This action might not be possible to undo. Are you sure you want to continue?

BooksAudiobooksComicsSheet Music### Categories

### Categories

Scribd Selects Books

Hand-picked favorites from

our editors

our editors

Scribd Selects Audiobooks

Hand-picked favorites from

our editors

our editors

Scribd Selects Comics

Hand-picked favorites from

our editors

our editors

Scribd Selects Sheet Music

Hand-picked favorites from

our editors

our editors

Top Books

What's trending, bestsellers,

award-winners & more

award-winners & more

Top Audiobooks

What's trending, bestsellers,

award-winners & more

award-winners & more

Top Comics

What's trending, bestsellers,

award-winners & more

award-winners & more

Top Sheet Music

What's trending, bestsellers,

award-winners & more

award-winners & more

P. 1

Option Pricing Incomplete Information|Views: 16|Likes: 1

Published by rahohe

See more

See less

https://www.scribd.com/doc/46299964/Option-Pricing-Incomplete-Information

01/29/2012

text

original

Mondher Bellalah

Professeur of finance, THEMA, University of Cergy, 33 boulevard du port,

95011 Cergy-Pontoise, (France). Mail: Mondher.Bellalah@eco.u-cergy.fr and

ESG Group, Paris

Sana Mahfoudh Besbes

Assistant professor in the University of Maine, UFR Exact Sciences, Le Mans

(France), and Phd Student, GREFI, University of Maine, UFR Management

sciences, Avenue OLivier Messiean, 72085 Le Mans (France), and Member

of THEMA. Mail: Sana.Mahfoudh@eco.u-cergy.fr

Abstract

Options are analyzed and valued in the context of Merton’s (1987) “Simple Model of Capital Market Equilibrium with Incomplete Information”. We show now the derivation of

the partial differential equation for options in the presence of shadow qcosts of incomplete information and stochastic volatility. We illustrate our approach by specific appli-

cations and show the dependancy of the option price on information and stochastic volatility. Then, we introduce information costs in a general diffusion model for asset prices

which allows the description of stochastic volatility in an incomplete market. As in Norbert, Platen and Schweizer (1992), we show that the investor’s choice of the minimal

equivalent martingale measure is not changing, but the process of the price of the asset depends on incomplete information.

Option Pricing Under Stochastic

Volatility with Incomplete Information

1 Introduction

Volatility is an important parameter in option pricing theory. Black and

Scholes (1973)proposed an option valuation equation under the assump-

tion of a constant volatility in a complete market without frictions.

Engle (1982) developed a discrete-time model, to show that the volatility

depends on its previous values.

The stochastic volatility problem has been examined by several

authors. For example, Hull and White (1987), Wiggins (1987), Johnson

and Shanno (1987) studied the general case in which the instantaneous

variance of the stock price follows some geometric process. Scott (1989)

and Stein and Stein (1991) used an arithmetic volatility in the study of

option pricing. All these models describe (with precision) the effects of

the volatility on the options prices. Stein and Stein (1991) and Heston

(1993) proposed a dynamic approach for the volatility which is repre-

sented by an Ornstein-Ulhenbeck. It is difficult to find an analytic solu-

tion for the stochastic volatility option pricing problem.

Merton (1987) proposed a capital asset pricing model in the presence

of the shadow costs of incomplete information. Bellalah (1990) applied

the Merton (1987) model to the valuation of options under incomplete

information. Bellalah and Jacquillat (1995) and Bellalah (1999) re derived

the Black and Scholes (1973) equation in the context of Merton (1987)

model, they obtained another version of the Black and Scholes equation

within information uncertainty.

In this paper, we propose a general context for the pricing of options

under stochastic volatility and information costs.

The first section provides a general concept for the valuation of

options with shadow costs when the volatility is random. The second

Wilmott magazine 51

Section examines some applications to several known models. The third

section investigates a general process of a compatible asset with incom-

pleteness in the market under information uncertainty and stochastic

volatility.

2 Valuation of Options In the Presence

of a Stochastic Volatility and Shadow

Costs of Incomplete Information

2.1 The valuation model

The pricing of derivative securities in the presence of a random volatil-

ity needs the use of two processes : one for the underlying asset and

one for the volatility. Consider the following dynamics for the under-

lying asset

dS = µSdt + σSdW

1

and the following process for the volatility

dσ = p(S, σ, t)dt + q(S, σ, t)dW

2

The two processes dW

1

dW

2

are Brownian-motions with a correlation

coefficient ρ. The functions p(S, σ, t) and q(S, σ, t) are specified in a way

that fits the dynamics of the volatility over time. Hence, the derivative

asset price V(S, σ, t) can be expressed as a function of the dynamics of the

underlying asset price S, the volatility σ and time t. Since the volatility is

not a traded asset, a problem arises because this new source of random-

ness can not be easily hedged away. The pricing of options in this context

needs the search for two hedging contracts. The first is the underlying

asset. The second can be an option that allows a hedge against volatility

risk. Following the same logic as in the original Black-Scholes model

(1973), consider a portfolio comprising a long position in the option V, a

short position of units of the underlying asset and a short position of

−

1

units of an other option with value V

1

(S, σ, t) :

= V − S −

1

V

1 (1)

Over a short interval of time dt, applying Ito’s lemma for the functions

S, σ and t gives the change in the value of this portfolio as:

d =

∂V

∂t

+

1

2

σ

2

S

2

∂

2

V

∂S

2

+ ρσqS

∂

2

V

∂S∂σ

+

1

2

q

2

∂

2

V

∂σ

2

dt

−

1

∂V

1

∂t

+

1

2

σ

2

S

2

∂

2

V

1

∂S

2

+ ρσqS

∂

2

V

1

∂S∂σ

+

1

2

q

2

∂

2

V

1

∂σ

2

dt

+

∂V

∂S

−

1

∂V

1

∂S

dS +

∂V

∂σ

−

1

∂V

1

∂σ

dσ

All the sources of randomness in the portfolio value resulting from dS

can be eliminated by setting the quantity before dS equal to zero, or

∂V

∂S

− −

1

∂V

1

∂S

= 0

and also by setting the quantity before dσ equal to zero, or

∂V

∂σ

−

1

∂V

1

∂σ

= 0

After eliminating the stochastic terms, the terms in dt must yield the

deterministic return as in a Black-Scholes “hedge” portfolio. Hence, the

instantaneous return on the portfolio must be the risk-free rate plus infor-

mation costs on each asset in the portfolio as in Bellalah (1999). This gives

d =

∂V

∂t

dt +

1

2

σ

2

S

2

∂

2

V

∂S

2

dt + ρσSq

∂

2

V

∂S∂σ

dt +

1

2

q

2

∂

2

V

∂σ

2

dt

−

1

∂V

1

∂t

dt +

1

2

σ

2

S

2

∂

2

V

1

∂S

2

dt + ρσSq

∂

2

V

1

∂S∂σ

dt +

1

2

q

2

∂

2

V

1

∂σ

2

dt

= [(r + λ

V

)V − (r + λ

S

)S − (r + λ

V1

)

1

V

1

]dt,

Isolating the terms in V and V

1

gives

∂V

∂t

+

1

2

σ

2

S

2

∂

2

V

∂S

2

+ ρσSq

∂

2

V

∂S∂σ

+

1

2

q

2

∂

2

V

∂σ

2

+(r + λ

S

)S

∂V

∂S

− (r + λ

V

)V

∂V

∂σ

=

∂V

1

∂t

+

1

2

σ

2

S

2

∂

2

V

1

∂S

2

+ ρσSq

∂

2

V

1

∂S∂σ

+

1

2

q

2

∂

2

V

1

∂σ

2

+(r + λ

S

)S

∂V

1

∂S

− (r + λ

V1

)V

1

∂V

1

∂σ

Since the two options differ by their strikes, payoffs and maturities, this

implies that both sides of the equation are independent of the contract

type. Since both sides are functions of the independent variables S, σ and

t, we have

∂V

∂t

+

1

2

σ

2

S

2

∂

2

V

∂S

2

+ ρσSq

∂

2

V

∂S∂σ

+

1

2

q

2

∂

2

V

∂σ

2

+ (r + λ

S

)S

∂V

∂S

− (r + λ

V

)V

= −(p − δq)

∂V

∂σ

,

for a function δ(S, σ, t) referred to as the market price for risk or volatility

risk. This equation can also be written as

∂V

∂t

+

1

2

σ

2

S

2

∂

2

V

∂S

2

+ ρσSq

∂

2

V

∂S∂σ

+

1

2

q

2

∂

2

V

∂σ

2

+ (r + λ

S

)S

∂V

∂S

+ (p − δq)

∂V

∂σ

− (r + λ

V

)V = 0

(2)

^

TECHNICAL ARTICLE 2

52 Wilmott magazine

This equation shows two hedge ratios

∂V

∂S

and

∂V

∂σ

. The term (p − δq) is

known as the risk-neutral drift rate.

2.2 Market price of volatility risk

Suppose the investor holds only the option V which is hedged only by the

underlying asset S in the following portfolio

= V − S

Over a short interval of time dt, the change in the value of this portfolio

can be written as

d =

∂V

∂t

+

1

2

σ

2

S

2

∂

2

V

∂S

2

+ ρσqS

∂

2

V

∂S∂σ

+

1

2

q

2

∂

2

V

∂σ

2

dt

+

∂V

∂S

−

dS +

∂V

∂σ

dσ

In the standard delta-hedging, the coefficient of dS is zero and we have

d− [(r + λ

V

)V − (r + λ

S

)S]dt =

¸

∂V

∂t

+

1

2

σ

2

S

2

∂

2

V

∂S

2

+ ρσqS

∂

2

V

∂S∂σ

+

1

2

q

2

∂

2

V

∂σ

2

+ (r + λ

S

)S

∂V

∂S

− (r + λ

V

)V

¸

dt +

∂V

∂σ

dσ

= q

∂V

∂σ

(δdt + dW

2

)

This results from equations (1) and (2). The term dW

2

represents a unit of

volatility risk. There are δ units of extra-return, given by dt for each unit

of volatility risk.

