T H E Q U A N T C L A S S R O O M B Y AT T I L I O M E U C C I

A Short,
Comprehensive, Practical
Guide to Copulas
Visually introducing a powerful risk management tool
to generalize and stress-test correlations.

T
he stochastic behavior of one single financial vari- retical results. In Section 3, we address copulas implementa-
able say, prices, or implied volatilities (etc.) tion issues.
is fully described by its probability distribution,
which is called the marginal distribution. In a Section 1: Univariate Results
market of multiple financial variables, all the in- In this section, we cover well-known
Consider an arbitraryresults random that prepare
variable ,thewith a fully arbitrar
formation on the stochastic behavior of the market is fully ground for the Alldefinition
the features of, and thedistribution
of the intuition behind, of copulas.
are described by its prob
described by the joint probability distribution. Consider an arbitrary
function (pdf)random variable
, dened inX, witha away
such fullythat,
arbi-for any set of pote
The multivariate distribution of a set of financial trary for
distribution. the variable
All the features
, the
Consider an arbitrary random variable , with a fully arbitrary distribution. are
variables of following
the identity
distribution of for
X the pdf holds
is fully specified by the separate marginalAll Consider
distributions an arbitrary
of the random
described variable
by its with
probability
, a fully
the features of the distribution of are described by its probabilityXdensity in
density arbitrary
function distribution.
(pdf) f , defined
Z
All the
variables and by their copula, or, loosely speaking,
function features
(pdf) of
the
the correla- distribution wayofthat,
suchina such
, dened a wayare described
forthat,
any for anyby
set of itsofprobability
potential
set P values
potential
{ Xvalues density
}for theXvari-
() .
tions among the variables. function (pdf) , dened
for the variable , the able
in such a way
X, theidentity
following followingthat, for
foridentityany
the pdffor set of potential
the pdf holds:
holds values X
X

Modeling the marginals and the copula for separately

the variable , the following identity
provides Zfor the pdf
Accordingly, withholds
mild abuse of notation we write to den
greater flexibility for the practitioner to model randomness. Z
P { Xa}distribution . pdf is . An alternative
()whose (1) (1) way to represent t
Therefore, copulas have been used extensively in finance, P { X }of isXits () .
cumulative distribution function (1) (cdf), dened as follo
X
both on the sell-side to price derivativesAccordingly,
(see, e.g., Li (2000)),
with mild abuse Accordingly,
of notation with wemild
writeabuse
of notation,
to denotewethat write
ZX~f has X to
and on the buy-side to model portfolioAccordingly,
risk (see e.g. with
Meucci,mild abuse
denote of notation
that X has wea write
distribution
a distribution whose pdf is . An alternative way to representthe whose
to denote
pdf is f
()
that
X . An
has
alternative
distribution () .
Gan, Lazanas and Phelps (2007)). a of
distribution
is its cumulative is to. represent
whose pdfdistribution
way An alternative
function way to
the distribution
(cdf), represent
dened of as the
X isfollows distribution
its cumulative distri-
Here, we provide a review of the theory of of iscopulas
its cumulative
proving distribution
bution function function(cdf),
ZIf (cdf),
defineddened as follows
as follows:
we feed the random variable into a generic function we
the most useful results. We also provide a guide to the practi-
1 Z (2)if we set () sin (
() random
variable
() . (). For instance,
cal implementation of copulas. For a detailed discussion of () random variable
() . sin (), which is bound (2) (2)in the interval [1

