R R ISK ISK & R & R ETURN ETURN The concept and measurement of Return: Realized and Expected return. Ex-ante and ex-post returns The concept of Risk: Sources and types of risk. Measurement of risk : Range, Std Deviation and Co-Efficient of Variation. Risk-return trade-off Risk-return trade-off 0 6 / 1 0 / 2 2 R i s k , R e t u r n a n d P o r t f o l i o T h e o r y
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
RRISKISK & R & RETURNETURN
The concept and measurement of Return:
Realized and Expected return. Ex-ante and ex-post returns
The concept of Risk: Sources and types of risk. Measurement of risk :
Range, Std Deviation and Co-Efficient of Variation.
The difference among the most important types of returns
How to estimate expected returns and risk for individual securities
What happens to risk and return when securities are combined in a portfolio
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
INTRODUCTION TO RISK AND RETURN
Tu
esd
ay, A
pril 1
1, 2
02
3R
isk, Re
turn
an
d P
ortfo
lio T
he
ory
INTRODUCTION TO RISK AND RETURNRisk and return are the two most
important attributes of an investment.
Research has shown that the two are linked in the capital markets and that generally, higher returns can only be achieved by taking on greater risk.
Risk isn’t just the potential loss of return, it is the potential loss of the entire investment itself (loss of both principal and interest).
Consequently, taking on additional risk in search of higher returns is a decision that should not be taking lightly.
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
Return %
RF
Risk
Risk Premium
Real Return
Expected Inflation Rate
RISK RETURN TRADE OFF
The concept of investment return is widely understood. For example, a 10% per annum return on a capital sum of $100,000 would result in $10,000 increase in value for the year. However, what exactly is ‘risk?’.
Risk is for the most part unavoidable – in life generally as much as in investing!
In investments, the term ‘risk’ is often expressed as ‘volatility’ or variations in returns.
In investment terms, the concept of ‘volatility’ is the measurement of fluctuation in the market values of various asset classes as they rise and fall over time.
The greater the volatility the more rises and falls are recorded by an individual asset class.
The reward for accepting greater volatility is the likely hood of higher investment returns over mid to longer term.
The disadvantage can mean lower returns in the shorter term. It must also be remembered that it can mean an increase or decrease in capital.
All investments involve some risk. In general terms the higher the risk, the higher the potential return, or loss. Conversely the lower the risk the lower the potential return, or loss.
The long-term risk/return trade off between different asset classes is illustrated in the following graph:
Risk, Return and Portfolio Theory04/11/23
Risk, Return and Portfolio Theory
04/11/23
MEASURING RETURNSRisk, Return and Portfolio Theory
Tu
esd
ay, A
pril 1
1, 2
02
3R
isk, Re
turn
an
d P
ortfo
lio T
he
ory
MANY DIFFERENT MATHEMATICAL DEFINITIONS OF "RETURNS"...
RETURNS =WHAT YOU'VE GOT WHAT YOU HAD TO BEGIN WITH, AS A
PROPORTION OF WHAT YOU HAD TO BEGIN WITH.
QUANTITATIVE RETURN MEASURES NECESSARY TO:
MEASURE PAST PERFORMANCE => "EX POST" OR HISTORICAL RETURNS;
MEASURE EXPECTED FUTURE PERFORMANCE => "EX ANTE" OR EXPECTED RETURNS.
MANY DIFFERENT MATHEMATICAL DEFINITIONS OF "RETURNS"...
TYPE 1: PERIOD-BY-PERIOD RETURNS . . .
"PERIODIC" RETURNSSIMPLE "HOLDING PERIOD RETURN"
(HPR)MEASURES WHAT THE INVESTMENT
GROWS TO WITHIN EACH SINGLE PERIOD OF TIME,
ASSUMING ALL CASH FLOW (OR VALUATION) IS ONLY AT BEGINNING AND END OF THE PERIOD OF TIME (NO INTERMEDIATE CASH FLOWS).
TYPE 1: PERIOD-BY-PERIOD RETURNS (CONT’D)
RETURNS MEASURED SEPARATELY OVER EACH OF A SEQUENCE OF REGULAR AND CONSECUTIVE (RELATIVELY SHORT) PERIODS OF TIME.
SUCH AS: DAILY, MONTHLY, QUARTERLY, OR ANNUAL RETURNS SERIES.
E.G.: RETURN TO IBM STOCK IN: 1990, 1991, 1992, ...
PERIODIC RETURNS CAN BE AVERAGED ACROSS TIME TO DETERMINE THE "TIME-WEIGHTED" MULTI-PERIOD RETURN.
TYPE 1: PERIOD-BY-PERIOD RETURNS (CONT’D)
THE PERIODS USED TO DEFINE PERIODIC RETURNS SHOULD BE SHORT ENOUGH THAT THE ASSUMPTION OF NO INTERMEDIATE CASH FLOWS DOES NOT MATTER.
TYPE 2: MULTIPERIOD RETURN MEASURES
PROBLEM: WHEN CASH FLOWS OCCUR AT MORE THAN TWO POINTS IN TIME, THERE IS NO SINGLE NUMBER WHICH UNAMBIGUOUSLY MEASURES THE RETURN ON THE INVESTMENT.
TYPE 2: MULTIPERIOD RETURN MEASURES (CONT’D)
NEVERTHELESS, MULTI-PERIOD RETURN MEASURES GIVE A SINGLE RETURN NUMBER (TYPICALLY QUOTED PER ANNUM) MEASURING THE INVESTMENT PERFORMANCE OF A LONG-TERM (MULTI-YEAR) INVESTMENT WHICH MAY HAVE CASH FLOWS AT INTERMEDIATE POINTS IN TIME THROUGHOUT THE "LIFE" OF THE INVESTMENT.
TYPE 2: MULTIPERIOD RETURN MEASURES (CONT’D)
THERE ARE MANY DIFFERENT MULTI-PERIOD RETURN MEASURES, BUT THE MOST FAMOUS AND WIDELY USED (BY FAR) IS:
THE "INTERNAL RATE OF RETURN" (IRR).
THE IRR IS A "DOLLAR-WEIGHTED" RETURN BECAUSE IT REFLECTS THE EFFECT OF HAVING DIFFERENT AMOUNTS OF DOLLARS INVESTED AT DIFFERENT PERIODS IN TIME DURING THE OVERALL LIFETIME OF THE INVESTMENT.
ADVANTAGES OF PERIOD-BY-PERIOD (TIME-WEIGHTED) RETURNS:
1)ALLOW YOU TO TRACK PERFORMANCE OVER TIME, SEEING WHEN INVESTMENT IS DOING WELL AND WHEN POORLY.
ADVANTAGES OF PERIOD-BY-PERIOD (TIME-WEIGHTED) RETURNS (CONT’D)
2)ALLOW YOU TO QUANTIFY RISK (VOLATILITY) AND CORRELATION (CO-MOVEMENT) WITH OTHER INVESTMENTS AND OTHER PHENOMENA.