2.3 The market price of risk for traded assets

In the Black-Scholes analysis, the hedging portfolio is constructed using

the option and its underlying tradable asset. Consider the construction

of a portfolio as before using two options V and V

1

with different charac-

teristics, the initial portfolio value would be

= V −

1

V

1

Note that there are none of the underlying asset in this portfolio. Using

the same methodology as before gives the following equation

∂V

∂t

+

1

2

σ

2

S

2

∂

2

V

∂S

2

+ (µ − δ

S

σ)S

∂V

∂S

− (r + λ

V

)V = 0 (3)

The variable asset S is the value of a traded asset. Then V = S must be a

solution to this last equation. Substituting V = S in the last equation

gives

(µ − δ

S

σ)S − (r + λ

S

)S = 0

The market price of risk for a traded asset in the presence of information

costs

δ

S

=

µ − (r + λ

S

)

σ

Substituting δ

S

in (3) gives the following equation

∂V

∂t

+

1

2

σ

2

S

2

∂

2

V

∂S

2

+ (r + λ

S

)S

∂V

∂S

− (r + λ

V

)V = 0

This is the Black-Scholes equation in the presence of information costs.

3 Generalization of Certain Model

with Stochastic Volatility

and Information Costs

3.1 Generalization of the Hull and White 1987 model

We consider the following model

dB

t

= (r + λ

B

)B

t

dt

dS

t

= µ(S

t

, σ

t

, t)S

t

dt + σ

t

S

t

dW

1

t

dν

t

= γ (σ

t

, t)ν

t

dt + δ(σ

t

, t)ν

t

dW

2

t

(4)

where S

t

denotes the stock price at time t, ν

t

= σ

2

t

its instantaneous vari-

ance, and r the riskless interest rate, which is assumed to be constant. W

1

and W

2

are Brownian motions under P, they are independent. ν

t

has no

systematic risk. This yields a unique option price which can be computed

as the (conditional) expectation of the discounted terminal payoff under

a risk-neutral probability measure

˜

P. Put differently,

˜

P is obtained from P

by means of a Girsanov transformation such that

dB

t

= (r + λ

B

)B

t

dt

dS

t

= (r + λ

S

)S

t

dt + σ

t

S

t

d

˜

W

1

t

dν

t

= γ (σ

t

, t)ν

t

dt + δ(σ

t

, t)ν

t

d

˜

W

2

t

(5)

under

˜

P, where

˜

W

1

,

˜

W

2

are independent Brownian motions under

˜

P. The

risk-neutral dynamics of the bond and the underlying asset are used in

Bellalah (1999). The portfolio value would be

= V − S −

B

t

with

**units of the bond. When we apply the methodology of the previ-
**

ous section to this model, equation (2) gives:

∂V

∂t

+

1

2

ν

t

S

2

t

∂

2

V

∂S

2

t

+ ρσ

3

t

S

t

ξ

∂

2

V

∂S∂ν

t

+ ξ

2

ν

2

t

1

2

∂

2

V

∂ν

2

t

+ (r + λ

S

)S

t

∂V

∂S

t

+ (γ − δξ)ν

t

∂V

∂ν

t

+ (r + λ

B

)B

t

∂V

∂B

t

− (r + λ

V

)V = 0

(6)

with

˜

W

1

,

˜

W

2

independent Brownian motion under the probability

˜

P (ρ = 0)and λ

S

is the information cost of the security S

t

. The investor paid

^

Wilmott magazine 53

the shadow cost λ

S

if he does not know the asset. Also λ

B

is the informa-

tion cost of the bond B

t

and it is equal to zero if the asset is reskless. We

suppose that δ = 0, The option price is then given by

V(t, S

t

) =

˜

E

¸

B

t

B

T

(S

T

− K)

+

|F

t

¸

= e

−(r+λB )(T−t)

˜

E[(S

T

− K)

+

|F

t

] (7)

To obtain a more specific form for V, we use the additional assumption

contained in (5) and the independence of W

1

, W

2

that the instanta-

neous variance ν is not influenced by the stock price S. Setting

ν

t,T

=

1

T − t

T

t

ν

s

ds (8)

They show that the conditional distribution of

ST

St

under

˜

P, given ν

t,T

, is log-

normal with parameters (r + λ

B

)(T − t) and ν

T−t

. This allows to write V as

V(t, S

t

, σ

2

t

) =

∞

0

u

BS

(t, S

t

, ν

t,T

)dF(ν

t,T

|S

t

, σ

2

t

) (9)

where V

BS

denotes the usual Black-Scholes (1973) price corresponding to

the variance ν

t,T

and F is the conditional distribution under

˜

P of ν

t,T

given S

t

and σ

2

t

. This is equivalent to write :

V(t, S

t

, σ

2

t

) =

∞

0

V

BS

(t, S

t

, ν

t,T

)h(ν

t,T

|S

t

, σ

2

t

)dν

t,T (10)

with

V

BS

(ν) = S

t

N(d

1

) − Xe

−(r+λS )(T−t)

N(d

2

)

and d

1

=

log(S

t

/K) + (r + λ

S

+ ν/2)(T − t)

ν(T − t)

, d

2

= d

1

−

√

ν(T − t) .

When µ = 0 and as in H and White (1987) we have:

V(S, σ

2

t

) = V

BS

(ν) +

1

2

S

√

T − tN

(d

1

)(d

1

d

2

− 1)

4σ

3

×

¸

2σ

4

(e

k

− k − 1)

k

2

− σ

4

¸

+

1

6

S

√

T − tN

(d

1

)[(d

1

d

2

− 1)(d

1

d

2

− 3) − (d

2

1

+ d

2

2

)]

8σ

5

× σ

6

¸

e

3k

− (9 + 18k)e

k

+ (8 + 24k + 18k

2

+ 6k

3

)

3k

3

¸

+ . . . ,

(11)

avec k = ξ

2

(T − t), N

(x) =

1

√

2π

e

−

x

2

2 .

3.2 Generalization of Wiggins’s model

Under the assumption of the continuous trading, without frictions, in a

complete market, Wiggins (1987) use the following dynamics for the

asset and the volatility:

dS

t

= µ(S

t

, σ

t

, t)S

t

dt + σ

t

S

t

dW

St

dσ

t

= f (σ

t

)dt + θσ

t

dW

σt

(12)

with dW

St

, dW

σt

are processes of Wiener, the correlation coefficient

between stock returns and volatility movements is ρdt = dW

St

dW

σt

and

(dP/P)(dS

t

/S

t

) = 0. The instantaneous rate of return on the hedge port-

folio P is

dP/P = wdV/V + (1 − w)dS

t

/S

t

with w the fraction invested in the contingent claim V and (1 − w) the

fraction invested in the stock S. Equation (2) is equivalent in this case to:

∂V

∂t

+

1

2

σ

2

t

S

2

t

∂

2

V

∂S

2

t

+ ρσ

2

t

θS

t

∂

2

V

∂S

t

∂σ

t

+ θ

2

σ

2

t

1

2

∂

2

V

∂σ

2

t

+ (r + λ

S

)S

t

∂V

∂S

t

+ (f (σ

t

) − δθσ

t

)

∂V

∂σ

t

− (r + λ

V

)V = 0

(13)

As in Wiggins (1987), we can write the following equation

∂V

∂t

+

1

2

σ

2

t

S

2

t

∂

2

V

∂S

2

t

+ ρσ

2

t

θS

t

∂

2

V

∂S∂σ

t

+ θ

2

σ

2

t

1

2

∂

2

V

∂σ

2

t

+ (r + λ

S

)S

t

∂V

∂S

t

+

¸

f (σ

t

) − (µ − r − λ

S

)ρθ + (.)θ σ

t

(1 − ρ

2

)

¸

∂V

∂σ

t

− (r + λ

V

)V = 0

(14)

We conclude that the market price of risk affects the term given by

Wiggins (1987)(.) = (µ

P

− r − λ

P

)/σ

P

. This term is the expected excess

return per unit risk, or the market price of risk, for the hedge portfolio. It

represents the return-to-risk tradeoff required by investors for bearing

the volatility risk of the stock.

(.) =

δσ

t

− (µ − r − λ

S

)ρ

σ

t

(1 − ρ

2

)

(15)

The market price of risk depends on the information cost of the stock and

the stochastic volatility.

3.3 Generalization of Stein and Stein’s model

In this model, the stock price dynamics are given by the following

process:

dS

t

= µ(S

t

, σ

t

, t)S

t

dt + σ

t

S

t

dW

1

The volatility follows an Ornstein-Uhlenbeck process:

dσ

t

= (σ

t

− θ)dt + kdW

2

The Weiner processes dW

1

, dW

2

are uncorrelated. When equation (2) is

applied in this context, we have:

∂V

∂t

+

1

2

σ

2

t

S

2

t

∂

2

V

∂S

2

t

+ k

2

1

2

∂

2

V

∂σ

2

t

+ (r + λ

St

)S

t

∂V

∂S

t

+ [−(σ

t

− θ) − δk]

∂V

∂σ

t

− (r + λ

V

)V = 0

(16)

TECHNICAL ARTICLE 2

54 Wilmott magazine

When δ = 0 or to be a constant, equation (16) has a solution with the

same form as in Stein and Stein (1991). The solution depends on informa-

tion costs of V and the underlying asset S. The option price has the fol-

lowing form:

V = e

−(r+λV )

∞

St =K

[S

t

− K]H(S

t

, t | , r + λ

St

, k, θ)dS

t (17)

with H(S

t

, t) is the price distribution of the underlying asset at the time t

with a non-zero drift of S

t

.

