MFE Study Guide (Fall 2007)

Notes from McDonald’s Derivative Markets

Written by Colby Schaeffer

MFE Study Guide Fall 2007

Introduction
The material in this quick study guide has been done to the best
of my knowledge. Some topics are only covered briefly (Delta-Hedging
and Caps/Floors) while other topics have been omitted (Equity Linked
Annuities and Compound Options w/One Discrete Dividend). All exam
tips are marked in red! This the 1st Edition of my MFE study guide, and
it may be updated in the future with better content. Comments and
questions may be directed via PM to colby2152 on the Actuarial
Outpost.

Acknowledgements: Day Yi, Abraham Weishus, Bill Cross and AO

Member “Jraven”

Written by Colby Schaeffer 2/17

MFE Study Guide Fall 2007

Chapter 9 - Parity and Other Option Relationships

Option Exercise Style

• American: any time
• European: end of maturity
Value of otherwise identical options: European < American

Put-Call Parity
General Formula: Call(K, T) – Put(K, T) = PV(FO,T – K)
K: strike price
T: exercise time
Put-Call parity usually fails for American-style options.

Currency: C(K, T) – P(K, T) = x0e-r€T – Ke-rT

Stock: C(K, T) – P(K, T) = S0 – PV0,T(DIV) – e-rTK
Bond: C(K, T) – P(K, T) = B0 – PV0,T(Coupons) – e-rTK
Different Assets: C(St, Qt, T – t) – P(St, Qt, T – t) = PV[Ft,T(S) - Ft,T(Q)]

x0: current exchange rate denominated as $/€

r€: euro-denominated interest rate
r: American or implied interest rate
DIV: stream of dividends paid on stock

Early Exercise for American Options

• American-style call options on a nondividend-paying stock should
never be exercised prior to expiration.
• Early exercise is not optimal if: C(St, K, T – t) > St – K
• When exercising calls just prior to a dividend, early exercise is
not optimal at any time where: K - PVt,T(K) > PVt,T(DIV)

Arbitrage Inequalities (for both American & European)

THERE IS NO FREE LUNCH!

K1 < K2 < K3

0 ≤ C(K1) – C(K2) ≤ K2 – K1
0 ≤ P(K2) – P(K1) ≤ K2 – K1
*if options are European, then the difference in option premiums must
be less than the present value of the difference in strikes

Premiums decline at a decreasing rate as we consider calls with

progressively higher strike prices. Premiums also decline for puts but
when the strike price monotonically decreases.

Written by Colby Schaeffer 3/17

MFE Study Guide Fall 2007

Convexity of option price w.r.t. strike price

[C(K1) – C(K2)]/(K2 – K1) ≤ [C(K2) – C(K3)]/(K3 – K2)
[P(K2) – P(K1)]/(K2 – K1) ≥ [P(K3) – P(K2)]/(K3 – K2)

Option Trends
American options become more valuable as time to expiration
increases, but the value of European options may go up or down.

As the strike price increases for calls or decreases for puts, the options
become less valuable with their price decreasing at a decreasing rate.

With dividends, longer term European options may be less valuable

than shorter term European options.

Written by Colby Schaeffer 4/17

MFE Study Guide Fall 2007

Chapters 10, 11 - Binomial Pricing

• assumes that the stock price can change to either an upper
value or to a lower value

If the observed option price differs from its theoretical price, arbitrage
is possible…
u ≥ e(r – δ)h ≥ d

Δ: delta, the number of shares to replicate the option payoff

δ: dividend rate
σ: volatility, the standard deviation of the rate of return on stock

At the prices Sh = Su, Sd, a replicating portfolio will satisfy:

δ
(∆ × Su × e h ) + (B × erh) = Cu
δ
(∆ × Sd × e h ) + (B × erh) = Cd

Risk Neutral Probability (of increase in stock)

* e( r −δ ) h − d
p =
u−d
u = e( r −δ ) h+σ h

d = e ( r −δ ) h−σ h

Suppose that the continuously compounded expected return on the

stock is α and that the stock does not pay dividends.

