Journal of

Risk and Financial

Management

Article
Utility Indifference Option Pricing Model with a Non-Constant
Risk-Aversion under Transaction Costs and Its
Numerical Approximation
Pedro Pólvora and Daniel Ševčovič *

Department of Applied Mathematics and Statistics, Comenius University, 842 48 Bratislava, Slovakia;
[email protected]
* Correspondence: [email protected]

Abstract: Our goal is to analyze the system of Hamilton-Jacobi-Bellman equations arising in

derivative securities pricing models. The European style of an option price is constructed as a
difference of the certainty equivalents to the value functions solving the system of HJB equations. We
introduce the transformation method for solving the penalized nonlinear partial differential equation.
The transformed equation involves possibly non-constant the risk aversion function containing the
negative ratio between the second and first derivatives of the utility function. Using comparison
principles we derive useful bounds on the option price. We also propose a finite difference numerical
discretization scheme with some computational examples.





 Keywords: option pricing; utility indifference pricing; transaction costs; Hamilton-Jacobi-Bellman
equation; penalty methods; finite difference approximation
Citation: Pólvora, Pedro, and
Daniel Ševčovič. 2021. Utility
Indifference Option Pricing Model
with a Non-Constant Risk-Aversion
under Transaction Costs and Its
1. Introduction
Numerical Approximation. Journal of In the last century the world witnessed a tremendous change and evolution in almost
Risk and Financial Management 14: 399. every industry and the financial one is no exception. One of the aspects that evolved greatly
https://doi.org/10.3390/ in finance was the financial derivatives, which saw their usage grow exponentially.
jrfm14090399 A financial derivative is a contract between two parties where they agree to future
financial exchanges and whose value depends on one more underlying assets. There are
Academic Editor: George Halkos multiple types of these contracts and they are used extensively in many industries, both
for hedging and speculation. Depending on the type of derivative and on the position
Received: 1 July 2021
(buyer vs. seller) they can be used to either limit or increase the financial exposure to
Accepted: 23 August 2021
a particular financial asset. Examples of uses of financial derivatives include: financial
Published: 25 August 2021
institutions transforming a pool of equally risky mortgages into multiple contracts with
different specific risk profiles, international enterprises reducing their foreign exchange
Publisher’s Note: MDPI stays neutral
risk, investors increasing their exposure to the increase in price of a stock by buying
with regard to jurisdictional claims in
financial derivatives.
published maps and institutional affil-
Financial options, a particular type of financial derivatives, are contracts where the
iations.
buyer has the option but not the obligation to transact an asset at predefined conditions such
as price or time. A key aspect of financial options is that their value, or price, is dependent
on the underlying assets and finding it is fundamental for trading and managing the option
and requires some type of mathematical modeling. Since the future payoff of the option is
Copyright: © 2021 by the authors.
uncertain at the time of the trade it is required to price it with probabilistic and statistical
Licensee MDPI, Basel, Switzerland.
considerations, and different approaches have been developed, examples include: Binomial
This article is an open access article
Trees, Monte Carlo simulations, Black-Scholes/PDEs. This was a key piece in the birth of
distributed under the terms and
what is now called financial mathematics.
conditions of the Creative Commons
Attribution (CC BY) license (https://
The usage of these type of contracts is very diverse, and they are important not only
creativecommons.org/licenses/by/
for the financial industry but for virtually every industry. They make it possible to manage
4.0/).
risk in a way that gives great financial flexibility to enterprises, consequently promoting

