IT Stocks-Which one outperformed?

Recently, the IT sector has been in news with major IT companies announcing their quarterly results (Q2, FY 21). Infosys reported a 20.5% growth in its consolidated net profit (after minority interest) for the second quarter ended September 30th, over the same period last year. For HCL tech it was a growth of 18.5% in net profit during the same period,  while for Wipro and TCS it was a decline of 3.4% and 7.05% respectively. In the first week of October TCS also became the second Indian firm after Reliance Industries to cross the Rs 10 lakh crore mark. With the announcement of financial results and with stock market moves, investors and analysts look to gauge the performance of stocks based on various metrics.                     

In this blog post, I will be doing the same i.e. try to analyze the performance of top Indian IT stocks (by market cap), however, the metric I will be using is Jensen’s alpha. This measure was first used by Michael Jensen in 1968 to evaluate mutual fund managers. 

Jensen’s Alpha

Simplistically putting, Jensen’s alpha is the return on a stock that is above (or below) its’ expected return based on the Capital Asset Pricing Model (CAPM). In general, Capital Asset Pricing Model (CAPM) is the model used to find the expected return on a stock based on its’ systematic risk (Beta). Any return (positive or negative) that is not explained by this model is the Jensen’s alpha of a stock.

Calculation of Jensen’s Alpha

Both equation-1 and equation-2 are used in the calculation of Jensen’s alpha. Equation-2  shows how Jensen’s alpha can be calculated through regression analysis. Equation-2 shows that, if we run the regression of the excess return on a stock (Ri -Rf) (dependent variable or “Y”) over the excess return on market (Rm-Rf)   (independent variable or “X”), the slope of the line will be the Beta (β) while the intercept of the line will be Jensen’s alpha (∝). This line is also known as Security Characteristic Line (SCL).

Comparison of IT stocks based on Jensen’s alpha.

I calculated Jensen’s alpha of top Indian IT stocks by market cap. For this, I ran a regression of excess returns on these stocks over the excess returns on market. I collected monthly returns for these stocks (listed on BSE) for the period between January 2015 to August 2020. I assumed a monthly risk-free rate of 0.096% over the period of regression. Excess returns over the risk-free rate were calculated for each month, by subtracting 0.096% from the monthly return of a particular stock over the period of regression. Excess returns on market were calculated by subtracting 0.096% from the monthly returns of S&P BSE 500, over the period of regression. The results of the regression for these 4 stocks are summarized below.

Regression Results

	R^2	Beta	Jensen's Alpha (%) (intercept)	P-value of Intercept (%)
Infosys	0.15	0.53	0.81%	33.00%
TCS	0.22	0.59	0.64%	37.70%
HCL Tech	0.27	0.73	0.60%	45.38%
Wipro	0.10	0.41	0.31%	70.52%
Assumed Risk-Free Rate: 0.096%, Regression period: Jan-15 to Aug-20, Dividends not included for stocks and market.

Interpretation of Jensen’s Alpha

As discussed earlier the intercept of these regression lines will be Jensen’s alpha (∝) for these stocks. Infosys had an ∝  of approx. 0.81%. This means that on average Infosys made approx. 0.81% extra each month over the period of regression. This extra return is over and above the return as predicted by CAPM. Since CAPM takes into consideration the riskiness of the stock with respect to the market (i.e. Beta), it can be said that this extra return was over and above, the expected, risk-adjusted return of Infosys. This extra return was earned on an average each month, so, it can be estimated that on an annual basis, Infosys earned an extra, 0.808% X 12= 9.70%.(approx). This calculation shows that Infosys earned the highest Jensen’s alpha over the period of regression and Wipro the lowest. However, based on this it would be difficult to conclude that the performance of these stocks was significantly different from the prediction as per CAPM.

Statistical Significance of Jensen’s Alpha (Intercept)

All four stocks have a high p-value of intercept which means that for all these four stocks we cannot reject the null hypothesis that intercept or Jensen’s alpha is zero. In other words, in none of these stocks, Jensen’s alpha is statistically significant and hence it cannot be concluded that the performance of these stocks was significantly different from the prediction as per the CAPM. 

Statistical Significance of differences between Jensen’s alpha

I tried to compare the regression lines of each stock with one another (i.e. a total of six pairs) to check the null hypothesis that the difference in intercepts of these stocks is equal to zero. The process involved (although I am not an expert in running such regressions),  combining the data of two stocks together and creating, categorical variables (0 and 1) for the two stocks. Then, I ran a regression on the combined dataset and checked for the significance of the difference in intercepts. All the six regressions show a high p-value for the difference in intercepts which means we cannot reject the null hypothesis that the difference in intercepts was equal to zero. 

This means that we can not conclude that any of these stocks earned a Jensen’s alpha which was significantly different from Jensen’s alpha of the other 3 stocks. 

