The Magic of Bottom-Up Betas

When analysing a stock, we can calculate its daily returns and create a histogram to visualize the data. The mean of the stock represents its long-term return, while the standard deviation indicates the range within which the stock operates, also known as total risk. This means that at any given point in time, the stock's movement will fall within the standard deviation range. When considering total risk, we must take into account both systematic and unsystematic risk. For instance, if the standard deviation of Maruti Suzuki's daily returns is 1.493% and the mean is 0.00868%, then we can expect that 95% of the time, the daily return will be between 3.0145% and 0.0106%.

Beta is a measure of systematic risk, specifically examining the market-specific risk associated with a particular stock. It provides insight into how a stock moves in relation to the overall market. For example, if the market moves by 1% on any given day and the beta of a stock is 1.6, then that stock would be expected to move by 1.6%.

To calculate beta, one must regress the stock returns against the returns of the market over a specific timeframe (such as monthly, daily, or weekly) and for a certain number of years. Selecting the appropriate timeframe is crucial, as going too far back in time may not accurately reflect current business dynamics, while choosing a short timeframe may not capture the full cyclical nature of the business. Ideally, a monthly timeframe over a period of 5 years would provide a comprehensive analysis.

During the regression analysis, it is important to consider the timeline for which returns are being analyzed. Additionally, it is crucial to account for any significant macroeconomic events, such as the impact of COVID-19. For instance, if a company is in the automobile business, its sales may have plummeted during the pandemic, whereas an IT business may have experienced a surge in sales. Therefore, it is necessary to conduct the analysis under normal circumstances. Similarly, if a company operates in the oil business, it is important to avoid accidentally selecting a timespan when oil prices have skyrocketed. This could significantly skew the results. The slope of the regression line represents the beta coefficient. Another way to obtain this coefficient is by calculating the covariance of market returns and stock returns divided by the variance of market returns.

There are certain issues associated with utilizing this beta range. When calculating a regression beta (levered beta), a standard error for the equation is also obtained, resulting in a range of beta values. For instance, 95% of the time beta will fall within the range of (beta-2*standard error, beta+2*standard error). Consequently, determining the best estimate becomes challenging. Some may argue in favor of selecting the extreme value (beta + 2*standard error) as the optimal estimator, but this could represent the worst-case scenario for the company under analysis.

Moreover, this beta solely takes into account historical events, failing to consider potential changes in the company's financial leverage. Additionally, if the company has recently ventured into a new industry or introduced a new product, the business mix may have altered significantly, which is not reflected in the levered beta.

Furthermore, in the case of a company operating across multiple segments like Reliance, which is involved in energy, petrochemicals, natural gas, retail, entertainment, telecommunications, mass media, and textiles, the regression analysis yields a single value. This value provides a combined risk assessment for all segments, disregarding the fact that each segment carries its own distinct risk profile.

To address these issues, one may employ the bottom-up beta approach. The triadic facets characterize the bottom-up beta phenomenon.

Business Characteristics

Due to the fact that beta is influenced by the nature of business operations, a company producing luxury goods will exhibit a higher beta as luxury products are considered more risky. Conversely, companies in the FMCG sector, where products are consumed daily, will have a lower beta due to lower risk associated with their products. Additionally, industries with cyclical patterns such as hotels, airlines, and tourism will have higher betas. The demand for a product also plays a role in determining beta, as products that are essential and have high demand will result in lower beta. Furthermore, some businesses may have negative betas, indicating that returns are lower than the risk-free rate. For instance, gold mining companies often have negative betas as gold serves as a hedge against inflation, leading to an increase in gold prices when currency devalues.

Operating Leverage

Sectors that have higher fixed costs generally possess a higher beta, as their revenue tends to rise sharply during periods of economic growth, leading to increased profitability. Nevertheless, in times of economic downturn, these industries encounter obstacles and complications.

Lastly financial leverage

The company aims to achieve an ideal financial structure, characterized by a favorable debt-to-equity ratio, throughout its life cycle. However, it is important to acknowledge that taking on debt is a two-edged sword. On one hand, it offers tax advantages, but on the other hand, it raises the anticipated return on equity. This is because when a company incurs debt, it must make fixed interest payments, resulting in lower returns for shareholders. Additionally, there is an increased risk of default, leading to greater volatility in earnings for equity holders

Financial leverage factor

Hence, by combining these three factors, we can determine a beta known as the bottom-up beta. In order to calculate the bottom-up beta, we must first gather a comprehensive list of publicly listed companies worldwide (approximately 40,000 companies as of now) that operate exclusively within a single industry. We will then perform a regression analysis, comparing these companies to a market index specific to their respective countries or any suitable substitute for the market portfolio. It is worth noting that some argue that beta may vary across countries and that country-specific risks should be considered. However, these risks are accounted for in the equity risk premium portion.

Once we have obtained the levered beta, we will eliminate the debt component to obtain the unlevered beta. This process will be repeated for as many countries as possible within the industry, resulting in multiple unlevered betas. By calculating the average of these unlevered betas, we can determine the industry-specific beta. To obtain the bottom-up levered beta for a particular company, the financial leverage factor can be added to the industry-specific beta.

This approach provides a comprehensive analysis and resolves the aforementioned issues. Previously, we relied on a single value as our estimate, but now we have a more accurate value by utilizing the law of large numbers. Even if one number cannot be the best estimate, the average of 100 less accurate numbers will yield a more reliable estimate. Consequently, we can calculate the bottom-up beta for a company's current business mix. For example, in the case of Reliance Industries, we can calculate the bottom-up beta for their oil, telecom, textile, etc. segments. By considering the debt proportions of each business segment and incorporating the financial leverage factor, we can obtain the bottom-up beta. Finally, we can utilize the weighted average beta of the revenue segments within the Reliance industry to gain a clearer understanding of the reliance beta.

In conclusion, calculating betas using the bottom-up approach is significantly more reliable than relying on a single number obtained through regression analysis.