Outline and readings

Weeks Topic Elements and reading

1-3 A. Basics 1. Real options (CWS 9; Topic A.1 Slides)
a. Valuing simple real options
b. Valuing compound options
c. Risk-neutral probabilities
2. The efficient markets hypothesis (CWS 10-11; Topic A.2 slides)
a. Market efficiency and the value of information
b. Rational expectations
c. Testing the EMH
4-5 B. Corporate 1. Roles, objectives and decisions (CWS 13; Topic B slides)
Financial 2. Capital structure (CWS 15; SI:86-113)
Policy 3. Pay-out policy (CWS 16; SII:1-34)
6-8 C. Applying 1. Hidden information and Actions (CWS 12; SII:35-55)
the theory a. Basic problem
b. Contract theory
c. Signalling
2. Choice of projects – risk-shifting, debt overhang (SII:56-69)
3. Executive compensation and effort (SII:70-101; SIII:1-11)
4. Competition and efficiency (SIII:1-11)
5. Changes in structure including M&A (CWS 18; SIII:12-34)
9-10 D. Topics 1. Automated financial markets
(subject 2. Brexit- and Covid-related topics
to change)

EC334 Topics in Financial Economics 2

 Definitions
• Corporate finance:
– Capital budgeting decisions – types and proportions
of real investments the corporation chooses
– Capital structure – financial instruments used
to finance investment
– Investor decisions - their impact on asset rates of return
and therefore cost of capital to corporations

EC334 Topics in Financial Economics 3

 Assets
• Asset - a claim on money linked to specific dates t, states (and places).
• Arrow security is one unit at a given date, place and state
– All assets are bundles of Arrow securities
• Equity - Common and preferred stock
– Claim on dividends (usually per share) on specific (regular or extraordinary) dates
– Claim also entitles the owner to vote on some issues (investment decisions, executive
pay, corporate ownership…)
– The claim can be sold for capital gains/losses
• Debt – corporate bonds, fixed-income securities
– Claim on periodic fixed payments (interest) and/or terminal payments on a given date
– Claim is enforced by law - if the firm does not pay, it is bankrupt
• Bank debt
– Like corporate debt, but generally not traded
– Short term, sometimes renegotiable
• Derivatives – generic term for compound assets
(e.g. debt is a derivative)

EC334 Topics in Financial Economics 4

 Net Present Value of the firm
• Value of the firm - the expected net present discounted value of the
returns to its activities
– linked to the value of owning the firm (equity)
• Which activities will the (management of the) firm choose?
– This is something management knows and investors anticipate
– They may not agree, so values may differ
• NPV (Net Present Value) rule:
– Firm estimates future revenues
and investment and other costs,
discounts at a suitable opportunity cost rate
(WACC= weighted average cost of capital)
– Pick projects that offer positive NPV.
• RoR (Rate of Return) rule:
– Estimate projects’ breakeven rate
(IRR= ‘internal’ rate of return)
– Pick projects with IRR > WACC

EC334 Topics in Financial Economics 5

 Actually, the value is non-monotone:
NPV
£600.00

£400.00

£200.00
IRR=4.588% IRR=74.028%
£0.00
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

(£200.00)

(£400.00)

(£600.00)
IRR is:
• Based on assumption that returns are reinvested at IRR
• Harder to use than rate of return or NPV based on cost of capital
• Better for e.g. internal planning
• Opportunity cost of a project is IRR of best unfunded project
• Can be extended to markets …
… but only if uncertainty is independent or ignored.

EC334 Topics in Financial Economics 7

 Real (and financial) options
• Example 1.1 – a project that can be expanded
if it goes well.
• Example 1.2 – a project that can be deferred

EC334 Topics in Financial Economics 8

 Evaluating a risky project or asset
• There are various ways to evaluate risky choices, depending on available
information. We work a simple example, then discuss more complex choices.
• Consider a project with current value that will be worth (with probability ) or
(with probability ) next period. The riskless discount rate is .
• One way to find is to discount next period’s expected payoff back to the present
using a risk-adjusted discount rate . In finance, this is often referred to as the
weighted average cost of capital (WACC). The corresponding equation is:

• If we know all but one of the variables in this equation, we can solve for the
remaining one. The above equation shows how to compute
– To find the WACC, compute
– It is often hard to find the WACC; using market risk-adjusted rates means matching the risks of the
choice at hand. Suppose we had independent estimates of , and and (usually the rate paid on gilts).
We can compute the risk-neutral probability () of getting the ‘good’ value which would give the
correct present value when used to compute the expected next period value:

EC334 Topics in Financial Economics 9

 Equivalent assets and replicating portfolios
• Where do we get the data from? In many problems, you will know the future
values of a choice (project, stock, etc.) in different states and will be interested in
finding , the current value. You can often do this using the law of one price: two
risky prospects that give the same return in every state must have the same value
– otherwise, you could buy (claims on) the cheaper one and sell (claims on) the
other for a riskless profit.
• In the above example, suppose you could find an equivalent asset; a stock
currently selling for per share, whose price will either rise to (with probability ) or
fall to (with probability ) next period. If

Then shares will exactly reproduce the pattern of returns of the choice, which is
thus worth .
• If the choice is enriched by options (e.g., to delay, expand or reverse the choice),
the pattern or returns will change. However, the equivalent asset can be combined
with riskless bonds to change the pattern of investment returns accordingly.

EC334 Topics in Financial Economics 10

 An option to delay a costly choice
• Consider a project (returning or next period) with a start-up cost of . The optimal decision
today is to pursue the project if and only if .
• As an alternative, you might be able to purchase an option to delay the decision until next
period, for a price of . This gives the following pattern of returns
– In the good state: now and next period
– In the bad state: now and next period
– In expectation: now and (next period)

– Then (net) returns next period would be in the good state and . This can be replicated by a portfolio
consisting of A shares of the equivalent asset and B riskless bonds if

– Suppose that the option would be exercised in both states; we can find the replicating portfolio
amounts of share and bonds:

– The cost of this portfolio in period 0 – and thus the value of the option – is

EC334 Topics in Financial Economics 11

 An option to delay a costly choice, 2
• Now suppose the option would only be exercised in the good state:

• Then (net) returns next period would be in the good state and otherwise. This can
be replicated by a portfolio consisting of A shares of the equivalent asset and B
riskless bonds if

– Suppose that the option would be exercised in both states; we can find the replicating portfolio
amounts of share and bonds:

– The cost of this portfolio in period 0 – and thus the value of the option – is
• Now let’s apply this numerically

EC334 Topics in Financial Economics 12

 Example 1.1: a project that can be expanded if it goes well
• The basic project will be worth £160M or £62.5M next year, with equal
probabilities
£160 M
Prob 0.5
£100 M
Prob 0.5
£62.5 M

Assumptions:
 The risk adjusted discount rate is 11.25%, giving a project value of £100

 The risk free rate is 5%, which would give

EC334 Topics in Financial Economics 13

 Option to Expand
• Project can be expanded by next year for additional (paid then).
• With this option, the project becomes
– {X,Y} where X = payoff if don’t expand, Y = payoff if expand
– Optimal choice in green (expand in ‘good state’):

Prob 𝑀𝑎𝑥 {£ 160 𝑀 ,1.5∗(£160 𝑀 )− £ 50 𝑀 }=£ 190 𝑀

Prob 𝑀𝑎𝑥 {£ 62.5 𝑀 ,1.5∗(£ 62.5 𝑀)− £50 𝑀 }=£62.5 𝑀

• How much is the option to expand worth?

Something else?

EC334 Topics in Financial Economics 14

 Risk Free Portfolio perspective
• We can separate the optimally exercised option to expand from the project as
show below. The option adds in one year in the good state and 0 otherwise.

£160 M
Prob 0.5 Prob 0.5 £30 M

£100 M
Prob 0.5 £62.5 M Prob 0.5
£0 M
Original project Expansion option
• Create a portfolio of the original project and n copies of a call option (C):
.

EC334 Topics in Financial Economics 15

 Risk Free Portfolio
• The value of the portfolio next year will be:
0.5
£ 160 𝑀 +𝑛∗(£ 30 𝑀 )
‘good’

𝑃𝑉 =£ 100 𝑀 +𝑛∗ 𝐶
0.5
‘bad’
£ 62.5 𝑀
• To make this risk free, pick to make returns the same in the ‘good’ and ‘bad’ states:

• The PV of the portfolio is therefore

• Because the portfolio is risk-free, we should discount it at the risk-free rate (5%)
• Compute C (effectively the value of the expansion option):

• The total value with the option is the project value plus C:

• To repeat, the value of the option is the value of the project with the option minus the value
of the project without the option.

