CF Session 6 - Introduction to Risk

Avijit Bansal
Indian Institute of Management Calcutta

January 11, 2023

1
Story so far

Opportunity Cost

Required rate of return

Hurdle rate

Risk-adjusted discount rate

So far we have hand-waved the topic of how to estimate risk. the cases directly
provided a value of r for us to go ahead with computations.

We will now develop a framework to measure the risk of an investment

2
Efficient Markets

Price incorporate all the available information

If any new piece of information comes up, markets react and incorporate the
information quickly

It is difficult to predict future returns by looking at the past data

Any apparent mispricing cannot persist for long - no arbitrage

In this course, we will assume that markets are efficient.

3
What is risk?

4
What is risk?

In finance, we think of risk in terms of

Correlation

Standard Deviation

Probability of an extreme negative event

5
Riskier assets give higher returns over the long
term

Equity has historically given higher than corporate bonds

Corporate bonds have historically given higher returns than government bonds

6
Average returns of assets in India (1978-2017)

Nominal Real Risk Premium

1 year G-Sec yield 9.12% 2.46% 0%

AAA-Rated Corporate Bonds 12.19% 5.54% 3.08%
Common Stocks 18.92% 12.27% 9.81%

Note: Risk premiums are not constant

7
Some statistical terms

Mean = µi = E [Ri,t ]

Variance = σi2 = E [(Ri,t − µi )2 ]

p
Standard Deviation = σi = σi2

Correlation
How two random variable move in conjunction?

Covariance = Cov [Ri,t , Rj,t ] = E [(Ri,t − µi )(Rj,t − µj )]

E [(Ri,t − µi )(Rj,t − µj )]
Correlation = Corr [Ri,t , Rj,t ] =
σi σj

8
Corelation coefficent = 0 Corelation coefficent = −0.5

Corelation coefficent = 0.5 Corelation coefficent = 0.9

9
Sensex returns - daily correlation
Corelation between current and previous day sensex returns

−10 −5 0 5 10

10
Sensex returns - monthly correlation
Corelation between current and previous month sensex returns

−10 −5 0 5 10

11
Tata Motors returns - daily correlation
Corelation between current and previous day return − Tata Motors

−10 −5 0 5 10

12
Tata Motors returns - monthly correlation
Corelation between current and previous month return − Tata Motors

−20 −10 0 10 20

13
Sensex and Tata Motors daily ret- Corr of 0.61
Corelation between Sensex and Tata Motors returns − daily

−10 −5 0 5 10

14
Sensex and Tata Motors monthly ret - Corr of 0.61
Corelation between Sensex and Tata Motors returns − Monthly

−10 0 10

15
Watch this video

Basant Maheshwari 2023 Outlook

16
What can we infer from the above plots?

It is difficult to predict future returns both at market and at stock level (in line
with the predictions of efficient markets)

However, there is some degree of correlation between individual stock and the
market

Mean Standard Deviation

Sensex returns - daily 0.06% 1.44%

Sensex returns - monthly 1% 6%
Tata Motors returns - daily 0.05% 2.68%
Tata Motors returns - monthly 0.8% 11.6%

Should we measure the risk of the stock in terms of its standard deviation or
correlation with the market?

17
What is more desirable - A single stock or a
portfolio of stocks?

18
What is a portfolio?

Portfolio is a collection of securities

Each asset in the portfolio has a particular weight, which in most cases sum to 1

Let there be n stocks in the portfolio with the weights of w1 , w2 , . . . , wn

w = {w1 , w2 , . . . , wn }

Ni × Pi
wi =
N1 × P1 + · · · + Nn × Pn

19
Why form portfolios? Why not pick best stocks?

Stock picking and market timing is not easy (for most people)

Portfolios can help reduce exposure to certain risks by diversification

20
What is a “good” portfolio?

Maximizes mean returns

Minimizes variance of the return

Mean-variance efficient

21
Assumptions about human behaviour

People like high returns

People dislike uncertainty and high fluctuations

People care only about the return and volatility of their overall portfolio not
individual stocks

People hold well-diversified portfolios

People only care about the risk that a particular stock adds to their portfolio and not
the risk of the stock itself.

The main objective is to find optimal weights such that combination is mean-variance
efficient.

