QUANTI TATI VE FI NANCE VOLUME 2 (2002) 31–44 RESEARCH PAPER

I NSTI TUTE O F PHYSI CS PUBLI SHI NG quant.iop.org
Deterministic implied volatility
models
P Balland
Merrill Lynch Financial Centre, 2 King Edward Street, London EC1, UK
E-mail: philippe balland@ml.com
Received 20 September 2001
Published 4 February 2002
Online at stacks.iop.org/Quant/2/31
Abstract
In this paper, we characterize two deterministic implied volatility models,
defined by assuming that either the per-delta or the per-strike implied
volatility surface has a deterministic evolution. Practitioners have recently
proposed these two models to describe two regimes of implied
volatility (see Derman (1999 Risk 4 55–9)). In an arbitrage-free sticky-delta
model, we show that the underlying asset price is the exponential of a process
with independent increments under the unique risk neutral measure and that
any square-integrable claim can be replicated up to a vanishing risk by
trading portfolios of vanilla options. This latter result is similar in nature to
the quasi-completeness result obtained by Bjork et al (1997 Finance
Stochastics 1 141–74) for interest rate models driven by Levy processes.
Finally, we show that the only arbitrage-free sticky-strike model is the
standard Black–Scholes model.
1. Introduction
In classic extensions of the Black–Scholes (1973) model that
accounts for the smile effect, the underlying asset price S is
driven by one or two Brownian motions. These models are
referredas local volatilitymodels as theydiffer fromthe Black–
Scholes model by simply allowing the local volatility σ
t
of the
underlying asset price to be stochastic:
dS
t
/S
t
= µ
t
dt + σ
t
dW
t
.
Two types of local volatility models have been proposed: the
so-called deterministic and stochastic local volatility models.
In the deterministic local volatility models, the local volatility
satisfies σ
t
= σ(t, S
t
) as in Dupire (1994) and Derman
and Kani (1994). Deterministic local volatility models
are complete and complex options are replicated using the
underlying S alone! In the stochastic local volatility models,
ν
t
= σ
2
t
is a stochastic process characterized by its mean-
reversion rate, its volatility and its correlation with S
t
as in
Hull and White (1987) and Heston (1993):
dv
t
= (α
t
−κv
t
) dt + ξ

v
t
dB
t
(Heston)
dv
t
/v
t
= α
t
dt + ξ dB
t
(Hull–White)
The processes B and W are Brownian motions with correlation
ρ and are defined under the risk-neutral measure. This risk-
neutral measure is unique and the stochastic local volatility
models are complete if call options are traded instruments.
The parameters α
t
, κ, ξ and ρ are implied by calibration to
the initial smile surface. When inferred from historical data,
these parameters are typically much lower. The reason for
this can be understood if we observe that in stochastic models,
d ln S
t
has a normal conditional distributionwithvariance σ
2
t
dt
as in the Black–Scholes model. Consequently, the calibration
of a stochastic local volatility model to a short-dated smile
curve, which typically has large variations near the at-the-
money strike, results in unrealistically large correlation ρ and
volatility ξ. To compensate for these large parameters, the
calibration to a long-dated smile curve, which is typically
flatter, implies a large mean-reversion rate κ. Deterministic
local volatility models have similar problems despite accurate
calibrations. The local volatility function inferred from the
initial smile surface, has typically ‘unrealistically’ large slope
and convexity for small maturity while it is almost flat for large
maturity.
1469-7688/02/010031+14$30.00 © 2002 IOP Publishing Ltd PII: S1469-7688(02)32504-1 31
P Balland QUANTI TATI VE FI NANCE
It is important to understand that a complex option will
be hedged efficiently using a smile model only if the model
implies a dynamic of the smile that is sufficiently ‘realistic’ and
‘stationary’ for the model to not require frequent re-calibration.
Local volatility models are, in this sense, unsatisfactory. The
dynamic of the implied volatility surface is entirely specified
once the model has been calibrated with the initial smile
surface, leaving thus no control on this dynamic.
To illustrate further this point, consider the foreign-
exchange market, where implied volatilities are quoted per
delta. The underlying S is typically traded more frequently
than vanilla options and thus, the implied volatilities do
not change as frequently as S does. In a deterministic
local volatility model such as the Dupire model, the implied
volatilities are however functions of S. Hence, over a short
time interval during which implied volatilities do not change,
the model will still need re-calibration every time S changes!
These frequent re-calibrations cause the hedging strategies
implied by the model to be ‘inefficient’ as reported by Dumas
et al (1998). In a stochastic local volatility model, implied
volatilities depend on t , the local volatility σ
t
and on the
percentage-in-the-money S
t
/K. This last dependence implies
a greater stability of the corresponding hedging strategies.
However, to be consistent with market smiles, such a model
over-estimates, as previously explained, the volatility of
volatility and the correlation of volatility with S and, as a
consequence, the cost of hedging is not accurately accounted
for.
The need for smile models that are calibrated to an
initial smile surface and that imply a ‘realistic’ evolution of
implied volatility explains the recent interest in the so-called
implied volatility models. These models are defined by direct
assumptions on the stochastic evolution of the smile surface
from an initial surface, as in Schonbucher (1999) and in Cont
and da Fonseca (2001). It is important to stress, however, that
these models need severe restrictions to be arbitrage-free.
In this paper, we characterize two recently proposed
deterministic implied volatility models, defined by assuming
that either the per-delta or the per-strike implied volatility
surface has a deterministic evolution. In arbitrage-free sticky-
delta models, we show that the underlying asset price is the
exponential of a process with independent increments under
the risk neutral measure and that any square-integrable claim
can be replicated up to a vanishing risk by trading portfolios of
vanilla options. Finally, we show that the only arbitrage-free
sticky-strike model is the Black–Scholes model.
2. The option market and the implied
volatility models
We consider a continuous trading economy on a finite horizon
[0, T ] with traded assets at time t , a financial asset S, the money
market account B, the call and the put options on S with strike
K > 0 and maturity in (t, t + x
m
). We also assume that static
portfolios consisting of a continuum of traded vanilla options
are traded instruments as in Breeden and Litzenberger (1978)
and Carr and Madan (1998).
We assume that only short-dated vanilla options are traded
assets because in typical option markets, only short-dated
options have liquid enough prices to be regarded as traded
instruments. In typical foreign exchange markets for example,
only the vanilla options with maturity less than a couple of
years have liquid prices.
We denote by S
t
> 0, C
t
(x, K) > 0 and P
t
(x, K) > 0
the price at time t of one unit of S, of one call and of one put
option on S with maturity t + x ∈ (t, t + x
m
) and with strike
K > 0.
We assume no transaction costs. We assume deterministic
interest rates and we denote by P
t x
the price at time t of a
discount bond with maturity t + x and by B
t
≡ 1/P
0t
the
money-market account. We denote by F
t x
the forward price at
time t of receiving one unit of S at time t + x t .
We assume that there exists a real bounded function µ(t )
such that
F
t x
/S
t
= exp
__
t +x
t
µ(s) ds
_
≡ m
t x
.
We assume that there exists a stochastic basis (, ` =
`
T
, {`
t
: t ∈ [0, T ]], P) with a right-continuous, complete,
increasing filtration with respect to which S
t
and 1/S
t
are
cadlag quasi-left continuous square-integrable processes and
sup{E[S
2
t
+ S
−2
t
] : 0 t T ] < ∞,
C
t
(x, K) = P
t x
E
t
[(S
t +x
−K)
+
],
P
t
(x, K) = P
t x
E
t
[(K −S
t +x
)
+
].
We assume that the map (x, K) .→ C
t
(x, K) from (0, x
m
)
[0, +∞) into (0, +∞) is of class C
1
(resp. C
2
) in the first
(resp. second) variable for all t ∈ (0, T ). The market filtration
{`
t
: t ∈ [0, T ]] is the filtration generated by all primary traded
asset prices.
Definition 2.1. A family {C
t
(x, K), P
t
(x, K), S
t
: (x, K)
∈ (0, x
m
) (0, +∞)} satisfying the above assumptions is
called an implied volatility model.
We use the terminology of implied volatility models
instead of option price models because implied volatilities
rather than option prices are the financial observable (see
section 1). It is clear, however, that a process S
t
and a
two-parameter family of implied volatility processes define
an implied volatility model providing that this specification is
arbitrage-free. Our definition of implied volatility models is
based on option prices to ensure our implied volatility models
are arbitrage-free.
We shall need the following technical restriction to ensure
that S has bounded local characteristics. This restriction is
fairly mild and met by most local volatility models.
Definition 2.2. An implied volatility model is regular if
(i) lim
x→0
E[f (S
t +x
)[S
t
= S] = f (S), (f ∈ C
0
),
(ii) sup
x

x
E[(S
t +x
/F
t x
)
2
[S
t
] is locally bounded,
(iii) sup
t
E[sup
x
[∂
x
E
t
[(S
t +x
/F
t x
)
±2
][] < ∞.
We write C
0
for the space of continuous functions
vanishing at zero and infinity. We refer to Derman and Kani
(1998), Schonbucher (1999) and Cont and da Fonseca (2001)
for examples of stochastic implied volatility models. We now
restrict our attention to two deterministic implied volatility
models.
32
QUANTI TATI VE FI NANCE Deterministic implied volatility models
3. The sticky-delta and the sticky-strike
implied volatility models
The Black–Scholes function is denoted C
BS
(V, F, K) with
C
BS
(V, F, K) = F N(d) −KN(d −

V),
d = ln(F/K)/

V +
1
2

V.
We denote

BS
(V, F, K) = ∂
F
C
BS
(V, F, K),

BS
(V, F, K) = ∂
2
F
C
BS
(V, F, K).
We recall that

V
C
BS
(V, F, K) =
1
2
F
2

BS
(V, F, K).
Definition 3.1. Let (x, I) ∈ (0, x
m
) (0, +∞). The
per-delta implied volatility denoted σ
t
(x, I) where I stands
for the forward moneyness ratio, is the unique positive
solution of the equation
P
t x
C
BS
(xσ
t
(x, I)
2
, F
t x
, F
t x
I) = C
t
(x, F
t x
I).
The per-strike implied volatility denoted
t
(x, K) is the
unique positive solution of the equation
P
t x
C
BS
(x
t
(x, K)
2
, F
t x
, K) = C
t
(x, K).
Since the function C
BS
(V, F, K) is strictly increasing
with respect to V, we conclude that

t
(x, K) = σ
t
(x, K/F
t x
),
lim
x↓0

t
(x, I)
2
= lim
x↓0
x
t
(x, K)
2
= 0.
At a given time t , the volatility functions (x, I) .→
σ
t
(x, I) and (x, K) .→
t
(x, K) are of class C
1
in the
first variable and C
2
in the second variable in (0, x
m
)
(0, +∞). For (x, I) ∈ (0, x
m
) (0, +∞), the implied
volatility processes σ
t
(x, I) and
t
(x, I) are cadlag quasi-left
continuous processes adapted to the market filtration.
Following Derman (1999) and Reiner (1999), we propose
the following definition of sticky-delta and sticky-strike
implied volatility models.
Definition 3.2. An implied volatility model is sticky-delta if
the per-delta volatility process σ
t
(x, I) is deterministic on
[0, T ] for all (x, I) in (0, x
m
) (0, +∞).
An implied volatility model is sticky-strike if the per-strike
volatility process
t
(x, K) is deterministic on [0, T ] for all
(x, K) in (0, x
m
) (0, +∞).
We observe that the Black–Scholes model is an implied
volatility model defined by σ
t
(x, I) =
t
(x, K) = σ
t
(x, 1).
4. Characterization of the sticky-delta
implied volatility models
For a sticky-delta implied volatility model, we define the
following deterministic function:
c(t, x, I) = C
BS
(xσ
t
(x, I)
2
, 1, I). (4.1)
Since y
2
< 2(e
y
+ e
−y
) for all real y and S
t
, 1/S
t
are
square-integrable, ln S
t
is square-integrable. Finally, we have
the following result.
Theorem 4.1. In a sticky-delta implied volatility model
defined on [0, T ], the market filtration {`
t
: 0 t T ]
coincides with the filtration generated by S, all risk-neutral
probability measures are equal to the probability measure P
on the σ-algebra `
T
and the stochastic process ln S
t
has
independent increments under P.
Proof. We note that P is a risk-neutral measure. Consider a
forward-start call option with fixing date t and maturity t + x
with x < x
m
. This option pays at maturity the quantity
(S
t +x
−kF
t x
)
+
/F
t x
.
At time t , this forward-start option is a regular call option
on S with strike kF
t x
, time-to-maturity x and notional 1/F
t x
.
At time t , the cost of replicating this option is P
t x
c(t, x, k).
This value is deterministic. Hence at time s t , the forward
value of this option is c(t, x, k) and thus independent of s.
Finally, we have proved that for any risk-neutral measure Q
and any s t , we have
E
Q
s
[(S
t +x
−kF
t x
)
+
/F
t x
] = c(t, x, k). (4.2)
Next, we note that
0 −
1
F
t x

k
(S
t +x
−kF
t x
)
+
1. (4.3)
We can thus permute differentiation with respect to k and
Q-expectation in (4.2):
Pr
Q
{S
t +x
< kF
t x
[ `
s
] = 1 + ∂
k
c(t, x, k). (4.4)
The cumulative distribution of ln S
t +x
−ln S
t
conditional
on `
s
is thus independent of s t , S
s
and Q.
Let u
1
= ln S
r
− ln S
s
and u
2
= ln S
u
− ln S
t
with
s < r < t < u < T , 0 < r − u < x
m
and u − t < x
m
.
Equation (4.4) implies that for any risk-neutral measure Q, we
have
E
Q
[e
iαu
1
+iβu
2
] = E
Q
[e
iαu
1
] E
Q
[e
iβu
2
]. (4.5)
The variables u
1
and u
2
are thus independent and have
the same joint distribution under any risk-neutral measure Q.
This result is extended by convolution to n arbitrary non-
overlapping increments of ln S
t
in [0, T ]. It thus follows
that ln S
t
has independent increment in [0, T ] under any risk-
neutral measure.
33
P Balland QUANTI TATI VE FI NANCE
Finally, we show by induction that for any integer n, any
sequence {T
i
] ∈ [0, T ]
n
and any real U
i
:
Q{S
T
1
< U
1
, . . . , S
T
n
< U
n
] = P{S
T
1
< U
1
, . . . , S
T
n
< U
n
].
All risk-neutral probability measures Q coincide with P
on the algebra A of cylindrical sets {ω ∈ : S(T
1
, ω) <
U
1
, . . . , S(T
n
, ω) < U
n
]. Adirect application of the monotone
class theorem as in Jacod and Protter (1991, p 32) implies that
these probabilitymeasures coincide onσ(A) andthus on`
T
. ¬.
We note that the uniqueness of a risk-neutral measure
does not, however, imply completeness of the model since
the number of traded assets is infinite (see Jarrow and Madan
1999). We have proved that in a sticky-delta implied volatility
model, there exists a cadlag square-integrable P-martingale X
with P-independent increments such that
S
t
= S
0
m
0t
exp(X
t
)/E[exp(X
t
)].
Examples of processes with independent increments are
Brownian motions, Poisson processes, Levy processes and
jump-diffusion processes. We refer to Protter (1995) for
the construction of stochastic integrals with respect to cadlag
square-integrable martingales. In this paper, the stochastic
integrals froma 0 to b are integrals on (a, b]. We recall that
the quadratic variationof a cadlagsquare-integrable martingale
Z is an increasing cadlag adapted process defined by
[Z, Z]
t
= Z
2
t
−2
_
t
0
Z
u−
dZ
u
.
To ease notation, this process will also be denoted [Z]
t
or [Z
t
].
Restricting our attention to regular sticky-delta models
will ensure that X has finite moments and satisfy the
representation property of Nualart and Schoutens. This will
allowus toprove the quasi-completeness of regular sticky-delta
models. But first, we needsome notationandsome preliminary
results.
Since S is assumed quasi-left continuous, the martingale
X has no fixed points of discontinuity and can be decomposed
as in Jacod and Shiryaev (1987, p 77)
X
t
= (σ • W)
t
+ (x ∗ (N −n))
t
,

0<st
X
t
= (x ∗ N)
t
=
_
t
0
_
xN(dx ds),
where X
t
= X
t
−X
t −
is the jump of Xat time t , σ is a square-
integrable deterministic function, W is a Brownian motion,
N is a Poisson measure independent of W with deterministic
compensator n such that n
s
(dx) ds = E[N(dx ds)] where
N(dx ds) is the number of non-zero jumps of X in (s, s +
ds) (x, x + dx).
Since ln S
t
and thus X
t
are square-integrable, we conclude
that for t in [0, T ]
E
_

0<ut
X
2
u
_
=
_
t
0
_
+∞
−∞
x
2
n
u
(dx) du < ∞,
_
t
0
σ
2
u
du < ∞.
We recall the following notation:
(H • W)
t
=
_
t
0
H
u
dW
u
,
(∗ (N −n))
t
=
_
t
0
_

u
(x){N(dx du) −n
u
(dx) du].
By applying Ito’s formula (see Jacod and Shiryaev
1987) to S
t
.→ E[(S
s
− K)
+
[S
t
], we derive the following
representation for the short-dated option prices:
ˆ
C
t,s−t,K
= C
0sK
+ (H
C
sK
• W)
t
+ (
C
sK
∗ (N −n))
t
(4.6)
ˆ
P
t,s−t,K
= P
0sK
+ (H
P
sK
• W)
t
+ (
P
sK
∗ (N −n))
t
. (4.7)
where
ˆ
X
t
= P
0t
X
t
. Jensen inequality implies that the
martingales P
0t
C
t
(s −t, K) and P
0t
P
t
(s −t, K) are square-
integrable. It follows that H
C
(s, K), H
P
(s, K) are in L
2
W
and

