Changing Nature of the Value Premium in the Indian Stock Market

Abstract

The article aims to test whether the nature of value premium has changed over time in the Indian stock market. Contrary to the existing literature on the Indian market, the analysis suggests that post the global crisis, the value premium has nearly ceased to exist. Using price-to-book (P/B) ratio as the proxy, the findings are robust to employing capital asset pricing model (CAPM) and the Fama and French three-factor (FF) model, and five alternative measures of returns: equal-weighted (whole and trimmed sample) and value-weighted (whole, trimmed and winsorized sample). Chow test suggests that the model parameters have changed over time. While the decrease in premium has been discovered for other countries, there was no such study in the Indian context. This is the first attempt of its kind. The findings have important implications for academia and investors.

Keywords

Value Premium India Chow test Asset Pricing Models Market Anomaly

Introduction

Results that seem to be discrepant with the theories of asset pricing behaviour are called anomalies (Schwert, 2003). If value stocks (stocks with low market price relative to their fundamentals) tend to outperform growth stocks (stocks with high market price relative to their fundamentals), it is called value anomaly. One group of researchers believes that value and other variables are priced factors and hence need to be incorporated in the asset pricing models. Development of new asset pricing models is the result of this school of thought. Another group of researchers believes that the existence of anomalies is a matter of semi-strong form of market inefficiency [1] and hence the anomalies are likely to vary over time with the level of market efficiency. Literature has documented mixed evidence of the nature of anomalies over time for various markets.

In the context of India, while there are studies which have examined the presence of value anomaly, surprisingly, almost all of them have overlooked the examination of how the anomaly has changed over time. Putting differently, studies have examined the implications of using alternative asset pricing models; however, they have overlooked the implications of the changing level of market efficiency over time. To the best of the authors’ knowledge, Mohanty (2002) is the only study to have examined this issue. However, the changes over the past decade, particularly post the global financial crisis remain unaddressed. This article aims to fill this gap.

The rationale for studying India can be justified by the facts that follow: according to the International Monetary Fund’s (IMF) projections in October 2015, the Indian economy, backed by strong macroeconomic fundamentals, is expected to become the fastest-growing major economy in the world (National stock exchange [NSE] website, 2015b). With the increasing global integration of the financial markets, investments in emerging markets are gaining relevance (Lischewski & Voronkova, 2012), and the Indian stock market has distinguished itself at the global level. The popularity of the Indian stock market is reflected by the fact that the foreign portfolio investment in India has risen from USD2.1 billion in the financial year 1996 to USD45.4 billion in the financial year 2015. According to the World Federation of Exchanges (WFE), as at the end of September 2015, two of the Indian stock exchanges ranked eleventh and twelfth globally in terms of domestic market capitalization (NSE website, 2015b). Hence, a take on how the level of efficiency is changing in the Indian stock market is warranted.

Literature Review

The value effect was first documented by Basu (1977) on US data from 1957 to 1971. Following its identification, numerous studies were conducted across countries confirming the presence of the anomaly (e.g., Ball, 1992; Barry, Goldreyer, Lockwood & Rodriguez, 2002; Cakici, Fabozzi & Tan, 2013; Cakici & Tan, 2014; de Groot, Pang & Swinkels, 2012; Fama & French, 1992, 1993, 1998, 2012; Griffin, 2002; Hawawini & Keim, 1998; Patari & Leivo, 2017; Rouwenhorst, 1999).

The discovery encouraged academicians to come up with various probable explanations for the anomaly. Roll (1981) contended that infrequent trading of value stocks led to mis-assessment of their risk by asset pricing models. Peavy III & Goodman (1983) argued that as firms of same industry tend to have similar price-to-book (P/B) ratios, value effect might actually be due to industry performance, and not the ratios. Fama and French (1993, 1995) contended that value was proxy for sensitivity to risk factors and held information related to the profitability of firms. Lakonishok, Shleifer and Vishny (1994) reasoned that investors wrongly over-estimated past performances of value and growth stocks in future, and correction of their over-estimation caused value premium. Wang (2000) argued that the studies suffered from data truncation bias or survival bias according to which only good performing value stocks remained part of the study sample and poor performers got dropped from the sample. Hwang and Rubesam (2013) presented behavioural biases of investors as the reason causing anomaly. Ebrahim, Girma, Shah and Williams (2014) reported that value and growth firms had different economic fundamentals, like investment patterns, and hence justified different premiums.

However, in the words of John F. Kennedy, ‘Change is the law of life’ (John F. Kennedy presidential library and museum website, 1963). There have been mixed evidence of persistence and attenuation of anomalies since their discovery (e.g., Brennan, Chordia & Subrahmanyam, 1998; Cakici & Tan, 2014; Chung, Hsu, Ke, Liao & Chiang, 2016; Fama & French, 2015; Gillan, 1990; Kothari, Shanken & Sloan, 1995; Loughran, 1997; Schwert, 2003; Shum & Tang, 2005). Reasons being cited for attenuation range from increased implementation of strategies to capitalize on anomalous patterns by practitioners (Schwert, 2003) to deregulation of financial markets and enhancement of trading facilities (Chung et al., 2016). Besides, there have been evidence of reverse anomaly effects (Claessens, Dasgupta & Glen, 1995); time-varying behaviour of anomalies (Wu, Liu & Chen, 2016); anomalies varying according to monetary policies of the government (Jensen, Johnson & Mercer, 1997).

