in the ex-
EGARCH and GJR-GARCH models to investigate tant literature, the following research hypotheses
the presence of volatility and leverage in the Indian are employed in this study:
stock market. Further, the study suggests applying
Student’s t and GED distribution for asymmetric H1: Volatility shocks are highly persistent in
GARCH models, since are proficient at capturing Bank Nifty returns.
fat tail and highly peaked properties. Subsequently,
Karmakar (2007), to model asymmetric volatility, H2: Bank Nifty returns are equally sensitive to
employs the EGARCH model with GED and cor- good news and bad news of the same size.
roborates the presence of asymmetry in the form
of leverage effect in India. Later, Tripathy and Gil-
Alana (2015) examine the volatility in India from 2. METHODOLOGY
August 1992 to September 2012, decomposing
the study period into pre-crisis (1992–2008) and 2.1. Data and study period
post-crisis (2008–2012). The findings reveal that
leverage effects and volatility persistence are more The closing prices of the Nifty Bank Index were
significant in the post-crisis period than in the collected from the National Stock Exchange from
pre-crisis period. Another study was carried out by June 10, 2005 to May 31, 2022, and used for mod-
Bhatia and Gupta (2020) on the volatility between eling volatility corresponding to 4,206 observa-
Indian Nifty Bank Index, Private Sector Bank tions. The 12 most liquid equities with large mar-
Index, and Public Sector Undertaking Banking ket capitalizations from the banking sector that
Index (PSUBI). In addition, they compare the vol- are traded on the NSE make up the Bank Nifty
atility between two events, namely, the great re- index. Further, the index gives market interme-
cession of 2008 and COVID-19. The results show diaries and investors a benchmark that represents
that asymmetric impact during the great recession the market performance of the banking sector in
2008 is lower for PSUBI compared to the other India.1 The study uses long-period data for a bet-
two sectors. However, during the COVID period, ter model fit, as the amount of time significantly
the results are insignificant, implying that bank- affects the model fit.
1 Bank Nifty Index of NSE is a benchmark for traders and market intermediaries as it reflects the performance of the Indian banking sector’s
secondary market.

http://dx.doi.org/10.21511/bbs.18(1).2023.11 129
Banks and Bank Systems, Volume 18, Issue 1, 2023

The returns data (Rt) are accrued as compounded 2.2.1. EGARCH model
returns:
 p  The linear GARCH models do not differenti-
Rt = ln  t  , (1) ate the influence of optimistic and pessimistic
 pt −1  news on the volatility of any time series. Nelson
(1991) developed Exponential GARCH, which
where pt = current days’ stock price; pt-1 = stock accounts for the impact of positive and negative
price of the previous day. shocks on time series volatility (the leverage ef-
fect). The EGARCH (1,1) model can be written
The financial time series may exhibit intensified as follows:
market volatility when the returns are negative,
known as the market asymmetry or the Leverage Variance equation:
effect (Black & Cox, 1976). The changes in the
market segment, transaction costs, and frictions ln(σ t2 ) = α 0 + β1 ln(σ t2−1 ) +
are the few causes of asymmetry in the financial ε 2 ε t −1 (3)
time series. Therefore, such behavior necessi- +α1  t −1 − −ϒ ,
tates employing asymmetric GARCH models to  σ t −1 π  σ t −1
capture non-linear and asymmetric volatility in
Indian stock returns. where ϒ denotes the leverage effects. ϒ > 0 implies
more volatility when the news is good. However,
2.2. GARCH models ϒ < 0 implies that the bad news is more disruptive
to returns when negative. The model is regarded
The increase in the asset price is always accom- to be symmetric when ϒ = 0. Further, ln(σ t2) may
panied by a decline to a greater extent and in the be negative, and the parameters in the EGARCH
financial time series such price fluctuations are model have no sign restrictions.
regarded as stylized facts (Lin, 2018). Early evi-
dence shows that the researchers predominantly 2.2.2. GJR-GARCH model
used ARCH (Autoregressive conditional hetero-
scedasticity) models to forecast stock returns’ dy- This model proposed by Glosten et al. (1993) ac-
namics. Later, Bollerslev (1986) suggested a superi- counts for the leverage effect. The model can be
or, Generalized ARCH model, well-known as the expressed as follows:
GARCH model. The symmetric GARCH model is
defined as follows: µt=
2
α 0 + α1µt2−1 + βσ t2−1 + ϒµt2−1 I t −1. (4)
p q
(2) A positive shock is indicated by µt-1 < 0, a negative
β 0 + ∑α i µt2−i + ∑β j µt2− j .
σ t2 = shock by µt-1 > 0, and a dummy variable by It-1. The
=i 1 = j 1
coefficients ϒ ≠ 0 and ϒ > 0 represent the asym-
In equation (2), ARCH and GARCH parameters metric shocks and leverage effect, respectively. α0
are represented by µt-i2 and µt-j2, respectively. The α > 0, α1 > 0, β ≥ 0 and α1+ ϒ ≥ 0 are the conditions
term denotes the ARCH coefficient, the β term de- for non-negativity (Brooks, 2008, p. 405).
notes the GARCH coefficient, and p and q values
indicate the model’s lag order. The GARCH model
with one lag order is considered as the best mod- 3. RESULTS
el to measure the volatility clustering in financial
time series (Brooks & Burke, 2003). However, the To eliminate biased regression in the model fit,
symmetric GARCH models do not support the confirming the stationary for the time series da-
asymmetry in the asset returns since such mod- ta is necessary. Furthermore, as GARCH mod-
els assume conditional variance to be constant. els are developed to describe the variations in
This necessitates the employment of asymmetric heteroscedastic data, it is essential to identify
GARCH models. whether heteroscedasticity exists in the data. As
a result, tests to determine whether the data are