C
(s, K),
P
(s, K) are in L
2
n
with
L
2
W
=
_
H ∈ : E
__
T
0
H
2
u
du
_
< ∞
_
,
L
2
n
=
_
∈ ⊗B(R) : E
__
T
0
_

2
u
(x) n
u
(dx) du
_
< ∞
_
,
where is the predictable σ-algebra on [0, T ] , that is the
smallest σ-algebra making all adapted processes that are left
continuous with right limits, measurable (see Protter 1995).
Let V
u
(t, K) = P
0u
P
u
(t − u, K) and recall that the
quadratic variation of this process satisfies with equation (4.7)
(see Protter 1995)
[V(t, K), V(t, K)]
u
= V
0
(t, K)
2
+
_
u
0
H
P
s
(t, K)
2
ds
+
_
u
0
_

P
s
(t, K)(x)
2
N(dx ds). (4.8)
Let λ(K) be a square-integrable function. The Lebesgue–
Fubini theorem for positive integrand (Malliavin and Airault
1994, p 46) implies that the following Lebesgue integrals, if
finite, satisfy
E
_ _
+∞
0
_
T
0
λ(K)
2
d[V
t,K
]
u
dK
_
= E
__
T
0
__
+∞
0

K
H
P
ut K
)
2
dK
_
du
_
+E
_ _
T
0
_ __
+∞
0

K

P
ut K
(x))
2
dK
_
n
u
(dx) du
_
.
Since S
t
and 1/S
t
are assumed to be square-integrable, we
obtain that for all t in [0, T ] and all i 0
E[[ ln S
t
[
i
exp([ ln S
t
[)] < i! E[S
2
t
+ S
−2
t
]. (4.9)
It follows that for all i 0:
sup
t ∈[0,T ]
{E[S
t
[ ln S
t
[
i
] + E[S
−1
t
[ ln S
t
[
i
]] < ∞. (4.10)
With equation (4.10), we prove the following lemma that
will be useful in establishing our quasi-completeness result.
34
QUANTI TATI VE FI NANCE Deterministic implied volatility models
Lemma 4.1. For any integers p 0 and any real α > 0, we
have
E
_ _
α
0
_
t
0
(ln K)
2p
K
−4
d
u
[
ˆ
P
u,t −u,K
]dK
_
< ∞, (4.11)
E
_ _
+∞
α
_
t
0
(ln K)
2p
K
−4
d
u
[
ˆ
C
u,t −u,K
]dK
_
< ∞, (4.12)
where [M] ≡ [M, M],
ˆ
M ≡ M/B for a cadlag process M.
Proof. Since
ˆ
P
u
(t −u, K) is a P-martingale, we observe that:
E
_ _
t
0
d
u
[
ˆ
P
u
(t −u, K)]
_
= E[
ˆ
P
t
(0+, K)
2
] −
ˆ
P
0
(t, K)
2
.
Therefore, we have
E
__
t
0
(ln K)
2p
K
4
d
u
[
ˆ
P
u,t −u,K
, ]
_
(ln K)
2p
K
−4
E[(K −S
t
)
+2
].
Using the Lebesgue–Fubini theorem for a positive
integrand, we obtain
_
α
0
(ln K)
2p
K
−4
E[(K−S
t
)
+2
]dK E[[ ln S
t
[∨[ ln α[
2p
S
−1
t
]
With (4.10), we conclude that
_
α
0
E
_ _
t
0
(ln K)
2p
K
−4
d
u
[
ˆ
P
u,t −u,K
]
_
< ∞
By Lebesgue–Fubini, we obtain (4.11). Similarly, we
derive inequality (4.12). ¬.
Since X has independent increments and E[exp(2X
t
)] +
E[exp(−2X
t
)] < ∞, the Laplace transform of X
t
is defined
for all [θ[ 2, t ∈ [0, T ] and satisfies the time-dependent
Levy–Khintchine formula for square-integrable processes:
E[exp(θ(X
t +x
−X
t
))] = exp
__
t +x
t
ψ
s
(θ) ds
_
,
ψ
s
(θ) =
1
2
σ
2
s
θ
2
+
_
+∞
−∞
(e
θz
−1 −θz) n
s
(dz).
Using the above formula, we show that the martingale X
for a regular sticky-delta model as in definition 2.2 is regular
in the following sense.
Definition 4.1. The martingale X is regular if there exists
λ 2:
sup
s∈(0,T )
_
σ
2
s
+
_
(−1,1)
c
exp(λ[x[)n
s
(dx)
_
< ∞, (4.13)
Thanks to (4.13), we derive for i 2:
_
+∞
−∞
[x[
i
n
s
(dx) <
i!
λ
i
_
(−1,1)
c
exp(λ[x[)n
s
(dx)+
_
1
−1
x
2
n
s
(dx)
Therefore for k, i 2, m
i
(t ) ≡
_
+∞
−∞
x
i
n
t
(dx) and
M
k
(t ) ≡ E
_
0<st
(X
s
)
k
_
=
_
t
0
m
k
(s) ds are uniformly
bounded in [0, T ].
In the case where the increments of X have stationary
distributions, we note that X is a Levy martingale and the
condition (4.13) is similar to the condition introduced by
Nualart and Schoutens (2000). With that restriction, we will
extend in section 6, the polynomial representation obtained by
Nualart and Schoutens for regular Levy martingales, to regular
martingales. With this representation property, we will derive
our quasi-completeness result in section 7. But first we need
to define precisely our admissible trading strategies and what
we mean by quasi-completeness, as this type of completeness
is not standard.
5. Trading strategies, attainable claims
and quasi-completeness
In this section, we define the trading strategies that will be used
to replicate contingent claims in a regular sticky-delta model.
We first define the static trading strategies, which are static
portfolios having a continuumof traded vanilla options. These
static trading strategies are traded instruments by assumption
(see section 2). As in the Black–Scholes theory, we then
define tradingstrategies as dynamic portfolios involvinga finite
number of traded instruments.
We highlight that in this paper, static portfolios with a
continuum of traded options, i.e. static trading strategies, are
assumedtobe tradedinstruments, as inCarr andMadan(1998).
We note that this assumption is equivalent to assuming that at
time t , all European claims, with maturity u ∈ (t, t + x
m
) and
payoffs f (S
u
), that can be decomposed in a continuum of call
and put option payoffs as in Breeden and Litzenberger (1978),
are traded instruments.
We define a static trading strategy λ to be a portfolio of
short-dated options with maturity t
1
and of the money market
account such that the portfolio holdings are constant over the
trading interval (t
0
, t
1
] ⊂ (t
0
, t
0
+ x
m
) and there are no flows
coming in or out of the strategy up to time t
1
.
A static strategy is characterized by an initial date t
0
, an
end date t
1
∈ (t
0
, t
0
+ x
m
), an initial endowment C
0
∈ `
t
0
,
two locally bounded functions λ
C
(K) and λ
P
(K) defined in
[α, +∞) and in (0, α] respectively, and two positive σ-finite
measures c
λ
(dK) and p
λ
(dK) defined on the positive half line.
The static trading strategy λ = (t
0
, t
1
, λ
P
, λ
C
, p, c, α) is
a portfolio that has constant holdings in (t
0
, t
1
] and that is
instantiated at zero cost using the initial endowment at time t
0
to purchase λ
C
(K)c
λ
(dK) unit(s) of the call option with strike
K ∈ (α, +∞) and maturity t
1
and λ
P
(K)p
λ
(dK) unit(s) of the
put option with strike K ∈ (0, α) and maturity t
1
.
We assume that the functions λ
C
(K) and λ
P
(K) satisfy
the following conditions:
E
_ _
α
0

P
K
[
ˆ
P
0t
1
K
p
λ
(dK) +
_
+∞
α

C
K
[
ˆ
C
0t
1
K
c
λ
(dK)
_
< ∞,
(5.1)
E
_ _
α
0
_
t
1
t
0
λ
P
(K)
2
d
s
[
ˆ
P
s,t
1
−s,K
]p
λ
(dK)
_
< ∞, (5.2)
E
_ _
+∞
α
_
t
1
t
0
λ
C
(K)
2
d
s
[
ˆ
C
s,t
1
−s,K
]c
λ
(dK)
_
< ∞ (5.3)
35
P Balland QUANTI TATI VE FI NANCE
where we recall [X] = [X, X]. Using the Cauchy–Schwartz
inequality, we show as in the proof of lemma 4.1, that
equations (5.1)–(5.3) are implied by the single equation
E
_ _
α
S
t
1
∧α

P
K
K)
2
p
λ
(dK) +
_
S
t
1
∨α
α

C
K
S
t
1
)
2
c
λ
(dK)
_
< ∞.
We impose the holding functions to satisfy (5.2) and (5.3)
in order to apply the stochastic Fubini theorem as in Protter
(1995, p 160), so as to permute strike and time integrations.
As we will see, this will allow us to define dynamic portfolios
based on static trading strategies.
The value of the static trading strategy at time t ∈ (t
0
, t
1
]
is obtained by adding the values of its constituents:
V
t
(λ) =
_
α
0
λ
P
(K)P
t
(t
1
−t, K)p
λ
(dK)
+
_
+∞
α
λ
C
(K)C
t
(t
1
−t, K)c
λ
(dK) + C
0
B
t
/B
t
0
.
The above two integrals are guaranteed to exist thanks
to (5.1). Since the static portfolio does not involve any cost at
initiation, we derive for t in (t
0
, t
1
):
ˆ
V
t
(λ) =
_
α
0
λ
P
(K)
__
t
t
0
d
s
ˆ
P
s
(t
1
−s, K)
_
p
λ
(dK)
+
_
+∞
α
λ
C
(K)
__
t
t
0
d
s
ˆ
C
s
(t
1
−s, K)
_
c
λ
(dK).
Using (5.2) and (5.3) together with (4.6)–(4.9), we show
that we can permute strike and time integration, by application
of the stochastic Fubini theoremas formulatedinProtter (1995)
for martingales and in Lebedev (1995) for random measures:
ˆ
V
t
(λ)=
___
α
0
λ
P
K
H
P
t
1
K
p
λ
(dK) +
_
+∞
α
λ
C
K
H
C
t
1
K
c
λ
(dK)
_
•W
_
t
t
0
+
___
α
0
λ
P
K

P
t
1
K
p
λ
(dK) +
_
+∞
α
λ
C
K

C
t
1
K
c
λ
(dK)
_
∗ (N −n)
_
t
t
0
where we have used the notation [X]
b
a
= X
b
− X
a
. In short
and with abuse of notations, we simply write
ˆ
V
t
(λ) =
_
t
t
0
_
α
0
λ
P
(K)p
λ
(dK) d
s
ˆ
P(t
1
−s, K)
+
_
t
t
0
_
+∞
α
λ
C
(K)c
λ
(dK) d
s
ˆ
C(t
1
−s, K).
We set
ˆ
V
t
(λ) =
ˆ
V
(t ∨t
0
)∧t
1
(λ) for t ∈ [0, T ]. This
discounted value process is a square-integrable cadlag
martingale which is zero before t
0
and constant after t
1
.
These static strategies which are portfolios with a
continuum of traded options, are traded instruments by
assumption.
A trading strategy λ is defined as a portfolio of n static
trading strategies and of the money market account. A trading
strategy is thus characterized by a finite sequence {
i
: 1
i n] of static trading strategies and by some predictable
processes {λ
i
t
: 1 i n] and by a progressively measurable
adapted process λ
B
t
such that
sup
t ∈(0,T )
E
__
n

i=1
λ
i
t
ˆ
V(
i
) + λ
B
t
_
2
_
< ∞, (5.4)
n

i,j=1
E
__
T
0
λ
i
s
λ
j
s
d[
ˆ
V(
i
),
ˆ
V(
j
)]
s
_
< ∞. (5.5)
At time t , the strategy λ is the portfolio consisting of λ
i
t
unit(s) of each of the static portfolios
i
and of λ
B
t
unit(s) of
the money market account.
The value of such a strategy is given at time t by the sum
of the values of its constituents:
V
t
(λ) =
n

i=1
λ
i
t
V
t
(
i
) + λ
B
t
B
t
.
Equation (5.4) guarantees that this process is a square-
integrable semimartingale. The gain cumulated up to time t
by the trading strategy is defined as usual by
G
t
(λ) ≡
n

i=1
_
t
0
λ
i
s
dV
s
(
i
) +
_
t
0
λ
B
s
dB
s
.
Equation (5.5) guarantees the existence of the above
integrals. A trading strategy is self-financing if and only if
the value process satisfies as usual
V
t
(λ) = V
0
(λ) + G
t
(λ).
Hence the discounted value process of a self-financing
trading strategy is a square-integrable martingale that satisfies
ˆ
V
t
(λ) = V
0
(λ) +
n

i=1
_
t
0
λ
i
s
d
ˆ
V
s
(
i
). (5.6)
A self-financing trading strategy is thus entirely
characterized by λ = (λ
i
t
,
i
) where λ
i
t
is a predictable process
satisfying (5.4) and (5.5). We denote by the space of all self-
financing trading strategies. We define the following linear
subspace of L
2
(, `), V() = {V
t
(λ) : t ∈ [0, T ], λ ∈ ].
As explained in Jarrow and Madan (1999), an arbitrage
strategy is a self-financing strategy λ such that
V
0
(λ) = 0, P(V
T
(λ) 0) = 1, P(V
T
(λ) > 0) > 0.
Proposition 5.1. Implied volatility models are arbitrage-free.
Proof. Suppose that λ is an arbitrage strategy then
E[1{V
T
(λ) < 0]] = 0 and thus, we have E[V
T
(λ) 1{V
T
(λ) <
0]] = 0. It follows that E[
ˆ
V
T
(λ)] = E[[
ˆ
V
T
(λ)[] > 0 =
ˆ
V
0
(λ).
This contradicts the fact that
ˆ
V
t
(λ) is a P-martingale. ¬.
Remark 5.1. This result is not surprising since we have
assumed the existence of at least one risk-neutral measure P.
We can now define attainability and attainability up to a
vanishing risk.
Definition 5.1. A claim ∈ L
2
(, `) with maturity t T is
attainable or can be replicated if there is a self-financing
trading strategy λ such that = V
t
(λ), i.e. ∈ V().
A claim ∈ L
2
(, `) with maturity t T is attainable up
to a vanishing risk or can be replicated up to a vanishing risk
if there is a sequence of self-financing strategies {λ
n
] such
that lim
n→∞
E[(−V
t

n
))
2
] = 0, i.e. ∈
¨
V().
36
QUANTI TATI VE FI NANCE Deterministic implied volatility models
A claim ∈ L
2
(, `) with maturity t is thus
attainable if there are some static trading strategies {φ
i
=
(t
0i
, t
1i
, φ
P
i
, φ
C
i
, p
i
, c
i
, α
i
) : 0 i n], some predictable
processes {λ
i
t
: 0 i n] satisfying (5.4) and (5.5) and a real
V
0
= E[P
0t
] which is the cost of replication, such that
P
0t
= V
0
+
n

