The current issue and full text archive of this journal is available on Emerald Insight at:

www.emeraldinsight.com/1746-8809.htm

Indian telecom
Impact of regulatory sector
announcements on systemic risk
in the Indian telecom sector
Sushma Priyadarsini Yalla 1395
National Institute of Industrial Engineering, Mumbai, India
Received 24 August 2017
Som Sekhar Bhattacharyya Revised 15 December 2017
25 February 2018
Strategic Management, National Institute of Industrial Engineering, 19 April 2018
Mumbai, India, and Accepted 29 April 2018

Karuna Jain
National Institute of Industrial Engineering, Mumbai, India

Abstract
Purpose – Post 1991, given the advent of liberalization and economic reforms, the Indian telecom sector
witnessed a remarkable growth in terms of subscriber base and reduced competitive tariff among the service
providers. The purpose of this paper is to estimate the impact of regulatory announcements on systemic risk
among the Indian telecom firms.
Design/methodology/approach – This study employed a two-step methodology to measure the impact of
regulatory announcements on systemic risk. In the first step, CAPM along with the Kalman filter was used to
estimate the daily β (systemic risk). In the second step, event study methodology was used to assess the
impact of regulatory announcements on daily β derived from the first step.
Findings – The results of this study indicate that regulatory announcements did impact systemic risk
among telecom firms. The study also found that regulatory announcements either increased or decreased
systemic risk, depending upon the type of regulatory announcements. Further, this study estimated the
market-perceived regulatory risk premiums for individual telecom firms.
Research limitations/implications – The regulatory risk premium was either positive or negative,
depending upon the different types of regulatory announcements for the telecom sector firms. Thus, this
study contributes to the theory of literature by testing the buffering hypothesis in the context of Indian
telecom firms.
Practical implications – The study findings will be useful for investors and policy-makers to estimate the
regulatory risk premium as and when there is an anticipated regulatory announcement in the Indian
telecom sector.
Originality/value – This is one of the first research studies in exploring regulatory risk among the Indian
telecom firms. The research findings indicate that regulatory risk does exist in the telecom firms of India.
Keywords India, Systemic risk
Paper type Research paper

1. Introduction
Regulation is defined as the imposition of rules by governments (Posner, 1974), which are
intended to maximize social welfare (Posner, 1974). According to Brito and Dudley (2012),
regulation is a set of administrative laws or rules and is the primary vehicle by which
governments implement laws and objectives. Regulations are specific standards concerning
what businesses, individuals and organizations can or cannot do (Brito and Dudley, 2012).
As a part of economic reforms, policy liberalization and privatization with the intent of
fueling growth in investments and capital flows, the Government Of India (GOI) had
International Journal of Emerging
permitted private players to become a part of the telecom sector since the year 1992. In the Markets
early 90s during the time of economic reforms, there was an enormous gap in the demand Vol. 13 No. 5, 2018
pp. 1395-1416
and supply of telecom services in India. This motivated many private telecom service © Emerald Publishing Limited
1746-8809
providers to enter into the sector, expecting substantial growth and higher returns. By the DOI 10.1108/IJoEM-08-2017-0307
IJOEM year 2000, the telecom sector consisted both private and public-sector firms. The two major
13,5 public-sector firms were Mahanagar Telephone Nigam Limited (MTNL) and Bharat
Sanchar Nigam Limited (BSNL). These public-sector companies have continued to be
market leaders in wired services and also have substantial presence in the wireless market
with a share of 0.35 and 8.26 percent, respectively (TRAI, 2016a, b). In addition to them,
there are five major private firms contributing to 70 percent of the market share of wireless
1396 services in India (TRAI, 2016a, b). These firms are Bharati Airtel (read Airtel), Vodafone
India, Idea Cellular (read Idea), Reliance Communications, and Tata Communications
Limited (read Tata) with market shares of 24, 18, 16, 7 and 5 percent, respectively (TRAI,
2016a, b). Of the above public sector and private firms, BSNL and Vodafone are not listed in
the Indian stock market.
The GOI established the Telecom Regulatory Authority of India (TRAI) in the year 1997
to regulate telecom services (TRAI, 2015). The primary objective of TRAI was to setup tariff,
license auctioning, spectrum management, setting up of benchmark standards for the
quality of service and creating a favorable environment for investments into the Indian
telecom sector (TRAI, 2015). TRAI makes necessary policy and regulatory amendments
time to time, ensuring the welfare of consumers and protecting the interest of investors.
Further, TRAI is also responsible for maintaining fair competition among telecom firms by
developing policies to encourage private participation in the Indian telecom services. It has
been argued that without the presence of regulatory bodies like TRAI, the sustainability of
telecom firms in India would be challenging (TRAI, 2015).
The telecom sector is capital intensive and requires substantial investments for the
development of basic infrastructure (Chalmeau, 2012). The major portion of investment is
sunk cost (once invested in infrastructure, cannot be recovered easily as salvage value is less)
in nature (Haucap, 2003). Telecom firms are primarily technology driven, wherein the
economic feasibility of the project cannot be deterministically ascertained (Haucap, 2003).
Thus, it has been advocated that telecom firms have to be judicious while making
investments, and might fail to earn the minimum required rate of return otherwise (Chalmeau,
2012). Elton and Gruber (1971) advocated that regulated firms operate with more constraints
due to the discretionary behavior of the regulator. Hence, there could be uncertainty in
regulated firms with respect to performance as compared to non-regulated firms. This
uncertainty could be modeled as regulatory risk (Ergas and Small 2001; Knieps and Weiß,
2007). Therefore, the estimation of cost of capital for the regulated firms is likely to be different
from that of non-regulated firms due to the existence of regulatory risk (Elton and Gruber,
1971). The Capital Asset Pricing Model (CAPM) theory (Sharpe, 1964; Lintner, 1965; Mossin,
1966) defines risk as the co-variability of a security’s return with the market’s return or
systemic risk. Systemic risk or market risk is the risk that cannot be eliminated through the
diversification of portfolio (Sharpe, 1964). The CAPM theory assumes that all investors hold
perfectly diversified portfolios, they are presumed to be exposed only to systemic risk and the
market (according to the model) will not reward them with a risk premium for unsystematic or
non-market risk (Sharpe, 1964; Lintner, 1965; Mossin, 1966).
Risk drivers are the primary factors while estimating the cost of capital, which is a
further key input in the regulatory rate base and price setting managed by regulators
(Pedell, 2006). The prices set by the regulator have an influence on the expected future cash
flows, which, in turn, influences regulatory rate base setting; this ultimately affects the
firm’s cash flows (Wright et al., 2003; Pedell, 2006). Firm’s cash flows further impact its
profitability and sustainability. Thus, there is a need to ascertain the level of regulatory risk.
This determination of regulatory risk would be useful for the regulator as the risk-adjusted
cost of capital is a key input for setting regulated prices (Pedell, 2006). There are arguments
that investors could also adjust their funds if they know how regulatory risk would impact
systemic risk, thereby impacting the cost of equity capital (Antoniou and Pescetto, 1997).
Cost of equity capital is defined as the minimum rate of return acquired on fund investment Indian telecom
for a given level of risk, below which investors would prefer to invest in other investments sector
that offer higher return with the same level of risk (Ross, 1976).
In regulated firms, regulatory interventions might impact the valuation of the firm and
the cost of capital in a more complex fashion unlike the case of non-regulated firms (Elton
and Gruber, 1971). Understanding the impact of regulation on systemic risk and cost of
capital would enable the telecom firms to estimate revenue requirements at the firm level 1397
(Chalmeau, 2012). Hence, it is important to understand the impact of regulation (regulatory
announcements) on systemic risk and their further effect on the cost of equity capital
(Gaggero, 2007). Theoretically, the impact of regulation on systemic risk has not been very
clear, whether it increases or decreases systemic risk (Robinson and Taylor, 1998).
Therefore, the objective of the present study is to understand the impact of regulatory
announcements on systemic risk and their further bearing on the cost of equity capital
among the Indian telecom firms. The regulatory announcements considered in this study
are primarily focused on “Price” or “Tariff,” “Service Quality” and “Competition among
telecom operators” in the Indian telecom sector.
Post a detailed literature review as a part of this study, it could be concluded that the
existing body of literature mostly explored the impact of regulatory announcements on
systemic risk and cost of equity capital in developed countries like the USA and UK (Wright
et al., 2003; Pedell, 2006). In some of the developed countries, the extent of regulation was not
too strong as regulatory announcements impacted systemic risk, i.e., regulatory risk existed
(Tapia, 2012). If regulatory announcements impact the systemic risk, then the extent of
regulation is not strong and vice versa (Tapia, 2012). It is, therefore, necessary to ascertain
the extent of regulation in developing countries, especially in India, in contrast with
developed counties. Hence, this study aims to analyze the existence of regulatory risk in
India, targeting the telecom firms. Moreover, there is a paucity of empirical research
evidence of this nature in developing countries, like India. Given the background, the
primary objective of this paper is to analyze the impact of regulatory announcements on
systemic risk and cost of equity capital for listed Indian telecom firms. In this study, the
Kalman filter algorithm has been applied to the CAPM to estimate the dynamic daily β of
telecom firms. Further, the event study methodology has been used to estimate the impact of
regulatory announcements on systemic risk.
The paper is organized into six sections: the introduction has been outlined in Section 1.
In Section 2, review of literature, identified gaps in collective knowledge and objectives of
the study have been enlisted. Data and methods have been detailed in Section 3. In Section 4,
analysis and results have been described, whereas in Section 5, the discussion of the results
has been carried out. In Section 6, research and practice implications, limitations of the study
and future research options have been provided.