3.4 Generalization of Heston’s model

The underlying asset and the volatility follow the diffusion process:

dS

t

= µ(S

t

, σ

t

, t)S

t

dt +

√

ν

t

S

t

dW

1

t

dν

t

= κ(θ − ν

t

)dt + σ

t

√

ν

t

dW

2

t

(18)

with ρ the correlation coefficient between dW

1

t

, dW

2

t

.

In this case, the value of any option V(S

t

, ν

t

, t) must satisfy the follow-

ing partial differential equation

∂V

∂t

+ ρσ

t

ν

t

S

t

∂

2

V

∂S

t

∂ν

t

+

1

2

ν

t

S

2

t

∂

2

V

∂S

2

t

+ σ

2

t

ν

t

1

2

∂

2

V

∂ν

2

t

+ (r + λ

S

)S

t

∂V

∂S

t

+ [κ(θ − ν

t

) − δσ

t

√

ν

t

]

∂V

∂ν

t

− (r + λ

V

)V = 0

(19)

Under the same assumption as in Heston (1993), it is possible to obtain

solution to equation (19). This solution depends on information costs λ

S

.

In fact, an European call with a strike price K and maturing at time T, sat-

isfies the equation (19) subject to the following boundary conditions

V(S, ν

t

, t) = Max(o, S − K)

V(0, ν

t

, t) = 0

∂V

∂S

t

(∞, ν

t

, t) = 1

(r + λ

S

)S

t

∂V

∂S

t

+ κ(θ)

∂V

∂ν

t

− (r + λ

V

)V +

∂V

∂t

= 0

V(S, ∞, t) = S

By analogy with the Black et Scholes (1973) formula, Heston (1993) gives a

solution of the form

V(S, ν

t

, t) = SP

1

− KP(t, T)P

2 (20)

with P(t, t + τ ) = e

−(r+λS )τ

the price at time t of a unit discount bond that

matures at time t + τ . The first term of the right side of the solution

V(S, ν

t

, t) is the present value of the underlying asset upon optimal exer-

cise. The second term is the present value of the strike-price. Both of

these terms must satisfy the equation. It is convenient to write them in

terms of the logarithm, (19)x = ln(S). By substitution of the solution in

equation (19), (20)P

1

and P

2

must satisfy the following equation:

∂P

j

∂t

+

1

2

ν

t

∂

2

V

∂x

+ ρσ

t

ν

t

S

t

∂

2

V

∂S

t

∂ν

t

+ σ

2

t

ν

t

1

2

∂

2

P

j

∂ν

2

t

+ (r + λ

S

+ u

j

ν

t

)

∂P

j

∂x

+ (a − b

j

√

ν

t

)

∂P

j

∂ν

t

= 0

(21)

for j = 1, 2 where u

1

= 1/2, u

2

= −1/2, a = κθ, b

1

= (κ − ρσ

t

)

√

ν

t

+ δσ

t

,

b

2

= κ

√

ν

t

+ δσ

t

Following the same resolution method in Heston (1993) for the equa-

tion (21), we obtain the solution of the characteristic function:

f

j

(x, ν

t

, t; φ) = exp[C(T − t; φ) + D(T − t; φ)ν

t

+ ixφ] (22)

when

C(τ ; φ) = i(r + λ

St

)φτ +

a

σ

2

t

¸

(b

j

− iρσ

t

φ + d)τ − 2 ln

¸

1 − ge

dτ

1 − g

¸¸

D(τ ; φ) =

b

j

− iρσ

t

φ + d

σ

2

t

¸

1 − e

dτ

1 − ge

dτ

¸

and g =

b

j

− iρσ

t

φ + d

b

j

− iρσ

t

φ − d

, d =

(iρσ

t

φ − b

j

)

2

− σ

2

t

(2iu

j

φ − φ

2

)

By inverting the characteristic functions f

j

, we obtain the desired

probabilities:

P

j

(x, ν

t

, t; ln K) =

1

2

+

1

π

∞

0

Re

¸

(e

−iφ ln K

)

∗

f

j

(x, ν

t

, T; φ)

iφ

¸

dφ (23)

with f

j

(x, ν

t

, T; φ) = e

iφx

.

3.5 Generalization of Johnson and Shanno’s model

We consider the following model:

dS

t

= µ

St

S

t

dt + σ

t

S

α

t

dW

1

t

dσ

t

= µ

σt

σ

t

dt + σ

t

σ

σt

dW

2

t

(24)

with dW

1

t

dW

2

t

= ρdt.

When equation (2) is applied to the model, we obtain:

∂V

∂t

+

1

2

σ

2

t

S

2α

t

∂

2

V

∂S

2

t

+ ρσ

3

t

S

α

t

σ

σt

∂

2

V

∂S∂σ

t

+ σ

2

t

σ

2

σt

1

2

∂

2

V

∂σ

2

t

+ (r + λ

S

)S

t

∂V

∂S

t

+ (µ

σt

− δσ

σt

)σ

t

∂V

∂σ

t

− (r + λ

V

)V = 0

(25)

Johnson and Shanno (1987), suppose that the risk premium of the volatil-

ity is zero. Consequently, we have:

δ =

µ

σt

σ

σt

^

Wilmott magazine 55

3.6 A general Markovian model with shadow costs of

incomplete information

In this subsection we present a general model in which we include infor-

mation cost and volatilities stochastic. We consider the following multi-

dimensional diffusion process:

dX

i

t

= a

i

(t, X

t

)dt +

n

¸

j=1

b

ij

(t, X

t

)W

j

t

(26)

for i = 1, . . . , m, where

a

i

, b

ij

: [0, T] × R

m+1

→R

are measurable functions. The process W = (W

1

, . . . , W

n

) is an n-dimen-

sional Brownian motion on a probability space (, F, Q ), and

F = (F

t

)

0≤t≤T

is the Q-augmentation of the filtration generated by W. We

assume that the coefficients a

i

, b

ij

satisfy appropriate growth and Lip-

schitz conditions so that the solution of (26) is a Markov process. We also

remark that under suitable continuity and nondegeneracy conditions on

the coefficients, F coincides with the natural filtration F

X

of X. This

model will be interpreted in the following way. The component X

0

describes the risk less asset; setting B := X

0

, we shall take b

0j

≡ 0 for

j = 1, . . . , n and a

0

(t, x) = (r(t, X

t

) + λ

B

)x

0

, so

dB

t

= (r(t, X

t

) + λ

B

)B

t

dt (27)

We assume that

T

0

|r(s, X

s

) + λ

X

|ds ≤ L < ∞ Q − a.s. (28)

for some L > 0 and X = (B, X

1

, . . . X

m

). According to the analysis of

Merton (1987),we assume that λ

X

= λ(s, X

s

) is measured in units of

expected return. We shall work with only one stock. The component X

1

describes its price process and is denoted by S. The other components of X

can then be used to model the additional structure of the market in

which S is embedded. In this general framework, an option or contingent

claim will be a random variable of the form g(X

T

). The classical example

is provided by European call option with strike price K which corre-

sponds to the claim (S

T

− K)

+

. Since the process X will usually contain

more components than just the bond B = X

0

and the stock price S = X

1

,

claim can depend on many things other than just the terminal stock

price S

T

. In fact, the only serious restriction is that the underling process

X (but not necessarily S) should be Markovian. This implies that (subject

to some integrability conditions) we can associate to any contingent

claim g(X

T

) an option pricing function

V : [0, T] × R

m+1

→R

defined by

V(t, x) = E

Q

¸

exp

−

T

t

r(s, X

s

) + λ

X

ds

g(X

t,x

T

)

(29)

where (X

t,x

s

)

t≤s≤T

denotes the solution of (26) starting from x at time t, i.e.,

with X

t,x

t

= x ∈ R

m+1

. To illustrate the previous analysis, we give the fol-

lowing example:

dB

t

= (r(t, X

t

) + λ

B

)B

t

dt

dS

t

= (r(t, X

t

) + λ

S

)S

t

dt + σ

t

S

t

dW

1

t

dσ

t

= −q(σ

t

− ς

t

)dt + pσ

t

dW

2

t

dς

t

=

1

α

(σ

t

− ς

t

)dt

(30)

with p > 0, q > 0, α > 0 and dW

1

t

, dW

2

t

are independent Brownian

motion under the probability Q . The processes of σ, ςare respectively the

instantaneous and weighted average volatility of the stock. The equation

for σ shows that the instantaneous volatility σ

t

is distributed by some

external noise (with an intensity p) and at the same time continuously

pulled back toward the average volatility ς

t

. The parameter q measures

the strength of this restoring force or speed of adjustment. The equation

(2) becomes in this case as follows

∂V

∂t

+

1

2

σ

t

S

2

t

∂

2

V

∂S

2

t

+ ρσ

2

pS

t

∂

2

V

∂S∂σ

t

+ p

2

σ

2

t

1

2

∂

2

V

∂σ

2

t

+ (r + λ

S

)S

t

∂V

∂S

t

+ (−q(σ

t

− ς

t

) − δpσ

t

)

∂V

∂σ

t

+ (r + λ

B

)B

t

∂V

∂B

t

+

1

α

(σ

t

− ς

t

)

∂V

∂ς

t

− (r + λ

V

)V = 0

(31)

Or

∂V

∂ς

t

=

∂V

∂σ

t

∂σ

t

∂ς

t

The equation (31) becomes

∂V

∂t

+

1

2

σ

t

S

2

t

∂

2

V

∂S

2

t

+ ρσ

2

pS

t

∂

2

V

∂S∂σ

t

+ p

2

σ

2

t

1

2

∂

2

V

∂σ

2

t

+ (r + λ

S

)S

t

∂V

∂S

t

+ (−q(σ

t

− ς

t

) − δpσ

t

+

1

α

(σ

t

− ς

t

)

∂σ

t

∂ς

t

)

∂V

∂σ

t

+ (r + λ

B

)B

t

∂V

∂B

t

− (r + λ

V

)V = 0

(32)

This equation can be written as follows:

∂V

∂t

+

1

2

σ

t

S

2

t

∂

2

V

∂S

2

t

+ ρσ

2

pS

t

∂

2

V

∂S∂σ

t

+ p

2

σ

2

t

1

2

∂

2

V

∂σ

2

t

+ (r + λ

S

)S

t

∂V

∂S

t

+

¸

1

α

∂σ

t

∂ς

t

− q

(σ

t

− ς

t

) − δpσ

t

¸

∂V

∂σ

t

+ (r + λ

B

)B

t

∂V

∂B

t

− (r + λ

V

)V = 0

(33)

with λ

B

, λ

S

and λ

V

indicate the information costs respectively for the

bond, the stock and the option.