130 http://dx.doi.org/10.21511/bbs.18(1).2023.11
 Banks and Bank Systems, Volume 18, Issue 1, 2023

stationary (ADF test and PP test) and whether Table 2. ARCH LM test results
ARCH effects are present (LM test) are used as Probability value
preconditions for GARCH modelling. t-statistics
[Chi-square (1)]
68.14172 0.0000
3.1. Test for stationarity
Note: H0: There are no ARCH effects in the returns data
The results of both ADF and PP tests are reported of the Bank Nifty.
in Table 1. Since the p-values are less than 0.05 for
the tests, the null hypothesis is rejected, i.e., data 3.3. Descriptive statistics
have a unit root, outlining that Bank Nifty returns
data are not non-stationary. Before processing the data, it is essential to
have an overview of the summary statistics.
Table 1. Unit root test results
The results of the descriptive statistics are re-
Critical value Probability ported in Table 3. The mean value for the 4206
Unit root tests t-statistic
(5%) values observations of the returns series is 0.000548,
Augmented Dickey-
–58.58497 –2.862031 0.0001 and the observed sample standard deviation is
Fuller test
Phillips-Perron test –58.45962 –2.862031 0.0001
0.018624. The mean value is small compared to
the standard deviation observed in the study,
Note: H0: Nifty Bank Returns have unit root one (non-stationary). indicating high volatility during the sample
period. The skewness coefficient shows a neg-
3.2. Test for ARCH effect ative value (–0.195874), outlining that the Bank
Nifty returns series is negatively skewed. The
The test for conditional heteroscedasticity is a kurtosis of the series exceeds the standard nor-
pre-requisite to determining volatility in the Bank mal distribution value (+3), demonstrating the
Nifty returns data. Accordingly, the study applies fat tail. These findings corroborate that the
the LM test (Engle, 1982). Table 2 displays a sum- distribution differs remarkably from normali-
mary of the ARCH test results. The findings of ty (refer to Figure 2). Accordingly, the Jarque-
the ARCH LM test offer strong evidence against Bera test for normality shows a t-statistic val-
the null hypothesis. This corroborates the pres- ue (9,539.506) with a significant p-value at a 1%
ence of the ARCH effect in the Bank Nifty returns significance level. As a result, the null hypothe-
data. Apart from this, the daily returns’ graphi- sis is rejected, confirming that the distribution
cal representation affirms time-varying volatility does not exhibit normality. Further, as a tool
and volatility clustering during the sample period to estimate the distributional properties, the
(Figure 1). Thus, the conditional variance with the study uses a Quantile-Quantile (QQ) graph (re-
homoscedasticity assumption is no longer valid fer to Figure 3). The graphical results confirm
for Bank Nifty returns, and the current research the findings of the Jarque-Bera test, i.e., Bank
can test conditional heteroscedasticity by employ- Nifty returns data do not adhere to the normal
ing the GARCH process. distribution.
Source: Computed Using EViews.

Price Logreturns
45,000 .20
40,000 .15
35,000 .10
30,000
.05
25,000
.00
20,000
-.05
15,000
10,000 -.10

5,000 -.15
0 -.20
06 08 10 12 14 16 18 20 22 06 08 10 12 14 16 18 20 22

Figure 1. The trend (Closing prices) and log-returns of the Bank NIFTY index during the sample period

http://dx.doi.org/10.21511/bbs.18(1).2023.11 131
Banks and Bank Systems, Volume 18, Issue 1, 2023

Source: Computed using EViews.

