Applying Stochastic Time Changes To Levy Processes

Applying stochastic time changes
to Levy processes
Liuren Wu
Zicklin School of Business, Baruch College
Option Pricing
Liuren Wu (Baruch) Stochastic time changes Option Pricing 1 / 38
Outline
1
Stochastic time change
2
Option pricing
3
Model Design
Outline
1
2
Option pricing
3
Model Design
What Levy processes can and cannot do
Levy processes can generate dierent iid return innovation distributions.
Any distribution you can think of, we can specify a Levy process, with
the increments of the process matching that distribution.
Caveat: The same type of distribution applies to all time horizons
you may not be able to specify the distribution simultaneously at
dierent time horizons.
Levy processes cannot generate distributions that vary over time.
Returns modeled by Levy processes generate option implied volatility
surfaces that stay the same over time (as a function of standardized
moneyness and time to maturity).
Levy processes cannot capture the following salient features of the data:
Stochastic volatility
Stochastic risk reversal (skewness)
Stochastic correlation
Capturing stochastic volatility via time changes
Discrete-time analog again: R
t+1
=
t
+
t
t+1
t+1
is an iid return innovation, with an arbitrary distribution
assumption Levy process.
t
is the conditional volatility,
t
is the conditional mean return, both
of which can be time-varying, stochastic...
In continuous time, how do we model stochastic mean/volatility tractably?
If the return innovation is modeled by a Brownian motion, we can let
the instantaneous variance to be stochastic and tractable, not volatility
(Heston(1993), Bates (1996)).
If the return innovation is modeled by a compound Poisson process, we
can let the Poisson arrival rate to be stochastic, not the mean jump
size, jump distribution variance (Bates(2000), Pan(2002)).
If the return innovation is modeled by a general Levy process, it is tractable
to randomize the time, or something proportional to time.
Variance of a Brownian motion, intensity of a Poisson process are both
proportional to time.
Randomize the time
Review the Levy-Khintchine Theorem:
(u) E
_
e
iuX
t
= e
t(u)
,
(u) = iu +
1
2
u
2
2
for diusion with drift and variance
2
,
(u) =
_
1 e
iu
J
1
2
u
2
v
J
_
for Mertons compound Poisson jump.
The drift , the diusion variance
2
, and the Poisson arrival rate are all
proportional to time t.
We can directly specify (
t
,
2
t
,
t
) as following stochastic processes.
Or we can randomize time t T
t
for the same result.
We dene T
t

