EPM944: Managing Risk and Uncertainty

Lecture 3: Interest rates and Present Value

analysis

Professor George Halikias

School of Mathematics, Computer Sciences and Engineering

City University London
Interest rates

I Suppose the amount P (“the principal”) is borrowed from a

bank and must be repayed after time T along with simple
interest rate r per time T .
I The amount to be repayed at time T is: P + rP = P(1 + r ).
I Suppose P must be repayed after one year with interest at
rate r per year compounded semi-annually.
I For the first half-year interest is charged at a rate r /2 per
half-year, i.e. amount owed at this time is P(1 + r /2). This is
regarded as new principal for the second half-year period with
interest rate r /2 per half-year.
I Thus amount owed at end of one-year period is P(1 + r /2)2 .
Example: Credit-card loan

I A credit-card company charges interest at rate of 18%

compounded monthly.
I This is equivalent to paying simple interest every month at a
rate of 18/12 = 1.5%.
I If amount P is initially charged, after one year amount owed:

P(1 + r /12)12 = P(1 + 0.015)12 = 1.1956P

I The “effective interest rate” is

1.1956P − P
reff = = 19.56%
P
per year.
Present value analysis

I Suppose one can borrow and lend money at a nominal interest

rate of r per period compounded periodically.
I What is the present worth of a payment v made at the end of
period i?
I Since a bank loan of v (1 + r )−i requires a payoff of v at
period i it follows that the present value of a payoff of v to be
made at time period i is v (1 + r )−i .
I The concept of present value allows us to compare different
income streams by reducing each stream to its present value.
Example: Present Value analysis
I Suppose that you are to receive payments at the end of each
of the next five years.
I Consider three interest rates: r1 = 10%, r2 = 20% and
r3 = 30%.
I Which of the following payment sequences is preferable?
P
A 12 14 16 18 20 x = 80
Pi i
B 16 16 15 15 15 x = 77
Pi i
C 20 16 14 12 10 i xi = 72
P5
I Present Value = i=1 (1 + r )−i xi

r A B C
0.1 59.21 58.60 56.33
0.2 45.70 46.39 45.69
0.3 36.49 37.89 38.12
I For large interest rates streams with up-front payments may
have higher PV despite being lower in total.
Note on Present Value Analysis

I Payment sequences can be compared according to their value

at any specified time.
I For instance, to compare them in terms of their time-5 values
in previous example:
5
X 5
X
(1 + r )5 (1 + r )−i xi = (1 + r )5−i xi
i=1 i=1

I Hence preference ordering remains invariant to time shifts.

Present Value equivalence
Let interest rate be r compounded annually. Any cash flow stream
a = (a1 , a2 , . . . , an ) that returns you ai at end of period i (for each
i = 1, 2, . . . , n) can be replicated by depositing
a1 a2 an
PV(a) = + 2
+ ... +
1+r (1 + r ) (1 + r )n

in a bank at time 0 and then successively withdraw a1 , a2 , . . . , an :

I Withdrawing a1 at end of year 1 would leave you with
 
a1 a2 an
(1 + r ) + + ... + − a1
1+r (1 + r )2 (1 + r )n

which is equal to
a2 an
+ ... +
(1 + r ) (1 + r )n−1

on deposit.
Present Value equivalence (cont.)
I Thus, withdrawing a2 at end of year 2 you would have:
 
a2 an a3 an
(1+r ) + ... + n−1
−a2 = +. . .+
1+r (1 + r ) (1 + r ) (1 + r )n−2
I In general withdrawing ai at end of year i (i < n) would leave
you with:
ai+1 an
+ ... +
(1 + r ) (1 + r )n−i
I This leaves an /(1 + r ) on deposit after withdrawing an−1 , just
enough to cover final withdrawal an at end of next year.
I Conversely the cash flow sequence a can be transformed into
initial capital PV(a) by borrowing this amount from a bank
and then using the cash-flow to pay-off the debt.
I Conclusion: Any cash flow is equivalent to initial reception of
its PV and hence a preferable to b iff PV(a) > PV(b).
Example: Machine replacement
I A company needs a certain type of machine for the next five
years.
I They presently own such a machine, now worth £6K but will
lose £2K in value in each of the next three years, after which
it will be worthless. The (beginning of the year)
yearly-operating cost is £9K, this amount expected to
increase by £2K in each subsequent year of use.
I A new machine can be bought at the beginning of any year
for a fixed cost of £22K. Its lifetime is 6 years and its value
decreases by £3K in each of the first two years of use and
then by £4K in each following year. The operating cost of a
new machine is £6K in the first year, with an increase of £1K
in each subsequent year.
I If interest rate is 10%, when should the company purchase a
new machine?
Example: Machine replacement (cont.)