Figure 2. Histogram for Bank Nifty Index returns

Source: Computed using EViews.
.08
.06
.04
Quantiles of Normal

.02
.00
-.02
-.04
-.06
-.08
-.2 -.1 .0 .1 .2

Quantiles of Logreturns

Figure 3. Normal QQ plot for daily Bank Nifty Index returns: 2005–2022
Table 3. Results showing summary statistics
Statistics Value
Number of Observations 4206
Mean 0.000548
Median 0.000804
Max 0.172394
Min –0.183130
Standard Deviation 0.018624
Skewness –0.195874
Kurtosis 10.36751
Jarque-Bera t-stat 9539.506
Probability value 0.000000
Sum 2.303677
Sum of Squared. Dev 1.458463

3.3.1. Empirical results to the normal distribution assumption, the study

employs Student’s t (Stud-t) and the Generalized
Following the results of the LM test, the study ap- Error Distribution (GED). Further, to choose an
plies asymmetric GARCH models (EGARCH and appropriate model and distribution, the study us-
GJR-GARCH). Because the data do not conform es the maximum log-likelihood (LnL) approach

132 http://dx.doi.org/10.21511/bbs.18(1).2023.11
 Banks and Bank Systems, Volume 18, Issue 1, 2023

Table 4. Estimation of results using GARCH models

Distribution GARCH (1,1) GJR-GARCH (1,1) EGARCH (1,1)
Normal 0.000930 *** 0.000596 *** 0.000510 ***
µ(MU) Stud-t 0.000818*** 0.000560 ** 0.000514 **
GED 0.000900 *** 0.000654 *** 0.000608 ***
Normal 0.000003 * 0.000003 ** –0.104000 ***
Ω (OMEGA) Stud-t 0.000002 0.000003 –0.099659 ***
GED 0.000002 0.000003 –0.104387 ***
Normal 0.077638 *** 0.028178 *** –0.065713 ***
α (ALPHA) Stud-t 0.076559 *** 0.022394 –0.072205 ***
GED 0.076803 *** 0.025149 –0.068601 ***
Normal 0.917038 *** 0.925181 *** 0.986795 ***
β (BETA) Stud-t 0.919136 *** 0.926487 *** 0.988018 ***
GED 0.917909 *** 0.925489 *** 0.987538 ***
Normal – 0.079641 *** 0.146704 ***
γ (GAMMA) Stud-t – 0.090630 *** 0.139638 ***
GED – 0.084824 *** 0.144127 ***

Note: *, **, and ***: statistically significant at the 10%, 5%, and 1% level of significance, respectively.

and AIC/BIC values congruent with prior studies pha” component indicates that negative volatility
(Franses & Van Dijk, 1996). The EGARCH model reduces the conditional variance for the following
with stud-t was found to be the best-fit model to period (Brooks, 2019).
the returns’ data as it exhibits lower forecast er-
rors and provides a better description of the con- A long memory in the variance is indicated by β
ditional volatility. However, the performance of coefficient. The correlation structure of a specific
Student’s t and GED performance is almost iden- series at long lags is referred to as having a long
tical due to the negligible difference in their AIC memory. Distance observations have persistent
and BIC values. temporal dependency if a series demonstrates long
memory. The estimated coefficient for the GARCH
Table 4 reports the findings of the symmetric term (β) is 0.988018. As the coefficients approach
GARCH (1,1), EGARCH (asymmetric), and GJR unity, previous news is likely to continue affecting
GACRH (asymmetric) models with GED and current volatility for a very long time.
Student’s t distribution. The α and β terms are sta-
tistically significant in the returns series for the For the Bank Nifty returns, the total of α and β
symmetric model. This suggests that regardless of terms is 0.915813, showing that volatility shocks
the sign, squared-lagged innovation considerably are extremely persistent. Because the returns se-
influences conditional variance. ries show their volatility reverting to half its mean
value within 8 days, investors should open a posi-
The results of the EGARCH model for the Bank tion on the 0th day and close it after 16 days2. If a
Nifty returns series are provided in Table 4. The shock to a given system is lasting, volatility per-
leverage coefficient γ for the data series is statis- sists for a prolonged time, and the behavior of vol-
tically significant at the 1% level, which supports atility in the past can be utilized to forecast future
the use of the asymmetric volatility model. In the volatility. Thus, the higher beta (β) coefficient val-
model, the stated α measures how much a volatil- ue (close to one) indicates that volatility shocks are
ity shock today affects volatility in the following highly persistent in the Indian Bank Nifty series;
period (Campbell et al., 1997). The α for the Bank supporting the study’s first hypothesis (H1).
Nifty series is –0.072205. This coefficient is statis-
tically significant, demonstrating that historical Using the EGARCH model, it has been found that
lags can affect future volatility only in the near the γ coefficient for Bank Nifty returns is 0.139638
term. However, the negative estimate on the “al- (positive) and statistically significant at 0.01 (Table
2 The half-life volatility of GARCH model is a statistic to measure the mean reverting time (in terms of average days). To measure mean
reverting time, we have employed the method followed by (Ahmed et al., 2018).

http://dx.doi.org/10.21511/bbs.18(1).2023.11 133
Banks and Bank Systems, Volume 18, Issue 1, 2023

