See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/325678567

An improvement of option pricing using the Finite difference method

Conference Paper · September 2014

CITATION READS

1 753

2 authors:

Lucia Svabova Marek Ďurica

University of Zilina University of Žilina
80 PUBLICATIONS 949 CITATIONS 53 PUBLICATIONS 432 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Lucia Svabova on 10 June 2018.

The user has requested enhancement of the downloaded file.

7th International Scientific Conference Managing and Modelling of Financial Risks Ostrava
VŠB-TU of Ostrava, Faculty of Economics, Finance Department 8th – 9th September 2014

An improvement of option pricing using the Finite

difference method
Lucia Švábová1, Marek Ďurica 2

Abstract
The paper deals with the improvement of pricing of the European options using the Finite
difference method, which is the numerical method for estimating the price of the financial
derivatives based on the known Black - Scholes model. In the paper we present the
improvement of setting the values in the endpoints of the grid using in the Finite difference
method, using which we get the estimates closer to the real prices of the options.

Key words
Finite difference method, Option pricing, Black – Scholes model, financial derivatives

JEL Classification: C02, G00

1. Introduction
Financial derivatives are instruments used in all financial markets: money, foreign
exchange, capital, and commodity market. Their main role is to protect stakeholders against
risks arising from future developments in prices of products, goods, interest rates, exchange
rates and so on. But more often is derivatives trading used for speculation purpose. Setting the
right price of financial derivatives is therefore very closely monitored. If there would be a
risk-free profit opportunities – arbitration by closing the positions in several contracts, traders
would not hesitate to use it. (Vlachynský, 2002)
There exist several methods of valuation of financial derivatives. Most of them take as
their basis the famous Black-Scholes model, which is based on partial differential equation.
Fischer Black, Myron Scholes and Robert Merton developed European option pricing model
in 1973. (Merton, 1973) This model has become a breakthrough in valuation of stock options
and M. Scholes and R. Merton (F. Black died in 1995) in 1997 received the Nobel Prize in
Economics for it. The Black – Scholes – Meton model for option pricing is at present the most
widely used tool for derivative contracts valuation. The prices of derivatives determined using
this model are nearly identical to the actual prices of contracts traded on the financial markets.
Since the introduction of the Black – Scholes – Merton model the volume of option trades
increased significantly. In addition to European vanilla options there were introduced various
kinds of more complex types of exotic options. Exotic options are important for derivatives
dealer because they are in general much more profitable than plain vanilla products. (Hull,

1
RNDr. Lucia Švábová, PhD. Katedra kvantitatívnych metód a hodpodárskej informatiky, Fakulta
prevádzky a ekonomiky dopravy a spojov, Žilinská univerzita v Žiline, e-mail:
[email protected]
2
RNDr. Marek Ďurica, PhD., Katedra kvantitatívnych metód a hodpodárskej informatiky, Fakulta
prevádzky a ekonomiky dopravy a spojov, Žilinská univerzita v Žiline, e-mail:
[email protected]
7th International Scientific Conference Managing and Modelling of Financial Risks Ostrava
VŠB-TU of Ostrava, Faculty of Economics, Finance Department 8th – 9th September 2014

2012) Exotics have the structure different from standard calls and puts. They are designed for
two reasons. The first reason is to reflect a view on potential future movements in particular
market variables and to capture better the needs of traders to ensure their portfolio. And the
second reason why exotic options are traded is that they appear to be more attractive tool for
the traders to achieve profit. (Hull, 2012) But some types of these exotic options are so
complicated that setting their prices is more difficult because the analytical formula for their
pricing do not exists. The prices of these types of exotic options could be estimated
numerically then.
One alternative method for options pricing is a Finite Difference Method, which is also
based on the Black – Scholes – Merton model. This numerical method consists of reduction of
the field of solutions to a finite number of points, and the replacement of partial derivations in
Black - Scholes model by given differences. This leads to a system of equations whose
solution will be the value of the option with given parameters.