ADVANTAGES OF PERIOD-BY-PERIOD (TIME-WEIGHTED) RETURNS (CONT’D)
3) ARE FAIRER FOR JUDGING INVESTMENT PERFORMANCE WHEN THE INVESTMENT MANAGER DOES NOT HAVE CONTROL OVER THE TIMING OF CASH FLOW INTO OR OUT OF THE INVESTMENT FUND (E.G., A PENSION FUND).
ADVANTAGES OF MULTI-PERIOD RETURNS:
1)DO NOT REQUIRE KNOWLEDGE OF MARKET VALUES OF THE INVESTMENT ASSET AT INTERMEDIATE POINTS IN TIME (MAY BE DIFFICULT TO KNOW FOR REAL ESTATE).
ADVANTAGES OF MULTI-PERIOD RETURNS (CONT’D)
2) GIVES A FAIRER (MORE COMPLETE) MEASURE OF INVESTMENT PERFORMANCE WHEN THE INVESTMENT MANAGER HAS CONTROL OVER THE TIMING AND AMOUNTS OF CASH FLOW INTO AND OUT OF THE INVESTMENT VEHICLE (E.G., PERHAPS SOME "SEPARATE ACCOUNTS" WHERE MGR HAS CONTROL OVER CAPITAL FLOW TIMING, OR A STAGED DEVELOPMENT PROJECT).
ADVANTAGES OF MULTI-PERIOD RETURNS (CONT’D)
NOTE: BOTH HPRs AND IRRs ARE WIDELY USED IN REAL ESTATE INVESTMENT ANALYSIS
CAN ALSO MEAN: "TOTAL YIELD", "YIELD TO MATURITY" THESE ARE IRRs, WHICH ARE TOTAL RETURNS,
NOT JUST INCOME. "BASIS POINT" = 1 / 100th PERCENT = .0001
CONTINUOUSLY COMPOUNDED RETURNS:
THE PER ANNUM CONTINUOUSLY COMPOUNDED TOTAL RETURN IS:
WHERE "Y" IS THE NUMBER (OR FRACTION) OF YEARS BETWEEN TIME "t-1" AND "t". )Yr(* V = CF+V Y)V( - )CF + V( = r t1-ttt1-tttt EXPLNLN
EXAMPLE:
01/01/98 V = 100003/31/99 V = 1100 & CF = 50 PER ANNUMr = (LN(1150) – LN(1000)) / 1.25
= 7.04752 – 6.90776= 11.18%
"REAL" VS. "NOMINAL" RETURNS
NOMINAL RETURNS ARE THE "ORDINARY" RETURNS YOU NORMALLY SEE QUOTED OR EMPIRICALLY MEASURED. UNLESS IT IS EXPLICITLY STATED OTHERWISE, RETURNS ARE ALWAYS QUOTED AND MEASURED IN NOMINAL TERMS. The NOMINAL Return is the Return in Current Dollars (dollars of the time when the return is generated).
REAL RETURNS ARE NET OF INFLATION. The REAL Return is the Return measured in constant purchasing power dollars ("constant dollars").
EXAMPLE:Suppose INFLATION=5% in 1992 (i.e., need
$1.05 in 1992 to buy what $1.00 purchased in 1991).
So: $1.00 in "1992$" = 1.00/1.05 = $0.95 in "1991$“
If rt = Nominal Total Return, year tit = Inflation, year tRt = Real Total Return, year t
Then: Rt = (1+rt)/(1+it) - 1 = rt - (it + it Rt ) rt - it ,
Thus: NOMINAL Return = REAL Return + Inflation Premium
Inflation Premium = it + it Rt It
IN THE CASE OF THE CURRENT YIELD
(Real yt)=(Nominal yt)/(1+it) (Nominal yt)
EXAMPLE:
1991 PROPERTY VALUE = $100,0001992 NET RENT = $10,0001992 PROPERTY VALUE = $101,0001992 INFLATION = 5%
WHAT IS THE REAL r, y, and g for 1992?
ANSWER:
Real g = (101,000/1.05)/100,000-1= -3.81% -4% (versus Nominal g=+1%)
Real y = (10,000/1.05)/100,000 = +9.52% 10% (versus Nominal y=10% exactly)
Real r = (111,000/1.05)/100,000-1=+5.71% 6% (versus Nominal r = 11%) = g + y =+9.52%+(-3.81%) 10% - 4%
RISK
INTUITIVE MEANING...THE POSSIBILITY OF NOT MAKING THE
EXPECTED RETURN:
rt Et-j[rt]
MEASURED BY THE RANGE OR STD.DEV. IN THE EX ANTE PROBABILITY DISTRIBUTION OF THE EX POST RETURN . . .
0%
25%
50%
75%
100%
-10% -5% 0% 5% 10% 15% 20% 25% 30%
Returns
Pro
bab
ility
A
B
C
Figure 1
C RISKER THAN B.B RISKIER THAN A.A RISKLESS.
WHAT IS THE EXPECTED RETURN? . . .
EXAMPLE OF RETURN RISK QUANTIFICATION:
SUPPOSE 2 POSSIBLE FUTURE RETURN SCENARIOS. THE RETURN WILL EITHER BE:+20%, WITH 50% PROBABILITYOR:-10%, WITH 50% PROBABILITY
"EXPECTED" (EX ANTE) RETURN
= (50% CHANCE)(+20%) + (50% CHANCE)(-10%)
= +5%
RISK (STD.DEV.) IN THE RETURN
= SQRT{(0.5)(20-5)2 + (0.5)(-10-5)2}= 15%
THE RISK/RETURN TRADEOFF..
INVESTORS DON'T LIKE RISK!
SO THE CAPITAL MARKETS COMPENSATE THEM BY PROVIDING HIGHER RETURNS (EX ANTE) ON MORE RISKY ASSETS . . .
Risk
ExpectedReturn
rf
RISK & RETURN:
TOTAL RETURN = RISKFREE RATE + RISK PREMIUM
rt = rf,t + RPt
RISK FREE RATE
RISKFREE RATE (rf,t)
= Compensation for TIME= "Time Value of Money" US Treasury Bill Return (For Real Estate, usually use Long Bond)
RISK PREMIUM
RISK PREMIUM (RPt):
EX ANTE: E[RPt]
= E[rt] - rf,t
= Compensation for RISK
EX POST: RPt
= rt - rf,t
= Realization of Risk ("Throw of Dice")
RELATION BETWEEN RISK & RETURN:
GREATER RISK <===> GREATER RISK PREMIUM (THIS IS EX ANTE, OR ON AVG. EX POST, BUT NOT NECESSARILY IN ANY GIVEN YEAR OR ANY GIVEN INVESTMENT EX POST)
EXAMPLE OF RISK IN REAL ESTATE:
PROPERTY "A" (OFFICE):VALUE END 1998 = $100,000POSSIBLE VALUES END 1999
$110,000 (50% PROB.)$90,000 (50% PROB.)
STD.DEV. OF g99 = 10%
EXAMPLE (CONT’D)
PROPERTY "B" (BOWLING ALLEY):VALUE END 1998 = $100,000POSSIBLE VALUES END 1999
$120,000 (50% PROB.)$80,000 (50% PROB.)
