FNCE650 | Jan – Mar 2025 | LKCSB | SMU
Session 5
Plan for Session 5
• Case debriefing
• A tension between growth and value
• Measuring investment performance
• The risk-return plane
• A two-stock portfolio
• Diversification
FNCE650#05
CASE DEBRIEFING
Q1: The problem Stryker faces
• Problem
◦ Performance of some PCB suppliers had been unsatisfactory with
respect to quality, delivery, and responsiveness
◦ In fact, several of the suppliers were in a weak financial position, close
to bankruptcy
• Options for improvement
#1: Maintain the current out-sourcing strategy (with modifications like a
safety stock or dual sourcing)
#2: Boost reliability with a single partner supplier
#3: “In-sourcing” the production of PCB
Q2: Incremental free cash flow
• The objective is to compare “#3 in-sourcing” with “status quo”
• Only need the incremental free cash flow triggered by “#3”
Incremental Free Cash Flow = FCF(“in-sourcing”) − FCF(“status quo”)
• Underlying assumption: “status quo” and #3 are mutually exclusive
• What cash items might be affected by “in-sourcing”?
◦ decrease in purchases from contract manufacturers
◦ construction and equipment costs
◦ architectural and engineering fees
◦ in-sourcing manufacturing costs
◦ depreciation tax shields
◦ [decrease in working capital requirement]
◦ [assuming no change in revenue, as the case does not mention that]
Q2: Incremental free cash flow – start from earnings
2003 2004 2005 2006 … 2009
Status quo
Purchasing costs 0 10,238 10,238 11,774 … 20,152
In-sourcing
Purchasing costs 0 10,238 3,800 0 … 0
Manufacturing costs 0 1,712 7,848 9,554 … 13,651
Other expenses 278 0 0 0 … 0
Incremental EBIT −278 −1,712 −1,409 2,220 … 6,502
Incremental taxes −100 −616 −507 799 … 2,341
Incremental NOPLAT −178 −1,096 902 1,421 … 4,161
Q2: Incremental free cash flow – apply adjustments
• Two adjustments on incremental NOPLAT (ΔWCR ignored)
◦ adding back incremental Depreciation (linear)
◦ subtracting incremental CapEx
2003 2004 2005 2006 … 2009
Incremental NOPLAT −178 −1,096 −902 1,421 … 4,161
Incremental Depr. 0 591 591 591 … 479
Incremental CapEx 6,009 0 0 0 … 0
Incremental FCF −6,187 −505 −311 2,011 … 4,640
• Have we forgotten anything?
◦ No incremental cash flow, in or out, from the 8-acre land in Kalamazoo
◦ Implicit assumption: No alternative use and cannot be sold
Q2: Incremental free cash flow
$8,000
$6,000
$4,000
$2,000
$0
-$2,000
-$4,000
-$6,000
-$8,000
2003 2004 2005 2006 2007 2008 2009
Q3: NPV and IRR
2003 2004 2005 2006 … 2009
Incremental FCF −6,187 −505 −311 2,011 … 4,640
Discount rate 15% 15% 15% 15% … 15%
Discount factor 1.00 0.87 0.76 0.66 … 0.43
Present values −6,187 −439 −235 1,322 … 2,006
NPV −76
• IRR can be easily found using the “IRR()” function in Excel; otherwise
505 311 2011 4640
0 = −6187 − − + + … +
1+IRR 1+IRR 2 1+IRR 3 1+IRR 6
⟹ IRR ≈ 14.72%
Q3: NPV and IRR
$4,000
$3,000
$2,000
$1,000
$0
NPV = −$76
(k=15%)
-$1,000
IRR = 14.7%
(NPV = 0)
-$2,000
6% 9% 12% 15% 18% 21%
Q3: Payback period
$8,000 cumulative
FCF
$6,000
$4,000
FCF
$2,000 cumulative
discounted
$0 FCF
-$2,000
-$4,000
DPP is beyond
the forecasting
-$6,000
PP ≈ 4.6 years horizon
-$8,000
2003 2004 2005 2006 2007 2008 2009
More details on payback period
2003 2004 2005 2006 2007 2008 2009
Incremental FCF −6,187 −505 −311 2,011 2,952 3,558 4,640