implementation issues and for the code, please If we refer
feed theto the
random variable situation arises function
into a generic when wetransformwe obtain with its own cdf, i.e.
another
companion paper Meucci (2011). If we feed
random the random
variable (). If weFor
variable feed the random
into
instance, if wevariable
a generic ()X
function
set into
sinaobtain
we generic
() function
we another
obtain a g,
This article is organized as follows. Inrandom
Section 1, we
variable review
we
(). obtain
For
random variable sin (), which is bound another
instance, random
if we set variable
()
in the interval Y
sin g(X).
() we For instance,
obtain a if
Key concept. If we [1
feed 1].theAarbitrary
special variable through
strictly univariate results that nonetheless naturally
random lead
variable to
the
sinwe(),
situation arises when we transform we set g(x)
which sin(X),
is bound
with we inobtain
its ownthe a random
interval
cdf,special
i.e. [1 variable
1].
. A Y
special sin(X).
obtain a very transformed random variable, which
multivariate concept of copulas. In Sectionsituation2, we introduce
arises when wewhich is bound
transform within the itsinterval
own cdf, [-1,1].
i.e. Aspecial
. situation arises
grade of
copulas, highlighting and proving the most important theo- when we transform X with its own cdf i.e., g FX.
Key concept. If we feed the arbitrary variable through itsown cdf, () .
Key concept. If we
we obtain a very special transformedfeed the arbitrary variable through its own cdf,
Therandom variable,
distribution of the whichgrade is called
is uniform the on the unit interval
we obtain
grade of a very special transformed random variable, which is called the
the original distribution
grade of
www.garp.org O () C T .O B E R 2 0 1 1 RISK PROFESSIONAL (3) 22
() . (3)
U ,
The distribution of the grade is uniform on the unit interval regardless of [01]
The
the distribution of the grade
original distribution is uniform on the unit interval regardless of
 P { X } X
P {
()
.X } () . (1) (1)
X
X U FX X U 0,1
with Accordingly,
mild abuse ofwithnotation we write to denote that has grade
with mild abuse of mildwe
notation abuse
writeof
notation wedenote
to write
that has
to denote that has
whose pdf is . An alternative way to
represent
the distribution U FX X U 0,1
a distribution
whose pdf is . Anto alternative way to represent the distribution
whose
mulative is . An
pdfdistribution
of distribution
mulative T H E Q U A N T C L A S S R O O M B Y AT T I L I O M E U C C I
alternative
function
is its cumulative
way
(cdf), represent
dened
distribution function
function
as the
followsdistribution
(cdf), dened as follows
Z (cdf), dened Z
as follows
x
() Z () .
() () . (2) (2)
() () . (2)
x
X random variable with arbitrary distribution
the random variable
If wevariable
feed the into a generic
random function
variablefunction into a we we obtain
generic another
function we obtain another x
he random
ble ().
random For into a generic
instance, if we set () sin obtain
() we another
obtain a
ble variable
For ().
instance, if wein For
set instance,
() sin if ()
we set we A () asin () we obtain a
obtain X random variable with arbitrary distribution
ble random
sin().
(),
variablewhich is
bound
sin (), the interval
which is bound [1 in 1].
the special [1 1]. A special
interval X random variable with arbitrary distribution
ble sin
es when we (),
transformwhichis with bound itsin ownthecdf, interval
i.e. [1
own 1]. A special
. cdf, i.e. Figure 1: The cdf maps an arbitrary random variable into a uniform variable
situation
s when we transform arises when we transform
with its own
concept.
cdf,feed with
i.e.the its .
arbitrary .
Key If we variable X through The above result (3)-(4) also works backwards: if we feed
its own cdf, we obtain a very special transformed Figure
random 1: Theacdf maps an
uniform arbitrary
random variablerandom variable
U into thefeed into a uniform
inverse -1 variable
cdf Frandom
X , we
ncept. If Key we feed the arbitrary variable through its owncdf, The above Figureresult cdf also
(3)-(4)
1: The mapsworks backwards:
an arbitrary random if wevariable ainto
uniform
a uniform variable
concept. If we feed the arbitrary variable through its own cdf, 1 variable X FX
ncept. variable, which is called thethrough
grade of Xown cdf, variable into theobtain a random (U)variable
with distribution 1 fX,
-1
If we feed
very special the arbitrary
transformed random variable
variable, which itscalled
is the inverse cdf we obtain a random ( )
we obtain
very special a very special
transformed random transformed
variable, which random variable,
is called the which is called the as follows:
with distribution
The above result
(3)-(4) also works backwards: if we feed a uniform random
grade of The above result (3)-(4) also 1 works backwards: if we feed a uniform random
() . . (3) variablevariable (3)into(3) the inverse cdf cdf we 1 obtain a random variable 1( )
1
() . () (3) with distribution U[01] 7 we
into the inverse 1 obtain
( ) a random.
variable (6)
(6) ( )
ution of thedistribution
grade is uniform on the unit interval regardless with distribution
of unit interval,
ution ofThethe grade The is distribution
of theon
uniform grade of the
the is
unitgrade
uniform
interval is on
uniform
the unit
regardless on interval
the
of regardless of , say Student
distribution distribution Even better, we can choose any desired target distribution
1 1
( ) .
the original
distribution regardless of theoriginal distribution fX : Even U
better, we
[01] 7 choose
can
U[01] 7 any desired
( ) target . distribution fX (6) (6)
or chi-square, etc.; then we compute the cdf and the inverse cdf 1
; and
U[01] , (4) say, Student t or chi-square, etc.; subsequently, we compute
U[01] , nally
Even transform
better,
Even better, a uniform
we
(4) canvariable
canwechoose todesired
any desired
choose any generate
targettarget a random
distribution
distribution
variable, ,say with
Student
say Student
U[01] , (4)
the ordesired (4)
orarbitrary
chi-square, the distribution.
chi-square,
etc.; cdf FX-we
etc.;
then andcompute
then the inverse
we compute thecdf cdfFcdf
the X ;
-1
and,
and andlastly,
the the we transform
inverse
inverse cdf
cdf 11a ; and
; and

nally nally
transform uniform
transform
a uniformvariable
a uniform U
variable to generate
variable a random
to generate
to generate variable
a randomvariable
a random X with
variable the
withwith
the the desired
desired desired
arbitrary arbitrary
arbitrary distribution.
distribution.
distribution.
e proof of The this resultThe proceeds
simple as follows
proof of this result isas follows Key concept. Starting from an arbitrary target distribution and a
e proof of thissimple
result proof
proceeds of this as result
follows proceeds
uniform random variable, we canStarting transform the
() P { } = P{ ()} }= P { () } (5) Key Key concept.concept. Starting from from anuniform
an arbitrary arbitrary variable distribu-
target target
into a
distribution and a
P variable with
Key the
concept.desired
(5) target
Starting distribution
from an arbitrary target distribution and a
() P { } 1 () = P { () {
} 1

1 (5) uniformuniform(5) tionrandom and avariable,
uniformwe random variable,the
can transform weuniform
can transformvariable the into a
= P 1 () =
P 1() = . random variable, we can transform the uniform variable into a
= P () = =
1 () = () = .
variable uniform
with thevariable
desiredinto target a variable
distribution with the desired target distri-
() = . variable with the desired target distribution
bution, as follows: 1
= , which
() Therefore is()the= cdf
Therefore, , Fofof a uniform
which isu, the distribution.
cdf of acdf uniform We distribution.
can We can 7 ( ) . (7)
() = ,
e univariate which
miracle is the
as cdf
follows. U (u)
a =
uniform which is the
distribution. of a
We uniform
can distribu- U
7 ( ) .
1
(7)
summarize the univariate miracle as follows. [01]
e1univariate
we sketch miracle
the tion. as follows.
intuition behind (4). First of all the variable U