_
t
0
v
s
ds as the (stochastic) time change, with v

t
being the
instantaneous activity rate.
Depending on the Levy specication, the activity rate has the same
meaning (up to a scale) as a randomized version of the instantaneous
drift, instantaneous variance, or instantaneous arrival rate.
Applying separate time changes
... to dierent Levy components
Consider a Levy process X
t
(,
2
, p(x)).
If we apply random time change to X
t
X
T
t
with T
t
=
_
t
0
v
s
ds, it is
equivalent to assuming that (
t
,
2
t
,
t
) are all time varying, but they
are all proportional to one common source of variation v
t
.
If (
t
,
2
t
,
t
) vary separately, we need to apply separate time changes
to the three Levy components.
Decompose X
t
into three Levy processes: X
1
t
(, 0, 0),
X
2
t
(0,
2
, 0), and X
1
t
(0, 0, p(x)), and then apply separate time
changes to the three Levy processes.
Interpretation I
We can think of t as the calendar time, and T
t
as the business time.
Business activity accumulates with calendar time, but the speed varies,
depending on the business activity.
At heavy trading hours, one hour on a clock generates two hours worth of
business activity (v
t
= 2).
At afterhours, one hour generates half hour of activity (v
t
= 1/2).
Business activity tends to intensify before earnings announcements, FOMC
meeting days...
In this sense, v
t
captures the intensity of business activity at a certain time t.
Ane & Geman (2000, JFE): Stock returns are not normally distributed, but
become normally distributed when they are scaled by number of trades.
Interpretation II
We use Levy processes to model return innovations and stochastic time
changes to generate stochastic volatility and higher moments...
We can think of each Levy process as capturing one source of economic
shock.
The stochastic time change on each Levy process captures the random
intensity of the impact of the economic shock on the nancial security.
Return
K
i =1
X
i
T
i
t
i =1
(Economic Shock From Source i)
Stochastic impacts
.
I like this interpretations. Many economic phenomena can be modeled in
terms of economic shocks and (time varying) intensities.
Classication
In 1949, Bochner introduced the notion of time change to stochastic
processes. In 1973, Clark suggested that time-changed diusions could be
used to accurately describe nancial time series.
At present, there are two types of clocks used to model business time:
1
Continuous clocks have the property that business time is always
strictly increasing over calendar time.
2
Clocks based on increasing jump processes have staircase like paths.
The rst type of business clock can be used to describe stochastic volatility
which is what we do in this section.
The second type of clock can transform a diusion into a jump process
All Levy processes considered in the previous section can be generated as
changing the clock of a diusion with an increasing jump process
(subordinator).
All semimartingales can be written as time-changed Brownian motion.
Outline
1
2
Option pricing
3
Model Design
Option pricing
To compute the time-0 price of a European option price with maturity at t,
we rst compute the Fourier transform of the log return ln S
t
/S
0
. Then we
compute option value via Fourier inversions.
The Fourier transform of a time-changed Levy process:
Y
(u) E
Q
_
e
iuX
T
t
= E
Q
_
e
iuX
T
t
+
x
(u)T
t
e
x
(u)T
t
= E
M
_
e
x
(u)T
t
, u D C,
where the new measure M is dened by the exponential martingale:
dM
dQ
t
= exp (iuX
T
t
+T
t
x
(u)) .
Without time-change, e
iuX
t
+t
x
(u)
is an exponential martingale by
Levy-Khintchine Theorem.
A continuous time change does not change the martingality.
Proof: Uwe Kuchler and Michael Sorensen, 1997, Exponential Families of Stochastic Processes, Springer.
M is complex valued (no longer a probability measure).
Complex-valued measure change
When X and T
t
=
_
t
0
v
s
ds are independent, we have
Y
(u) E
Q
_
e
iuX
T
t
= E
Q
_
E
Q
e
iuX
Z
T
t
= Z
= E
Q
_
e
x
(u)T
t
,
by law of iterated expectations.
No measure change is necessary.
The operation is similar to Hull and White (1987): The option value of
an independent stochastic volatility model is written as the expectation
of the BMS formula over the distribution of the integrated variance,
T
t
=
_
t
0
v
s
ds.
A continuous time change does not change the martingality.
When X and v
t
are correlated, the measure change from Q to M hides the
correlation under the new measure.
If we must give a name, lets call M the correlation neutral measure.
Fourier transform
The Fourier transform of a time-changed Levy process:
Y
(u) E
Q
_
e
iuX
T
t
= E
M
_
e
x
(u)T
t
_
Tractability of the transform (u) depends on the tractability of
1
The characteristic exponent of the Levy process
x
(u)
Tractable Levy specications include: Brownian motion, (Compound)