i=0
_
t
1i
t
0i
_
α
i
0
λ
i
s
g
i
(K)p
i
(dK) d
s
ˆ
P
s
(t
1i
−s, K)
+
n

i=0
_
t
1i
t
0i
_
+∞
α
i
λ
i
s
f
i
(K)c
i
(dK) d
s
ˆ
C
s
(t
1i
−s, K).
We say that a claim ∈ L
2
(, `) with maturity t is
attainable at time s if there exists a self-financing strategy λ
and C
s
∈ `
s
such that P
0t
=
ˆ
V
t
(λ) −
ˆ
V
s
(λ) + C
s
. Similarly,
we define attainable up to a vanishing risk, at time s.
With the above notations, we define quasi-completeness
as in Jarrow and Madan (1999) and in Bjork et al (1997).
Definition 5.2. An implied volatility model is complete up to
a vanishing risk or quasi-complete if any ∈ L
2
(, `) is
attainable up to a vanishing risk i.e. L
2
(, `) −
¨
V().
We now prove the following classic result to be used later.
Proposition 5.2. Let a ∈ (0, x
m
), b = t + a ∈ [0, T ] and let
f be a C
2
(0, +∞) function with f (S
b
) ∈ L
2
(, `) such that
there is 0 < α < ∞satisfying
E
__
α
S
b
∧α
f
//
(K)
2
K
2
dK +
_
S
b
∨α
α
f
//
(K)
2
S
2
b
dK
_
< ∞.
(5.7)
Then the claim with payoff f (S
b
) at time b can be replicated
at time t at a cost P
t b
E
t
[f (S
b
)].
Proof. Following Carr and Madan (1998), we decompose the
payoff into four components
f (S
b
) =
_
α
0
(K −S
b
)
+
f
//
(K) dK
+
_
+∞
α
(S
b
−K)
+
f
//
(K) dK + f (α)
+f
/
(α)(S
b
−α) ≡ A + B + C + D. (5.8)
Thanks to our assumptions on f
//
, we derive by Cauchy–
Schwartz:
_
α
0
P
0bK
[f
//
(K)[dK +
_
+∞
α
C
0bK
[f
//
(K)[dK < ∞. (5.9)
[f
//
(K)[(K −S)
+
1{K < α] + [f
//
(K)[(S −K)
+
1{K > α] is
thus P{S
b
∈ dS]dK integrable on (0, +∞)
2
and the Fubini–
Lebesgue theoremimplies (Malliavin and Airault (1993, p 46))
P
t b
E
t
[A+B] =
_
α
0
f
//
(z)P
t
(a, z) dz+
_
+∞
α
f
//
(z)C
t
(a, z) dz.
Thanks to our assumption on f
//
, we conclude that the
claim A + B is square-integrable and attainable at t by the
static trading strategy (t, b, f
//
, f
//
, dK, dK, α):
P
t b
(A + B) = P
t b
E
t
[A + B] +
_
b
t
_
α
0
f
//
(z) dz d
s
ˆ
P
s
(b −s, z)
+
_
b
t
_
+∞
α
f
//
(z) dz d
s
ˆ
C
s
(b −s, z).
The third component is trivially replicated by taking
position in the money market account. The fourth component
is replicated by buying a call option and selling a put option
since it is the terminal value of a forward contract. ¬.
Corollary 5.1. The above result still holds if we replace in
(5.7) the squares by absolute values and replicated by
replicating up to a vanishing risk.
Proof. Consider the sequence of smooth functions f
n
defined
by replacing in (5.8) f
//
(K) by f
//
(K)1{[f
//
(K)[ < n]. By
applying the Lebesgue dominated convergence theorem, we
obtain lim
n→∞
E[(f
n
−f )
2
] = 0.
By application of proposition 5.2 to f
n
(S
b
), we conclude
that f
n
(S
b
) is attainable and thus at time t , the claim f (S
b
) ∈
L
2
(, `) can be replicated up to a vanishing risk at a cost
P
t b
E
t
[f (S
b
)]. ¬.
Proposition 5.3. Let U
t
be a square-integrable martingale
such that the claim U
T
with maturity T is attainable and let
γ
t
be a predictable bounded process. Then the claim with
payoff at time a, Z
a
=
_
a
0
γ
t
dU
t
∈ L
2
(, `), can be
replicated at zero cost.
Proof. The claim with payoff U
T
/P
0T
is attained by a
self-financing trading strategy λ = (λ
i
s
, φ
i
). We note that
ˆ
V
t
(λ) = E
t
[U
T
] = U
t
satisfies
U
t
= U
0
+
n

i=1
_
t
0
λ
i
s
d
ˆ
V
s

i
).
Since γ
t
is a predictable bounded process, (γ
s
λ
i
s
, φ
i
)
defines a self-financing trading strategy and we have
_
a
0
γ
t
dU
t
=
n

i=1
_
a
0
γ
t
λ
i
t
d
ˆ
V
t

i
).
Therefore, the claim Z
a
with maturity a is attainable at
zero cost. ¬.
We conclude this section with a first illustration of
how hedging in a sticky-delta model works in practice (see
remark 7.2).
Proposition 5.4. All call and put options with maturity in
[0, T ] are attainable.
Proof. Consider the long-dated call option with strike K > 0
and maturity t = kx
m
/2 + x in [0, T ] where k is an integer
and 0 < x x
m
/2. Given the set of dates {t
i
= (ix
m
/2) ∧ t :
1 i k + 1], we shall construct a sequence of static trading
strategies φ
i
having trading intervals (t
i
, t
i+1
] such that (1, φ
i
)
replicates the long-dated call option.
At time t
k
= kx
m
/2, the call option is a short-dated
option and thus can be replicated by purchasing the call option
itself. The discounted value process associated with this static
37
P Balland QUANTI TATI VE FI NANCE
strategy is
ˆ
V
u
=
ˆ
C
k
(t −u, K) for u ∈ [t
k
, t ]. Given the sticky-
delta assumption, we obtain
ˆ
V
t
k
= P
0t
E[(S
t
−K)
+
[S
t
k
] = P
0t
S
t
k
c(t −t
k
, K/S
t
k
),
where c(x, y) = E[(S
t
/S
t −x
−y)
+
] is a C
2
function which is
decreasing and convex with respect to the second argument.
The density of S
t
/S
t −x
is ∂
2
y
c(x, y) and y
2

2
y
c(x, y) is
uniformly bounded in (0, +∞). We note that
ˆ
V ∈ L
2
(, `)
and ∂
2
S
t
k
ˆ
V
t
k
= K
2

2
y
c(t
k
, K/S
t
k
)/S
3
t
k
0. We derive
E
__
1
S
k
∧1

2
y
c(t
k
, K/S
k
)
2
K
6
/S
6
k
dK
_
< ∞,
E
__
S
k
∨1
1

2
y
c(t
k
, K/S
k
)
2
K
4
/S
4
k
dK
_
< ∞.
We can thus apply proposition 5.2 and replicate the claim
ˆ
V
t
k
between time t
k−1
and t
k
using a static portfolio with
positive holdings in short-dated call and put options having
maturity t
k
, at a discounted cost
ˆ
V
t
k−1
= P
0t
E[(S
t
−K)
+
[S
t
k−1
] = P
0t
S
t
k−1
c(t −t
k−1
, K/S
t
k−1
).
By repeating the above argument, we obtain a set of static
trading strategy {φ
i
: 0 i k] with positive holdings such
that the self-financing trading strategy (1, φ
i
) replicates the
long-dated call option at the following initial cost:
ˆ
V
0
= P
0t
E[(S
t
−K)
+
].
We obtain the result for the long-dated put options by put–
call parity. ¬.
In order to construct similar hedging strategies for square-
integrable claims, we need a representation property for X.
6. Representation property for regular
martingales with independent increments
Inthis sectionandinthe appendix, we extendthe representation
property obtained by Nualart and Schoutens (2000) for Levy
processes, to regular martingales with independent increments.
We adapt the notation and the approach taken by Nualart and
Schoutens to our purpose.
We define the Teugels martingales on [0, T ] as in Nualart
and Schoutens (2000):
Y
(1)
t
= X
t
,
Y
(k)
t

0<st
(X
s
)
k

_
t
0
m
k
(s) ds, (k 2).
It is clear that these martingales are square-integrable
with independent increments and finite moments of all orders.
In the appendix and by [ ]-orthogonalization of the Teugels
martingales, we construct the following family of pairwise [ ]-
orthogonal martingales having independent increments:
H
(1)
t
= Y
(1)
t
,
H
(i)
t
=
_
t
0
dY
(i)
s
+ a
i,i−1
(s) dY
(i−1)
s
+ . . . + a
i,1
(s) dY
(1)
s
.
The deterministic functions a
ij
(s) are bounded in [0, T ]
and are such that the martingales H
(i)
t
are square-integrable
and pairwise strongly orthogonal.
We show in the appendix that the [ ]-orthogonal family
{H
(i)
: i 1] forms a complete basis of L
2
(, `). More
precisely, we have the following representation property.
Proposition 6.1. Let F ∈ L
2
(, `) then there are
predictable processes {φ
(i)
t
: i 1] such that
F = E[F] +
+∞

i=1
_
T
0
φ
(i)
s
dH
(i)
s
,
where {φ
(i)
t
: i 1] belongs to ⊕
+∞
i=1
L
2
H
(i)
.
Proof. See appendix. ¬.
Using this representation property, we prove in the next
section that in regular sticky-delta models, all square integrable
claims can be replicated up to a vanishing residual risk by
trading portfolios of vanilla options. We cannot typically
replicate, in the classic sense, contingent claims because the
family of martingales underlying the representation property
of proposition 6.1 has in general an infinite dimension and thus
exact replication would be possible only if strategies based on
an infinite number of traded instruments were admissible.
7. Quasi-completeness of regular
sticky-delta implied volatility models
In this section, we prove that all regular sticky-delta implied
volatility models are quasi-complete. But first, we show that
the claims H
(i)
T
are attainable by application of Ito’s lemma to
polynomial functions. We define f
0b
≡ F
0b
/E[exp(X
b
)] and
observe that X
b
= ln(S
b
/f
0b
).
Lemma 7.1. Let i 2. The claim with payoff (X
b
)
i
at time b
is attainable at time a > b −x
m
using the static strategy:
X
i
b
= E
a
[(X
b
)
i
] +
_
b
a
_
f
0b
0
x
(i)
(K)K
−2
dKd
u
ˆ
P
u
(b −u, K)
+
_
b
a
_
+∞
f
0b
x
(i)
(K)K
−2
dK d
u
ˆ
C
u
(b −u, K),
x
(i)
(K) = {i(i −1)(ln{K/f
0b
])
i−2
−i(ln{K/f
0b
])
i−1
]/P
0b
.
Proof. Using equation (4.10) and proposition 5.2 with f (S) =
ln(S/f
0b
)
i
, we derive the result. ¬.
38
QUANTI TATI VE FI NANCE Deterministic implied volatility models
Proposition 7.1. For i 1 and b ∈ (a, a + x
m
), H
(i)
b
−H
(i)
a
satisfies
H
(i)
b
−H
(i)
a
=
_
b
a
_
f
0b
0
h
(i)
u
(K)K
−2
dKd
u
ˆ
P
u
(b −u, K)
+
_
b
a
_
+∞
f
0b
h
(i)
u
(K)K
−2
dK d
u
ˆ
C
u
(b −u, K)
+
_
b
a
g
(i)
u
d
u
ˆ
C
u
(b −u, f
0b
) −
_
b
a
g
(i)
u
d
u
ˆ
P
u
(b −u, f
0b
),
h
(i)
u
(K) =

0jki−1
h
jk
u,i
X
j
u−
(ln(K/f
0b
))
k
,
g
(i)
u
=

0ki−1
g
k
u,i
X
k
u−
.
The coefficients h
jk
u,i
and g
k
u,i
are deterministic and uniformly
bounded on [0, T ]. For all i 1, the claim H
(i)
T
is attainable,
i.e. H
(i)
T
∈ V().
Proof. Y
(1)
t
= ln(S
t
/f
0t
) is a square-integrable martingale. By
proposition 5.2, we conclude that Y
(1)
b
is attainable using the
static trading strategy defined by
P
0b
Y
(1)
t
= −
_
f
0b
0
ˆ
P
t
(b −t, K)K
−2
dK

_
+∞
f
0b
ˆ
C
t
(b −t, K)K
−2
dK
+(
ˆ
C
t
(b −t, f
0b
) −
ˆ
P
t
(b −t, f
0b
))/f
0b
.
This equation implies for t ∈ [a, b]
Y
(1)
t
−Y
(1)
a
=
_
t
a
_
f
0b
0
y
(1)
u
(K)K
−2
dK d
u
ˆ
P
u
(b −u, K)
+
_
t
a
_
+∞
f
0b
y
(1)
u
(K)K
−2
dK d
u
ˆ
C
u
(b −u, K)
+
_
t
a
f
(1)
u
d
u
ˆ
C
u
(b −u, f
0b
) −
_
t
a
f
(1)
u
d
u
ˆ
P
u
(b −u, f
0b
),
where y
(1)
u
(K) = −
1
P
0b
and f
(1)
u
=
1
P
0b
f
0b
.
With Ito’s formula for a real function f of class C
2
:

0<ut
{f (X
u
) −f
/
(X
u−
)X
u
] = f (X
t
) −f (X
0
)

_
t
0
f
/
(X
u−
) dX
u

1
2
_
t
0
f
//
(X
u−

2
u
du.
Taking f (x) = x
2
, we derive:
Y
(2)
b
= (X
b
)
2
−E[(X
b
)
2
] −2
_
b
0
X
u−
dY
(1)
u
.
Finally, the above equation and lemma 7.1 imply that
Y
(2)
b
−Y
(2)
a
is attainable:
Y
(2)
b
−Y
(2)
a
=
_
b
a
_
f
0b
0
y
(2)
u
(K)K
−2
dK d
u
ˆ
P
u
(b −u, K)
+
_
b
a
_
+∞
f
0b
y
(2)
u
(K)K
−2
dK d
u
ˆ
C
u
(b −u, K)
+
_
b
a
f
(2)
u
d
ˆ
C
u
(b −u, f
0b
) −
_
b
a
f
(2)
u
d
ˆ
P
u
(b −u, f
0b
)
y
(2)
u
(K) = 2{X
u−
−ln(K/f
0b
) + 1]/P
0b
f
(2)
u
= −2X
u

f
(1)
u
Similarly, we obtain by applying Ito’s lemma to
f (x) = x
3
:

0<ub
(X
u
)
3

_
b
0
m
3
(s) ds = (X
b
)
3
−E[(X
b
)
3
]
−3
_
b
0
X
u−
dY
(2)
u
−3
_
b
0
(X
2
u−
+ V
b
−V
u
+ M
(2)
b
−M
(2)
u
) dY
(1)
u
where V
t
=
_
t
0
σ
2
s
ds and M
(2)
t
=
_
t
0
m
2
(s) ds.
It follows that:
Y
(3)
b
−Y
(3)
a
=
_
b
a
_
f
0b
0
y
(3)
u
(K)K
−2
dK d
u
ˆ
P
u
(b −u, K)
+
_
b
a
_
+∞
f
0b
y
(3)
u
(K)K
−2
dK d
u
ˆ
C
u
(b −u, K)
+
_
b
a
f
(3)
u
d
ˆ
C
u
(b −u, f
0b
) −
_
b
a
f
(3)
u
d
ˆ
P
u
(b −u, f
0b
)
where y
(3)
u
(K) = Q
(2)
u
(X
u−
, ln(K/f
0b
)), f
(3)
u
= R
(2)
u
(X
u

)
and the polynomials Q
(k)
u
(X, Y) =

0ijk
q
ij
u,k
X
i
Y
j
and R
(k)
u
(X) =

0ik
r
i
u,k
X
i
have coefficients that are
deterministic and uniformly bounded in [0, T ].
By induction, we extend the above formula to all Y
(k)
b

Y
(k)
a
. We recall that:
[H
(k)
]
b
a
=
_
b
a
dY
(k)
u
+ a
k,k−1
(u) dY
(k−1)
u
+ . . . + a
k,1
(u)dY
(1)
u
,
where the coefficients are deterministic and bounded in [0, T ].
We finally obtain the promised expression for H
(k)
b
− H
(k)
a
by using the previous equation for Y
(k)
b
− Y
(k)
a
. By adding
the above decompositions obtained for a = t
l−1
, b = t
l
with
l = 1, . . . , [2T/x
m
] + 1, t
l
= (lx
m
/2) ∧ T , we obtain a self-
financing trading strategy that replicates H
(k)
T
. ¬.
Finally, we have the following quasi-completeness result.
Theorem 7.1. Any square-integrable claim ∈ L
2
(, `)
with maturity T can be replicated up to a vanishing risk at a
cost P
0T
E[] by trading the underlying, the money-market
account and some portfolios of traded call and put options. A
regular sticky-delta model is quasi-complete or complete up
to a vanishing risk.
Proof. According to proposition 6.1, the random variable
∈ L
2
(, `) can be represented as follows:
= E[] +
+∞

i=1
_
T
0
ξ
(i)
s
dH
(i)
s
,
where ξ
(i)
t
is in L
2
H
(i)
.
We define
n
= E[] +

n
i=1
_
T
0
ξ
(i,n)
s
dH
(i)
s
∈ V()
with ξ
(i,n)
t
= ξ
(i)
y
1{[ξ
(i)
t
[ < n].
39
P Balland QUANTI TATI VE FI NANCE
Thanks to the strong orthogonality of H
(i)
t
, we obtain
E[(−
n
)
2
] E
__
−E[] −
n

i=1
_
T
0
ξ
(i)
s
dH
(i)
s
_
2
_
+E
_
+∞

i=1
_
T
0

(i)
s
−ξ
(i,n)
s
]
2
d[H
(i)
, H
(i)
]
s
_
.
The first sequence on the RHS converges to zero since

(i)
t
: i 1] belongs to ⊕
+∞
j=1
L
2
H
(j)
. Observe next that:
lim
n→∞

(i,n)
s
−ξ
(i)
s
)
2
= 0, (ξ
(i,n)
s
−ξ
(i)
s
)
2
4(ξ
(i)
s
)
2
.
By three applications of the dominated convergence
Lebesgue theorem, we derive
lim
n→∞
E
_
+∞

i=1
_
T
0

(i)
s
−ξ
(i,n)
s
)
2
d[H
(i)
, H
(i)
]
s
_
= 0.
Finally, we have lim
n→∞
E[(−
n
)
2
] = 0. Each claim

n
is attainable by application of propositions 7.1 and 5.3 at a
cost P
0T
E[]. ¬.
Remark 7.1. The concept of quasi-completeness is due to
Jarrow and Madan (1999) and to Bjork et al (1997).
The next proposition shows that some square-integrable
claims can be replicated in the classical sense, by rolling a
portfolio of vanilla options.
Proposition 7.2. Consider a claim with square-integrable
payoff f (S
T
0
, . . . , S
T
n
) at time T
n
. We assume that this payoff
is such that S
T
i
.→ E[f (S
T
0
, . . . , S
T
n
)[`
T
i−1
∨ S
T
i
] satisfies,
after subtraction of a finite number of call and put payoffs,
equation (5.7) with b = T
i
. Then the claim can be replicated
by trading the underlying, the money-market account and
some portfolios of traded call and put options.
Proof. Consider a claim with maturity T
n
such as an Asian
option, a discrete barrier option, a Parisian option or a volatility
swap, that has a finite number of fixing dates T
i
with 0 <
T
i
−T
i−1
< x
m
and a square-integrable payoff f (S
T
0
, . . . , S
T
n
)
as in proposition 7.2. At time T
n−1
, the payoff can be replicated
using a portfolio of call and put options, a forward contract and
a zero coupon bond as in proposition 5.2. The value V
T
n−1
of
the complex option at time T
n−1
, is thus
V
T
n−1
= P
T
n−1
,T
n−1
[℘
T
n−1
,T
n−1
f (S
T
0
, . . . , S
T
n−1
, •)](S
T
n−1
),
where the linear operator ℘
t x
is defined by
[℘
t x
f ](S) ≡
_
+∞
0
f (I Sm
t x
)∂
2
I
C
BS
(xσ
t
(x, I)
2
, 1, I) dI.
The sticky-delta assumption implies that the operator