7

1 ( ) . (7) (7)
1 we sketch In Figure 1 we sketch
the intuition behind the intuition First behind
of all (4). First of all the variable [01]

terval [0 1]
lives in because
the
In
intervaltheFigure
cdf
[0 1]
1 (see(4).
satises
because
below),
0 the cdf
we
() 1.the
sketch
satises 0
variable
the

intuition
Furthermore,
()
behind (4).
1. Furthermore, U[01]
erval [0 1] because
is the
uniform on [0 Firstthe
1] ofcdfall,
because satises
thethe 0
variable
cdf U ()
islivesin1. the
steeper Furthermore,
interval
where

there[0,1], because
is uniform
ntial variable
on 1]
[0 is because
uniform the on cdf [0 1] because
is the cdf
steeper where is steeper
there The proof where there
of (7) reads
outcomes forthe cdfand satisfies
thus these 0 outcomes
FX (x)1. are
Furthermore,
spread out the variable U The
overare The proof
is proof of(7) of reads(7) reads
ntialonare more
outcomes potential
foruniform
andthe outcomes
thus these for and
outcomes thus these
are Fspread outcomes
out over spread out over
val; a the
wider other
interval;hand, on on
the [0,1]
cdf
other because
is
hand, atterthe
the cdf
where
cdf is steeper
X there
is are
atter where
few
where there
P {
there are
are } few= P 1
( ) = P () = () . (8)
The proof of reads
P { } = P ( ) = P () = () . (8) (8)
(7)
dal; on the
the cdfother hand, the allcdfthe is atter where there are few
outcomes 1
thusoutcomes, more
concentrates potential
and thus theallcdf
outcomes for X, and
occurring thus
over these
a largeoutcomes are
a thus theset.cdf concentrates theconcentrates
outcomes all the outcomes
occurring over a largeoccurring over a large
Thisof result iscdf 1generate Monte from arbitrary
smallinterval The
into balance
spread
a small between
out
set. over
The athese
wider
balance two eects,on
interval;
between i.e.the
these dilution
other
two hand,
eects, theThis
i.e. Pextremely
{
dilution
resultof }
is
useful
= P to
extremely ( ) to
useful = P Carlo
generate scenarios
Monte = ()
()scenarios
Carlo from. arbitrary (8)
smalland
narios set.concentration
The balance
F is between
of
flatter scarce
where these
scenarios,
there twoare eects,
gives
few i.e.todilution
rise
outcomes, a uniform
and ofthus distributions
desired the cdf using
This as
result

input
is only
extremely a uniform
useful to number
generate generator.
Monte Carlo sce-
abundant scenarios
X and concentration of scarce scenarios, gives rise desired
to a distributions using as input only a uniform number generator.
uniform
arios and concentration of scarce scenarios, gives rise to a uniformThis result is extremely useful to generate Monte Carlo scenarios from arbitrary
distribution. concentrates all the outcomes occurring over a large interval narios from arbitrary desired distributions, using as input only
into a small set. The balance between these two effects desired distributions
i.e., a uniform using as input
number only a uniform number generator.
generator.
3 3
dilution 2of abundant scenarios and concentration of scarce
scenarios2 gives rise to a uniform 2 distribution. Section 2: Copulas (Theory)
2 Copulas: theory 3
Now we are fully equipped to introduce the copula, by ex-
Figure 1: Mapping an Arbitrary Random Now we are tending to the multivariate
fully equipped to introduce framework
the copula, the by univariate
extending results
to the multi-
Variable into a Uniform Variable via cdf variate framework (3)-(4) and the(7).univariate results (4) and (7).
Consider Consider a N-dimensional
a -dimensional vector of vector
random of random
variables variables
( X 1 )
with a fully (X1general
,...,XN) with a fully general
multivariate multivariate
distribution distribution
represented by itsrepre-
pdf .
We recall that sented in by
theitsmultivariate
pdf X~ fX. We case recall
the pdfthat in the multivariate
is dened in suchcase, a way that,
for any setthe ofpdf fX is defined
potential joint invaluessuch aXway R for
that, for any
(1 set of potential
) the following
identity holds joint values X N
for (X1,...,XN), the following identity holds:
Z
P {(1 ) X } (1 ) 1 . (9) (9)
X

From the jointFrom the joint distribution

distribution we can in fX, principle
we can inextract
principle extract
all the marginal
all the N marginal distributions X ~ f , where n=1,...,N,
distributions , where = 1 , by n computing the marginal
n X by pdfs
as follows computing the marginal pdfs, as follows:
Z
( ) = (1 ) 1 1 +1 . (10)
R1

Then we can compute the marginal cdfs as in (2). Finally, we can feed
23 RISK PROFESSIONAL O C T O B E R 2 0 1 1 each cdf , which is a function, with the respectivewww.garp.org
entry of the vector ,
namely the random variable . The outcome of this operation are the grades,
which we know from (4) have a uniform distribution on the unit interval
 workNowtheweunivariate
are fully equipped to introduce
results (4) and (7). the copula, by extending to the multi- 1

ade U2
pdf fU1
a variate framework the univariate results (4) and (7).
-dimensional vector of random variables (1 ) x2
Consider a -dimensional vector of random variables (1 ) x2
eneral multivariate distribution represented by its pdf . .
with a fully general multivariate distribution represented by its pdf X fX cdf FX 2
inWe
therecall
multivariate
that in thecase
multivariate isthe
the pdf case dened
pdf inissuchdened a way that,
in such a way that,
potential
for any setjointT H E Q U A N T C L A S S R O O M B Y AT T I L I O M E U C C I
of potential Rvalues
values Xjoint
forX(1 R for ) 1the
( following
) the following
identity holds
Z Z pdf f X 2