J. Risk Financial Manag. 2021, 14, 399. https://doi.org/10.3390/jrfm14090399 https://www.mdpi.com/journal/jrfm

J. Risk Financial Manag. 2021, 14, 399 2 of 12

economic and business growth and development. For these reasons, it is not surprising
that the number and type of financial options issued and traded has grown immensely over
the last decades. This growth required more and more elaborate models to accommodate
the new complexity of the contracts. Also, the greater size and impact of the usage of
these contracts made evident the first models widely used were not good enough. Recent
very unfortunate economic events such as the sub-prime crisis in the United States in 2008
were partially caused by the misuse of financial derivatives, and this demonstrated the
importance of properly modeling and understand these contracts. The research proposed
in this document aims to study new mathematical models that take into account some
often neglected features of financial markets.
The most well known model for pricing financial options is called the Black-Scholes
(BS) model and, although still largely used, it has multiple shortfalls like the fact that it
does not account for feedback effects or transaction costs. The BS model does not consider
that a trade can have an impact on the price, however, it has been empirically verified
that a very large trader (such as an investment bank) can affect the assets’ prices upon
performing a large trade. Also, virtually every market has transaction costs or a different
price for buying versus selling an asset (bid-ask spreads), which is not considered by the
BS model as it assumes continuous cost-free trading to perfectly hedge the portfolio. Due
to these shortfalls, in many situations, the Black-Scholes model is not sufficient for a robust
application. Consequently a lot of research has been made on new models that extend
Black-Scholes considering at least one of the previous characteristics. These extended
models often result in non-linear pricing equations, which we introduce below.
The Leland (1985) model was one of the first and most popular extensions of the
Black-Scholes to accommodate transaction costs. This model assumes only discrete trading
at a pre-specified time intervals as opposed to the Black-Scholes model where trading is
made continuously. Following Leland’s approach, a model involving variable transaction
costs has been introduced and analyzed in Ševčovič and Žitňanská (2016). For an overview
of nonlinear option pricing models of the Leland type under transaction costs we refer to
Ševčovič (2017).
A model that considers proportional transaction costs was introduced by Barles and
Soner (1998). The authors apply a utility maximization approach, where they consider
an economic agent with a constant risk-aversion. Using asymptotic analysis where the
transaction costs were taken to zero and the risk-aversion to infinity they found that, in the
limit, price of an European style option is given by a PDE of the Black-Scholes type where
the volatility nonlinearly depends on the Gamma of the option price.
The concept of a utility function and even expected utility has been known and
used for several decades in economics in general and its usage to price financial deriva-
tives has gained a lot of momentum in recent years. The idea was first formulated by
Hodges and Neuberger (1989), however, their work not fully formalized and mathemati-
cally proved which was then done by other authors such as Davis et al. (1993) that proposed
a numerical scheme for solving the equation. Barles and Soner worked on that same model
and provided an analytical study by take asymptotic limits on the transaction costs and
risk-aversion coefficient (c.f. (Barles and Soner 1998)). This has become a very well-known
model due to many practical reasons. Indifference pricing theory was presented in the
book Carmona and Çinlar (2009) which covers, in a very deep and comprehensive matter,
the subject of pricing via utility maximization.
In Davis et al. (1993) the authors investigated the problem of pricing European options
in a Black-Scholes model with proportional costs on stock transactions and they defined
the option writing price as the difference between the utilities achievable by going into the
market to hedge the option and by going into the market on one’s own account. Without
transaction costs, this definition is shown to yield the usual Black-Scholes price. To compute
the option price under transaction costs, one has to solve two stochastic control problems,
corresponding to the two utilities compared above. The value functions of these problems
are shown to be the unique viscosity solutions of one fully nonlinear quasi-variational
J. Risk Financial Manag. 2021, 14, 399 3 of 12

inequality, with two different boundary and terminal conditions. They constructed a stable
and convergent discretization scheme based on the binomial approximation of the stock
price process. A generalization of this model was done by Cantarutti et al. (2017), where
besides having proportional transaction costs, the underlying stock price dynamics was
considered to have the form of a general exponential Lévy process. Numerical results are
obtained by Markov chain approximation methods when the returns follow a Brownian
motion and a variance gamma process.
In Monoyios (2004), an efficient algorithm is developed to price European options
in the presence of proportional transaction costs, using the optimal portfolio framework
of Davis et al. in Dempster and Pliska (1997). In this approach, the fair option price is
determined by requiring that an infinitesimal diversion of funds into the purchase or
sale of options has a neutral effect on achievable utility. This results in a general option
pricing formula, in which option prices are computed from the solution of the investor’s
basic portfolio selection problem, without the need to solve a more complex optimisation
problem involving the insertion of the option payoff into the terminal value function.
The option prices are computed numerically using a Markov chain approximation to the
continuous time singular stochastic optimal control problem, for the case of exponential
utility. Comparisons with approximately replicating strategies are made. The method
results in a uniquely specified option price for every initial holding of stock, and the price
lies within bounds which are tight even as transaction costs become large. A general
definition of an option hedging strategy for a utility maximising investor is developed that
involved calculating the perturbation to the optimal portfolio strategy when an option
trade is executed.
In Kallsen and Muhle-Karbe (2015), asymptotic formulas for utility indifference
prices and hedging strategies in the presence of small transaction costs were obtained.
Perrakis and Lefoll (2000) derived optimal perfect hedging portfolios in the presence of
transaction costs. In the paper Dong and Lu (2021) the price of a European option with pro-
portional transaction costs has been determined using a utility indifference approach where
the resulting Hamilton-Jacobi-Bellman equation for the portfolio without option is two-
dimensional instead of three-dimensional as in standard utility indifference approaches
(c.f. (Davis et al. 1993)).
Furthermore, Li and Wang (2009) study the application of the penalty method to solve
the resulting variational inequality. Song Wang and Wen Li have published numerical re-
sults of an implementation of the penalty method to price for both American and European
style options (c.f. (Li and Wang 2014)). They considered exponential utility which is by far
one of the most studied utilities but still slightly restrictive.
Our goal is to analyze the system of two Hamilton-Jacobi-Bellman (HJB) equations.The
option price is constructed as a difference of the certainty equivalents to the value functions
solving the system of HJB equations. We introduce a transformation method for solving
the penalized nonlinear partial differential equation. The transformed equation involves
possibly non-constant and non-zero risk aversion function containing the negative ratio
between the second and first derivatives of the utility function. Using the parabolic
comparison principles we derive useful bounds on the option price. We also propose a
finite difference numerical discretization scheme with some computational examples.
The paper is organized as follows. The next section is focused on generalization of
the utility indifference option pricing model. We consider a general class of concave utility
functions. A system of two Hamilton-Jacobi-Bellman equations is derived. The option
price is then obtained in terms of a difference of their solutions. In Section 3 we present a
transformation method for solving the penalized nonlinear partial differential equation.
The penalized equation involves the risk aversion function. Section 4 is devoted to con-
struction of a numerical scheme which is based on time implicit backward Euler method
in combination with an upwind finite difference method for spatial discretizations. The
Hamilton-Jacobi-Bellman equations are solved by means of the penalty method utilizing
J. Risk Financial Manag. 2021, 14, 399 4 of 12

the policy iteration method. It contains numerical examples of option prices for various
concave utility functions.