2. Review of literature
To conduct the literature review, EBSCO and ProQuest databases were searched. The
keywords used for the search were “regulation and cost of capital,” “regulatory risk,”
“impact of regulation on cost of capital” and such others. Around 60 articles were collated,
which further narrowed down to 30 on the basis of relevance. During the literature review, it
was observed that there were several studies exploring the effect of regulation and
valuation of firms. Elton and Gruber (1971) advocated that in the case of regulated firms,
regulatory interventions impacted the valuation of the firm and the cost of capital.
Therefore, the valuation and measurement of cost of capital of regulatory firms was
different from that of non-regulated firms. Pedell (2006) noted that the measurement of the
effect of regulation on the cost of capital was challenging. Pedell (2006) segregated
regulatory factors that affect systemic risk and cost of capital, some of which are “time
IJOEM elapsed since last regulatory review,” “actual value of rate of return in the past,” “elections”
13,5 (since the issue of election notification to the period till results are declared) and such others.
Literature on regulation reflected that regulators address the concerns of both consumers
and firms (Stigler, 1971). This has often led to a trade-off between the consumer’s interest
(high consumer surplus with low firm price realization) and the firm’s interest (high firm
profit realization – low consumer price). Stigler (1971) wrote that if a demand and cost shock
1398 affected the firm’s profit negatively, the regulator paid off this effect by increasing the price
of service and vice versa. There is redistribution because the regulator is balancing
consumer surplus and firm profit through regulation (Stigler, 1971).
Peltzman (1976) and Riddick (1992) argued that regulated firms faced less of a risk than
their unregulated counterparts because regulation buffers the demand and cost shocks. On
the contrary, Joskow and MacAvoy (1975) argued that regulated firms could also face more
risk than unregulated firms due to “regulatory lag effect (also known as reinforcing effect),”
which reinforces the risk.
A significant number of studies were found to have adopted the event study
methodology ( Joskow and Rose, 1989; Johanning and Ruhle, 2004) and CAPM along with
the Kalman filter algorithm (Buckland and Fraser, 2000; Paleari and Redondi, 2005;
Barcelos and da Silveira Bueno, 2010; Chalmeau, 2012) for ascertaining the impact of
regulatory announcements on systemic risk and cost of equity. In the event study
methodology, the business, political and economic events are considered as discrete
dummy variables in the model.
For instance, Binder (1985, 1998) conducted an event study methodology with monthly
and daily stock returns and found significant stock price changes owing to the chance of a
regulatory announcement. Antoniou and Pescetto (1997) measured the effect of regulation
on the cost of equity capital for British telecom. They evaluated time-varying β using CAPM
along with the Kalman filter algorithm and regressed the evaluated time-varying β against
daily dummy variables of regulatory announcements to analyze the impact of regulatory
announcements on systemic risk. Their analysis reported that there was a significant
impact of regulatory announcements on systemic risk. Trillig (2012) investigated whether
political decisions in Europe influence the risk/return profile of technology-based companies
from the clean tech industry. The author calculated time-varying β and applied the event
study methodology to examine the impact of regulatory announcements on the company’s
market value. Results deliberated that capital markets showed significant reactions to
regulatory announcements in Europe.
On the other hand, other methodologies have also been adopted for finding the impact of
regulation on systemic risk, and thereby cost of equity capital. One such methodology is the
EGARCH model adopted by Schiereck and Trillig (2014). In their study, they examined the
impact of political risk on the German solar energy industry and found that the volatility
response varied with the exposure to political risk. Companies with higher exposure to
political risk showed significant volatility response.
Also, a good many studies have attempted to test the buffering hypothesis propounded
by Peltzman (1976). One of the interesting studies by Davidson et al. (1997) stated that a
more light-handed regulation increased systemic risk, and consequently, the risk-adjusted
cost of capital, thereby opposing the buffering hypothesis. On the other hand, Nwaeze (2000)
and Johanning and Ruhle (2004) explicitly supported the buffering hypothesis that
regulation decreased the risk but did not reduce it to zero, which was observed from the
β factors of most utilities. Hence, it could be concluded that stronger regulation has a
buffering effect on risk-adjusted cost of capital. However, Boubakri et al. (2012) examined
the cost of equity capital of politically connected firms using multivariate regression,
yielding strong evidence that politically connected firms were less at risk than
non-connected ones. Hence, this study strongly supported the Peltzmen hypothesis.
 In a recent study, Angeles (2017) stated that regulatory risk increased the return on Indian telecom
investment and caused underinvestment in industries with high sunk costs in the context of sector
Philippines. Furthermore, Pham et al. (2017) tested the effect of financial regulatory
announcements on risk and return in the Vietnamese equity market and found evidentiary
support for both positive and negative market reactions to bank regulatory announcements
across financial and non-financial sectors.
Certain studies communicated about the type of regulatory regime and how it impacted 1399
the level of risk in those regulated industries; one such study was by Gaggero (2012) who
tested empirically that whether regulation characterized by high incentives (price cap
regulation) implies more risk to firms than regulation characterized by low incentives (rate
of return regulation). The author found that different regulatory regimes did not result in
different levels of risk to the regulated firms and stated that these findings could be driven
by a higher level of development of financial markets combined with a sophisticated
diversifying behavior of regulated firms. On the contrary, the study of Camacho and
Menezes (2013) reported a higher cost of capital under price cap regulation and a lower cost
of capital under rate of return regulation.
Several other research attempts have targeted the uncertainty caused by the
discretionary behavior of the regulator in terms of firm’s other performance parameters
like indirect costs with respect to regulatory risk and the level of debt with respect to the
extent of regulation. For instance, Maegli and Jaag (2009) studied the construct of indirect
costs of regulatory governance with respect to regulatory risk and uncertainty and stated
that there exist regulatory risk and uncertainty on how the regulators behave as well as on
how the regulatory framework evolved and what the effects on their business were. Also,
Moore et al. (2014) examined the capital structure of regulated infrastructure firms. They
developed a model showing that the ratio of liabilities to assets, and leverage was lower
under high-powered regulation and that firms operating under high-powered regulation
made proportionally more reduction in leverage when the cost of the debt increased. They
tested the model using a panel data set of 124 transport concessions in Brazil, Chile,
Colombia and Peru from 1992 to 2011. Findings from the results of panel data have
supported the model predictions.
Several authors have used the Kalman filter and event study methodology for finding the
impact of regulation (regulatory announcements) on systemic risk (Antoniou and Pescetto,
1997; Buckland and Fraser, 2000; Barcelos and da Silveira Bueno, 2010; Chalmeau, 2012) in
the context of both developed and developing countries. Hence, the Kalman filter and event
study methodology were deemed suitable techniques to determine the impact of regulation
on systemic risk in this study.
As stated earlier, only a handful of studies have addressed the impact of regulatory risk
on equity capital in developing countries, though a few authors have attempted so. For
instance, Ghosh and Kathuria (2016) investigated the impact of regulatory governance on
the performance of thermal power plants in India. They hypothesized that greater the
quality of regulation in a federal Indian state, greater was the efficiency of electric
generation utilities. Importantly, the present study did not find any literature that talked
about the impact of regulation on systemic risk and thereby, the cost of equity capital in the
Indian context. Hence, the present study aims to assess the impact of regulation on systemic
risk in the context of India.
Table I depicts the most relevant studies pertaining to the impact of regulation on systemic
risk and thereby, the cost of equity capital. Literature review reveals that the impact of
regulation is highly variable, being positive, negative or neutral (no impact) in different
contexts around the world. However, the most prevalent regulatory effect (buffering or
reinforcing risk) appeared to be dependent on the design of the regulatory system and process;
ultimately, this could only be answered empirically (Pedell, 2006). Hence, there is compelling
IJOEM Author Direction of
13,5 S. No. and year Findings Methodology significance Context

