6/03/2023

Econ 122 Financial Economics I

Lecture 3: Return and Risk

Jianxin Wang

Team Project
100 marks for 30% of the subject grade.
Form teams of 3-5 members and select two stocks
Send team info and stock names to Thad by Mar 13
Team report:
Max 8 pages. Each section covers a required task
Focus on explaining and interpreting (not reporting) the
statistical results
Know and rely on the characteristics of your stock/company
Consultation:
Mid-April and mid-May
Submission deadline: May 29
Team report and an Excel file.

1
 6/03/2023

Today’s Topics
Measuring investment returns
Holding-period return, annual percentage return, and
effective annual return.
Continuously compounded return
Arithmetic and geometric average returns
Nominal and real returns
Expected return, risk, and risk premium
Risk-free return
Risk measures
Historical returns

Holding-Period Return (HPR)

Holding period: hours, days, weeks, years, etc.
Rate of return over an investment period t:
Pt and Dt are price and dividend at the end of t
HPRt = rt 1
Return before deducting transaction costs
If Pt-1 = $25, Pt = $27, and Dt = $1.25
$ .
rt 13%
.
Capital gain = 8%, dividend yield = 5%

2
 6/03/2023

Annualized Returns
n periods in a year, e.g. 4 quarters, 12 months, etc.
Annualizing: the annual return if HPR = r is earned
in every period of a year
Annual percentage rate (APR)
APR ≡ HPR × n
Effective annual rate (EAR)
If prices are P0 and Pn at the start and the end of the year,
HPRyear = EAR = Pn/P0 – 1 = (1+HPR)n – 1
If return over 100 days is HPR = 4%,
n = 365/100 = 3.65
APR = 4% x 3.65 = 14.6%, EAR = (1.04)3.65 – 1 = 15.4%
In economics, annualize monthly interest rate.

APR, EAR, and Continuous Compounding

By definition HPR=APR/n=(1+EAR)1/n –1, therefore
1+EAR = [1+APR/n]n
What if reinvest in every second, millisecond,…?
n → ∞: continuous compounding (CC).
Under CC: 1+EAR = lim 1 e
→
Alternatively, APR = ln(1+EAR)
Since EAR = Pn/P0 – 1, the implied continuously
compounded return is rCC = ln(Pn/P0) = ln(Pn) - ln(P0).
For any period t, rCC,t = ln(1+rt) rt if rt is small.

3
 6/03/2023

APR, EAR, and Continuous Compounding

EAR includes earnings from reinvestment.
APR does not.
The difference between APR and EAR grows as the
frequency of compounding/reinvestment increases.

Multi-Period Return
T: number of investment periods (e.g. years)
n: number of compounding/reinvestment periods
(e.g. months) within an investment period T.
r is the (constant) HPR in one investment period.
If the initial investment is V0, the value after T
investment periods is VT = V0(1+r/n)nT
If V0 = $10, T = 3, n = 2, and r = 4%
End value V3=$10x(1+0.04/2)2x3 = $11.26
return r(3)=V3/V0-1=(1+0.04/2)6-1 = 12.6%
If n → ∞, VT = lim V0 1 r/n V0e
→
T-period return under CC is r(T) = VT/V0 – 1 = erT – 1
r(3) = e0.04x3 – 1 = 12.75%

4
 6/03/2023

Average Return over T Periods

Now HPRt = rt varies with t.
Arithmetic average: r̅ ∑ r
Geometric average
Compound return over t periods = (1+r1)(1+r2)…(1+rT) – 1
Average compound return or geometric average
(1 + r̅ )T = (1+r1)(1+r2)…(1+rT)
r̅ [(1+r1)(1+r2)…(1+rT)]1/T - 1
r̅ r̅ - var r
Quarterly returns are 10%, 26%, -20%, and 20%.
r̅ =(0.10+0.26-0.20+0.20)/4 = 9%
r̅ [(1.1)(1.26)(0.8)(1.2)]1/4 – 1 = 7.4%

Nominal and Real Returns

Economics: nominal and real interest rates
Consumer price index CPIt
Base period t = 1: CPI1 = 1
CPIt = ∏ 1 i
Inflation it = (CPIt-CPIt-1)/CPIt-1 = CPIt/CPIt-1 – 1
Observed nominal price Pnom,t
Nominal return rnom,t = Pnom,t/Pnom,t-1 – 1
Real price (in t=1 $) Preal,t = Pnom,t/CPIt
,
Real return rreal,t = Preal,t/Preal,t-1 – 1 = –1
,

Fisher equation: 1 + rreal,t = (1 + rnom,t)/(1 + it)

10

5
 6/03/2023

Fisher Equation
Alternatively: 1+rnom,t = (1+rreal,t)(1+it)
rnom,t = rreal,t + it + rreal,t it
If rreal,t and it are small, rnom,t rreal,t + it
Example
If rnom,t = 4% and it = 2%, rreal,t = 1.04/1.02 – 1 = 1.96%.
Or rreal,t 4% - 2% = 2%
Fisher hypothesis: rnorm,t = rreal,t + E(it)
E(it) is the expected inflation by investors
How good are investors in forecasting inflation?