2. Utility Indifference Option Pricing Model

In economics, a utility function is a function measuring the economic agent’s prefer-
ences on different goods. In a financial context the utility function is usually applied to
monetary quantities, and it can be used to measure the agent’s risk aversion when in a
context of uncertainty.
The usual requirements for an utility function are that it is continuous and non-
decreasing function. Additional properties such as concavity or convexity can be shown to
be directly related with the investor’s risk aversion (see below).

2.1. Risk Aversion and the Concept of a Certainty Equivalent

If an investor’s wealth at a future time T is affected by a source of uncertainty then it
can be modeled by a random variable, say WT which we assume to have finite expectation
E[WT ] =: w. Then, we know from Jensen’s inequality that if U : R → R is a concave
function then
E[U (WT )] − U (w) ≤ 0.
This difference can be seen as how much an investor prefers to hold an uncertain amount
(which can turn out to be greater or lower that its average) or its average. The greater the
concavity of U the greater that difference, and that leads one to define the Arrow-Pratt
measure of absolute risk-aversion (also referred to as the coefficient of risk-aversion),

R(ξ ) ≡ −U 00 (ξ )/U 0 (ξ ).

For a concave increasing utility function U we have R ≥ 0. Now, for pricing financial
options one still needs one more concept, the concept of certainty equivalent which we
denoted by v. It is defined as follows:

v : U (v) = E[U (WT )], v = U −1 (E[U (WT )]).

Throughout the paper we shall consider various types of utility functions with different risk-
aversion profiles. Their profiles U ( x ) and the risk aversion coefficients
R(ξ ) = −U 00 (ξ )/U 0 (ξ ) are shown in Table 1.

Table 1. Utility functions, their inverse functions and risk aversion functions.

Type Utility Function Parameter Inverse Utility Function Risk Aversion

Linear U (ξ ) =ξ — U −1 ( y ) =y R(ξ ) =0
Exp. U (ξ ) = 1 − e−γξ γ>0 U −1 ( y ) = − ln(1 − y)/γ R(ξ ) =γ
Power U (ξ ) = ξa a<1 U −1 ( y ) = y1/a R(ξ ) = (1 − a ) ξ −1
Log. U (ξ ) = ln(bξ + 1) b>0 U −1 ( y ) = (ey − 1)/b R(ξ ) = b/(bξ + 1)

The power utility function U (ξ ) = ξ a and the logarithmic utility function

U (ξ ) = ln(bξ + 1) belong to the wide class of the so-called decreasing absolute risk
aversion (DARA) utility function characterized by a decreasing risk aversion coefficient
R(ξ ) considered as a function of the wealth ξ. In the context of dynamic stochastic portfolio
optimization the importance of DARA utility functions has been investigated in papers by
Post et al. (2015); Kilianová and Ševčovič (2018).

2.2. Utility Indifference Option Pricing Model under Transaction Costs

To price a derivative in a market with proportional transaction costs using this frame-
work we proceed as follows.
J. Risk Financial Manag. 2021, 14, 399 5 of 12

Firstly, we consider that we have an investor that can invest in either a risky asset or a
risk-free asset following the dynamics of a geometric Brownian motion and exponential
deterministic process, respectively. That is

dS = µSdt + σSdwt , dB = rBdt, (1)

where {wt , t ≥ 0}, stands for the standard Wiener stochastic process.
We model the transactions costs by introducing two different prices for the risky
asset, depending on weather the investor is buying (ask price) or selling (bid price) the
underlying asset,
Sask = (1 + θ )S, Sbid = (1 − θ )S,
where θ = (Sask − Sbid )/(2S) represents the bid-ask spread factor.
We assume that the investor can buy or sell shares of the risky asset (shares account
αt ) by increasing or decreasing his holdings in the risk-free asset (money account β t ). We
model the cumulative purchase of risky assets by a process Lt and the sale by the process
Mt . The goal of an investor is to maximize his terminal utility and will chose his trades Lt ,
Mt accordingly. The resulting dynamics for the investors portfolio are,

dα = dL − dM, dβ = rβdt − (1 + θ )SdL + (1 − θ )SdM. (2)

Now, we can define the liquid wealth Wt of the investor as follows:

Wt = β t + St (αt − θ |αt |).