Table 5. Model selection criteria

Criteria Distribution GARCH (1,1) GJR-GARCH (1,1) EGARCH (1,1)
Normal –5.4669 –5.4901 –5.4917
AIC
Stud-t –5.5119 –5.5340 –5.5350
(Akaike)
GED –5.5080 –5.5283 –5.5291
Normal –5.4609 –5.4796 –5.4811
BIC
Stud-t –5.5029 –5.5205 –5.5214
(Bayes)
GED –5.5005 –5.5163 –5.5171
Normal 11500.91 11552.76 11555.97
Log-Likelihood Stud-t 11597.56 11647.09 11649.120
GED 11588.38 11634.11 11635.78

Note: *, **, and ***: statistically significant at the 10%, 5%, and 1% level of significance, respectively.

4). Therefore, it becomes clear that the influence Since information decays slowly in the Bank Nifty
of good news/shocks is greater than the impact returns series, historical information is more sig-
of bad news/shocks on the volatility of the Bank nificant than new knowledge for market partici-
Nifty returns. This suggests that positive shocks pants. This demonstrates what is referred to be
increase volatility in the next period relative to long-memory behavior. Therefore, rather than the
negative shocks of equal magnitude in the Indian nature of the information, these trends are proba-
banking industry, known as the “anti-leverage ef- bly caused by the market microstructures. The re-
fect.” Thus, the result of the leverage coefficient sults are consistent with those reported by Patton
(γ) provides substantial proof to reject the study’s and Sheppard (2015) and Katsiampa et al. (2019),
second hypothesis (H2), confirming that the dis- while low volatility persistence was found by Yaya
ruption in the volatility due to positive shocks is et al. (2019). However, an asset with a high β is no
higher than that of negative shocks, and their im- longer considered a safe haven or good hedge due
pact is not uniform for the Indian Banking series. to its inability to effectively protect investors from
volatile market conditions (Elder & Serletis, 2008).
Because of the persistent nature of volatility, inves-
4. DISCUSSION tors need to consider the volatility shocks for fore-
casting long-term returns behavior and deciding
Recent shreds of evidence on the inconclusive optimal hedging (Abakah et al., 2020).
findings of volatility, demonstrating the impet-
uous retortion of stock returns to abrupt news The result for the EGARCH asymmetry term (γ)
events, is an academic puzzle for discerning asym- shows a positive coefficient for Bank Nifty returns.
metric returns behavior (Thazhungal Govindan The “anti-leverage effect,” which is recognized in
Nair, 2022). The literature lacks a comprehensive extant research, has been theoretically document-
theoretical basis that simultaneously captures ed by Nelson (1991) and Glosten et al. (1993). Over
such data patterns. However, a large body of lit- time, various studies have shown the significance
erature in finance and economics demonstrates of the “anti-leverage effect” (Ghysels et al., 2005;
the persistence of volatility in asset returns within Harrison & Zhang, 1999; Ludvigson & Ng, 2007).
the financial time series (Baillie & Morana, 2009; According to the EGARCH results, positive shocks/
Charfeddine & Khediri, 2016; Greene & Fielitz, news cause more volatility in the subsequent peri-
1977). In other words, the market reacts to infor- od than negative shocks/news of the same size. In
mation gradually over time rather than respond- light of this, it is apparent that volatility rises when
ing immediately to it when it enters the financial returns unexpectedly increase compared to when
market. The findings of the EGARCH model sug- returns decline. However, a stylized feature of fi-
gest that the volatility of Bank Nifty returns is per- nancial volatility is that bad news impacts volatili-
sistent over time; when it rises, it remains high for ty more than good news (Black & Cox, 1976).
a considerable time and returns to its mean only
gradually. Therefore, a new shock will have a long- According to Veronesi (1999), investors’ behav-
term influence on the return’s series. ior is related to the state of the business cycle

134 http://dx.doi.org/10.21511/bbs.18(1).2023.11
 Banks and Bank Systems, Volume 18, Issue 1, 2023

and the impacts of macroeconomic news on value was expected during the study period. In
the financial market instruments. In principle, line with the findings of Zhang et al. (2020), the
positive economic advancements are projected recent outbreak of the COVID-19 pandemic has
to significantly influence the volatility of the wobbled uncertainty in the financial markets.
Bank Nifty returns. This explains why the pres- However, Ledwani et al. (2021) shred evidence
ent study observed a positive gamma (γ) value for a faster market correction in economies like
corresponding to the inverse leverage effect in India after a negative shock. As a result, the
the Indian Bank Nifty returns. Due to the lev- Indian stock market tends to bounce back more
eraging effect, a negative gamma value may be quickly after negative shocks. The market cor-
observed in risky assets (but not all). However, rection hypothesis and the market overreaction
due to the financial and economic turmoil in- theory support the decline and speedy recovery
terwoven with COVID-19, a negative gamma during such periods.