T
n−1
,T
n−1
is deterministic and thus V
T
n−1
is a deterministic
function of S
T
0
, . . . , S
T
n−1
. At time T
n−2
, the complex option
can thus be replicated by the use of calls, puts, forwards and
zero coupon bonds with maturity T
n−1
. Finally, we derive by
induction that the complex option can be replicated at a cost
V
0
= P
0T
n
_

0,T
0
n−1

i=0

T
i
,T
i
_
(f )(S
0
) = P
0T
n
E[f ].
¬.
Remark 7.2. By modifying the above strategy, we obtain a
super-replication strategy for the complex option when there
are transaction costs on the implied volatility or when the
short-dated implied volatility smile is ‘uncertain’ but
bounded. Observe that the transition operators ℘
T
k
,T
k
are
nonlinear in this case.
The previous expectation can be estimated by using the
Monte Carlo method, a fast Fourier transform as in Carr and
Madan (1999) or by solving the integro-differential equation

t
V + A
t
V = r
t
V where A
t
is the generator defined by
A
t
= (∂
x

t x
)
x=0
+ .
8. Regular geometric Levy models
We consider a regular sticky-delta implied volatility model
and we suppose that the per-delta implied volatility processes
are independent of time. Our previous analysis in section 4
shows that the increments of ln S
t
are independent and have a
stationary distribution under the probability measure P, i.e.
S
t
= F
0t
exp(L
t
−ln E[e
L
t
]),
where L is a regular P-Levy martingale (see Levy 1965).
We define a regular geometric Levy under P by:
S
t
= F
0t
exp(L
t
−ln E[e
L
t
]),
C
t
(x, K) = P
t x
E
t
[(S
t +x
−K)
+
],
P
t
(x, K) = P
t x
E
t
[(K −S
t +x
)
+
],
where x < x
m
and Lis an adapted regular P-Levy martingale.
Proposition 8.1. Regular geometric Levy models are
arbitrage-free and quasi-complete in the sense that all
square-integrable claims can be replicated up to a vanishing
risk by trading the underlying, the money-market account and
portfolios of short-dated call and put options.
Proof. By direct calculation, we show that regular geometric
Levy models are regular stationary sticky-delta implied
volatility models. Therefore, these models are arbitrage-free
and quasi-complete by application of theorem 7.1. ¬.
There is a long list of geometric Levy models proposed
as alternatives to the Black–Scholes model. We mention,
in particular, Mandelbrot (1963), Merton (1976), Madan and
Seneta (1990), the continuous-time formulation by Koponen
(1995) of the ‘truncated Levy flight’ introduced by Mantegna
and Stanley (1994), Eberlein and Keller (1995), Barndorff-
Nielsen (1995), Bouchaud, Cont and Potters (1998), Bjork
et al (1997) and Schoutens (2001).
40
QUANTI TATI VE FI NANCE Deterministic implied volatility models
9. Characterization of sticky-strike
implied volatility models
As with sticky-delta models, we consider only regular sticky-
strike models. We denote v
t x
(K) ≡ x
t
(x, Km
t x
)
2
, the den-
sity of S
t
by p
t
and
t x
(S, K) ≡
BS
(x
t
(x, Km
t x
)
2
, S, K).
Lemma 9.1. For each t < T , there exists a positive, locally
bounded, P-integrable function h
t
such that:
lim
k→+∞

x
v
t x
k
(S
t
) = h
t
(S
t
) P a.s.,
where x
k
is a sequence with limit zero.
Proof. First, we note by Jensen inequality:
E
t
[(S
t +x
/m
t x
−K)
+
] = E
t
[(S
t +x+y
/m
t,x+y
−K)
+
].
Hence v
t x
(K) is increasing with x. To simplify notation,
we assume that p
t
> 0 in (0, +∞). The case where p
t
vanishes
is treated similarly since we have P(p
t
(S
t
) = 0) = 0.
Using the Black–Scholes equation, we derive:

x
E
t
[S
2
t +x
/F
2
t x
] =
_
+∞
0

x
v
t x
(K)
t x
(S
t
, K)dK.
For
¨
I ⊂ (0, +∞), we define the increasing function:
H
It x
: z .→
_
I

x
v
t x
(K)E[
t x
(S
t
, K) 1{S
t
< z]dK.
We observe that for any positive z:
H
It x
(z) E[sup
x

x
E
t
[S
2
t +x
/F
2
t x
]].
Definition 2.2 implies that the increasing functions H
It x
are
uniformly bounded with respect to x and I. Helly’s theorem
implies that there is a sequence x
I,n
with limit zero and an
increasing function H
I,t
such that (see Doob 1994):
lim
n→+∞
H
It x
In
(z) = H
I,t
(z).
The increasing function H
I,t
satisfies for 0 a < b:
H
I,t
(b) −H
I,t
(a)
_
b
a
sup
x

x
E[S
2
t +x
/F
2
t x
[S
t
]p
t
(S
t
)dS
t
.
The Radon–Nikodym theorem implies that there is a positive,
locally bounded, P-integrable function h
I,t
such that:
H
I,t
(b) −H
I,t
(a) =
_
b
a
h
I,t
(z)p
t
(z)dz (a, b > 0).
0 h
I,t
(z) sup
0<x<T −t

x
E[S
2
t +x
/F
2
t x
[S
t
= z] P a.s.
Finally, we conclude that for any compact A ⊂ (0, +∞):
lim
n→+∞
_
I

x
v
t x
I,n
E[
t x
I,n
(S
t
, K)1
A
(S
t
)]dK =
_
A
h
I,t
p
t
dz.
Since E[
t,x
I,n
1
A
]/p
t
converges uniformly to 1
A
on I:
lim
n→+∞
_
I

x
v
t x
I,n
1
A
(K)p
t
(K)dK =
_
A
h
I,t
p
t
dz.
Hence ∂
x
v
t x
I,n
(S
t
) converges in probability to h
I,t
(S
t
) in I and
there exists x
I,σ(n)
with limit zero, such that (see Doob 1994):
lim
n→+∞

x
v
t x
I,σ(n)
(S
t
) = h
I,t
(S
t
) (S
t
∈ I, P a.s.).
Application of Fatou’s theorem gives (see Doob 1994):
lim
n→+∞
_
I
E[
t x
I,n
(S
t
, K)1
A
(S
t
)] [∂
x
v
t x
I,n
−h
I,t
[dK = 0.
(9.1)
We define I
k
=
_
1
k+1
,
1
k
_
∪ [k, k + 1) and construct, as
previously, a family of positive P-integrable functions h
¨
I
k
,t
and sub-sequences x
σ
k
(n)
⊂ x
σ
k−1
(n)
converging to zero such
that:
lim
n→+∞

x
v
t x
σ
k
(n)
(S
t
) = h
¨
I
k
,t
(S
t
) (k > 0, S
t
∈ I
k
, P a.s.).
Using the diagonal procedure, we define the sequence x
φ(n)
=
x
σ
n
(n)
with limit zero, and the positive, locally bounded, P-
integrable function:
h
t
(S) =
+∞

k=1
1
I
k
(S)h
¨
I
k
,t
(S) sup
x

x
E[S
2
t +x
/F
2
t x
[S
t
= S].
Using equation (9.1), we obtain for any compact sets A, C:
lim
n→+∞
_
C
E[
t x
φ(n)
(S
t
, K)1
A
(S
t
)][∂
x
v
t x
φ(n)
−h
t
[dK = 0.
(9.2)
Since

k
I
k
= (0, +∞), we finally conclude that:
lim
n→+∞

x
v
t x
φ(n)
(S
t
) = h
t
(S
t
) P a.s.
¬.
In fact, we have a stronger result.
Theorem 9.1. In a regular sticky-strike implied volatility
model, the per-strike implied volatility
t
(x, K) is
independent of K. Hence the Black–Scholes model is the only
arbitrage-free regular sticky-strike model.
Proof. The implied volatility model with zero-drift underlying
S
t
/m
0t
and implied volatility
t
(x, Km
0t
) is regular and
sticky-strike. Therefore, there is no loss of generality in
assuming zero drift i.e. m
0t
= 1.
For any bounded, Borel measurable function f , we have:
P
t x
E
P
t
[f (S
t +x
)] =
_
+∞
0

2
K
C
t
(x, K)f (K)dK.
With our assumptions, the asset price process S
t
is a Markov
process entirely characterized by the transition function


t x
(see Revuz and Yor 1991):
(


t x
f )(S
t
) = E[f (S
t +x
)[S
t
].
41
P Balland QUANTI TATI VE FI NANCE
The transition function is Feller and it is thus associated with
an infinitesimal generator

A
t
defined on D
A
t
satisfying:
[

A
t
f ](S
t
) = lim
x↓0

x
[


t x
f ](S
t
).
Let f ⊂ D
A
t
∩ C
2
with compact support . We obtain:

x
[


t x
f ](S
t
) =
_


x
C
BS
(v
t x
(K), S
t
, K)f
//
(K)dK. (9.3)
The Black–Scholes function satisfies:

V
C
BS
(V, F, K) =
1
2
F
2

BS
(V, F, K).
Hence, equation (9.3) can be written as follows:

x
[


t x
f ](S
t
) =
1
2
_

S
2
t

t x
(S
t
, K) ∂
x
v
t x
(K)f
//
(K)dK.
According to lemma 9.1, there is a positive, locally bounded
P-integrable function h
t
such that:
lim
n→∞

x
v
t x
n
(S
t
) = h
t
(S
t
) P a.s.
By taking first the expectation of equation (9.3) on a compact
C and then the limit as x
n
tends to zero, we obtain with (9.2):
_
C
[

A
t
f ](K)p
t
(K)dK =
1
2
_
C
h
t
(K)K
2
f
//
(K)p
t
(K)dK.
We finally obtain the following expression for the generator:
[

A
t
f ](S
t
) =
1
2
h
t
(S
t
)S
2
t
f
//
(S
t
) P a.s. (9.4)
The forward price V(t, S
t
) of a call option with maturity a and
strike K is a P-martingale. Therefore, V satisfies the backward
equation (see Revuz and Yor 1991):

t
V(t, S
t
) + [

A
t
V](S
t
) = 0. (9.5)
On the other hand, V satisfies:
V(t, S
t
) = C
BS
((a −t )
t
(a −t, K)
2
, S
t
, K).
Since V has the same second-order derivatives with respect to
S as the Black–Scholes function, it follows that:

t
V(t, S
t
) =
1
2

t
[(a −t )
t
(a −t, K)
2
]S
2

2
S
V(t, S
t
). (9.6)
Finally, equations (9.4)–(9.6) imply that we have for all
K, t, a ∈ (t, t + x
m
):

t
(a −t, K)
2
=
1
a −t
_
a
t
E[h
u
(S
u
)]du.
Hence, the implied volatility
t
(x, K) is independent of K
and the regular sticky-strike model coincides with the Black–
Scholes model! ¬.
Acknowledgments
I would like to thank L P Hughston, D B Madan, T Bjork,
W Schoutens, D Nualart, R Cont, an anonymous referee, the
participants of the EURANDOM seminar on Levy processes
(2001), the participants of the AMS Meeting (2001) and
my colleagues at Merrill Lynch for stimulating and fruitful
discussions.
Appendix. Representation property for
regular martingales
In this appendix, we extend the Schoutens–Nualart
representation property obtained for regular Levy martingales
to regular martingales with independent increments. We use
the notations of section6andwe followcloselythe presentation
of Nualart and Schoutens (2000).
We recall that two square-integrable martingales M and N
are said to be strongly orthogonal or [ ]-orthogonal if [M, N]
t
is a martingale (see Protter 1995).
Proposition A.1. There exists a family of pairwise strongly
orthogonal square-integrable martingales {H
(i)
: i 1].
Proof. We define the following inner product acting on the
space of real polynomials with time-dependent coefficients in
L
2
(0, T ):
¸Q, R) =
_
T
0
_
+∞
−∞
Q(x, s)R(x, s)(x
2
n
s
(dx)−σ
2
s
δ(x)dx)ds.
By ¸)-orthogonalization of the total family {1{s < t ]x
i
:
i 0, t ∈ (0, T )], we find a family of pairwise ¸)-orthogonal
polynomials {R
i
(x, s)] with bounded coefficients such that:
R
i
(x, s) = a
i+1,1
(s) + a
i+1,2
(s)x + . . . + a
i+1,i
(s)x
i−1
+ x
i
,
¸R
i
, 1{s < t ]x
j
) = 0, (t < T, 1 j i −1).
The coefficients of the above polynomial are bounded
because the measure n(s, dx) has finite moments of order
i 2, uniformly bounded in [0, T ].
Define Q
i+1
(x, s) ≡ x(R
i
(x, s) −a
i+1,1
(s)), a
i+1,i+1
≡ 1
and the following square integrable martingale:
H
(i+1)
t

_
t
0
a
i+1,1
(s)dY
(1)
s
+ . . . + a
i+1,i+1
(s) dY
(i+1)
s
.
By direct calculation, we derive:
H
(i+1)
t
=
_
t
0
a
i+1,1
(s)dX
s
+

0<st
Q
i+1
(S
s
, s)

i+1

k=2
_
t
0
a
i+1,k
(s)m
k
(s)ds.
We note that the processes [Y
(k)
, Y
(j)
]
t
and [H
(i+1)
, Y
(j)
]
t
have independent increments. Furthermore, we have for
t ∈ [0, T ] and j = 1, . . . , i:
E[H
(i+1)
, Y
(j)
]
t
= E
__
t
0
i

k=1
a
i+1,k
(s) d[Y
(k)
, Y
(j)
]
s
_
= ¸R
i
(x, s), 1{s < t ]x
j−1
) = 0.
42
QUANTI TATI VE FI NANCE Deterministic implied volatility models
Therefore H
(i+1)
is strongly orthogonal to all Y
(j)
for
j = 1, . . . , i. The martingales H
(j)
are consequently pairwise
strongly orthogonal. ¬.
Remark A.1. We observe that if the polynomial R
i
is such
that ¸R
i
, R
i
) = 0 then the martingale H
(i+1)
satisfies
E[H
(i+1)
, H
(i+1)
]
T
= 0 and thus H
(i+1)
= 0 almost
everywhere. If the discontinuous component of the regular
martingale is the sum of a finite number of Poisson processes
with deterministic intensity then only a finite number of the
martingales H
(k)
will be non-zero.
We define the following families of L
2
(, `) variables:

t
1
...t
n
= {X
k
1
t
1
(X
t
2
−X
t
1
)
k
2
. . . (X
t
n
−X
t
n−1
)
k
n
: k
i
0],
= {X
k
1
t
1
. . . (X
t
n
−X
t
n−1
)
k
n
: 0 t
i
< t
i+1
T, k
i
0].
Lemma A.2. The family
t
1
...t
n
is total in the space
L
2
(, σ(X
t
1
, X
t
2
−X
t
1
, . . . , X
t
n
−X
t
n−1
)).
Proof. Let us prove the result for the family
t
. The case with
n non-overlapping independent increments is treated similarly.
Since X is regular, the linear hull
ˆ

t
of
t
is in L
2
(P{X
t

dx]). The space C of functions in L
2
(P{X
t
∈ dx]) having
compact support is dense in L
2
(P{X
t
∈ dx]). Let F ∈ C
with support in [−a, a] and weakly orthogonal to
ˆ

t
, i.e.
E[(X
t
)
k
F(X
t
)] = 0, (k 0). For any real Z, we have:
+∞

k=0
[Z[
k
k!
E[[X
t
[
k
[F(X
t
)[] < E[F(X
t
)
2
]
1/2
exp([Za[).
By the dominated convergence theorem, we derive
E[exp(iZX
t
)F(X
t
)] = 0 for all Z and thus F(X
t
) = 0
(Malliavin and Airault (1994, p 110)). Therefore
ˆ

t
is dense
in C and thus in L
2
(P{X
t
∈ dx]). ¬.
Proposition A.2. The family is total in L
2
(, `).
Proof. A variable Z in L
2
(, `) can be approximated
arbitrarily closely by an element Y of L
2
(, σ(X
t
1
, X
t
2

X
t
1
, . . . , X
t
n
− X
t
n−1
)) for some sequence {t
i
]. Following
lemma A.2
t
1
...t
n
is total in L
2
(, σ(X
t
1
, . . . , X
t
n
− X
t
n−1
)).
Therefore, Y and thus Z can be approximated arbitrarily
closely by an element of
ˆ
. ¬.
We have the following representation property for
elements of .
Lemma A.3. For any integer k and any power-increment
(X
s
1
−X
s
0
)
k
, there is a sequence of predictable processes

