drbeatak. bierut(dnb) seminarium narodowy bank polski 13 ... · global liquidityas ewi...
TRANSCRIPT
Glo
ba
l li
qu
idit
ya
s E
WI
for
ass
et
pri
ce
bo
om
s: G
5 v
ers
us
bro
ad
er
me
asu
res
Dr
Be
ata
K.
Bie
rut
(DN
B)
Se
min
ari
um
Na
rod
ow
yB
an
k P
ols
ki
13
No
ve
mb
er
20
12
Th
e v
iew
s e
xp
ress
ed
he
re a
re s
ole
ly t
he
re
spo
nsi
bil
ity
of
the
au
tho
r a
nd
sho
uld
no
t b
e i
nte
rpre
ted
as
refl
ect
ing
th
e v
iew
s o
f D
e N
ed
erl
an
dsc
he
Ba
nk
.
Pla
n o
f th
e p
rese
nta
tio
n
•S
ho
rt in
tro
du
ctio
nto
EW
I: N
tSa
nd
LDV
ap
pro
ach
•Li
tera
ture
revie
w:
liq
uid
ity
as
EW
I
•R
HS
va
ria
ble
s: l
iqu
idit
ym
ea
sure
s, g
lob
al
liq
uid
ity
(G5
vs.
bro
ad
er
cove
rag
e)
•LH
S v
ari
ab
les:
ass
et
pri
ceb
oo
ms,
pre
-bo
om
ep
iso
de
s
•E
WI
mo
de
l
•B
ase
lin
e r
esu
lts
•R
ob
ust
ne
ssch
eck
s
•C
on
clu
sio
ns
Ab
stra
ct
We
te
st t
he
pe
rfo
rma
nce
of
vari
ou
s g
lob
al
liq
uid
ity
me
asu
res
as
ea
rly w
arn
ing
in
dic
ato
rs f
or
bo
om
s in
ho
use
an
d e
qu
ity
pri
ces
in 2
0 O
EC
D c
ou
ntr
ies
be
twe
en
19
70
a
nd
20
10
. W
e u
se a
pa
ne
l p
rob
ita
pp
roa
ch t
o lin
k g
lob
al
liq
uid
ity m
ea
sure
s w
ith
a
sse
t p
rice
bo
om
s. W
e t
est
th
e r
ela
tive
pe
rfo
rma
nce
of
glo
ba
l li
qu
idit
y m
ea
sure
s b
ase
d o
n t
wo
ag
gre
ga
tio
n s
che
me
s: t
he
tra
dit
ion
al
me
asu
res,
ba
sed
on
G5
da
ta,
an
d
bro
ad
er
me
asu
res,
ba
sed
on
da
ta f
or
up
to
26
co
un
trie
s/cu
rre
ncy
are
as.
Ou
r re
sult
s sh
ow
th
at
in t
he
la
st d
eca
de
glo
ba
l li
qu
idit
y m
ea
sure
s m
ore
oft
en
o
utp
erf
orm
ed
do
me
stic
me
asu
res
as
ea
rly w
arn
ing
in
dic
ato
rs t
ha
n o
the
rwis
e.
Be
twe
en
th
e t
wo
glo
ba
l li
qu
idit
y m
ea
sure
s, G
5 a
gg
rega
tes
ou
tpe
rfo
rme
d b
roa
de
r g
lob
al
liq
uid
ity m
ea
sure
s. T
he
se
arc
h f
or
the
be
st e
arl
y w
arn
ing
in
dic
ato
rs s
ho
ws
tha
t th
e G
5 r
ea
l n
arr
ow
mo
ne
y g
ap
pe
rfo
rms
be
st f
or
bo
om
s in
ho
use
pri
ces,
wh
ile
th
e
glo
ba
l re
al
pri
va
te c
red
it g
row
th g
ap
pe
rfo
rms
be
st f
or
bo
om
s in
eq
uit
y p
rice
s, e
ith
er
wh
en
ag
gre
ga
ted
ove
r G
5 o
r o
ver
a b
roa
de
r sa
mp
le o
f co
un
trie
s.G
ive
n t
he
ris
ing
im
po
rta
nce
of
the
em
erg
ing
ma
rke
t e
con
om
ies
an
d t
he
de
clin
ing
sh
are
of
G5
in
glo
ba
l li
qu
idit
y, t
he
cu
rre
nt
sup
eri
or
pe
rfo
rma
nce
of
G5
me
asu
res
ma
y
no
t w
arr
an
t th
eir
su
pe
rio
r p
erf
orm
an
ce i
n t
he
fu
ture
. T
he
refo
re,
giv
en
th
e
imp
ort
an
ce o
f g
lob
al
liq
uid
ity m
ea
sure
s in
de
tect
ing
ass
et
pri
ce b
oo
ms,
th
e n
ee
d f
or
con
stru
ctin
g b
roa
de
r g
lob
al
liq
uid
ity
me
asu
res
rem
ain
s w
arr
an
ted
.
