Stochastic Optimization: Algorithms and Applications (S. Uryasev and P. M.

Pardalos, Editors),
pp. 1-2
c 2000 Kluwer Academic Publishers

Option pricing in a world with arbitrage

Xin Guo (xinguo@us.ibm.com)

IBM, T.J. Watson Research Center
Yorktown, NY 10598, USA

Larry Shepp (shepp@stat.rutgers.edu)

Rutgers, Statistics
Piscataway, NJ 08854, USA

Abstract

We discuss option pricing problems under a new model of stock fluctuations. This
model captures the information distribution among investors by adjoining a hidden
Markov process to the Black-Scholes exponential Brownian motion model. We pro-
vide new valuations for various standard hedge options, such as European, perpetual
American and look-back options.
Keywords: Information, Option pricing, Arbitrage, Black-Scholes

1 Introduction
The Black-Scholes model [Black and Scholes (73)] is widely accepted as the cornerstone
of option pricing theory. This model is popular and successful, yet simple. Unfortunately,
it is far from reality because it assumes that information plays no role in the market. Thus
the use of the assumption that “arbitrage” can never exist is unrealistic and too rigid. We
give a model for option pricing that is as simple as the Black-Scholes model and is closer
to reality.
The new model focuses on the major factor that distinguishes the stock market from
other natural phenomena—human activity. The stock market cannot exist independently
of investors. Yet standard models focus on modeling the stock price and do not provide
a simple way of modeling human activity. Human activity is the very source of stock

1
fluctuations, hence the chance of arbitrage, if it ever exists. Moreover, the opportunity for
arbitrage is the catalyst of human activities. Arbitrage will never exist for long; it appears,
but is removed immediately by active tradings. Existing models do not exploit the latent
connection between human activity and arbitrage.
In [Guo (99b)], it is postulated that information is the factor that quantifies human ac-
tivity. Information is rarely shared by everyone and even if, simultaneously. This time
difference gives rise to an arbitrage opportunity, which will disappear quickly once every-
one gets the “information”.
In this context, we like to note that lot of work has been done to capture the features that
are present in reality but missing in the Black-Scholes model. Several papers have studied
financial markets with different information levels among investors [Duffie and Huang (86),
Ross (89), Anderson (96), Grorud and Pontier (98), Karatzas and Pikovsky (96)]. In par-
ticular, Kyle [Kyle (85)] considered a dynamic model of insider trading with sequential
auctions, in which the informational content of prices and the values of private informa-
tion to an insider are examined. Option pricing and related problems are not addressed,
however.
We now proceed to describe the new model. The existence of increased or non-uniform
information at certain times is incorporated by modeling the fluctuations of a single stock
price Xt with an equation of the form

dXt = Xt µε(t ) dt + Xt σε(t ) dWt (1)

where ε(t ) is an additional stochastic process, independent of Wt , which represents the state
of information in the investor community.
We assume that ε = ε(t ) is a Markov process which moves among a few states and
for each state i, there is a known drift parameter µi and a known volatility parameter σi .
(µε(t ) ; σε(t ) ) take different values when ε(t ) is in different states. For example, ε(t ) = 0 at
those times t at which the price change is not abnormal and people believe that they are all
well-informed in a seemingly complete market. ε(t ) = 1 when there are wild fluctuations in
the stock price and people suspect that some individuals or groups have extra information
which is not circulating among the mass of investors and thus would possibly bring wilder
fluctuations depending on the reaction of the investors. µ1 may be larger or smaller than µ0
depending on the nature of inside information, therefore this state may divide into two extra
states where informed investors believe the company will prosper or decline. Furthermore,
some inside groups may actually be misled and the model could include a state which
would indicate that there is a group of investors who erroneously believe that the company’s
fortunes are going to change for the positive, and another state for the negative.
If the σ’s are distinct then it is no loss of generality to assume that ε(t ) is actually
observable, since the local quadratic variation of Xt in any small interval to the left of t
will yield σε(t ) exactly ([McKean (69)]). Nevertheless, it is conceivable that sometimes
insiders will try to manipulate their buying and selling in such a way that the existence of
such information is not detectable from the change of volatilities, namely σ’s are identical.
It is an open mathematical problem to detect the state change of ε(t ) when σ’s remain