CONCLUSIONS
The objective of the study was to model the conditional variance of Bank Nifty Index returns in India.
The primary analysis demonstrates that the sampling distribution of mean among Bank Nifty returns
is non-normal. Therefore, to examine the long-memory and asymmetric effects, the current study has
focused on a battery of GARCH specifications. However, the model selection criteria exhibit that among
the GARCH-type models, the EGARCH model with Student’s t distribution provides a better descrip-
tion for conditional variance exhibiting lower forecasting errors.

The results demonstrated that the degree of volatility of returns has a tendency to persist and re-
turn to the mean gradually. Besides, the volatility of Bank Nifty returns tends to rise in reaction to
positive shocks as opposed to negative shocks of equal magnitude, suggesting the possibility of an
“anti-leverage effect.” This result supports the rapid market correction following the panic-induced
decline in the Indian Bank Nifty returns. On the other hand, good earnings reports, an announce-
ment of a new product, corporate acquisitions, government policies, and other positive economic
indicators induce investors’ behavior, causing a burlier movement in the Indian Bank Nifty returns.
However, the sample period used in this study corresponds to a period under stress (a bear market);
thus, interesting insights can be drawn from a comparison with a euphoric period (a bull market).
Therefore, a more comprehensive examination of such comparison should be conducted in future
research.

AUTHOR CONTRIBUTIONS
Conceptualization: Nikhil M. N., Lithin B. M., Sanket Ledwani.
Data curation: Nikhil M. N., Lithin B. M., Sanket Ledwani.
Formal analysis: Nikhil M. N., Suman Chakraborty, Satyakam.
Funding acquisition: Suman Chakraborty.
Investigation: Suman Chakraborty, Satyakam.
Methodology: Nikhil M. N., Lithin B. M., Sanket Ledwani.
Project administration: Suman Chakraborty, Satyakam.
Resources: Suman Chakraborty, Satyakam.
Software: Nikhil M. N., Lithin B. M., Sanket Ledwani.
Supervision: Suman Chakraborty, Satyakam.
Validation: Suman Chakraborty, Lithin B. M., Sanket Ledwani.
Visualization: Suman Chakraborty, Satyakam
Writing – original draft: Nikhil M. N., Lithin B. M., Sanket Ledwani.
Writing – reviewing & editing: Suman Chakraborty, Satyakam.

http://dx.doi.org/10.21511/bbs.18(1).2023.11 135
Banks and Bank Systems, Volume 18, Issue 1, 2023