TECHNICAL ARTICLE 2

56 Wilmott magazine

4 The Incomplete Market and the

Minimal Equivalent Martingale Measure

with Information Costs

We shall work with a model considerably more general than. It contains

one risk less asset B and m risky assets (26)S

i

, i = 1, . . . , m. The bond

price B and the stock prices S

i

are given by the stochastic differential

equation

dB

t

= (r + λ

B

)B

t

dt

dS

i

t

= µ

i

t

S

i

t

dt + S

i

t

n

¸

j=1

σ

i,j

t

dW

j

t

(34)

Here, W = (W

1

, . . . , W

n

)∗ is an n-dimensional Brownian motion on a

probability space (, F, P) and F = (F

t

)

0≤t≤T

denotes the P-augmentation

of the filtration generated by W. We take n ≥ m so that there are at least

as many sources of uncertainty as there are stocks available for trading.

All processes will be defined on [0, T], where the constant T > 0 denotes

the terminal time for our problem. We assume that the interest rate

r = (r

t

)

0≤t≤T

, the vector µ = (µ

t

)

0≤t≤T

= (µ

1

t

, . . . , µ

m

t

)

0≤t≤T

of stock appre-

ciation rates, the volatility matrix σ = (σ

t

)

0≤t≤T

= (σ

ij

t

)

0≤t≤T,i=1,... ,n,j=1,... ,m

and the vector (λ

X

)

0≤t≤T

= (λ

B

, λ

S

1 .., λ

S

m )

0≤t≤T

of the assets’information

costs , are progressively measurable with respect to F. The interest rate

r and λ

X

satisfies

T

0

|r

u

+ λ

X

|du ≤ L < ∞ P − a.s. for some L > 0

This implies that the bond price process B is bounded above and away

from 0, uniformly in t and ω. We also assume that the matrix σ

t

has full

rank m for every t so that the matrix (σ

t

σ

∗

t

)

−1

is well defined. This means

that the basic assets, namely the stock prices, have been chosen in such a

way that they are all nonredundant. Consider a “small investor”, i.e. , an

economic agent whose actions do not influence prices, who trades in the

stocks and the bond. His trading strategy can be described at any time t

by his total wealth V

t

and by the amounts π

i

t

invested in the ith stock for

i = 1, . . . , m. The amount invested in the bond is then given by

V

t

−

¸

m

i=1

π

i

t

. We shall call π = (π

t

)

0≤t≤T

= (π

1

t

, . . . , π

m

t

)

∗

0≤t≤T

a portfolio

process if π is progressively measurable with respect to F and satisfies

T

0

||σ

∗

u

π

u

||

2

du < ∞ P − a.s.

and

T

0

|π

∗

u

(µ

u

− r

u

1 − λ

S

|du < ∞ P − a.s

where 1 = (1, . . . , 1)

∗

∈ R

m

and λ

S

= (λ

S

1 , . . . , λ

S

m ) the vector of shadow

costs of the risky assets. The trading strategy is called self-financing if all

changes in the wealth process are entirely due to gains or losses from trad-

ing in the stocks and bond. For such a strategy, we denote two predictable

processes in R

m+1

the first is (η

t

)

0≤t≤T

the quantity of the riskless asset or

the bond and the second one is ξ

t

= (ξ

1

t

, . . . , ξ

m

t

)

0≤t≤T

the quantity of the

risky assets hold in such portfolio. the wealth process V must satisfy the

following equation:

dV

t

= η

t

dB

t

+

m

¸

i=1

ξ

i

t

dS

i

t

(35)

Substituting equation (34) in equation (35) and setting π

i

t

= ξ

i

t

S

i

t

we have:

dV

t

=

m

¸

i=1

π

i

t

µ

i

t

dt +

n

¸

j=1

σ

i,j

t

dW

j

t

+ (V

t

−

m

¸

i=1

π

i

t

)(r

t

+ λ

B

)dt

(36)

⇐⇒

dV

t

=

m

¸

i=1

π

i

t

µ

i

t

dt +

n

¸

j=1

σ

i,j

t

dW

j

t

+ V

t

(r

t

+ λ

V

) −

m

¸

i=1

π

i

t

(r

t

+ λ

i

)dt

With no arbitrage, we have:

dV

t

=

m

¸

i=1

π

i

t

(µ

i

t

− r

t

− λ

i

)dt +

m

¸

i=1

π

i

t

n

¸

j=1

σ

i,j

t

dW

j

t

+ V

t

(r

t

+ λ

V

)dt

With λ

V

is the vector of information costs (λ

B

, λ

1

, . . . , λ

m

)because the

value of the option is equal to the value of the portfolio in this strategy.

Equation (36) is equal to:

dV

t

=

π

∗

t

[µ

t

− r

t

1 − λ

S

] + (r

t

+ λ

V

)V

t

dt

+ π

∗

t

σ

t

dW

t

, 0 ≤ t ≤ T

(37)

The discounted wealth process V

= V/B is then given by

dV

t

= π

∗

t

[µ

t

− r

t

1 − λ

S

]dt + V

t

(λ

V

− λ

B

1) + π

∗

t

σ

t

dW

t (38)

⇐⇒

dV

t

= π

∗

t

[µ

t

− r

t

1 − λ

S

]dt + V

t

(λ

S

− λ

B

1) + π

∗

t

σ

t

dW

t (39)

with π

∗

t

= π

∗

t

/B

t

. Thus, any portfolio, process π uniquely determines a

wealth process V such that π and V together constitute a self-financing

strategy. We remark that the process give by the equation (37) depend

on the vector of the shadow costs λ

S

[because that the value of the

market price depend on λ

S

] and on λ

B

. If we interpret the process V as

the price of some assets, we can remind the definition of any “general”

asset:

Definition: A general asset is any asset whose value A is a semi martin-

gale with respect to P and F.

^

Wilmott magazine 57

Remark: For the following study, the general asset A has the following

form

dA

t

= υ

∗

t

dW

t

+ dF

t

0 ≤ t ≤ T (40)

with F un F-adapted process with paths of finite variation and the

process υ = (υ

1

, . . . , υ

n

)

∗

is progressively measurable with respect to F

and satisfies:

T

0

||υ

u

||

2

du < ∞ P − a.s.

We recall also the concept of an equivalent martingale for S:

Definition 1: A probability measure

˜

P on (, F) is called equivalent

martingale measure for S if

i)-

˜

P and P have the same null sets,

˜

P ≈ P. In particular, this implies

˜

P = P on F

0

.

ii)- The discounted price process S

**= S/B is a vector martingale under
**

˜

P.

Definition 2: An equivalent martingale measure

ˆ

P for S is called mini-

mal if any local P-martingale, orthogonal to S

i

, i = 1, . . . , m 0remains a

local martingale under

ˆ

P.

We begin by describing more precisely the equivalent martingale

measures for S. If

˜

P is any equivalent martingale measure for S and

˜

Z

t

= E

P

¸

d

˜

P

dP

F

t

¸

=

d

˜

P

dP

Ft

, 0 ≤ t ≤ T (4.5)

denotes a continuous version of the density process of

˜

P with respect to P,

then

˜

Z can be written as

˜

Z

t

= exp

−

t

0

˜ γ

∗

u

dW

u

−

1

2

t

0

|| ˜ γ

u

||

2

du

, 0 ≤ t ≤ T (41)

where ˜ γ = ( ˜ γ

1

, . . . , ˜ γ

n

)

∗

is adapted to F and satisfies the following

condition

T

0

|| ˜ γ

u

||

2

du < ∞ P − a.s. (42)

and

σ

t

˜ γ

t

= [µ

t

− r

t

1 − λ

S

], 0 ≤ t ≤ T (43)

if we suppose the existence and the uniqueness of the minimal equiva-

lent martingale measure

1

ˆ

P, we have

ˆ γ

t

= σ

∗

t

(σ

t

σ

∗

t

)

−1

[µ

t

− r

t

1 − λ

S

], 0 ≤ t ≤ T (44)

If ˜ γ ∈ L

2

a

[0, T] satisfies (44), then ˜ γ can be written as

2

˜ γ = ˆ γ + ϑ for some ϑ ∈ K(σ)

Indeed, decomposing ˜ γ as ˜ γ = ϑ + σ

∗

π with ϑ ∈ K(σ)yields by (43)

[µ

t

− (r

t

+ λ

B

)1] = σ ˜ γ = σσ

∗

π

Then the equation (43) becomes

˜ γ = ϑ + σ

∗

(σσ

∗

)

−1

[µ

t

− r

t

1 − λ

S

]

This allows us to prove the following result.

Lemma: Every compatible asset has a value process A of the form

dA

t

=

π

∗

t

[µ

t

− r

t

1 − λ

π

∗ ] + (r

t

+ λ

A

)A

t

dt

+ π

∗

t

σ

t

dW

t

+ ν

∗

t

dW

t

+ ν

∗

t

ϑ

t

dt

(45)

for some portfolio process π and some process ν, ϑ ∈ K(σ).

proof: Let A and his equivalent A

a continuous processes.

˜

P an

equivalent martingale measure for S such that A

is a local

˜

P-martingale.