Cumulative incr. FCF −6,187 −6,692 −7,003 −4,992 −2040 1,518 6,158
• The cumulative (incremental) FCF is the total amount of the free cash
flow generated by the project
◦ Starting from year 2008, it will become positive, meaning that the
project has covered all its initial investment
◦ The same can be said if using discounted FCF
• Here, Stryker will have recovered its investment by 2008; hence the
payback period is 5 years
• Assuming that the FCF is equally distributed throughout a year, the
payback period can be estimated as 4 + 2040/3558 ≈ 4.57 years
• Using DPP, the payback time is beyond the 6 years’ forecasting horizon
Q4: The business case
• The business case for “#3 in-sourcing” is mostly driven by the weak
performance, financially and quality-wise, of Stryker’s PCB suppliers
• These problems pose a risk factor to Stryker as revenue might be
negatively impacted in case suppliers fail to deliver
• The business case is therefore mostly about reducing risk
The business case is about reducing risk, not the risks associated with
the cash flow of in-sourcing, but the risks associated with Stryker’s overall
operation, which heavily hinges on the PCB supply. While NPV analysis is
mainly about value creation, it is intentionally conservative. Additional
adjustments to relax the penalizing assumptions tend to easily suggest a
“go” for in-sourcing.
Summary
• The NPV and the IRR analysis compares the option #3 “in-sourcing” to “status
quo”; our assumptions penalized the “NPV” a lot:
◦ Ignoring terminal value and effects on working capital
◦ Not adjusting the risk aspects
◦ Not accounting for the potential in increasing revenue
◦ …
• We go along such a hand-tied approach because the business case is not really
about value creation but about risk
◦ Given that #3 reduces the risk associated with PCB suppliers, as long as the
value destruction is not too much, we feel okay with the project
◦ Essentially, we set a lower bound for the value to be created/destroyed – close
to a worst scenario
• What happened
◦ The in-source PCB project was approved and implemented
◦ There turned out to be a small delay and the cost rose a bit
◦ Eventually the PCB facility performed as expected and delivered significant
value to Stryker
FNCE650#05
A TENSION BETWEEN
GROWTH AND VALUE
Beyond forecast horizon
… 2006 2007 2008 2009 2010 and beyond
Incremental FCF … 2,011 2,952 3,558 4,640 ???
• A terminal value (aka continuation value) is often assumed for free cash
flows beyond the forecast horizon
• Typically, it is a growing perpetuity with growth g:
FCFt+1
Vt =
WACC − g
which appears monotone increasing in the assumed growth g
• While the intuition appears sound (faster growth, higher value), it begs
the question: How growth is achieved?
• Key: Invest today to grow tomorrow! But the above formula does not
show this explicitly…
Some algebra (to gain insights)
−CapEx + Depr = change in PP&E
• Relabel FCF components: FCF = NOPLAT + Depr − CapEx − ΔWCR
This alternative formula says that FCF = NOPLAT − ΔPP&E − ΔWCR
is the true cash income (NOPLAT) less
the true cash invested (NIC) FCF = NOPLAT − NIC
ΔPP&E + ΔWCR = New Invested Capital
• Compare FCFt+1 with FCFt+2: The new investment last period, NICt+1, should
FCFt+1 = NOPLATt+1 − NICt+1 directly affect the additional net cash income,
ΔNOPLATt+2 (“invest today to grow tomorrow”)!