STD.DEV. OF g99 = 20%
EXAMPLE (CONT’D)
B IS MORE RISKY THAN A. T-BILL RETURN = 7%
EXAMPLE (CONT’D)
A: Office BuildingKnown as of end
1998 Value = $100,000 Expected value end
99 = $100,000 Expected net rent
99 = $11,000 Ex ante risk
premium = 11% - 7% = 4%
B: Bowling AlleyKnown as of end
1998 Value = $100,000 Expected value end
99 = $100,000 Expected net rent
99 = $15,000 Ex ante risk
premium = 15% - 7% = 8%
EXAMPLE (CONT’D) – SUPPOSE THE FOLLOWING OCCURRED IN 1999
A: Office BuildingNot known until end
1999 End 99 Value =
$110,000 99 net rent =
$11,000 99 Ex post risk
premium = 21% - 7% = 14%
(“The Dice Rolled Favorably”)
B: Bowling AlleyNot known until end
1999 End 99 Value =
$80,000 99 net rent =
$15,000 99 Ex post risk
premium = -5% - 7% = -12%
(“The Dice Rolled Unfavorably”)
SUMMARY:
THREE USEFUL WAYS TO BREAK TOTAL RETURN INTO TWO COMPONENTS...1) TOTAL RETURN = CURRENT YIELD
+ GROWTHr = y + g
2) TOTAL RETURN = RISKFREE RATE + RISK PREMIUM
r = rf + RP
3) TOTAL RETURN = REAL RETURN + INFLATION PREMIUM
r = R + (i+iR) R + I
"TIME-WEIGHTED INVESTMENT". . .
SUPPOSE THERE ARE CFs AT INTERMEDIATE POINTS IN TIME WITHIN EACH “PERIOD” (E.G., MONTHLY CFs WITHIN QUARTERLY RETURN PERIODS).
THEN THE SIMPLE HPR FORMULAS ARE NO LONGER EXACTLY ACCURATE.
"TIME-WEIGHTED INVESTMENT". . .
A WIDELY USED SIMPLE ADJUSTMENT IS TO APPROXIMATE THE IRR OF THE PERIOD ASSUMING THE ASSET WAS BOUGHT AT THE BEGINNING OF THE PERIOD AND SOLD AT THE END, WITH OTHER CFs OCCURRING AT INTERMEDIATE POINTS WITHIN THE PERIOD.
THIS APPROXIMATION IS DONE BY SUBSTITUTING A “TIME-WEIGHTED” INVESTMENT IN THE DENOMINATOR INSTEAD OF THE SIMPLE BEGINNING-OF-PERIOD ASSET VALUE IN THE DENOMINATOR.
"TIME-WEIGHTED INVESTMENT". . .
ii
i
CFwBegVal
CFBegValEndValr
where:= sum of all net cash flows occurring in period t,
wi = proportion of period t remaining at the time when net cash flow "i" was received by the investor.
(Note: cash flow from the investor to the investment is negative; cash flow from the investment to the investor is positive.)
iCF
EXAMPLE . . .
CF: Date:- 100 12/31/98+ 10 01/31/99+100 12/31/99
Simple HPR: (10 + 100-100) / 100
= 10 / 100= 10.00%
TWD HPR: (10 + 100-100) / (100 – (11/12)10)
= 10 / 90.83= 11.01%
IRR: = 11.00% . . .
EXAMPLE (CONT’D)
%00.111)0087387.1(/
%87387.0//1
100
/1
0
/1
101000
12
12
11
2
yrIRR
moIRRmoIRRmoIRRmoIRR j
j
THE DEFINITION OF THE "NCREIF" PERIODIC RETURN FORMULA . . . THE MOST WIDELY USED INDEX OF
PERIODIC RETURNS IN COMMERCIAL REAL ESTATE IN THE US IS THE "NCREIF PROPERTY INDEX" (NPI).
NCREIF = "NATIONAL COUNCIL OF REAL ESTATE INVESTMENT FIDUCIARIES“
“INSTITUTIONAL QUALITY R.E.” QUARTERLY INDEX OF TOTAL RETURNS PROPERTY-LEVEL APPRAISAL-BASED
NCREIF FORMULAFORMULA INCLUDES A TIME-WEIGHTED INVESTMENT DENOMINATOR, ASSUMING:
ONE-THIRD OF THE QUARTERLY PROPERTY NOI IS RECEIVED AT THE END OF EACH CALENDAR MONTH;
PARTIAL SALES RECEIPTS MINUS CAPITAL IMPROVEMENT EXPENDITURES ARE RECEIVED MIDWAY THROUGH THE QUARTER...
SUPPOSE YOU WANT TO KNOW WHAT IS THE RETURN EARNED OVER A MULTI-PERIOD SPAN OF TIME, EXPRESSED AS A SINGLE AVERAGE ANNUAL RATE?... YOU COULD COMPUTE THE AVERAGE OF THE HPRs ACROSS THAT SPAN OF TIME. THIS WOULD BE A "TIME-WEIGHTED" AVERAGE RETURN.
MULTI-PERIOD RETURNS (CONT’D)
IT WILL:=>Weight a given rate of return more
if it occurs over a longer interval or more frequently in the time sample.
=>Be independent of the magnitude of capital invested at each point in time; Not affected by the timing of capital flows into or out of the investment.
MULTI-PERIOD RETURNS (CONT’D)
YOU CAN COMPUTE THIS AVERAGE USING EITHER THE ARITHMETIC OR GEOMETRIC MEAN... Arithmetic average return over 1992-94:= (r 92+ r93+ r94)/3 Geometric average return over 1992-94:= [(1+r 92)(1+r93)(1+r94)]
(1/3) – 1
ARITHMETIC VS. GEOMETRIC MEAN…
Arithmetic Mean: => Always greater than geometric mean. => Superior statistical properties:
* Best "estimator" or "forecast" of "true" return.=> Mean return components sum to the mean total return=> Most widely used in forecasts & portfolio analysis.
ARITHMETIC VS. GEOMETRIC MEAN (CONT’D)
Geometric Mean:=> Reflects compounding ("chain-linking") of returns:
* Earning of "return on return". => Mean return components do not sum to mean total return
* Cross-product is left out. => Most widely used in performance evaluation.
ARITHMETIC VS. GEOMETRIC MEAN (CONT’D)
The two are more similar:- The less volatility in returns
across time- The more frequent the
return intervalNote: "continuously compounded"
returns (log differences) side-steps around this issue. (There is only one continuously-compounded mean annual rate: arithmetic & geometric distinctions do not exist).
CAN’T COMPUTE HPRs IF YOU DON’T KNOW ASSET VALUE AT INTERMEDIATE POINTS IN TIME (AS IN REAL ESTATE WITHOUT REGULAR APPRAISALS) SO YOU CAN’T COMPUTE TIME-WEIGHTED AVERAGE RETURNS.
You need the “IRR”.
IRR
SUPPOSE YOU WANT A RETURN MEASURE THAT REFLECTS THE EFFECT OF THE TIMING OF WHEN (INSIDE OF THE OVERALL TIME SPAN COVERED) THE INVESTOR HAS DECIDED TO PUT MORE CAPITAL INTO THE INVESTMENT AND/OR TAKE CAPITAL OUT OF THE INVESTMENT.