Sh
ort
in
tro
du
ctio
nto
EW
I
•E
arl
y W
arn
ing
In
dic
ato
r
•In
dic
ato
r =
a v
ari
ab
le
•W
arn
ing
= t
ha
t (c
orr
ect
ly)
sig
na
ls a
(b
ad
) e
ven
t
•E
arl
y =
in
ad
van
ce (
as
far
as
po
ssib
le)
EW
I: N
tSa
pp
roa
ch
•A
sta
tist
ica
l ap
pro
ach
•B
ina
ry e
ven
t/b
ina
ry s
ign
al (
issu
ed
by t
he
in
dic
ato
r,
dis
cre
te t
hre
sho
ld)
•Ty
pe
1 e
rro
r: n
o s
ign
al i
s is
sue
d w
he
n e
ven
t o
ccu
rs (
C);
Typ
e 2
err
or:
a s
ign
al i
s is
sue
d b
ut
no
eve
nt
occ
urs
(B
).
•S
ele
ct in
dic
ato
r a
nd
th
resh
old
th
at
min
imiz
es
NtS
rati
o =
[B
/(B
+D
)]/[
A/(
A+
C)]
EW
I: L
DV
ap
pro
ach
•A
n e
con
om
etr
ic a
pp
roa
ch
•B
ina
ry e
ven
t/co
nti
nu
ou
s in
dic
ato
r, l
inke
d b
y a
CD
F
P(Y
=1
) =
P(Y
*>
0)
= C
DF
(b0+
b1X
1+
b2X
2)
•P
rob
itm
od
el:
CD
F o
f a
sta
nd
ard
no
rma
l d
istr
ibu
tio
n
•S
ele
ct in
dic
ato
r th
at
is s
tati
stic
ally s
ign
ific
an
t, y
ield
s th
e
be
st m
od
el
fit,
etc
.
EW
I: L
DV
ap
pro
ach
•A
llo
ws
for
test
ing
th
e s
tati
stic
al si
gn
ific
an
ce o
f in
div
idu
al
ind
ica
tors
an
d t
he
co
nst
an
cy o
f co
eff
icie
nts
ove
r ti
me
•A
llo
ws
for
a m
ore
sa
tisf
act
ory
ag
gre
ga
tio
n o
f th
e
ind
ica
tors
in
to o
ne
co
mp
osi
te in
dic
ato
r, t
ak
ing
in
to
acc
ou
nt
corr
ela
tio
ns
am
on
g d
iffe
ren
t va
ria
ble
s
•e
tc.