1 Peltzman Regulation reduced conventional Lagrangian Negative –

(1976) measures of owner risk by buffering the multiplier
firm against demand and cost changes; method
also the variability of profits (and stock
1400 prices) was lowered
2 Binder Examined the ability of stock return event Multivariate Both Positive USA
(1985) studies to measure the effect of regulation. regression and Negative
A series of tests were conducted with model, event depending on
monthly and daily return and found study regulatory
significant stock price changes about as methodology event
often as is expected owing to the chance of
an announcement
3 Robinson Tested the presence of regulatory risk by Event study Positive UK
and Taylor modeling the conditional volatility of methodology
(1998) equity returns before and after 30 ARCH model
significant regulatory events using an
ARCH process; found that the conditional
variance increased after the intervention
for most of these events
4 Buckland Utilities regulation attempted to attenuate CAPM along Positive UK
and Fraser the effects of market failure. Also, risk in Kalman Filter
(2000) the case of electricity utilities was time-
variant; political and regulatory
influences significantly affected the
systematic risk faced by electricity utility
shareholders
Results also stated that impact was not
uniform across electricity utilities
5 Paleari and Analyzed the impact of regulation on Kalman Negative UK
Redondi company risk Filter, Event
(2005) Systemic risk varied significantly during study
the analysis period and established methodology
negative relationships between abnormal
returns and overall risk variations and
market correlation
6 Barcelos Found evidence of regulatory risk in CAPM along No Brazil
and da electricity, telecommunications and all Kalman filter, significance
Silveira other regulated sectors in Brazil Random
Bueno Further, regulatory changes in the effect panel
(2010) country neither reduced nor increased the data
β of the regulated sectors, opposing the
buffering hypothesis proposed by
Peltzman (1976)
7 Chalmeau Regulation had impacted both systemic CAPM, Both positive France
(2012) risk and abnormal returns for two French Kalman filter, and negative
telecom operators Event study depending on
Further, found that results on mobile methodology regulatory
wholesale price regulation supported the event
theory of Peltzman (1976) that regulation
lowers risk (1976) that regulation
Table I. lower risk
Summary of findings 8 Ben-Ami Found that the risk to which the regulated Multivariate No South Africa
from most relevant (2015) company was exposed to in terms of regression significance
literature for the variability of returns was not significantly
present study different from that of the market
evidence that there is a dearth of research addressing the impact of regulation on systemic Indian telecom
risk in the context of India. Thus, there is a necessity to study whether regulation increases or sector
decreases the systemic risk of a regulated firm. This may be answered by understanding
whether regulation has any impact on systemic risk. This systemic risk is captured by β in the
CAPM (Morin and Hillman, 1994; Grayburn et al., 2002). In addition to the sector-related
regulatory announcements, literature had also reported that general elections can induce
regulatory risk through impacting systemic risk (Antoniou and Pescetto, 1997; Strausz, 2017). 1401
Therefore, the present study also aims to understand the impact of general elections on
systemic risk faced by telecom firms in India.
Based on the gaps observed in the existing body of literature, the authors have identified
that impact of regulation on systemic risk in Indian regulated industries like the telecom
sector can be addressed. Hence, the objective of the study would be to: test the dynamic
nature of systemic risk (β) faced by telecom firms (reflecting the regulatory risk); ascertain
the impact of regulatory announcements on systemic risk faced by telecom firms and how
the impact is across different firms; and estimate the impact of general elections on systemic
risk faced by telecom firms.

3. Data and methods

The data required for this study comprised of four variables, namely, stock price data, risk-free
rate return, regulatory announcements and general elections, and the studied sample period
was from January 1999 to June 2015. Data pertaining to regulatory announcements were
collected from the TRAI website (the major regulatory announcements data has been presented
in Table AI). All the regulatory announcements made by TRAI pertaining to price, service
quality and competition were considered for a period between January 1999 and June 2015 for
analysis. Regular, non-regulatory events like the announcement of quarterly earnings by the
respective companies were not considered, such has been advocated by Antoniou and Pescetto
(1997) and Buckland and Fraser (2000). The exact date of regulatory announcements released
to the public (as reported in TRAI website) was considered for the event study proposed in this
paper (Binder, 1985). The classification of regulatory announcements analyzed in the study has
been presented in Table II. In this study, researchers did not identify any service negative or
competition negative regulatory announcements as per data collected from the TRAI website.
The second data required for this study were stock price data of listed Indian telecom
firms. The stock price data used in this study were the daily adjusted closing stock price
(adjusted for bonus and rights issue) of the selected telecom firms (Tata, Airtel and MTNL as
only these firms had the data of sample period of at least 10 years). These data have been
collected from the Centre for Monitoring Indian Economy database where the researchers had
institutional access. The National Stock Exchange of India index NIFTY INDEX 50 was
considered as a proxy for computing the overall market return.

S. No. Announcement Sub classification Description of regulatory announcements

1 Price Price positive (PRICE POSt) Expected to increase the price of telecom services
Price negative (PRICR NEGt) Expected to decrease the price of telecom services
2 Service Service positive (SERVICE POSt) Expected to increase quality of service and range
of services
Service negative (SERVICE NEGt) Expected to decrease quality of service and range
of services
3 Competition Competition positive (COMP POSt) Expected to increase competition among telecom Table II.
operators Classification of
Competition negative (COMP POSt) Expected to decrease competition among telecom regulatory
operators announcements
IJOEM The third data required for this study were risk-free rate return. The risk-free rate considered
13,5 was the 91-day Treasury bill (T-bill) rate, which was collected from the Reserve Bank of India
website (RBI, 2015). The fourth data required for this study were that related to the Indian
general elections during our studied sample period from January 1999 to June 2015. This was
collected from the Election Commission of India website (ECI, 2015).

1402 3.1 Descriptive statistics

The descriptive statistics of the data considered in the present study has been presented in
Table III. It deliberates the number of observations, mean, standard deviation, minimum
and maximum values of the stock return and market return.