2. Methodology and Data

One of the basic terms mentioned here is undoubtedly the term option. There are two basic
types of options. A call option gives the holder the right to buy the underlying asset by a
certain date in the future in a certain price. A put option on the other hand gives the holder the
right to sell the underlying asset by a certain price in a certain time. The price agreed by the
seller and the buyer in this contract is known as the strike price. The date, when the right
given in the option will expire, is known as the expiration date or maturity. It is important to
emphasize, that the option gives his holder the right to do something and the holder is not
obligated to exercise this right. (Hull, 2012)
The ownership of the option provides only the right to do something and this is why the
owner of the option has to pay something for conclusion of such a contract. An option buyer
pays the option premium for the right to buy or sell the underlying asset, and, like the buyer of
any other asset, faces carry costs. (Whaley, 2006)

2.1 Black – Scholes – Merton model

As we mentioned earlier, setting the correct price of each option traded on the financial
market is very important for ensuring no arbitrage opportunities on the market. One of the
basic models for pricing the financial derivatives is already mentioned Black – Scholes –
Merton model. The first model was proposed in the early 1970s by Black and Scholes as
a basic model of behavior of option prices which underlying asset is non – dividend paying
stock. F. Black and M. Scholes herewith achieved a major breakthrough in the pricing of
European stock options, because their model has had a huge influence on the way that traders
price the derivatives. The importance of their research was recognized in 1997 and they were
awarded the Nobel Prize for economics. (Hull, 2012)
The Black – Scholes – Merton (BSM) model was introduced by R. Merton in 1973.
Merton´s approach was more general and different from that of Black and Scholes. His
approach did not rely on some assumptions of the underlying asset. Merton includes into his
model the assumption of the possibility of payment the dividend yield from the underlying
share of the derivative. (Hull, 2012)
The BSM model is a solution of BSM differential equation. This equation has to be
satisfied by the price of every type or financial derivative, whose underlying asset is a non-
dividend paying stock. In other words, the prices of the options dependent on non-dividend
paying stock are governed by this model. To derive the BSM differential equation, several
assumptions have to be satisfied. The stock price has to follow the geometric Brownian
7th International Scientific Conference Managing and Modelling of Financial Risks Ostrava
VŠB-TU of Ostrava, Faculty of Economics, Finance Department 8th – 9th September 2014

motion with constant expected return μ and constant stock price volatility σ. The short selling
of securities is permitted with full use of the proceeds. No transaction costs and taxes exist
and all the securities are perfectly divisible. No dividends are paid from the underlying share
during the lifetime of the derivative. No arbitrage opportunities exist and trading with
securities is continuous. The risk – free interest rate r is constant and the same for all times of
maturity. (Hull, 2012)
Some of these assumptions can be relaxed by some way. For example the parameters of the
stock price process can be known functions of the time, the interest rate can be stochastic
under some conditions and the underlying share can bring dividend yield during the lifetime
of the derivative. (Hull, 2012)
Under the assumptions mentioned hereabove, we obtain the BSM differential equation:
(Hardik, 2008)
(1)
The BSM equation (1) is derived using the portfolio consisting of the long position in a
share and short position in its derivative. The portfolio which must be riskless during a small
interval of time, so that it must instantaneously earn the same rate of return as the other short
term risk – free securities. In equation (1) is used the Merton´s approach, which allows the
existence of continuous dividend yield q, paid from the price of the underlying asset. The
following notation was used in equation (1): f is the price of the derivative (option), t is the
time, S is the price of the underlying share, r is risk – free interest rate, q is dividend yield
from the asset, σ is the stock price volatility.
This equation has infinite number of solutions based on the selected initial conditions. For
a European call option the solution of the BSM equation is a formula: (Hull, 2012)
, (2)
where and are the values of the distribution function of a standard normally
distributed random variable in and given by