FCFt+2 = NOPLATt+2 − NICt+2 Empirical approximation:
= NOPLATt+1 + ΔNOPLATt+2 − NICt+2 ΔNOPLATt+2 ≈ NICt+1×R
Rewriting NOPLAT t+2 in terms of the last
NOPLATt+1 and the additional ΔNOPLATt+2 The “R” here is the return on NlC,
1 RONIC, an intrinsic character of
• Hence, FCFt+1 = NOPLATt+1 − ΔNOPLATt+2
R the company’s operation efficiency
1 ΔNOPLATt+2
= NOPLATt+1 1− ⋅
R NOPLATt+1
At the target growth, the
g
FCFt+1 = NOPLATt+1 1− additional ΔNOPLATt+2 must be
R g-percent of the last NOPLATt+1
A tension between growth and value
FCFt+1 1 − g/R
Vt = = NOPLATt+1
WACC − g WACC − g
• Example: NOPLATt+1 = $120 and WACC = 15%
900 Value, Vt
850 R=20%
800 R=15%
750
700 R=10%
0.0% 0.5% 1.0% 1.5% 2.0% 2.5% 3.0%
Growth, g
• Growing is costly: Short-term FCFs are consumed by NIC
◦ if R(ONIC) > WACC, growth creates value (Vt increases with g)
◦ if R(ONIC) < WACC, growth destroys value (Vt decreases with g)
Burning cash (value) to grow
There was a time when WeWork was a hot, venture-backed company.
[…] It was not until the company filed to go public that we really learned
how it had financed its impressive growth rates. And grow WeWork
did. It went from revenue of $436.1 million in 2016 to $886 million in 2017, $1.82
billion in 2018, and $1.54 billion in the first half of 2019. That growth came at a
massive cost. The company’s operating losses swelled from $931.8 million in
2017 to $1.69 billion in 2018, and then to $1.37 billion in the first half of 2019. In
short, WeWork grew quickly but had to burn piles of cash to do it.
WeWork’s bankruptcy is proof that its core business
never actually worked, TechCruch, November 8, 2023
Advice for calculating terminal value
1 − g/R
• Use the more insightful formula: Vt = NOPLATt+1
WACC − g
• With the explicit tension between Vt and g,
◦ you are forced to justify your implicit assumption on the key variable
(but usually hidden) RONIC – a key efficiency measure of operation
◦ you will not be able to blindly manipulate g to create value
• Reality check:
◦ For mature companies (large, long-lived), RONIC in general is not too
high, comparable with WACC
◦ For new, growing companies, RONIC can be high – but make sure you
have concrete competitive advantage to justify it
▪ technology breakthrough? natural monopoly? patents? …
◦ Compare with industry average (Is the company really a unicorn?
Why?)
FNCE650#05
MEASURING INVESTMENT
PERFORMANCE
Your investment performance
• Suppose you bought one share of Disney stock at $90.71 at the end of
2023. After receiving a dividend of $0.95 at the end of 2024, you the
share for $111.55
• How to measure the performance of such an investment?
• Dollar return
P1 + δ1 − P0 = $111.55 + $0.95 − $90.71 = $21.79
• What are the problems of such a measure?
◦ Time value of money inconsistency
◦ Dependent on the initial investment scale
⟹ Difficult to compare across investments
Percentage return
To be consistent with timing, we typically
choose a period to end when a dividend is paid.
• Holding period return (sometimes just briefed as “return”):
δ1+P1−P0 0.95 + 111.55 − 90.71
R= = = 24.02%
P0 90.71
◦ δ1/P0 is called dividend yield (assuming δ1 is paid at the end of the period)
◦ (P1 − P0)/P0 is called capital gains yield
⟹ holding period return R = dividend yield + capital gains yield
• Insight: What really is the holding period return? Rewrite the definition:
δ1+P1−P0 δ +P
R= ⟹ 0 = −P0 + 1 1
P0 1+R
◦ R is the IRR of “buying the stock and holding it for one period”
▪ It only depends on the realization of cash flows, not the discount rate
▪ It is also scale-free: not depending on how much was invested
• In finance, when the investment can be scaled up or down freely, we
prefer measuring performance by such holding period returns
$1 invested in 1951
$10,000.0
S&P 500
$1,000.0
$100.0
US 1-month T-bill
$10.0
$1.0
$0.1
1951 1961 1971 1981 1991 2001 2011 2021
Stock return is random
E[R] ≈ 10% per year
σ[R] ≈ 15% per year
Historical frequency
100
80 These numbers will
prove handy in some
60 back-of-the-envelope
calculations.