You need the “IRR”.
IRR
FORMAL DEFINITION OF IRR "IRR" (INTERNAL RATE OF RETURN) IS THAT SINGLE RATE THAT DISCOUNTS ALL THE NET CASH FLOWS OBTAINED FROM THE INVESTMENT TO A PRESENT VALUE EQUAL TO WHAT YOU PAID FOR THE INVESTMENT AT THE BEGINNING:
IRR
CFt = Net Cash Flow to Investor in Period "t“CF0 is usually negative (capital outlay).Note: CFt is signed according to the convention: cash flow from investor to investment is
negative, cash flow from investment to investor is
positive.Note also: Last cash flow (CFN) includes two
components: The last operating cash flow plus The (ex dividend) terminal value of the asset
("reversion").
)IRR + (1
CF + . . . +
)IRR + (1
CF + IRR) + (1
CF + CF = 0N
N2
210
WHAT IS THE IRR?...
(TRYING TO GET SOME INTUITION HERE . . .) A SINGLE ("BLENDED") INTEREST RATE, WHICH IF ALL THE CASH IN THE INVESTMENT EARNED THAT RATE ALL THE TIME IT IS IN THE INVESTMENT, THEN THE INVESTOR WOULD END UP WITH THE TERMINAL VALUE OF THE INVESTMENT (AFTER REMOVAL OF CASH TAKEN OUT DURING THE INVESTMENT):
WHAT IS THE IRR? (CONT’D)
where PV = -CF0, the initial cash "deposit" in the "account" (outlay to purchase the investment).
IRR is "internal" because it includes only the returns earned on capital while it is invested in the project.
Once capital (i.e., cash) is withdrawn from the investment, it no longer influences the IRR.
This makes the IRR a "dollar-weighted" average return across time for the investment, because returns earned when more capital is in the investment will be weighted more heavily in determining the IRR.
NNNN CFIRRCFIRRCFIRRPV
111 11
1
THE IRR INCLUDES THE EFFECT OF:
1. THE INITIAL CASH YIELD RATE (INITIAL LEVEL OF CASH PAYOUT AS A FRACTION OF THE INITIAL INVESTMENT;
2. THE EFFECT OF CHANGE OVER TIME IN THE NET CASH FLOW LEVELS (E.G., GROWTH IN THE OPERATING CASH FLOW);
3. THE TERMINAL VALUE OF THE ASSET AT THE END OF THE INVESTMENT HORIZON (INCLUDING ANY NET CHANGE IN CAPITAL VALUE SINCE THE INITIAL INVESTMENT WAS MADE).
IRR
THE IRR IS THUS A TOTAL RETURN MEASURE (CURRENT YIELD PLUS GROWTH & GAIN).
NOTE ALSO: THE IRR IS A CASH FLOW BASED RETURN
MEASURE... DOES NOT DIFFERENTIATE BETWEEN
"INVESTMENT" AND "RETURN ON OR RETURN OF INVESTMENT".
INCLUDES THE EFFECT OF CAPITAL INVESTMENTS AFTER THE INITIAL OUTLAY.
DISTINGUISHES CASH FLOWS ONLY BY THEIR DIRECTION: POSITIVE IF FROM INVESTMENT TO INVESTOR, NEGATIVE IF FROM INVESTOR TO INVESTMENT (ON SAME SIDE OF "=" SIGN).
IRR
In general, it is not possible to algebraically determine the IRR for any given set of cash flows. It is necessary to solve numerically for the IRR, in effect, solving the IRR equation by "trial & error". Calculators and computers do this automatically.
ADDITIONAL NOTES ON THE IRR . . .
TECHNICAL PROBLEMS: IRR MAY NOT EXIST OR NOT BE UNIQUE (OR GIVE MISLEADING RESULTS) WHEN CASH FLOW PATTERNS INCLUDE NEGATIVE CFs AFTER POSITIVE CFs. BEST TO USE NPV IN THESE CASES. (SOMETIMES “FMRR” IS USED.)
ADDITIONAL NOTES ON THE IRR (CONT’D)
THE IRR AND TIME-WEIGHTED RETURNS: IRR = TIME-WTD GEOMEAN HPR IF (AND ONLY IF) THERE ARE NO INTERMEDIATE CASH FLOWS (NO CASH PUT IN OR TAKEN OUT BETWEEN THE BEGINNING AND END OF THE INVESTMENT).
ADDITIONAL NOTES ON THE IRR (CONT’D)
THE IRR AND RETURN COMPONENTS: IRR IS A "TOTAL RETURN" IRR DOES NOT GENERALLY BREAK OUT EXACTLY INTO A SUM OF: y + g: INITIAL CASH YIELD + CAPITAL VALUE GROWTH COMPONENTS. DIFFERENCE BETWEEN THE IRR AND THE INITIAL CASH YIELD IS DUE TO A COMBINATION OF GROWTH IN THE OPERATING CASH FLOWS AND/OR GROWTH IN THE CAPITAL VALUE.
THE IRR AND RETURN COMPONENTS (CONT’D)
IF THE OPERATING CASH FLOWS GROW AT A CONSTANT RATE, AND IF THE
ASSET VALUE REMAINS A CONSTANT MULTIPLE OF THE CURRENT OPERATING
CASH FLOWS, THEN THE IRR WILL INDEED EXACTLY EQUAL THE SUM OF THE INITIAL
CASH YIELD RATE PLUS THE GROWTH RATE (IN BOTH THE CASH FLOWS AND THE ASSET CAPITAL VALUE), AND IN THIS CASE THE IRR
WILL ALSO EXACTLY EQUAL BOTH THE ARITHMETIC AND GEOMETRIC TIME-
WEIGHTED MEAN (CONSTANT PERIODIC RETURNS): IRRt,t+N=rt,t+N=yt,t+N+gt,t+N.
ADDITIONAL NOTES ON THE IRR (CONT’D)
THE IRR AND TERMINOLOGY: IRR OFTEN CALLED "TOTAL YIELD" (APPRAISAL) "YIELD TO MATURITY" (BONDS) EX-ANTE IRR = "GOING-IN IRR".
DOLLAR-WEIGHTED & TIME-WEIGHTED RETURNS:A NUMERICAL EXAMPLE . . .
"OPEN-END" (PUT) OR (CREF). INVESTORS BUY AND SELL "UNITS" ON THE BASIS OF THE APPRAISED VALUE OF THE PROPERTIES IN THE FUND AT THE END OF EACH PERIOD. SUPPOSE THE FUND DOESN'T PAY OUT ANY CASH, BUT REINVESTS ALL PROPERTY INCOME. CONSIDER 3 CONSECUTIVE PERIODS. . .
INVESTMENT PERIODIC RETURNS: HIGH, LOW, HIGH . . .
GEOM MEAN TIME-WTD RETURN = (1.089)(1/3)-1 = 2.88%
DOLLAR-WTD RETURN BEST FOR MEASURING INVESTOR PERFORMANCE IF INVESTOR CONTROLLED TIMING OF CAP. FLOW. TIME-WTD RETURN BEST FOR MEASURING PERFORMANCE OF THE UNDERLYING INVESTMENT (IN THIS CASE THE PUT OR CREF), AND THEREFORE FOR MEASURING INVESTOR PERFORMANCE IF INVESTOR ONLY CONTROLS WHAT TO INVEST IN BUT NOT WHEN.