While a region-level study is preferred for asset pricing tests, a country-level study is reported to be more suitable for examining the strength of anomalies (Cakici & Tan, 2014). The aim of this article is the latter. There has been mixed evidence about the presence of value anomaly in India. While Claessens et al. (1995) reported insignificant ratio of book-to-market (B/M) premium, de Groot and Verschoor, 2002; Mohanty, 2002 and Kumar and Sehgal, 2004 reported weak evidence of the value effect. Significant effect was reported by Barry et al., 2002; Sehgal, Subramaniam and Morandiere, 2012; Sehgal and Balakrishnan, 2013; Aziz and Ansari, 2014; Ebrahim et al., 2014; Sehgal and Pandey, 2014 and Das and Barai, 2016. Taneja (2010) reported a high correlation between the size and value factors and hence argued that either of the two factors could explain the variability in stock returns.

Research Gaps

Although the literature seems rife with studies on testing the presence of value anomaly and/or testing the performance of asset pricing models, there are certain lacunae. The issue of the changing level of Indian market efficiency with respect to anomalies over the past 15 years remains an under-researched area. While studies have been done on an aggregate basis, the less attended sub-period analysis seems essential to capture the dynamism in the market. Specifically, while the impact of financial crisis has been studied with respect to other phenomena like weak form market efficiency (Jain, Vyas & Roy, 2013), investment strategies (Vardhan, Sinha & Vij, 2015) and price volatility (Goh, Tan, Khor & Ng, 2016), its impact on the Indian stock market with respect to market anomalies has not been addressed. Besides, while robustness has been checked with respect to different measures of value —book-to-market/earnings-to-price/dividends-to-price/cash’’flow-to-price (Sehgal & Balakrishnan, 2013) —robustness to different methods of calculating returns (equal-weighted/value-weighted based on trimmed/winsorized data) remain under-attended (an exception to this is the study by Manjunatha & Mallikarjunappa [2011] who study both equal-weighted and value-weighted returns).

This study aims to address these erstwhile neglected issues.

Objectives, Data and Methodology

Objectives

The overall objective of this article is to test the level of semi-strong form market efficiency of the Indian stock market based on the presence of value anomaly. The objective is further divided into two sub-objectives: the first sub-objective is to test whether the value anomaly exists in the Indian stock market and the second sub-objective is to test whether the level of anomaly has remained unchanged pre and post the global financial crisis. Based on the objectives, the null hypotheses are as follows:

H₀₁: There is no value anomaly in the Indian stock market.

H₀₂: The level of value anomaly has remained unchanged in the Indian stock market post the global financial crisis.

Data

The study sample comprises of the constituent companies of Nifty 500 index. The study period ranges from October 1995 to September 2015 and the date of sample selection is 31 March 2014. Only those firms whose financial year ends in March form part of the study. The rationale behind this choice is that the financial year of the government of India spans from April to March, and nearly all the sample companies follow the same. The Nifty 500 index is the first broad-based benchmark of the Indian capital market. The 500 constituent stocks of the index cover 18 sectors and represent about 95.77 per cent of the free float market capitalization of the stocks listed on the NSE as on 31 March 2015 (NSE website, 2015a). Hence, the study sample is a fair representative of the Indian market. The data are obtained from Ace Equity® database.

Methodology

Every year at September end, companies are sorted on the basis of P/B ratio and divided into 10 univariate sorted portfolios. Monthly returns are calculated on the portfolios from October end of the year till September end of the next year. To maintain consistency, every September end, only those companies form part of the portfolio for which monthly records are available for every month for the following 1 year. Further, cases with negative P/B ratios are excluded (Dash & Mahakud, 2013; Ebrahim et al., 2014; Fama & French, 1993).

Following the practice in literature, P/B ratio is taken as the proxy for value (Barry et al., 2002; Cakici et al., 2013; Fama & French, 1993). Every year, P/B ratio is based on the closing prices of the stocks and number of paid up equity shares as at March end. It is customary in literature to use 6-month lagged values of variables for portfolio creation (Agarwalla, Jacob & Varma, 2013; Das & Barai, 2016). This is to ensure that the variables are known at the time of portfolio creation.

Returns on individual stocks are calculated as simple returns:

R_{t} = \frac{p_{t} - p_{t - 1}}{p_{t - 1}} \times 100 %

(1)

where R_t is the return on a stock at time t, p_t is the price of a stock at time t and p_t–1 is the price of a stock at time t−1.