Poisson, DPL, NIG, ... (done in previous section)
2
The Laplace transform of T
t
under M.
Tractable Laplace comes from activity rate dynamics: ane, quadratic,
Wishart, 3/2
The measure change from Q to M is dened by an exponential
martingale.
The two (X, T
t
) can be chosen separately as building blocks, for dierent
purposes.
The Laplace transform of the stochastic time T
t
We have solved the characteristic exponent of the Levy process (by the
Levy-Khintchine Theorem).
Now we try to solve the Laplace transform of the stochastic time,
L
T
() E
_
e
T
t
= E
_
e
t
0
v
s
ds
_
(1)
Recall the pricing equation for zero-coupon bonds:
B(0, t) E
Q
_
e
t
0
r
s
ds
_
(2)
The two pricing equations look analogous (even though they are not related)
Both v
t
and r
t
need to be positive.
If we set r
t
= v
t
, L
T
() is essentially the bond price.
The similarity allows us to borrow the vast literature on bond pricing:
Ane class: Zero-coupon bond prices are exponential ane in the
state variable.
Quadratic: Zero-coupon bond prices are exponential quadratic in the
state variable.
...
Review: Bond pricing dynamic term structure models
A long list of papers propose dierent dynamic term structure models:
Specic examples:
Vasicek, 1977, JFE: The instantaneous interest rate follows an
Ornstein-Uhlenbeck process.
Cox, Ingersoll, Ross, 1985, Econometrica: The instantaneous interest
rate follows a square-root process.
Many multi-factor examples ...
Classications (back-lling)
Due, Kan, 1996, Mathematical Finance: Spot rates are ane
functions of state variables.
Due, Pan, Singleton, 2000, Econometrica: Ane with jumps.
Due, Filipovic, Schachermayer, 2003, Annals of Applied Probability:
Super mathematical representation and generalization of ane models.
Leippold, Wu, 2002, JFQA: Spot rates are quadratic functions of state
variables.
Filipovic, 2002, Mathematical Finance: How far can we go?
Gabaix: Bond prices are ane.
Identifying dynamic term structure models:
The forward and backward procedures
The traditional procedure:
First, we make assumptions on factor dynamics (Z), market prices (),
and how interest rates are related to the factors r (Z), based on what
we think is reasonable.
Then, we derive the fair valuation of bonds based on these dynamics
and market price specications.
The back-lling (reverse engineering) procedure:
First, state the form of solution that we want for bond prices.
Then, gure out what dynamics specications generate the pricing
solutions that we want.
The dynamics are not specied to be reasonable, but specied to
generate a form of solution that we like.
It is good to be able to go both ways.
It is important not only to understand existing models, but also to
derive new models that meet your work requirements.
Back-lling ane models
What we want: Zero-coupon bond prices are exponential ane functions of
state variables.
Continuously compounded spot rates are ane in state variables.
It is simple and tractable. We can use spot rates as factors.
Let Z denote the state variables, let B(Z
t
, ) denote the time-t fair value of
a zero-coupon bond with time to maturity = T t, we have
B(Z
t
, ) = E
Q
t
_
exp
_
_
T
t
r (Z
s
)ds
__
= exp
_
a() b()
Z
t
_
By writing B(Z
t
, ) and r (Z
t
), and solutions a(), b(), I am implicitly
focusing on time-homogeneous models. Calendar dates do not matter.
This assumption is for (notational) simplicity more than anything else.
With calendar time dependence, the notation can be changed to,
B(Z
t
, t, T) and r (Z
t
, t). The solutions would be a(t, T), b(t, T).
Questions to be answered:
What is the short rate function r (Z
t
)?
Whats the dynamics of Z
t
under measure Q?
Diusion dynamics
To make the derivation easier, lets focus on diusion factor dynamics:
dZ
t
= (Z)dt + (Z)dW
t
under Q.
We want to know: What kind of specications for (Z), (Z) and r (Z)
generate the ane solutions?
For a generic valuation problem,
f (Z
t
, t, T) = E
Q
t
_
exp
_
_
T
t
r (Z
s
)ds
_
T
_
,
where
T
denotes terminal payo, the value satises the following partial
dierential equation:
f
t
+Lf = rf , Lf innitesimal generator
with boundary condition f (T) =
T
.
Apply the PDE to the bond valuation problem,
B
t
+ B
Z
(Z) +
1
2
B
ZZ
(Z)(Z)
= rB
with boundary condition B(Z
T
, 0) = 1.
Back lling
Starting with the PDE,
B
t
+ B
Z
(Z) +
1
2
B
ZZ
(Z)(Z)
= rB, B(Z
T
, 0) = 1.
If B(Z
t
, ) = exp(a() b()
Z
t
), we have
B
t
= B
_
a
() + b
()
Z
t
_
, B
Z
= Bb(), B
ZZ
= Bb()b()
,
y(t, ) =
1
_
a() + b()
Z
t
_
, r (Z
t
) = a
(0) + b
(0)
Z
t
= a
r
+ b
r
Z
t
.
Plug these back to the PDE,
a
() +b
()
Z
t
b()
(Z) +
1
2
b()b()
(Z)(Z)
= a
r
+b
r
Z
t
Question: What specications of (Z) and (Z) guarantee the above PDE
to hold at all Z?
Power expand (Z) and (Z)(Z)
around Z and then collect