MEASURING RETURNSINTRODUCTION
Ex Ante Returns Return calculations may be done ‘before-the-
fact,’ in which case, assumptions must be made about the future
Ex Post Returns Return calculations done ‘after-the-fact,’ in
order to analyze what rate of return was earned.
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
MEASURING RETURNSINTRODUCTION
According to the constant growth DDM can be decomposed into the two forms of income that equity investors may receive, dividends and capital gains.
WHEREAS
Fixed-income investors (bond investors for example) can expect to earn interest income as well as (depending on the movement of interest rates) either capital gains or capital losses.
Yield loss)(or Gain Capital Yield Dividend / Income
0
1
g
P
Dkc
Risk, Return and Portfolio Theory
MEASURING RETURNSINCOME YIELD
Income yield is the return earned in the form of a periodic cash flow received by investors.
The income yield return is calculated by the periodic cash flow divided by the purchase price.
Where CF1 = the expected cash flow to be received
P0 = the purchase price
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
yield Income 0
1
P
CF[8-1]
INCOME YIELD STOCKS VERSUS BONDS
Figure 8-1 illustrates the income yields for both bonds and stock in Canada from the 1950s to 2005
The dividend yield is calculated using trailing rather than forecast earns (because next year’s dividends cannot be predicted in aggregate), nevertheless dividend yields have exceeded income yields on bonds.
Reason – risk The risk of earning bond income is much less than the
risk incurred in earning dividend income.
(Remember, bond investors, as secured creditors of the first have a legally-enforceable contractual claim to interest.)
(See Figure 8 -1 on the following slide)
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
EX POST VERSUS EX ANTE RETURNSMARKET INCOME YIELDS
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
8-1 FIGURE
Insert Figure 8 - 1
MEASURING RETURNSCOMMON SHARE AND LONG CANADA BOND YIELD GAP
Table 8 – 1 illustrates the income yield gap between stocks and bonds over recent decades
The main reason that this yield gap has varied so much over time is that the return to investors is not just the income yield but also the capital gain (or loss) yield as well.
MEASURING RETURNSDOLLAR RETURNS
Investors in market-traded securities (bonds or stock) receive investment returns in two different form:
Income yield Capital gain (or loss) yield
The investor will receive dollar returns, for example: $1.00 of dividends Share price rise of $2.00
To be useful, dollar returns must be converted to percentage returns as a function of the original investment. (Because a $3.00 return on a $30 investment might be good, but a $3.00 return on a $300 investment would be unsatisfactory!)
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
MEASURING RETURNSCONVERTING DOLLAR RETURNS TO PERCENTAGE RETURNS
An investor receives the following dollar returns a stock investment of $25:
$1.00 of dividends Share price rise of $2.00
The capital gain (or loss) return component of total return is calculated: ending price – minus beginning price, divided by beginning price
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
%808.$25
$25-$27 return (loss)gain Capital
0
01
P
PP[8-2]
MEASURING RETURNSTOTAL PERCENTAGE RETURN
The investor’s total return (holding period return) is:
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
%1212.008.004.025$
25$27$
25$
00.1$
yield loss)(or gain Capital yield Income return Total
0
01
0
1
0
011
P
PP
P
CF
P
PPCF
[8-3]
MEASURING RETURNSTOTAL PERCENTAGE RETURN – GENERAL FORMULA
The general formula for holding period return is:
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
yield loss)(or gain Capital yield Income return Total
0
01
0
1
0
011
P
PP
P
CF
P
PPCF
[8-3]
MEASURING AVERAGE RETURNSEX POST RETURNS
Measurement of historical rates of return that have been earned on a security or a class of securities allows us to identify trends or tendencies that may be useful in predicting the future.
There are two different types of ex post mean or average returns used: Arithmetic average Geometric mean
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
MEASURING AVERAGE RETURNSARITHMETIC AVERAGE
Where:ri = the individual returnsn = the total number of observations
Most commonly used value in statistics Sum of all returns divided by the total number of
observations
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
(AM) Average Arithmetic 1
n
rn
ii
[8-4]
MEASURING AVERAGE RETURNSGEOMETRIC MEAN
Measures the average or compound growth rate over multiple periods.
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
11111(GM)Mean Geometric 1
321 -)]r)...(r)(r)(r [( nn[8-5]
MEASURING AVERAGE RETURNSGEOMETRIC MEAN VERSUS ARITHMETIC AVERAGE
If all returns (values) are identical the geometric mean = arithmetic average.
If the return values are volatile the geometric mean < arithmetic average
The greater the volatility of returns, the greater the difference between geometric mean and arithmetic average.
(Table 8 – 2 illustrates this principle on major asset classes 1938 – 2005)
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
MEASURING AVERAGE RETURNSAVERAGE INVESTMENT RETURNS AND STANDARD DEVIATIONS
Annual Arithmetic
Average (%)
Annual Geometric Mean (%)
Standard Deviation of Annual Returns
(%)
Government of Canada treasury bills 5.20 5.11 4.32Government of Canada bonds 6.62 6.24 9.32Canadian stocks 11.79 10.60 16.22U.S. stocks 13.15 11.76 17.54
Source: Data are from the Canadian Institute of Actuaries
Table 8 - 2 Average Investment Returns and Standard Deviations, 1938-2005
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
The greater the difference, the greater the volatility of
annual returns.
MEASURING EXPECTED (EX ANTE) RETURNS
While past returns might be interesting, investor’s are most concerned with future returns.
Sometimes, historical average returns will not be realized in the future.
Developing an independent estimate of ex ante returns usually involves use of forecasting discrete scenarios with outcomes and probabilities of occurrence.
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
ESTIMATING EXPECTED RETURNSESTIMATING EX ANTE (FORECAST) RETURNS
The general formula
Where:ER = the expected return on an investmentRi = the estimated return in scenario i
Probi = the probability of state i occurring
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
)Prob((ER)Return Expected 1
i
n
iir[8-6]
ESTIMATING EXPECTED RETURNSESTIMATING EX ANTE (FORECAST) RETURNS
Example:
This is type of forecast data that are required to make an ex ante estimate of expected return.
ESTIMATING EXPECTED RETURNSESTIMATING EX ANTE (FORECAST) RETURNS USING A FORMULA APPROACH
Example Solution:
Sum the products of the probabilities and possible returns in each state of the economy.
7.25%
)25.0(-25%0.5)(12% .25)0(30%
)Prob(r)Prob(r )Prob(r
)Prob((ER)Return Expected
332211
1i
n
iir
Risk, Return and Portfolio Theory
MEASURING RISKRisk, Return and Portfolio Theory
Tu
esd
ay, A
pril 1
1, 2
02
3R
isk, Re
turn
an
d P
ortfo
lio T
he
ory
RISK
Probability of incurring harm For investors, risk is the probability of earning an
inadequate return. If investors require a 10% rate of return on a given
investment, then any return less than 10% is considered harmful.