To test the robustness of the findings to different methods of returns calculations, returns on portfolios are calculated as (a) equal-weighted returns and (b) value-weighted returns, where for every October of year t to September of year t+1, market capitalization as at September end of year t are used as weights. Further, the equal-weighted returns are calculated on (a) the whole sample and (b) trimmed sample (Fabozzi, Focardi & Kolm, 2010). Similarly, the value-weighted returns are calculated on (a) the whole sample; (b) trimmed sample; and (c) winsorized sample (Fabozzi et al., 2010) where the weights are winsorized at 98 per cent [2]. Trimming and winsorization helps to mitigate the influence of extreme observations on the results. Only the results on the whole sample (equal-weighted) have been presented for the purpose of this article.

To test whether the differences in portfolio returns are statistically significant, the returns on extreme portfolios are compared using paired t-test (de Groot & Verschoor, 2002).

To examine whether the magnitude of the relationship between returns and value has changed over time, the study period is divided into two phases: phase I (October 1999 to September 2008) and phase II (October 2008 to September 2015). The rationale for choosing this break point of the sample period is that the global financial crisis began in mid-2008 (NSE website, 2015b). Thus, phase I is depicting the pre-recession phase while phase II is depicting the post-recession phase.

Dividends are not taken into consideration while calculating portfolio returns to match the returns calculated on the market portfolio which are based only on capital gains and losses due to price movements (NSE website, 2015c). This is consistent with other previous studies (Sehgal & Tripathi, 2005; Sehgal & Balakrishnan, 2013).

To test whether the relationship of returns with value survives on risk-adjusted basis, risk adjustment is done based on asset pricing models —the capital asset pricing model (CAPM) and the Fama and French three-factor (FF) model. The two models are mathematically represented as follows:

The CAPM

R_{P t} - R_{F t} = β_{P} (R_{M t} - R_{F t}) + ε_{t}

(2)

where R_Pt is the return on a portfolio P for time t, R_Ft is the risk-free rate during time t, β_P is the degree of sensitivity of P to the market risk premium, (R_Mt – R_Ft) is the market risk premium, defined as the excess of return on market portfolio for time t over the risk-free rate and ε_t is the error term.

The FF model

\begin{matrix} R_{P t} - R_{F t} = α_{P} + β_{P} (R_{M t} - R_{F t}) + γ_{P} (S M B_{t}) + \\ θ_{P} (L M H_{t}) + ε_{t} \end{matrix}

(3)

where α_P is the intercept, γ_P is the degree of sensitivity of P to the size premium, SMB_t is the size premium (it is the difference between returns on portfolio of small stocks and portfolio of big stocks), θ_P is the degree of sensitivity of P to the value premium, LMH_t is the value premium (it is the difference between returns on portfolio of value stocks and portfolio of growth stocks) [3] and R_Pt, R_Ft, β_P, R_Mt – R_Ft, and ε_t are as defined in the previous model.

Nifty 500 is used as the proxy for market portfolio. Implicit yields on 91-day treasury bills issued by the government of India are taken as the proxy for risk-free returns. The yield on the bill auctioned in the first week of every month is taken as the measure of risk-free return for that month.

The small minus big (SMB) and low minus high (LMH) portfolios are constructed in line with the Fama and French (1993) model [4].

The Chow test (Brooks, 2014) is then performed to test the stability of parameters ( α, β, γ, and θ here) of the asset pricing models over time. For this, following steps are followed: First, three set of regressions are run (in the form of asset pricing models) —one for the entire period and then for the two sub-periods. Then, the residual sum of squares (RSS) is calculated for each and the test statistic based on RSS is calculated. Finally, the test statistic is compared with the critical value from the F-distribution, which is F(k, T−2k).

T e s t s t a t i s t i c = \frac{R S S - (R S S_{1} + R S S_{2})}{(R S S_{1} + R S S_{2})} \times \frac{T - 2 k}{k}

(4)

where RSS is residual sum of squares for the whole sample, RSS₁ is residual sum of squares for sub-sample 1, RSS₂ is residual sum of squares for sub-sample 2, T is total number of observations and k is number of parameters in the model.

Analysis

Table 1 summarizes the time-series averages of monthly excess returns on the value-sorted portfolios. Excess returns are actual returns net of risk-free returns. Portfolio P1 contains 10 per cent stocks with the lowest P/B ratios while portfolio P10 contains 10 per cent stocks with the highest P/B ratios.

Table 1.