REFERENCES
1. Abakah, E. J. A., Gil-Alana, L. A., 33(8), 1577-1592. https://doi. 18. Brooks, C., & Burke, S. P.
Madigu, G., & Romero-Rojo, F. org/10.1016/j.jedc.2009.02.009 (2003). Information criteria
(2020). Volatility persistence in for GARCH model selection.
9. Begley, T. A., Purnanandam,
cryptocurrency markets under European Journal of Finance,
A., & Zheng, K. (2017). The
structural breaks. International 9(6), 557-580. https://doi.org
Strategic Underreporting of Bank
Review of Economics and Risk. Review of Financial Studies, /10.1080/1351847021000029188
Finance, 69, 680-691. https://doi. 30(10), 3376-3415. https://doi. 19. Bryant, J. (1980). A model of
org/10.1016/j.iref.2020.06.035 org/10.1093/rfs/hhx036 reserves, bank runs, and deposit
2. Acharya, V., Engle, R., & 10. Bennett, R. L., Hwa, V., & Kwast, insurance. Journal of Banking and
Richardson, M. (2012). Capital M. L. (2015). Market discipline by Finance, 4(4), 335-344. https://doi.
shortfall: A new approach to bank creditors during the 2008– org/10.1016/0378-4266(80)90012-
ranking and regulating systemic 2010 crisis. Journal of Financial 6
risks. American Economic Stability, 20, 51-69. https://doi. 20. Campbell, J. Y., Lo, A. W., &
Review, 102(3), 59-64. https://doi. org/10.1016/j.jfs.2015.06.003 MacKinlay, A. C. (1997). The
org/10.1257/aer.102.3.59
11. Bhatia, P., & Gupta, P. (2020). Econometrics of Financial Markets.
3. Ahmed, R. R., Vveinhardt, J., Sub-prime Crisis or COVID-19: Princeton University Press. https://
Streimikiene, D., & Channar, A Comparative Analysis of doi.org/10.2307/J.CTT7SKM5
Z. A. (2018). Mean reversion in Volatility in Indian Banking 21. Candelon, B., Carare, A., & Miao,
international markets: evidence Sectoral Indices. FIIB Business K. (2016). Revisiting the new
from GARCH and half-life Review, 9(4), 286-299. https://doi. normal hypothesis. Journal of
volatility models. Economic org/10.1177/2319714520972210 International Money and Finance,
Research-Ekonomska Istrazivanja,
12. Bhattacharyay, B. N. (2013). 66, 5-31. https://doi.org/10.1016/j.
31(1), 1198-1217. https://doi.org/1
Determinants of bond market jimonfin.2015.12.005
0.1080/1331677X.2018.1456358
development in Asia. Journal
22. Charfeddine, L., & Khediri, K.
4. Âhou, R. B. A., & Gueyie, J.-P. of Asian Economics, 24, 124-137.
B. (2016). Time varying market
(2001). Canadian chartered banks’ https://doi.org/10.1016/j.asie-
efficiency of the GCC stock
stock returns and exchange rate co.2012.11.002
markets. Physica A: Statistical
risk. Management Decision, 39(4),
285-295. https://doi.org/10.1108/ 13. Black, F., & Cox, J. C. (1976). Mechanics and Its Applications,
Valuing Corporate Securities: 444, 487-504. https://doi.
EUM0000000005463
Some Effects of Bond Indenture org/10.1016/j.physa.2015.09.063
5. Anagnostopoulos, I. (2018). Provisions. The Journal of Finance,
23. Christoffersen, S. E. K., & Simutin,
Fintech and regtech: Impact on 31(2), 351-367. https://doi.
M. (2017). On the demand for
regulators and banks. Journal of org/10.1111/J.1540-6261.1976.
high-beta stocks: Evidence from
Economics and Business, 100, 7-25. TB01891.X
mutual funds. Review of Financial
https://doi.org/10.1016/j.jecon-
14. Bollerslev, T. (1986). Generalized Studies, 30(8), 2596-2620. https://
bus.2018.07.003
autoregressive conditional doi.org/10.1093/rfs/hhx022
6. Anginer, D., Demirguc-Kunt, heteroskedasticity. Journal of
24. Crockett, A. (2002). Market
A., & Zhu, M. (2014). How Econometrics, 31(3), 307-327.
discipline and financial stability.
does competition affect bank https://doi.org/10.1016/0304-
Journal of Banking and Finance, 26,
systemic risk? Journal of 4076(86)90063-1
977-987. Retrieved from https://
Financial Intermediation, 23(1),
15. Brealey, R., Leland, H. E., & nscpolteksby.ac.id/ebook/files/
1-26. https://doi.org/10.1016/j.
Pyle, D. H. (1977). Informational Ebook/Journal/2015/Banking%20
jfi.2013.11.001
Asymmetries, Financial and%20Finance/Surya%20Neg-
7. Baele, L., de Jonghe, O., & Structure and Financial ara%20Vol.%2026/Volume%20
vander Vennet, R. (2007). Does Intermediation. The Journal of 26%20Issue%205/Market%20dis-
the stock market value bank Finance, 32(2), 371-387. https:// cipline%20and%20financial%20
diversification? Journal of Banking doi.org/10.1111/j.1540-6261.1977. stability.pdf
and Finance, 31(7), 1999-2023. tb03277.x
25. De Jonghe, O. (2010). Back to
https://doi.org/10.1016/j.jbank- 16. Brooks, C. (2008). Introductory the basics in banking? A micro-
fin.2006.08.003 econometrics for finance (2nd ed.). analysis of banking system
8. Baillie, R. T., & Morana, C. (2009). Cambridge University Press. stability. Journal of Financial
Modelling long memory and 17. Brooks, C. (2019). Intermediation, 19(3), 387-
structural breaks in conditional Introductory Econometrics 417. https://doi.org/10.1016/j.
variances: An adaptive for Finance. Cambridge jfi.2009.04.001
FIGARCH approach. Journal of University Press. https://doi. 26. Demirer, M., Diebold, F. X.,
Economic Dynamics and Control, org/10.1017/9781108524872 Liu, L., & Yilmaz, K. (2018).

136 http://dx.doi.org/10.21511/bbs.18(1).2023.11
 Banks and Bank Systems, Volume 18, Issue 1, 2023