3.2 Research methods

This study uses the two step procedure for finding the effect of regulation on systemic risk.
The two steps are:
(1) The first step was to estimate systemic risk (β) using Kalman filter along with CAPM
(Barcelos and da Silveira Bueno, 2010; Chalmeau, 2012; Messis and Zapranis, 2016). The β
values were not expected to be constant due to the regulatory and policy announcements
during the study period, which is confirmed by applying the stationarity tests.
(2) The second step was the use of event study methodology for finding the effect of
regulatory announcements on daily β (Binder, 1985; Chalmeau, 2012). Seemingly,
uncorrelated regression (SUR) was used to regress daily β against regulatory
announcements. SUR accommodates the error term of one stock possibly correlating
with the error terms of other stocks (Zellner, 1963).
3.2.1 Research method first step. In this section, stock return calculation and the estimation
of daily β through Kalman filter were being explained.
3.2.1.1 Stock return calculation. For equities and market index, the return was calculated as:


P i;t
Ri;t ¼ ln ; (1)
P i;t1
where Pi,t is the price of the equity i at time t and Pi,t−1 is the price of the equity i at time t−1.
For risk-free rate, 91days T-bill rate was the bond yield equivalent which gives annual
return. Risk-free return at time t is Rf,t.
3.2.1.2 Kalman filter equations. Kalman filter approach was used to estimate the
parameters of the state equation of the model Γi,t (Harvey, 1982).
The difference between the best estimation of the state given the information up to t −1
and the result obtained at time t is as follows:
d
ei;t ¼ Y i;t Y i;t ;

¼ Y i;t Z i;t Gd
i;t1 ;

S. No. Variable No. Obs. Mean SD Min Max

1 Tata Return 4,101 −0.0001 0.318 −0.457 0.168

2 MTNL Return 4,109 −0.0008 0.031 −0.211 0.181
Table III. 3 Airtel Return 2,095 −0.0002 0.017 −0.108 0.094
Descriptive statistics 4 Market Return 4,109 0.0004 0.0158 −0.131 0.163
where Γi,t−1 is the estimator at time t given the information up to time t−1 and is called the Indian telecom
prediction error. sector
Consider ai,t as the optimal estimator of Γi,t based on all of the information at time t. Then,
the estimator could be written as ai,t ¼ Ei,t (Γi,t), i.e. the conditional expectation of the state
variables up to time t for stocks i ¼ 1, 2, …, n. The covariance of the estimators denoted as
Pi,t is defined by Pi,t ¼ Ei,t[(ai,t−Γi,t)(ai,t−Γi,t)′]. Therefore, the optimal estimator of Γi,t based
on all the observations at time t−1 could be denoted by ai,t ¼ Ei,t−1 (Γi,t) and consequently 1403
the covariance of this estimator is defined by Pi,t/t−1.
The Kalman filtering consists of the following recursive set of equations for stocks i ¼ 1,
2, …, n:
• at,/t−1 ¼ Øat−1 (State prediction).
• Pt/t−1 ¼ ØPt−1Ø' + (RQ)' (Prediction dispersion).
• yt/t−1 ¼ zt at/t−1.
• ηt ¼ yt - yt/t−1 (prediction error).
0
• F t ¼ zt P t=t1 zt þH (Error dispersion).
0
1
• Gt ¼ P t=t1 zt F t (Kalman gain).
0
1
• at ¼ at=t1 þP t=t1 zt F t Zt (State estimate).
• Pt ¼ (It−Ptzt) Pt/t−1 (Estimate dispersion).
3.2.1.3 Estimation of β through Kalman filter. In this section, β are estimated using a modified
version of the CAPM model using Kalman Filter (Harvey, 1982; Wells, 1996; Punales, 2011).
The model has been described as follows.
AR (1) (Auto Regressive One) structure of CAPM for the evaluation of β is given as:
   
Ri;t Rf ;t ¼ ai;t þbi;t Rm;t Rf ;t þei;t
ai;t ¼ M 2i ai;t1 þ oi;t
bi;t ¼ M 1i bi;t1 þvi;t ; (2)
where Ri,t is the return of ith stock at time t; Rf,t the risk-free rate return at time t; αi,t the
intercept of ith stock at time t; βi,t the systemic risk of ith stock at time t; εi,t the error term of
ith stock at time t; ωi,t the intercept error term of ith stock at time t; vi,t the systemic risk error
term of ith stock at time t.
In Equation (2), the parameters α and β evolve as a random walk process and ωi,t, vi,t
contains the variation of α and β due to new information, which is confirmed by unit root or
stationarity test.
The Kalman filter algorithm uses OLS (ordinary least square) regressed values as initial
conditions that have been represented as:
OLS
bi;t ¼ b^
OLS
ai;t ¼ a^ OLS i;t ; s2e ¼ sb2 e
i;t ;
   
OLS
s2ot ¼ var aOLS
i;t and s 2
vt ¼ var bi;t

M 2i ¼ 1 and M 1i ¼ 1; (3)
where s2e is the variance of CAPM regression equation; s2ot the variance of αi,t and s2vt the
variance of βi,t; M1i the coefficient of βi,t−1 and M2i the coefficient of αi,t−1. For each of the
stock, the Kalman filter model is optimized to estimate daily β for studied sample period.
IJOEM 3.2.2 Research method second step: effect of regulatory announcements on systemic risk
13,5 (β). In this section, authors have demonstrated event study methodology which was used to
find the effect of regulatory announcements on systemic risk of telecom firms. SUR
regression method is used to regress daily β against dummy variables of regulatory
announcements for an event window period.
During the early 1980s, event studies used expected accounting data, especially earnings
1404 data to test the impact of regulation (Beaver, 1968). The use of financial data for measuring
the effect of regulatory announcements has explanatory power as the frequency of data was
more (Schwert, 1981; Binder, 1985, 1998). Therefore, a well-accepted approach to test the
impact of regulatory announcement is by analyzing the stock price data, as the stock price
data are available at more frequencies. Stock returns are used to estimate β value using
CAPM. This is considered as more appropriate compared to examining the impact using
accounting information. The following are the reasons (Binder, 1985):
• Stock price data are more accurate than accounting data in event study analysis.
• Stock data can provide more number of observations than accounting data in a given
period of time.
• Accounting data communicate only current earnings, whereas stock data relate to
future earnings.
Thus, measuring the effect of regulation on systemic risk was found to be effective with
stock price data (Schwert, 1981; Binder, 1985; Antoniou and Pescetto, 1997). This study
applied event study methodology on stock price data to analyze the effect of changes in
regulation (regulatory announcements) on systemic risk of telecom firms.
Event study methodology was adopted to capture the qualitative nature of regulatory
announcements (Brown and Warner, 1985). To capture the qualitative nature of regulatory
announcements, dummy variables of an event window of 7 days (the day of the event, three
days after the event and three days before the event) were used (Binder, 1985; Paleari and
Redondi, 2005; Barcelos and da Silveira Bueno, 2010). Its value was equal to “1” during event
window period, and otherwise “0.” The event window was not extended beyond this period,
since it may increase the risk of overlapping with other unknown events which may have
occurred during the time period covered by a larger window (Antoniou and Pescetto, 1997;
Chalmeau, 2012).
The daily β were regressed against dummy variables of regulatory announcements for an
event window of 7 days by SUR regression method. This study also included dummies which
captured the effect of 1999, 2004, 2009 and 2014 Indian general elections. General elections
dummy variable value was “1” over the period from when the election notification was issued
until the day after election results have been declared, and “0” otherwise. The estimation of β
is as follows (Antoniou and Pescetto, 1997; Barcelos and da Silveira Bueno, 2010);

b^ i;t ¼ g0 þg1 COM P POS t þg2 PRI CE NEGt þg3 PRI CE POS t þg4 SERV I CE POS t

þ d1 GEN 1999 þd2 GEN 2004 þd3 GEN 2009 þd4 GEN 2014 þci;t ; (4)
where b^ i;t is the β corresponding to each stock i in time t, estimated in the first step by the
Kalman filter. g0 is a constant coefficient for all firms, g1 ; . . .g4 are the coefficients
corresponding to the respective regulatory announcement groups. δ1, …., δ4 are the
coefficients corresponding to the general elections and gi;t is the error term.
To form a meaningful interpretation of the analysis, referring to Equation (4), this
study has tested the following hypotheses to examine the impact of each category of
regulatory announcements on systemic risk under the assumption that the primary
concern of the regulators was to protect the customers and there is no regulatory capture. Indian telecom
This information may be useful for making policy conclusions on the effect of different sector
announcement groups on systemic risk and the cost of equity capital (Antoniou and
Pescetto, 1997).
H1. Announcements that capture an increase of competition (COPM POS) are also
expected to increase β, and thus increase the cost of equity capital.
1405
H 0 : g1 ¼ 0;