Then the option pricing problem can be formulated as a stochastic control problem. Firstly,
we introduce a portfolio in which the investor is optimizing their expected utility by trading
either stock or bonds,
v0 (α, β, s, t) = sup E[U (WT )|α, β, s]. (3)
L,M

Secondly, we consider that the investor has at his disposal another portfolio, which is
equally comprised of risk-free and risky assets but also a short buyer position with δ = −1
or long seller position with δ = +1 on a derivative with the terminal payoff CT . The value
vδ of this second portfolio is given by,

vδ (α, β, s, t) = sup E[U (WT + δCT (s))|α, β, s]. (4)

L,M

Let us denote the certainty equivalents of portfolios by zδ = zδ (α, β, s, t) and

z0 = z0 (α, β, s, t), respectively. The functions zδ , z0 satisfy the system of equations:

U ( w − z0 ) = v0 , U (w − zδ ) = vδ , w = β + s(α − θ |α|). (5)

For further details of utility indifference option pricing we refer to the book Carmona and
Çinlar (2009), the price V of an option with a payoff diagram CT is given as the discounted
difference of certainty equivalents, i.e.,

V = e −r ( T − t ) ( z δ − z0 ), where zδ − z0 = U −1 (v0 ) − U −1 (vδ ). (6)

In order to determine solutions zδ and z0 we have to solve a pair of stochastic optimal

control problems for vδ , and v0 , by means of the dynamic programming principle. Mathe-
matical representation leads to a system of two Hamilton-Jacobi-Bellman (HJB) equations
that we introduce and analyze in the next subsection.
J. Risk Financial Manag. 2021, 14, 399 6 of 12

2.3. Hamilton-Jacobi-Bellman Equations for Value Functions

Following Davis et al. (1993) the functions vδ and v0 satisfy the system of variational
inequalities of the form:
min(V A (v), V B (v), VC (v)) = 0, (7)
where the linear differential operators V A , V B , VC are defined as follows:

σ2 2 2
V A ≡ ∂t + s ∂s + µs∂s + rβ∂ β , V B ≡ −∂α + s(1 + θ )∂ β , VC ≡ ∂α − s(1 − θ )∂ β ,
2
and terminal conditions,

v0 (α, β, s, T ) = U (w(α, β, s)), vδ (α, β, s, T ) = U (w(α, β, s) + δCT (s)). (8)

Here CT (S) denotes the prescribed pay-off diagram, i.e., CT (S) = (S − K )+ in the case of
a plain vanilla call option, or CT (S) = (K − S)+ in the case of a put option. Here K > 0
denotes the strike price, ( β)+ = max( β, 0), and ( β)− = min( β, 0) denote the positive
and negative parts of a real number β. Recall that the no-transaction region of values
(α, β, s, t) is characterized by the equation V A (v) = 0. The Buy and Sell regions correspond
to equations V B (v) = 0 and VC (v) = 0, respectively (c.f. (Davis et al. 1993)).
The minimal Equation (7) is equivalent to the following linear complementarity
problem for the functions v = v0 , and v = vδ :

V A (v) ≥ 0, V B (v) ≥ 0, VC (v) ≥ 0, V A (v) · V B (v) · VC (v) = 0. (9)

2.4. Penalty Method for Solving HJB Equations

With penalty methods, the initial variational inequality is replaced by one single
equation which has a term parameterized by a small parameter. One should prove that
the solution of this new equation will converge to the initial one. Besides convergence to
the initial problem, the perturbed equation’s solution will always respect the constrains
posed by the initial problem. Next, introduce the specific implementation of the penalty
method for our pricing model following Li and Wang (2009, 2014). The penalty method has
been successfully adopted for solving various nonlinear option pricing model by Lesmana
and Wang (2015), or Chernogorova et al. (2018), and others. The optimal time dependent
penalty function has been proposed recently by Clevenhaus et al. (2020).
Next we introduce the penalty method in a more detail. Let us define the following
penalized perturbed equation for the function v = vλB ,λC (α, β, s, t),

V A (v) + λ B [V B (v)]− + λC [VC (v)]− = 0. (10)

Here λ B , λC  0 are sufficiently large penalty parameters. In what follows, we will

drop the subscripts for the sake of simplicity v := vλB ,λC . In the limit λ B , λC → ∞ we
formally deduce that the limiting solution v solves the linear complementarity problem.
Indeed, V A (v) = −λ B [V B (v)]− − λC [VC (v)]− ≥ 0, and V B (v), VC (v) ≥ 0 in the limit
λ B , λC → ∞. Taking λ B , λC → ∞ such that λ B /λC → ∞ we obtain V A (v)V B (v) = 0.
Similarly, V A (v)VC (v) = 0.