Balance sheet (Decision: Buy at beginning of year 1)

Cash flow Justification

Year 1 £22K cost new machine bought
-£6K value old machine replaced
£6K operating cost (new)
——
£22K Total cost
Year 2 £7K operating cost (new)
Year 3 £8K operating cost (new)
Year 4 £9K operating cost (new)
Year 5 £10K operating cost (new)
Year 6 -£4K value new machine sold
(£22K-(£3K+£3K+£4K+£4K+£4K))
Example: Machine replacement (cont.)

Balance sheet (Decision: Buy at beginning of year 3):

Cash flow Justification

Year 1 £9K operating cost (old)
Year 2 £11K operating cost (old)
Year 3 £22K cost new machine bought
-£2K value old machine replaced
£6K operating cost (new)
——
£26K Total cost
Year 4 £7K operating cost (new)
Year 5 £8K operating cost (new)
Year 6 -£12K value new machine sold
(£22K-(£3K+£3K+£4K))
Example: Machine replacement (cont.) - Cash flows

Decision: Buy new machine at the beginning of Year i,

i = 1, 2, 3, 4. Assume r = 0.1.

Decision i/Yj Y1 Y2 Y3 Y4 Y5 Y6 PV
1 22K 7K 8K 9K 10K -4K 46.08K
2 9K 24K 7K 8K 9K -8K 43.79K
3 9K 11K 26K 7K 8K -12K 43.76K
4 9K 11K 13K 28K 7K -16K 45.63K

P6 xij
Note that PV(i) = j=1 (1+r )j−1 .

Therefore the company should purchase a new machine at the

beginning of year 3.
Geometric series

I Let: Sn = 1 + b + b 2 + . . . + b n . Then:

bSn = b(1 + b + b 2 + . . . + b n ) = b + b 2 + . . . + b n + b n+1

Hence:
1 − b n+1
(1 − b)Sn = 1 − b n+1 ⇒ Sn =
1−b
I If we let n → ∞ then b n → 0 as long as |b| < 1 and in this
case:
1
S∞ = lim Sn = 1 + b + b 2 + . . . =
n→∞ 1−b
Example: Pension plan

I An individual planning to retire in 20 years has decided to put

an amount A in the bank at the beginning of each of the next
240 months to enable her to withdraw £1, 000 at the
beginning of each of the following 360 months.
I Assuming the interest rate is 6% compounded monthly how
large should A be?
I Let r = 0.06/12 = 0.005 the monthly interest rate. With
1
β = 1+r the present value of her deposit is:

1 − β 240
PVd = A + Aβ + Aβ 2 + . . . + Aβ 239 = A
1−β
Example: Pension plan (cont.)

I Similarly, if W is the amount withdrawn in each of the

following 360 months, then the Present Value of all
withdrawals is
1 − β 360
PVw = W β 240 + W β 241 + . . . + W β 599 = W β 240
1−β
I Hence she will be able to (just) fund all withdrawals if
PVd = PVw , i.e.

1 − β 240 1 − β 360
A = W β 240 ⇒ A = £360.99
1−β 1−β

for W =£1000 and β = 1/1.005.

Example (perpetuity)
I A perpetuity entitles its holder to be paid a constant amount
c at the end of an infinite sequence of years, i.e. at the end of
each year i, i = 1, 2, . . ..
I If the interest rate is r compounded annually what is the PV
of such a cash flow sequence?
I Suppose A is deposited at the beginning of each rear yielding
interest rA after one year. If only the interest is withdrawn
(c = rA ⇒ A = c/r ) the capital is left intact and the process
can be perpetuated.
I This intuition can be checked mathematically by calculating
the PV of this (infinite) cash flow:
c c c
PV = + 2
+ + ...
1+r (1 + r ) (1 + r )3
 
c 1 1
= 1+ + + ...
1+r 1+r (1 + r )2
c 1 c
= 1
=
1 + r 1 − 1+r r
Example: Mortgage loan

I Suppose one takes a mortgage loan for the amount L to be

paid back over n months with equal payments A at the end of
each month.
I The interest rate is r per month compounded monthly.
I What is the value of A in terms of L, n and r ?
Example: Mortgage loan (cont.)

The PV of the n-month payment is

A A A
PV = + 2
+ ... +
1+r (1 + r ) (1 + r )n
 n
1
A 1 − 1+r A
1 − (1 + r )−n

= 1
=
1 + r 1 − 1+r r

This must equal the loan amount L, hence:

Lr L(α − 1)αn
A= =
1 − (1 + r )−n αn − 1

where α = 1 + r .
Rate of Return

I Consider an investment that for an initial payment a returns

amount b after one period. The rate of return of the
investment is the interest rate r which makes the PV of the
return equal to a, i.e.
b b
= a or r = − 1
1+r a
I More generally consider an investment that for an initial
payment a (a > 0) yields a string of non-negative returns
(b1 , b2 , . . . , bn ) where bi is received at end of period i and
bn > 0.
Rate of Return (cont.)
I The rate of return per period of the investment is the interest
rate such that the PV of the cash flow sequence is equal to
zero when compounded periodically at that interest rate.
I If we define the function P by
n
X
P(r ) = −a + bi (1 + r )−i
i=1

then the rate of return per period of the investment is the

value r ∗ > −1 such that P(r ∗ ) = 0.
I It follows from the assumptions that P(r ) is a strictly
decreasing function when r > −1 and since
limr →−1+ P(r ) = ∞, limr →∞ P(r ) = −a < 0 that the root
P(r ∗ ) = 0 is unique. Further
n
X n
X
P(0) = bi − a > 0 ⇔ bi > a ⇔ r ∗ > 0
i=1 i=1

and vice versa.