{(1 P
) {( )(
X1} X1}
)(
1 1
) .
. 1 (9) (9) x1
X X
joint X=(X1,X2) marginal X2
From the joint
t distribution distribution
we can in principle we can extract in principle all extract
the marginalall the marginal
distributions , where = 1 , by computing the marginal pdfs
, where = 1 , by computing the marginal pdfs
as follows
Z Figure 2: Copulas are non-linear standardizations of multivariate distributions
Z ( ) = and what is left is the pure joint information amongst the Xn's
( 1 ) 1 1 +1 . (10)(10)
) = (1 ) 1 1 +1 .
R 1
(10) i.e., the copula fU. We summarize this statement in an al-
R1 Then we can compute the marginal cdf s FX , as in (2). Fi- ternative, intuitive formulation of the copula.
Then we can compute the marginal cdfs as in (2). Finally, n we canKey feedconcept. the copula is the information missing from the individual
each cdfthe nally, we can feed each cdf F Xn, which is a function, with the re-
compute , which iscdfs
marginal a function,
aswith in the
(2).respective
Finally, entry we can of the
feedvector ,
spective entry of the vector X namely, the random marginals
variable to
Key complete
concept. theThejointcopula
distribution
is the information missing from
ulas:namely theory
the random variable . The
which is a function, with the respective entry of the vector ,
outcome of this operation are the grades,
which
ndom weXknow
variable n. The
from. outcome
The haveofa this
(4)outcome of operation
uniform thisdistribution
operation are the on
arethegrades,
theunit grades,which
intervalwe the individual marginals to complete the joint distribution:
" joint = copula + marginals " (13)
ully equipped know to from
introduce
w from (4) have a uniform distribution (4) havethe a uniform
copula, distribution
by
on extending
the unit onto
intervalthe
the unit
multi- interval
( ) U[01] . (11)
work the univariate results (4) and (7). " joint = copula + marginals " (13)
a -dimensional
However,the vector
mainpoint (of random U[01]variables
) remember
to . is that the (1of
entries 1) (11)
(11)
( )
general
are not multivariate
independent. distribution
Therefore, the represented
joint distributionby its pdf ofthe . The
grades intuitive
is not The denition (13) can (13)
intuitive definition be made can be rigorous. From theFrom
made rigorous. denition of
the main
inuniform point
the multivariate todomain,
on However,
its remember
case the the
which is pdf
main that
is the
point
theunit entries
isto dened
cube
remember ofin
[0 such
1] is ( a1 way
that
the
[0 that,
1], copula
see
)Figure
entries (12)
of 2. the definition of copula (12),
fendent.
potential Therefore,
Ujoint
(Uvalues the joint not
1,...,UN)Xare
distribution
Rindependent. of
the
for (1 Therefore, ) grades
the the is
following
joint notdistri-
domain, which isfUthe unitgradescube [0 1] [0distribution
1],itsseedomain,
Figure 2.the joint () P {1 1 } (14)
Key bution
concept. of The the copula is not
of an uniform
arbitrary on iswhich is
the unit
distribution cube grades... [0,1] (see Figure 2, below).
of itsZ[0,1] = P {1 (1 ) 1 ( ) }

{(
ncept.
1 ) X }
Thecopula of an arbitrary ( 1 )
distribution
1

is . the joint (9) = P 1 1
(1 ) 1
( ) (14)
1 1 (1 ) 1 1

of its gradesFigure 2: Copulas X

.. Non-linear Standardiza- = (1 ) 1
( ) .
( ) (12)
tions of we Multivariate . Distributions 1
t distribution can in principle extract all the marginal
( ) ( ) Dierentiating this expression
, where1 = 11 1 , by computing the marginal pdfs

Differentiating thisweexpression,
obtain Sklars theorem
we obtain (for two
Sklars theoremvariables, the
( .. ) (12) general
(for case
two follows immediately)
variables, the general case follows immediately):
.
Z 1
( ) ( 2 ) = 21 2 (1 2 ) = 21 2 (1 ) 1
(2 )
( ) =The grades of (the
1 distribution
) 1 can be
1
interpreted
+1 as .a sort(10)

of non-linear
generalcase1 follows immediately)
2
1 1

1
2
1
z-score, which forces all the entries to have a uniform distribution on the
R1 = 1 2 1 (1 ) 2 ( 2 )11 1 ( 1 ) 22 (2 )
2 2 1 (1 ) 1
unit interval [0 1]. By feeding each random variable into its own cdf, all (1 2 ) = 1 2 2(1 2 ) =1
12 1 ( 1 2)
1 2 ( 2 ) 2
(2 )
compute the marginal
the information containedcdfs
in each as in (2).
marginal Finally, we
distribution can
is feed
swept away, and = 2
= 12 1 1
(1 ) 2 ( 1 1 (15)
s of the distribution can be interpreted as a sort ofnon-linear
1 1 1
1
1 (1 ) 2 )
2 2
1 (
1 (1 ) 2 2 (2 )
2 12 ) (15)
, which is left
what is a function,
is the purewith theinformation
joint respectiveamongst
entry ofthethevector
s, i.e. ,
the copula 2
1 2 1( 1
(1 )1 1
( (2 )
forces allsummarize
. We the entries to haveinaanuniform
thisstatement distributionformulation
alternative, on the 1 1) 22)
(15)
ndom variable . The outcome of this operationintuitive
are the grades, of the == 1