H 1 : g1 40:

H2. Announcements that capture decrease in price of service (PRICE NEG) are also
expected to increase β, and subsequently, the cost of equity capital.
H 0 : g2 ¼ 0;

H 1 : g2 40:

H3. Announcements that capture the increase the price of service (PRICE POS) are also
expected to decrease β, and subsequently, the cost of equity capital.
H 0 : g3 ¼ 0;

H 1 : g3 o0:

H4. Announcements which capture increase in the range of services and quality of
service (SER POS) are also expected to decrease β, and subsequently, the cost of
equity capital.

H 0 : g4 ¼ 0;

H 1 : g4 o0:

4. Analysis and results

In this section, researchers discussed the findings of data analysis. The central theme of the
paper is to compute dynamic systemic risk (β) and find regulatory announcements and
general elections impact on β. Section 4.1 describes time-varying systemic risk and Section
4.2 describes event study analysis with SUR regression and Section 4.3 describes firm
specific regulatory risk premiums.

4.1 Time-varying β
This section deliberates the findings of daily β. For this study, researchers have used
Kalman filter to obtain the daily β. The values of standard deviation of systemic risk (β) of
Tata, MTNL and Airtel are 0.119, 0.090 and 0.171, respectively. As the value standard
deviation of Airtel is higher than Tata and MTNL, one can conclude that Airtel stock is
IJOEM more volatile than the stocks of Tata and MTNL. The descriptive statistics of the obtained
13,5 daily β has been presented in Table AII. The authors also found that β of all stocks are
volatile and non-stationary in nature which was confirmed by stationary test. This
phenomenon is captured and presented in Figure 1.
From Figure 1, it is evident that systemic risk (β) of all firms is time-varying in nature and
would be impacted by regulatory announcements. As variation in systemic risk with respect
1406 to time is not same for all firms, regulatory impact may not be uniform for all the firms.
To check the robustness of the non-stationary β, ADF test (stationary test) has been
performed on daily β of the stocks. The results and findings of the ADF test have been
presented in Table IV. The finding column of the Table IV depicts that time series under test
is non-stationary.

4.2 Event study analysis with SUR regression

This section describes the results of SUR regression conducted upon systemic risk against
regulatory announcements and general elections dummy variables. As explained in
Equation (4), systemic risk (β) is dependent variable and dummy variables of regulatory
announcements (in an event window of 7 days) and general elections were independent
variables. The regression analysis of the three telecom firms has been depicted in Table V.
The correlation matrix of the dependent and independent variables among the three telecom
firms has been presented in Table AIII (Table V ).

Time series 
1.4
MTNL 
1.3 TATA 
Airtel 

1.2

1.1
Daily 

0.9

0.8

0.7

Figure 1.
Plot of daily β for 0.6
1998 2000 2002 2004 2006 2008 2010 2012 2014 2016
telecom stocks
Time Period

S. No. Variable z-statistic p-value p-critical Finding

1 Tata β −1.96 0.59 0.05 Non-Stationary

Table IV. 2 MTNL β −2.51 0.36 0.05 Non-Stationary
Results of ADF test 3 Airtel β −1.72 0.69 0.05 Non-Stationary
 Tata Airtel MTNL
Variables Co-efficient SE p-value Co-efficient SE p-value Co-efficient SE p-value

Intercept 1.038 0.002 0.000*** 1.032 0.003 0.000*** 1.013 0.002 0.000***
(PRICE POSt) −0.034 0.015 0.001** −0.056 0.014 0.000*** −0.022 0.008 0.004**
(PRICE NEGt) 0.017 0.010 0.092 −0.102 0.013 0.000*** 0.019 0.008 0.01*
(SERVICE POSt) −0.044 0.008 0.000*** 0.125 0.010 0.000*** 0.067 0.006 0.000***
(COMP POSt) 0.068 0.010 0.000*** −0.077 0.013 0.000*** −0.001 0.007 0.925
GEN1999 −0.073 0.019 0.000*** Data not available 0.014 0.014 0.300
GEN2004 −0.237 0.021 0.000*** −0.183 0.028 0.000*** −0.059 0.016 0.000***
GEN2009 0.066 0.021 0.002** −0.092 0.026 0.000*** −0.311 0.0157 0.000***
GEN2014 −0.10 0.020 0.000*** 0.104 0.026 0.000*** 0.040 0.015 0.005**
Notes: Significance codes of P-Critical: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’
sector
Indian telecom

1407

Table V.
Results of SUR –
seven-day window
IJOEM 4.3 Regulatory risk premiums of telecom firms
13,5 This section deliberates the regulatory risk premiums of individual telecom firm. From the
results of regression analysis of this study, it has been found that regulatory risk exists in Indian
telecom firms. It was also found that different categories of regulatory announcements impacted
systemic risk differently, that impact was either increased systemic risk or decreased systemic
risk based on the kind of regulatory announcement. The impact of regulatory announcement
1408 was not always uniform. It is dependent on the type of regulatory announcement and also
varied from firm to firm that demonstrated inter-firm variations. The firm-wise regulatory
premiums are tabulated in Table VI–VIII.

Coefficient
Aggregate
Regulatory Increased Decreased Neutral to Regulatory regulatory
S. No. announcements systemic risk systemic risk systemic risk premium premium

1 Price-positive | −0.034 0.008

2 Price-negative | 0.017
3 Service-positive | −0.043
4 Competition-positive | 0.068
5 General elections 1999 | −0.073 −0.086
Table VI. 6 General elections 2004 | −0.237
Regulatory risk 7 General elections 2009 | 0.066
premium for Tata 8 General elections 2014 | −0.100

Coefficient
Aggregate
Regulatory Increased Decreased Neutral to Regulatory regulatory
S. No. announcements systemic risk systemic risk systemic risk premium premium

1 Price-positive | −0.056 −0.110

2 Price-negative | −0.102
3 Service-positive | 0.125
4 Competition-positive | −0.077
5 General elections 1999 Data not available –
Table VII. 6 General elections 2004 | −0.184 −0.057
Regulatory risk 7 General elections 2009 | −0.092
premium for Airtel 8 General elections 2014 | 0.104

Coefficient
Aggregate
Regulatory Increased Decreased Neutral to Regulatory regulatory
S. No. announcements systemic risk systemic risk systemic risk premium premium