Multiple periods: work with future values and compute option prices
retrospectively
*Take step-by-step answers to six digits!
C / P = e− rt E[C / Pbinomial (value)]

American Options
• For an American call, the value of the option at a node is given
by
Call Value = max[S – K *, e–rh(p* Value(Up) + (1 - p*)Value(Down)]

Written by Colby Schaeffer 5/17

MFE Study Guide Fall 2007

*switch to K – S for puts

• The valuation of American options proceeds as follows:

– At each node, we check for early exercise.
– If the value of the option is greater when exercised, we
assign that value to the node. Otherwise, we assign the
value of the option unexercised.
• Early Exercise: receive dividends, advance payment of strike
(interest), and lose insurance
Kr > Stδ Call goes Put goes up
down
Kr < Stδ Call goes up Put goes
down

Pricing Options on Other Assets

Stock Index – similar to nondividend-paying stocks
Currency – replace stock price with currency exchange rate and
dividend rate with foreign risk-free rate:
( r − r f ) h +σ h
u=e
Commodities – replace dividend rate with lease rate
Bonds – volatility decreases over time and interest rates are variable
Forwards – forwards aren’t risk-free…
σ h
u =e
Stocks Paying Discrete Dividends
The dividend is taken off the first node. The tree does not completely
recombine after a discrete dividend unless it is a percentage of the
stock. Another solution is to use:

Schroder’s Method
F = S – PV(Div)
S
σF = σS
F

Alternative Trees
Cox Ross-Rubinstein
u = eσ h

d = e −σ h

Lognormal

Written by Colby Schaeffer 6/17

MFE Study Guide Fall 2007

2) + σ h
u = e (r −δ− 0.5σ
2) −σ h
d = e (r −δ− 0.5σ

McDonald
Assumes S = Sud

Written by Colby Schaeffer 7/17

MFE Study Guide Fall 2007

Chapter 12 – Black-Scholes model

The Black-Scholes formula is a limiting case of the binomial formula for
the price of a European option.

ln(Se−δ t / Ke−rt ) + 0.5σ 2t

d1 =
σ t
d2 = d1 − σ t

Contrary to other forms of the d1 equation that you will see, this is the
only one that you need to know. Currency and Futures options replace
variables of this equation, but it remains the same.

C = Se−δ tN(d1) − Ke−rtN(d2 )

P = Ke−rtN(−d2 ) − Se−δ tN(−d1)

Where N(x) is the cumulative normal distribution function

Assumptions/Properties
• returns on stock are normally distributed and independent over
time
• volatility and risk-free rate are both known and constant
• future dividends are known
• there are no transaction costs/taxes

Currency Options
Replace stock price with currency exchange rate and the dividend rate
with foreign risk-free rate – known as Garman-Kohlhagen model

Futures
Replace stock price with forward price and the dividend rate with risk-
free rate.

Option Greeks
Formulas that express the change in the option price when an input to
the formula changes, taking all other inputs as fixed.

Δ, delta: option price change w.r.t stock price change

Delta is the only Greek that you are expected to compute
Δcall = e-δt N(d1)
Δput = e-δt N(-d1)
Δcall = Δput + e-δt
*Delta of a stock is always equal to 1

Г, gamma: measures convexity OR change in delta, always > 0

Written by Colby Schaeffer 8/17
MFE Study Guide Fall 2007

Vega: tests if volatility is sufficient, always > 0

θ, theta: option price change w.r.t. time to maturity change, usually <
0
ρ, rho: sensitivity to discounted strike price, + for call, - for put
ψ, psi: sensitivity to discounted stock, - for call, + for put

The Greek measure of a portfolio is the sum of the Greeks of the

individual portfolio components

Elasticity
• tells us the risk of the option relative to the stock in %terms
S∆
Ω=
C
For a call Ω ≥ 1, while for a put Ω ≤ 0
σ option = σ stock | Ω |

Risk Premium: γ – r = (α – r) Ω
α −r
Sharpe Ratio:
σ

Perpetual Options
σ 2x 2 + 2(r − δ − 0.5σ 2 )x − 2r = 0
Each x value is the present value of 1 when a stock of value S rises or
falls to price H, where the value is (S/H)x

h1 = lower value of x, and h2 = higher value of x

Calls
h1
S 
Value: (H − K )  
H 
 h 
Maximum H: H * = K  1 
 h1 − 1 

Puts – just reverse H and K and h1 and h2

Written by Colby Schaeffer 9/17

MFE Study Guide Fall 2007

Chapter 13 – Delta Hedging

Market makers want stable portfolios, so they use delta hedging as a
method of controlling risk.

Overnight Profit
OP = C 0 − C1 + ∆(S1 − S0 ) − (e r / 365 − 1)(∆S0 − C 0 )

Delta-Gamma-Theta Approximation
*Delta and Delta-Gamma approximations are contained within the
formula
1
C (St + h ) = C (St ) + ∆ε + Γε2 + h θ
2
Where: ε = St + h − St

Black-Scholes formula
This is different than the Black-Scholes EQUATION that was used for
pricing options. Rather, this equation is a function of the greeks, stock
price, volatility, and risk-free rate.

