Price-to-Earnings Ratio Anomalies: A Comprehensive Analysis of Portfolio Performance in the Indian Stock Market

Introduction

When it comes to investing money, everyone wants to make as much profit as possible while mitigating risks. That\'s where the idea of \'Value Investing\' comes in. Value investing is all about finding securities trading below their intrinsic value, thereby offering a margin of safety for investors. One important concept in value investing is called the \'price-earnings (P/E) ratio anomaly\', or \'earnings yield anomaly\'. This idea challenges the usual way of looking at how stocks are valued. Instead of just looking at stock prices, the P/E ratio anomaly looks at a company\'s earnings compared to its overall value.

The genesis of the P/E ratio anomaly can be traced back to pioneering research conducted by Basu in 1977. Basu\'s groundbreaking study revealed that stocks with high earnings yield (low price to earnings), often categorized as value stocks, consistently outperformed their counterparts with lower earnings yield (high price to earnings), termed growth stocks. This discovery changed how people thought about investing and led to more research on the topic.

The interest in this topic stems from the significance of P/E ratios in guiding investment decisions and understanding market dynamics. As posited by Graham and Dodd (1934), the P/E ratio serves as a key determinant in assessing the intrinsic value of a stock, offering insights into its growth potential and future earning prospects. However, despite its widespread usage, anomalies in P/E ratios have long intrigued scholars due to their potential to influence investment outcomes and market efficiency.

Our examination of this problem is driven by the desire to shed light on the impact caused by P/E ratio anomalies on portfolio performance in the Indian context. By exploring this issue, we aim to contribute to the existing body of literature on behavioural finance and market efficiency by drawing upon the foundational theories of Fama and French (1992) to frame our analysis.

In the subsequent sections of this paper, we will delve into the methodology employed for our analysis, present the empirical findings, and discuss their implications. Finally, we will conclude with a summary of our key findings.

Review of Literature

The examination of \'low P/E ratio stocks\', commonly referred as \'value stocks\', has garnered significant attention over investment research after the passing of Benjamin Graham, the pioneer of value investing. Basu (1977) conducted ground-breaking research on the performance of value stocks compared to growth stocks on the New York Stock Exchange (NYSE) revealing that portfolios comprised of low P/E ratio stocks, consistently outperformed their high P/E ratio counterparts. Capaul et al. (1993) extended this analysis across six countries and affirmed the superior performance of value stocks, even after accounting for risk. Further empirical evidence by Brouwer et al. (1997) in European markets reinforced the notion that value strategies, based on earnings-to-price valuation metrics, yield higher annual returns than glamour strategies.

Bauman and Miller (1997) investigated the investment performance of high P/E ratio stocks listed on various stock exchanges, including NYSE, AMEX, and NASDAQ, reaffirming the outperformance of value stocks overgrowth portfolios. Similarly, Bauman et al. (1998) expanded this analysis to 20 established markets represented in the Morgan Stanley Capital International (MSCI) EAFE index and Canada, consistently finding value stocks outperforming growth stocks on both total return and risk-adjusted bases.

Anderson et al. (2003) observed the presence of a value premium in Mongolia, while Dunis and Reilly (2004) demonstrated the superior risk-adjusted performance of low P/E ratio stocks in the UK market. Ding et al. (2005) uncovered a positive value premium in East Asian countries before the 1997 Asian financial crisis, highlighting the resilience of value strategies across diverse market conditions. Anderson and Brooks (2006) corroborated these findings in the UK market, emphasizing the persistent outperformance of value stocks based on P/E ratios.

Despite the historical prevalence of the value premium, recent research has indicated a diminishing significance of asset pricing anomalies, including the value and size premiums. Schwert (2002) noted a reduction in anomalies over different sample periods, particularly observing the disappearance of the value and size premiums. Chordia et al. (2012) further supported this trend, suggesting a decrease in the economic and statistical significance of the asset pricing anomalies due to increased market efficiency and anomaly-based arbitrage activity.

While the presence of asset pricing anomalies may be diminishing in larger capital markets, studies on smaller capital markets suggest that premiums may still exist. Norges et al. (2009) and Grimeland (2018) examined the Oslo Stock Exchange and, found evidence of a size effect but limited support for a value effect, indicating that the dynamics of asset pricing anomalies may vary across different market environments.