2
unchanged.
More generally, one can use the state space S = f0; 1; 2; : : : ; N g for ε(t ) to model more
complex information structures. In such cases, it is harder to obtain closed-form solution
for option pricing problems.
It is worth pointing out that this model is different from being merely a model with
stochastic volatility. This is because the drift µε(t ) is also driven by the hidden Markov
process ε(t ) to reflect the simple fact that investors’ expectation varies with the arrival of
informations. Indeed, this very difference exemplifies exactly what is usually missing in
the standard stochastic volatility model and as we shall see in the following discussions,
for certain choices of µε(t ) ; σε(t ) , there is an “arbitrage ” in option pricings. The model also
differs from the model with “uncertain volatility” [Avellaneda, et. al. (95)] which does not
connect information structure with stock fluctuations in an explicit way and is unable to
provide explicit solutions to option pricings.
We will concentrate on the two-state case in which ε(t ) alternates between 0 and 1 such
that
(
0; when the market seems complete, and
ε(t ) = (2)
1; when some people have or believe they have inside information;

where σ0 6= σ1 .
More concretely, let λi denote the rate of leaving state i, τi the time to leave state i, so
that the exponential law holds

λit
P(τi > t ) = e ; i = 0; 1: (3)

Remark 1. This formulation is drawn from a basic yet key observation that “inside infor-
mation” plays a role in changes of volatilities. If we imagine that the flow of each piece of
information is a random process Xi , and X1 ; X2; : : : ; Xn being i.i.d process, then it is known
that their super-imposed process is Poisson. The choice of this memoryless process has
more justifications than its pure mathematical one. Indeed, it would be easily identified in
practice if the flow of information is without the memoryless property.

Remark 2. A main characteristic of this model is that X (t ) is not Markovian, but (X (t ); ε(t ))
is jointly Markovian. By formulating stock fluctuations in the form of Eq. (1), it provides a
simple way to connect historical data and current situation. To consider (X (t ); ε(t )) jointly
reflects the idea that market never exists independently of the investors’ activity. It is worth
pointing out that “inside information” is a convenient way to describe the hidden Markov
process ε(t ), which can be interpreted in a broader way to reflect the “noise” that goes
beyond the Black-Scholes and other standard models.

3
2 Option pricings under the new model
It is easy to see that the model is not “complete” (according to [Harrison and Kreps (79),
Harrison and Pliska (81)]) because of the additional information structure provided by ε(t ).
In other words, ε(t ) is a bounded adapted process with respect to the σ-algebra Ft generated
by Xt (denoted as F X ), but is not adapted to the σ-algebra generated by Wt (written as F W ).

Completing the market—New securities COS. One way to complete the market is,
suggested by Darrell Duffie, as follows: at each time t, there is a market for a security
(COS) that pays one unit of account (say, dollar) at the next time τ(t ) = inffu > t g when ε(t )
changes state. That (COS) contract then becomes worthless (i.e., has no future dividends),
and a new contract is issued that pays at the next change of state, and so on.
It was shown [Harrison and Kreps (79), Harrison and Pliska (81)] that the absence of
arbitrage is effectively the same as the existence of a probability measure Q, equivalent to
P, under which the price of any derivative is the expected discounted value of its future
cash flow.
Given such a measure Q, price process S must have the form
dS(t ) = (r δε(t ) )S(t ) dt + St σε(t ) dBQ ; (4)

where BQ is a standard Brownian motion under Q.

Now the usual techniques from [Harrison and Kreps (79), Harrison and Pliska (81)] can
be applied to get complete market and unique pricing for any derivatives with appropriate
square-integrable cash flows.

European options. As a consequence, the unique arbitrage price at time 0 of a European

call option is obtained in [Guo (99b)]:
Theorem 1 Given Eq(1), COS contract, and a riskless interest rate r, the arbitrage free
price of a European call option with expiration date T and strike price K is
Vi (T ; K ; r) = EQ [e rT
(XT K )+ jε(0) = i] (5)
Z ∞Z T
rT
= e yρ(ln(y + K ); m(t ); v(t )) fi(t ; T )dtdy; (6)
0 0

where ρ(x; m(t ); v(t )) is the normal density function with expectation m(t ) and variance
v(t ), and
λ1 T T t 1=2
f0 (t ; T ) = e e(λ1 λ0 )t
t ] J 1 [2( λ0λ1 T t + λ0 λ1 T 2 )1=2 ]
([
λ0 λ1
+ λ0 J0 [2( λ0 λ1 T t + λ0 λ1 T 2 )1=2 ]); (7)
T t 1=2
f1 (t ; T ) = e λ0 T e(λ0 λ1 )t ([ t ] J 1 [2( λ0λ1 T t + λ0 λ1 T 2 )1=2 ]
λ0 λ1
+ λ1 J0 [2( λ0 λ1 T t + λ0 λ1 T 2 )1=2 ]); (8)

4
 m(t ) = (d1 1=2(σ20 σ21 ))t + (r d1 1=2σ21 )T ; (9)
(σ0 σ21 )t + σ21 T ;
2
v(t ) = (10)

where Ja (z) is the Bessel function such that

∞
1)n (z=2)2n
(
Ja (z) = (1=2z)
a
∑ ; (11)
n=0 n!Γ(a + n + 1)
Ya (z) = cot(πa)Ja(z) csc(πa)J a (z): (12)

In particular, when µ0 = µ1 ; σ1 = σ0 , we have fi (t ; T ) = δt ;T , and therefore above the equa-

tions reduce to the classical Black-Scholes formula for European options.