22
Mean-Variance Analysis of a portfolio

Mean return of each stock = E [Ri ] = µi

Variance of returns of each stock = E [(Ri − µi )2 ] = σi2

p
Standard Deviation of returns of each stock = Var [Ri ] = σi

If there are n stocks in the portfolio with weights w = {w1 , w2 , . . . , wn }

E [Rp ] = w1 µ1 + w2 µ2 + · · · + wn µn = µp

Mean expected return of the portfolio is the weighted average of the mean return of
the portfolio constituents

23
Mean-Variance Analysis of a portfolio
Var [Rp ] = E [(Rp − µ)2 ] = E [(w1 (R1 − µ1 ) + · · · + wn (Rn − µn ))2 ]

n
X n n
X X
Var [Rp ] = wi2 σi2 + wi wj Cov [Ri , Rj ]
i=1 i=1 j=1,j̸=i

n n n
X X X
Var [Rp ] = wi2 σi2 + wi wj σi σj ρij
i=1 i=1 j=1,j̸=i

w12 σ12 w1 w2 σ1 σ2 ρ12 ··· w1 wn σ1 σn ρ1n

 
w2 w1 σ2 σ1 ρ21 w22 σ22 ··· w2 wn σ2 σn ρ1n
Covariance Matrix = 
 
.. .. .. .. 
. . . .
wn w1 σn σ1 ρn1 wn w2 σn σ2 ρn2 ··· wn2 σn2

n terms representing variances

n2 − n terms representing covariances

If ρ′ s < 1, there are benefits of forming a portfolio

24
Consider a portfolio of two stocks

Let there be two stocks in a portfolio

E [Rp ] = w1 µ1 + w2 µ2

Var [Rp ] = w12 σ12 + w22 σ22 + 2w1 w2 σ1 σ2 ρ12

When ρ12 = 1, Var [Rp ] = (w1 σ1 + w2 σ2 )2 , no benefit from diversification

When ρ12 = −1, Var [Rp ] = (w1 σ1 − w2 σ2 )2 , money making machine

25
Mean Variance analysis with two stocks

From 2000 to 2022, Asian Paints has delivered an average monthly return of 2% with
a standard deviation of 7.43%. Over the same period, Pidilite has delivered average
monthly return of 2.5% with a standard deviation of 9%. The correlation between the
monthly returns of the two stocks is 0.44. How can you benefit from diversifying your
investments between these two stocks?

26
Consider an equally weighted portfolio
n n n
X X X
Var [Rp ] = wi2 σi2 + wi wj Cov [Ri , Rj ]
i=1 i=1 j=1,j̸=i

1
Let wi =
n

n n n
X σ2 i 1 X X
Var [Rp ] = + Cov [Ri , Rj ]
n2 n2
i=1 i=1 j=1,j̸=i

n n2 − n
= × Average Variance + × Average Covariance (1)
n 2 n2
1 n−1
= × Average Variance + × Average Covariance
n n
≈ Average Covariance (If n is large)

For a portfolio with large n, the risk is largely driven by average covariance of
the constituent stocks

You can think of Average Covariance as a risk that can not be eliminated by
adding more stocks in your portfolio
27
Systematic risk (Undiversifiable risk)

28
Systematic risk

Systematic risk is also known as undiversifiable risk or market risk

Stocks usually have some degree of commonality in their movement

29
Tangency portfolio

30
Sharpe Ratio

The line connecting risk-free asset and the tangency portfolio is called the
capital market line / capital asset line

Tangency portfolio has the highest Sharpe Ratio

E [Rp ] − rf
Sharpe Ratio =
σp

31
Why will the tangency portfolio be equal to market
portfolio?

Market portfolio is the portfolio of all possible equities containing them in

proportion of their market capitalization

If each investor demands the same tangency portfolio with the highest sharpe
Ratio, and

If every asset that is traded has to be held by someone, then M will be equal to
value weighted portfolio of all stocks

The combined holdings of stocks held by all investors must be equal to M

(demand equal supply)

Also known as the market portfolio

Demand of Efficient Portfolio = Supply of all the traded stocks

32
Risk-return trade-off of other portfolios on CML

σp
E [Rp ] = rf + (E [Rm ] − rf )
σm

Every portfolio that can be formed using rf and Market portfolio will satisfy the
above equation

Note: The above equation is true only for efficient portfolios

How will be compute the risk-return trade-off of non-efficient portfolios?

33
Risk-return trade-off of non-efficient portfolios

σp × ρpm
E [Rp ] = rf + (E [Rm ] − rf )
σm

For efficient portfolios (portfolios on CML), ρpm = 1

34
Major points

Diversification reduces risk, as measured by standard deviation of the portfolio

returns

In a well diversified portfolio, the source of risk is the covariances between the
constituent stocks. This risk is systematic (undiversifiable)

A rational investor will always invest in a portfolio on the upper half of the
mean-variance efficient frontier

If a risk-less asset is available, then any investment which is a combination of

risk-free asset and the tangency portfolio provides the best risk-return tradeoff

35
How individual stock impacts portfolio risk?

Risk of a well-diversified portfolio depends on the market risk

(systematic/undiversifiable risk)

Hence, a stock’s contribution to portfolio risk is only to the extent of that

particular stocks’s market risk

The market-risk of stock or the sensitivity of a stock’s returns to market returns

is also known as beta (β)

σim
βi = 2
σm
You can think of βi as the proportion of the market variance that arises due to stock
i ′ s co-movement with the market.

36
References I

37