Then A

˜

P is a local P-martingale and therefore continuous, since F is a

Brownian filtration. We denote by ˜ γ the process corresponding to

˜

P by

the relation (42). Under P, A

**= A/B has the form:
**

dA

t

=

1

B

t

dF

t

− A

t

[r

t

+ λ

B

]dt +

υ

∗

t

B

t

dW

t

where we have used equation. If we decompose (39)υ ∈ L

2

a

[0, T] as:

υ = ν + σ

∗

t

π with ν ∈ K(σ)

then applying Girsanov’s theorem to W shows that A

can be written

under

˜

P as

dA

t

=

1

B

t

dF

t

−

A

t

[r

t

+ λ

B

] +

υ

∗

t

B

t

˜ γ

t

dt +

υ

∗

t

B

t

d

˜

W

t

for some

˜

P-Brownian motion

˜

W

t

.

Since A

**is a continuous local
**

˜

P-martingale, by substitution of

˜ γ = ˆ γ + ϑ and using equation (44) in the previous equation we obtain:

dF

t

=

A

t

[r

t

+ λ

A

] + υ

∗

t

˜ γ

t

dt

=

A

t

[r

t

+ λ

At

] + (ν

∗

t

+ π

∗

t

σ

t

)( ˆ γ

t

+ ϑ

t

)

dt

=

A

t

[r

t

+ λ

At

] + ν

∗

t

ˆ γ

t

+ ν

∗

t

ϑ

t

+ π

∗

t

[µ

t

− r

t

1 − λ

π

∗ ]

dt

with F un F-adapted process with paths of finite variation.

This equation confirm that the process of a general asset defined in

equation is equivalent to the process defined in equation (45).

5 Conclusion

This paper developes a general context for the valuation of options with

stochastic volatility and information costs. The shadow costs are integrat-

ed in the investor’s portfolio wealth process in the same vein as in

Merton (1987), Bellalah and Jacquillat (1995) and Bellalah (1999). The

TECHNICAL ARTICLE 2

W

58 Wilmott magazine

TECHNICAL ARTICLE 2

information costs appear naturally in the derivation proposed in this

analysis. There is also another reformulation of the compatible asset’s

process, that gives more information for the drift term, which depends

on information costs. In the same way, several extensions of existing

models can be used for the development of the option valuation with sto-

chastic volatility and information costs.

1

see Norbert, Platen and Schweizer[1992] (page 162, 163)

2

see Norbert, Platen and Schweizer[1992] (page 165)

I Bellalah M : “Quatre Essais Sur Lévaluation des Options: Dividendes, Volatilités des

Taux d’intérˆ et et Information Incompléte .”, Doctorat de l’université de Paris-Dauphine,

(June 1990).

I Bellalah M, Jacquillat B : ”Option Valuation with Information Costs :Theory and Tests.”,

The Financial Review , Vol 30, N 3, (August 1995), 617–635.

I Bellalah M : ”The Valuation of Futures and Commodity Options with Information

Costs.“, Journal of Futures Markets (September 1999).

I Black F, Scholes M : ”The Pricing of Options and Corporate Liabilities.”, Journal of

Political Economy, 81 (May-June 1973), 637–659.

I Cox J C, Ross S A : ”A survey of Some New Results In financial Option Pricing Theory.”,

Journal of Finance, Vol 31, N 2 (May 1976), 383–402.

I Engle R : ”Autoregressive Conditional Hetero-elasticity with Estimates of the Variance

of U.K.Inflation .”, Econometrica 50, (1982), 987–1008.

I Heston S : ”A closed-Form Solution for Options with Stochastic Volatility with

Applications to Bond and Currency Options.”, The Review of Financial Studies, Vol 6

(1993), 327–343.

I Hofmann N, Platen E, Schweizer M : ”Option Pricing Under Incompleteness”,

Mathematical Finance, Vol 2, N 3, (July 1992), 153–187.

I Hull J, White A : ”The Pricing of Options on Assets with Stochastic Volatilities.”,

Journal of Finance,(June 1987), 281–320.

I Hull J, White A : ”An Analysis of The Bias in Option Pricing Caused by a Stochastic

Volatility.”, Advances in Futures and Options Research, Vol 3, (1988), 29–61.

I Johnson H, Shanno D : ”Option Pricing When The Variance is Changing.”, Journal of

Financial Quant. Anal, Vol 22, (1987), 143–151.

I Karatzas I : ”Optimization Problems In The Theory of Continuous Trading.”, SIAM.

Journal. Control Optima; Vol 27, (November 1989), 1221–1259.

I Merton R C : ”A Simple Model of Capital Market Equilibrium with Incomplete

Information.”, Journal of Finance, N 3 (July 1987), 483–509.

I Scott L : ”Option Pricing when The Variance Changes Randomly: Theory, Estimation

and an Application.”, Journal of Financial and Quantitative Analysis, Vol 22 (December

1987), 419–437.

I Stein E M, Stein J C : ”Stock Price Distributions with Stochastic Volatility: An Analytic

Approach.”, The Review of Financial Studies, Vol 4 (1991), 727–752.

I Wiggins J B :”Option Values Under Stochastic Volatility.” The Journal of Financial

Economics, Vol 19 (1987), 351–372.

FOOTNOTES & REFERENCES

Hence. Since both sides are functions of the independent variables S. σ and t gives the change in the value of this portfolio as: 1 1 ∂2V ∂2V ∂2V ∂V + σ 2 S2 2 + ρσ qS + q2 2 ∂t 2 ∂S ∂S∂σ 2 ∂σ − + 1 d = dt dt for a function δ(S. σ. the terms in dt must yield the deterministic return as in a Black-Scholes “hedge” portfolio. σ. the derivative asset price V(S. σ. a short position of units of the underlying asset and a short position of − 1 units of an other option with value V1 (S. All the sources of randomness in the portfolio value resulting from dS can be eliminated by setting the quantity before dS equal to zero. or ∂V − ∂σ 1 ∂V1 =0 ∂σ After eliminating the stochastic terms. Hence. σ and t. the volatility σ and time t. The functions p(S. a problem arises because this new source of randomness can not be easily hedged away. the instantaneous return on the portfolio must be the risk-free rate plus information costs on each asset in the portfolio as in Bellalah (1999). This equation can also be written as 1 1 ∂2V ∂2V ∂V ∂V ∂2V + σ 2 S2 2 + ρσ Sq + q2 2 + (r + λS )S ∂t 2 ∂S ∂S∂σ 2 ∂σ ∂S ∂V − (r + λV )V = 0 + (p − δq) ∂σ 1 1 ∂ 2 V1 ∂ 2 V1 ∂ 2 V1 ∂V1 + σ 2 S2 + q2 + ρσ qS ∂t 2 ∂S2 ∂S∂σ 2 ∂σ 2 ∂V ∂V1 ∂V1 ∂V − 1 dS + − 1 dσ ∂S ∂S ∂σ ∂σ (2) Wilmott magazine 51 ^ .TECHNICAL ARTICLE 2 Section examines some applications to several known models. Since the volatility is not a traded asset.1 The valuation model The pricing of derivative securities in the presence of a random volatility needs the use of two processes : one for the underlying asset and one for the volatility. ∂σ (1) Over a short interval of time dt . t) : =V− S− 1 V1 Since the two options differ by their strikes. t)dt + q(S. The pricing of options in this context needs the search for two hedging contracts. Consider the following dynamics for the underlying asset dS = µSdt + σ SdW 1 and also by setting the quantity before dσ equal to zero. applying Ito’s lemma for the functions S. The first is the underlying asset. t)dW 2 2 2 2 ∂V + 1 σ 2 S2 ∂ V + ρσ Sq ∂ V + 1 q2 ∂ V ∂t 2 ∂S∂σ 2 ∂σ 2 ∂S2 ∂V − (r + λV )V +(r + λS )S ∂S ∂V ∂σ 2 2 2 ∂V1 + 1 σ 2 S2 ∂ V1 + ρσ Sq ∂ V1 + 1 q2 ∂ V1 2 ∂t 2 ∂S ∂S∂σ 2 ∂σ 2 ∂V1 − (r + λV1 )V1 +(r + λS )S ∂S = ∂V1 ∂σ The two processes dW 1 dW 2 are Brownian-motions with a correlation coefficient ρ . this implies that both sides of the equation are independent of the contract type. t) and q(S. t) are specified in a way that fits the dynamics of the volatility over time. This gives d = 1 1 ∂2V ∂V ∂2V ∂2V dt + σ 2 S2 2 dt + ρσ Sq dt + q2 2 dt ∂t 2 ∂S ∂S∂σ 2 ∂σ 1 2 2 ∂ 2 V1 1 ∂ 2 V1 ∂ 2 V1 ∂V1 dt + σ S dt + q2 dt + ρσ Sq dt − 1 ∂t 2 ∂S2 ∂S∂σ 2 ∂σ 2 1 V1 ]dt. σ. σ. consider a portfolio comprising a long position in the option V . Following the same logic as in the original Black-Scholes model (1973). payoffs and maturities. The second can be an option that allows a hedge against volatility risk. t) can be expressed as a function of the dynamics of the underlying asset price S. = [(r + λV )V − (r + λS ) S − (r + λV1 ) and the following process for the volatility Isolating the terms in V and V1 gives dσ = p(S. σ. we have 1 1 ∂2V ∂2V ∂V ∂V ∂2V + σ 2 S2 2 + ρσ Sq + q2 2 + (r + λS )S − (r + λV )V ∂t 2 ∂S ∂S∂σ 2 ∂σ ∂S = −(p − δq) ∂V . The third section investigates a general process of a compatible asset with incompleteness in the market under information uncertainty and stochastic volatility. σ. or ∂V − ∂S − 1 ∂V1 =0 ∂S 2 Valuation of Options In the Presence of a Stochastic Volatility and Shadow Costs of Incomplete Information 2. t) referred to as the market price for risk or volatility risk.