. (3)
The formula (2) is the solution of the BSM differential equation (1) and is known as the
Black – Scholes – Merton model (pricing formulas) for the price of European call options.
The notation is as before, in addition here T is the time of maturity of the option.
2.2 Finite difference method
The BSM equation is a partial differential equation which solution will be significantly
easier if we reduce the domain of the definition into a finite number of points. This
consideration is the keynote of the Finite difference method.
First, suppose that the lifetime of the option is T. We divide the time interval into N
equally spaced intervals of length . Second, suppose that a maximum stock price is
which is so high that, when it is reached, the corresponding put option has virtually no value.
We divide the whole stock price interval into M equally spaced subintervals with
length .
Than we reduce all the considerations into only grid points, which is defined as the time
points and stock price points. The variable will denote the value of the option in time
with stock price . The whole grid consists of points. (Hull, 2012)
The keynote the Finite difference method is to replace the partial derivations in the BSM
differential equation (1) with given differences. Every interior point of the grid has the value
of the partial differential equations approximated with the differences of the values of the
7th International Scientific Conference Managing and Modelling of Financial Risks Ostrava
VŠB-TU of Ostrava, Faculty of Economics, Finance Department 8th – 9th September 2014

option in neighboring nodes. There exist two basic ways to use the Finite difference method,
depending on the differences used:
 The Implicit Finite difference method uses the approximations, which by being
substituted to equation (1) leads to a system of equations for the unknown value of
the derivative at internal points of the grid. By repeating the procedure for solving
the system we obtain the derivative price at the time zero.
 The Explicit Finite difference method uses such approximations of partial
derivations, which by being substituted to equation (1) leads to the formula for
calculating the price of the derivative at that point directly:
(4)
The equation (4) gives the relationship between three values of the option at time in
points , and and one value of the option in time in the point as
shown in following figure of the Explicit Finite difference method.

Figure 1 Explicit Finite difference method grid

S

Smax

fi,j+1

fi,j fi+1,j

fi,j-1

S =0
t
t=0 t=T

2.3 The improvement of using the Finite difference method for option pricing
In the Finite difference method grid showed on the Figure 1 we move from the end of the
grid to the beginning. Or by other words we move from the right side of the grid to the left. At
each node we calculate the value of the derivative using the expression (4). The main
assumption is that the values of the option in the endpoints of the grid are known:
 For a share with zero prices the call option has zero prices too. So in the lower
points on the bottom of the grid the call option has zero prices.
 At the time of expiration T the value of the option is given by the payoff function.
So in the endpoints on the right side of the grid the call option has the value given
by t
 he difference .
 Using the first assumption hereabove, for a share with maximum price in the
upper points of the grid, the corresponding put option has zero prices too. So that
the price of call option follows from the put – call parity. (Švábová, Ďurica, 2013)
The third assumption described hereabove is the main difference between the approach
given in (Hull, 2012) and the improvement suggested by authors of this paper. In (Hull, 2012)
and other sources the third assumption using for the values of the option in the upper points of
the grid is given by for . So that the value of the call option in the upper
7th International Scientific Conference Managing and Modelling of Financial Risks Ostrava
VŠB-TU of Ostrava, Faculty of Economics, Finance Department 8th – 9th September 2014

points of the grid was still constant equal to exercise price of the option. The improvement of
using the Finite difference method for option pricing is based on using the put – call parity for
the values of call option in the upper point of the grid (or for lower points of the grid for put
option respectively).
The put – call parity provides the relationship between the prices of the European call and
put options when they have the same strike price X and time to maturity T. If the underlying
asset has a dividend yield of q, the relationship is
. (5)
Using (5) and the assumption that for a share with maximum price the put option has
no value, the value of the call option in the upper point of the grid will be given by
(6)
for . In next section we will show the example of the valuation of the call
option using the Explicit finite difference method with this improvement of the assumption
for the values of it in the upper endpoints of the grid. (Švábová, Ďurica, 2013), (Švábová,
2011).
We will compare the numerical estimates of the call option prices with given parameters
obtained from the Explicit finite difference method with the values of the call option given by
(2). We will show the differences of the results between the real values of the option given by
(2) with the values estimated by Finite difference method using (4). Especially we will
compare the differences between these two values using the improvement suggested
hereabove with the differences between these two prices without the improvement.