Average Monthly Excess Returns on Value-sorted Portfolios, 1995–2015

			P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P1–P10
equal-weighted	whole	Whole (1995–2015)	2.62	1.90	2.12	1.92	1.41	2.04	1.41	1.17	1.07	1.04	1.58**
		Phase I (1995–2008)	3.30	2.27	2.76	2.40	1.58	2.51	1.48	1.14	1.03	0.94	2.36***
		Phase II (2008–2015)	1.35	1.21	0.93	1.02	1.08	1.17	1.28	1.24	1.13	1.22	0.12
	trimmed	Whole (1995–2015)	2.48	1.64	2.10	1.84	1.60	1.89	1.43	1.17	1.13	1.05	1.43**
		Phase I (1995–2008)	3.04	1.90	2.71	2.30	1.82	2.36	1.44	1.16	1.10	0.95	2.09**
		Phase II (2008–2015)	1.44	1.17	0.97	0.98	1.21	1.02	1.40	1.18	1.20	1.22	0.22
value-weighted	whole	Whole (1995–2015)	2.32	1.32	2.93	1.44	0.81	1.16	1.01	0.47	0.65	0.96	1.36
		Phase I (1995–2008)	3.24	1.80	4.31	2.01	1.12	1.62	0.99	0.62	0.60	1.03	2.21*
		Phase II (2008–2015)	0.60	0.42	0.38	0.38	0.23	0.29	1.04	0.17	0.73	0.82	−0.22
	trimmed	Whole (1995–2015)	1.89	1.26	2.81	1.28	1.02	1.07	0.94	0.53	0.64	0.88	1.02
		Phase I (1995–2008)	2.56	1.72	4.06	1.83	1.38	1.56	0.90	0.69	0.55	0.85	1.71*
		Phase II (2008–2015)	0.65	0.40	0.51	0.25	0.35	0.15	1.01	0.22	0.79	0.92	−0.27
	winsorized	Whole (1995–2015)	2.32	1.32	2.93	1.43	0.81	1.17	1.02	0.50	0.65	0.96	1.37
		Phase I (1995–2008)	3.25	1.80	4.30	2.00	1.12	1.60	1.00	0.62	0.60	1.03	2.22*
		Phase II (2008–2015)	0.60	0.42	0.38	0.38	0.24	0.36	1.06	0.28	0.74	0.82	−0.22

Source: Authors’ calculations.

Notes: All the figures are in per cent; *** indicates significance at 1 per cent; ** indicates significance at 5 per cent; * indicates significance at 10 per cent level.

The results show while the negative relation between returns and value is evident in phase I and overall period taken together, the relation appears almost constant in phase II. While the difference between returns on extreme portfolios (equal-weighted, whole sample) is 2.36 per cent in phase I, it reduces to 0.12 per cent in phase II. Interestingly, when the value-weighted portfolios are created for phase II, P10 provides slightly higher returns than P1 in all the three cases —whole, trimmed and winsorized samples. The paired t-tests show while the difference was significant in phase I, it is statistically insignificant in phase II. This implies that relationship between returns and P/B ratio has almost disappeared from the Indian market.

On the 10 portfolios formed, CAPM is applied to see whether the value relates with the expected returns in the same manner on risk-adjusted basis. The intercept of the regression, CAPM alpha (also called Jensen’s alpha) will capture the average return on a portfolio in excess of the return according to CAPM (Bodie, Kane & Marcus, 2014). Table 2 summarizes the values of the CAPM alphas.

Table 2.

CAPM (Equal-weighted, Whole Sample) Excess Returns (Jensen’s alpha), Goodness of Fit (adjusted R²), and Test Statistic of Chow Test, 1999–2015

Portfolios	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10
	Jensen’s alpha
Whole	2.30***	1.57***	1.08***	0.99***	1.16**	1.55***	0.87***	0.36	0.47	0.20
1999–2015	(3.37)	(2.85)	(2.68)	(2.73)	(2.11)	(3.08)	(3.32)	(1.20)	(1.60)	(0.81)
Phase I	3.68***	2.49***	1.82***	1.56***	1.84**	2.37***	1.12***	0.28	0.46	−0.10
(1999–2008)	(3.42)	(2.82)	(3.16)	(2.98)	(2.03)	(2.75)	(2.97)	(0.69)	(1.05)	(−0.28)
Phase II	0.53	0.40	0.14	0.27	0.30	0.50*	0.55*	0.47	0.48	0.60**
(2008–2015)	(0.84)	(0.84)	(0.32)	(0.63)	(0.80)	(1.75)	(1.90)	(1.32)	(1.58)	(2.02)
	Adjusted R²
Whole	0.48	0.58	0.68	0.74	0.54	0.61	0.85	0.80	0.77	0.83
Phase I	0.36	0.46	0.60	0.68	0.38	0.50	0.81	0.76	0.70	0.79
Phase II	0.76	0.84	0.85	0.85	0.89	0.91	0.92	0.90	0.90	0.89
	Test statistic - Chow test
Whole	5.36***	5.55***	14.42***	9.27***	6.54***	1.96	12.38***	18.69***	8.77***	7.10***

Source: Authors’ calculations.

Notes: Figures in brackets indicate the t-stats; *** indicates significance at 1 per cent level; ** indicates significance at 5 per cent level; * indicates significance at 10 per cent level.

Similar pattern is evident between returns and P/B ratio as discovered in Table 1. While in phase I, P1 provided statistically significant alpha and P10 did not, the situation changes in phase II when P1 provides insignificant alpha, whereas P10 provides alpha not only higher than P1 but also statistically significant. This evidence is more an indication of a reverse of value anomaly in the Indian market.

The findings suggest that in the post-recession period (that is phase II), value effect in India nearly ceases to exist.

Chow test is now performed to confirm the findings that the dynamics of the market indeed have changed over time. Table 2 summarizes the findings for the same.

The test statistic for the hypothesis of stability of parameters over time gets rejected in all but one case (the exception being P6), implying the CAPM parameters have indeed not been stable over the two phases.