Optimal stopping time and option pricings. In [Guo (99c)], more option pricing formu-
lae were derived for perpetual American put options, Russian options (perpetual look-back
options). We illustrate here with the example of Russian options.
The Russian option was coined by [Shepp and Shiryaev (93)]. It is a perpetual look-
back option. The owner of the option can choose any exercise date, represented by the
stopping time τ (0  τ  ∞) and gets a payoff of either s (a fixed constant) or the maximum
stock price achieved up to the exercise date, whichever is larger, discounted by e rτ , where
r is a fixed number.
To price Russian options is to consider the following optimal stopping time problem:
Let X = fXt ; t  0g be the price process for a stock with X0 = x > 0, and

St = max fs ; sup Xu g; (13)

0ut

where s > x is a given constant, what is the value of V ,

rτ
V = sup Ex;s e Sτ ? (14)
τ

where τ is a stopping time with respect to the filtration FXt = fX (s); s  t g, meaning no
clairvoyance is allowed.
This is an optimal stopping problem with an infinite time horizon and with state space
f(ε; x; s)jx  sg. The key is to find the so called “free boundary ” x = f (s; µ; σ; ε) such
that if x  f (s; µ0; µ0 ; σ0 ; σ1 ; λ0 ; λ1 ) we should stop immediately and exercise the option,
while if s  x  f (s; µ0; µ0 ; σ0 ; σ1 ; λ0 ; λ1 ), we should keep observing the underlying stock
fluctuations. By extending the technique of the “Principle of Smooth Fit”, which should
be attributed to A. N. Kolmogorov and H. Chernoff, to allow discontinuous jumps, an
explicit closed-form solution was derived in [Guo (99c)]. It showed that if we define Ωi =
f(x; s)jcis  x  sg, then the optimal stopping time τi will be the first time to leave Ωi at
state i, i.e.,
τi = infft  0j = ci ; ε(t ) = ig;
Xt
St

5
Namely, when the hidden Markov process ε(t ) switches from one state to another, there is
a discontinuous jump over the boundary, which is also called the “regime switching”. The
proof is via martingale theory.
In [Guo (99c)], the Markov property of (Xt =St ; ε(t )) is intensively exploited. It was also
proved that for a N-state case (N > 2), an algebraic equation of order 2N must be solved.
The methodology of extending the “Principle of Smooth Fit” also applies to the pricing
of perpetual American put options, in which the buyer pays a certain amount of money V
(to be decided) and in turn receives the right to obtain at any time t of his choice a share of
the stock at an fixed price K. His profit is then (Xt K )+ . To determine V is to solve the
corresponding optimal stopping problem:

V (x) = sup Ex [(Xτ K )+ ] :

τ

Using the same technique developed in [Guo (99c)], it is not hard to see that the free bound-
aries are two points, from which the closed-form solution is derived.

Arbitrage in the new model: Comparing Eq. (4) with a geometric Brownian motion
process with drift r and variance σ, δε(t ) is of special interest to us. It provides a way to
“measure” the flow of the inside information. The drift differs from the riskless interest
rate r by µ0 µ1 when there is an inside information and hence the arbitrage opportunity
emerges. It also suggests the difference between the case of “pure noise” (i.e., σ0 6= σ1 ; δ0 =
δ1 = 0) and the case when there may exist an inside information (i.e., σ0 6= σ1 ; δ1 6= 0).
When the state change in ε(t ) is invisible to outsiders, i.e., when σ0 = σ1 , it will give
insiders a chance of taking the profit that is equal to the price of the COS contract. More-
over, when the state change is visible yet with the δε(t ) less than the level of transaction
cost, then the arbitrage opportunity is not worth taking. This is not uncommon.
There are also cases in which the emergence of “inside information” will have a certain
delay time, such that σε(t ) and µε (t ) will be generated by different processes ε(t ) and ε0 (t ).
0

This is another interesting problem that is yet to be tackled.

References
[Anderson (96)] Anderson, T. G, 1996, Return volatility and trading volume: an information flow
interpretation of stochastic volatility Journal of Finance 51, 169–204.

[Avellaneda, et. al. (95)] Avellaneda, M., Levy, P., and Parás, A., 1995, Pricing and hedging deriva-
tive securities in markets with uncertain volatilities Applied Mathematical Finance 2, 73–88.

[Back (92)] Back, K., 1992, Insider trading in continuous time, Review of Financial Studies 5,
387–409.