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
RISKILLUSTRATED
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
Possible Returns on the Stock
Probability
-30% -20% -10% 0% 10% 20% 30% 40%
Outcomes that produce harm
The range of total possible returns on the stock A runs from -30% to more than +40%. If the required return on the stock is 10%, then those outcomes less than 10% represent risk to the investor.
A
RANGE
The difference between the maximum and minimum values is called the range Canadian common stocks have had a range of annual
returns of 74.36 % over the 1938-2005 period Treasury bills had a range of 21.07% over the same
period. As a rough measure of risk, range tells us that
common stock is more risky than treasury bills.
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
DIFFERENCES IN LEVELS OF RISKILLUSTRATED
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
Possible Returns on the Stock
Probability
-30% -20% -10% 0% 10% 20% 30% 40%
Outcomes that produce harm The wider the range of probable outcomes the greater the risk of the investment.
A is a much riskier investment than BB
A
HISTORICAL RETURNS ON DIFFERENT ASSET CLASSES
Figure 8-2 illustrates the volatility in annual returns on three different assets classes from 1938 – 2005.
Note: Treasury bills always yielded returns greater than 0% Long Canadian bond returns have been less than 0% in
some years (when prices fall because of rising interest rates), and the range of returns has been greater than T-bills but less than stocks
Common stock returns have experienced the greatest range of returns
(See Figure 8-2 on the following slide)
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
MEASURING RISKANNUAL RETURNS BY ASSET CLASS, 1938 - 2005
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
FIGURE 8-2
REFINING THE MEASUREMENT OF RISKSTANDARD DEVIATION (Σ)
Range measures risk based on only two observations (minimum and maximum value)
Standard deviation uses all observations. Standard deviation can be calculated on forecast or
possible returns as well as historical or ex post returns.
(The following two slides show the two different formula used for Standard Deviation)
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
MEASURING RISKEX POST STANDARD DEVIATION
nsobservatio ofnumber the
year in return the
return average the
deviation standard the
:
_
n
ir
r
Where
i
Risk, Return and Portfolio Theory
1
)(post Ex 1
2_
n
rrn
ii
[8-7]
MEASURING RISKEXAMPLE USING THE EX POST STANDARD DEVIATION
ProblemEstimate the standard deviation of the historical returns on investment A that were: 10%, 24%, -12%, 8% and 10%.
Step 1 – Calculate the Historical Average Return
Step 2 – Calculate the Standard Deviation
%0.85
40
5
10812-2410 (AM) Average Arithmetic 1
n
rn
ii
Risk, Return and Portfolio Theory
%88.121664
664
4
404002564
4
2020162
15
)814()88()812()824(8)-(10
1
)(post Ex
22222
222221
2_
n
rrn
ii
EX POST RISKSTABILITY OF RISK OVER TIME
Figure 8-3 (on the next slide) demonstrates that the relative riskiness of equities and bonds has changed over time.
Until the 1960s, the annual returns on common shares were about four times more variable than those on bonds.
Over the past 20 years, they have only been twice as variable.
Consequently, scenario-based estimates of risk (standard deviation) is required when seeking to measure risk in the future. (We cannot safely assume the future is going to be like the past!)
Scenario-based estimates of risk is done through ex ante estimates and calculations.
Risk, Return and Portfolio Theory
RELATIVE UNCERTAINTYEQUITIES VERSUS BONDS
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
FIGURE 8-3
MEASURING RISKEX ANTE STANDARD DEVIATION
A Scenario-Based Estimate of Risk
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
)()(Prob anteEx 2
1i ii
n
i
ERr
[8-8]
SCENARIO-BASED ESTIMATE OF RISKEXAMPLE USING THE EX ANTE STANDARD DEVIATION – RAW DATA
MODERN PORTFOLIO THEORYRisk, Return and Portfolio Theory
Tu
esd
ay, A
pril 1
1, 2
02
3R
isk, Re
turn
an
d P
ortfo
lio T
he
ory
PORTFOLIOS
A portfolio is a collection of different securities such as stocks and bonds, that are combined and considered a single asset
The risk-return characteristics of the portfolio is demonstrably different than the characteristics of the assets that make up that portfolio, especially with regard to risk.
Combining different securities into portfolios is done to achieve diversification.
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
DIVERSIFICATION
Diversification has two faces:
1. Diversification results in an overall reduction in portfolio risk (return volatility over time) with little sacrifice in returns, and
2. Diversification helps to immunize the portfolio from potentially catastrophic events such as the outright failure of one of the constituent investments.
(If only one investment is held, and the issuing firm goes bankrupt, the entire portfolio value and returns are lost. If a portfolio is made up of many different investments, the outright failure of one is more than likely to be offset by gains on others, helping to make the portfolio immune to such events.)
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
EXPECTED RETURN OF A PORTFOLIOMODERN PORTFOLIO THEORY
The Expected Return on a Portfolio is simply the weighted average of the returns of the individual assets that make up the portfolio:
The portfolio weight of a particular security is the percentage of the portfolio’s total value that is invested in that security.
wA = weight of security A = $2,000 / $7,000 = 28.6%
wB = weight of security B = $5,000 / $7,000 = (1-28.6%)= 71.4%
RANGE OF RETURNS IN A TWO ASSET PORTFOLIO
In a two asset portfolio, simply by changing the weight of the constituent assets, different portfolio returns can be achieved.
Because the expected return on the portfolio is a simple weighted average of the individual returns of the assets, you can achieve portfolio returns bounded by the highest and the lowest individual asset returns.
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
RANGE OF RETURNS IN A TWO ASSET PORTFOLIO
Example 1:
Assume ERA = 8% and ERB = 10%
(See the following 6 slides based on Figure 8-4)
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
EXPECTED PORTFOLIO RETURNAFFECT ON PORTFOLIO RETURN OF CHANGING RELATIVE WEIGHTS IN A AND B
Risk, Return and Portfolio Theory
Ex
pe
cte
d R
etu
rn %
Portfolio Weight
10.50
10.00
9.50
9.00
8.50
8.00
7.50
7.00
0 0.2 0.4 0.6 0.8 1.0 1.2
8 - 4 FIGURE
ERA=8%
ERB= 10%
EXPECTED PORTFOLIO RETURNAFFECT ON PORTFOLIO RETURN OF CHANGING RELATIVE WEIGHTS IN A AND B
Risk, Return and Portfolio Theory
8 - 4 FIGURE
Ex
pe
cte
d R
etu
rn %
Portfolio Weight
10.50
10.00
9.50
9.00
8.50
8.00
7.50
7.00
0 0.2 0.4 0.6 0.8 1.0 1.2
ERA=8%
ERB= 10%
A portfolio manager can select the relative weights of the two assets in the portfolio to get a desired return between 8% (100% invested in A) and 10% (100% invested in B)
EXPECTED PORTFOLIO RETURNAFFECT ON PORTFOLIO RETURN OF CHANGING RELATIVE WEIGHTS IN A AND B
Risk, Return and Portfolio Theory
Ex
pe
cte
d R
etu
rn %
Portfolio Weight
10.50
10.00
9.50
9.00
8.50
8.00
7.50
7.00
0 0.2 0.4 0.6 0.8 1.0 1.2
8 - 4 FIGURE
ERA=8%
ERB= 10%
The potential returns of the portfolio are bounded by the highest and lowest returns of the individual assets that make up the portfolio.