To summarize we cannot conclude that any of these stocks, performed significantly different from the prediction as per the CAPM, and any of these stocks significantly outperformed each other on a risk-adjusted basis.

Limitations of Jensen’s Alpha.

The analysis is based on CAPM and is subject to the limitations of the CAPM. There are limitations of the linear regression model also (such as the presence of outliers) which can impact the precision of this analysis. Moreover, the analysis is based on past data and there is no guarantee that the past performance of stock would be repeated in the future. The future performance of a stock is likely to depend on several other factors (for example, growth in the number of clients which will further drive revenue growth) rather than relying only on past data.  

Disclaimer: This blog post is not a stock or investment recommendation. It is not meant to provide any professional advice and is written with the purpose to discuss an analytical methodology. The methodology has various limitations and some of them have been discussed above. Any investment action you take based upon the analysis presented here is strictly at your own risk. The analysis and views presented here do not reflect the ideologies or points of view of any organization, I am affiliated with, or potentially affiliated with. Despite best efforts to present authentic information, the blog post is likely to suffer from errors and omissions. I am eligible to modify, update, or delete the information on this blog post.

The Mysterious Terminal Value Formula

                                          

One of the most important formulas in Discounted Cash Flow Valuation is the terminal Value formula which is stated as follows:

When I saw this formula for the first time (long back!), I was surprised that how can one formula calculate the entire terminal value of a firm. The terminal value formula assumes that the firm will continue to infinity, hence this single formula is calculating the entire terminal value of the firm, for all cash flows after year "t" to infinity.


In discounted cash flow valuation, after a certain number of years, we stop forecasting the annual cash flow to the firm and calculate the entire value of the firm thereafter, through this single formula.

In this blog, I will try to show how this formula is derived which can be useful to demystify this formula.

I have used three steps to explain the formula wherein, I explain the finite geometric series, the infinite geometric series, and then finally the application of infinite geometric series to derive the terminal value formula.   

Please click on the below link to read the full blog post.

The Mysterious Terminal Value Formula

The document contains mathematical formulas and I thought it would be better to provide a link to the document rather than posting all formulas as images.



I hope the blog has helped clarify the concept of terminal value formula.

In case of any queries, you can leave your comment or contact me at tulsyananimesh@gmail.com

Traps of Relative Valuation

                                                    

In this blog, I would be discussing a few traps that a beginner analyst is likely to fall into while doing relative valuation and how to overcome them. I am writing this blog to teach myself about relative valuation and to share some of my learnings on relative valuation.

In the realms of equity research/corporate finance relative valuation is used for identifying mispriced securities based on how comparable securities are priced in the market. One can also use relative valuation to derive a price for a particular security based on the price of similar securities in the market. While in a discounted cash flow (DCF) valuation, an analyst looks to derive the intrinsic value of a company, in relative valuation a security is viewed in comparison to other securities. This blog will not discuss the complete mechanism of relative valuations, however, it will be discussing some of the traps a novice is likely to fall into, if he is not careful. There are likely to be several other traps that may be more important than the ones discussed here. One can also have different viewpoints on the issues discussed here based on his method of analysis.

Trap-1-Negative EPS, Zero EPS

Price Earnings Ratio (P/E) multiple is one of the most commonly used multiples in relative valuation and it is sometimes used for ranking companies where a lower P/E implies relatively under-priced security and a higher P/E implies relatively overpriced security. However, P/E multiple breaks down when dealing with negative or zero EPS and ranks the companies incorrectly. To overcome this glitch, we can use Earnings Yield (E/P) which is the inverse of the Price Earnings ratio. E/P helps to correctly rank the companies when the EPS for a few companies is negative or zero. This can be illustrated with the help of the following example:

				Ranking           (Lowest to Highest)		Ranking                                    (Highest to Lowest)
Company	Current Market Price	EPS	P/E		E/P
A	10	5	2	2	0.5	1
B	10	2	5	3	0.2	2
C	10	-5	-2 (NM)	1	-0.5	3

Let us consider three companies A, B and C with the same current Market Price but different EPS as shown in the table above. While looking for cheapest stocks based on P/E we rank the companies from the lowest P/E to the highest. As shown above, company C turns out to be the cheapest stock and company B turns out to be the most expensive. This ranking is not meaningful because buying a security with a negative EPS is not cheaper than buying security with a positive EPS, at the same price. To overcome this glitch, we can take the inverse ratio i.e. E/P and rank the results from highest to lowest. Now the ranking is consistent with company B ranking above company C. Since, negative P/Es do imply any meaning they are considered not meaningful (NM).

