In this assignment we looked at financial ratios of same industry companies and at the CAPM applied to companies from different industries.
I chose the same companies as in assignment 1 in order to be consistent with my analysis.
After looking at the financial ratios of selected companies in the insurance sector we roughly analyzed the results and constructed a “buy” recommendation for Phoenix stock.
We used the Capital Asset Pricing Model in order to find the expected cost of equity for 4 companies from different industries. Also here I chose the same companies as in assignment 1 in order to be consistent with my analysis.
I chose to look at 4 companies from the financial services/insurance industry.
The chosen companies were:
| Company | Assets | Leverage | EPS | Market_to_book | Profit_Multiplier | Equity_to_book | ROE | MVA | Stock_price |
|---|---|---|---|---|---|---|---|---|---|
| Phoenix | 100,893,156 | 6.70% | 1.19 | 0.75 | 6.15 | 4.40% | 12.10% | (1,177,778) | 13.38 |
| Migdal | 132,357,064 | 0.70% | (0.18) | 0.65 | 0.00 | 3.80% | -0.40% | (1,753,410) | 3.11 |
| Clal | 96,491,866 | 0.80% | (6.35) | 0.60 | 0.00 | 4.30% | -5.00% | (1,537,329) | 48.50 |
| Harel | 94,050,808 | 0.70% | 0.76 | 0.75 | 9.27 | 5.30% | 8.10% | (1,222,582) | 17.32 |
The insurance sector as a whole is traded at a discount (Market to book < 1). This is because of negative market expectations due to the raising regulatory environment (solvency - not enabling the distribution of dividends,Bituach Hova (Motor bodily injury) reform,health insurance reform and others) and current and expected low interest rates.
It can also be seen that Phoenix’s gearing is substantially higher than is competitors. This seems a bit strange to me since certain debt instruments can be used as solvency capital. Also, we saw in the previous assignment - the interest rates for Clal,Migdal were higher than the 2.5% that the latest Phoenix bond was issued at. Maybe they are waiting for better times to raise debt capital.
I also just heard that Phoenix recently raised another 460 Million shekel in debt at a 3.6% interest rate.
Phoenix stock is traded at the moment at 13.38 shekels giving a 40% raise in the past months (after a 40% drop in the months before). I even though the market might have already incorporated it’s expectations in the current price I would give a “buy” recommendation due to following reasons:
# Loading Necessary libraries
library(quantmod)
library(zoo)
library(tseries)
library(dplyr)
library(knitr)
library(dygraphs)
# Function to calculate weekly returns on a stock
weekly_stock_returns <- function(ticker, start_date) {
# Download the data from Yahoo finance
symbol <- getSymbols(ticker, src = 'yahoo', from = start_date,
auto.assign = FALSE, warnings = FALSE)
# Tranform it to weekly returns using the periodReturn function from quantmod
data <- periodReturn(symbol, period = 'weekly', type = 'log')
# Let's rename the column of returns to something intuitive because the column
colnames(data) <- as.character(ticker)
return(data)
}Data is downloaded at a daily price rate. The log of the ratio of the adjusted closed price of each week is then used to calculate the weekly returns. This function builds upon the quantmod R package. I used the same function as in assignment 1 except changed the returns to weekly returns instead of monthly.
I chose the following companies:
The mathematical definition of \(\beta\) is \[\beta_{i}=\rho_{i,m}\cdot\frac{\sigma_{i}}{\sigma_{m}}\]
In other words it’s the correlation between an asset’s return and the market times the ratio of standard deviation of the returns. Therefore three things can affect the \(\beta\) - the correlations and the marginal volatilities.
Having said that:
El al - I would expect the correlation to be at a medium level and the ratio of volatilities to be greater than 1. El al’s returns are linked to foreign demands as well and FX and commodities (fuel etc’) movements.
Clal bio - I wouldn’t expect correlation to be high and the ratio of volatilities to be greater than 1. This is since clal is an exporter and shouldn’t be so correlated to the TA 100. It’s in a risky sector should be more volatile. In overall would expect the \(\beta_{clal-bio}\) to be lower than 1.
Strauss - I wouldn’t expect correlation to be so high and the ratio of volatilities should a bit larger than 1. This is since strauss is a big exporter and shouldn’t be so correlated to the TA 100. It’s also in a fairly “safe” industry that should be very volatile. Therefore \(\beta_{Strauss}\) should be lower than 1.
Phoenix - I would expect Phoenix to have a beta that is close to 1 or a bit higher since the profits are linked to portfolio returns which are correlated to the TA 100 as well as demand for insurance products grows with economic grow and claim tend to grow with financial distress.
# download return data
ta100 <- weekly_stock_returns("^TA100","2016-08-30")
elal <- weekly_stock_returns("ELAL.TA","2016-08-30")
clalbio <- weekly_stock_returns("CBI.TA","2016-08-30")
strauss <- weekly_stock_returns("STRS.TA","2016-08-30")
fnx <- weekly_stock_returns("PHOE.TA","2016-08-30")
# running regressions
capm_elal <- lm(elal~ta100)
capm_clalbio <- lm(clalbio~ta100)
capm_strauss <- lm(strauss~ta100)
capm_fnx <- lm(fnx~ta100)
# results
betas <- cbind(coef(capm_elal),coef(capm_clalbio),coef(capm_strauss),coef(capm_fnx))
colnames(betas) <- c("EL AL","Clal bio","Strauss","FNX")
rownames(betas) <- c("alpha","beta")
kable(betas)| EL AL | Clal bio | Strauss | FNX | |
|---|---|---|---|---|
| alpha | -0.0104563 | -0.0090934 | -0.0001565 | 0.0119509 |
| beta | -1.2301773 | 0.5100862 | 0.3535192 | 1.0871395 |
WE can see in the return graph above that the returns of El Al weren’t correlated with the returns of the TA 100. But this wasn’t because Elal’s business isn’t linked to the Israeli market but due to internal Elal affairs that caused the stock to fall even when the TA100 gave positive returns.
The time period chosen there isn’t a good representation - will discuss this issue more below.
We will use the following relationship in order to get our company specific cost of capital: \[r_{i}=r_{f}+\beta_{i}(r_{m}-r_{f})\]
Risk free rate taken from here.
# Calculating the Cost of Equity
rm <- 0.07
rf <- 0.0325
cost_equity <- rf+betas[2,]*(rm-rf)
DF <- data.frame(Cost_of_Equity=percent(cost_equity))
formattable(DF)| Cost_of_Equity | |
|---|---|
| EL AL | -1.36% |
| Clal bio | 5.16% |
| Strauss | 4.58% |
| FNX | 7.33% |
I would assume that Strauss would have less risk than Clal biotechnology.
Problems with this estimation method is the trade off of having enough data (variance) and having relevant representable data (bias). I’m not sure that the last 3 months are enough. Also a period of weekly return might be too noisy for this exercise.
El Al is good example where short term data leads to short term variance. This variance could be reduced with more data but it might add bias. If we believe that return’s distribution is stationary is shouldn’t cause a problem. But this is a strong assumption.
The definition of the market can be misleading. It is can be claimed that the TA100 is an appropriate market for Phoenix (even though 40% of investment portfolio is invested abroad…) , but what is the market for Strauss and Clal bio?
Also, the significance of estimated parameters need to be checked as well as if the assumptions of a linear model hold in our case.