1 2 2
rC (S ) = σ S Γ + rS ∆ + θ
2
1
Var (Rh ,i ) = (S 2σ2Γh )2
2

Written by Colby Schaeffer 10/17

MFE Study Guide Fall 2007

Chapter 14 – Exotic Options

Asian Options – based on the arithmetic/geometric average of

underlying asset/strike price
*useful for hedging currency exchange, variable annuities, and
reducing volatility
Geometric(S) < Arithmetic(S)

Barrier Options – payoff depends if price of asset reaches a barrier

level
• payoff and option premium is less valuable than those of
standard options
1. Knock Out – option goes out of existence if price reaches
barrier
2. Knock In – option “comes into play” if price reaches barrier
3. Rebate – fixed payment if asset price reaches barrier

Knock-In + Knock-Out = Standard Option

Compound Option – option whose underlying asset is another option

that expires later

Compound Option Parity

x: strike price of compound option
t1: expiry of compound option
t2: expiry of underlying option

CallOnOption – PutOnOption = Option − xe−rt1

Gap Options – option with trigger K2 (price that option must be

exercised) and strike price K1 that differ, election is not optimal… Use
K1 for put-call parity

Exchange Options – lets you receive an asset in exchange for another

at time T, pays off only if the option asset outperforms the asset it is
being exchanged for

Volatility depends on both assets: σ = σ s2 + σ k2 − 2ρσ sσ k

Written by Colby Schaeffer 11/17

MFE Study Guide Fall 2007

Chapter 20 – Brownian Motion & Ito’s Lemma

Introduction of terms
1) Stochastic process is a random process that is also a function
of time.
2) Brownian motion is a continuous stochastic process
3) Diffusion process is Brownian motion where uncertainty
increases over time
4) Martingale is a stochastic process for which E[Z(t2)] = Z(t1) if t2
> t1

Arithmetic Brownian Motion

Z(0) = 0, Z(t + s) – Z(t) ~ N(0, s), Z(t) is continuous

dX(T) = α dt + σ dZ(t)

X(t) = X(a) + α (t − a) + σ t − aξ

σ: volatility or variance factor

α: drift factor

Ornstein-Uhlembeck Process
Variation of Arithmetic Brownian motion…

dX(t) = λ (α − X(t))dt + σ X(t)dZ

Geometric Brownian Motion

dX (t )
d ln[X (t )] = = α dt + σ dZ (t )
X (t )

2 )(t − a) + σ t − aξ
X(t) = X(a)e(α − 0.5σ

Ito’s Lemma
δC ∂ 2C δC
dC = dS + 0.5 2 (dS )2 + dt
δS ∂S δt

Multiplication Table

dt dZ
dt 0 0
dZ 0 dt

Written by Colby Schaeffer 12/17

MFE Study Guide Fall 2007

Sharpe Ratio
α −r
= “expected return per unit risk”
σ

If dS1 = α1S1dt + σ 1S1dZ

AND dS2 = α 2S2dt + σ 2S2dZ

α1 − r α 2 − r
THEN =
σ1 σ2

Written by Colby Schaeffer 13/17

MFE Study Guide Fall 2007

Chapter 24 – Interest Rate Models

A stochastic interest-rate model that assumes a flat yield curve cannot
be arbitrage-free.

Arithmetic: dr = α dt + σ dZ (similar to Arithmetic Brownian Motion)

Problems:
• r < 0 is possible
• drift is positive, so r can goto infinity
• volatility is independent of interest rate

Rendleman-Barter Model: dr = r α dt + r σ dZ (similar to Geometric

Brownian Motion)

Problems:
• drift sends the interest rate to infinity

Vasicek: dr = a(b − r )dt + σ dZ

Problems:
• volatility is independent

P (t ,T , r (t )) = A(t ,T )e −B (t ,T )r (t )
B 2σ 2
r (B (t ,T ) +t −T ) −
A(t ,T ) = e 4a

(1 − e −a (T −t ) )
B (t ,T ) =
a
2
σφ σ
r =b+ −
a 2a

φ: Sharpe Ratio
r : yield to maturity on infinitely lived bond

These are fairly extensive formulas to memorize, but there is no way

around it. Skip it if short on time or space in your head.

Cox-Ingersoll-Rand (CIR) Model: dr = a(b − r )dt + σ r dZ

CIR yield to maturity formulas are too much for an SOA exam which
says a lot.

Model doesn’t have problems like the other models have… “Mean
reversion” prevents interest rate from going to infinity.