EXPECTED PORTFOLIO RETURNAFFECT ON PORTFOLIO RETURN OF CHANGING RELATIVE WEIGHTS IN A AND B
Risk, Return and Portfolio Theory
Ex
pe
cte
d R
etu
rn %
Portfolio Weight
10.50
10.00
9.50
9.00
8.50
8.00
7.50
7.00
0 0.2 0.4 0.6 0.8 1.0 1.2
8 - 4 FIGURE
ERA=8%
ERB= 10%
The expected return on the portfolio if 100% is invested in Asset A is 8%.
%8%)10)(0(%)8)(0.1( BBAAp ERwERwER
EXPECTED PORTFOLIO RETURNAFFECT ON PORTFOLIO RETURN OF CHANGING RELATIVE WEIGHTS IN A AND B
Risk, Return and Portfolio Theory
8 - 4 FIGURE
Ex
pe
cte
d R
etu
rn %
Portfolio Weight
10.50
10.00
9.50
9.00
8.50
8.00
7.50
7.00
0 0.2 0.4 0.6 0.8 1.0 1.2
ERA=8%
ERB= 10%
The expected return on the portfolio if 100% is invested in Asset B is 10%.
%10%)10)(0.1(%)8)(0( BBAAp ERwERwER
EXPECTED PORTFOLIO RETURNAFFECT ON PORTFOLIO RETURN OF CHANGING RELATIVE WEIGHTS IN A AND B
Risk, Return and Portfolio Theory
8 - 4 FIGURE
Ex
pe
cte
d R
etu
rn %
Portfolio Weight
10.50
10.00
9.50
9.00
8.50
8.00
7.50
7.00
0 0.2 0.4 0.6 0.8 1.0 1.2
ERA=8%
ERB= 10%
The expected return on the portfolio if 50% is invested in Asset A and 50% in B is 9%.
%9%5%4
%)10)(5.0(%)8)(5.0(
BBAAp ERwERwER
RANGE OF RETURNS IN A TWO ASSET PORTFOLIO
Example 1:
Assume ERA = 14% and ERB = 6%
(See the following 2 slides )
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
RANGE OF RETURNS IN A TWO ASSET PORTFOLIOE(R)A= 14%, E(R)B= 6%
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
A graph of this relationship is found on the following slide.
Expected return on Asset A = 14.0%Expected return on Asset B = 6.0%
RANGE OF RETURNS IN A TWO ASSET PORTFOLIO E(R)A= 14%, E(R)B= 6%
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
Range of Portfolio Returns
0.00%2.00%4.00%6.00%8.00%
10.00%12.00%14.00%16.00%
Weight Invested in Asset A
Ex
pe
cte
d R
etu
rn o
n T
wo
A
ss
et
Po
rtfo
lio
EXPECTED PORTFOLIO RETURNSEXAMPLE OF A THREE ASSET PORTFOLIO
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
K. Hartviksen
Relative Weight
Expected Return
Weighted Return
Stock X 0.400 8.0% 0.03Stock Y 0.350 15.0% 0.05Stock Z 0.250 25.0% 0.06 Expected Portfolio Return = 14.70%
RISK IN PORTFOLIOSRisk, Return and Portfolio Theory
Tu
esd
ay, A
pril 1
1, 2
02
3R
isk, Re
turn
an
d P
ortfo
lio T
he
ory
MODERN PORTFOLIO THEORY - MPT
Prior to the establishment of Modern Portfolio Theory (MPT), most people only focused upon investment returns…they ignored risk.
With MPT, investors had a tool that they could use to dramatically reduce the risk of the portfolio without a significant reduction in the expected return of the portfolio.
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
EXPECTED RETURN AND RISK FOR PORTFOLIOSSTANDARD DEVIATION OF A TWO-ASSET PORTFOLIO USING COVARIANCE
))()((2)()()()( ,2222
BABABBAAp COVwwww
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
[8-11]
Risk of Asset A adjusted for weight
in the portfolio
Risk of Asset B adjusted for weight
in the portfolio
Factor to take into account comovement of returns. This factor
can be negative.
EXPECTED RETURN AND RISK FOR PORTFOLIOSSTANDARD DEVIATION OF A TWO-ASSET PORTFOLIO USING CORRELATION COEFFICIENT
))()()()((2)()()()( ,2222
BABABABBAAp wwww
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
[8-15]
Factor that takes into account the degree of
comovement of returns. It can have a negative value if correlation is
negative.
GROUPING INDIVIDUAL ASSETS INTO PORTFOLIOS The riskiness of a portfolio that is made of different
risky assets is a function of three different factors: the riskiness of the individual assets that make up the
portfolio the relative weights of the assets in the portfolio the degree of comovement of returns of the assets
making up the portfolio The standard deviation of a two-asset portfolio may
be measured using the Markowitz model:
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
BABABABBAAp wwww ,2222 2
RISK OF A THREE-ASSET PORTFOLIO
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
The data requirements for a three-asset portfolio grows dramatically if we are using Markowitz Portfolio selection formulae.
We need 3 (three) correlation coefficients between A and B; A and C; and B and C.
A
B C
ρa,b
ρb,c
ρa,c
CACACACBCBCBBABABACCBBAAp wwwwwwwww ,,,222222 222
RISK OF A FOUR-ASSET PORTFOLIO
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
The data requirements for a four-asset portfolio grows dramatically if we are using Markowitz Portfolio selection formulae.
We need 6 correlation coefficients between A and B; A and C; A and D; B and C; C and D; and B and D.
A
C
B D
ρa,b ρa,d
ρb,c ρc,d
ρa,c
ρb,d
COVARIANCE
A statistical measure of the correlation of the fluctuations of the annual rates of return of different investments.
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
)-)((Prob _
,1
_
,i BiB
n
iiiAAB kkkkCOV
[8-12]
CORRELATION
The degree to which the returns of two stocks co-move is measured by the correlation coefficient (ρ).
The correlation coefficient (ρ) between the returns on two securities will lie in the range of +1 through - 1.
+1 is perfect positive correlation-1 is perfect negative correlation
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
BA
ABAB
COV
[8-13]
COVARIANCE AND CORRELATION COEFFICIENT
Solving for covariance given the correlation coefficient and standard deviation of the two assets:
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
BAABABCOV [8-14]
IMPORTANCE OF CORRELATION
Correlation is important because it affects the degree to which diversification can be achieved using various assets.
Theoretically, if two assets returns are perfectly positively correlated, it is possible to build a riskless portfolio with a return that is greater than the risk-free rate.
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
AFFECT OF PERFECTLY NEGATIVELY CORRELATED RETURNSELIMINATION OF PORTFOLIO RISK
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
Time 0 1 2
If returns of A and B are perfectly negatively correlated, a two-asset portfolio made up of equal parts of Stock A and B would be riskless. There would be no variabilityof the portfolios returns over time.