•N
tSa
pp
roa
ch c
an
be
in
corp
ora
ted
in
LD
V f
ram
ew
ork
(Be
rg a
nd
Pa
ttillo
(19
99
))
Liq
uid
ity
as
EW
I
•B
aks
an
d K
ram
er
(19
99
): G
7 e
xce
ss b
roa
d /
na
rro
w m
on
ey
gro
wth
->
re
al s
tock
ma
rke
t re
turn
s, lo
ng
-a
nd
sh
ort
-te
rm in
tere
st r
ate
s•
Ag
ne
llo
an
d S
chu
kn
ech
t(2
00
9):
pa
ne
l pro
bit
mo
de
l: g
lob
al
liq
uid
ity,
me
asu
red
as
a w
eig
hte
d a
vera
ge
of
bro
ad
mo
ne
y g
row
th
for
all s
am
ple
co
un
trie
s, m
inu
s th
e c
orr
esp
on
din
g d
om
est
ic
ag
gre
ga
te->
ho
use
pri
ce b
oo
ms
in 1
8 in
du
stri
alize
d c
ou
ntr
ies
•G
erd
esm
eie
re
t a
l. (
20
10
): p
an
el p
rob
itm
od
el:
do
me
stic
cre
dit
a
gg
reg
ate
s ->
am
on
g t
he
be
st E
WIs
of
ass
et
pri
ce b
ust
s in
18
ma
in
ind
ust
ria
l co
un
trie
s•
Ale
ssi
an
d D
etk
en
(20
11
): g
lob
al p
riva
te c
red
it g
ap
an
d g
lob
al M
1
ga
p (
de
fin
ed
as
de
tre
nd
ed
rati
os
to G
DP
) ->
th
e b
est
EW
Is f
or
ass
et
pri
ce b
oo
ms
in 1
8 O
EC
D c
ou
ntr
ies
•E
WIs
fo
r fi
na
nci
al/
ba
nk
ing
cri
ses
(Bo
rio
et
al.
, d
om
est
ic c
red
it g
ap
)
No
pa
pe
rs c
om
pa
rin
g p
erf
orm
an
ce o
f g
lob
al
liq
uid
ity
me
asu
res
de
pe
nd
ing
on
co
un
try
co
ve
rag
e
Me
asu
rin
g(g
lob
al)
liq
uid
ity
•C
GFS
Ad
-ho
c G
rou
p t
o i
nve
stig
ate
th
e m
ea
sure
me
nt,
d
rive
rs a
nd
po
licy
im
pli
cati
on
s o
f g
lob
al l
iqu
idit
y•
Ea
se o
f fi
na
nci
ng
•G
lob
al p
riva
te c
red
it +
co
mp
lem
en
tary
ind
ica
tors
So
urc
e:
CG
FS P
ap
er
No
45
“G
lob
al
liq
uid
ity
–co
nce
pt,
me
asu
rem
en
t a
nd
po
licy
im
pli
cati
on
s”,
p.
20
Me
asu
rin
g(g
lob
al)
liq
uid
ity
•R
ea
l b
roa
d m
on
ey
•R
ea
l n
arr
ow
mo
ne
y
•R
ea
l b
an
k c
red
it t
o t
he
pri
vate
se
cto
r
•R
ea
l sh
ort
-te
rm in
tere
st r
ate
•R
ea
l lo
ng
-te
rm i
nte
rest
ra
te
•Te
rm s
pre
ad
-> 3
0 t
ran
sfo
rma
tio
ns/
ind
ica
tors
Glo
ba
l li
qu
idit
y:
G5
vs
bro
ad
er
cove
rag
e
•G
5:
Un
ite
d S
tate
s, e
uro
are
a,
Un
ite
d K
ing
do
m,
Jap
an
a
nd
Ca
na
da
•B
roa
de
r: G
5 +
Au
stra
lia
, B
razi
l, C
hil
e,
Ch
ina
, C
zech
R
ep
ub
lic,
De
nm
ark
, H
un
ga
ry,
Ind
ia,
Ice
lan
d,
Ind
on
esi
a,
Isra
el,
Me
xico
, N
ew
Ze
ala
nd
, N
orw
ay,
Po
lan
d,
Ru
ssia
n
Fed
era
tio
n,
So
uth
Afr
ica
, S
ou
th K
ore
a,
Sw
ed
en
, Sw
itze
rla
nd
, Tu
rke
y
•A
gg
reg
ati
on
: G
DP
-we
igh
ted
ave
rag
es,
usi
ng
PP
P-b
ase
d
GD
P w
eig
hts
(fr
om
IM
F W
EO
). T
he
we
igh
ts a
re
con
sta
nt
(ave
rag
ed
ove
r 1
99
5-2
01
0).