Overall, the literature underscores the enduring appeal of value investing strategies based on earnings to price yield metrics, while also highlighting the evolving nature of asset pricing anomalies in global financial markets. This paper contributes to this body of research by examining the persistence of the value premium in the Indian stock market context, providing valuable insights into the efficacy of value-based investment strategies amidst changing market conditions.

Data and Methodology

Selection of Companies

A total of 60 companies were randomly selected from those listed on the National Stock Exchange (NSE). P/E ratios for all 60 companies were collected for the years 2021, 2022, and 2023 to ensure a comprehensive dataset for analysis. The selection of companies from 2021 aimed to mitigate the potential impact of the COVID-19 pandemic on the analysis.

Data Collection

Closing prices of the selected securities were collected from July 1st, 2021, to February 28th, 2024. Data collection for each year spanned from July 1st of the current year to June 30th of the following year, aligning with the annual portfolio construction timeline.

By selecting July 1st as the starting point for data collection, the analysis aimed to capture the impact of P/E ratios calculated as of March 31st, allowing for a lag of three months to reflect the availability of P/E ratio information to investors.

Portfolio Construction

Portfolios were constructed based on predefined P/E ratio ranges, categorizing companies into five distinct groups (<5, 5-10, 10-15, 15-25, and 25<). This categorization facilitated the comparison of companies with similar valuation metrics.

Calculation of Performance Metrics

Key performance metrics, including annualized return, annualized risk, Sharpe ratio, Beta, Treynor ratio, and Alpha (Jensen\'s), were calculated for each portfolio. These metrics provided insights into the risk-adjusted returns and abnormal returns generated by each portfolio relative to the market.

Comparison Analysis

Pairwise comparisons were conducted between portfolios and the market index (NIFTY 500) using t-tests to assess the statistical significance of differences in performance. The t-statistic and corresponding p-values were calculated to determine whether differences in returns were statistically significant.

Hypothesis

H0 - The absolute returns of the low P/E ratio portfolios are not statistically higher than the absolute returns of the high P/E ratio portfolios and the NIFTY 500 benchmark index.

Empirical Evidence

Performance Comparison of Portfolios and Market Index from 2021 to 2024

Table 1 presents a comprehensive analysis of the performance of various portfolios based on P/E ratios, along with comparisons against the market index NIFTY 500. This analysis aims to evaluate the impact of P/E ratio anomalies on investment outcomes and provide insights into the relative performance of different portfolios.

Table 1

Note: Jensen\'s Alpha for the market (NIFTY 500) is always zero, as it represents the benchmark against which portfolio performance is measured. Therefore, only the Jensen\'s Alpha values for individual portfolios are calculated and presented in the table.

Portfolio A (P/E ratio < 5): Portfolio A, consisting of stocks with a P/E ratio of less than 5, demonstrates the highest annualized return among all portfolios at 9.51%. This result suggests that stocks with lower P/E ratios have the potential to deliver higher returns compared to those with higher P/E ratios. However, it\'s important to note that with higher returns comes higher risk. Portfolio A also exhibits the highest annualized risk, or volatility, at 5.35%, indicating greater fluctuations in returns compared to other portfolios. Despite the higher risk, Portfolio A achieves a notable Sharpe Ratio of 1.14, signifying a favourable risk-adjusted return compared to the risk-free rate. Additionally, Portfolio A\'s Treynor Ratio of 5.22% implies that it generates excess return per unit of systematic risk, further highlighting its superior performance relative to the market index. Furthermore, Portfolio A\'s Jensen\'s Alpha of 6.74% indicates that it has generated significant abnormal returns beyond what would be expected based on its risk exposure.