EC334 Topics in Financial Economics 16

 Example 1.2a: a project that can be deferred
• An investment costs , payable next period
• It will pay or (equally likely) at the end of next period (net: )
• Evaluate by NPV, discounting at
– the expected NPV is negative () so investment should be rejected.
• Now consider a 2-period investment.
– “trial” investment of £2 today improves information of future success (if pursuing original project)
– If true state is good, then trial indicates probability of success next period
– If true state is bad, then trial indicates probability of success next period

Probability 2/3 +£50

-£2
Probability ½
Probability 1/3

-£55

-£2 Probability 1/3 +£50

Probability ½
Probability 2/3

-£55

EC334 Topics in Financial Economics 17

 Example 1.2b: deferral
• The expected present value is

• Why is this different from the project without ‘testing’?

– The first period lets you sample the state with relatively small stakes
– Invest if outcome of the trial is good - expected payoff to investment
– Don’t invest if outcome of the trial is bad - expected payoff to investment
– This is adaptive behaviour
– You can play with different combinations of
trial cost (here, £2)
trial payoff (here, £0 in both states)
and trial informativeness (here, in good/bad state)
(perfect trial gives 1 or 0 in good or bad states resp.)
To see when option of trial increases value.

EC334 Topics in Financial Economics 18

 Option types
• Option types - use learning about risky investments to
– Expand or shrink the investment size
– Extend or truncate the time-frame
– Delay decision (not delaying a project that has been accepted)

EC334 Topics in Financial Economics 19

 What are real options?
• Financial options (e.g. European puts and calls) have a deferred decision (the exercise
decision) – the value of flexibility is measured in terms of the expected gain from not buying
or selling today at the exercise price (if it was available)
• Real options are derived from an underlying – a physical asset whose value is affected by
management decisions; the underlying for a financial option is another security
• Financial options have no direct connection with the firm (they can be created as side bets
between any two people – a party and a counterparty on the other side)
• Two essential ingredients:
– Time (otherwise now-or-never like NPV – though implementation may be delayed or staged)
– Uncertainty (otherwise no reason not to nail down future decisions)
• The tale of Thales (2nd C BCE):
– He expected a big harvest of olives and offered his life savings to the owners of oil presses in
exchange for the right to rent their presses at the normal rate;
– They expected a normal harvest and hoped to
a) keep Thales’ savings and
b) rent their presses out at the normal rate as well;
– Thales was right, and rented the presses (which they needed)
at the ‘normal’ rate and renting them back to the owners
… at an extortionate rate!
• What was the underlying risky asset?
• Not the harvest, but the rental price of the presses
EC334 Topics in Financial Economics 23
 Differences between NPV and Real Options (RO)
• Discounting the future
– NPV uses a constant discount rate (WACC) for all future cash flows;
– RO disregards cash flows that would be avoided by optimal decisions, and
evaluates projects using replicating portfolios that give the same returns
as the optimally-managed project in every state.
• Choosing a delay period
– NPV analysis treats decisions to delay by one period, two periods, etc. as
mutually exclusive – choose the optimal precommitted delay at.
– Under RO you construct a decision tree, start at the end and work
backwards
– The RO value is higher than any of the fixed-period NPV values; under RO
there are no mutually exclusive alternatives, just a single value for the
American Call (the right to defer) and state-contingent rules for when (if
at all) to invest.
• But RO is more than just a decision tree (DT)
– DT payoffs are discounted at the same discount rate (WACC or )
– In RO, replicating portfolios and the law of one price effectively produce a
sequence of risk-adjusted rates – because the risk changes at each stage!

EC334 Topics in Financial Economics 28

 Example 1.5: Changing risks, changing rates

• Extend the previous example (1.4 a and b) by adding another period, to

compare Real Option (RO) and Decision tree (DT) approaches
• Consider holding the equivalent security for two periods instead of one;
– New ‘draw’ of the state before period 2;
– Discounting at WACC = 16.7% in each period, this is worth
 In period 0: (market cost of 1 share of equivalent security)
 In period 1:
• Good draw: £40 with 50% probability of original value
• Bad draw: £16 with 50% probability of original value
• [Expected: ]
 In period 2:
• Two good draws (prob. 25%) -
• One good, one bad draw (prob. 50%_ -
• Two bad draws (prob. 25%) -
• [Expected: £32.7]