Glo
ba
l li
qu
idit
y:
G5
vs
bro
ad
er
cove
rag
e
•G
5 a
gg
reg
ate
s e
asi
er
to c
on
stru
ct,
ba
sed
on
mo
re r
eli
ab
le
da
ta a
nd
ava
ila
ble
wit
h a
mu
ch l
on
ge
r h
isto
ry.
•S
ha
re o
f G
5 i
n b
roa
de
r a
gg
reg
ate
s st
ea
dil
y d
ecr
ea
sed
, fr
om
a
bo
ut
81
% i
n t
he
la
te 1
99
0s
to a
bo
ut
69
% a
t th
e e
nd
of
20
10
.•
G5
an
d b
roa
de
r li
qu
idit
y m
ea
sure
s d
isp
lay
sim
ila
r b
eh
av
ior
ove
r ti
me
: th
e o
vera
ll c
orr
ela
tio
n b
etw
ee
n t
he
tw
o
ag
gre
ga
tio
ns
–fo
r re
al
gro
wth
or
inte
rest
ra
tes
-is
b
etw
ee
n 0
.47
an
d 0
.86
, w
ith
an
up
wa
rd t
ren
d i
n
corr
ela
tio
n o
ver
tim
e.
=>
op
po
site
co
ncl
usi
on
s re
ga
rdin
g t
he
ne
ed
fo
r a
nd
th
e
use
fuln
ess
of
bro
ad
er
ag
gre
ga
tes
Ass
et
pri
ceb
oo
ms
•P
eri
od
s o
f la
rge
sw
ing
s in
ass
et
pri
ces
his
tori
call
y
ass
oci
ate
d w
ith
pe
rio
ds
of
fin
an
cia
l in
sta
bilit
y (
Bo
rio
an
d L
ow
e (
20
02
), D
etk
en
et
al.
(2
01
0),
Dre
hm
an
ne
t a
l.
(20
11
))•
40
% o
f h
ou
se p
rice
bo
om
s h
isto
rica
lly f
oll
ow
ed
by
bu
sts,
le
ad
ing
to
ou
tpu
t lo
sse
s o
f a
bo
ut
8%
of
GD
P.
25
% o
f e
qu
ity p
rice
bo
om
s fo
llo
we
d b
y b
ust
s, w
ith
lo
sse
s o
f a
bo
ut
4%
of
GD
P(H
elb
lin
ga
nd
Te
rro
ne
s(2
00
3))
•R
ece
ssio
ns
ali
gn
ed
wit
h f
ina
nci
al c
rise
s te
nd
ed
to
be
a
bo
ut
1/3
mo
re c
ost
ly t
ha
n n
orm
al o
ne
s (J
ord
á,
Sch
ula
rick
an
d T
aylo
r (2
01
0))
De
fin
ing
ass
et
pri
ceb
oo
ms
•A
n u
nu
sua
lly s
wif
t a
nd
pe
rsis
ten
t a
sse
t p
rice
in
cre
ase
co
mp
are
d t
o t
ren
d (
Ad
ali
da
nd
De
tke
n(2
00
7),
Ale
ssi
an
d D
etk
en
(20
11
))->
Id
en
tifi
cati
on
of
a b
oo
m/b
ust
?•
Re
al h
ou
se p
rice
s a
nd
re
al e
qu
ity p
rice
s•
20
OE
CD
co
un
trie
s: A
ust
rali
a,
Ca
na
da
, Sw
itze
rla
nd
, D
en
ma
rk,
Un
ite
d K
ing
do
m,
Jap
an
, N
orw
ay,
Ne
w
Ze
ala
nd
, Sw
ed
en
, K
ore
a,
Un
ite
d S
tate
s, G
erm
an
y,
Fra
nce
, It
aly
, S
pa
in,
Gre
ece
, Ir
ela
nd
, F
inla
nd
, B
elg
ium
a
nd
Ne
the
rla
nd
s; Q
1 1
97
0 t
o Q
4 2
01
0
•E
WI:
len
gth
of
the
pre
-bo
om
ho
rizo
n?