coecients of Z
p
for p = 0, 1, 2 . These coecients have to be zero
separately for the PDE to hold at all times.
Back lling
a
() + b
()
Z
t
b()
(Z) +
1
2
b()b()
(Z)(Z)
= a
r
+ b
r
Z
t
Set (Z) = a
m
+ b
m
Z + c
m
ZZ
+ and
[(Z)(Z)
]
i
=
i
+
i
Z +
i
ZZ
+ , and collect terms:

constant a
() b()
a
m
+
1
2
b()b()

i
= a
r
Z b
()
b()
b
m
+
1
2
b()b()
i
= b
r
ZZ
b()
c
m
+
1
2
b()b()

i
= 0
The quadratic and higher-order terms are almost surely zero.
We thus have the conditions to have exponential-ane bond prices:
(Z) = a
m
+ b
m
Z, [(Z)(Z)
]
i
=
i
+
i
Z, r (Z) = a
r
+ b
r
Z.
We can solve the coecients [a(), b()] via the following ordinary
dierential equations:
a
() = a
r
+ b()
a
m
1
2
b()b()

i
b
() = b
r
+ b
m
b()
1
2
b()b()

i
starting at a(0) = 0 and b(0) = 0.
Quadratic and others
Add jumps to the ane dynamics: The arrival rate of jumps need to be
ane in the state vector.
Can you identify the conditions for quadratic models: Bond prices are
exponential quadratic in state variables?
Can you identify the conditions for cubic models: Bond prices are
exponential cubic in state variables?
Ane bond prices: Recently Xavier Gabaix derive a model where bond
prices are ane (not exponential ane!) in state variables.
From ane DTSM to ane activity rates
Review of ane DTSM: B(Z
0
, t) E
_
e
t
0
r
s
ds
_
= e
a(t)b(t)
Z
0
if
r
t
= a
r
+ b
r
Z
t
, (Z) = ( Z), [(Z)(Z)
]
ii
=
i
+
i
Z
By analogy, if we want:
L
T
(z) E
_
e
zT
t
= E
_
e
z
t
0
v
s
ds
_
= e
a(t)b(t)
Z
0
, we can set
v
t
= a
v
+ b
v
Z
t
, (Z) = ( Z), [(Z)(Z)
]
ii
=
i
+
i
Z
Problem: We need the ane dynamics under the complex-valued measure
M. Correlations between the Levy process X and the state vector Z can
make the whole thing messy. Ane dynamics under Q are no guarantee for
exponential ane solution.
We use some concrete examples to show how this works.
Example: The Heston (1996) stochastic volatility model
SDE: dS
t
/S
t
= (r q)dt +
v
t
dW
t
with
dv
t
= ( v
t
)dt +
v
v
t
dW
v
t
, E[dW
t
dW
v
t
] = dt.
We can write the security return as a time-changed Levy process,
ln S
t
/S
0
= (r q)t + W
T
t

1
2
T
t
, T
t
=
_
t
0
v
s
ds.
The Fourier transform of the return,
s
(u) E
Q
_
e
iu ln S
t
/S
0
= e
iu(r q)t
E
Q
_
e
iu(W
T
t
1
2
T
t
)
_
= e
iu(r q)t
E
Q
_
e
iu(W
T
t
1
2
T
t
)+(u)T
t
(u)T
t
_
= e
iu(r q)t
E
M
_
e
(u)T
t
where (u) =
1
2
(iu + u
2
) is the characteristic exponent of the Levy process
(W
t