(i)
s
0
s
1
k
(s) : i 1] such that:
(X
s
1
−X
s
0
)
k
= E[(X
s
1
−X
s
0
)
k
] +
+∞

i=1
_
s
1
s
0
θ
(i)
s
0
s
1
k
(s) dH
(i)
s
.
Proof. We prove the above equation by induction on k and by
application of Ito’s formula to power functions as in Nualart
and Schoutens (2000). ¬.
Proposition A.3. Let R ∈ ; then there are predictable
processes {
(i)
t
: i 1] such that:
R = E[R] +
+∞

i=1
_
T
0

(i)
s
dH
(i)
s
.
Proof. With lemma A.3, we derive that the product of non-
overlapping power-increments Y
kl
≡ ([X]
s
1
s
0
)
k
([X]
s
3
s
2
)
l
with
s
0
< s
1
s
2
< s
3
can be represented as follows:
Y
kl
= E[Y
kl
] +
+∞

i=1
_
T
0

(i)
s
dH
(i)
s
.
The process
(i)
s
is a predictable process defined by:

(i)
s
= 1{s ∈ (s
2
, s
3
)]θ
(i)
s
2
s
3
l
(s)
+∞

j=1
_
s
1
s
0
θ
(j)
s
0
s
1
k
(u) dH
(j)
u
.
Hence, we have proved the result for the product of two
non-overlapping power-increments. By induction, we prove
the result for the product of an arbitrary number of non-
overlapping power-increments. ¬.
Since the linear space
ˆ
spanned by is dense in
L
2
(, `), we deduce the following representation property
of square-integrable variables.
Proposition 6.1. Let F ∈ L
2
(, `) then there is a family of
predictable processes {φ
(i)
t
: i 1] such that:
F = E[F] +
+∞

i=1
_
T
0
φ
(i)
s
dH
(i)
s
,
where φ
(i)
t
belongs to L
2
H
(i)
.
Proof. For a square-integrable adapted martingale m, we recall
that:
L
2
m
=
_
H ∈ :
_
T
0
H
s
dm
s
∈ L
2
(, `)
_
,
where is the predictable σ-algebra on [0, T ] . We need
to prove that:
L
2
(, `) =
+∞

i=1
L
2
H
(i)
,
+∞

i=1
L
2
H
(i)

_
X ∈ L
2
(, `) : X=
+∞

i=1
φ
(i)
• H
(i)
, φ
(i)
∈ L
2
H
(i)
_
.
We observe that ⊕
+∞
i=1
L
2
H
(i)
is closed in L
2
(, `) since the
martingales H
(i)
are pairwise strongly orthogonal. Thanks to
proposition A.3, we have:
ˆ

+∞

i=1
L
2
H
(i)
⊂ L
2
(, `).
Since
¨
ˆ
= L
2
(, `) by proposition A.2, we derive the
result by closure of the above inclusion. ¬.
43
P Balland QUANTI TATI VE FI NANCE
References
Balland P and Hughston L P 1999 Sticky-delta model Derivatives
Week
Barndorff-Nielsen O E 1995 Normal inverse Gaussian distributions
and the modeling of stock returns Research Report no 300
Department of Theoretical Statistics, Aarhus University
Bjork T, Masi D G, Kabanov Y and Runggaldier W 1997 Towards a
general theory of bond markets Finance Stochastics 1 141–74
Black F and Scholes M 1973 The pricing of options and corporate
liabilities J. Political Economy 81 637–59
Bouchaud J P, Cont R and Potters M 1998 Financial markets as
adaptive systems Europhys. Lett. 41 3
Breeden D and Litzenberger R 1978 Prices of state-contingent
claims implicit in option prices J. Business 51 621–51
Brezis H 1992 Analyse Functionelle. Theorie et Applications (Paris:
Masson)
Carr P and Madan D 1998 Towards a theory of volatility trading
Volatility ed R A Jarrow (London: Risk) pp 417–27
Carr P and Madan D 1999 Option valuation using the fast Fourier
transform J. Comput. Finance 61–73
Cont R and da Fonseca J 2001 Deformation of implied volatility
surfaces: an empirical analysis Empirical Approaches to
Financial Fluctuations ed H Takayasu (Tokyo: Springer)
Derman E 1999 Regimes of volatility Risk 4 55–9
Derman E and Kani I 1994 Riding on a smile Risk 7 32–9
Derman E and Kani I 1998 Stochastic implied trees: Arbitrage
pricing with stochastic term and strike structure of volatility
Int. J. Theor. Appl. Finance 1 61–110
Doob J L 1994 Measure Theory (New York: Springer)
Dumas B, Fleming J and Whaley R 1998 Implied volatility
functions: empirical tests J. Finance 53 2059
Dupire B 1994 Pricing with a smile Risk 7 18–20
Eberlein E and Keller U 1995 Hyperbolic distributions in finance
Bernoulli 1 281–99
Heston S 1993 A closed form solution for options with stochastic
volatility with applications to bond and currency options Rev.
Financial Studies 6 327–43
Hull J C and White A 1987 The pricing of options with stochastic
volatilities J. Finance 42 281–300
Jacod J and Protter P 1991 Probability Essentials (Berlin: Springer)
Jacod J and Shiryaev A N 1987 Limit Theorems for Stochastic
Processes (Berlin: Springer)
Jarrow R A and Madan D 1999 Valuing and hedging contingent
claims on semimartingales Finance Stochastics 3 111–34
Koponen I 1995 Phys. Rev. E 52 1197
Lebedev V A 1995 Fubini-theorem for parameter-dependent
stochastic integrals with respect to L
0
-valued random measures
Probab. Theory Appl. 40 313–23
Levy P 1965 Processus Stochastiques et Mouvement Brownien
(Paris: Gauthier-Villars)
Madan D B and Seneta E 1990 The variance gamma model for share
market returns J. Business 63 511–24
Malliavin P and Airault H 1994 Integration et Analyse de Fourier
(Paris: Masson)
Mandelbrot B B 1963 The variation of certain speculative prices J.
Business 36 394–419
Mantegna R N and Stanley H E 1994 The truncated levy flights
Phys. Rev. Lett. 73 2946
Merton R 1976 Option pricing when underlying stocks returns are
discontinuous J. Financial Economics 3 125–44
Nualart D and Schoutens W 2000 Chaotic and predictable
representations for Levy processes Stochastic Processes
Applications 90 109–22
Protter P 1995 Stochastic Integration and Differential Equations
(Berlin: Springer)
Reiner E 1999 Volatility rules and implied processes Global
Derivative Conf.
Revuz D and Yor M 1991 Continuous Martingales and Brownian
Motion (Berlin: Springer)
Schonbucher P J 1999 A market model for stochastic implied
volatility Phil. Trans. R. Soc. A 357 2071–92
Schoutens W 2001 Meixner processes in finance EURANDOM
Report 2001-002
Schoutens W 2000 Stochastic Processes and Orthogonal
Polynomials (Lecture Notes in Statistics) vol 146 (Berlin:
Springer)
44

xm ) × (0. that a process St and a two-parameter family of implied volatility processes define an implied volatility model providing that this specification is arbitrage-free. Pt (x. we characterize two recently proposed deterministic implied volatility models. T ]}. To illustrate further this point. in this sense. K). K) → Ct (x. the model will still need re-calibration every time S changes! These frequent re-calibrations cause the hedging strategies implied by the model to be ‘inefficient’ as reported by Dumas et al (1998). We also assume that static portfolios consisting of a continuum of traded vanilla options are traded instruments as in Breeden and Litzenberger (1978) and Carr and Madan (1998). +∞) is of class C 1 (resp. complete. as in Schonbucher (1999) and in Cont and da Fonseca (2001). µ(s) ds ≡ mtx . K) from (0. K) = Ptx Et [(K − St+x )+ ]. Finally. We denote by Ftx the forward price at time t of receiving one unit of S at time t + x t. C 2 ) in the first (resp. 32 . (f ∈ C0 ). the implied volatilities are however functions of S. however. T ] with traded assets at time t. We use the terminology of implied volatility models instead of option price models because implied volatilities rather than option prices are the financial observable (see section 1). T ). This last dependence implies a greater stability of the corresponding hedging strategies. In this paper. 2. Hence. to be consistent with market smiles. It is important to stress. we show that the only arbitrage-free sticky-strike model is the Black–Scholes model. of one call and of one put option on S with maturity t + x ∈ (t. only short-dated options have liquid enough prices to be regarded as traded instruments. implied volatilities depend on t. Definition 2. consider the foreignexchange market. K). We denote by St > 0. It is clear. The market filtration { t : t ∈ [0. leaving thus no control on this dynamic. such a model over-estimates. Local volatility models are. the volatility of volatility and the correlation of volatility with S and. The dynamic of the implied volatility surface is entirely specified once the model has been calibrated with the initial smile surface. In typical foreign exchange markets for example. We write C0 for the space of continuous functions vanishing at zero and infinity. The option market and the implied volatility models We consider a continuous trading economy on a finite horizon [0. { t : t ∈ [0. K) ∈ (0. Ct (x.1. We assume that the map (x. Ct (x. Pt (x. = . St : (x. xm ) × [0. +∞) into (0. (ii) supx ∂x E[(St+x /Ftx )2 |St ] is locally bounded. only the vanilla options with maturity less than a couple of years have liquid prices. This restriction is fairly mild and met by most local volatility models. that these models need severe restrictions to be arbitrage-free. In a deterministic local volatility model such as the Dupire model. Schonbucher (1999) and Cont and da Fonseca (2001) for examples of stochastic implied volatility models. A family {Ct (x. However. The underlying S is typically traded more frequently than vanilla options and thus. K) > 0 and Pt (x. We now restrict our attention to two deterministic implied volatility models. second) variable for all t ∈ (0. we show that the underlying asset price is the exponential of a process with independent increments under the risk neutral measure and that any square-integrable claim can be replicated up to a vanishing risk by trading portfolios of vanilla options. P ) with a right-continuous. An implied volatility model is regular if (i) limx→0 E[f (St+x )|St = S] = f (S). In a stochastic local volatility model. These models are defined by direct assumptions on the stochastic evolution of the smile surface from an initial surface. the cost of hedging is not accurately accounted for. T increasing filtration with respect to which St and 1/St are cadlag quasi-left continuous square-integrable processes and sup{E[St2 + St−2 ] : 0 t T } < ∞. (iii) supt E[supx |∂x Et [(St+x /Ftx )±2 ]|] < ∞. We assume deterministic interest rates and we denote by Ptx the price at time t of a discount bond with maturity t + x and by Bt ≡ 1/P0t the money-market account. defined by assuming that either the per-delta or the per-strike implied volatility surface has a deterministic evolution.2. We shall need the following technical restriction to ensure that S has bounded local characteristics. We assume that there exists a stochastic basis ( . t + xm ). however. where implied volatilities are quoted per delta. The need for smile models that are calibrated to an initial smile surface and that imply a ‘realistic’ evolution of implied volatility explains the recent interest in the so-called implied volatility models. We assume no transaction costs. In arbitrage-free stickydelta models. over a short time interval during which implied volatilities do not change. as a consequence. unsatisfactory. the money market account B. the local volatility σt and on the percentage-in-the-money St /K. Definition 2. +∞)} satisfying the above assumptions is called an implied volatility model. We refer to Derman and Kani (1998). a financial asset S.P Balland Q UANTITATIVE F I N A N C E We assume that only short-dated vanilla options are traded assets because in typical option markets. Our definition of implied volatility models is based on option prices to ensure our implied volatility models are arbitrage-free. K) = Ptx Et [(St+x − K)+ ]. the implied volatilities do not change as frequently as S does. as previously explained. We assume that there exists a real bounded function µ(t) such that Ftx /St = exp t+x t It is important to understand that a complex option will be hedged efficiently using a smile model only if the model implies a dynamic of the smile that is sufficiently ‘realistic’ and ‘stationary’ for the model to not require frequent re-calibration. the call and the put options on S with strike K > 0 and maturity in (t. K) > 0 the price at time t of one unit of S. t + xm ) and with strike K > 0. T ]} is the filtration generated by all primary traded asset prices.

F. 2 (V . √ √ 1 d = ln(F /K)/ V + 2 V . the forward value of this option is c(t. K) = ∂F C BS (V . K) = ∂F C BS (V . K) 2 At a given time t. 33 . The sticky-delta and the sticky-strike implied volatility models The Black–Scholes function is denoted C BS (V . xm ) × (0. we have proved that for any risk-neutral measure Q and any s t. xm ) × (0. I ) where I stands for the forward moneyness ratio. K) = 2 F 2 BS (V . (V . An implied volatility model is sticky-delta if the per-delta volatility process σt (x. It thus follows that ln St has independent increment in [0. F. xm ) × (0. K) is strictly increasing with respect to V . Ftx . K) with √ C BS (V . (4. (4. 1). We denote BS 4. F. F. 1. Ss and Q. ln St is square-integrable. K). x.1. xm ) × (0. This value is deterministic. K) in (0. Finally. I ) → σt (x. time-to-maturity x and notional 1/Ftx . This option pays at maturity the quantity (St+x − kFtx )+ /Ftx . At time t. (4. F. Hence at time s t. Finally. k) and thus independent of s.2): Pr Q {St+x < kFtx | on s} = 1 + ∂k c(t. F. I ) = C BS (xσt (x. +∞). we propose the following definition of sticky-delta and sticky-strike implied volatility models. K) = σt (x. k).4) The cumulative distribution of ln St+x − ln St conditional t. The per-delta implied volatility denoted σt (x.Q UANTITATIVE F I N A N C E Deterministic implied volatility models 3. 1/St are square-integrable. 0 < r − u < xm and u − t < xm . x. x. Since the function C BS (V . K) → t (x. we have Q Es [(St+x − kFtx )+ /Ftx ] = c(t. T ] for all (x. we conclude that t (x. F. I )2 . +∞). I ) ∈ (0. At time t. We observe that the Black–Scholes model is an implied volatility model defined by σt (x. we define the following deterministic function: c(t. I ) are cadlag quasi-left continuous processes adapted to the market filtration. s is thus independent of s Let u1 = ln Sr − ln Ss and u2 = ln Su − ln St with s < r < t < u < T . Consider a forward-start call option with fixing date t and maturity t + x with x < xm . xm ) × (0. I ) and t (x. K) = σt (x. +∞). (4. (4. This result is extended by convolution to n arbitrary nonoverlapping increments of ln St in [0.2) Next. all risk-neutral probability measures are equal to the probability measure P on the σ -algebra T and the stochastic process ln St has independent increments under P .3) lim xσt (x. T ]. the implied volatility processes σt (x. T ] under any riskneutral measure. T ]. Characterization of the sticky-delta implied volatility models For a sticky-delta implied volatility model. T ] for all (x. K) = Ct (x.5) The variables u1 and u2 are thus independent and have the same joint distribution under any risk-neutral measure Q. For (x. BS We recall that 1 ∂V C BS (V . K) is deterministic on [0.1. K) = F N (d) − KN (d − V ). the volatility functions (x. An implied volatility model is sticky-strike if the per-strike volatility process t (x. We note that P is a risk-neutral measure. Let (x. this forward-start option is a regular call option on S with strike kFtx . I ) in (0. In a sticky-delta implied volatility model defined on [0. F. Definition 3. Equation (4. we have E Q [eiαu1 +iβu2 ] = E Q [eiαu1 ] E Q [eiβu2 ].4) implies that for any risk-neutral measure Q. x. F. K). Ftx . we have the following result. Ftx I ) = Ct (x. Theorem 4. +∞). k). K) are of class C 1 in the first variable and C 2 in the second variable in (0. Following Derman (1999) and Reiner (1999). x. Ftx I ). I ) is deterministic on [0. the market filtration { t : 0 t T } coincides with the filtration generated by S. I ) = t (x. k). 0 − 1 ∂k (St+x − kFtx )+ Ftx 1. BS 2 The per-strike implied volatility denoted unique positive solution of the equation Ptx C BS (x t (x. Proof.1) Since y 2 < 2(ey + e−y ) for all real y and St . +∞). K/Ftx ). K) 2 t (x. K). Definition 3. the cost of replicating this option is Ptx c(t. We can thus permute differentiation with respect to k and Q-expectation in (4. I ) ∈ (0. I ) and (x. is the unique positive solution of the equation Ptx C (xσt (x.2. I ) . K) is the . K). we note that = 0. I ). I )2 = lim x x↓0 x↓0 t (x.