11 ) 2 1 1( . 2 )
0 1]. By feeding each random variable into its own cdf, all
copula
(
1 1 1( )1 2 2 (2)
w from (4) have a uniform distribution on the unit interval 1 1 1 2 2 2
on contained in each marginal distribution is swept away, and 1
(1 ) 1
(2 )
= 1 1 52 1 .
s the pure joint information U[01] . the s, i.e. the copula
( ) amongst (11) Key ( ) 2 ( 2 )original joint distribution ,
Keyconcept.
concept. 1 Sklars
1 theorem links
Sklars theorem
1 2 the
links the original joint dis-
arize this statement in an alternative,4intuitive formulation of the the copula , and the marginals , or equivalently , as follows
tribution fX, the copula fU and the marginals fXn or, equiva-
the main point to remember is that the entries of (1 ) 1 Sklars theorem
Key concept. the original joint distribution ,
links
lently,
FX n,and
as(follows: 1
1 ) ( )
endent. Therefore, the joint distribution of the grades is not the copula
| , 1 the {z marginals } , or
= equivalently , as follows (16)
s domain, which is the unit cube [0 1] [0 1], see Figure 2. 1 joint
4 1 (1 ) 1
( ) 1 ( ) 1
| ({z 1 ) ( )
(16)
} } | =1 1
1
|
joint
{z {z } .
pure joint 1 pure m arginal 1
ncept. The Keycopula of an arbitrary
concept. distribution
The copula is the
of an arbitrary joint fX
distribution is (1 ) 1 1 (1 ) ( )
of its grades |
Sklars theorem {z
justies }theintuitive
| {z (13).
copula denition } .
the joint distribution fU of its grades, as follows: pure joint pure m arginal

1 1 (1 ) Sklars theorem justies the intuitive copula denition (13).

.. Sklars
Sklars theorem
theorem justifies
provides thethe
pdfintuitive copula
of the copula definition
from (13).and the
the joint pdf
( . ) (12) (12)marginal pdfs. This allows us to use maximum likelihood to t copulas to
empirical
Sklars data. provides the pdf of the copula from the joint pdf and the
theorem
( ) We
Sklars
now
theorem
derive another
provides theIf pdf of the
thegrades
copula fromarbitrary
the
marginal pdfs. This allows useful
us to result. we feedlikelihood
use maximum t into
to(12) copulas to
joint
inverse pdf
cdfs
empirical data.
and
1
, the
we marginal
obtain new pdfs. This
transformed allows
random us to
variablesuse maxi-
with a given
The grades of the distribution fX can be interpreted as ajoint
mum likelihood
distribution, whichto fit
we copulas
denote by to
empirical data.
We now derive another useful result. If we feed the grades (12) into arbitrary

sort of non-linear
s of the distribution can bez-score, which as
interpreted forces allofthe
a sort entries Xn inverse
non-linear to We now
cdfs , derive
1
another
we obtain useful
new transformed
1 1 ( (1 ))
result.random If we feed the
variables withgrades
a given
joint distribution, which we denote by1 1
have a uniform distribution on the unit interval
h forces all the entries to have a uniform distribution on the [0,1]. By feed- (12) into arbitrary inverse cdf
.. s F Y n
-1
, we obtain new trans-
( . ) (17)
ing eacheach
[0 1]. By feeding random variable
random Xn into
variable itsinto
ownits
cdf,own
all the
cdf,informa-
all formed random variables 1
1 (
1 with( a
)) given joint distribution,
1 (1 ( ))
1
on containedtion
in contained in each
each marginal marginal distribution
distribution is sweptfXnaway,
is swept
andaway, which we denote ( by fY .. ) (17)
.
s the pure joint information amongst the s, i.e. the copula The joint distribution has marginals
1 whose cdfs are , which follows from
( ( ))
applying (11) and (6) insequence. Furthermore, the copula of is the same as
marize this statement in an alternative, intuitive formulation of the the copula of , because from the denition of in (17) and the denition of
The joint distribution has marginals whose cdfs are , which follows from
the copula
applying (11) (12) we obtain
and (6) in sequence. Furthermore, the copula of is the same as
www.garp.org the copula of , because from
OC Tthe
O denition
B E R 2of0 1 in
1 (1 ) = 1 (1 )
RISK
1 (17) PROFESSIONAL
and the denition of 24
the copula (12) we obtain ..
4 ( . ) . (18)
( ) = ( )
 (
la , and |the 1 joint
marginals (1) , as
) , or equivalently follows
( )
{z } | 1 1

{z 1 } .
1 pure
( 1 joint
) 1
( pure
) m arginal
( )
1 (1 ) | {z( ) } | 1 1
1 1
{z } = {z (16) } .
lars theorem
joint justies the intuitive copula denition
pure joint (13).
pure m arginal
T H E Q U A N T C L A S S R O O M B Y A T T I L I O M E U C C I
(1 justies
) 1 (1 ) copula 1
denition
(13). ( ) 1
klars
| theorem{z } |the1 intuitive {z } .
klars theorem
pure joint provides the pdf of the copula
pure m arginal from the joint pdf and the
ginal pdfs. This allows us to use maximum likelihood to t copulas to
Sklars
heorem
irical theoremthe
justies
data. provides
intuitivethe pdf of
copula the copula
denition (13). from the joint pdf and the
We now derive another useful result. If we feed thelikelihood
rginal pdfs. This allows us to use maximum grades (12) to into
t copulas
arbitraryto
pirical
rse data.
cdfs 1
, wereturns
obtain new transformed random variables with a given
theorem Compounded
provides the C=(C
pdf of1,C2) copula from the
the Linear
joint returns
pdf R=(R
and1,R 2)
the
We now derive
distribution, another
which we useful
denote result.
by If we feed the grades (12) into arbitrary
dfs. This allows c2 us to use maximum likelihood to r2t copulas to
erse cdfs
data.
f C2 1 , we obtain newreturns
Compounded transformed
C=(C1,C2) random variables withR=(R
Linear returns a given
1,R2)
C f
t distribution, which we denote1 by 1
1 c Ifwe
C
(
R
e C