t)νt dt + ˜ δ(σt . νt has no systematic risk. The risk-neutral dynamics of the bond and the underlying asset are used in Bellalah (1999). equation (2) gives: 1 ∂2V ∂V ∂2V 1 ∂2V + νt S2 2 + ρσt3 St ξ + ξ 2 νt2 t ∂t 2 ∂S∂νt 2 ∂νt2 ∂St (6) ∂V ∂V ∂V + (r + λS )St + (γ − δξ )νt + (r + λB )Bt − (r + λV )V = 0 ∂St ∂νt ∂Bt ˜ ˜ with W 1 . P is obtained from P by means of a Girsanov transformation such that dBt = (r + λB )Bt dt ˜ dSt = (r + λS )St dt + σt St dWt1 dνt = γ (σt . and r the riskless interest rate. When we apply the methodology of the previous section to this model. Substituting V = S in the last equation gives (µ − δS σ )S − (r + λS )S = 0 52 .1 Generalization of the Hull and White 1987 model dBt = (r + λB )Bt dt dSt = µ(St . νt = σt2 its instantaneous variance.2 Market price of volatility risk The market price of risk for a traded asset in the presence of information costs µ − (r + λS ) δS = σ Substituting δS in (3) gives the following equation 1 ∂V ∂V ∂2V + σ 2 S2 2 + (r + λS )S − (r + λV )V = 0 ∂t 2 ∂S ∂S Suppose the investor holds only the option V which is hedged only by the underlying asset S in the following portfolio =V− S This is the Black-Scholes equation in the presence of information costs. W 1 and W 2 are Brownian motions under P . Then V = S must be a solution to this last equation. The investor paid Wilmott magazine (3) The variable asset S is the value of a traded asset. ∂V ∂S and ∂V ∂σ . The term (p − δq) is 2. σt . Using the same methodology as before gives the following equation 1 ∂V ∂V ∂2V + σ 2 S2 2 + (µ − δS σ )S − (r + λV )V = 0 ∂t 2 ∂S ∂S with units of the bond. where St denotes the stock price at time t. t)νt dWt2 2. t)νt dt + δ(σt . The term dW 2 represents a unit of volatility risk. W 2 are independent Brownian motions under P . the hedging portfolio is constructed using the option and its underlying tradable asset. Over a short interval of time dt . given by dt for each unit of volatility risk. where W 1 . W 2 independent Brownian motion under the probability ˜ P (ρ = 0)and λS is the information cost of the security St . t)νt dW t2 In the standard delta-hedging. There are δ units of extra-return. the coefficient of dS is zero and we have d − [(r + λV )V − (r + λS ) S]dt = + 1 ∂2V ∂2V ∂V + σ 2 S2 2 + ρσ qS ∂t 2 ∂S ∂S∂σ 2 We consider the following model 1 2∂ V ∂V q + (r + λS )S 2 ∂σ 2 ∂S ∂V dσ − (r + λV )V dt + ∂σ ∂V (δdt + dW 2 ) =q ∂σ (4) This results from equations (1) and (2). Consider the construction of a portfolio as before using two options V and V1 with different characteristics. the initial portfolio value would be =V− 1 V1 (5) ˜ ˜ ˜ ˜ under P . the change in the value of this portfolio can be written as d 1 1 ∂2V ∂2V ∂2V ∂V + σ 2 S2 2 + ρσ qS + q2 2 = ∂t 2 ∂S ∂S∂σ 2 ∂σ ∂V ∂V + − dS + dσ ∂S ∂σ dt 3 Generalization of Certain Model with Stochastic Volatility and Information Costs 3. t)St dt + σt St dW t1 dνt = γ (σt .This equation shows two hedge ratios known as the risk-neutral drift rate. they are independent.3 The market price of risk for traded assets In the Black-Scholes analysis. Put differently. This yields a unique option price which can be computed as the (conditional) expectation of the discounted terminal payoff under ˜ ˜ a risk-neutral probability measure P . which is assumed to be constant. The portfolio value would be =V− S− Bt Note that there are none of the underlying asset in this portfolio.

σt2 )dν t. d2 = d1 − ν(T − t) . dW 2 are uncorrelated.TECHNICAL ARTICLE 2 the shadow cost λS if he does not know the asset. σt . t)St dt + σt St dW St dσt = f (σt )dt + θ σt dW σt Wilmott magazine The Weiner processes dW 1 .) = (µP − r − λP )/σ P . When equation (2) is applied in this context.T )dF(ν t.T = 1 T−t T with w the fraction invested in the contingent claim V and (1 − w) the fraction invested in the stock S. St .T given St and σt2 . Equation (2) is equivalent in this case to: 1 ∂V ∂2V ∂2V 1 ∂2V + σt2 S2 2 + ρσt2 θ St + θ 2 σt2 t ∂t 2 ∂St ∂σt 2 ∂σt2 ∂St + (r + λS )St ∂V ∂V + (f (σt ) − δθ σt ) − (r + λV )V = 0 ∂St ∂σt νs ds t (8) (13) ˜ They show that the conditional distribution of STt under P. This allows to write V as V(t.T and F is the conditional distribution under ˜ P of ν t. σt . 3.)θ σt (1 − ρ 2 ) − (r + λV )V = 0 ∂σt where VBS denotes the usual Black-Scholes (1973) price corresponding to the variance ν t.2 Generalization of Wiggins’s model Under the assumption of the continuous trading. This is equivalent to write : V(t. dW σt are processes of Wiener. we can write the following equation (9) 1 ∂V ∂2V ∂2V 1 ∂2V ∂V + σt2 S2 2 + ρσt2 θ St + θ 2 σt2 + (r + λS )St t 2 ∂t 2 ∂S∂σt 2 ∂σt ∂St ∂St (14) ∂V + f (σt ) − (µ − r − λS )ρθ + (. ν(T − t) When µ = 0 and as in H and White (1987) we have: We conclude that the market price of risk affects the term given by Wiggins (1987) (. The instantaneous rate of return on the hedge portfolio P is dP/P = wdV /V + (1 − w)dSt /St (7) To obtain a more specific form for V . × σ6 3k3 N (x) = x √1 e− 2 2π 2 The market price of risk depends on the information cost of the stock and the stochastic volatility. or the market price of risk. We suppose that δ = 0. we have: 1 ∂V ∂2V 1 ∂2V + σt2 S2 2 + k2 t ∂t 2 2 ∂σt2 ∂St ∂V ∂V + (r + λSt )St + [− (σt − θ ) − δk] − (r + λV )V = 0 ∂St ∂σt (16) (12) 53 ^ .T |St ... . t)St dt + σt St dW 1 avec k = ξ 2 (T − t). the correlation coefficient between stock returns and volatility movements is ρdt = dW St dW σt and (dP/P)(dSt /St ) = 0 .3 Generalization of Stein and Stein’s model In this model.) = δσt − (µ − r − λS )ρ σt (1 − ρ 2 ) and d1 = (15) V(S. It represents the return-to-risk tradeoff required by investors for bearing the volatility risk of the stock. (. we use the additional assumption contained in (5) and the independence of W 1 .. in a complete market. St . St ) = E BT with dW St . ν t. σt2 ) = 0 ∞ VBS (t. St . ν t. σt2 ) = 0 ∞ uBS (t. Also λB is the information cost of the bond Bt and it is equal to zero if the asset is reskless.T |St . the stock price dynamics are given by the following process: dSt = µ(St . W 2 that the instantaneous variance ν is not influenced by the stock price S.T )h(ν t. is logS normal with parameters (r + λB )(T − t) and ν T−t . The volatility follows an Ornstein-Uhlenbeck process: dσt = (σt − θ )dt + kdW 2 3. without frictions. σt2 ) As in Wiggins (1987). for the hedge portfolio. This term is the expected excess return per unit risk. σt2 ) = VBS (ν) + √ 1 S T − tN (d1 )(d1 d2 − 1) 2 4σ 3 4 k 2σ (e − k − 1) × − σ4 k2 (11) √ 1 S T − tN (d1 )[(d1 d2 − 1)(d1 d2 − 3) − (d2 + d2 )] 1 2 + 6 8σ 5 3k k e − (9 + 18k)e + (8 + 24k + 18k2 + 6k3 ) + . Setting ν t. Wiggins (1987) use the following dynamics for the asset and the volatility: dSt = µ(St . The option price is then given by ˜ ˜ Bt (ST − K)+ |Ft = e−(r+λB )(T−t) E[(ST − K)+ |Ft ] V(t.T . St . given ν t.T (10) with VBS (ν) = St N(d1 ) − Xe−(r+λS )(T−t) N(d2 ) √ log(St /K) + (r + λS + ν/2)(T − t) .