11

Scenario Analysis and Probability Distributions

BKM page 129 – 130
Stock return rt is a random variable
There are st = 1,…,Kt possible economic scenarios or
states. Economic state st has HPR r . Probability of
state s is P(r ).
Expected return E(rt) = ∑ P r r
Unexpected return = rt – E(rt)
Variance of rt
σ Er E r E r E r ∑ P r r E r
A measure of return uncertainty or risk
Standard deviation σt has the same unit as rt.

12

6
 6/03/2023

Excess Return and Risk Premium

Risk-free assets
Debt without default risk held to maturity
Treasury bonds, commercial papers, bank deposits, etc.
Risk-free return rf,t
All risk-free assets are assumed to have the same return rf.
Their actual returns are (almost) identical.
rf,t is known at the start of a period, i.e. risk-free.
rf,t changes over time.
Excess return = rt – rf,t
Risk premium = E(rt) – rf,t = expected excess return
Depends on investors’ risk tolerance.
Relatively stable over time.

13

Concept Check 5.3 Holding 30 bonds

14

7
 6/03/2023

Concept Check 5.3

Exp ret = 0.2*0.0278 + 0.5*0.1 + 0.3*0.1778

= 0.1089
Risk prem = 0.1089 – 0.05 = 0.0589
If interest rate is low, the unexpected return is
0.1778 – 0.1089 = 0.0689

15

Sample Return Statistics

Sample average return r̅ and variance s2
Skewness/asymmetry: skew = ∑
skew > 0: long right tail; skew < 0: long left tail

Kurtosis: kurtosis = ∑
Kurtosis of normal distribution = 3
Excess kurtosis = ∑ -3
If excess kurtosis > 0, return distribution has fatter tails,
i.e. higher probabilities, than normal distribution
The central limit theorem
In large samples, the average return has near-normal
distributions.

16

8
 6/03/2023

S&P 500 Daily Return (MV weighted)

0.3

0.25

0.2

0.15

0.1

0.05

0
-0.06 -0.04 -0.02 0 0.02 0.04 0.06

Daily returns are not independent across stocks. The return of the S&P
500 Index is the weighted average returns of 500 stocks. Its distribution
is approximately normal, confirming the CLT.

17

Queensland Mining Corp (QMN)

A mining firm with market cap A$7m
No trading on most days.
High probability of extreme returns
0.6

0.5

0.4

0.3

0.2

0.1

0.0
‐0.1
‐0.09
‐0.08
‐0.07
‐0.06
‐0.05
‐0.04
‐0.03
‐0.02
‐0.01
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
More

18

9
 6/03/2023

Risk Measures
Return variance in T periods, e.g. days.
σ2 = ∑ r r̅ : an unbiased estimate of the average
daily variance over T days.
Lower partial standard deviation (LPSD)
rt = observed returns, rf,t = risk-free rate, t = 1,…,T.
Risk is when rt < rf,t, not when rt rf,t.
Dummy for low return: Dt = 1 if rt < rf; Dt = 0 otherwise.
/
LPSD = ∑ r r D
Tail risk: BKM page 140
Value at risk
Expected shortfall

19

Risk-Adjusted Return
Investors care about excess return per unit of risk
Average excess return in T periods
r̅ ∑ r r, r̅ r̅
rf,t = risk-free rate in period t.
Sharpe ratio =
Excess return per unit of total risk
Sortino ratio =
Excess return per unit of down-side risk

20

10
 6/03/2023

US T-bill Return Distribution

21

US 30-Year T-bond Return Distribution

22

11
 6/03/2023

US Individual Stock Return Distribution

23

US Stock Index Return Distribution

24

12
 6/03/2023

Return and Risk of US Treasuries and Stocks

25

US Stock Return Volatility

26

13
 6/03/2023

Risk Premium in 1900-2017

27

Readings and Exercises

Readings
Ch 5: 5.1, 5.2 (p123-125), 5.3, 5.4, 5.6-5.7 (p141-145)
Recommended exercises
Ch 5: 3, 6, 7, 11, 18 (a & b); CBA: 1, 4, 6.

28

14