Estimating global bank network 10370). Retrieved from https:// relationship in the Indian stock
connectedness. Journal of Applied cepr.org/publications/dp10370 market. South Asia Economic
Econometrics, 33(1), 1-15. https:// Journal, 8(1), 99-116. https://doi.
36. Ghysels, E., Santa-Clara, P., &
doi.org/10.1002/jae.2585 org/10.1177/139156140600800106
Valkanov, R. (2005). There is
27. Diamond, D. W. (1984). Financial a risk-return trade-off after all. 45. Kasman, S., Vardar, G., & Tunç, G.
Intermediation and Delegated Journal of Financial Economics, (2011). The impact of interest rate
Monitoring. The Review of 76(3), 509-548. https:// and exchange rate volatility on
Economic Studies, 51(3), 393-414. doi.org/10.1016/J.JFINE- banks’ stock returns and volatility:
https://doi.org/10.2307/2297430 CO.2004.03.008 Evidence from Turkey. Economic
Modelling, 28(3), 1328-1334.
28. Diamond, D. W., & Dybvig, P. 37. Glosten, L. R., Jagannathan, R.,
https://doi.org/10.1016/j.econ-
H. (1983). Bank Runs, Deposit & Runkle, D. E. (1993). On the
mod.2011.01.015
Insurance, and Liquidity. Journal Relation between the Expected
of Political Economy, 91(3), 401-419. Value and the Volatility of the 46. Katsiampa, P., Corbet, S., &
Retrieved from http://www.jstor. Nominal Excess Return on Lucey, B. (2019). High frequency
org/stable/1837095 Stocks. The Journal of Finance, volatility co-movements in
48(5), 1779-1801. https://doi. cryptocurrency markets. Journal
29. Dickey, D. A., & Fuller, W. A.
org/10.1111/J.1540-6261.1993. of International Financial Markets,
(1981). Likelihood Ratio Statistics
TB05128.X Institutions and Money, 62, 35-52.
for Autoregressive Time Series
https://doi.org/10.1016/j.in-
with a Unit Root. Econometrica, 38. Greene, M. T., & Fielitz, B. D.
tfin.2019.05.003
49(4), 1057-1072. https://doi. (1977). Long-term dependence
org/10.2307/1912517 in common stock returns. Journal 47. Kumar, K., & Prakash, A. (2019).
of Financial Economics, 4(3), 339- Examination of sustainability
30. Elder, J., & Serletis, A. (2008).
349. https://doi.org/10.1016/0304- reporting practices in Indian
Long memory in energy futures
405X(77)90006-X banking sector. Asian Journal
prices. Review of Financial
of Sustainability and Social
Economics, 17(2), 146-155. https:// 39. Gupta, A. (2010). Financial
Responsibility, 4(1). https://doi.
doi.org/10.1016/j.rfe.2006.10.002 crisis enforcing global banking
org/10.1186/s41180-018-0022-2
reforms. Business Strategy Series,
31. Elnahass, M., Trinh, V. Q., & Li, T.
11(5), 286-294. https://doi. 48. Laeven, L., Ratnovski, L., & Tong,
(2021). Global banking stability
org/10.1108/17515631011080696 H. (2016). Bank size, capital, and
in the shadow of Covid-19
systemic risk: Some international
outbreak. Journal of International 40. Harrison, P., & Zhang, H. H.
evidence. Journal of Banking
Financial Markets, Institutions and (1999). An investigation of the
and Finance, 69(1), S25-S34.
Money, 72, 101322. https://doi. risk and return relation at long
https://doi.org/10.1016/j.jbank-
org/10.1016/j.intfin.2021.101322 horizons. Review of Economics and
fin.2015.06.022
Statistics, 81(3), 399-408. https://
32. Engle, R. F. (1982). Autoregressive
doi.org/10.1162/003465399558337 49. Ledwani, S., Chakraborty, S., &
Conditional Heteroscedasticity
Shenoy, S. S. (2021). Spatial tale
with Estimates of the Variance 41. Hasan, I., Siddique, A., & Sun, X.
of G-7 and Brics stock markets
of United Kingdom Inflation. (2015). Monitoring the “invisible”
during COVID-19: An event
Econometrica, 50(4), 987-1007. hand of market discipline: Capital
study. Investment Management
Retrieved from https://www.jstor. adequacy revisited. Journal of
and Financial Innovations, 18(2),
org/stable/1912773 Banking and Finance, 50, 475-492.
20-36. https://doi.org/10.21511/
https://doi.org/10.1016/j.jbank-
33. Flannery, M. J. (2001). The Faces imfi.18(2).2021.03
fin.2014.03.029
of “Market Discipline”. Journal
50. Lin, Z. (2018). Modelling and
of Financial Services Research, 42. Hoggarth, G., Reis, R., & Saporta,
forecasting the stock market
20(3), 107-119. https://doi. V. (2002). Costs of banking
volatility of SSE Composite Index
org/10.1023/A:1012455806431 system instability: Some empirical
using GARCH models. Future
evidence. Journal of Banking
34. Franses, P. H., & Van Dijk, D. Generation Computer Systems, 79,
& Finance, 26(5), 825-8556.
(1996). Forecasting stock market 960-972. https://doi.org/10.1016/j.
Retrieved from https://personal.
volatility using (non-linear) Garch future.2017.08.033
lse.ac.uk/reisr/papers/02-Hoggar-
models. Journal of Forecasting,
thReisSaporta.pdf 51. Ludvigson, S. C., & Ng, S. (2007).
15(3), 229-235. https://doi.
The empirical risk-return relation:
org/10.1002/(SICI)1099- 43. Karmakar, M. (2005). Modeling
A factor analysis approach. Journal
131X(199604)15:3%3C229::AID- Conditional Volatility of the
of Financial Economics, 83(1),
FOR620%3E3.0.CO;2-3 Indian Stock Markets. Vikalpa:
171-222. https://doi.org/10.1016/J.
The Journal for Decision
35. Fratzscher, M., & Rieth, M. (2015). JFINECO.2005.12.002
Makers, 30(3). https://doi.
DP10370 Monetary Policy. Bank
org/10.1177/0256090920050303 52. Mahajan, V., Thakan, S., &
Bailouts and the Sovereign-Bank
Malik, A. (2022). Modeling and
Risk Nexus in the Euro Area 44. Karmakar, M. (2007). Asymmetric
Forecasting the Volatility of
(CEPR Discussion Papers No. volatility and risk-return