Trap-2- Relying only upon P/E ratios

 P/E multiples have a drawback that they only measure the value of equity and not the value of the whole business. P/E multiples do not take into account the debt carried on by a company. Moreover, P/E multiples are affected by the accounting estimates used in reporting for items such as depreciation. Enterprise Value to EBITDA (EV/EBITDA) is another multiple that can be used to overcome these drawbacks up to an extent. Enterprise value (EV) is the value of the operating assets of the firm and is calculated simplistically as:

Market value of equity + Market Value of Debt (short term and long term) - Cash.

The advantage of EV/EBITDA over P/E ratio can be explained with the help of the following example:

Alpha				Beta
Price/Share	10			Price/Share	10
Cash	50			Cash	50
Other Assets	150			Other Assets	150
Equity	100	No of Shares@10/share=	10	Equity	200	No of Shares@10/share=	20
Debt	100			Debt	0
Revenue	100			Revenue	100
COGS	50			COGS	40
Gross Profit	50	Equity	100	Gross Profit	60	Equity	200
SG&A	10	Add: Debt	100	SG&A	10	Add: Debt	0
EBITDA	40	Less: Cash	50	EBITDA	50	Less: Cash	50
Depreciation	10	Enterprise Value	150	Depreciation	15	Enterprise Value	150
EBIT	30			EBIT	35
Interest	12.5			Interest	0
PBT	17.5			PBT	35
Tax	5.25			Tax	10.5
PAT	12.25	EPS	1.23	PAT	24.5	EPS	1.23
		P/E	8.16			P/E	8.16
		EV/EBITDA	3.75			EV/EBITDA	3

Note: Book Value of Debt is assumed to be equal to Market Value

In the above example, the companies Alpha and Beta are similar. Alpha is financed 50% by equity and 50% by debt. Alpha has an additional interest cost of 12.5 in comparison to Beta. Beta is financed 100% by equity. Beta manages its operations better than Alpha as therefore has a lower COGS and higher EBITDA. Beta also uses more conservative estimates for charging depreciation and therefore has a higher depreciation. Comparing based on P/E; both companies are identically priced; however comparing based on EV/EBITDA, Beta is cheap relative to Alpha.

Although, P/E ratio has its advantages, complementing the P/E ratio with EV/EBITDA multiple can uncover more insights. One can also complement the analysis by using other multiples such as Price to Book, (P/B), Enterprise Value to Sales (EV/Sales), etc. which have their advantages and disadvantages and may be more suitable in a particular situation.

Trap -3 – Not comparing apples to apples.

For a more logical comparison across firms, we need to exercise caution that we are comparing the same multiples across firms. For example, P/E ratios can be calculated in many ways (e.g., P/E ratio based on trailing twelve months of earnings, P/E ratio based on earnings reported for most recent financial year or P/E ratio based on forward earnings.). For consistency, similar ratios should be used across all comparable firms.

In the below example, Alpha and Beta are similar firms with the same market price. They reported the same EPS in the last financial year. However, subsequently Beta grew more rapidly and it’s trailing twelve months EPS is higher than Alpha. Comparing P/E ratios based on last reported EPS, both firms have similar P/Es; however, comparing P/Es based on trailing twelve months of EPS, Beta is cheap relative to Alpha. Caution needs to be exercised here that we are comparing P/Es based on trailing twelve months of earnings for both the companies.

Alpha		P/E	Beta		P/E
Current Market Price	100		Current Market Price	100
EPS (Last financial year)	20	5	EPS (Last financial year)	20	5
EPS (TTM)	20	5	EPS (TTM)	25	4

The EPS number for a company can also vary depending upon whether Basic EPS is used or Diluted EPS is used.

For comparison, it is also a general practice to compare firms in the same sector. Further filtering can be done by dividing firms based on size such as large-cap stocks, mid-cap stocks, etc.

Trap-4-Not removing the outliers

Defined simply, an outlier is a data point that differs significantly from other data points. While comparing firms based on a multiple, one can take an average of multiples observed across firms. This average is then used to identify overvalued/undervalued securities. However, this average may be affected by outliers making it too high or too low and the remaining firms may start appearing relatively undervalued or overvalued. It is better to remove these outliers from the list of comparable firms before calculating averages. Another method would be to use median values, instead of averages which will remove the effect of outliers.   

Trap-5 –Not doing further analysis

One may make the mistake of believing that an overvalued/undervalued company based on a comparison of its multiple with its peer group is the final step of his analysis. However, this is just the beginning of the process. For example, if a company appears undervalued based on its P/E multiple relative to its peers, it does not make it a good investment by default. There may be a reason for that undervaluation. The company may have growth prospects and ROE lower than its peers or it may have a higher risk. A more in-depth analysis of that company is required to find out if the company is trading below its fair value. As a result, the relative valuation process should be used as a screening and its’ results should be used for a more in-depth analysis of the company.

To sum it up, it can be said that although relative valuation is one of the most widely used tools for valuing and comparing securities, it should be used with caution as there are likely to be traps in the process.

References: Aswath Damodaran.