)) f R (12) into arbitrary

(1grades R fR r the pitfalls of disregarding such differences). For example, if
w derive another usefulf result. feed
1 the
2
2
2

the prices distribution fP is multivariate log-normal, the linear

C2
1
. C f
s 1 f (17)
, we obtain new transformed random e 1 with a given
R
variables C
. C ln
( 1. ( ( )) )
1 R C
f R (17)

bution, which we denote by 1 1
c11 1 1
r
R2 R
returns distribution fR is multivariate shifted-lognormal and
( (... ( )) )
1
C
ln 1 R
(17) the compounded returns distribution fC is multivariate nor-
c1 f R1 r1
fC 1 ( ( )) mal, as illustrated in Figure 3 (see below) for two negatively
1
joint distribution has
1
11
1 . marginals 1
(whose
( cdfs
)) are , which follows from
( (6)The joint
.. distribution ) fY has marginals whose cdfs f are FYn correlated stocks.

f
ying (11) and in sequence. ,F
F follows
Furthermore,
C1
the copula of (17) is the same as R1

ecopula
joint distribution which has from
marginals applying
whose (11) FR, FR (6)
and in sequence. Fur-
of inare
cdfs , which follows from
1 C1 C2

of , because
from ( the( ))
denition (17) and the denition
1
of 2

Figure 3: Linear and Compounded Returns

thermore, the copula F , F of Y isthe the same as theF copula of X,

lying (11) and (6) in sequence. Furthermore,
Copula U=(U ,U
C1
) copula
C2 of is , F same
the as
copula (12) we obtain 1 2
R1 R2

istribution
copula of , has marginals
because
because from whose
fromthethe u2 cdfs
denition are
definition , which
ofof YininU=(U
Copula
(17)follows
(17) ,U
1 2
and
and
) from
the
the definition
denition of
of display Same Copula
11) and (6)
(12)in we
sequence. Furthermore,
the copula the copula u of U is fthe same as
copula obtain 1 ((12),
1 ) = we obtain
1 (1 )
2 U
of , because from the denition fU 2
.. of in (17) )and the denitionU of f
. f 1 (1 ) . (18) U
(12) we obtain ( 1 (1 ) = U2

( ) .. ( )
( ( )= ( .) ) u1 . (18)
(18)
1 1 1 1 u1
.. (copula fU
refore, we derive ( thatthe . ) of )an( arbitrary
.) random variable
(18)f
1
U1
) does not(change when we transform each into a new variable
erefore, ) the copula
( ) of an arbitrary
(we derive
that random
) by means Therefore,
of functions we derive
() that the 1 copula of anvariable
( ()), where are
arbitraryran-
we ) does
derive that not
dom change
variable when we
Xarbitrarytransform
(X1,...,Xreturns
) does each
not into
change a new
when
we variable
trans-
3: the copulaandofcompounded
an Nrandom variable

Figure Linear have samecopula These two types of returns, though calculated on the same securities prices
(not
) does ) by means
changeform
when of
Figure
each functions
we 3:transform
XLinear
n into and
a each
()
new 1
compounded
variable
into
(aY returns
new ()),
n variable
where
gn (Xhave
n) bysamemeans are
copula ofare dierent. Therefore, their distributions are dierent, refer to

,
functions () 61
+1
) by means of functions ( ()), where where Fare arearbitrary cdfs.(2010) for more details on the pitfalls of disregarding such dier-
Yn Meucci
itrary cdfs.arbitrary
It It
is iseasy
cdfs. to verify
easy toItverifyis easy that
that to such
such functions
functions
verify that g
such
are are
a a
very
functions very
broad broad
class
6 n ences.a For
are veryexample,
broad if the prices distribution is multivariate log-normal,
ss, namely class, all the increasing
namely,
namely alltransformations,
all6 the the increasing
increasing also known also
transformations,
transformations, as also
co-monotonic
known known as as
co-monotonic
the linear returns distribution is multivariate shifted-lognormal and the
nsformations. Thus we obtain
transformations.
co-monotonic Thus the wefollowing
obtain the following
transformations. Thus, we obtain the following: compounded returns distribution is multivariate normal, as illustrated in
Figure 3 for two negatively correlated stocks.
Key concept. KeyKeyCo-monotonic
concept. transformations
concept.Co-monotonic
Co-monotonic transformations
= (
transformations )Y= the
nof gHowever,
(Xn)) of the
n( the copula of the linear returns and the copula of the com-
ntries of doentries not
of alter
the
of the
entries copula
do not ofalter
X do ofthenotcopula
alter the of copula of X pounded returns However, the copula
are identical, seeofagain
the linear returns
Figure 3. Thisand the copula
result of sur-
is not
prising, because the compounded
= returns
1 is anare identical transformation
increasing (see, again, Figure 3).
of and
1 1 1 1
is an increasing transformation of Cn .
.
. = ln (1 + This
)result is not surprising, because R n = e
1 is an increas-
( ... ) ( .. ) ( ... ) ( .. )
1 1
ing transformation of Cn, and Cn =1n (1+ Rn) is an increasing
transformation of Rn.
1 1 1 1 3 (19)
Copulas:
(19) practice
& & . .
1 Section 3: Copulas (Practice)
1 The implementation of the copula-marginal decomposition in practice relies on
(19) The implementation of the copula-marginal decomposition
( ... )
( ... ) two distinct processes, which appear in multiple steps in the theoretical discus-
in practice relies on two distinct processes, which appear in
sion of Section 2.
multiple steps
First, the separation process,in thewhich
theoretical
led us discussion of Sectionof2.the copula
to the denition
(12). First, the separation process, which led us to the definition
To illustrate how the copula is not affected by increasing of the copula (12).
To illustrate how consider,
transformations, the copulaasis follows,
not aected
the by increasing
linear returnstransformations,
and
To illustrate how
considerthe copula
the linear isreturns
not aected
and the bycompounded
increasing transformations,
returns between time Key
andThe
the compounded Key theconcept. separation The process S strips an Sarbitrary
strips an distribution
consider the linear
time returns
+ 1 for and thereturns
the same
betweenreturns
compounded
securities
time t and time ttime
between + 1 for and concept. separation process arbitrary
same securities: into its marginals
distribution f and
into its
its copula
marginals f and its copula f , as follows:
time + 1 for the same securities X X n U
+1 +1
+1 1
+1
ln( ). (20) 1
1 ln(
). (20)
(20)