As a final capping, the table presents the measure of the goodness of fit (adjusted R²) of the CAPM to show the contrast in its ability to explain the variability in the portfolio returns in phases I and II.

The table reveals that the explanatory power of the CAPM has increased for the recent period (phase II) across all the portfolios. While the model could explain only 36 per cent variability for P1 in phase I, it could explain 76 per cent of the variability in phase II. For P10, the explanatory power rises from 79 per cent to 89 per cent.

The same set of tests is now performed using the FF model. The results are presented in Table 3. The main findings are summarized below.

Table 3.

FF Model (equal-weighted, whole sample) Excess Returns (intercept), Goodness of Fit (adjusted R²), and Test Statistic of Chow Test, 1999–2015

Portfolios	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10
	Intercept
Whole	0.81	0.23	0.20	0.18	0.59	0.73	0.52**	0.21	0.33	0.19
1999–2015	(1.46)	(0.60)	(0.60)	(0.58)	(1.07)	(1.54)	(2.03)	(0.67)	(1.15)	(0.83)
Phase I	1.44	−0.03	0.47	0.46	0.90	1.40*	0.71*	0.11	0.43	0.15
(1999–2008)	(1.49)	(−0.05)	(0.96)	(0.99)	(0.97)	(1.69)	(1.86)	(0.24)	(0.92)	(0.43)
Phase II	−0.01	0.07	−0.23	−0.13	−0.19	0.27	0.20	−0.01	0.02	0.19
(2008–2015)	(−0.04)	(0.24)	(−0.74)	(−0.44)	(−0.56)	(1.10)	(0.75)	(−0.03)	(0.09)	(0.79)
	Adjusted R²
Whole	0.68	0.82	0.81	0.84	0.58	0.68	0.86	0.80	0.80	0.87
Phase I	0.54	0.77	0.75	0.78	0.42	0.59	0.83	0.76	0.71	0.83
Phase II	0.95	0.95	0.94	0.93	0.92	0.94	0.94	0.92	0.94	0.93
	Test statistic—Chow test
Whole	2.20*	6.69***	10.31***	5.83***	3.89***	1.62	6.61***	12.23***	5.39***	3.52***

Source: Authors’ calculations.

Notes: Figures in brackets indicate the t-stats; *** indicates significance at 1 per cent level; ** indicates significance at 5 per cent level; * indicates significance at 10 per cent level.

None of the phases show a significant difference between the intercepts of P1 and P10. This is quite intuitive and expected because this model incorporates the value premium as an explanatory variable. So, after being incorporated as part of the model, getting no discernible pattern in intercepts is intuitive. What is of interest here is the decrease in the magnitude of the intercepts from phase I to phase II, which is evident from the results. These results are similar to the results obtained using the CAPM as the underlying model, signifying the weakening magnitude of the value premium.

Similar to the results of the CAPM, the Chow test for the FF model also rejects the hypothesis of the stability of the parameters over the two phases for almost all the cases: The parameters are found to change from one phase to the other for 9 out of 10 cases at 10 per cent significance level, eight of which are significant at 1 per cent level of significance. This further reinforces the previous findings of changing dynamics of the Indian market.

For all the portfolios, the explanatory power of the FF model in explaining the variability in the portfolio returns has increased in phase II vis-à-vis phase I of the study period. The goodness of fit increases from 54 per cent to 95 per cent for P1 and from 83 per cent to 93 per cent for P10.

Findings

The above analysis can be summarized as follows.

Value anomaly or value premium seems to have ceased to exist in India over time. While phase I demonstrates value stocks giving higher returns than growth stocks (for different measures of returns and also on a risk-adjusted basis), the evidence ceases to exist for phase II. The latter phase shows a flattening of the relationship between returns and P/B ratio. When the underlying data are trimmed, even the extreme portfolios fail to show any relation. Findings are similar when returns are taken after adjusting for risk, although the extreme value portfolio actually provides lower returns than extreme growth portfolio. The attenuation, however, is not identifiable when the whole study period is taken together. This explains why the pattern remains undiscovered even in recent studies on India like Sehgal and Balakrishnan (2013), Aziz and Ansari (2014), Sehgal and Pandey (2014) and Ebrahim et al. (2014). The findings of this article are more in tune with findings of other countries, such as Indonesia and Taiwan (Ding, Chua & Fetherston, 2005), UK, France, Germany and Italy (Abhyankar, Ho & Zhao, 2009), and Australia and New Zealand (Chung et al., 2016).

The robustness of the finding that the level of anomaly has changed over time is further backed by Chow test, which supports the intuitive implication that parameters of the asset pricing models are not similar in the two phases. This is also revealed by a change in the explanatory capacity of the models in the two phases. That model performances that can vary over time has been documented before by Barclay, Fletcher and Marshall (2010) for emerging markets.