1 Price-positive | −0.022 0.064

2 Price-negative | 0.019
3 Service-positive | 0.067
4 Competition-positive | Nil
5 General elections 1999 | 0.014 −0.079
Table VIII. 6 General elections 2004 | −0.059
Regulatory risk 7 General elections 2009 | −0.311
premium for MTNL 8 General elections 2014 | 0.040
It was found that for Tata, alternate hypothesis (H1) of H1 has been accepted that denotes Indian telecom
competitive-positive regulatory announcements have increased systemic risk with a sector
regulatory premium of 0.068. Also, alternate hypothesis (H1) of H2 has been accepted that
describes that price-negative regulatory announcements increased systemic risk with a
regulatory premium of 0.017. Similarly, alternate hypothesis (H1) of H3 has been accepted
that denotes price-positive regulatory announcements decreased systemic risk with
regulatory premium of −0.034. Finally, alternate hypothesis (H1) of H4 has also been 1409
accepted that indicates service-positive regulatory announcements decreased systemic risk
with a regulatory premium of −0.044. Thus, during studied sample period, the overall
aggregate regulatory premium of Tata is 0.008. General elections 1999, 2004 and 2014 have
also shown impact on systemic risk by decreasing it with risk premiums of −0.073, −0.237
and −0.100, respectively. General election 2009 has also shown impact on systemic risk by
increasing it with a risk premium of 0.066. Hence, the average general elections premium for
Tata is −0.086.
It was found that for Airtel, H1 has been rejected that indicates competition-positive
regulatory announcements decreased systemic risk with regulatory premium of −0.077. H2
has also been rejected that denotes price-negative regulatory announcements decreased
systemic risk with regulatory premium of −0.102. An alternate hypothesis (H1) of H3 has
been accepted that denotes price-positive regulatory announcements decreased systemic
risk with regulatory premium of −0.056. Finally, H4 has been rejected that signifies service-
positive regulatory announcements increased systemic risk with regulatory premium of
0.125. Hence, during studied sample period, the overall aggregate regulatory premium of
Airtel is −0.110. General elections 2004 and 2009 have shown impact on systemic risk by
decreasing it with risk premiums of −0.184 and −0.092, respectively. General election 2014
has also shown impact on systemic risk by increasing it with a risk premium of 0.104. Hence,
average general elections premium for Airtel is −0.057.
It was found that for MTNL, the null hypothesis (H0) of H1 has been accepted that
indicates competition-positive regulatory announcements did not show any impact on
systemic risk. Moreover, alternate hypothesis (H1) of H2 has been accepted that denotes
price-negative regulatory announcements increased systemic risk with a regulatory
premium of 0.019. Also, alternate hypothesis (H1) of H3 has been accepted that price-
positive regulatory announcements decreased systemic risk with a regulatory premium
of −0.022. On the contrary, H4 has been rejected that denotes service-positive regulatory
announcements increased systemic risk with a regulatory premium of 0.067. Thus, during
the studied sample period, the aggregate regulatory premium of MTNL is 0.064. General
elections 2004 and 2009 have shown impact on systemic risk by decreasing it with risk
premiums of −0.060 and −0.311, respectively. General elections 1999 and 2014 have also
shown impact on systemic risk by increasing it with risk premiums of 0.014 and 0.040,
respectively. Hence, average general elections risk premium is −0.079.

5. Discussion
In this section, the authors have deliberated on the findings of the study with extant
literature. This study found that systemic risk (β) is not constant over the studied sample
period as it is time varying in nature, which indicated that developing countries stock
markets are volatile and dynamic in nature. This was in line with the study finding of Das
and Barai (2015) that Indian stock markets are volatile in nature.
It was reflected from the study that price-positive regulatory announcements against
systemic risk (β) were found to be significant for all the three telecom firms in an event
window of 7 days, with decreased β and decreased cost of equity capital. Clearly, such
regulations supported Peltzman (1976) buffer hypothesis (regulation buffers demand and
cost shocks and reduces risk). On the other hand, price-negative regulatory announcements
IJOEM against systemic risk were found to be significant for all the three telecom firms; however,
13,5 these announcements increased systemic risk, thereby increased the cost of equity capital
for Tata and MTNL stocks, which were supporting reinforcing effect ( Joskow and
MacAvoy, 1975) (regulation increases systemic risk due to regulatory lag). On the contrary,
these announcements decreased systemic risk thereby decreased the cost of equity for Airtel
stock, which were supporting Peltzman (1976) buffer hypothesis.
1410 Competitive-positive regulatory announcements against systemic risk were found to be
significant for Tata and Airtel. For Tata, the β value and hence, the cost of equity capital
increased. On the contrary, for Airtel, the β value decreased, thereby decreasing the cost of
equity capital. In the case of MTNL, significance levels were not attained for
competition-positive regulatory announcements; this may be explained by the prevalent
faith in the market that MTNL being a PSU would be immune to the impact of
competition-positive regulatory announcements. Hence, such regulatory announcements
were also found to support the buffering hypothesis (Peltzman, 1976) as well as the
reinforcing effect proposed by Joskow and MacAvoy (1975), depending upon factors such
as internal strategies and stakeholder structure of individual firms.
The study also reported that service-positive regulatory announcements against
systemic risk were found to be significant for all the three telecom firms. In the case of Tata,
these announcements decreased systemic risk and thereby decreased the cost of equity,
which were supporting Peltzman (1976) buffer hypothesis. On the contrary, these
announcements increased systemic risk thereby increased the cost of equity for Airtel &
MTNL, which were supporting reinforcing effect ( Joskow and MacAvoy, 1975).
Therefore, one can argue that regulatory announcements of all three categories have
either supported Peltzman’s (1976) buffering hypothesis (regulation buffers demand and
cost shocks and reduces systemic risk) or supported reinforcing effect (increases systemic
risk due to regulatory lag) ( Joskow and MacAvoy, 1975), depending on pertinent issues like
internal strategies such as defensive, reactive, anticipatory or proactive strategies (Oliver
and Holzinger, 2008), firm’s other circumstances and stakeholder structure.
It was also indicated from the study that general elections of 2004, 2009 and 2014 have
shown a significant impact for all the three telecom firms (Tata, Airtel and MTNL). This is in
line with the study finding of Antoniou and Pescetto (1997) and Strausz (2017) that elections
have induced regulatory risk through impacting systemic risk. This indicates that Indian
telecom stocks are sensitive to the political scenario of the nation.
This study findings suggested that firm’s regulatory risk (regulatory premium) level is
not same for all telecom firms even though the category of regulatory announcements is
same. This finding is in line with the finding of Buckland and Fraser (2000). One of the
reasons for the varying level of regulatory risk among the same sector’ firms could be that
the risk absorbing capacity and risk mitigating strategies are different from firm to firm
though they operate in the same sector due to their firm-level strategies such as defensive
strategy, reactive strategy, anticipatory strategy and proactive strategy (Oliver and
Holzinger, 2008).

5.1 Research implications

Over the last three decades, financial literature confirms that regulatory risk exists in
developed countries as well as in some of the developing countries. This study has tested
and established that regulatory risk does exist in telecom sector of India. Consequently, this
study also found that systemic risk (β) is not constant over the studied sample period, and is
time varying in nature, which indicates that stock markets of developing countries are
volatile and dynamic. This study also established that Indian telecom sector stocks are
sensitive to the national political scenario since general elections had shown a significant
impact on systemic risk (β) for all the three telecom firms.
 This research pointed out that the impact of regulation (regulatory announcements) on Indian telecom
systemic risk is either positive by increasing systemic risk which supported reinforcing sector
effect ( Joskow and MacAvoy, 1975) or negative by decreasing systemic risk which
supported Peltzman’s (1976) buffer hypothesis. This study also found the aggregate
regulatory premiums and individual regulatory premiums of different type of regulatory
announcements, namely price, quality of service and competition among operators for all
the three telecom firms. 1411
Overall, this study has concluded that regulation has an impact on systemic risk which
led to an effect on the cost of equity capital. Therefore, regulatory risk existed in telecom
firms of India. However, the impact of regulation and regulatory premiums is not uniform
across firms though they operate in the same sector. Also, the firm’s regulatory risk level is
not same for all telecom firms even though the categories of regulatory announcements are
same. This may be due to the firm’s own internal capabilities, stakeholder structure and
firm-level decision to adopt defensive, reactive, anticipatory or proactive strategies (Oliver
and Holzinger, 2008). Hence, the analysis of regulatory impact on systemic risk is firm
specific, not a sector specific. Moreover, regulatory risk premium is either positive or
negative related to different categories of regulatory announcements for telecom sector
firms. Thus, this study contributes to theory by testing Peltzman’s (1976) buffering
hypothesis in the context of Indian telecom sector.