3. Transformation of the HJB Equation Involving Risk Aversion Function

For a general utility function U we search the solution v = vδ , δ = 0, ±1, in the
following form:

v(α, β, s, t) = U (eµ(T −t) (w(α, β, s) + A (t) + V (α, β, s, t))),

where
β
A (t) = (r − µ)(1 − e−µ(T −t) )
µ
J. Risk Financial Manag. 2021, 14, 399 7 of 12

is a time dependent shift function such that A ( T ) = 0. Notice that ∂2s w = 0, ∂ β w = 1, and
s∂s w = w − β. Hence

V A (eµ(T −t) (w + A )) = eµ(T −t) (−µ(w + A ) + A 0 + µ(w − β) + rβ)

= eµ(T −t) (A 0 − µA + β(r − µ)) = 0,

because of the definition of the auxiliary function A (t). For any function z = z(α, β, s, t)
we have

σ2
V A (U (z)) = U 0 (z)(∂t z + µs∂s z + rβ∂ β z + s2 ∂s (U 0 (z)∂s z))
2

σ 2 
0 2
= U (z) V A (z) − R(z)(s∂s z)) ,
2

where R(z) = −U 00 (z)/U (z) ≥ 0 is the risk aversion function. Taking z = eµ(T −t) (w +
A + V ) we obtain

σ2
 
µ( T −t) 2
V A (v) = F (α, β, s, t) V A (V ) − µV − R(z)e (s∂s w + s∂s V ) ,
2

where F (α, β, s, t) = U 0 (z)eµ(T −t) > 0 is a positive factor. For V B (w), and VC (w) we have
 
2sθ, if α > 0, 0, if α > 0,
V B (w) = VC (w ) =
0, if α ≤ 0, 2sθ, if α ≤ 0.

Furthermore, as V B (A ) = VC (A ) = 0, we have

V B (v) = F (α, β, s, t)(V B (w) + V B (V )), VC (v) = F (α, β, s, t)(VC (w) + VC (V )).

A solution V = V δ is subject to the terminal condition at t = T:


 0, if δ = 0,
V (α, β, s, T ) = CT ( s ) , if δ = 1, (long seller position), (11)
− CT ( s ) , if δ = −1, (short buyer position).


Summarizing, we deduce that the penalized problem (10) can be reformulated in terms of
the function V as follows:

σ2
V A (V ) − µV − R(z)eµ(T −t) (s∂s w + s∂s V )2
2
+ λ B [V B (w + V )]− + λC [VC (w + V )]− = 0. (12)

Recall that the option price V is obtained as the difference between certainity equiva-
lents z0 and zδ . It means that

V = e−r(T −t) (zδ − z0 ) = e−r(T −t) (U −1 (v0 ) − U −1 (vδ )) = e(µ−r)(T −t) (V 0 − V δ ).

In the next proposition we compare a solution to the system of transformed HJB equations
with the explicit solution to the linear Black-Scholes equation:

σ2 2 2
∂t V + s ∂s V + µs∂s V − µV = 0, V (s, T ) = δCT (s), δ = 0, ±1.
2
In the call option case where CT (s) = (s − K )+ the price V (s, t) = VBS (s, t) is given by an
explicit formula:  
VBS (s, t) = δ sΦ(d1 ) − Ke−µ(T −t) Φ(d2 ) ,
J. Risk Financial Manag. 2021, 14, 399 8 of 12

√ Rd 2
where d1,2 = (ln(s/K ) + (µ ± σ2 /2)( T − t))/(σ T − t) and Φ(d) = (2π )−1/2 −∞ e−ξ /2 dξ.
A similar formula is available for pricing of put options.

Proposition 1. Assume U is an exponential (γ > 0) or linear (γ = 0) utility function, i.e., its risk
aversion function R(ξ ) = −U 00 (ξ )/U 0 (ξ ) ≡ γ where γ ≥ 0 is a non-negative constant. Then the
solution V of the penalized problem (12) satisfying the terminal condition (11) is independent of
the factor β, i.e., V = V (α, s, t). Consequently, the option price V = V (α, s, t) is independent of
β. Moreover,
V (α, s, t) ≤ VBS (s, t), for all α, s > 0, t ∈ [0, T ].

Proof. Notice that R(ξ ) ≡ γ, and s∂s w, V B (w), VC (w), as well as the terminal condition (11)
are independent functions of the factor β. The penalized Equation (12) can be rewritten
as follows:

σ2 2 2
∂t V + s ∂s V + µs∂s V + µβ∂ β V − µV (13)
2
σ2 µ( T −t)
= γe (s∂s w + s∂s V )2 − λ B [V B (w) + V B (V )]− − λC [VC (w) + VC (V )]− .
2
The right-hand side of (13) is nonnegative and it does not explicitly depend on β, so does
the solution V = V (α, s, t). Furthermore,

σ2 2 2
∂t V + s ∂s V + µs∂s V − µV ≥ 0.
2
As the Black-Scholes solution VBS satisfies

σ2 2 2
∂t VBS + s ∂s VBS + µs∂s VBS − µVBS = 0,
2
then, taking into account V (α, s, T ) = VBS (s, T ), applying the maximum principle for parabolic
equations on unbounded domains due to ((Meyer and Needham 2014), Theorem 3.4), we
obtain the inequality V (α, s, t) ≤ VBS (s, t) for a given parameter α and all s > 0, t ∈ [0, T ],
as claimed.

The following proposition is a direct consequence of Proposition 1.