Example: Rate of return
I Find the rate of return from an investment that, for an initial
payment of £100, yields returns of £60 at the end of each of
the first two periods.
I The rate of return is the solution of
60 60
100 = +
1+r (1 + r )2
1
I Setting x = gives 60x 2 + 60x − 100 = 0 and hence
1+r
p √
−60 ± 602 + 4(60)(100) 27, 600 − 60
x= =
120 120
or x ≈ 0.8844 (x > 0 since r > −1, hence the positive root
was taken). Hence
1
1 + r∗ ≈ ≈ 1.131
0.8844
and the rate of return is approximately 13%.
NPV Analysis for investments’ evaluation

I This is based on discounted cash flows of the project including

capital costs, operating costs, R&D costs, depreciation rates,
tax returns, expected sales, etc.
I Use random distributions, random scenarios to model
uncertain parameters of each cash-flow component (e.g.
customer demand, exchange rate fluctuations, etc). This
makes NPV a random variable with a probability distribution.
NPV Analysis for investments’ evaluation (cont.)

I Estimate life-horizon of project and terminal value of project

assets (“sarvage value”). This could be negative (e.g.
potential legal liabilities or recycling costs).
I Discount factor is a critical parameter: Normally this is an
estimate of the ”opportunity cost” of capital, i.e. the
expected return demanded by investors purchasing securities
with risks comparable to the risks of the investment.
I Note: The higher the discount rate the more difficult it is for
revenues generated at later years to offset the capital costs,
normally arising in early stages.
NPV Analysis for investments’ evaluation (cont.)

I Alternatively discount rate can be estimated separately for

each cash flow component.
I Model can include various options or decisions (e.g.
management may decide to disinvest after a number of years
if revenues fall below a certain threshold). Each option or
decision effectively initiates a new NPV analysis.
I Carry out Monte-Carlo simulations over all uncertainty
parameters to obtain empirical probability distribution of NPV
based on which project viability is decided. Validate results via
sensitivity analysis.
Real option calculations using loss functions

I Suppose that the outcome of some investment is uncertain

but there is a guarantee that it will not be lower than a given
value a.
I This may be because we have the option of not going ahead
with the venture (in which case the outcome can never be less
than zero)
I Alternatively this may be because we can sell the investment
for a given amount a since we hold a sell option for this
amount.
I In any case if X is a r.v. giving the profit in the absence of
the option, then our expected profit, which guarantees an
amount a is: E(max(X , a)) = E(max(X − a, 0)) − a.
I We can work out the value of this quantity for different
distributions X .
Real option calculations using loss functions
I Suppose X is uniformly distributed in the interval (b, c) and
b < a < c:
Z c
1
E(max(X − a, 0)) = max(u − a, 0) du
b c −b
Z c
1
= (u − a)du
c −b a
 2 c
1 u
= − au
c −b 2 a
 2
a2

1 c 2
= − ac − +a
c −b 2 2
(c − a)2
=
2(c − b)
(a−b)2
I Similarly: E(max(a − X , 0)) = 2(c−b) (Check!)
Example
I A salvage company considers carrying out a difficult deep
water salvage operation.
I The salvage value of the boat is £3 million (20% of a £15
million insured value).
I There is a cost of £300, 000 to carry out a preliminary
investigation which will determine the actual cost of the
salvage operation.
I This cost is estimated to be between £2 million and £4
million and the company regards any amount between these
numbers as equally likely.
I If the cost is found to be too large then the company will not
go ahead with the salvage operation, but they will not recover
the £300, 000 preliminary costs.
I What is the expected value of this project to the salvage
company?
Example (cont.)
I Once the £300, 000 has been spent it is a question of whether
or not to go ahead with the salvage and this will be
worthwhile if the cost is no more than 3million.
I Hence the expected profit (in millions) is:
−0.3 + E(max(3 − X , 0))
where X is uniform in the interval (2, 4).
I From previous formula (a=3, b = 2, c = 4) this is equal to:
(3 − 2)2
−0.3 + = −0.05
2(4 − 2)
I Hence this is a project that is not worth going ahead with.
I Note: 50% of the time the costs are more than £3M making
the salvage not worth it (and thus giving a loss of £0.3M)
and the other 50% of the time costs are less than £3M, so
average profit is £0.5 − 0.3 = 0.2M.
I With equal chances of a £0.3M loss or a £0.2M profit the
company should not take the contract.