Returns on Stock A
Returns on Stock B
Returns on Portfolio
Returns%
10%
5%
15%
20%
EXAMPLE OF PERFECTLY POSITIVELY CORRELATED RETURNSNO DIVERSIFICATION OF PORTFOLIO RISK
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
Time 0 1 2
If returns of A and B are perfectly positively correlated, a two-asset portfolio made up of equal parts of Stock A and B would be risky. There would be no diversification (reduction of portfolio risk).
Returns%
10%
5%
15%
20%
Returns on Stock A
Returns on Stock B
Returns on Portfolio
AFFECT OF PERFECTLY NEGATIVELY CORRELATED RETURNSELIMINATION OF PORTFOLIO RISK
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
Time 0 1 2
If returns of A and B are perfectly negatively correlated, a two-asset portfolio made up of equal parts of Stock A and B would be riskless. There would be no variabilityof the portfolios returns over time.
Returns%
10%
Returns on Portfolio5%
15%
20%
Returns on Stock B
Returns on Stock A
AFFECT OF PERFECTLY NEGATIVELY CORRELATED RETURNSNUMERICAL EXAMPLE
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
%10%5.2%5.7
) %5(.5)%15(.5)( n
1i
iip ERwER
Weight of Asset A = 50.0%Weight of Asset B = 50.0%
The potential of an asset to diversify a portfolio is dependent upon the degree of co-movement of returns of the asset with those other assets that make up the portfolio.
In a simple, two-asset case, if the returns of the two assets are perfectly negatively correlated it is possible (depending on the relative weighting) to eliminate all portfolio risk.
This is demonstrated through the following series of spreadsheets, and then summarized in graph format.
Risk of the portfolio is almost eliminated at 70% invested in asset A
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ryDiversification of a Two Asset Portfolio
Demonstrated Graphically
The Effect of Correlation on Portfolio Risk:The Two-Asset Case
Expected Return
Standard Deviation
0%
0% 10%
4%
8%
20% 30% 40%
12%
B
AB= +1
A
AB = 0
AB = -0.5
AB = -1
IMPACT OF THE CORRELATION COEFFICIENT
Figure 8-7 (see the next slide) illustrates the relationship between portfolio risk (σ) and the correlation coefficient The slope is not linear a significant amount of
diversification is possible with assets with no correlation (it is not necessary, nor is it possible to find, perfectly negatively correlated securities in the real world)
With perfect negative correlation, the variability of portfolio returns is reduced to nearly zero.
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
EXPECTED PORTFOLIO RETURNIMPACT OF THE CORRELATION COEFFICIENT
Risk, Return and Portfolio Theory
8 - 7 FIGURE
15
10
5
0
Sta
nd
ard
De
via
tio
n (
%)
of
Po
rtfo
lio
Re
turn
s
Correlation Coefficient (ρ)
-1 -0.5 0 0.5 1
ZERO RISK PORTFOLIO
We can calculate the portfolio that removes all risk. When ρ = -1, then
Becomes:
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
BAp ww )1( [8-16]
))()()()((2)()()()( ,2222
BABABABBAAp wwww [8-15]
AN EXERCISE TO PRODUCE THE EFFICIENT FRONTIER USING THREE ASSETSRisk, Return and Portfolio Theory
Tu
esd
ay, A
pril 1
1, 2
02
3R
isk, Re
turn
an
d P
ortfo
lio T
he
ory
AN EXERCISE USING T-BILLS, STOCKS AND BONDS
Base Data: Stocks T-bills BondsExpected Return(%) 12.73383 6.151702 7.0078723
Rational, risk averse investors will only want to hold portfolios such as B.
The actual choice will depend on her/his risk preferences.
DIVERSIFICATIONRisk, Return and Portfolio Theory
Tu
esd
ay, A
pril 1
1, 2
02
3R
isk, Re
turn
an
d P
ortfo
lio T
he
ory
DIVERSIFICATION We have demonstrated that risk of a portfolio can
be reduced by spreading the value of the portfolio across, two, three, four or more assets.
The key to efficient diversification is to choose assets whose returns are less than perfectly positively correlated.
Even with random or naïve diversification, risk of the portfolio can be reduced. This is illustrated in Figure 8 -11 and Table 8 -3
found on the following slides. As the portfolio is divided across more and more
securities, the risk of the portfolio falls rapidly at first, until a point is reached where, further division of the portfolio does not result in a reduction in risk.
Going beyond this point is known as superfluous diversification.
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
DIVERSIFICATIONDOMESTIC DIVERSIFICATION
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
8 - 11 FIGURE
14
12
10
8
6
4
2
0
Sta
nd
ard
Dev
iati
on
(%
)
Number of Stocks in Portfolio
0 50 100 150 200 250 300
Average Portfolio RiskJanuary 1985 to December 1997
Source: Cleary, S. and Copp D. "Diversification with Canadian Stocks: How Much is Enough?" Canadian Investment Review (Fall 1999), Table 1.
Table 8-3 Monthly Canadian Stock Portfolio Returns, January 1985 to December 1997
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
TOTAL RISK OF AN INDIVIDUAL ASSETEQUALS THE SUM OF MARKET AND UNIQUE RISK
This graph illustrates that total risk of a stock is made up of market risk (that cannot be diversified away because it is a function of the economic ‘system’) and unique, company-specific risk that is eliminated from the portfolio through diversification.
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
[8-19]
Sta
nd
ard
Dev
iati
on
(%
)
Number of Stocks in Portfolio
Average Portfolio Risk
Diversifiable (unique) risk
Nondiversifiable (systematic) risk
risk )systematic-(non Uniquerisk c)(systematiMarket risk Total [8-19]
INTERNATIONAL DIVERSIFICATION
Clearly, diversification adds value to a portfolio by reducing risk while not reducing the return on the portfolio significantly.
Most of the benefits of diversification can be achieved by investing in 40 – 50 different ‘positions’ (investments)
However, if the investment universe is expanded to include investments beyond the domestic capital markets, additional risk reduction is possible.
(See Figure 8 -12 found on the following slide.)
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
DIVERSIFICATIONINTERNATIONAL DIVERSIFICATION
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
8 - 12 FIGURE
100
80
60
40
20
0
Pe
rce
nt
ris
k
Number of Stocks
0 10 20 30 40 50 60
International stocks
U.S. stocks
11.7
SUMMARY AND CONCLUSIONS
In this chapter you have learned:How to measure different types of returnsHow to calculate the standard deviation
and interpret its meaningHow to measure returns and risk of
portfolios and the importance of correlation in the diversification process.
How the efficient frontier is that set of achievable portfolios that offer the highest rate of return for a given level of risk.
Risk, R
etu
rn a
nd
Po
rtfolio
Th
eo
ry
CONCEPT REVIEW QUESTIONSRisk, Return and Portfolio Theory
Tu
esd
ay, A
pril 1
1, 2
02
3R
isk, Re
turn
an
d P
ortfo
lio T
he
ory
CONCEPT REVIEW QUESTION 1EX ANTE AND EX POST RETURNS
What is the difference between ex ante and ex post returns?