Written by Colby Schaeffer 14/17

MFE Study Guide Fall 2007

Interest rate models allow us to generate stochastic yield curves, but

the models themselves may be too restrictive.

Binomial Interest Rate Model

All that is needed is a tree with short rates and p*. Unless given,
assume p* = 0.5.

Value bonds and options just like we did before, but it MUST BE
DISCOUNTED AT EACH NODE due to varying interest rates.

Types of Options
1) Calls/Puts – American/European
2) Caps/Floors
a. Strike Rate
b. Notional Amount
c. Frequency of Payment
d. Length of contract

Caps & floors control risk and promote parity. Simply, caps are like
calls and floors are like puts. Caplet values are equal to the difference
in the strike rate and given interest rate at a node multiplied by the
notional amount.

Black-Derman-Toy Model
BDT Model is actually not that difficult. It is an binomial evaluation of
the yield curve calibrated to actual results.

Pricing follows the same method as binomial interest rate trees.

Written by Colby Schaeffer 15/17

 MFE Study Guide Fall 2007

MFE Equations Sheet

Put-Call Parity Risk Neutral Probability (of increase in

General Formula: C - P = PV(FO,T – K) stock)
Currency: C-P = x0e-r€T – Ke-rT e( r −δ ) h − d
*
Stock: C-P = S0 – PV0,T(DIV) – e-rTK p =
Bond: C - P = B0 – u−d
PV0,T(Coupons) – e-rTK u = e( r −δ ) h+σ h
Different Assets: C - P = PV[Ft,T(S) - Ft,T(Q)]
d = e ( r −δ ) h−σ h
Early exercise is not optimal if: C(St, K, T – t) > St
–K u ≥ e(r – δ)h ≥ d
At the prices Sh = Su, Sd, a replicating portfolio Cox Ross-Rubinstein
will satisfy: u = eσ h
δ
(∆ × Su × e h ) + (B × erh) = Cu
d = e −σ h

δ
(∆ × Sd × e h ) + (B × erh) = Cd
Lognormal
2) + σ h
Black-Scholes Pricing u = e (r −δ− 0.5σ
2) −σ h
ln(Se−δ t / Ke−rt ) + 0.5σ 2t d = e (r −δ− 0.5σ
d1 =
σ t
Schroder’s Method
d2 = d1 − σ t
F = S – PV(Div)
S
C = Se−δ tN(d1) − Ke−rtN(d2 ) σF = σS
F
P = Ke−rtN(−d2 ) − Se−δ tN(−d1)
Path-dependent options
Option Greeks Asian – based on average price
Δ = e-δt N(d1)
Barrier
Elasticity Knock-In + Knock-Out = Standard Option
S∆
Ω= Other Exotic Options
C
Compound
σ option = σ stock | Ω | CallOnOption – PutOnOption =
Risk Premium: γ – r = (α – r) Ω Option − xe−rt1
α −r
Sharpe Ratio:
σ Gap: use trigger in d1, strike in PC parity

Overnight Profit Exchange: Volatility depends on both

OP = C 0 − C1 + ∆(S1 − S0 ) − (e r / 365 − 1)(∆S0 − C 0 ) assets: σ = σ s2 + σ k2 − 2ρσ sσ k

Written by Colby Schaeffer 16/17

 MFE Study Guide Fall 2007

Perpetual Options
Delta-Gamma-Theta Approximation σ 2x 2 + 2(r − δ − 0.5σ 2 )x − 2r = 0
1 h1
C (St + h ) = C (St ) + ∆ε + Γε2 + h θ S 
2 Value: (H − K )  
Where: ε = St + h − St H 
 h 
1 2 2 Maximum H: H * = K  1 
rC (S ) = σ S Γ + rS ∆ + θ
2  h1 − 1 
1
Var (Rh ,i ) = (S 2σ2Γh )2
2 Vasicek: dr = a(b − r )dt + σ dZ

dX(T) = α dt + σ dZ(t) Cox-Ingersoll-Rand (CIR) Model:

dr = a(b − r )dt + σ r dZ

X(t) = X(a) + α (t − a) + σ t − aξ Black-Derman-Toy tree

(α − 0.5σ 2 )(t − a) + σ t − aξ
X(t) = X(a)e
Ito’s Lemma
δC ∂ 2C δC
dC = dS + 0.5 2 (dS )2 + dt
δS ∂S δt

Ornstein-Uhlembeck Process
Variation of Arithmetic Brownian motion…

dX(t) = λ (α − X(t))dt + σ X(t)dZ

Written by Colby Schaeffer 17/17