EC334 Topics in Financial Economics 29

 Changing risks, changing rates example, 2
• If the required investment/the exercise price of the option () is the same whenever
the decision to invest is made, we get the following decision tree (on each branch
probabilities of good (G) or bad (B) states are both 50%):
– Note: Using the riskless rate, the decision tree would be worth (at WACC)
– Both approaches are wrong; choice of discount rate reflects an arbitrary assumption about riskiness.
– First ‘compare’ number is value of investing at that point; second is value of deferring for one more
period (or abandonment in the final period).
Compare £208.3 to 0:
G invest in period 2

Compare £75 to £80.9:

defer to period 2 WACC: 33.9%

G
B

Compare £8.3 to 0:
Compare -£5 to £31.4:
WACC: 33.6% invest in period 2
defer to period 1
G

B
Compare -£45 to £3: WACC: 38.3%
defer in period 2 Compare -£71.7 to 0:
B abandon option

EC334 Topics in Financial Economics 30

 How did we get the payoffs in the tree?
• We’ll work this out for the shaded decision node on the previous slide.
• Each (state-dependent) value is computed by constructing a replicating portfolio using the
equivalent security (ES - one unit is worth exactly of the project)
• Consider the shaded box on the previous slide (period 1 decision after good draw)
• Choose between
• exercising the call - invest £125for net payoff of
• deferring (keep option open) for value of
• How do we get deferral value? If we defer in period 1 we get the following in the final period
(see ex 1.4 for ES values; exercise price £125)
• In top state (G, then G = GG) state, project is worth (5 units of ES at each);
net return (if option lapses) so invest in period 2.
• In middle state (BG or GB), project is worth
(5 units of ES at each);
net return ,
so again invest in period 2

EC334 Topics in Financial Economics 31

 Computing payoffs in the tree
• Using the payoffs from the slide 31, form a replicating portfolio consisting
of M shares of the equivalent security and B risk-free bonds as before:

• This can be solved for M and B:

• M is the hedge ratio;
• Numerator = change in the option’s value between the two states of nature
• Denominator = change in value of equivalent security between the two states;
• 5 units will hedge the call (make it riskless)

EC334 Topics in Financial Economics 32

 Pricing the option
• To prevent arbitrage, PV of replicating portfolio must = value of the option.
• The prices of the two assets in the portfolio are £40 for the security and £1 for the
bond, so the value at the beginning of the period is:

• Repeat this process for each decision tree node to price the options they represent.
• The probabilities, payoffs and values (which obey the law of one price) are all known,
so we can also compute the risk adjusted discount rates for each node.
• Consider the decision node in the shaded box on slide 31: we have

 The risk of an option always exceeds that of the

underlying, so cash flows from the option are
discounted at a higher rate than the underlying.
 Also, a prospect with an option must
be worth at least as much as the underlying.

EC334 Topics in Financial Economics 33

 Real Option (RO) vs. decision tree (DT) analysis
• Both RO and DT allow adaptive behaviour; decisions conditioned on prior outcomes
• DT allows multiple outcomes at each node with associated probabilities; RO models
typically have two alternatives (e.g. ‘invest’ or ‘delay’) and suppress probabilities
(determined by optimal behaviour)
• DT generally uses the same (risk-adjusted) discount rate at each node; RO discount
rates vary (decisions take place in ‘real’ time, not thinking time)
– For example, the discount rate used to evaluate payoffs and the associated cost of capital (WACC) in the
business case for HS2 should change with oil prices (which affect costs of competing transport modes)

– When evaluating e.g. environmental policy, options that give future generations more control of their
own destiny (more flexibility) should discount their utility more steeply than one that limits their choices

EC334 Topics in Financial Economics 34

 What’s wrong with NPV and DT approaches?

In practice, managers often accept negative NPV projects.

• They may value flexibility – which is ruled out by NPV approach
– we will consider other reasons (debt overhang, personal kickbacks) later
• ‘Real’ projects have at least 5 potential options available:
– American expansion - expand the investment if it is doing well (‘American’ means there
is not a fixed ‘decision date’, just an upper limit)
– American extension - extend project life if it is doing well
– Contraction - shrink the project if it is doing badly
– Abandonment - abort the project if it is doing badly
– Deferral - delay the decision
• NPV ignores these; because each option potentially avoids a loss, NPV
undervalues every project (unless other parties are involved)
• Related to discounting – riskier payoffs
discounted steeply (they may not occur)
• Learning
limits (downside) risk
expands (upside) potential gains
EC334 Topics in Financial Economics 35
 Example 1.6: risk neutral probabilities
A risky project generates cash flows next year of £70 or £40 with equal probability.