Proposition 2. Assume U is the linear utility function, i.e., its risk aversion function R(ξ ) ≡ 0.
Suppose that there are no transaction costs, i.e., θ = 0. Then the solution V of the penal-
ized problem (12) satisfying the terminal condition (11) is independent of the factors α, β, i.e.,
V = V (s, t) = VBS (s, t), i.e., V is the Black-Scholes price of a European style option.

Proof. Since θ = 0 we have V B (v) = −VC (v) for any function v. Hence v is a solution to (9)
if and only if V B (v) = VC (v) = 0 and V A (v) ≥ 0. Therefore, a solution V to the penalized
problem (12) satisfies the linear Black-Scholes equation. Hence, V = V (s, t) = VBS (s, t),
as claimed.

4. Construction of a Numerical Discretization Upwind Finite Difference Scheme

In this section we propose a numerical discretisation scheme and several compu-
tational examples. The scheme is based on the finite difference method proposed in
Li and Wang (2009). The resulting scheme is of upwind type in the space discretization
and the backward Euler implicit scheme in time.
J. Risk Financial Manag. 2021, 14, 399 9 of 12

4.1. Finite Difference Approximation of a Solution to the Penalized Problem

We first introduce Ωb the truncated domain corresponding to the solvency region
where β + S(α − θ |α|) > 0 as follows:

Ωb = {(α, β, S) ∈ ( L− + − + +
α , Lα ) × ( L β , L β ) × (0, S ) : β + S ( α − θ | α |) > 0}.

We consider a simple 3D uniform mesh grid:

( α i , β j , Sk ) ∈ Ω b , i = 0, · · · , Nα , j = 0, · · · , Nβ , k = 0, · · · , NS ,

αi = L −
α + ihα , β j = L−
β + jh β , Sk = khS ,
− −
hα = ( L+
α − Lα ) /Nα , h β = ( L+
β − L β ) /Nβ , hS = S+ /NS ,

where Nα , Nβ , and NS are the numbers of discretization steps in the α, β, and S variables.
Notice that the spatial discretization can be easily adopted to a non-uniform grid, e.g., by
considering uniform discretization for the logarithmic variable xk = ln(Sk /K ). We consider
a uniform time discretization with time steps n∆t for n = N, · · · , 1, 0, where ∆t = T/N,
and N is the number of time discretization steps. The solution v = v(α, β, S, t) will be
n at ( α , β , S ) and time t = n∆t.
discretized by the value Vijk i j k
We define the following finite difference discretization operators:
n +1
Vijk n
− Vijk V(ni±1) jk − Vijk
n Vin( j±1)k − Vijk
n
n
Dt Vijk = , Dα± Vijk
n
=± , Dβ± Vijk
n
=± .
∆t hα hβ
(14)
Vijn(k+1) − Vijk
n Vijn(k+1) − 2Vijk
n + Vn
ij(k −1)
n n
DS Vijk = , DSS Vijk = ,
hS h2S

σ2 2
 
−
n
LA Vijk = − Dt + r ( β j ) +
+ r( β j )Dβ+ Dβ− + µSk DS + Sk DSS Vijk n
,
2 (15)
   
n
LB Vijk = − Dα+ + (1 + θ )Sk Dβ− Vijk
n
, n
LC Vijk = Dα− − (1 − θ )Sk Dβ+ Vijk
n
.

Clearly, for any λ > 0, and L ∈ R, we have


− 0, if L > 0,
λ[L] = min mL =
m∈[0,λ] λ L, if L ≤ 0.

The numerical discretization scheme is then given by:

n n n
LA Vijk + min m̄LB Vijk + min n̄LC Vijk = 0, ∀i, j, k. (16)
m̄∈[0,λ B ] n̄∈[0,λC ]

Terminal conditions.
For the last terminal time level n = N we have, for the call (put) option case with the
pay-off diagram CT (S) = (S − K )+ (CT (S) = (K − S)+ ),
N
Vijk = U ( β j + Sk (αi − θ |αi |) + δCT (Sk )), for (αi , β j , Sk ) ∈ Ωb .

Boundary conditions. We apply the Dirichlet boundary conditions, i.e.,

n
Vijk = U ( β j + Sk (αi − θ |αi |) + δCT (Sk )), for (αi , β j , Sk ) ∈ ∂Ωb , n = N − 1, · · · , 1, 0.

Here we set δ = 0 in the case of numerical approximation of the value function v0 , and
δ = ±1 in the case of approximation of the value function vδ , δ = ±1.
J. Risk Financial Manag. 2021, 14, 399 10 of 12

Next, we present the full numerical discretization algorithm involving the policy itera-
tion method for solving the penalized PDE (16). Notice that ( β j )+ + ( β j )− = β j . It yields
the following system of linear equations for the unknown vector V n for n = N − 1, · · · , 1, 0,
!
σ2 2

r µ n,p+1
1 + ∆t β + S + S Vijk
h β j hS k h2S k
σ2 ∆t 2 n,p+1 σ2 ∆t 2 n,p+1
 
µ∆t
− Sk + S V − S V
hS 2h2S k ij(k+1) 2h2S k ij(k−1)
! !
(1 − θ )