1
2
t).
To solve the Laplace transform, we need the dynamics of v under M.
Example: Heston
SDE: dS
t
/S
t
= (r q)dt +
v
t
dW
t
with
dv
t
= ( v
t
)dt +
v
v
t
dW
v
t
, E[dW
t
dW
v
t
] = dt.
The measure change:
dM
dQ
= exp(iu(W
T
t

1
2
T
t
) + (u)T
t
).
The v dynamics under M:
dv
t
= ( v
t
)dt +E[iudW
T
t
v
t
dW
v
t
] +
v
v
t
dW
v
t
= ( v
t
)dt + iu
v
v
t
dt +
v
v
t
dW
v
t
= (
M
v
t
)dt +
v
v
t
dW
v
t
with
M
= iu
v
.
Note that dW
T
t
and

v
t
dW
t
are equivalent in distribution.
Since the v dynamics are ane under M, we have the Laplace transform
exponential ane in v,
s
(u) E
Q
_
e
iu ln S
t
/S
0
= e
iu(r q)t
E
Q
_
e
iu(W
T
t
1
2
T
t
)
_
= e
iu(r q)t
E
M
_
e
(u)T
t
= e
iu(r q)ta(t)b(t)v
0
with a
(t) = b(t) and b
(t) = (u)
M
b(t)
1
2
b(t)
2
2
v
, starting at
a(0) = b(0) = 0.
Example: HestonNot all ane works
The key for tractability is to maintain the v dynamics ane under M.
Suppose we extend the Heston specication:
dv
t
= ( v
t
)dt +
v
+ v
t
dW
v
t
, which is still ane under Q, but it is
no longer ane under M:
dv
t
= ( v
t
)dt +E[iudW
T
t
,
v
+ v
t
dW
v
t
] +
v
v
t
dW
v
t
unless we redene the time change as T
t
=
_
t
0
( + v
s
)ds or we set = 0.
The constant volatility specication does not work either:
dv
t
= ( v
t
)dt +
v
dW
v
t
(in addition to the fact that v
t
can go to zero
this time).
A general ane specication v
t
= a
v
+ b
v
Z
t
does not always work for the
same reason, but the following two-factor specication works:
dv
t
= (m
t
v
t
)dt +
v
v
t
dW
v
t
, dm
t
=
m
(
m
m
t
)dt +
m
m
t
dW
m
t
with E[dW
m
t
dW
v
t
] = E[dW
m
t
dW
t
] = 0.
Example: Bates jump-diusion stochastic volatility model
SDE: dS
t
/S
t
= (r q)dt +
v
t
dW
t
+ dJ() (e
mu
J
+
1
2
v
J
1)dt with
dv
t
= ( v
t
)dt +
v
v
t
dW
v
t
, E[dW
t
dW
v
t
] = dt. The stock price
process includes compound Poisson jump process, with arrival rate .
Conditional on a jump occurring, the jump size in return has a normal
distribution (
J
, v
J
).
We can write the security return as a time-changed Levy process,
ln S
t
/S
0
= (r q)t + [W
T
t

1
2
T
t
] + [X
t
k
X
(1)t], T
t
=
_
t
0
v
s
ds.
where X denotes a pure-jump Levy process with Levy density given by
(x) =
1
2v
J
e
(x
J
)
2
2v
J
The cumulant exponent is
k
x
(s) =
_
R
0
(e
sx
1)(x)dx =
_
e
s
J
+
1
2
s
2
v
J
1
_
.
Example: Bates (1996)
The Fourier transform of the return,
s
(u) E
Q
_
e
iu ln S
t
/S
0
= e
iu(r q)t
E
Q
_
e
iu(W
T
t
1
2
T
t
)
_
E
Q
_
e
iu(X
t
k
X
(1)t)
= e
iu(r q)t
e
a(t)b(t)v
0
e
J
(u)
where [a(t), b(t)] come directly from the Heston model and the
characteristic exponent of the pure-jump Levy process (concavity adjusted)
is:
J
(u) =
_
R
0
(1e
iux
)(x)dx+iuk
X
(1) = (1e
iu
J
1
2
u
2
v
J
)+iu
_
e
J
+
1
2
v
J
1
_
By denition, jumps are orthogonal to diusion. Hence, the two components
can be processed separately.
Question: Why just time change diusion W? Why not also time change
the jump X?
ln S
t
/S
0
= (r q)t + [W
T
t