http://dx.doi.org/10.21511/bbs.18(1).2023.11 137
Banks and Bank Systems, Volume 18, Issue 1, 2023

NIFTY 50 Using GARCH and Shock Spillover Over Different 63. Tripathy, T., & Gil-Alana, L.
RNN Models. Economies, 10(5), Maturities: A Journey from A. (2015). Modelling time-
102. https://doi.org/10.3390/econ- Normal to Covid-19 Period. varying volatility in the Indian
omies10050102 Asia-Pacific Financial Markets, 29, stock returns: Some empirical
697-734. https://doi.org/10.1007/ evidence. Review of Development
53. Moshirian, F., & Wu, Q. (2009).
S10690-022-09371-X Finance, 5(2), 91-97. https://doi.
Banking industry volatility
org/10.1016/j.rdf.2015.04.002
and banking crises. Journal of 59. Stiroh, K. J. (2006). New evidence
International Financial Markets, on the determinants of bank 64. Veronesi, P. (1999). Stock Market
Institutions and Money, 19(2), risk. Journal of Financial Services Overreaction to Bad News
351-370. https://doi.org/10.1016/j. Research, 30(3), 237-263. https:// in Good Times: A Rational
intfin.2008.02.002 doi.org/10.1007/s10693-006- Expectations Equilibrium Model.
0418-5 The Review of Financial Studies,
54. Nelson, D. B. (1991). Conditional
12(5), 975-1007. Retrieved
Heteroskedasticity in Asset 60. Suhadak, S., Kurniaty, K.,
Returns: A New Approach. from https://www.jstor.org/
Handayani, S. R., & Rahayu, S. M.
Econometrica, 59(2), 347. https:// stable/2645973
(2019). Stock return and financial
doi.org/10.2307/2938260 performance as moderation 65. Williams, B. (2014). Bank risk
55. Padhi, P. (2006). Persistence variable in influence of good and national governance in
and Asymmetry Volatility in corporate governance towards Asia. Journal of Banking and
Indian Stock Market. Journal corporate value. Asian Journal of Finance, 49, 10-26. https://
of Quantitative Economics, 4, Accounting Research, 4(1), 18-34. doi.org/10.1016/j.jbank-
103-113. https://doi.org/10.1007/ https://doi.org/10.1108/AJAR-07- fin.2014.08.014
BF03546451 2018-0021
66. Yaya, O. O. S., Ogbonna, A. E., &
56. Patton, A. J., & Sheppard, K. 61. Thazhungal Govindan Nair, Olubusoye, O. E. (2019). How
(2015). Good volatility, bad S. (2022). Price extremes persistent and dynamic inter-
volatility: Signed jumps and the and asymmetric dependence dependent are pricing of Bitcoin
persistence of volatility. Review structures in stock returns: to other cryptocurrencies before
of Economics and Statistics, 97(3), the emerging market evidence. and after 2017/18 crash? Physica
683-697. https://doi.org/10.1162/ Journal of Economic Studies, A: Statistical Mechanics and
REST_a_00503 49(8), 1502-1523. https://doi. Its Applications, 531, 121732.
org/10.1108/JES-10-2021-0507 https://doi.org/10.1016/j.
57. Phillips, P. C. B., & Perron, P. physa.2019.121732
(1988). Testing for a unit root in 62. Thiripalraju, M., & Acharya, R.
time series regression. Biometrika, H. (2010). Modeling Volatility for 67. Zhang, D., Hu, M., & Ji, Q. (2020).
75(2), 335-346. https://doi. the Indian Stock Market. The IUP Financial markets under the
org/10.1093/biomet/75.2.335 Journal of Applied Economics, 9(1), global pandemic of COVID-19.
79-105. Retrieved from https:// Finance Research Letters, 36,
58. Rout, S. K., & Mallick, H. papers.ssrn.com/sol3/papers. 101528. https://doi.org/10.1016/j.
(2022). Sovereign Bond Market cfm?abstract_id=1535856 frl.2020.101528

138 http://dx.doi.org/10.21511/bbs.18(1).2023.11