1 (1 + θ ) 1 n,p
+∆t m̃ + Sk + ñ + Sk Vijk (17)
hα hβ hα hβ
 rβ+  rβ−
(1 − θ )
 
j n,p j (1 + θ ) n,p
−∆t + ñ Sk Vi( j+1)k − ∆t + m̃ Sk Vi( j−1)k
hβ hβ hβ hβ
∆t n,p ∆t n,p n +1
−m̃ V(i+1) jk − ñ V(i−1) jk = Vijk ,
hα hα
n,p n,p
where p = 0, · · · , pmax is the policy iteration parameter, m̃ = m̃ijk , ñ = ñijk are arguments
of the minimum in (16). The above system of linear equations for the unknown stacked
n,p+1
vector V = (Vijk ) can be rewritten as a system of linear equations of the form AV = b
where A is a sparse matrix with at most 3 nonzero elements in each row. The right-hand
side vector b consists of the known vector V n+1 complemented by the boundary conditions.
It is important to note that the coefficients of the matrix A depend on the coefficients m̃, ñ
which has to be computed within each policy iteration step. The full algorithm for the
computation of the value function is as in the Algorithm 1.

Algorithm 1: The algorithm for computing the value function V for δ = 0, ±1.
Initialization of model parameters and numerical parameters pmax , tolmax > 0;
Compute the terminal conditions for n = N.
N = U (w
Set Vi,j,k i,j,k + δCT ( Sk )) for each i, j, k;
Set n = N − 1;
while n > 0 do
n,0 n +1
Initiate policy iteration p = 0 with Vijk = Vijk ;
n +1
Compute the right-hand side vector b from Vijk and boundary conditions ;
while p < pmax and tol ≥ tolmax do
n,p n,p
Compute the penalty terms. Set m̃ijk = 0, ñijk = 0
n,p
if V B (Vijk ) < 0 then
n,p
m̃ijk = λ B
end
n,p
if VC (Vijk ) < 0 then
n,p
ñijk = λC
end
Compute the elements of the matrix A ;
Solve the linear system of equations A V = b;
n,p
Compute the difference tol = max |Vijk − Vijk |;
n,p+1
Vijk ← Vijk ;
p ← p+1
end
n ← n − 1;
end
J. Risk Financial Manag. 2021, 14, 399 11 of 12

Remark 1. Our numerical scheme is based on solving the system of linear Equation (16) for the
unknown stacked vector (Vijk n ). Its matrix representation contains at most 3 nonzero elements

in each row. It has a block matrix structure with Nα × Nβ tridiagonal NS × NS matrices on

the block diagonal. For each NS × NS tridiagonal we can employ the fast Thomas algorithm
with time complexity O( NS ). The overall complexity of computation of the system is therefore
O( Nα × Nβ × NS × pmax ) where pmax is the maximal number of policy iterations. Recall that there
are other fast and robust numerical methods for solving problems of the form (16). Among them
there are alternating direction explicit (ADI) methods for linear, nonlinear and multi-dimensional
Black-Scholes models (c.f. (Bučková et al. 2017) and references therein).

4.2. Results of Numerical Approximation of Option Prices

In this part, we present results of computation of European style call options for the
exponential utility function with a risk parameter γ > 0 and linear utility function.
The model and numerical parameters used can be found in Table 2. We used the
exponential mesh Sk = K ln( xk ) where { xk , k = 1, · · · , NS } is an equidistant mesh of
the interval [K/2, 2K ]. The plot of the call option price as a function of the underlying
asset price S is shown in Figure 1, for the buyer call option prices, i.e., δ = −1. We used
Matlab framework for computation of the solution on 3GHz Intel single Core machine.
For numerical discretization parameters shown in Table 2 The computational time was
8.1 sec per one policy iteration. According to Remark 1 the overall complexity is of the
order O( Nα × Nβ × NS × pmax ).

Table 2. Model and numerical parameters of the numerical solution.

Model Parameters Value Num Params Value Num Params Value

Strike price K 50 Nα 6 NS 100
Transaction costs θ 0.01 L−
α 0.2 S+ 100
Volatility σ 0.3 L+
α 0.6 λ B = λC 10
Risk-free rate r 0.05 Nβ 6 N 10
Drift µ 0.1 L−
β −100 T 1
Risk-aversion γ 0.1 L+
β 100

30 4

3.5
25

20
2.5

V V
15 2

1.5
10

5
0.5

0 0
35 40 45 50 55 60 65 70 75 42 44 46 48 50 52 54
S S

Figure 1. Linear (left) and exponential (right) utility indifference call option buyer price as a function
of the underlying asset price S with the wealth function parameters α = 0.467, β = 33.3.

5. Conclusions
In this paper we investigated and analyzed the system of Hamilton-Jacobi-Bellman
equations arising in pricing financial derivatives. We followed the utility indifference
option pricing model in which the option price is constructed as a difference of the certainty
equivalents to the value functions solving the system of HJB equations. We analyzed
solutions to the transformed nonlinear partial differential equation involving a possibly
non-constant risk aversion function. Useful bounds on the option price were obtained
J. Risk Financial Manag. 2021, 14, 399 12 of 12

using parabolic comparison principle. We also proposed a finite difference numerical

discretization scheme. Various computational examples were also presented.