40
20
0
-15% -10% -5% 0% 5% 10% 15%
• Because stock return is random, we focus on its statistical properties
◦ Expectation: 1
E[R] = (R + R2 + … +RT) often abbreviated as “return”
T 1
◦ Volatility:
1 also called “risk”
σ[R] = [(R1 − μ)2 +(R2 − μ)2 + … + (RT − μ)2]
T−1
Future return as a random variable
• We are more interested in the future than in the history
• Since the future is unknown, we tend to think of it as a random variable
• If we denote the future return of an investment as R, then
◦ E[R] is the expectation of R, the “expected return” (or simply “return”)
▪ The difference, R − E[R], is (the realization of) the unexpected return
◦ σ[R] is the (expected) volatility of R, the “risk”
• Example: What can be an expectation of S&P500 return in 2025? How
volatile you think will it be?
◦ A naïve (but valid!) forecast:
E[R] = 10% and σ[R] = 15% – it is just an average year
◦ You can produce different forecasts, based on different information
Note: There are other important statistical moments like skewness and kurtosis. But we only
focus on E[R] and σ[R] here because we know R is approximately normally distributed.
FNCE650#05
THE RISK-RETURN PLANE
The risk-return plane
• Consider Disney (DIS) vs. Netflix (NFLX) for the coming year:
DIS: E[R] = 10.2% and σ[R] = 13.2%
NFLX: E[R] = 6.1% and σ[R] = 17.4%
• We can put the estimates as points in a risk-return plane:
E[R]
DIS We call it “risk-return”
10.2% • plane for simplicity, but
the return actually refers
to the “expected return”
NFLX
6.1% •
σ[R]
0 13.2% 17.4%
• As a normal investor, which stock do you prefer?
Investor preference on a risk-return plane
E[R]
better depends
G•
• C
• H
B A
• • •D
F• •I
depends •E worse
σ[R]
0
• Investors are generally risk-averse
◦ for the same expected return, they prefer less risk
◦ for the same risk, they prefer higher return
• Stocks in the upper-left dominate those in the lower-right
Indifference curves
• An investor’s preference for risk and return can be characterized as a set
of indifference curves (IC) in the risk-return plane
• ICs are a set of parallel lines that characterize the investor’s preference
between E[R] and σ[R]
• An investor is indifferent between
any two points on the same IC
E[R]
• For a risk-averse investor, his ICs
typically look like the one sketched
• to the left, but the curvature differs
from one individual to another
• Along the curve, the investor trades
• off higher expected return E[R] with
higher risk σ[R], hence “indifferent”
σ[R] • Investments on the upper-left are
strictly preferred to those on the
lower-right
FNCE650#05
A TWO-STOCK PORTFOLIO
A two-stock portfolio
• A portfolio is a weighted combination of some assets
• Example: Consider a portfolio P with 69% money invested in stock A and
31% in stock B, with the following characteristics:
Expected Return Risk
Stock A 8% 15%
Stock B 12% 19%
Correlation ρ[RA,RB] = 0.40
• We want to know: What are the risk-return characteristics of the
portfolio? That is, what are E[RP] and σ[RP]? How are they affected by
the weighting of the component stocks? Is there a “best” weight?
• But we first need to understand how the two stocks are related
◦ This is characterized by their correlation ρ[RA,RB]
Correlation (a quick review)
• The correlation between two random variables, X and Y, is defined as
cov[X,Y]
ρ[X,Y] =
σ[X]σ[Y]
where cov[X,Y] = E[XY] − E[X]E[Y] is the covariance between X and Y
• ρ[X,Y] (= ρ[Y,X]) is always between −1 and +1:
X and Y are … On average, if X ↑↑
ρ[X,Y] = +1 perfectly positively correlated Y↑↑
0 < ρ[X,Y] < +1 positively correlated Y↑
ρ[X,Y] = 0 uncorrelated Y=
−1 < ρ[X,Y] < 0 negatively correlated Y↓
ρ[X,Y] = −1 perfectly negatively correlated Y↓↓
• If ρ[X,Y] > 0, does X cause Y or Y cause X to move together?
◦ Neither: correlation does NOT imply causation
Correlation ≠ causation
• There are several situations that can lead to ρ[X,Y] > 0:
◦ X causes Y
◦ Y causes X
◦ Z causes both
◦ Pure luck (“spurious” correlation)
• Most stocks are positively correlated
◦ Try it!
◦ Why? Pure luck?