[Bachelier (00)] Bachelier, L., 1900, Theorie de la Speculation Annales Scientifiques de L’École
Normale Supérieure, 3d ser., 17, 21–88.

6
[Ball (93)] Ball, C. A., 1993, A review of stochastic volatility models with applications to option
pricing Financial Markets, Institutions and Instruments 2, 55–69.

[Black and Scholes (73)] Black, F. and Scholes, M., 1973, The pricing of options and corporate
liabilities Journal of Political Economy 81, 637–654.

[Chernoff (61)] H. Chernoff 1961, Sequential tests for the mean of a normal distribution. Proc. 4th.
Berk. Symp. Stat. Prob. 1: 79–91.

[Cox, et. al. (79)] Cox, J., Ross, S., and Rubinstein, M., 1979, Option pricing, a simplified approach
Journal of Financial Economics 7, 229–263.

[Delbaen and Schachermayer (94)] Delbaen, F. and Schachermayer, W., 1994, A general version
of the fundamental theorem of asset pricing Mathematique Annales 463–520.

[Duffie and Harrison (93)] D. Duffie and M. Harrison 1993, Arbitrage Pricing of Perpetual Look-
back Options Annals of Applied Probability 3: 641-651.

[Duffie and Huang (86)] Duffie, D., and Huang, C. F., 1986, Multiperiod security markets with
differential information Journal of Mathematical Economics 15, 283–303.

[Grorud and Pontier (98)] Grorud, A., and Pontier, M., 1998, Insider trading in a continuous time
market model International Journal of Theoretical and Applied Finance 1, 331–347.

[Guo (99a)] Guo, X., 1999, Insider Information and Incomplete Market, Ph. D. dissertation, Math-
ematics Department, Rutgers University.

[Guo (99b)] Guo, X. 1999, Information and option pricings. (Submitted).

[Guo (99c)] Guo, X. 1999, An explicit solution to an optimal stopping problem with “regime
switching”. (Submitted).

[Harrison and Kreps (79)] Harrison, M., and Kreps, D., 1979, Martingales and arbitrage in multi-
period securities markets Journal of Economics Theory 20, 381–408.

[Harrison and Pliska (81)] Harrison, M. and Pliska, S., 1981, Martingales and stochastic integrals
in the theory of continuous trading Stochastic Processes and Their Applications 11, 215–260.

[Hull and White (87)] Hull, J. and White, A., 1987, The pricing of options on assets with stochastic
volatility Journal of Finance 2, 281–300.

[Jacka (91)] S. D. Jacka (1991): Optimal Stopping and the American put Mathematical Finance
Vol. 1, No. 2: 1-14.

[Jacka (93)] S. D. Jacka 1993, Local Times, Optimal Stopping and Semimartingales Annals of
Probability Vol. 21, No. 1: 329-339.

[Karatzas and Pikovsky (96)] Karatzas, I., and Pikovsky, I., 1996, Anticipative portfolio optimiza-
tion Advances in Applied Probability 28, 1095–1122.

[Kyle (85)] Kyle, A., 1985, Continuous auctions and insider trading Econometrica 53, 1315–1335.

7
[Lakner (98)] Lakner, P., 1998, Optimal trading strategy for an investor, the case of partial infor-
mation Stochastic Processes and Their Applications 76, 77–97.

[Di Masi, et. al. (94)] Di Masi, G. B., Yu, M., Kabanov, and Runggaldier, W. J., 1994, Mean-
variance hedging of options on stocks with Markov volatility Theory of Probability and Its
Applications 39, 173–181.

[McKean (69)] McKean, H. P., 1969, Stochastic Integrals Academic Press.

[Merton (73)] Merton, R., 1973, The theory of rational option pricing Bell Journal of Economics
and Management Science 4, 141–183.

[Pagès (87)] Pagès, H., 1987, Optimal consumption and portfolio when markets are incomplete,
Ph. D. dissertation, Department of Economics, Massachusetts Institute of Technology.

[Robbins, et. al. (71)] H. Robbins, D. Sigmund and Y. Chow 1971, Great Expectations: The Theory
of Optimal Stopping Houghtoon-Nifflin, Boston.

[Ross (89)] Ross, S. A., 1989, Information and volatility, the no-arbitrage martingale approach to
timing and resolution irrelevancy Journal of Finance 44, 1–8.

[Samuelson (73)] Samuelson, P., 1973, Mathematics of speculative price (with an appendix on
continuous-time speculative processes by Merton, R. C.) SIAM Review 15, 1–42.

[Shepp and Shiryaev (93)] L. Shepp and A. N. Shiryaev 1993, The Russian option: Reduced Regret
Annals of Applied Probability 3: 631-640.