5.2 Practice implications

The findings of this study suggest that regulation is definitely impacting the performance of
telecom firms as it affects systemic risk, thereby impacting the cost of equity capital. The
findings of this study will be useful for regulators in framing new policies by considering
the interests of investors as well as consumers, as regulators’ actions assuredly affect the
firm’s systemic risk. Investors can use the findings to forecast their systemic risk levels
(increase or decrease of systemic risk) by estimating regulatory premiums as and when
there is an expected change in regulation. The findings would also be useful for investors
and senior management of telecom firms to anticipate their risk level in case of an expected
policy & regulatory amendment.

5.3 Limitations & future research

The study has been conducted only for telecom firms of India; hence, there is a scope of
extending it to other regulated sector firms of India such as power sector, oil &gas sector
and aviation sector to ascertain the existence of regulatory risk herein. Further, one can
estimate regulatory premiums in the firms of these sectors as well. Further, the studies can
also be undertaken in other emerging economies like Russia and China. This would help in
conducting a comparative analysis of regulated industries in emerging economies. The
present study only considered regulatory announcements made by TRAI. The study did not
consider any other sources of information like daily newspaper or magazines. Also, the
present study adopted event study methodology, which would be useful only if the
announcements contain unanticipated information (Binder, 1985).

References
Angeles, J.E.L. (2017), “Assessing impact of regulatory risk on shareholders’ wealth”, Applied
Economics, Vol. 49 No. 2, pp. 202-212.
Antoniou, A. and Pescetto, G.M. (1997), “The effect of regulatory announcements on the cost of equity
capital of British Telecom”, Journal of Business Finance & Accounting, Vol. 24 No. 1, pp. 1-26.
IJOEM Barcelos, L. and da Silveira Bueno, R. (2010), “Regulatory risk in the securities markets: a CAPM
13,5 model approach to regulated sectors in Brazil”, 10th Meeting of the Brazilian Finance Society,
Sao Paulo.
Beaver, W.H. (1968), “The information content of annual earnings announcements”, Journal of
Accounting Research, Vol. 6, pp. 67-92.
Ben-Ami, Z. (2015), “Regulatory risk-does the national energy regulator of South Africa correctly price
1412 the cost of equity”, doctoral dissertation, University of Pretoria.
Binder, J. (1998), “The event study methodology since 1969”, Review of Quantitative Finance and
Accounting, Vol. 11 No. 2, pp. 111-137.
Binder, J.J. (1985), “Measuring the effects of regulation with stock price data”, The RAND Journal of
Economics, Vol. 16 No. 2, pp. 167-183.
Boubakri, N., Guedhami, O., Mishra, D. and Saffar, W. (2012), “Political connections and the cost of
equity capital”, Journal of Corporate Finance, Vol. 18 No. 3, pp. 541-559.
Brito, J. and Dudley, S.E. (2012), Regulation: A Primer, Mercatus Center, George Mason University,
Washington, DC, available at: http://mercatus.org/sites/default/files/RegulatoryPrimer_
DudleyBrito_0.pdf (accessed March 3, 2015).
Brown, S.J. and Warner, J.B. (1985), “Using daily stock returns: the case of event studies”, Journal of
Financial Economics, Vol. 14 No. 1, pp. 3-31.
Buckland, R. and Fraser, P. (2000), “Risk and returns sensitivity in UK electricity utilities,
1990-1999”, Aberdeen Papers in Accountancy, Finance & Management No. 00-1, AB24
3FX, Aberdeen.
Camacho, F.T. and Menezes, F.M. (2013), “The impact of price regulation on the cost of capital”, Annals
of Public and Cooperative Economics, Vol. 84 No. 2, pp. 139-158.
Chalmeau, O. (2012), “Impact of regulation on the risk and returns of French telecommunication
operators”, 23rd European Regional ITS Conference (No. 60343). International
Telecommunications Society (ITS), Vienna, June.
Das, S. and Barai, P. (2015), “Time-varying industry beta in Indian stock market and forecasting
errors”, International Journal of Emerging Markets, Vol. 10 No. 3, pp. 521-534.
Davidson, W.N., Rangan, N. and Rosenstein, S. (1997), “Regulation and systemic risk in the
electric utility industry: a test of the buffering hypothesis”, Financial Review, Vol. 32 No. 1,
pp. 163-184.
ECI (2015), April 14, available at: https://eci.nic.in/
Elton, E.J. and Gruber, M.J. (1971), “Valuation and the cost of capital for regulated industries”,
The Journal of Finance, Vol. 26 No. 3, pp. 661-670.
Ergas, H. and Small, J. (2001), “Price caps and rate of return regulation”, Network Economics
Consulting Group, Kingston.
Gaggero, A.A. (2007), “Regulatory risk in the utilities industry: an empirical study of the English-
speaking countries”, Utilities Policy, Vol. 15 No. 3, pp. 191-205.
Gaggero, A.A. (2012), “Regulation and risk: a cross-country survey of regulated companies”, Bulletin of
Economic Research, Vol. 64 No. 2, pp. 226-238.
Ghosh, R. and Kathuria, V. (2016), “The effect of regulatory governance on efficiency of thermal power
generation in India: a stochastic frontier analysis”, Energy Policy, Vol. 89, pp. 11-24.
Grayburn, J., Hern, R. and Lay, H. (2002), “A report for the national audit office on regulatory risk”,
Pipe and Wires, National Audit Office.
Harvey, A.C. (1982), “The Kalman filter and its applications in econometrics and time series analysis”,
Methods of Operations Research, Vol. 44 No. 1, pp. 3-18.
Haucap, J. (2003), “The economics of mobile telephone regulation”, discussion paper, University of the
Federal Armed Forces Hamburg, Hamburg.
Johanning, L. and Ruhle, E.O. (2004), Assessment of Major Regulatory Decisions in France: An Event Indian telecom
Study for France Telecom by Hermann-Josef Piepenbrock, Fabian Schuster and Ernst-Olav sector
Ruhle, Lohmar and Cologne, pp. 143-164.
Joskow, P.L. and MacAvoy, P.W. (1975), “Regulation and the financial condition of the electric power
companies in the 1970s”, The American Economic Review, Vol. 65 No. 2, pp. 295-301.
Joskow, P.L. and Rose, N.L. (1989), “The effects of economic regulation”, Handbook of Industrial
Organization, Vol. 2, pp. 1449-1506. 1413
Knieps, G. and Weiß, H. (2007), “Reduction of regulatory risk: a network economic approach”, Institut
für Verkehrswissenschaft und Regionalpolitik No. 117, Universität Freiburg Platz der Alten
Synagoge, September.
Lintner, J. (1965), “The valuation of risk assets and the selection of risky investments in stock portfolios
and capital budgets”, The Review of Economics and Statistics, Vol. 47 No. 1, pp. 13-37.
Maegli, M. and Jaag, C. (2009), “Regulatory risk and uncertainty in British postal regulation-a
governance perspective”, Second Annual Conference on Competition and Regulation in Network
Industries CRNI 2009, (No. EPFL-CONF-152519).
Messis, P. and Zapranis, A. (2016), “Forecasting time-varying daily betas: a new nonlinearapproach”,
Managerial Finance, Vol. 42 No. 2, pp. 54-73.
Moore, A., Straub, S. and Dethier, J.J. (2014), “Regulation, renegotiation and capital structure: theory
and evidence from Latin American transport concessions”, Journal of Regulatory Economics,
Vol. 45 No. 2, pp. 209-232.
Morin, R.A. and Hillman, L.T. (1994), “Regulatory finance: utilities’ cost of capital”, Public Utilities
Reports, Incorporated, Arlington, TX, pp. 409-411.
Mossin, J. (1966), “Equilibrium in a capital asset market”, Econometrica, Vol. 34 No. 4, pp. 768-783.
Nwaeze, E.T. (2000), “Deregulation of the electric power industry: the earnings, risk, and return effects”,
Journal of Regulatory Economics, Vol. 17 No. 1, pp. 49-67.
Oliver, C. and Holzinger, I. (2008), “The effectiveness of strategic political management: a dynamic
capabilities framework”, Academy of Management Review, Vol. 33 No. 2, pp. 496-520.
Paleari, S. and Redondi, R. (2005), “Regulation effects on company beta components”, Bulletin of
Economic Research, Vol. 57 No. 4, pp. 317-346.
Pedell, B. (2006), Regulatory Risk and The Cost of Capital: Determinants and Implications for Rate
Regulation, Vol. 3823, Springer Science & Business Media, New York, NY.
Peltzman, S. (1976), “Toward a more general theory of regulation”, The Journal of Law and Economics,
Vol. 19 No. 2, pp. 211-240.
Pham, H.N., Ramiah, V., Moosa, I. and Nguyen, J.H. (2017), “The effects of regulatory announcements
on risk and return: the Vietnamese experience”, Pacific Accounting Review, Vol. 29 No. 2,
pp. 152-170.
Posner, R. (1974), “Theories of economic regulation”, Bell Journal of Economics and Management
Science, Vol. 5 No. 2, pp. 335-358.
Punales, A.G. (2011), “Time-varying coefficient models and the Kalman filter: applications to hedge
funds”, masters dissertation, Ryerson University, Toronto, available at: http://digitalcommons.
ryerson.ca/dissertations/1657 (accessed April 26, 2015).
RBI (2015), April 15, available at: www.rbi.org.in/
Riddick, L.A. (1992), “The effects of regulation on stochastic systemic risk”, Journal of Regulatory
Economics, Vol. 4 No. 2, pp. 139-157.
Robinson, T.A. and Taylor, M.P. (1998), “The effects of regulation and regulatory risk in the UK
electricity distribution industry”, Annals of Public and Cooperative Economics, Vol. 69 No. 3,
pp. 331-346.
Ross, S.A. (1976), “The arbitrage theory of capital asset pricing”, Journal of Economic Theory, Vol. 13
No. 3, pp. 341-360.
IJOEM Schiereck, D. and Trillig, J. (2014), “Regulatory changes and the volatility of stock returns – the
13,5 German solar energy sector”, International Journal of Energy Sector Management, Vol. 8 No. 2,
pp. 160-177.
Schwert, G.W. (1981), “Using financial data to measure effects of regulation”, Journal of Law and
Economics, Vol. 24 No. 1, pp. 121-158.
Sharpe, W.F. (1964), “Capital asset prices: a theory of market equilibrium under conditions of risk*”,
1414 The Journal of Finance, Vol. 19 No. 3, pp. 425-442.
Stigler, G.J. (1971), “The theory of economic regulation”, The Bell Journal of Economics and
Management Science, Vol. 2 No. 1, pp. 3-21.
Strausz, R. (2017), “Politically induced regulatory risk and independent regulatory agencies”,
International Journal of Industrial Organization, Vol. 54, pp. 215-238.
Tapia, J. (2012), “The ‘duty to finance’, the cost of capital and the capital structure of regulated utilities:
lessons from the UK”, Utilities Policy, Vol. 22, pp. 8-21.
TRAI (2015), June 2, available at: www.trai.gov.in
TRAI (2016a), January 15, available at: www.trai.gov.in
TRAI (2016b), January 15, available at: www.trai.gov.in/sites/default/files/NCAER-Report08june12.pdf
Trillig, J. (2012), “Regulatory changes and market reactions – the European renewable energy
market”, International Journal of Entrepreneurship and Small Business 9, Vol. 15 No. 1,
pp. 116-129.
Wells, C. (1996), The Kalman Filter in Finance (Advanced Studies in Theoretical and Applied
Econometrics, Vol. 32, Kluwer Academic Publishers, Dordrecht.
Wright, S., Mason, R. and Miles, D. (2003), “A study into certain aspects of the cost of capital for the
regulated utilities”, Report commissioned by the United Kingdom economic regulators and the
Office of Fair Trading, London, available at: www.ofcom.org.uk/static/archive/Oftel/publications/
pricing/2003/capt0203.pdf (accessed April 18, 2013).
Zellner, A. (1963), “Estimators for seemingly unrelated regression equations: Some exact
finite sample results”, Journal of the American Statistical Association, Vol. 58 No. 304,
pp. 977-992.
Appendix Indian telecom
sector