t) must satisfy the following partial differential equation 1 ∂V ∂2V ∂2V 1 ∂2V + ρσt νt St + νt S2 2 + σt2 νt t ∂t ∂St ∂νt 2 2 ∂νt2 ∂St ∂V √ ∂V + (r + λS )St + [κ(θ − νt ) − δσt νt ] − (r + λV )V = 0 ∂St ∂νt C(τ . 3. it is possible to obtain solution to equation (19). νt . It is convenient to write them in terms of the logarithm. t + τ ) = e−(r+λS )τ the price at time t of a unit discount bond that matures at time t + τ . This solution depends on information costs λS . t) = Max(o. 2 where u1 = 1/2. νt . (20)P1 and P2 must satisfy the following equation: ∂Pj 1 ∂2V ∂2V 1 ∂ 2 Pj + νt + ρσt νt St + σt2 νt ∂t 2 ∂x ∂St ∂νt 2 ∂νt2 ∂Pj √ ∂Pj + (a − bj νt ) + (r + λS + uj νt ) =0 ∂x ∂νt (21) [St − K]H(St . In fact. d = (iρσt φ − bj )2 − σt2 (2iuj φ − φ 2 ) bj − iρσt φ − d By inverting the characteristic functions fj .When δ = 0 or to be a constant. φ) + D(T − t. By substitution of the solution in Johnson and Shanno (1987). T)P2 (20) (25) with P(t. we obtain the solution of the characteristic function: fj (x. an European call with a strike price K and maturing at time T. T. dW t2 . k. the value of any option V(St . t. S − K) V(0. σt . φ)νt + ixφ] The underlying asset and the volatility follow the diffusion process: when (22) (18) with ρ the correlation coefficient between dW t1 . νt . The first term of the right side of the solution V(S. νt . φ) dφ iφ (23) with fj (x. νt . √ b2 = κ νt + δσt Following the same resolution method in Heston (1993) for the equation (21). r + λSt . νt . The solution depends on information costs of V and the underlying asset S. φ) = eiφx . we have: δ= µσ t σσ t Wilmott magazine 54 . b1 = (κ − ρσt ) νt + δσt . suppose that the risk premium of the volatility is zero. satisfies the equation (19) subject to the following boundary conditions V(S. ln K) = 1 1 + 2 π ∞ Re 0 (e−iφ ln K )∗ fj (x. t) = S Pj (x. The option price has the following form: V = e−(r+λV ) ∞ S t =K equation (19). t)St dt + νt St dW t1 √ dνt = κ(θ − νt )dt + σt νt dW t2 √ for j = 1. Both of these terms must satisfy the equation. t) = 0 ∂V (∞. (19)x = ln(S) . νt . Heston (1993) gives a solution of the form V(S. φ) = exp[C(T − t. θ )dSt (17) with H(St . Consequently.5 Generalization of Johnson and Shanno’s model We consider the following model: dSt = µSt St dt + σt Sα dW t1 t dσt = µσt σt dt + σt σσt dW t2 (24) with dW t1 dW t2 = ρdt . t) is the present value of the underlying asset upon optimal exercise. T. t. νt . 3. When equation (2) is applied to the model. νt . t) = 1 ∂St ∂V ∂V ∂V (r + λS )St =0 + κ(θ ) − (r + λV )V + ∂St ∂νt ∂t V(S. ∞. t) is the price distribution of the underlying asset at the time t with a non-zero drift of St . The second term is the present value of the strike-price. t) = SP1 − KP(t. φ) = a 1 − gedτ (bj − iρσt φ + d)τ − 2 ln 1−g σt2 bj − iρσt φ + d 1 − edτ 1 − gedτ σt2 (19) bj − iρσt φ + d . we obtain the desired probabilities: and g = Under the same assumption as in Heston (1993). equation (16) has a solution with the same form as in Stein and Stein (1991). In this case. νt . t | . we obtain: 2 1 ∂V ∂2V ∂2V 2 1 ∂ V + σt2 S2α 2 + ρσt3 Sα σσt + σt2 σσt t t ∂t 2 ∂S∂σt 2 ∂σt2 ∂St ∂V ∂V + (r + λS )St + (µσt − δσσt )σt − (r + λV )V = 0 ∂St ∂σt By analogy with the Black et Scholes (1973) formula. φ) = i(r + λSt )φτ + D(τ . a = κθ .4 Generalization of Heston’s model √ dSt = µ(St . u2 = −1/2 .

The other components of X can then be used to model the additional structure of the market in which S is embedded. m . The processes of σ. λS and λV indicate the information costs respectively for the bond. . This implies that (subject to some integrability conditions) we can associate to any contingent claim g(XT ) an option pricing function V : [0. q > 0. the stock and the option. X1 . bij : [0. . claim can depend on many things other than just the terminal stock price ST . According to the analysis of Merton (1987). Xs ) + λX ds g(XT ) (29) with λB . an option or contingent claim will be a random variable of the form g(XT ). α > 0 and dW t1 . In fact. The equation (2) becomes in this case as follows 1 ∂V ∂2V ∂2V 1 ∂2V ∂V + σt S2 2 + ρσ 2 pSt + p2 σt2 + (r + λS )St t ∂t 2 ∂S∂σt 2 ∂σt2 ∂St ∂St ∂V 1 ∂V ∂V + (−q(σt − ςt ) − δpσt ) + (r + λB )Bt + (σt − ςt ) ∂σt ∂Bt α ∂ςt − (r + λV )V = 0 (27) (31) We assume that T |r(s. The classical example is provided by European call option with strike price K which corresponds to the claim (ST − K)+ . and F = (Ft )0≤t≤T is the Q-augmentation of the filtration generated by W. . T] × R m+1 → R The equation (31) becomes 1 ∂V ∂2V ∂2V 1 ∂2V ∂V + σt S2 2 + ρσ 2 pSt + p2 σt2 + (r + λS )St t ∂t 2 ∂S∂σt 2 ∂σt2 ∂St ∂St 1 ∂σt ∂V ∂V + (−q(σt − ςt ) − δpσt + (σt − ςt ) ) + (r + λB )Bt α ∂ςt ∂σt ∂Bt − (r + λV )V = 0 (32) This equation can be written as follows: 1 ∂V ∂2V ∂2V 1 ∂2V ∂V + σt S2 2 + ρσ 2 pSt + p2 σt2 + (r + λS )St t ∂t 2 ∂S∂σt 2 ∂σt2 ∂St ∂St ∂V ∂V 1 ∂σt + − q (σt − ςt ) − δpσt + (r + λB )Bt α ∂ςt ∂σt ∂Bt − (r + λV )V = 0 (33) defined by T V(t.e. The component X1 describes its price process and is denoted by S. . . Wilmott magazine 55 ^ . W n ) is an n-dimensional Brownian motion on a probability space ( . . The equation for σ shows that the instantaneous volatility σt is distributed by some external noise (with an intensity p) and at the same time continuously pulled back toward the average volatility ςt . We also remark that under suitable continuity and nondegeneracy conditions on the coefficients.x where (Xs )t≤s≤T denotes the solution of (26) starting from x at time t. Xt )dt + i j=1 b ij j (t.x with Xt = x ∈ R m+1 . . . Since the process X will usually contain more components than just the bond B = X0 and the stock price S = X1 . F . Xs ) is measured in units of expected return. T] × R m+1 → R are measurable functions. Xt ) + λB )Bt dt with p > 0. . x) = EQ exp − t t. t. n and a0 (t. (28) Or ∂V ∂σt ∂V = ∂ςt ∂σt ∂ςt for some L > 0 and X = (B. i.we assume that λX = λ(s. We consider the following multidimensional diffusion process: n t. Xm ) . Xs ) + λX |ds ≤ L < ∞ 0 Q − a. The process W = (W 1 . Xt ) + λS )St dt + σt St dW t1 dσt = −q(σt − ςt )dt + pσt dW t2 1 dςt = (σt − ςt )dt α dXti = a (t. To illustrate the previous analysis. . We assume that the coefficients ai .. where ai . we give the following example: dBt = (r(t. Q ) . . we shall take b0j ≡ 0 for j = 1.6 A general Markovian model with shadow costs of incomplete information In this subsection we present a general model in which we include information cost and volatilities stochastic. F coincides with the natural filtration F X of X. . dW t2 are independent Brownian motion under the probability Q . This model will be interpreted in the following way. Xt )Wt (26) (30) for i = 1. .TECHNICAL ARTICLE 2 3. In this general framework. The parameter q measures the strength of this restoring force or speed of adjustment. We shall work with only one stock. Xt ) + λB )x0 . the only serious restriction is that the underling process X (but not necessarily S) should be Markovian. x) = (r(t. Xt ) + λB )Bt dt dSt = (r(t.x r(s. . so dBt = (r(t. The component X0 describes the risk less asset. ς are respectively the instantaneous and weighted average volatility of the stock. . bij satisfy appropriate growth and Lipschitz conditions so that the solution of (26) is a Markov process.s. setting B := X0 .

. . Thus. . We remark that the process give by the equation (37) depend on the vector of the shadow costs λS [because that the value of the market price depend on λS ] and on λB . . we have: m m n dVt = i=1 πti (µit − rt − λi )dt + i=1 πti j=1 σt dW t + Vt (rt + λV )dt i.e.s. If we interpret the process V as the price of some assets. . .. .i=1. namely the stock prices. .. 0≤t≤T (37) The discounted wealth process V = V/B is then given by dVt = πt ∗ [µt − rt 1 − λS ]dt + Vt (λV − λB 1) + πt ∗ σt dW t ⇐⇒ dVt = πt ∗ [µt − rt 1 − λS ]dt + Vt (λS − λB 1) + πt ∗ σt dW t (38) (39) P − a. This means that the basic assets. It contains one risk less asset B and m risky assets (26)Si . an economic agent whose actions do not influence prices. .j j + Vt (rt + λV ) − i=1 πti (rt + λi )dt With no arbitrage. λm ) because the value of the option is equal to the value of the portfolio in this strategy. πtm )∗ i=1 0≤t≤T a portfolio process if π is progressively measurable with respect to F and satisfies T 0 ∗ ||σu πu ||2 du < ∞ With λV is the vector of information costs (λB . the volatility matrix σ = (σt )0≤t≤T = (σt )0≤t≤T.. . P) and F = (Ft )0≤t≤T denotes the P-augmentation of the filtration generated by W. We take n ≥ m so that there are at least as many sources of uncertainty as there are stocks available for trading. Consider a “small investor”. for some L > 0 This implies that the bond price process B is bounded above and away from 0. His trading strategy can be described at any time t by his total wealth Vt and by the amounts πti invested in the ith stock for i = 1. T]. .. λS1 . . The bond price B and the stock prices Si are given by the stochastic differential equation dBt = (r + λB )Bt dt n processes in R m+1 the first is (ηt )0≤t≤T the quantity of the riskless asset or the bond and the second one is ξt = (ξt1 . . are progressively measurable with respect to F . Wilmott magazine 56 . any portfolio. have been chosen in such a way that they are all nonredundant. . we can remind the definition of any “general” asset: Definition: A general asset is any asset whose value A is a semi martingale with respect to P and F . 1)∗ ∈ R m and λS = (λS1 . . . W n )∗ is an n-dimensional Brownian motion on a probability space ( . The amount invested in the bond is then given by Vt − m πti . . we denote two predictable with πt ∗ = πt∗ /Bt ..n. λSm )0≤t≤T of the assets’information costs . the vector µ = (µt )0≤t≤T = (µ1 . the wealth process V must satisfy the following equation: m dVt = ηt dBt + i=1 ξti dSit (35) Substituting equation (34) in equation (35) and setting πti = ξti Sit we have: m n dVt = σt dW t i. . . .. F . The trading strategy is called self-financing if all changes in the wealth process are entirely due to gains or losses from trading in the stocks and bond. . .4 The Incomplete Market and the Minimal Equivalent Martingale Measure with Information Costs We shall work with a model considerably more general than.j=1. . ξtm )0≤t≤T the quantity of the risky assets hold in such portfolio. . . m .. . m . .. i. The interest rate r and λX satisfies T dVt = i=1 πti µit dt + j=1 m σt dW t i.s where 1 = (1. who trades in the stocks and the bond. . uniformly in t and ω .. µm )0≤t≤T of stock appret t ij ciation rates.m and the vector (λX )0≤t≤T = (λB . .. . . i = 1. Equation (36) is equal to: dVt = πt∗ [µt − rt 1 − λS ] + (rt + λV )Vt dt + πt∗ σt dW t . . where the constant T > 0 denotes the terminal time for our problem.j j |ru + λX |du ≤ L < ∞ 0 P − a. For such a strategy. . We also assume that the matrix σt has full rank m for every t so that the matrix (σt σt∗ )−1 is well defined.j j dSit = µit Sit dt + Sit j=1 (34) (36) + (Vt − i=1 m πti )(rt ⇐⇒ n Here. λSm ) the vector of shadow costs of the risky assets. process π uniquely determines a wealth process V such that π and V together constitute a self-financing strategy. All processes will be defined on [0. . λ1 . We shall call π = (πt )0≤t≤T = (πt1 .j j πti µit dt + i=1 m j=1 σt dW t + λB )dt i. . and T 0 ∗ |πu (µu − ru 1 − λS |du < ∞ P − a. . W = (W 1 . . We assume that the interest rate r = (rt )0≤t≤T .s.