1
2
T
t
] + [X
T
x
t
k
X
(1)T
x
t
].
Also, replace the compound Possion jump with any type of jump you like
SV4 in Huang and Wu (2004).
Outline
1
2
Option pricing
3
Model Design
Model design: General principles
Start with the risk-neutral (Q) process Thats where tractability is
needed the most dearly.
Identify the economic sources, model each with a Levy process (X
k
t
).
Decide whether to apply separate time changes: X
k
t
X
k
T
k
t
to make
the impact of this economic source stochastic over time.
Concavity adjust each component to guarantee the martingale
condition: E
Q
[S
t
/S
0
] = e
(r q)t
.
ln S
t
/S
0
= (r q)t +
K
k=1
_
b
k
X
k
T
k
t
k
x
k (b
k
)T
k
t
_
,
For tractability,
Use Levy processes that generate tractable characteristic exponents
(
X
(u)).
Use time changes that generate tractable Laplace transforms
Orthogonalize the economic shocks X
k
such that
s
(u) = e
iu(r q)t
k
E
_
e
iu
b
k
X
k
T
k
t
k
x
k
(b
k
)T
k
t
_
Market prices and statistics dynamics
Since we can always use Euler approximation for model estimation,
tractability requirement is not as strong for the statistical dynamics.
We can specify pretty much any forms for the market prices subject to (i)
technical conditions, (ii) economic sensibility, and (iii) identication
concerns.
Simple/parsimonious specication: Constant market prices of return and vol
risks (
k
,
kv
)
M
t
= e
rt
K
k=1
exp
_
k
X
k
T
k
t
x
k (
k
) T
k
t