Portfolio B (P/E ratio 5-10): Moving on to Portfolio B, which encompasses stocks with P/E ratios ranging from 5 to 10, we observe a slightly lower annualized return of 6.51% compared to Portfolio A. This indicates that as the P/E ratio increases, the potential for returns diminishes. However, Portfolio B also exhibits a higher annualized risk of 5.67%, reflecting increased volatility in returns compared to Portfolio A. Despite the lower return and higher risk, Portfolio B maintains a Sharpe Ratio of 0.55, suggesting a modest risk-adjusted performance relative to the risk-free rate. Similarly, its Treynor Ratio of 2.72% indicates some excess return per unit of systematic risk, albeit lower than Portfolio A. Furthermore, Portfolio B\'s Jensen\'s Alpha of 3.73% suggests that it has generated positive abnormal returns beyond what would be expected based on its risk exposure.

Portfolio C (P/E ratio 10-15): Transitioning to Portfolio C, which includes stocks with P/E ratios between 10 and 15, we observe a further decline in annualized return to 5.11%. This downward trend in returns as P/E ratios increase corroborates the notion that higher P/E ratios are associated with lower potential returns. Despite the decrease in return, Portfolio C maintains a relatively high annualized risk of 5.29%, indicating significant volatility in returns. The Sharpe Ratio for Portfolio C is 0.32, suggesting a less favourable risk-adjusted return compared to Portfolios A and B. Similarly, its Treynor Ratio of 1.49% indicates a lower excess return per unit of systematic risk compared to previous portfolios. However, Portfolio C still exhibits a positive Jensen\'s Alpha of 2.32%, suggesting some degree of abnormal returns beyond what would be expected based on its risk exposure.

Portfolio D (P/E ratio 15-25): Moving on to Portfolio D, which comprises stocks with P/E ratios ranging from 15 to 25, we continue to observe a decrease in annualized returns to 5.05%. This reinforces the inverse relationship between P/E ratios and potential returns. Despite the decrease in return, Portfolio D exhibits a lower annualized risk of 4.50% compared to Portfolios B and C, indicating relatively lower volatility in returns. However, Portfolio D\'s Sharpe Ratio of 0.37 suggests a slightly less favourable risk-adjusted return compared to Portfolio C. Similarly, its Treynor Ratio of 1.58% indicates a moderate excess return per unit of systematic risk. Additionally, Portfolio D maintains a positive Jensen\'s Alpha of 2.22%, suggesting some degree of abnormal returns beyond what would be expected based on its risk exposure.

Portfolio E (25 < P/E ratio): Finally, Portfolio E represents stocks with P/E ratios greater than 25, where we observe a further decline in annualized return to 5.16%. This reinforces the trend of diminishing returns as P/E ratios increase. Despite the lower return, Portfolio E exhibits a moderate annualized risk of 4.89%, suggesting relatively lower volatility as compared to Portfolios B and C. However, Portfolio E\'s Sharpe Ratio of 0.36 indicates a less favourable risk-adjusted return as compared to Portfolios A, B, and D. Similarly, its Treynor Ratio of 1.54% suggests a moderate excess return per unit of systematic risk. Nonetheless, Portfolio E maintains a positive Jensen\'s Alpha of 2.37%, indicating some degree of abnormal returns beyond what would be expected based on its risk exposure.

Market Index (NIFTY 500): In comparison to the portfolios, the market index NIFTY 500 exhibits an annualized return of 2.86%. This suggests that while the portfolios may outperform or underperform the market index. The market index itself represents the average performance of the broader market. Additionally, the market index displays a lower annualized risk of 3.03% compared to the portfolios, indicating relatively lower volatility in returns. However, the market index\'s negative Sharpe Ratio of -0.17 implies a less favourable risk-adjusted return as compared to risk-free assets. Similarly, its negative Treynor Ratio of -0.53% suggests a negative excess return per unit of systematic risk.

Comparative Analysis of Portfolio Performance and Market Index from 2021 to 2024

Table 2 presents the t-stats and p-values using a significance level of 5%. We reject the null hypothesis (the absolute returns of the low P/E ratio portfolios are not statistically higher than the absolute returns of the high P/E ratio portfolios and the NIFTY 500 benchmark index) if the p-value is below the significance level.

Table 2

A vs E (Portfolio A vs Portfolio E):

The comparison between Portfolio A and Portfolio E aims to assess whether there is a statistically significant difference in their performance. Portfolio A consists of stocks with a P/E ratio of less than 5, while Portfolio E comprises stocks with a P/E ratio greater than 25.