Ass
et
pri
ceb
oo
ms/
pre
-bo
om
pe
rio
ds
•B
oo
ms:
ass
et
pri
ce i
nd
ex
exc
ee
ds
the
re
curs
ive
tre
nd
plu
s 0
.5 t
ime
s th
e r
ecu
rsiv
e S
TD
. Tr
en
d c
alc
ula
ted
ove
r a
m
ov
ing
win
do
w o
f 4
0 q
ua
rte
rs (
qu
asi
re
al-
tim
e a
pp
roa
ch),
w
ith
a v
ery
slo
wly
ad
just
ing
HP
fil
ter
(λ=
10
00
00
).
•C
ross
-ch
eck
ed
wit
h H
elb
lin
ga
nd
Te
rro
ne
s(2
00
3),
Bra
cke
(20
11
), C
lae
sse
ns
et
al.
(2
01
1)
an
d A
less
ia
nd
De
tke
n(2
01
1)
•Tw
o p
re-b
oo
m h
ori
zon
s: 2
-6 a
nd
6-1
0 q
ua
rte
rs f
or
rea
l h
ou
se p
rice
s, 1
-5 a
nd
5-9
qu
art
ers
fo
r re
al
eq
uit
y p
rice
s•
Focu
s o
n d
ete
ctin
g g
en
era
l b
uil
d-u
p o
f v
uln
era
bil
itie
s ra
the
r th
an
pre
dic
tin
g t
he
exa
ct t
imin
g o
f a
bo
om
(D
reh
ma
nn
et
al.
(2
01
1))
•B
ori
oe
t a
l.:
pe
rfo
rma
nce
of
EW
Is f
or
fin
an
cia
l/b
an
kin
g
cris
es
imp
rove
s co
nsi
de
rab
ly a
s ti
me
ho
rizo
n i
s lo
ng
er.
EW
I m
od
el
•B
ase
lin
e m
od
el:
pa
ne
l ra
nd
om
eff
ect
s p
rob
it•
20
OE
CD
co
un
trie
s•
Two
pre
-bo
om
ho
rizo
ns
•B
oo
tstr
ap
pe
d e
rro
rs
•Fo
ur
est
ima
tio
ns
sam
ple
s:•
Wh
en
bro
ad
er
liq
uid
ity m
ea
sure
s a
re a
vail
ab
le o
r fr
om
a f
ixe
d p
oin
t in
tim
e (
sho
rte
r sa
mp
les)
•W
ho
le s
am
ple
or
wh
en
G5
ag
gre
ga
tes
are
ava
ila
ble
(l
on
ge
r sa
mp
les)
Re
sult
s: 1
99
9-2
01
0
Rea
lhou
sepr
ices
1R
ealh
ouse
pric
es2
Rea
lequ
ity
pric
es1
Rea
lequ
ity
pric
es2
Bro
adm
oney
Bro
ader
outp
erfo
rms
G5
No
No
Yes
Yes
Bro
ader
outp
erfo
rms
dom
esti
cY
esN
oN
oN
o
G5
outp
erfo
rms
dom
esti
cY
esY
esN
oN
o
Nar
row
mon
eyB
road
erou
tper
form
sG
5N
oN
oX
No
Bro
ader
outp
erfo
rms
dom
esti
cN
oY
esX
No
G5
outp
erfo
rms
dom
esti
cY
esY
esX
No
Pri
vate
cred
itB
road
erou
tper
form
sG
5Y
esY
esN
oY
es
Bro
ader
outp
erfo
rms
dom
esti
cY
esY
esY
esY
es
G5
outp
erfo
rms
dom
esti
cN
oY
esY
esY
es
Shor
tte
rmra
teB
road
erou
tper
form
sG
5N
oN
oY
esN
o
Bro
ader
outp
erfo
rms
dom
esti
cY
esN
oN
oY
es
G5
outp
erfo
rms
dom
esti
cY
esN
oN
oY
es
Lon
gte
rmra
teB
road
erou
tper
form
sG
5N
oN
oN
oN
o
Bro
ader
outp
erfo
rms
dom
esti
cN
oN
oY
esY
es
G5
outp
erfo
rms
dom
esti
cN
oN
oY
esY
es
Ter
msp
read
Bro
ader
outp