3. Results and Discussion

The influence of the improvement suggested hereabove to the results of the numerical
estimates of the option prices in comparison with prices calculated by BSM model (2) is
clearly visible in the example of Exchange option.
Option to Exchange One Asset for Another (or simply Exchange Option) is an exotic
option that gives the holder the right to exchange an asset "B" for an asset "A" at given time in
the future. Exchange option was first described by W. Margrabe in 1978. (Margrabe, 1978)
The holder of this option is therefore entitled to change the asset A worth S1, with the
volatility σ1 and dividend yield q1 for the asset B, worth S2, volatility σ2, and dividend yield q2.
3.1 Valuing the Exchange option using the analytical formula
A formula for valuing European option to give up an asset A worth S1 and receive in return
an asset B worth S2 was first produced by Magrabe in 1978. (Margrabe, 1978) Analytic
formula for the price of American type of Exchange option, which can be exercised at any
time up to the expiration date of the option, was presented in (Gounden, O´Hara, 2009). The
value of an Exchange option at the time zero is given by
, (7)
where

. (8)
and
(9)
is the volatility of . (Hull, 2012)
Let the initial value of the asset A be , the initial value of the underlying asset B be
. Let the volatilities be , the risk - free interest rate and the
time to maturity be year. The asset A brings the dividend yield and the asset B
7th International Scientific Conference Managing and Modelling of Financial Risks Ostrava
VŠB-TU of Ostrava, Faculty of Economics, Finance Department 8th – 9th September 2014

brings the dividend yield during the lifetime of the Exchange option. The correlation
between the assets is .
The total volatility will be calculated by (9). Substituting this values to the
analytical formula for the price of Exchange option (7) with (8) we get the value
.
3.2 Valuing the Exchange option using the Explicit Finite difference method
Again, we can use the example from above with values , , ,
, , , and . For the application of Explicit Finite
Difference Method we create the modified value of the underlying stock price S and strike
price X of classic European option with the same parameters. Using these modified values we
could value the Exchange option just like a classic European option. The modified values of
stock price and strike price is given by
, . (10)
This modification was suggested in (Švábová, 2013). Thus the price of the Exchange
Option will be estimated by Finite Difference Method using the same procedure, as reported
in (Švábová, 2011) for classic European option with continuous dividend yield with
underlying stock price and the strike price . The modified
dividend yield is .
The value of the Exchange option in the endpoints of the grid is calculated in the same way
as for European option with the underlying asset with continuous dividend yield:
 In the points on the right side of the grid the Exchange option prices are given by
the payoff function. It is the difference which for the Exchange option
represents the difference between the values of assets A and B, bearing with interest
rate :
. (11)
 In the lower points for zero share prices the corresponding Exchange option has
zero prices too.
 In the upper points for maximum share prices the values of Exchange option are
given by (6) implies from put – call parity with the modified values (10).
The last item represents the different approach from (Hull, 2012), where the upper points
values of the option are equal to exercise price X.
We calculate the option price for the modified value which corresponds to
Exchange option with given parameters. To obtain the numerical estimate of the Exchange
option price we will use the grid of 100 x 500 dividing points. The stock price interval is
divided into 100 subintervals and the time interval is divided into 500 subintervals. This
density of the grid has proven to be the most appropriate to obtain the best estimates of the
Exchange option price.
In Table 1 the values of Exchange option price estimated by Finite difference method are
compared with the real value calculated from analytical formula (7). The results and their
difference from the real value with and without the improvement of option price values in
upper points of the grid suggested hereabove are shown.