kv
X
kv
T
k
t
x
kv (
kv
) T
k
t
_
,
W
t
constant drift adjustment =
2
.
Pure jump Levy process
P
(x) = e
x
Q
(x), drift adjustment:
=
P
J
(1)
Q
J
(1) =
Q
J
(1 + )
Q
J
()
Q
J
(1).
Time change: instantaneous risk premium (v
t
) proportional to the
risk level v
t
.
Example: Return on a stock
Model the return on a stock to reect shocks from two sources:
Credit risk: In case of corporate default, the stock price falls to zero.
Model the impact as a Poisson Levy jump process with log return
jumps to negative innity upon jump arrival.
Market risk: Daily market movements (small or large). Model the
impact as a diusion or innite-activity (innite variation) Levy jump
process or both.
Apply separate time changes to the two Levy components to capture (1) the
intensity variation of corporate default, (2) the market risk (volatility)
variation.
Key: Each component has a specic economic purpose.
Carr and Wu, Stock Options and Credit Default Swaps: A Joint Framework for Valuation and Estimation, JFEC, 2010
Example: A CAPM model
Example: A CAPM model :
ln S
j
t
/S
j
0
= (r q)t +
_
j
X
m
T
m
t
x
m(
j
)T
m
t
_
+
_
X
j
T
j
t
x
j (1)T
j
t
_
.
Estimate and market prices of return and volatility risk using index
and single name options.
Cross-sectional analysis of the estimates:
Are the beta estimates similar to estimates from time-series stock
return regressions?
An international CAPM:
Henry Mo, and Liuren Wu, International Capital Asset Pricing: Evidence from Options, Journal of Empirical Finance, 2007,
14(4), 465498.
Example: Return on an exchange rate
Exchange rate reects the interaction between two economic forces.
Use two Levy processes to model the two economic forces separately.
Consider a negatively skewed distribution (downside jumps) from each
economic source (crash-o-phobia from both sides). Use the dierence to
model the currency return between the two economies.
Apply separate time changes to the two Levy processes to capture the
strength variation of the two economic forces.
Stochastic time changes on the two negatively skewed Levy processes
generate both stochastic volatility and stochastic skew.
Key: Each component has its specic economic purpose.
Peter Carr, and Liuren Wu, Stochastic Skew in Currency Options, Journal of Financial Economics, 2007, 86(1), 213247.
Example: Currencies returns and sovereign CDS
Dollar price of peso drops by a signicant amount when Mexico defaults on
its sovereign debt.
Currency return on peso contains both market risk and credit risk.
The intensities of both types of risks are stochastic (and probably
correlated).
Peter Carr, and Liuren Wu, Theory and Evidence on the Dynamic Interactions Between Sovereign Credit Default Swaps
and Currency Options, Journal of Banking and Finance, 2007, 31(8), 23832403.
Exchange rates and pricing kernels
Exchange rate reects the interaction between two economic forces.
The economic meaning becomes clearer if we model the pricing kernel of
each economy.
Let m
US
0,t
and m
JP
0,t
denote the pricing kernels of the US and Japan.
Then the dollar price of yen S
t
is given by
ln S
t
/S
0
= ln m
JP
0,t
ln m
US
0,t
.
If we model the negative of the logarithm of each pricing kernel
(ln m
j
0,t
) as a time-changed Levy process, X
j
T
j
t
(j = US, JP) with
negative skewness. Then, ln S
t
/S
0
= ln m
JP
0,t
ln m
US
0,t
= X
US
T
US
t
X
JP
T
JP
t
Consistent and simultaneous modeling of all currency pairs (not limited
to 2 economies).
Bakshi, Carr, and Wu, Stochastic Risk Premium, Stochastic Skewness, and Stochastic Discount Factors in International
Economies JFE, 2008, 87(1), 132-156.
Reverse engineer the pricing kernel of US, UK, and Japan using currency options on dollar-yen, dollar-pound, and
pound-yen.
Limitations and extensions
Recall the rule of thumb for model design:
ln S
t
/S
0
= (r q)t +
K
k=1
_
b
k
X
k
T
k
t
k
x
k (b
k
)T
k
t
_
,
X
k
are orthogonal. Generate correlation among dierent factors via
factor loadings b
k
.
Limitation: Constant factor loading (b
k
), together with orthogonality
assumption, puts restrictions on co-movements across assets.
Future work: How to allow exible correlation dynamics (with independent
variation)?
Each asset (economic shock) has its own business clock.
How to model the co-movements of business clocks of multiple assets
(economic sources)?
If economics ask for it, we may also need to get out of the
exponential-ane setup:
Carr&Wu: Leverage Eect, Volatility Feedback, and Self-Exciting Market Disruptions, wp.
Concluding remarks
Modeling security returns with (time-changed) Levy processes enjoys three
key virtues:
Generality: Levy process can be made to capture any return innovation
distribution; applying time changes can make this distribution vary
stochastic over time.
Explicit economic mapping: Each Levy component captures shocks
from one economic source. Time changes capture the relative variation
of the intensities of these impacts.
Tractability: Combining any tractable Levy process (with tractable
(u)) with any tractable activity rate dynamics (with a tractable
Laplace) generates a tractable Fourier transform for the time changed
Levy process. The two specications are separate.
It is a nice place to start with for generating security return dynamics that
are parsimonious, tractable, economically sensible, and statistically
performing well.

Applying Stochastic Time Changes To Levy Processes

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Applying Stochastic Time Changes To Levy Processes

Hochgeladen von

Copyright:

Verfügbare Formate

Applying stochastic time changes

ds as the (stochastic) time change, with v

around Z and then collect

+ , and collect terms:

(t) = b(t) and b

Das könnte Ihnen auch gefallen