1

S : ( .. )
7 1 (21)
These two types of returns, though 7 calculated on the same .

.. ) , (21)
securities prices Pn,t+17, are different. Therefore, their distribu-
( .


tions are different (refer to Meucci (2010) for more details on

where are the grades

( ) . (22)

25 RISK PROFESSIONAL O C T O B E R 2 0 1 1 www.garp.org

The separation process can be reverted, similarly to the univariate case (6).
By feeding each grade back into the respective inverse cdf 1 we obtain
 irst, the separation process, which led us to the denition of the copula
ration process, which led us to the denition of the copula

The T H E Q U A N T C L A S S R O O M B Y AT T I L I O M E U C C I
Key concept. The separation process S strips an arbitrary distribution
pt.into itsseparation
marginalsprocess
andS strips an arbitrary
its copula distribution
ginals and its copula


1
1
1




1

. S: ( ... )
7 1 (21)
: ( .. ) 7 1
( .. ) ,(21)
.
.

( .. ) ,

e grades
where Un are the grades
here are the grades
( ) . (22)
The implementation of
( ) . (22) (22)
the theory of copulas
processeach
can
The separation process can be reverted, similarly to the
The separation process can be reverted, similarly to the univariate case (6).
univariate
be reverted, case similarly
(6). By feeding each gradecase Un back
relies on two fundamental
into the
eeding grade back into theto the univariate
-1 respective 1 inverse cdf (6). 1
we obtain
rade
enal

(1(X
respective
back into
1,...,X
distribution
theinverse
N) )whose
whose
cdf FXinverse
respective
joint
joint
n , we obtain
ndom variable (1 ) whose joint distribution
distribution
cdf
distribution

is
a
is

random
we
exactly
obtain
exactlythe
variable

original
the
X
steps: "separation"
is exactly the
dis-
:
on : tribution fX:
1 1 1
(1 )
and "combination."
1
1
.. ) 1 71 ( 1 ( ) ..
1

.. ) (
7. (
.. ) .1
)
(23) (23)(23)
. . Implementing all the steps involved in such processes ana-
( )
1

( ) lytically is impossible, except in trivial cases, such as within the
However, we do not need to limit
However, we do not need to limit ourselves to reverting to the original dis-ourselves to reverting to normal family.
otion
notneed the
to limitoriginal distribution
ourselves fX. Copulas
to reverting to the areoriginal
so powerful dis- because Therefore, all practical applications of copulas rely on nu-
. Copulas are so powerful because they can be glued with arbitrary
ulas are so they
powerful can bebecause
glued theywith can be glued
arbitrary with arbitrary
marginal distributions, with merical techniques, most notably the representation of distri-
a technique that generalizes the univariate case (7), which we butions by means of Monte Carlo scenarios.
used to prove 8
rginal distributions, 8 the
with a
co-monotonic
technique that
invariance of the copula (19).
generalizes the1.univariate case (7),
Within the Monte Carlo framework, scenarios that repre-
Accordingly,
_ we start with two ingredients: an arbitrary sent copulas are obtained by feeding joint distributions sce-
d which we copula
used tofUprove thegrades
i.e., co-monotonic
U (U1,...,U invariance of the copula (19).
N), each of which has a narios into the respective marginal cdfs, as in (22), and joint
Accordingly, we start with two ingredients:
uniform distribution and joint distribution structure an arbitrary copula , i.e.
specified
_ scenarios with a given copula are obtained by feeding grades
des (1 ) each of which has a uniform distribution, and joint
by fU; and 2. arbitrary marginal
_ distributions f Xn. Then, we
_ scenarios into the respective inverse cdfs. as in (25).
tribution structure specied by ; and arbitrary marginal distributions .
compute the marginal cdfs' F Xn and their inverses F Xn , and we
-1
en we compute the marginal cdfs , their inverses and we feed eachHowever, even the above numerical operations present dif-
1
feed each grade into the respective marginal cdf. The output ficulties. First, the computation of the cdfs in (22) requires
de into the respective marginal cdf. The output is a -variate random
is a N-varitate random variable X (X1,...,X
iable (1 ) that has the desired copula _ N)that has the univariate integrations, as in (2). Second, the computation
and the desired
desired
rginals . copula f U and the desired marginals f X n . of the inverse cdf in (25) requires univariate integrations fol-
We summarize Weassummarize,
follows this assecond
follows,process.
this second process. lowed by search algorithms. Finally the extraction of the mar-
ginal distributions from the joint distribution in (21) requires
Key concept. _ The combination process C glues arbitrary multivariate integrations, as in (10).
Key concept.
marginalsThe combination
fXn and an arbitrary process
copula glues
C f arbitrary marginals The first two problems, namely the univariate integration
U into a new joint dis-
and an arbitrary copula into a new joint distribution
tribution fX, as follows: and inversion, have been addressed for a broad variety of
parametric distributions, for which cdf and quantile are avail-
1


1 able either analytically or in terms of efficient quadratures.