Table 1 The values of Exchange option price with and without the improvement
Without the With the
improvement improvement
Real value 45,42465133 45,42465133
Estimated value 45,42953812 45,42952520
Difference -0,00488679 -0,00487386
7th International Scientific Conference Managing and Modelling of Financial Risks Ostrava
VŠB-TU of Ostrava, Faculty of Economics, Finance Department 8th – 9th September 2014

Real value listed if Table 1 is calculated form (7). The estimated value is obtained from
Explicit Finite difference method using (4) and the procedure described hereabove. The
difference is between the real value and the estimated value.
As shown in Table 1, using the improvement suggested in this paper leads to better
estimates, which are closer to real values of the option prices. The estimated values are
smaller than the real values, so that the Explicit finite difference method use to underestimate.
But as shown in Table 1, the estimates are more accurate with the improvement of the option
price values in the upper points of the grid. The difference is not a very big number, but in fact
moneys of investors are in a play. And more, in case of higher volumes of option trades the
difference of their prices estimated increases.

4. Conclusions
In this paper we described the improvement of estimating the option price using the
Explicit Finite difference method. This numerical method is very appropriate and useful for
calculating the prices of the options. Especially in such cases of exotic option, which are so
complex, that the analytical formula for their valuation does not exist. Then come into play
some numerical method for their valuation and the Explicit finite difference method is one of
them. The aim of the method is that the whole domain of the solution of Black – Scholes –
Merton partial differential equation is reduced only into a finite number of grid points. The
values of the option in every interior point of the grid are dependent on the values in
neighboring points, where the endpoints values are known. In this paper we suggested the
improvement of the values of known upper points which leads to more accurate estimates of
the option prices. The situation was described using the example of Exchange option. But the
procedure is suitable also for other exotic options whose prices we need to estimate.

This paper Acknowledgments

was designed in response to the project co-funded by the EU called "The quality of
education and human resources development as pillars of a knowledge society at the Faculty
PEDAS, ITMS 26110230083”.

References
[1] Gounden, S. and O´Hara, J.G. (2009). An Analytic Formula for the Price of
American Exchange Options. [online]. Available at:
<http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1403746> [Accessed
10.5.2013].
[2] Hardik, D. (2008). Numerical Method for Pricing Exotic Options. MSc. Thesis.
Imperial College London. Available at:
<http://www3.imperial.ac.uk/pls/portallive/docs/1/55071696.PDF> [Accessed
15.5.2014].
[3] Hull, J. C. (2012). Options, Futures, and Other Derivatives, 8th ed. New York:
Pearson Prentice Hall.
[4] Margrabe, W. (1978). The Value of an Option to Exchange One Asset to
Another. Journal of Finance, 33(1), pp. 177–186.
 7th International Scientific Conference Managing and Modelling of Financial Risks Ostrava
VŠB-TU of Ostrava, Faculty of Economics, Finance Department 8th – 9th September 2014

[5] Švábová, L. (2011). Using of the Finite Difference Method for European Call
Option Pricing. In: Proceedings of the 8th International Scientific Conference
Actuarial Science in theory and in practice 2011, Bratislava, Slovakia, pp. 115-122.
[6] Švábová, L. (2013). Exchange option pricing using the finite difference method
In: Proceedings of the 6th international scientific conference Finance and the
performance of firms in science, education, and practice 2013, Ostrava, Czech
Republic, pp. 782 – 789.
[7] Švábová, L., Ďurica, M. (2013). Using the finite difference method for Chooser
option pricing. In Proceedings of 9th international scientific conference Financial
management of firms and financial institutions 2013. Ostrava, Czech Republic,pp.
943-950.
[8] Vlachynský, K., Markovič, P. (2001). Finančné inžinierstvo. Bratislava: IURA
EDITION.
[9] Whaley, R. E. (2006). Derivatives: Markets, Valuation, and Risk Management.
New Yersey: John Wiley & Sons.

View publication stats