(4. the martingale X has no fixed points of discontinuity and can be decomposed as in Jacod and Shiryaev (1987. .P Balland Q UANTITATIVE F I N A N C E t 0 2 σu du < ∞. K)2 ds (4. this process will also be denoted [Z]t or [Zt ]. . Examples of processes with independent increments are Brownian motions. . It follows that H C (s. x 2 nu (dx) du < ∞. any sequence {Ti } ∈ [0.10) sup {E[St | ln St |i ] + E[St−1 | ln St |i ]} < ∞.9) 0: (4. x + dx). T ] × . 34 . We have proved that in a sticky-delta implied volatility model. Since S is assumed quasi-left continuous. K) are in L2 and W C (s. The Lebesgue– Fubini theorem for positive integrand (Malliavin and Airault 1994. With equation (4. Poisson processes. Levy processes and jump-diffusion processes. V (t. K)(x) N (dx u To ease notation. imply completeness of the model since the number of traded assets is infinite (see Jarrow and Madan 1999). the stochastic integrals from a 0 to b are integrals on (a. du) − nu (dx) du}. measurable (see Protter 1995). . STn < Un }. b]. K) are squareintegrable. we conclude that for t in [0. there exists a cadlag square-integrable P -martingale X with P -independent increments such that St = S0 m0t exp(Xt )/E[exp(Xt )]. Since St and 1/St are assumed to be square-integrable. however. if finite.s−t.10).6) sK L2 = H ∈ W L2 = n ∈ ⊗B(R) : E :E 0 T 0 T 2 Hu du < ∞ . K) and recall that the quadratic variation of this process satisfies with equation (4.K ]u dK +∞ P (λK HutK )2 dK du +∞ 0 =E +E xN(dx × ds).7) (see Protter 1995) [V (t. p 77) Xt = (σ • W )t + (x ∗ (N − n))t . We refer to Protter (1995) for the construction of stochastic integrals with respect to cadlag square-integrable martingales.7) sK ˆ t = P0t Xt . 2 u (x) nu (dx) du < ∞ . Jensen inequality implies that the where X martingales P0t Ct (s − t. W is a Brownian motion. It follows that for all i t∈[0. We recall that the quadratic variation of a cadlag square-integrable martingale Z is an increasing cadlag adapted process defined by [Z. ( ∗ (N − n))t = 0 t u (x){N (dx P ˆ Pt. Xt = (x ∗ N )t = 0<s t 0 t HsP (t. s + ds) × (x. we obtain that for all t in [0. Restricting our attention to regular sticky-delta models will ensure that X has finite moments and satisfy the representation property of Nualart and Schoutens. But first. In this paper. P (s. Zu− dZu . Z]t = Zt2 − 2 t 0 We recall the following notation: (H • W )t = 0 t Hu dWu . K)2 + u 0 P 2 s (t. . we derive the following representation for the short-dated option prices: C ˆ Ct. we show by induction that for any integer n. p 32) implies that these probability measures coincide on σ (A) and thus on T . σ is a squareintegrable deterministic function. p 46) implies that the following Lebesgue integrals.T ] (4. A direct application of the monotone class theorem as in Jacod and Protter (1991. . Let Vu (t. T ] E 0<u t 2 Xu = t 0 +∞ −∞ 0 P 2 utK (x)) dK nu (dx) du . This will allow us to prove the quasi-completeness of regular sticky-delta models. . N is a Poisson measure independent of W with deterministic compensator n such that ns (dx) ds = E[N (dx × ds)] where N(dx × ds) is the number of non-zero jumps of X in (s. where is the predictable σ -algebra on [0. ω) < U1 . S(Tn . K). Finally. K) are in L2 with n By applying Ito’s formula (see Jacod and Shiryaev 1987) to St → E[(Ss − K)+ |St ]. Since ln St and thus Xt are square-integrable.8) + 0 × ds). . K). All risk-neutral probability measures Q coincide with P on the algebra A of cylindrical sets {ω ∈ : S(T1 . We note that the uniqueness of a risk-neutral measure does not. STn < Un } = P {ST1 < U1 . we prove the following lemma that will be useful in establishing our quasi-completeness result. T ]n and any real Ui : Q{ST1 < U1 . K) and P0t Pt (s − t.K = C0sK + (HsK • W )t + ( C ∗ (N − n))t (4. . K)]u = V0 (t. . K). (λK where Xt = Xt −Xt− is the jump of X at time t.s−t. we need some notation and some preliminary results. K) = P0u Pu (t − u. Let λ(K) be a square-integrable function. H P (s. satisfy E 0 +∞ 0 T 0 T 0 T λ(K)2 d[Vt. . T ] and all i 0 E[| ln St |i exp(| ln St |)] < i! E[St2 + St−2 ]. ω) < Un }. .K = P0sK + (HsK • W )t + ( P ∗ (N − n))t . that is the smallest σ -algebra making all adapted processes that are left continuous with right limits.

are assumed to be traded instruments. the Laplace transform of Xt is defined for all |θ | 2. ˆ ˆ ˆ du [Pu (t − u. we note that X is a Levy martingale and the condition (4. Since Pu (t − u. we derive inequality (4. we (4. +∞) and in (0. which are static portfolios having a continuum of traded vanilla options.t−u. we have E 0 t 5.K ]dK < ∞. M]. λC .K .t1 −s.13). K) is a P -martingale. M ≡ M/B for a cadlag process M. t ∈ [0. Since X has independent increments and E[exp(2Xt )] + E[exp(−2Xt )] < ∞.T ) σs2 + exp(λ|x|)ns (dx) < ∞.10). t1 (5. to regular martingales. 2: 1 −1 (4. A static strategy is characterized by an initial date t0 . static portfolios with a continuum of traded options.K ]pλ (dK) < ∞.13) (−1.12) ˆ (ln K)2p K −4 du [Pu. +∞) and maturity t1 and λP (K)pλ (dK) unit(s) of the put option with strike K ∈ (0.12). These static trading strategies are traded instruments by assumption (see section 2). As in the Black–Scholes theory. t1 . The martingale X is regular if there exists λ 2: sups∈(0. two locally bounded functions λC (K) and λP (K) defined in [α. t1 ] and that is instantiated at zero cost using the initial endowment at time t0 to purchase λC (K)cλ (dK) unit(s) of the call option with strike K ∈ (α. T ]. c. all European claims. we obtain (4.t−u. p. we then define trading strategies as dynamic portfolios involving a finite number of traded instruments. T ] and satisfies the time-dependent Levy–Khintchine formula for square-integrable processes: E[exp(θ (Xt+x − Xt ))] = exp ψs (θ ) = 1 2 2 σ θ + 2 s +∞ −∞ t+x t ψs (θ ) ds .K ]dK < ∞. we show that the martingale X for a regular sticky-delta model as in definition 2.1)c exp(λ|x|)ns (dx)+ +∞ x ns (dx) 2 Therefore for k.t−u.K ]cλ (dK) < ∞ 35 .2 is regular in the following sense. With this representation property. an end date t1 ∈ (t0 .1. λP . We first define the static trading strategies.t1 −s. For any integers p have E 0 α 0 +∞ α 0 t t Deterministic implied volatility models 0 and any real α > 0. an initial endowment C0 ∈ t0 . α] respectively. α) is a portfolio that has constant holdings in (t0 . t + xm ) and payoffs f (Su ). K)] = E[Pt (0+. we define the trading strategies that will be used to replicate contingent claims in a regular sticky-delta model. ] K4 (ln K)2p K −4 E[(K − St )+2 ]. t0 + xm ). we will extend in section 6.e. With that restriction. K)2 ] − P0 (t. K)2 . The static trading strategy λ = (t0 . Similarly.3) E +∞ α t0 ˆ λC (K)2 ds [Cs. But first we need to define precisely our admissible trading strategies and what we mean by quasi-completeness. E 0 α ˆ λP (K)2 ds [Ps. the polynomial representation obtained by Nualart and Schoutens for regular Levy martingales. with maturity u ∈ (t. as this type of completeness is not standard. static trading strategies.11). (eθz − 1 − θ z) ns (dz). we derive for i +∞ −∞ ˆ |λP |P0t1 K pλ (dK) + K t1 t0 +∞ α ˆ |λC |C0t1 K cλ (dK) < ∞. Trading strategies. we observe that: E 0 t In the case where the increments of X have stationary distributions. i.Q UANTITATIVE F I N A N C E Lemma 4. E ˆ where [M] ≡ [M.1)c Thanks to (4.1.11) (4. i 2. are traded instruments. We assume that the functions λC (K) and λP (K) satisfy the following conditions: E 0 α (ln K)2p ˆ du [Pu. t1 ] ⊂ (t0 .K ] < ∞ By Lebesgue–Fubini. We highlight that in this paper. Using the above formula.t−u. attainable claims and quasi-completeness In this section. we conclude that α 0 E[| ln St |∨| ln α|2p St−1 ] E 0 t ˆ (ln K)2p K −4 du [Pu.13) is similar to the condition introduced by Nualart and Schoutens (2000). α) and maturity t1 .2) (5. ˆ Proof.1) i! |x| ns (dx) < i λ i (−1. we will derive our quasi-completeness result in section 7. ˆ (ln K)2p K −4 du [Cu. We note that this assumption is equivalent to assuming that at time t. Definition 4. we obtain α 0 (ln K)2p K −4 E[(K −St )+2 ]dK With (4. mi (t) ≡ −∞ x i nt (dx) and t k Mk (t) ≡ E 0<s t ( Xs ) = 0 mk (s) ds are uniformly bounded in [0. as in Carr and Madan (1998). t0 + xm ) and there are no flows coming in or out of the strategy up to time t1 . Therefore. that can be decomposed in a continuum of call and put option payoffs as in Breeden and Litzenberger (1978). Using the Lebesgue–Fubini theorem for a positive integrand. We define a static trading strategy λ to be a portfolio of short-dated options with maturity t1 and of the money market account such that the portfolio holdings are constant over the trading interval (t0 . and two positive σ -finite measures cλ (dK) and pλ (dK) defined on the positive half line. K (5.

As we will see. Using (5. A claim ∈ L2 ( .T ) sup E n i=1 ˆ λi V ( t B i ) + λt 2 < ∞. P (VT (λ) > 0) > 0. ) with maturity t T is attainable or can be replicated if there is a self-financing trading strategy λ such that = Vt (λ). K). Hence the discounted value process of a self-financing trading strategy is a square-integrable martingale that satisfies ˆ Vt (λ) = V0 (λ) + n i=1 0 t ˆ ds Ps (t1 − s. that equations (5. Suppose that λ is an arbitrage strategy then E[1{VT (λ) < 0}] = 0 and thus. A trading strategy is self-financing if and only if the value process satisfies as usual Vt (λ) = V0 (λ) + Gt (λ).1). A trading strategy λ is defined as a portfolio of n static trading strategies and of the money market account. T ]. i. V ( ˆ j )]s < ∞. p 160).5). K)pλ (dK) Equation (5. This discounted value process is a square-integrable cadlag martingale which is zero before t0 and constant after t1 .4) 36 . ˆt (λ) is a P -martingale.e. t1 ] is obtained by adding the values of its constituents: Vt (λ) = + α α 0 +∞ At time t. (5.4) guarantees that this process is a squareintegrable semimartingale. The value of the static trading strategy at time t ∈ (t0 .1. (5.e. t1 ): ˆ Vt (λ) = + α α 0 +∞ λP (K) t t0 t t0 Equation (5.1)–(5.6) λP HtPK pλ (dK) + K 1 +∞ +∞ α λC HtCK cλ (dK) •W K 1 ∗ (N − n) t t0 t t0 + 0 λP K P t1 K pλ (dK) + α λC K C t1 K cλ (dK) where we have used the notation [X]b = Xb − Xa . The value of such a strategy is given at time t by the sum of the values of its constituents: Vt (λ) = n i=1 λi Vt ( t i) + λ B Bt . the strategy λ is the portfolio consisting of λi t unit(s) of each of the static portfolios i and of λB unit(s) of t the money market account. A claim ∈ L2 ( . so as to permute strike and time integrations.1. ∈ V ( ). Implied volatility models are arbitrage-free. we have E[VT (λ) 1{VT (λ) < ˆ ˆ ˆ 0}] = 0. In short a and with abuse of notations.2) and (5. As explained in Jarrow and Madan (1999).3) in order to apply the stochastic Fubini theorem as in Protter (1995.2) and (5. i ) where λi is a predictable process t t satisfying (5. s The above two integrals are guaranteed to exist thanks to (5. this will allow us to define dynamic portfolios based on static trading strategies. The gain cumulated up to time t by the trading strategy is defined as usual by Gt (λ) ≡ n i=1 0 t λC (K)Ct (t1 − t. ˆ λP (K)pλ (dK) ds P (t1 − s.3) are implied by the single equation E α St1 ∧α E 0 T ˆ λi λj d[V ( s s i ). Proof. an arbitrage strategy is a self-financing strategy λ such that V0 (λ) = 0. by application of the stochastic Fubini theorem as formulated in Protter (1995) for martingales and in Lebedev (1995) for random measures: ˆ Vt (λ) = 0 α α ˆ λi d V s ( s i ). we show that we can permute strike and time integration. T ]. Proposition 5. we show as in the proof of lemma 4. K)cλ (dK) + C0 Bt /Bt0 . We impose the holding functions to satisfy (5. Since the static portfolio does not involve any cost at initiation.9).1. We define the following linear subspace of L2 ( .6)–(4.5) guarantees the existence of the above integrals. We denote by the space of all selffinancing trading strategies. X]. This contradicts the fact that V Remark 5. Using the Cauchy–Schwartz inequality. These static strategies which are portfolios with a continuum of traded options. λ ∈ }. we simply write ˆ Vt (λ) = t t t0 α 0 +∞ α A self-financing trading strategy is thus entirely characterized by λ = (λi .3) together with (4. ∈ V ( ). ) with maturity t T is attainable up to a vanishing risk or can be replicated up to a vanishing risk if there is a sequence of self-financing strategies {λn } such ¯ that limn→∞ E[( − Vt (λn ))2 ] = 0. + t0 ˆ ˆ We set Vt (λ) = V(t∨t0 )∧t1 (λ) for t ∈ [0.4) and (5.P Balland Q UANTITATIVE F I N A N C E n i. This result is not surprising since we have assumed the existence of at least one risk-neutral measure P . are traded instruments by assumption. ). t λP (K)Pt (t1 − t. K) pλ (dK) λC (K) ˆ ds Cs (t1 − s.5) (λP K)2 pλ (dK) K St1 ∨α + α (λC St1 )2 cλ (dK) K < ∞.j =1 where we recall [X] = [X. (5. A trading strategy is thus characterized by a finite sequence { i : 1 i n} of static trading strategies and by some predictable processes {λi : 1 i n} and by a progressively measurable t adapted process λB such that t t∈(0. i. K) ˆ λC (K)cλ (dK) ds C(t1 − s. K) cλ (dK).1. V ( ) = {Vt (λ) : t ∈ [0. λi dVs ( s i) + 0 t λB dBs . Definition 5. We can now define attainability and attainability up to a vanishing risk. It follows that E[VT (λ)] = E[|VT (λ)|] > 0 = V0 (λ). P (VT (λ) 0) = 1. we derive for t in (t0 .

At time tk = kxm /2. The fourth component is replicated by buying a call option and selling a put option since it is the terminal value of a forward contract. We conclude this section with a first illustration of how hedging in a sticky-delta model works in practice (see remark 7. T ] are attainable. we define quasi-completeness as in Jarrow and Madan (1999) and in Bjork et al (1997). s (K − Sb )+ f (K) dK Since γt is a predictable bounded process. we decompose the payoff into four components f (Sb ) = 0 +∞ α ˆ λi dVs (φi ). +∞) function with f (Sb ) ∈ L2 ( . s We say that a claim ∈ L2 ( . Proof. +∞)2 and the Fubini– Lebesgue theorem implies (Malliavin and Airault (1993. (5. dK. we conclude that fn (Sb ) is attainable and thus at time t. An implied volatility model is complete up to a vanishing risk or quasi-complete if any ∈ L2 ( . With the above notations. z). 37 . f . at time s. Following Carr and Madan (1998). ci . ) − V ( ). Given the set of dates {ti = (ixm /2) ∧ t : 1 i k + 1}.4. Definition 5. By application of proposition 5. (γs λi . Proposition 5. K). Similarly.3. The discounted value process associated with this static P0bK |f (K)|dK + +∞ α C0bK |f (K)|dK < ∞. ). Za = 0 γt dUt ∈ L2 ( . ) can be replicated up to a vanishing risk at a cost Ptb Et [f (Sb )]. b. can be replicated at zero cost. ) with maturity t is attainable at time s if there exists a self-financing strategy λ ˆ ˆ and Cs ∈ s such that P0t = Vt (λ) − Vs (λ) + Cs . By applying the Lebesgue dominated convergence theorem. Proposition 5. Consider the long-dated call option with strike K > 0 and maturity t = kxm /2 + x in [0. φi ) replicates the long-dated call option. z) + t ˆ f (z) dz ds Cs (b − s. f . We now prove the following classic result to be used later. t +f (α)(Sb − α) ≡ A + B + C + D. φi ).5) and a real t V0 = E[P0t ] which is the cost of replication. All call and put options with maturity in [0. ti+1 ] such that (1. ) with maturity t is thus attainable if there are some static trading strategies {φi = (t0i . α): Ptb (A + B) = Ptb Et [A + B] + b +∞ α b t 0 α ˆ f (z) dz ds Ps (b − s. the claim f (Sb ) ∈ L2 ( . xm ). z) dz.8) f (K) by f (K)1{|f (K)| < n}. αi ) : 0 i n}. (5. Proof. (5. such that P0t = V0 + n n t1i t0i 0 αi i=0 t1i +∞ t0i αi Deterministic implied volatility models The third component is trivially replicated by taking position in the money market account.7) Then the claim with payoff f (Sb ) at time b can be replicated at time t at a cost Ptb Et [f (Sb )]. T ] where k is an integer and 0 < x xm /2. Let Ut be a square-integrable martingale such that the claim UT with maturity T is attainable and let γt be a predictable bounded process.Q UANTITATIVE F I N A N C E A claim ∈ L2 ( .1. K) s + i=0 ˆ λi fi (K)ci (dK) ds Cs (t1i − s. Proof. φiC . z) dz+ +∞ α f (z)Ct (a. L2 ( . ) is ¯ attainable up to a vanishing risk i. p 46)) Ptb Et [A+B] = 0 α f (z)Pt (a. pi . φiP . we shall construct a sequence of static trading strategies φi having trading intervals (ti . The claim with payoff UT /P0T is attained by a self-financing trading strategy λ = (λi .2. Corollary 5. Let a ∈ (0. ) such that there is 0 < α < ∞ satisfying E α Sb ∧α f (K)2 K 2 dK + Sb ∨α α 2 f (K)2 Sb dK < ∞. the call option is a short-dated option and thus can be replicated by purchasing the call option itself.2). We note that s ˆ Vt (λ) = Et [UT ] = Ut satisfies Ut = U0 + n i=1 0 t ˆ λi gi (K)pi (dK) ds Ps (t1i − s.4) and (5.e.9) |f (K)|(K − S)+ 1{K < α} + |f (K)|(S − K)+ 1{K > α} is thus P {Sb ∈ dS}×dK integrable on (0. Proposition 5.2 to fn (Sb ). Thanks to our assumption on f . some predictable processes {λi : 0 i n} satisfying (5. we derive by Cauchy– Schwartz: α 0 Therefore.8) Thanks to our assumptions on f . b = t + a ∈ [0. we conclude that the claim A + B is square-integrable and attainable at t by the static trading strategy (t. T ] and let f be a C 2 (0. Proof. Consider the sequence of smooth functions fn defined by replacing in (5. φi ) s defines a self-financing trading strategy and we have a 0 + α (Sb − K)+ f (K) dK + f (α) γt dUt = n i=1 0 a ˆ γt λi dVt (φi ).2. we define attainable up to a vanishing risk. The above result still holds if we replace in (5. the claim Za with maturity a is attainable at zero cost. dK.7) the squares by absolute values and replicated by replicating up to a vanishing risk. t1i . we obtain limn→∞ E[(fn − f )2 ] = 0. Then the claim with a payoff at time a.