The reason for attenuation of anomalies in India over time could be the increase in informational efficiency of the market. Rejeb and Boughrara (2013) reported that informational efficiency indeed had increased in emerging markets in recent years, which, in turn, was possible because of factors, such as level of development, the degree of liquidity and quality of investment. Another argument is that the premiums evaporate as more and more practitioners know about them and exploit them in their strategies (Schwert, 2003). In the context of India, some developments in the past few years are worth mentioning. Steps taken on educational front include (a) establishment of National Institute of Securities Market to promote education and research on securities market; (b) setting up of Investor Protection and Education Fund; (c) introduction of courses related to financial markets at school level; and (d) conducting of Investor Awareness Programmes. Steps taken on regulatory front include (a) amendment of the Securities Contracts Regulation Act to help development of the market in terms of depth and liquidity; (b) introduction of volatility index to help determine overall market volatility; (c) replacing the Companies Act 1956 by the Companies Act 2013, further strengthening the laws on corporate governance; and (d) introducing liquidity enhancement scheme for illiquid securities. Besides, simplifications and modifications are done in different processes from time to time to make the environment more investor-friendly (e.g., using Permanent Account Number as the only identification number for all transactions in securities market; processing investors’ complaints through web-based system; simplification of trading account opening process); steps are taken to increase investor protection (e.g., stock exchanges sending transaction alerts to investors and system audit of stock brokers). In 2016, the World Bank ranked India eighth globally from the point of view of ‘protecting minority investors’. Measures have also been taken to encourage the flow of savings of investors into the stock market. In addition, with development on the technology front, facilities such as Smart Order Routing have been introduced; norms for algorithmic trading have been amended; circuit-breaker systems have been revised; software used for trading and risk management are periodically tested; and robust Cyber Security and Cyber Resilience framework have been brought in. Last, but not the least, the Foreign Institutional Investors (FII) policy has been amended to relax the investment limits and eligibility conditions, and liberalize the investment instruments accessible to FIIs; this and other policy measures have been taken at an economy level to attract foreign investment. All these factors seem to be contributing to making the Indian market more efficient and moving away from evincing anomalous patterns in returns.

Conclusion

This study tests the change in magnitude of value premium over time in the Indian stock market, on both absolute and risk-adjusted basis. Application of multiple ways for data cleaning (i.e., full sample, trimmed sample and winsorized sample), multiple ways for calculation of returns (i.e., equal-weighted and value-weighted), more than one asset pricing model (i.e., the CAPM and the FF model) and further, employment of Chow test for stability acts as checks for the robustness of the results. The results show that the value premium has almost ceased to exist in recent years. These findings are similar to the findings of studies on other countries. The Indian context, however, has been reported for the first time.

With respect to the first sub-objective of testing the presence of value anomaly in the Indian stock market, the null hypothesis gets rejected. The data exhibit the presence of value anomaly. These findings are similar to the findings of Sehgal and Balakrishnan (2013), Dash and Mahakud (2013), Aziz and Ansari (2014), Sehgal and Pandey (2014) and Ebrahim et al. (2014).

With respect to the second sub-objective of testing the change in level of market anomalies post the financial crisis, the null hypothesis of no change again gets rejected. The value anomaly exhibits change over time: it almost ceases to exist. These findings are in line with the findings of Shum and Tang (2005), Ding et al., (2005), Abhyankar et al., (2009) and Chung et al., (2016).

The decrease in anomalies is an indication of improving market efficiency for the period studied. Steps taken by the Indian government and stock market regulators such as improving awareness and educating intermediaries and investors, simplifying processes, making regulations more robust, and increasing investor protection seem to be the forces behind this development.

The findings have important implications for academia and investors. For the academia, the changing nature of value premium in the Indian stock market implies that the debate on market efficiency is not over yet and more studies are required before a conclusion can be drawn. As already stated, while the findings are similar to those reported for other countries, they serve as fresh evidences for the Indian stock market. For the investors, the findings imply that with the increase in the level of market efficiency, there is a diminishing role of the value anomaly in designing of the investment strategies. In other words, preferring to invest in value stocks (stocks with lower P/B ratios) to growth stocks (stocks with higher P/B ratios) with an anticipation to earn value premium might not result in earning of higher returns. As the premium is not consistent across study periods, basing an investment strategy to exploit this anomaly is not advisable. In words of Horowitz, Loughran and Savin (2000), ‘money managers should not place bets on…any…market anomaly that cannot reliably be shown to work’. Further, the increasing efficiency should make the Indian stock market a more attractive destination for the international investors.

Footnotes

Notes

References

Abhyankar

, Ho

, & Zhao

(2009). International value versus growth: Evidence from stochastic dominance analysis. International Journal of Finance & Economics, 14(3), 222–232.

Agarwalla

S. K.

, Jacob

, & Varma

J. R.

(2013, September). Four factor model in Indian equities market (Working paper No. 2013-09-05). Ahmedabad: Indian Institute of Management.

Aziz

, & Ansari

V. A.

(2014). Size and value premiums in the Indian stock market. Pacific Business Review International, 7(4), 74–80.

Ball

(1992). The earnings-price anomaly. Journal of Accounting and Economics, 15(2–3), 319–345.

Barclay

, Fletcher

, & Marshall

(2010). Pricing emerging market stock returns: An update. Emerging Markets Review, 11(1), 49–61.