• The calculated t-statistic for this comparison is 1.0611, indicating the magnitude of difference between the mean returns of the two portfolios. A positive t-statistic suggests that Portfolio A has, on average, higher returns as compared to Portfolio E.
• The corresponding p-value is 0.2888, which represents the probability of observing such results if the null hypothesis (no difference in performance) were true. With a p-value greater than the commonly used significance level of 0.05, we fail to reject the null hypothesis.
• Conclusion: There is no statistically significant difference between the performance of Portfolio A and Portfolio E. This suggests that, based on the sample data, both portfolios exhibit similar performance characteristics despite the difference in their P/E ratios.

B vs D (Portfolio B vs Portfolio D)

• The comparison between Portfolio B and Portfolio D aims to evaluate whether there is a significant disparity in their performance. Portfolio B consists of stocks with a P/E ratio between 5 and 10, while Portfolio D comprises stocks with a P/E ratio between 15 and 25.
• The t-statistic for this comparison is 0.4150, indicating the magnitude of difference in the mean returns between the two portfolios. A positive t-statistic suggests that Portfolio B has, on average, higher returns as compared to Portfolio D, although the difference is less pronounced.
• The associated p-value is 0.6782, which exceeds the significance level of 0.05. Consequently, we do not have sufficient evidence to reject the null hypothesis.
• Conclusion: There is no statistically significant difference between the performance of Portfolio B and Portfolio D. This implies that, based on the available data, both portfolios demonstrate comparable performance characteristics despite their different P/E ratios.

A vs M (Portfolio A vs Market Index)

• The comparison between Portfolio A and the Market Index assesses whether Portfolio A outperforms or underperforms the broader market represented by the NIFTY 500 index.
• The t-statistic for this comparison is 2.0038, indicating a substantial difference in mean returns between Portfolio A and the Market Index. A positive t-statistic suggests that Portfolio A, on average, has higher returns as compared to the market index.
• The corresponding p-value is 0.0453, which is less than the significance level of 0.05. Consequently, we reject the null hypothesis, indicating a statistically significant difference in performance.
• Conclusion: Portfolio A exhibits a statistically significant difference in performance compared to the Market Index. This suggests that Portfolio A outperforms the market.

E vs M (Portfolio E vs Market Index)

• The comparison between Portfolio E and the Market Index aims to determine whether Portfolio E differs significantly from the broader market.
• The t-statistic for this comparison is 0.8051, indicating a modest difference in mean returns between Portfolio E and the Market Index. A positive t-statistic suggests that Portfolio E, on average, has higher returns as compared to the market index, although the difference is less pronounced.
• The associated p-value is 0.4209, which exceeds the significance level of 0.05. As a result, we lack sufficient evidence to reject the null hypothesis.
• Conclusion: There is no statistically significant difference between the performance of Portfolio E and the Market Index. This implies that Portfolio E exhibits performance characteristics similar to those of the broader market, as evidenced by the non-significant difference in mean returns.

Conclusion

The analysis of P/E ratio anomalies in the Indian stock market provides valuable insights on the performance of different category of portfolios, based on P/E ratios. The findings reveal that portfolios with lower P/E ratios tend to exhibit higher annualized returns, suggesting potential opportunities for investors seeking higher returns. However, these portfolios also come with higher volatility, indicating increased risk. Despite the higher risk, certain portfolios, particularly Portfolio A with P/E ratio < 5, demonstrate favourable risk-adjusted returns, as reflected in their higher Sharpe ratios as compared to the market index. Additionally, Portfolio A exhibits significant abnormal returns beyond what would be expected based on its risk exposure, suggesting potential for alpha generation.

Furthermore, the comparison analysis between portfolios and the market index NIFTY 500 sheds light on their relative performance. While some portfolios, such as Portfolio A, exhibit statistically significant differences in performance as compared to the market index, others show no significant deviations. This indicates that while certain portfolios may offer opportunities for outperformance relative to the broader market, others tend to align closely with market trends. Overall, the analysis underscores the importance of considering P/E ratio anomalies in investment decision-making and highlights the potential for strategic portfolio construction and allocation strategies to capitalize on these anomalies for generating alpha and optimizing investment outcomes in the Indian stock market.