) then there are predictable processes {φt(i) : i 1} such that +∞ E E 1 Sk ∧1 Sk ∨1 1 2 ∂y c( tk . we show that (i) the claims HT are attainable by application of Ito’s lemma to polynomial functions. + ai. ) 2 ˆ 2 2 3 and ∂St Vtk = K ∂y c( tk . +∞).1 (s) dYs(1) . . we derive the result. . F = E[F ] + i=1 0 T (i) φs dHs(i) . 38 . K/Sk ) K 2 6 6 /Sk dK < ∞. y) = E[(St /St−x − y)+ ] is a C 2 function which is decreasing and convex with respect to the second argument. By repeating the above argument. we obtain ˆ Vtk = P0t E[(St − K)+ |Stk ] = P0t Stk c(t − tk . But first. we prove that all regular sticky-delta implied volatility models are quasi-complete. Let F ∈ L2 ( .1 has in general an infinite dimension and thus exact replication would be possible only if strategies based on an infinite number of traded instruments were admissible. to regular martingales with independent increments. More {H (i) : i precisely. 6. t]. Yt(k) ≡ ( X s )k − 0<s t 0 t 7. x (i) (K) = {i(i − 1)(ln{K/f0b })i−2 − i(ln{K/f0b })i−1 }/P0b . Representation property for regular martingales with independent increments In this section and in the appendix. We define f0b ≡ F0b /E[exp(Xb )] and observe that Xb = ln(Sb /f0b ). we construct the following family of pairwise [ ]orthogonal martingales having independent increments: Ht(1) = Yt(1) . where c(x. all square integrable claims can be replicated up to a vanishing residual risk by trading portfolios of vanilla options.2 with f (S) = ln(S/f0b )i . Using this representation property. We define the Teugels martingales on [0. ). See appendix. K). T ] and are such that the martingales Ht(i) are square-integrable and pairwise strongly orthogonal.2 and replicate the claim ˆ Vtk between time tk−1 and tk using a static portfolio with positive holdings in short-dated call and put options having maturity tk . Proof. K/Sk )2 K 4 /Sk dK < ∞. In the appendix and by [ ]-orthogonalization of the Teugels martingales. K) mk (s) ds. The deterministic functions aij (s) are bounded in [0. We obtain the result for the long-dated put options by put– call parity. Using equation (4. we obtain a set of static trading strategy {φi : 0 i k} with positive holdings such that the self-financing trading strategy (1. K/Stk )/Stk 0. It is clear that these martingales are square-integrable with independent increments and finite moments of all orders. The claim with payoff (Xb )i at time b is attainable at time a > b − xm using the static strategy: i Xb = Ea [(Xb )i ] + b a 0 f0b ˆ x (i) (K)K −2 dKdu Pu (b − u. we have the following representation property. Given the stickydelta assumption. Proposition 6. We note that V ∈ L2 ( . K/Stk ). K) for u ∈ [tk .1. Let i 2. We can thus apply proposition 5.10) and proposition 5. We derive k dYs(i) + ai. i=1 H Proof. b +∞ f0b + a ˆ x (i) (K)K −2 dK du Cu (b − u. K/Stk−1 ). we extend the representation property obtained by Nualart and Schoutens (2000) for Levy processes. at a discounted cost ˆ Vtk−1 = P0t E[(St − K)+ |Stk−1 ] = P0t Stk−1 c(t − tk−1 . In order to construct similar hedging strategies for squareintegrable claims.i−1 (s) dYs(i−1) + . in the classic sense. We adapt the notation and the approach taken by Nualart and Schoutens to our purpose.1. φi ) replicates the long-dated call option at the following initial cost: ˆ V0 = P0t E[(St − K)+ ]. (k 2). where {φt(i) : i 1} belongs to ⊕+∞ L2 (i) . 2 2 The density of St /St−x is ∂y c(x. y) and y 2 ∂y c(x. We show in the appendix that the [ ]-orthogonal family 1} forms a complete basis of L2 ( . 2 4 ∂y c( tk . contingent claims because the family of martingales underlying the representation property of proposition 6.P Balland Q UANTITATIVE F I N A N C E Ht(i) = t 0 ˆ ˆ strategy is Vu = Ck (t − u. Lemma 7. y) is ˆ uniformly bounded in (0. T ] as in Nualart and Schoutens (2000): Yt(1) = Xt . we prove in the next section that in regular sticky-delta models. We cannot typically replicate. we need a representation property for X. Quasi-completeness of regular sticky-delta implied volatility models In this section.

. + a (i) ˆ gu du Cu (b − u. f0b ) + a (3) ˆ fu dCu (b − u. K) b a (3) ˆ fu dPu (b − u. − f (Xu− ) dXu − 1 2 2 f (Xu− )σu du. = 0 k i−1 k k gu. we obtain by applying Ito’s lemma to f (x) = x 3 : ( X u )3 − 0<u b 0 b 0 (2) Xu− dYu − 3 b 0 (2) 2 (2) (1) (Xu− + Vb − Vu + Mb − Mu ) dYu t 0 b ˆ h(i) (K)K −2 dKdu Pu (b − u. tl = (lxm /2) ∧ T . T ]. the claim HT is attainable. T ]. f0b ).k X Y i (k) i and Ru (X) = 0 i k ru.i are deterministic and uniformly (i) bounded on [0. . For all i 1. . f0b ). H We define n = E[ ] + (i) with ξt(i. (k) By induction. HT ∈ V ( ). + a (1) ˆ fu du Cu (b − u. K) b a (2) fu = E[ ] + i=1 0 T ξs(i) dHs(i) . + a ˆ dCu (b − u.i Xu− (ln(K/f0b ))k . 0b With Ito’s formula for a real function f of class C 2 : { f (Xu ) − f (Xu− ) Xu } = f (Xt ) − f (X0 ) 0<u t t 0 t 0 where the coefficients are deterministic and bounded in [0. A regular sticky-delta model is quasi-complete or complete up to a vanishing risk. + ak. K) u −3 m3 (s) ds = (Xb )3 − E[(Xb )3 ] + a ˆ h(i) (K)K −2 dK du Cu (b − u. ) can be represented as follows: +∞ + a (2) ˆ yu (K)K −2 dK du Cu (b − u. ˆ Ct (b − t. K)K −2 dK ˆ ˆ +(Ct (b − t.1 imply that (2) (2) Yb − Ya is attainable: (2) (2) Yb − Ya = b +∞ f0b b (2) fu b f0b a 0 (2) ˆ yu (K)K −2 dK du Pu (b − u.e. f0b ) − (1) (1) where yu (K) = − P1 and fu = P0b1f0b . Finally. f0b ) − Pt (b − t. we obtain a self(k) financing trading strategy that replicates HT . Taking f (x) = x 2 . the random variable ∈ L2 ( . a + xm ). we have the following quasi-completeness result. According to proposition 6. K) (k) Ya . (i) i.i and gu. f0b ) − Proof. T ]. f0b ) − h(i) (K) u (i) gu = 0 j k i−1 jk j hu.i Xu− . (k) (k) We finally obtain the promised expression for Hb − Ha (k) (k) by using the previous equation for Yb − Ya . b = tl with l = 1.2.1. Finally.Q UANTITATIVE F I N A N C E Proposition 7. ln(K/f0b )). K) u b a (i) ˆ gu du Pu (b − u. Hb − Ha Deterministic implied volatility models Similarly. [2T /xm ] + 1. By (1) proposition 5.k−1 (u) dYu + . Any square-integrable claim ∈ L2 ( . T n i=1 0 ξs(i. f0b ) − ˆ dPu (b − u. K) Proof. K)K −2 dK (3) (2) (3) (2) where yu (K) = Qu (Xu− . . + a (3) ˆ yu (K)K −2 dK du Cu (b − u. the above equation and lemma 7.n) = ξy 1{|ξt(i) | < n}. K) t a (1) ˆ fu du Pu (b − u. This equation implies for t ∈ [a. [H (k) ]b = a + a (1) ˆ yu (K)K −2 dK du Cu (b − u. fu = Ru (Xu− ) ij i j and the polynomials Q(k) (X.1 (u)dYu . By adding the above decompositions obtained for a = tl−1 . we extend the above formula to all Yb − We recall that: b a (k) (k−1) (1) dYu + ak.n) dHs(i) ∈ V ( ) 39 . K) k The coefficients hu. For i satisfies (i) (i) Hb − H a = b +∞ f0b b b a 0 f0b (i) (i) 1 and b ∈ (a.k X have coefficients that are deterministic and uniformly bounded in [0. f0b ))/f0b . . Yt(1) = ln(St /f0t ) is a square-integrable martingale. Y ) = u 0 i j k qu. Theorem 7. where Vt = 0 σs2 ds and Mt(2) = It follows that: (3) (3) Yb − Ya = b +∞ f0b b b a 0 f0b t m2 (s) ds.1. we conclude that Yb is attainable using the static trading strategy defined by P0b Yt(1) = − − +∞ f0b f0b 0 ˆ Pt (b − t. jk (3) ˆ yu (K)K −2 dK du Pu (b − u. f0b ) (2) yu (K) = 2{Xu− − ln(K/f0b ) + 1}/P0b (2) (1) fu = −2Xu− fu where ξt(i) is in L2 (i) . we derive: (2) Yb = (Xb )2 − E[(Xb )2 ] − 2 b 0 (1) Xu− dYu . . the money-market account and some portfolios of traded call and put options. b] (1) Yt(1) − Ya = t +∞ f0b t t f0b a 0 (1) ˆ yu (K)K −2 dK du Pu (b − u. ) with maturity T can be replicated up to a vanishing risk at a cost P0T E[ ] by trading the underlying.1.

2. where the linear operator ℘tx is defined by [℘tx f ](S) ≡ 0 +∞ 2 f (I × Smtx )∂I C BS (xσt (x. There is a long list of geometric Levy models proposed as alternatives to the Black–Scholes model.2. . the money-market account and some portfolios of traded call and put options. a fast Fourier transform as in Carr and Madan (1999) or by solving the integro-differential equation ∂t V + At V = rt V where At is the generator defined by At = (∂x ℘tx )x=0+ . a forward contract and a zero coupon bond as in proposition 5. At time Tn−2 .P Balland Q UANTITATIVE F I N A N C E zero coupon bonds with maturity Tn−1 . The first sequence on the RHS converges to zero since {ξt(i) : i 1} belongs to ⊕+∞ L2 (j ) . Mandelbrot (1963). . . . Therefore. We assume that this payoff is such that STi → E[f (ST0 . by rolling a portfolio of vanilla options.1. .n) }2 d[H (i) . Consider a claim with maturity Tn such as an Asian option. we show that regular geometric Levy models are regular stationary sticky-delta implied volatility models. Our previous analysis in section 4 shows that the increments of ln St are independent and have a stationary distribution under the probability measure P . STn−1 .1 and 5.n) )2 d[H .2. Observe next that: j =1 H n→∞ lim (ξs(i.2. Consider a claim with square-integrable payoff f (ST0 . we derive by induction that the complex option can be replicated at a cost V0 = P0Tn ℘0.e. STn )| Ti−1 ∨ STi ] satisfies. . . K) = Ptx Et [(St+x − K)+ ]. Pt (x. Tn−1 f (ST0 .1. STn ) at time Tn . we have limn→∞ E[( − n )2 ] = 0. we obtain a super-replication strategy for the complex option when there are transaction costs on the implied volatility or when the short-dated implied volatility smile is ‘uncertain’ but bounded. equation (5. 1. × [℘Tn−1 . lim E i=1 0 T (ξs(i) − ξs(i. •)](STn−1 ). Then the claim can be replicated by trading the underlying. The next proposition shows that some square-integrable claims can be replicated in the classical sense. Observe that the transition operators ℘Tk . . the continuous-time formulation by Koponen (1995) of the ‘truncated Levy flight’ introduced by Mantegna and Stanley (1994). we obtain E[( − +E i=1 0 n) +∞ 2 ] T E − E[ ] − n i=1 0 T ξs(i) dHs(i) 2 ℘T i . Tn−1 is deterministic and thus VTn−1 is a deterministic function of ST0 . . {ξs(i) − ξs(i. St = F0t exp(Lt − ln E[eLt ]). The sticky-delta assumption implies that the operator ℘Tn−1 . where x < xm and L is an adapted regular P -Levy martingale. . Regular geometric Levy models We consider a regular sticky-delta implied volatility model and we suppose that the per-delta implied volatility processes are independent of time. Finally. .1. H ]s = 0.T0 n−1 i=0 Thanks to the strong orthogonality of Ht(i) . At time Tn−1 .n) − ξs(i) )2 = 0. Cont and Potters (1998). H (i) ]s . . The previous expectation can be estimated by using the Monte Carlo method. (ξs(i. The concept of quasi-completeness is due to Jarrow and Madan (1999) and to Bjork et al (1997). that has a finite number of fixing dates Ti with 0 < Ti −Ti−1 < xm and a square-integrable payoff f (ST0 . . By modifying the above strategy. Tk are nonlinear in this case. (i) (i) Finally. Bjork et al (1997) and Schoutens (2001). Regular geometric Levy models are arbitrage-free and quasi-complete in the sense that all square-integrable claims can be replicated up to a vanishing risk by trading the underlying. STn ) as in proposition 7. By three applications of the dominated convergence Lebesgue theorem. Ct (x. these models are arbitrage-free and quasi-complete by application of theorem 7. . We mention. STn−1 . is thus VTn−1 = PTn−1 . . I )2 . Bouchaud. the complex option can thus be replicated by the use of calls. Proof. forwards and 40 . we derive +∞ n→∞ Remark 7. Madan and Seneta (1990). where L is a regular P -Levy martingale (see Levy 1965).7) with b = Ti . . a discrete barrier option. a Parisian option or a volatility swap. Proposition 8. We define a regular geometric Levy under P by: St = F0t exp(Lt − ln E[eLt ]). Each claim n is attainable by application of propositions 7. puts. Proposition 7. By direct calculation. . the payoff can be replicated using a portfolio of call and put options. Proof. Tn−1 8. . BarndorffNielsen (1995). Eberlein and Keller (1995).n) − ξs(i) )2 4(ξs(i) )2 . Remark 7. The value VTn−1 of the complex option at time Tn−1 . i. Merton (1976). I ) dI . . Ti (f )(S0 ) = P0Tn E[f ].3 at a cost P0T E[ ]. K) = Ptx Et [(K − St+x )+ ]. the money-market account and portfolios of short-dated call and put options. . after subtraction of a finite number of call and put payoffs. in particular.