Barry

C. B.

, Goldreyer

, Lockwood

, & Rodriguez

(2002). Robustness of size and value effects in emerging equity markets, 1985–2000. Emerging Markets Review, 3(1), 1–30.

Basu

(1977). Investment performance of common stocks in relation to their price-earnings ratios: a test of the efficient market hypothesis. The Journal of Finance, 32(3), 663–682.

Bodie

, Kane

, & Marcus

A. J.

(2014). Investments. New York, NY: McGraw-Hill Education.

Brennan

M. J.

, Chordia

, & Subrahmanyam

(1998). Alternative factor specifications, security characteristics, and the cross-section of expected stock returns. Journal of Financial Economics, 49(3), 345–373.

10.

Brooks

(2014). Introductory econometrics for finance. New York, NY: Cambridge University Press.

11.

Cakici

, & Tan

(2014). Size, value, and momentum in developed country equity returns: Macroeconomic and liquidity exposures. Journal of International Money and Finance, 44(8), 179–209.

12.

Cakici

, Fabozzi

F. J.

, & Tan

(2013). Size, value, and momentum in emerging market stock returns. Emerging Markets Review, 16(September), 46–65.

13.

Chung

, Hsu

, Ke

, Liao

T. L.

, & Chiang

(2016). The weakening value premium in the Australian and New Zealand stock markets. Pacific-Basin Finance Journal, 36(February), 123–133.

14.

Claessens

, Dasgupta

, & Glen

(1995, September). The cross-section of stock returns-evidence from emerging markets (Policy Research Working Paper No. 1505). Washington D.C.: The World Bank.

15.

Das

, & Barai

(2016). Size, value and momentum in stock returns: Evidence from India. Macroeconomics and Finance in Emerging Market Economies, 9(3), 284–302.

16.

Dash

S. R.

, & Mahakud

(2013). Conditional multifactor asset pricing model and market anomalies. Journal of Indian Business Research, 5(4), 271–294.

17.

de Groot

C. G. M.

, & Verschoor

W. F. C.

(2002). Further evidence on Asian stock return behaviour. Emerging Markets Review, 3(2), 179–193.

18.

de Groot

, Pang

, & Swinkels

(2012). The cross-section of stock returns in frontier emerging markets. Journal of Empirical Finance, 19(5), 796–818.

19.

Ding

D. K.

, Chua

J. L.

, & Fetherston

T. A.

(2005). The performance of value and growth portfolios in East Asia before the Asian financial crisis. Pacific-Basin Finance Journal, 13(2), 185–199.

20.

Ebrahim

M. S.

, Girma

, Shah

M. E.

, & Williams

(2014). Rationalizing the value premium in emerging markets. Journal of International Financial Markets, Institutions & Money, 29(March), 51–70.

21.

Fabozzi

F. J.

, Focardi

S. M.

, & Kolm

P. N.

(2010). Quantitative equity investing: Techniques and strategies. New Jersey, NJ: Wiley.

22.

Fama

E. F.

, & French

K. R.

(1992). The cross-section of expected stock returns. The Journal of Finance, 47(2), 427–465.

23.

Fama

E. F.

, & French

K. R.

(1993). Common risk factors in the returns on stocks and bonds. Journal of Financial Economics, 33(1), 3–56.

24.

Fama

E. F.

, & French

K. R.

(1995). Size and book-to-market factors in earnings and returns. The Journal of Finance, 50(1), 131–155.

25.

Fama

E. F.

, & French

K. R.

(1998). Value versus growth: The international evidence. The Journal of Finance, 53(6), 1975–1999.

26.

Fama

E. F.

, & French

K. R.

(2012). Size, value and momentum in international stock returns. Journal of Financial Economics, 105(3), 457–472.

27.

Fama

E. F.

, & French

K. R.

(2015). A five-factor asset pricing model. Journal of Financial Economics, 116(1), 1–22.

28.

Gillan

S. L.

(1990). An investigation into CAPM anomalies in New Zealand: The small firm and price-earnings ratio effects. Asia Pacific Journal of Management, 7(2), 63–78.

29.

Goh

H. H.

, Tan

K. L.

, Khor

C. Y.

, & Ng

S. L.

(2016). Volatility and market risk of rubber price in Malaysia: Pre and post global financial crisis. Journal of Quantitative Economics, 14(2), 323–344.

30.

Griffin

J. M.

(2002). Are the Fama and French factors global or country specific? The Review of Financial Studies, 15(3), 783–803.

31.

Hawawini

, & Keim

D. B.

(1998, 14 May. The cross-section of common stock returns: A review of the evidence and some new findings (Working Paper No. 008-99). Philadelphia, PA: The Rodney L. White Centre for Financial Research.

32.

Horowitz

J. L.

, Loughran

, & Savin

N. E.

(2000). Three analyses of the firm size premium. Journal of Empirical Finance, 7(2), 143–153.

33.

Hwang

, & Rubesam

(2013). A behavioural explanation of the value anomaly based on time-varying return reversals. Journal of Banking & Finance, 37(7), 2367–2377.

34.