Year Regulatory announcements

1999 National telecom policy

Telecom interconnect usage charges 1415
2000 National long-distance service opened for competition
BSNL and MTNL permitted to enter as third cellular operator in their respective circles
2001 Fixed service providers permitted to provide limited mobility in the form of Wireless in Local Loop on a
restricted basis
TRAI issued its first Interconnection Regulation for Basic and Cellular services
2002 International long-distance service opened up to competition
Tariffs for national roaming service brought under regulation. Ceiling tariffs notified
Universal Service Obligation Fund (USOF) established
2003 Basic service tariffs except rural fixed-line forborne by TRAI
Government approves intra-circle mergers
2004 Broadband Policy 2004 formulated
2005 FDI limit increased from 49 to 74 per cent
TRAI announces new Access Deficit Charges (ADC) regime
2006 DoT announces new criteria for additional spectrum
TRAI issues recommendations on number portability and NGN
2007 TRAI orders reduction in tariff for national roaming services
Domestic Leased Circuits Regulations, 2007 issued
2008 Access Deficit Charges (ADC) abolished
2009 Interconnect usage charges reduced
DoT releases guidelines on Voice over Internet Protocol (VoIP)
2010 3G and Broadband Wireless Access (BWA) spectrum auctions completed
2011 Draft National Telecom Policy 2011 released
2012 Intelligent network services in multi operator and multi network scenario
2013 Mobile number portability Table AI.
Short message services termination charges Major regulatory
2014 International calling card services announcements in the
2015 Mobile number portability telecom sector of India

S. No Variable No. Obs. Mean SD Min Max

1 Tata β 4,101 1.023 0.119 0.767 1.361 Table AII.

2 MTNL β 4,109 1.028 0.171 0.777 1.337 Descriptive statistics
3 Airtel β 2,095 1.016 0.090 0.652 1.311 of daily β
IJOEM Correlation matrix
13,5 Price Price Comp Gen Gen Gen Gen
Pos Neg Service Pos Ele Ele Ele Ele
Tata β Airtel β MTNL β Reg Reg Pos Reg Reg 1999 2004 2009 2014

Tata β 1.000
Airtel β −0.374 1.000
1416 MTNL β −0.050 0.149 1.000
Price Pos Reg −0.041 −0.064 −0.032 1.000
Price Neg Reg 0.043 −0.121 0.060 −0.043 1.000
Service Pos Reg −0.097 0.183 0.195 0.043 0.056 1.000
Comp Pos Reg 0.111 −0.103 0.016 0.118 0.102 0.079 1.000
Table AIII. Gen Ele 1999 −0.048 0.000 0.006 0.139 −0.023 −0.027 0.127 1.000
Correlation matrix Gen Ele 2004 −0.192 −0.094 −0.045 −0.027 −0.023 0.044 −0.024 0.000 1.000
of β and regulatory Gen Ele 2009 0.057 −0.055 −0.323 −0.022 −0.023 −0.031 −0.011 0.000 −0.110 1.000
announcements Gen Ele 2014 −0.089 0.084 0.056 −0.025 −0.025 0.036 −0.026 0.000 −0.012 −0.012 1.000

Corresponding author
Sushma Priyadarsini Yalla can be contacted at: priyadarsini.sushma@gmail.com

For instructions on how to order reprints of this article, please visit our website:
www.emeraldgrouppublishing.com/licensing/reprints.htm
Or contact us for further details: permissions@emeraldinsight.com