▪ Later we will see the reason is a common driver, some “Z”
Risk and return of a two-stock portfolio
• Suppose a portfolio P has 2 stocks
◦ with returns RA and RB
◦ with respective weights wA and wB, such that wA + wB = 100%
• Then RP = wARA + wBRB and
E[RP] = wAE[RA] + wBE[RB]
σ[RP] = √(wAσ[RA])2 + 2ρ[RA,RB](wAσ[RA])(wBσ[RB])+ (wBσ[RB])2
• Apply to our example:
◦ E[RP] = 0.69×8% + 0.31×12% = 9.24%
◦ σ[RP] = (0.69×15%)2+2×0.69×0.31×0.4×15%×19%+(0.31×19%)2
= 13.81%
Varying portfolio weights: The opportunity set
• Construct portfolios with stocks A and B varying wA from 0% to 100%
• Then plot the portfolios into the risk-return plane:
E[R]
12% •
B, wA = 0%
10%
• P, wA = 69%
8% •
A, wA = 100%
σ[R]
6%
12% 14% 16% 18% 20%
• The opportunity set of a set of stocks is the collection of all possible
risk-return combinations of these stocks
◦ In the example above, it is the locus connecting point A and point B
FNCE650#05
REDUCING RISK THROUGH
DIVERSIFICATION
Reducing risk through diversification
• We have shown in the previous slide that
◦ while E[RP] is always between E[RA] and E[RB]
◦ the risk σ[RP] can be lower than the lower of σ[RA] and σ[RB]
• This effect is called diversification
• Risk-averse investors benefit from diversification because the
opportunity set expands up-and-leftward
• Can we do better? Two experiments:
◦ Varying the correlation
◦ Adding more stocks
Varying the correlation
E[R]
12% •B
10%
8% •
A
σ[R]
6%
0% 5% 10% 15% 20%
• As the correlation decreases (to −1), the opportunity set perturbs
towards more to the left, i.e. to the zero-risk axis
• This is favorable: Risk-averse investors prefer low-correlation securities!
Seeking negative correlation
Gold has long been a safe
haven in times of turbulence, 6500
S&P 500 Gold
3000
6000 index 2800
and that is no less true today.
5500 2600
Coronavirus, the war in Ukraine, 5000 2400
geopolitical tensions, inflation fears, 4500 2200
mounting global debt, high interest rates 4000 2000
and the banking crisis have all prompted 3500 1800
3000 1600
investors to re-evaluate safe-haven
2500 1400
assets. Gold has been the beneficiary. 2000 1200
The new gold boom: how long can it last?, 1500 1000
2019 2020 2021 2022 2023 2024
Financial Times, May 24, 2023
Gold prices have rallied sharply since the Hamas-Israel conflict broke
out, […] as investors flee to the haven asset. Prices of the precious
metal have surged as much as 10 per cent to $1,996 per troy ounce,
hitting a five-month high […] With the region at risk of tipping into wider conflict,
investors have bought gold, which is regarded as a store of value during times
of geopolitical and market uncertainty.
Gold rallies as geopolitical turmoil overshadows
rising bond yields, Financial Times, Oct 21, 2023
Adding another stock
• We add another stock, C, to the portfolio:
Expected Return Risk
Stock A 8% 15%
Stock B 12% 19%
Stock C 10% 16%
Correlations ρ[RA,RB] = ρ[RA,RC] = ρ[RB,RC] = 0.40
• How does the new opportunity set look like?
Adding another stock
E[R]
12% •B
10% •C
8% •
A
σ[R]
6%
11% 13% 15% 17% 19%
Adding another stock
E[R]
12% •B
•
10% •C
•
8% •
A
σ[R]
6%
11% 13% 15% 17% 19%
Adding another stock
E[R]
•B
•C
•
A
σ[R]
Indifference curves combined
• Given an opportunity set, a risk-averse investor’s optimal investment
choice lies on the most upper-left point that tangents with his/her IC
• In the example below, the optimal investment is shown in the red dot
E[R]
σ[R]
Diversification with more stocks
σ[RP]
diversifiable diversifiable total risk of the portfolio
roughly 15%
to 20%
(annualized)
non-
# of stocks
in portfolio
0 30
• Portfolio risk falls as the number of stock in the portfolio increases
◦ But there is a limit to risk-reduction through diversification
◦ Magic number: 30 stocks (for randomly picked stocks!)