Author Contributions: The authors contributed equally to this research. Both authors have read and
agreed to the published version of the manuscript.
Funding: This research received no external support.
Acknowledgments: The authors gratefully acknowledge the contribution of the Slovak Research
and Development Agency under the project APVV-20-0311.
Conflicts of Interest: The authors declare no conflict of interest.

References
Barles, Guy, and Halil Mete Soner. 1998. Option pricing with transaction costs and a nonlinear Black-Scholes equation. Finance and
Stochastics 2: 369–97. [CrossRef]
Bučková, Zuzana, Matthias Ehrhardt, Michael Günther, and Pedro Pólvora. 2017. Alternating direction explicit methods for linear,
nonlinear and multi-dimensional Black-Scholes models. In Novel Methods in Computational Finance. Cham: Springer, pp. 333–71.
Cantarutti, Nicola, Joao Guerra, Manuel Guerra, and Maria do Rosário Grossinho. 2017. Indifference pricing in a market with transaction
costs and jumps. In Novel Methods in Computational Finance. Cham: Springer, pp. 31–46. [CrossRef]
Carmona, René, and Erhan Çinlar. 2009. Indifference Pricing: Theory and Applications. Princeton: Princeton University Press.
Chernogorova, Tatiana P., Miglena N. Koleva, and Radoslav L. Valkov. 2018. A two-grid penalty method for American options. Journal
of Computational and Applied Mathematics 37: 2381–98. [CrossRef]
Clevenhaus, Anna, Matthias Ehrhardt, Michael Günther, and Daniel Ševčovič. 2020. Pricing American Options with a Non-Constant
Penalty Parameter. Journal of Risk and Financial Management 13: 124. [CrossRef]
Davis, Mark H. A., Vassilios G. Panas, and Thaleia Zariphopoulou. 1993. European option pricing with transaction costs. SIAM Journal
on Control and Optimization 31: 470–93. [CrossRef]
Dempster, Michael A. H., and Stanley R. Pliska, eds. 1997. Mathematics of Derivative Securities. Volume 15 of Publications of the Newton
Institute. Cambridge: Cambridge University Press.
Dong, Yan, and Xiaoping Lu. 2021. Utility-indifference pricing of European options with proportional transaction costs. Journal of
Computational and Applied Mathematics 397: 12. [CrossRef]
Hodges, Stewart D., and Anthony Neuberger. 1989. Optimal replication of contingent claims under transaction costs. Review of Futures
Markets 8: 222–39.
Kallsen, Jan, and Johannes Muhle-Karbe. 2015. Option pricing and hedging with small transaction costs. Mathematical Finance 25:
702–23. [CrossRef]
Kilianová, Soňa, and Daniel Ševčovič. 2018. Expected utility maximization and conditional value-at-risk deviation-based Sharpe ratio
in dynamic stochastic portfolio optimization. Kybernetika (Prague) 54: 1167–83.
Leland, Hayne E. 1985. Option pricing and replication with transactions costs. The Journal of Finance 40: 1283–301. [CrossRef]
Lesmana, Donny Citra, and Song Wang. 2015. Penalty approach to a nonlinear obstacle problem governing American put option
valuation under transaction costs. Applied Mathematics and Computation 251: 318–30. [CrossRef]
Li, Wen, and Song Wang. 2009. Penalty approach to the hjb equation arising in european stock option pricing with proportional
transaction costs. Journal of Optimization Theory and Applications 143: 279–93. [CrossRef]
Li, Wen, and Song Wang. 2014. A numerical method for pricing european options with proportional transaction costs. Journal of Global
Optimization 60: 59–78. [CrossRef]
Meyer, John, and David Needham. 2014. Extended weak maximum principles for parabolic partial differential inequalities on
unbounded domains. Proceedings of the Royal Society A Mathematical Physical and Engineering Sciences 478: 20140079. [CrossRef]
Monoyios, Michael. 2004. Option pricing with transaction costs using a Markov chain approximation. Journal of Economic Dynamics
and Control 28: 889–913. [CrossRef]
Perrakis, Stylianos, and Jean Lefoll. 2000. Option pricing and replication with transaction costs and dividends. Journal of Economic
Dynamics and Control 24: 1527–61. [CrossRef]
Post, Thierry, Yi Fang, and Martin Kopa. 2015. Linear tests for DARA stochastic dominance. Management Science 61: 1615–29.
[CrossRef]
Ševčovič, Daniel. 2017. Nonlinear parabolic equations arising in mathematical finance. In Novel Methods in Computational Finance.
Cham: Springer, pp. 3–15.
Ševčovič, Daniel, and Magdaléna Žitňanská. 2016. Analysis of the nonlinear option pricing model under variable transaction costs.
Asia-Pacific Financial Markets 23: 153–74. [CrossRef]
Reproduced with permission of copyright owner. Further reproduction
prohibited without permission.