Jain

, Vyas

, & Roy

(2013). A study on weak form of market efficiency during the period of global financial crisis in the form of random walk on Indian capital market. Journal of Advances in Management Research, 10(1), 122–138.

35.

Jensen

G. R.

, Johnson

R. R.

, & Mercer

J. M.

(1997). New evidence on size and price-to-book effects in stock returns. Financial Analysts Journal, 53(6), 34–42.

36.

John F. Kennedy Presidential Library and Museum. (1963). John F. Kennedy quotations. Retrieved on 8 March 2016, from http://www.jfklibrary.org/Research/Research-Aids/Ready-Reference/JFK-Quotations.aspx

37.

Kothari

S. P.

, Shanken

, & Sloan

R. G.

(1995). Another look at the cross-section of expected stock returns. The Journal of Finance, 50(1), 185–224.

38.

Kumar

, & Sehgal

(2004). Company characteristics and common stock returns: The Indian experience. Vision: The Journal of Business Perspective, 8(2), 33–45.

39.

Lakonishok

, Shleifer

, & Vishny

R. W.

(1994). Contrarian investment, extrapolation, and risk. The Journal of Finance, 49(5), 1541–1578.

40.

Lischewski

, & Voronkova

(2012). Size, value and liquidity, do they really matter on an emerging stock market? Emerging Markets Review, 13(1), 8–22.

41.

Loughran

(1997). Book-to-market across firm size, exchange, and seasonality: Is there an effect? The Journal of Financial and Quantitative Analysis, 32(3), 249–268.

42.

Manjunatha

, & Mallikarjunappa

(2011). Does three-factor model explain asset pricing in Indian capital market? Decision, 38(1), 119–140.

43.

Mohanty

(2002). Evidence of size effect on stock returns in India. Vikalpa, 27(3), 27–37.

44.

National Stock Exchange of India Limited (2015a). About Nifty 500. Retrieved on 20 September 2016, from https://www1.nseindia.com/products/content/equities/indices/nifty_500.htm

45.

National Stock Exchange of India Limited (2015b). Indian securities market: A review. Retrieved on 18 April 2016, from http://www.nseindia.com/research/dynaContent/ismr.htm

46.

National Stock Exchange of India Limited (2015c). Methodology document: Nifty Broad Market Indices. Retrieved on 24 February 2016, from http://www.nseindia.com/content/indices/Method_Nifty_Broad_Indices.pdf

47.

Patari

, & Leivo

(2017). A closer look at value premium: Literature review and synthesis. Journal of Economic Surveys, 31(1), 79–168.

48.

Peavy III

J. W.

, & Goodman

D. A.

(1983). The significance of P/Es for portfolio returns. The Journal of Portfolio Management, 9(2), 43–47.

49.

Rejeb

A. B.

, & Boughrara

(2013). Financial liberalization and stock markets efficiency: New evidence from emerging economies. Emerging Markets Review, 17(December), 186–208.

50.

Roll

(1981). A possible explanation of the small firm effect. The Journal of Finance, 36(4), 879–888.

51.

Rouwenhorst

K. G.

(1999). Local return factors and turnover in emerging stock markets. The Journal of Finance, 54(4), 1439–1464.

52.

Schwert

G. W.

(2003). Anomalies and market efficiency. In Constantinides

G. M.

, Harris

and Stulz

(Eds.), Handbook of the economics of finance —Financial markets and asset pricing (pp. 939–974), Amsterdam, the Netherlands: Elsevier.

53.

Sehgal

, & Balakrishnan

(2013). Robustness of Fama–French three factor model: Further evidence for Indian stock market. Vision: The Journal of Business Perspective, 17(2), 119–127.

54.

Sehgal

, & Pandey

(2014). Profitable trading strategies based on price multiple information: Evidence from India. Afro-Asian Journal of Finance and Accounting, 4(4), 408–425.

55.

Sehgal

, & Tripathi

(2005). Size effect in Indian stock market: Some empirical evidence. Vision: The Journal of Business Perspective, 9(4), 27–42.

56.

Sehgal

, Subramaniam

, & Morandiere

L. P. D.

L. (2012). A search for rational sources of stock return anomalies: Evidence from India. International Journal of Economics and Finance, 4(4), 121–134.

57.

Shum

W. C.

, & Tang

G. Y. N.

(2005). Common risk factors in returns in Asian emerging stock markets. International Business Review, 14(6), 695–717.

58.

Taneja

Y. P.

(2010). Revisiting Fama French three-factor model in Indian stock market. Vision: The Journal of Business Perspective, 14(4), 267–274.

59.

Vardhan

, Sinha

, & Vij

(2015). Behaviour of Indian sectoral stock price indices in the post subprime crisis period. Journal of Advances in Management Research, 12(1), 15–29.

60.

Wang

(2000). Size effect, book-to-market effect, and survival. Journal of Multinational Financial Management, 10(3–4), 257–273.

61.

, Liu

, & Chen

(2016). Re-examining risk premiums in the Fama–French model: The role of investor sentiment. North American Journal of Economics and Finance, 36(April), 154–171.