• The eliminated is the diversifiable, firm-specific, or unsystematic risk
• The remaining is the non-diversifiable, market risk, or systematic risk
How does diversification work (and not work)?
• Loosely speaking, the total risk of a stock has 2 components:
1) A diversifiable risk: firm-specific news (a strike, higher than expected
profits), industry-wide news (regulation, resource)
2) A market risk: non-diversifiable, systemic, caused by broad economic
and market news that affect all companies in the same direction
▪ Examples: monetary policy, market regulation…
• Portfolio diversification eliminates diversifiable risk because
◦ the firm-specific risks of all individual stocks cancel out
◦ good news is offset, on average, by bad news
• Portfolio diversification cannot reduce market risk because
◦ all stocks are exposed to the same market risk
◦ there is no “averaging out”
Diversification as a business strategy
• Some examples we have seen in FNCE650 already:
◦ “Bart 100” to help raise money for medical research (Session 1)
◦ Italian companies’ bonds outperformed their sovereign debt, because
they had “internationally diversified asset bases which would protect
them to some extent from the regional economic fallout” (Session 4)
Singapore has a new source of live broiler chickens – an avian influenza-
free farm in Indonesia. […] This is the first time live chickens have been
delivered to Singapore from Indonesia. Previously, Singapore imported
live chickens from only Malaysia. The move […] will diversify Singapore’s
import sources and strengthen the resilience of its chicken supply. Between
June and October in 2022, Malaysia halted the export of chickens to Singapore to
address surging prices and its shortage of poultry.
Singapore begins importing live broiler chickens
from Indonesia, The Straits Times, May 14, 2023
FNCE650#05
RISK REVISITED
How many correlations?
• In order to find a portfolio’s risk σ[R], we need information about the
components’ risks σ and their correlations ρ
How many σs? How many ρs?
a 1-stock portfolio 1 0
a 2-stock portfolio 2 1
a 3-stock portfolio 3 3
a 4-stock portfolio 4 6
a 5-stock portfolio 5 10
… … …
an n-stock portfolio n (n-1)n/2
• Examples:
◦ A 10-stock portfolio’s risk is determined by 10 σs and 45 ρs
◦ A 30-stock portfolio’s risk is determined by 30 σs and 435 ρs
◦ A 100-stock portfolio’s risk is determined by 100 σs and 4,950 ρs
What drives the risk of a portfolio?
• For a well-diversified portfolio:
◦ The risks of individual component assets do not matter much
◦ Their correlations matter a lot
⟹ It is the correlation structure that “determines” the risk
• General insight:
◦ How assets comove is much more important than how volatile they
individually are
◦ In certain occasions, only the non-diversifiable risk, i.e., the “correlation
risk,” matters
◦ Only such non-diversifiable “correlation risks” deserve compensation
FNCE650#05
OUTLOOK AND SUMMARY
Outlook
• This session introduces the basics of portfolio analysis: risk and return,
investor preference, and diversification
• In the next session, we shall explore the implications on our investment
decisions in the stock market
• To-dos
◦ Review basic statistics: random variable, expectation, mean, standard
deviation, correlation, covariance, and regressions
◦ Weekly Exercise #5
Summary
• Stock return is random and can be largely characterized by the first two
statistic moments: mean and volatility; we will be mainly interested in
their forecasts, E[R] and σ[R]
• We use the (expected) return-and-risk plane to describe stock returns
• Most investors (risk-averse) prefer investments to the upper-left of the
chart: higher return and lower risk
• A portfolio is a combination of assets (stocks, commodities, etc.)
• Diversification (reducing risks) can be achieved by
◦ introducing low-correlation assets
◦ simply adding more assets
• Diversification pushes the opportunity set to expand left-and-upward
• Intuition: When one asset’s return goes down, it can be averaged out by
some other assets’ return that go up
• The total volatility of a stock has two components: diversifiable risk and
(non-diversifiable) market risk