On the other hand, the multivariate integration to extract
C: 1 7 ( ... ) , (24)
.
( .. )


(24) the marginal from the joint distribution represents a signifi-

cant hurdle, because numerical integration is practical only

in markets of very low dimension N. For larger markets, one
where must resort to analytical or quasi-analytical formulas. Such
where
( ) .
1 formulas are available only for a handful of distributions,
(25) (25)
most notably the elliptical family, which includes the normal
and the Student t distributions.
To implement copulas in practice, we must be able to im- An alternative to avoid the multivariate integration is to
To implementplement copulas in practice,process
the separation we must (21),beinable to to
order implement
obtain suit-the sep-
draw scenarios directly from parametric copulas. However,
tion process (21), in order to obtain suitable copulas, and the combination
able copulas, and the combination process (24), in order to the parametric specifications that allow for direct simulation
cess (24), in order to glue those copulas with suitable marginals.
glue those copulas with suitable marginals. are limited to the Archimedean family (see Genest and Rivest
Implementing all the steps involved in such processes analytically is impos-
e, except in trivial cases, such as within the normal family.
Therefore, all practical applications of copulas rely on numerical techniques,
st notably the representation of distributions by means of Monte Carlo sce-
ios. www.garp.org O C T O B E R 2 0 1 1 RISK PROFESSIONAL 26
Within the Monte Carlo framework, scenarios that represent copulas are
ained by feeding joint distributions scenarios into the respective marginal
 T H E Q U A N T C L A S S R O O M B Y AT T I L I O M E U C C I
The Copula-Marginal
Algorithm is a flexible tool
to implement in practice the
crucial separation step and
combination step with fully
flexible distributions.
(1993)) and few other extensions. Furthermore, the param-
eters of the Archimedean family are not immediate to inter-
pret. Finally, simulating grades scenarios from the Archime-
dean family when the dimension N is large is computationally
challenging.
To summarize, traditional implementations of copulas
mainly proceed as follows: first, Monte Carlo scenarios are
drawn from elliptical or related distributions; next, the sce-
narios are channelled through the respective (quasi- )analyti-
cal marginal cdfs, as in (22), thereby obtaining grade scenar-
ios; then, the grade scenarios are fed into flexible parametric
quantiles, as in (25), thereby obtaining the desired joint sce-
narios.
To avoid the restrictive assumptions of the traditional cop-
ula implementation and circumvent all the above problems,
Meucci (2011) proposes the Copula-Marginal Algorithm
(CMA), which simulates Monte Carlo scenarios with flex-
ible probabilities from arbitrary distributions, computes the
marginal cdfs without integrations, and avoids the quantile
computation. CMA is numerically extremely efficient. For all
the details, the code, and an application to stress-testing with
panic markets, please refer to Meucci (2011).
FOOTNOTE
1. For more background on the subject, the reader is referred to
articles such as Embrechts, A. and Straumann (2000); Durrleman,
:
Nikeghbali and Roncalli (2000); and Embrechts, Lindskog and Mc-
Attilio Meucci is the chief risk officer at Kepos Capital LP. He runs the 6-day "Ad-
Neil (2003). The reader is also referred to monographs, such as
vanced Risk and Portfolio Management Bootcamp" (see www.symmys.com). He is
Nelsen (1999); Cherubini, Luciano and Vecchiato (2004); Brigo,
grateful to Garli Beibi.
Pallavicini, and Torresetti (2010); and Jaworski, Durante, Haerdle
and Rychlik (2010).
27 RISK PROFESSIONAL O C T O B E R 2 0 1 1
REFERENCES
Brigo, D., A. Pallavicini and R. Torresetti, 2010. Credit Models
and the Crisis: A Journey Into CDOs, Copulas, Correlations and
Dynamic Models (Wiley).
Cherubini, U., E. Luciano and W. Vecchiato, 2004. Copula Meth-
ods in Finance (Wiley).
Durrleman, V., A. Nikeghbali and T. Roncalli, 2000. "Which Cop-
ula is the Right One?" Working Paper.
Embrechts, P., A. McNeil and D. Straumann, 2000. "Correlation:
Pitfalls and Alternatives," Working Paper.
Embrechts, P., F. Lindskog and A. J. McNeil, 2003. "Modelling
Dependence with Copulas and Applications to Risk Management,
Handbook of Heavy-Tailed Distributions in Finance."
Genest, C., and R. Rivest, 1993. "Statistical Inference Procedures
for Bivariate Archimedean Copulas," Journal of the American Sta-
tistical Association 88, 1034 1043.
Jaworski, P., F. Durante, W. Haerdle and T. Rychlik (Editors), 2010.
Copula Theory and its Applications (Springer, Lecture Notes in
Statistics - Proceedings).
Li, D. X., 2000. "On Default Correlation: A Copula Function Ap-
proach," Journal of Fixed Income 9, 4354.
Meucci, A., 2010. "Linear vs. Compounded Returns - Common
Pitfalls in Portfolio Management," Risk Professional, "The Quant
Classroom by Attilio Meucci," April, 5254. Article and code
available at http://symmys.com/node/141.
ibid, 2011. "New Breed of Copulas for Risk and Portfolio Manage-
ment," Risk 24, September, 122-126. Article and code available at
http://symmys.com/node/335.
Meucci, A., Y Gan, A. Lazanas and B. Phelps, 2007. A Portfolio
Manager's Guide to Lehman Brothers Tail Risk Model, Lehman
Brothers Publications.
Nelsen, R. B., 1999. An Introduction to Copulas (Springer).
www.garp.org