a family of positive P -integrable functions hI¯k . b > 0). we assume that pt > 0 in (0. Using the Black–Scholes equation. we note by Jensen inequality: Et [(St+x /mtx − K)+ ] = Et [(St+x+y /mt. the asset price process St is a Markov process entirely characterized by the transition function ℘ tx (see Revuz and Yor 1991): (℘ tx f )(St ) = E[f (St+x )|St ]. the density of St by pt and tx (S.σ (n) with limit zero. K) ht (S) = k=1 1Ik (S)hI¯k . Characterization of sticky-strike implied volatility models As with sticky-delta models. P integrable function: +∞ ¯ For I ⊂ (0.). P a. locally bounded.1. k ∪ [k. such that (see Doob 1994): n→+∞ lim ∂x vtxI. The implied volatility model with zero-drift underlying St /m0t and implied volatility t (x. +∞): n→+∞ I With our assumptions. For each t < T . we finally conclude that: n→+∞ Definition 2. Kmtx )2 . K) ≡ BS (x t (x.s. C: n→+∞ C We observe that for any positive z: HI tx (z) 2 2 E[sup ∂x Et [St+x /Ftx ]]. x lim E[ txφ(n) (St . K)1A (St )]|∂x vtxφ(n) − ht |dK = 0. m0t = 1.t and sub-sequences xσk (n) ⊂ xσk−1 (n) converging to zero such that: n→+∞ lim ∂x vtxσk (n) (St ) = hI¯k .t |dK = 0. lim ∂x vtxk (St ) = ht (St ) P a.σ (n) (St ) = hI. locally bounded. St ∈ Ik .xI.t (z)pt (z)dz = z] (a. x I × 1{St < z}dK. Using equation (9.t (a) = 0 hI. Km0t ) is regular and sticky-strike. n→+∞ I lim ∂x vtxI. K)dK.n − hI. K). P -integrable function hI.e. (9. P -integrable function ht such that: k→+∞ Since E[ t.2 implies that the increasing functions HI tx are uniformly bounded with respect to x and I . we define the sequence xφ(n) = xσn (n) with limit zero. Lemma 9. P a.1.n (St . Hence the Black–Scholes model is the only arbitrage-free regular sticky-strike model.t (S) 2 2 sup ∂x E[St+x /Ftx |St = S].n 1A (K)pt (K)dK = A Hence ∂x vtxI.2) Since k Ik = (0. there exists a positive. In a regular sticky-strike implied volatility model. and the positive.. K)f (K)dK. P a.t pt dz.s. a < b: The increasing function HI. K)1A (St )]dK = A hI. 0<x<T −t 2 2 ∂x E[St+x /Ftx |St Finally. we have a stronger result. we derive: 2 2 ∂x Et [St+x /Ftx ] = +∞ 0 lim E[ txI.Q UANTITATIVE F I N A N C E Deterministic implied volatility models 9.t (St ) (St ∈ I. Proof. Application of Fatou’s theorem gives (see Doob 1994): n→+∞ I where xk is a sequence with limit zero.t such that (see Doob 1994): n→+∞ lim ∂x vtxφ(n) (St ) = ht (St ) P a. We denote vtx (K) ≡ x t (x.t such that: HI.s. Theorem 9. the per-strike implied volatility t (x.t (z). Therefore. Proof.n 1A ]/pt converges uniformly to 1A on I : hI. +∞). Helly’s theorem implies that there is a sequence xI.n (St ) converges in probability to hI. Hence vtx (K) is increasing with x.t pt dz.x+y − K)+ ]. we have: Ptx EtP [f (St+x )] = +∞ 0 2 ∂K Ct (x. ∂x vtx (K) tx (St . For any bounded. Borel measurable function f . we obtain for any compact sets A.t satisfies for 0 HI.t (b) − HI. we conclude that for any compact A ⊂ (0.n E[ txI.1).n with limit zero and an increasing function HI. First.). K) is independent of K. we consider only regular stickystrike models.s.t (b) − HI. we define the increasing function: HI tx : z → ∂x vtx (K)E[ tx (St . To simplify notation. as previously. +∞). +∞). lim ∂x vtxI. (9.t (a) b a 2 2 sup ∂x E[St+x /Ftx |St ]pt (St )dSt . lim HI txI n (z) = HI.s. there is no loss of generality in assuming zero drift i. S.t (z) sup b a hI. x The Radon–Nikodym theorem implies that there is a positive. Using the diagonal procedure. In fact.t (St ) (k > 0. locally bounded. K)1A (St )] × |∂x vtxI. The case where pt vanishes is treated similarly since we have P (pt (St ) = 0) = 0.t (St ) in I and there exists xI.1) 1 1 We define Ik = k+1 . 41 . k + 1) and construct.n (St . Kmtx )2 .

K) = 2 F 2 BS Appendix. [At f ](St ) = lim ∂x [℘ tx f ](St ). 0 k=1 j −1 42 .4) By -orthogonalization of the total family {1{s < t}x i : i 0. St . K) × ∂x vtx (K)f (K)dK.3) can be written as follows: ∂x [℘ tx f ](St ) = 1 2 St2 tx (St . the implied volatility t (x. locally bounded P -integrable function ht such that: n→∞ lim ∂x vtxn (St ) = ht (St ) P a. .3) on a compact C and then the limit as xn tends to zero. s)} with bounded coefficients such that: Ri (x. Therefore. T Bjork. s) ≡ x(Ri (x. uniformly bounded in [0. W Schoutens. we obtain with (9. R Cont. We note that the processes [Y (k) . we have for t ∈ [0. Hence. St ). an anonymous referee. s). K)f (K)dK.5) − t. a ∈ (t. dx) has finite moments of order i 2. s) = ai+1. V satisfies: V (t. St ) = 2 ∂t [(a − t) t (a 2 − t. equation (9.2): [At f ](K)pt (K)dK = 1 2 C C ht (K)K 2 f (K)pt (K)dK. s)R(x. i: E[H (i+1) .k (s)mk (s)ds.i+1 ≡ 1 and the following square integrable martingale: Ht(i+1) ≡ t 0 t 0 t 0 The forward price V (t. We use the notations of section 6 and we follow closely the presentation of Nualart and Schoutens (2000). x↓0 Let f ⊂ D At ∩ C 2 with compact support . s)(x 2 ns (dx)−σs2 δ(x)dx)ds. D Nualart. (9.1 (s)dYs(1) + . the participants of the AMS Meeting (2001) and my colleagues at Merrill Lynch for stimulating and fruitful discussions. F. T ): Q. Y (j ) ]t = E = Ri (x. The coefficients of the above polynomial are bounded because the measure n(s. Proposition A. R = 0 T +∞ −∞ (V . We obtain: ∂x [℘ tx f ](St ) = ∂x C BS (vtx (K).P Balland Q UANTITATIVE F I N A N C E The transition function is Feller and it is thus associated with an infinitesimal generator At defined on D satisfying: At Acknowledgments I would like to thank L P Hughston. 1{s < t}x t i Hence. K). t ∈ (0. Y (j ) ]s = 0. K). t + xm ): 2 t (a − t.k (s) d[Y (k) . Ri .6) imply that we have for all K. On the other hand.1. s) − ai+1. . K) = ai+1. Since V has the same second-order derivatives with respect to S as the Black–Scholes function.i+1 (s) dYs(i+1) . Representation property for regular martingales In this appendix. F. + ai+1. T )}. K) is independent of K and the regular sticky-strike model coincides with the Black– Scholes model! ai+1. 1{s < t}x j = 0. 1 j i − 1). K)2 ]S 2 ∂S V (t. . St . We recall that two square-integrable martingales M and N are said to be strongly orthogonal or [ ]-orthogonal if [M.1. We finally obtain the following expression for the generator: 1 [At f ](St ) = 2 ht (St )St2 f (St ) P a.1 (s) + ai+1. K)2 .1 (s)).3) The Black–Scholes function satisfies: 1 ∂V C BS (V . (9. ai+1. + ai+1. Y (j ) ]t have independent increments. the participants of the EURANDOM seminar on Levy processes (2001). . There exists a family of pairwise strongly orthogonal square-integrable martingales {H (i) : i 1}. . s) Finally. T ] and j = 1. 1 a−t a t E[hu (Su )]du. (9. According to lemma 9. equations (9.4)–(9.s. Proof. Define Qi+1 (x. ai+1. T ]. t. V satisfies the backward equation (see Revuz and Yor 1991): ∂t V (t. N ]t is a martingale (see Protter 1995). it follows that: 1 ∂t V (t. Furthermore.1 (s)dXs + 0<s t Qi+1 ( Ss . St ) + [At V ](St ) = 0. . By taking first the expectation of equation (9. St ) of a call option with maturity a and strike K is a P -martingale. Y (j ) ]t and [H (i+1) .6) By direct calculation. we extend the Schoutens–Nualart representation property obtained for regular Levy martingales to regular martingales with independent increments. (t < T . Q(x. there is a positive. D B Madan. . . We define the following inner product acting on the space of real polynomials with time-dependent coefficients in L2 (0.s. we find a family of pairwise -orthogonal polynomials {Ri (x. we derive: Ht(i+1) = − i+1 k=2 ai+1.i (s)x i−1 + x i . St ) = C BS ((a − t) t (a (9.2 (s)x + .

Remark A. (i) φs dHs(i) . . . The family t1 . E[(Xt )k F (Xt )] = 0. i.2. ). Let F ∈ C with support in [−a. We need to prove that: +∞ L2 ( . ). we recall that: L2 = H ∈ m T : 0 Hs dms ∈ L2 ( . (j ) Lemma A.3.tn is total in the space L2 ( . 0}. Y and thus Z can be approximated arbitrarily closely by an element of ˆ . Xtn − Xtn−1 )). We observe that if the polynomial Ri is such that Ri . ) : X = H i=1 φ (i) • H (i) . H Proof. . The family is total in L2 ( . We have the following representation property for elements of . ). . . a] and weakly orthogonal to ˆ t . The case with n non-overlapping independent increments is treated similarly. i. Proof. σ (Xt1 . φ (i) ∈ L2 (i) .1. we have: ˆ ⊂ +∞ i=1 L2 (i) ⊂ L2 ( . Proposition A. Ykl = E[Ykl ] + i=1 0 T (i) (i) s dHs .2. ) since the i=1 H martingales H (i) are pairwise strongly orthogonal. H (i+1) ]T = 0 and thus H (i+1) = 0 almost everywhere. Ri = 0 then the martingale H (i+1) satisfies E[H (i+1) . Proposition A. .. . the linear hull ˆ t of t is in L2 (P {Xt ∈ dx}).tn is total in L2 ( . where φt(i) belongs to L2 (i) . The martingales H (j ) are consequently pairwise strongly orthogonal. Since the linear space ˆ spanned by is dense in L2 ( .Q UANTITATIVE F I N A N C E Therefore H (i+1) is strongly orthogonal to all Y (j ) for j = 1. σ (Xt1 . If the discontinuous component of the regular martingale is the sum of a finite number of Poisson processes with deterministic intensity then only a finite number of the martingales H (k) will be non-zero. p 110)). Let us prove the result for the family t . For any integer k and any power-increment (Xs1 − Xs0 )k .2. T ] × . ) can be approximated arbitrarily closely by an element Y of L2 ( . ) variables: t1 . Therefore.tn Deterministic implied volatility models Proof.1. A variable Z in L2 ( . Proposition 6. H +∞ L2 (i) ≡ X ∈ L2 ( . Let R ∈ . For any real Z. we have: +∞ k=0 = 1{s ∈ (s2 .e. 0s ˆ Since ¯ = L2 ( . we derive E[exp(iZXt )F (Xt )] = 0 for all Z and thus F (Xt ) = 0 (Malliavin and Airault (1994. Thanks to proposition A. = {Xtk11 . . .. we derive that the product of nonoverlapping power-increments Ykl ≡ ([X]s1 )k ([X]s3 )l with s0 s2 s0 < s1 s2 < s3 can be represented as follows: +∞ = {Xtk11 (Xt2 − Xt1 )k2 . . With lemma A. we have proved the result for the product of two non-overlapping power-increments. We prove the above equation by induction on k and by application of Ito’s formula to power functions as in Nualart and Schoutens (2000). H We observe that ⊕+∞ L2 (i) is closed in L2 ( . Proof.3. The space C of functions in L2 (P {Xt ∈ dx}) having compact support is dense in L2 (P {Xt ∈ dx})... By induction. 43 . there is a sequence of predictable processes {θs(i)1 k (s) : i 1} such that: 0s (Xs1 − Xs0 )k = E[(Xs1 − Xs0 )k ] + +∞ i=1 s1 s0 where is the predictable σ -algebra on [0.3. ) . . . . . we prove the result for the product of an arbitrary number of nonoverlapping power-increments. ) = i=1 +∞ i=1 L2 (i) . Lemma A.3.. For a square-integrable adapted martingale m. .. (Xtn − Xtn−1 )kn : ki ti < ti+1 T . . . ) by proposition A.2 t1 . . Xt2 − Xt1 . we deduce the following representation property of square-integrable variables. Let F ∈ L2 ( . s3 )}θs(i)3 l (s) 2s Hence. H θs(i)1 k (s) dHs(i) . then there are predictable processes { t(i) : i 1} such that: +∞ R = E[R] + i=1 0 T (i) s dHs(i) . Xtn − Xtn−1 )). (k 0). Since X is regular. Proof. . Following lemma A. σ (Xt1 . We define the following families of L2 ( . Xtn − Xtn−1 )) for some sequence {ti }. Therefore ˆ t is dense in C and thus in L2 (P {Xt ∈ dx}). Xt2 − Xt1 . (Xtn − Xtn−1 )kn : 0 The process (i) s (i) s is a predictable process defined by: +∞ j =1 s1 s0 (j θs0 s1 k (u) dHu ) . . we derive the result by closure of the above inclusion. ki 0}. ) then there is a family of predictable processes {φt(i) : i 1} such that: +∞ |Z|k E[|Xt |k |F (Xt )|] < E[F (Xt )2 ]1/2 exp(|Z a|). k! F = E[F ] + i=1 0 T By the dominated convergence theorem.

Finance 61–73 Cont R and da Fonseca J 2001 Deformation of implied volatility surfaces: an empirical analysis Empirical Approaches to Financial Fluctuations ed H Takayasu (Tokyo: Springer) Derman E 1999 Regimes of volatility Risk 4 55–9 Derman E and Kani I 1994 Riding on a smile Risk 7 32–9 Derman E and Kani I 1998 Stochastic implied trees: Arbitrage pricing with stochastic term and strike structure of volatility Int. J. Kabanov Y and Runggaldier W 1997 Towards a general theory of bond markets Finance Stochastics 1 141–74 Black F and Scholes M 1973 The pricing of options and corporate liabilities J. E 52 1197 Lebedev V A 1995 Fubini-theorem for parameter-dependent stochastic integrals with respect to L0 -valued random measures Probab. 73 2946 Merton R 1976 Option pricing when underlying stocks returns are discontinuous J. Theorie et Applications (Paris: Masson) Carr P and Madan D 1998 Towards a theory of volatility trading Volatility ed R A Jarrow (London: Risk) pp 417–27 Carr P and Madan D 1999 Option valuation using the fast Fourier transform J. Trans. Finance 1 61–110 Doob J L 1994 Measure Theory (New York: Springer) Dumas B. Rev. Political Economy 81 637–59 Bouchaud J P. Theory Appl. Finance 53 2059 Dupire B 1994 Pricing with a smile Risk 7 18–20 Eberlein E and Keller U 1995 Hyperbolic distributions in finance Bernoulli 1 281–99 Heston S 1993 A closed form solution for options with stochastic volatility with applications to bond and currency options Rev. 40 313–23 Levy P 1965 Processus Stochastiques et Mouvement Brownien (Paris: Gauthier-Villars) Madan D B and Seneta E 1990 The variance gamma model for share market returns J. Lett. Lett.P Balland Q UANTITATIVE F I N A N C E Hull J C and White A 1987 The pricing of options with stochastic volatilities J. Appl. Rev. 41 3 Breeden D and Litzenberger R 1978 Prices of state-contingent claims implicit in option prices J. Aarhus University Bjork T. Masi D G. Revuz D and Yor M 1991 Continuous Martingales and Brownian Motion (Berlin: Springer) Schonbucher P J 1999 A market model for stochastic implied volatility Phil. Financial Economics 3 125–44 Nualart D and Schoutens W 2000 Chaotic and predictable representations for Levy processes Stochastic Processes Applications 90 109–22 Protter P 1995 Stochastic Integration and Differential Equations (Berlin: Springer) Reiner E 1999 Volatility rules and implied processes Global Derivative Conf. Theor. Fleming J and Whaley R 1998 Implied volatility functions: empirical tests J. Cont R and Potters M 1998 Financial markets as adaptive systems Europhys. Soc. A 357 2071–92 Schoutens W 2001 Meixner processes in finance EURANDOM Report 2001-002 Schoutens W 2000 Stochastic Processes and Orthogonal Polynomials (Lecture Notes in Statistics) vol 146 (Berlin: Springer) References Balland P and Hughston L P 1999 Sticky-delta model Derivatives Week Barndorff-Nielsen O E 1995 Normal inverse Gaussian distributions and the modeling of stock returns Research Report no 300 Department of Theoretical Statistics. Comput. Finance 42 281–300 Jacod J and Protter P 1991 Probability Essentials (Berlin: Springer) Jacod J and Shiryaev A N 1987 Limit Theorems for Stochastic Processes (Berlin: Springer) Jarrow R A and Madan D 1999 Valuing and hedging contingent claims on semimartingales Finance Stochastics 3 111–34 Koponen I 1995 Phys. Business 51 621–51 Brezis H 1992 Analyse Functionelle. Financial Studies 6 327–43 44 . Business 36 394–419 Mantegna R N and Stanley H E 1994 The truncated levy flights Phys. R. Business 63 511–24 Malliavin P and Airault H 1994 Integration et Analyse de Fourier (Paris: Masson) Mandelbrot B B 1963 The variation of certain speculative prices J.

Sign up to vote on this title
UsefulNot useful