The discounted price process S = S/B is a vector martingale under P . F ) is called equivalent martingale measure for S if ˜ ˜ i). The ˜ ˜ Indeed. T] satisfies (44). by substitution of γ = γ + ϑ and using equation (44) in the previous equation we obtain: ˜ ˆ dFt = At [rt + λA ] + υt∗ γt dt ˜ = At [rt + λAt ] + (νt∗ + πt∗ σt )(γt + ϑt ) dt ˆ ||γu ||2 du < ∞ ˜ 0 P − a. then γ can be written as2 a γ =γ +ϑ ˜ ˆ for some ϑ ∈ K(σ ) 5 Conclusion This paper developes a general context for the valuation of options with stochastic volatility and information costs. . ϑ ∈ K(σ ). . ˜ proof: Let A and his equivalent A a continuous processes. Bellalah and Jacquillat (1995) and Bellalah (1999). ˜ can be written as then Z ˜ Zt = exp − 0 t then applying Girsanov’s theorem to W shows that A can be written ˜ under P as dAt = υ∗ ˜ 1 υ∗ dFt − At [rt + λB ] + t γt dt + t dWt ˜ Bt Bt Bt γu∗ dW u − ˜ 1 2 t ||γu ||2 du . . A = A/B has the form: dAt = 1 υ∗ dFt − At [rt + λB ]dt + t dW t Bt Bt where we have used equation. since F is a ˜ ˜ Brownian filtration. We denote by γ the process corresponding to P by the relation (42). γ n )∗ is adapted to F and satisfies the following condition T ˜ ˜ for some P -Brownian motion Wt . . ˜ Since A is a continuous local P -martingale. υ n )∗ is progressively measurable with respect to F and satisfies: T This allows us to prove the following result. . . ˆ Definition 2: An equivalent martingale measure P for S is called minii mal if any local P -martingale. Ft for some portfolio process π and some process ν. decomposing γ as γ = ϑ + σ ∗ π with ϑ ∈ K(σ )yields by (43) [µt − (rt + λB )1] = σ γ = σ σ ∗ π ˜ Wilmott magazine 57 ^ .5) ˜ denotes a continuous version of the density process of P with respect to P . .TECHNICAL ARTICLE 2 Remark: For the following study. ˜ Then A P is a local P-martingale and therefore continuous. this implies ˜ P = P on F0 . (42) and σt γt = [µt − rt 1 − λS ]. If we decompose (39)υ ∈ L2 [0.P and P have the same null sets. we have γt = σt∗ (σt σt∗ )−1 [µt − rt 1 − λS ]. . If P is any equivalent martingale measure for S and ˜ Zt = EP ˜ ˜ dP dP Ft = dP dP .s. ˜ 0 0≤t≤T (41) ˜ ˜ ˜ where γ = (γ 1 . This equation confirm that the process of a general asset defined in equation is equivalent to the process defined in equation (45). P ≈ P . ˜ ii). The shadow costs are integrated in the investor’s portfolio wealth process in the same vein as in Merton (1987). P an ˜ equivalent martingale measure for S such that A is a local P -martingale.s. . We begin by describing more precisely the equivalent martingale ˜ measures for S. (45) We recall also the concept of an equivalent martingale for S: ˜ Definition 1: A probability measure P on ( . ˆ 0≤t≤T (44) with F un F -adapted process with paths of finite variation. Under P. ˜ 0≤t≤T (43) ˆ = At [rt + λAt ] + νt∗ γt + νt∗ ϑt + πt∗ [µt − rt 1 − λπ ∗ ] dt if we suppose the existence and the uniqueness of the minimal equivaˆ lent martingale measure1 P . . i = 1. the general asset A has the following form dAt = υt∗ dW t + dFt 0≤t≤T Then the equation (43) becomes γ = ϑ + σ ∗ (σ σ ∗ )−1 [µt − rt 1 − λS ] ˜ (40) with F un F -adapted process with paths of finite variation and the process υ = (υ 1 . . Lemma: Every compatible asset has a value process A of the form dAt = πt∗ [µt − rt 1 − λπ ∗ ] + (rt + λA )At dt + πt∗ σt dW t + νt∗ dW t + νt∗ ϑt dt ||υu ||2 du < ∞ 0 P − a. In particular. T] as: a υ = ν + σt∗ π with ν ∈ K(σ ) 0≤t≤T (4. orthogonal to S . m 0remains a ˆ local martingale under P . ˜ ˜ If γ ∈ L2 [0. .

Advances in Futures and Options Research. (1988). I Scott L : ”Option Pricing when The Variance Changes Randomly: Theory. Schweizer M : ”Option Pricing Under Incompleteness”. Journal of Financial and Quantitative Analysis. Doctorat de l’université de Paris-Dauphine. 637–659. FOOTNOTES & REFERENCES see Norbert. (1982).”. Vol 19 (1987). I Hull J. Vol 27. I Johnson H. Vol 31. Econometrica 50. (August 1995). There is also another reformulation of the compatible asset’s process. 2 1 W 58 Wilmott magazine . The Financial Review . The Review of Financial Studies. SIAM.”. Vol 4 (1991). 163) see Norbert. N 2 (May 1976).”. I Bellalah M : ”The Valuation of Futures and Commodity Options with Information Costs. Journal of Finance.K. I Cox J C. Control Optima. 383–402. that gives more information for the drift term. Volatilités des ˆ Taux d’intéret et Information Incompléte .”. In the same way. I Bellalah M. Ross S A : ”A survey of Some New Results In financial Option Pricing Theory. I Hofmann N.”. Mathematical Finance. The Review of Financial Studies.”. Vol 3. I Merton R C : ”A Simple Model of Capital Market Equilibrium with Incomplete Information.” The Journal of Financial Economics. Vol 30. Journal. 327–343. Journal of Finance. Estimation and an Application. I Stein E M.“. Shanno D : ”Option Pricing When The Variance is Changing.”. which depends on information costs. (June 1990). 351–372. 1221–1259. Platen E. Journal of Financial Quant. (July 1992). I Black F. Scholes M : ”The Pricing of Options and Corporate Liabilities. Stein J C : ”Stock Price Distributions with Stochastic Volatility: An Analytic Approach. several extensions of existing models can be used for the development of the option valuation with stochastic volatility and information costs. Journal of Futures Markets (September 1999). Jacquillat B : ”Option Valuation with Information Costs :Theory and Tests. Vol 6 (1993). 143–151. Journal of Finance. N 3. (November 1989). I Wiggins J B :”Option Values Under Stochastic Volatility. Platen and Schweizer[1992] (page 162. (1987). 29–61. Vol 2.”. 281–320.”. Platen and Schweizer[1992] (page 165) I Bellalah M : “Quatre Essais Sur Lévaluation des Options: Dividendes. White A : ”An Analysis of The Bias in Option Pricing Caused by a Stochastic Volatility. Vol 22. 727–752. 153–187.(June 1987).Inflation . N 3 (July 1987). Anal. I Karatzas I : ”Optimization Problems In The Theory of Continuous Trading. I Hull J. I Heston S : ”A closed-Form Solution for Options with Stochastic Volatility with Applications to Bond and Currency Options. White A : ”The Pricing of Options on Assets with Stochastic Volatilities. 617–635. 81 (May-June 1973). N 3.”.TECHNICAL ARTICLE 2 information costs appear naturally in the derivation proposed in this analysis.”. 483–509. I Engle R : ”Autoregressive Conditional Hetero-elasticity with Estimates of the Variance of U.”.”. Vol 22 (December 1987). 419–437. Journal of Political Economy. 987–1008.

- Read and print without ads
- Download to keep your version
- Edit, email or read offline

Are you sure?

This action might not be possible to undo. Are you sure you want to continue?

CANCEL

OK

You've been reading!

NO, THANKS

OK

scribd