1. Adjust for risk using 50% probability and a risk adjusted discount rate (assume 10%). £70
0.5

The value of the project is:

PV
0.5

£40

2. Adjust for risk by using the risk-free rate (5%) and adjusting the probability to . £70
The value of the project is: p
PV
1-p
£40

The two formulations must give the same value, hence

EC334 Topics in Financial Economics 38

 Risk Neutral Probability
• The Risk Neutral Probability () gives the same PV when we discount cash flows ()
using the risk-free discount rate () as discounted cash flows using the objective
(real) probabilities () and discounting at the risk-adjusted rate ():

• can be computed from the existing relationship between the risk-adjusted discount
rate, the objective probabilities, the cash flows and the Present Value.
• The risk neutral probability method is equivalent to the replicating portfolio
method but is easier to manage for complicated problems.
• Note: risk-neutral probabilities are not “real” probabilities;
they don’t reflect the actual odds of any particular cash flow.
They are simply another way
of determining the project’s market value.

EC334 Topics in Financial Economics 39

 Example 1.7: Risk neutral probabilities and options
𝑉 𝑢=£ 160 𝑀
• This method can also be used to determine 𝑝 𝑟𝑛
the value of a project with options. 𝑉 0=£ 100 𝑀
1 −𝑝 𝑟𝑛
𝑉 𝑑 =£ 62.5 𝑀
The risk neutral probability satisfies:

=£30 M
𝑝 𝑟𝑛
• We can compute the value of the call through:
𝐶0
1 −𝑝 𝑟𝑛 =£0 M

EC334 Topics in Financial Economics 40

 Value of the expansion option
• As with example 1.1, we can evaluate 𝑟𝑛 𝑉 𝑢=£ 190 𝑀
𝑝
an option to expand:
𝑉0
𝑟𝑛
1 −𝑝 𝑉 𝑑 =£ 62.5 𝑀

– Note: this uses the prn computed without the option to expand (why?)
• The risk neutral probability approach gives the same results as both the
replicating portfolio and replicating portfolio approaches
• The risk neutral probability (prn )is constant throughout the project’s timeline
• There is a fixed relationship among cash flows, NPV, prn and .
• Given any 3, we can solve for the fourth.

EC334 Topics in Financial Economics 42

 Example 1.8 Binomial options and Black-Scholes
• A firm can invest in a project over two years.
• Assumptions:
– The value of the project today is
– In each period, value will increase by or fall by
– Note: so
– Next year the project will be worth or
– In two years, the value will be , or .
– WACC is
– Risk free discount rate is
• The firm has an option to expand by at a cost of in two years.
• What is the value of this option?

EC334 Topics in Financial Economics 43

 Solution by Risk Neutral Probability
• Model the underlying asset
£1821
p

£1350
1-p
p

£1000
£1000

p
1-p

£741

1-p
£549

EC334 Topics in Financial Economics 44

 Solve using risk neutral probability
V (1  r )  Vd 1000*1.07  741
p 0  0.54045
Vu  Vd 1350  741
 Model underlying asset
 Model exercise of the options Max(1821,1821*1.3-250)=£2118.8
p
 Determine risk neutral
probability p that incorporates p
£1521.2
1-p
the project’s risk in each node.
£1097.4
Max(1000,1000*1.3-250)=£1050
 Solve binomial tree by rolling p
back project payoffs discounted 1-p
£766.1
at the risk-free rate.
1-p
 One advantage; in most cases Max(549,549*1.3-250=£548.8

the probability p is the same for

all the nodes in the project.

EC334 Topics in Financial Economics 45

 Advantages

 Both the binomial and the Black-Scholes models can be used to

value options
 Black-Scholes only works for a limited set of problems
 The binomial model is more flexible and allows you to model and
resolve more, and more complicated practical problems, especially
American options
 Solving using risk neutral probabilities is also much simpler than
solving through the replicating portfolio

EC334 Topics in Financial Economics 46