erfo
rms
G5
Yes
Yes
No
No
Bro
ader
outp
erfo
rms
dom
esti
cN
oY
esN
oY
es
G5
outp
erfo
rms
dom
esti
cN
oY
esY
esY
es
Re
sult
s: 1
97
0-2
01
0
Rea
lhou
sepr
ices
1R
ealh
ouse
pric
es2
Rea
lequ
ity
pric
es1
Rea
lequ
ity
pric
es2
Bro
adm
oney
G5
outp
erfo
rms
dom
esti
cN
oN
oY
esY
es
Nar
row
mon
eyG
5ou
tper
form
sdo
mes
tic
Yes
Yes
No
No
Pri
vate
cred
itG
5ou
tper
form
sdo
mes
tic
No
No
Yes
Yes
Shor
tte
rmra
teG
5ou
tper
form
sdo
mes
tic
No
No
No
No
Lon
gte
rmra
teG
5ou
tper
form
sdo
mes
tic
No
No
No
No
Ter
msp
read
G5
outp
erfo
rms
dom
esti
cY
esN
oY
esX
Re
sult
s: s
um
ma
ry
•In
16
ou
t o
f 2
4 c
ase
s g
lob
al l
iqu
idit
y m
ea
sure
s o
utp
erf
orm
do
me
stic
va
ria
ble
s (b
roa
de
r m
ea
sure
s 1
2
tim
es,
G5
ag
gre
ga
tes
14
tim
es)
.•
In 8
ou
t o
f 2
4 c
ase
s b
roa
de
r li
qu
idit
y o
utp
erf
orm
s G
5
ag
gre
ga
tes.
•T
he
win
ne
rs:
�R
ea
l G
5 n
arr
ow
mo
ne
y g
ap
(d
evia
tio
ns
fro
m t
ren
d,
rea
l h
ou
se p
rice
s 1
,2)
�R
ea
l G
5 p
riva
te c
red
it g
row
th g
ap
(re
al e
qu
ity p
rice
s 1
, re
al h
ou
se p
rice
s 1
)�
Re
al b
roa
de
r p
riva
te c
red
it g
row
th g
ap
(re
al e
qu
ity
pri
ces
2)
Ro
bu
stn
ess
che
cks
1.
Est
ima
tio
ns
for
G5
co
un
trie
s o
nly
2.
Pa
ne
l fi
xed
eff
ect
s p
rob
itm
od
el
3.
Ro
llin
g e
stim
ati
on
s
4.
Ou
t-o
f-sa
mp
le f
ore
cast
ing
(la
test
bo
om
s)
Re
sult
sro
bu
stn
ess
che
ck 1
: 1
99
9-2
01
0
Rea
lhou
sepr
ices
1R
ealh
ouse
pric
es2
Rea
lequ
ity
pric
es1
Rea
lequ
ity
pric
es2
Bro
adm
oney
Bro
ader
outp
erfo
rms
G5
Ful
lsam
ple
(20)
No
No
Yes
Yes
G5
coun
trie
s(1
3)X
No
Yes
Yes
Nar
row
mon
eyB
road
erou
tper
form
sG
5
Ful
lsam
ple
(20)
No
No
XN
o
G5
coun
trie
s(1
3)Y
esX
XX
Pri
vate
cred
itB
road
erou
tper
form
sG
5
Ful
lsam
ple
(20)
Yes
Yes
No
Yes
G5
coun
trie
s(1
3)Y
esY
esN
oY
es
Shor
tte
rmra
teB
road
erou
tper
form
sG
5
Ful
lsam
ple
(20)
No
No
Yes
No
G5
coun
trie
s(1
3)N
oY
esY
esY
es
Lon
gte
rmra
teB
road
erou
tper
form
sG
5
Ful
lsam
ple
(20)
No
No
No
No
G5
coun
trie
s(1
3)N
oN
oN
oN
o
Ter
msp
read
Bro
ader
outp
erfo
rms
G5
Ful
lsam
ple
(20)
Yes
Yes
No
No
G5
coun
trie
s(1
3)Y
esY
esN
oN
o
Re
sult
sro
bu
stn
ess
che
ck 2
: 1
99
9-2
01
0
Rea
lhou
sepr
ices
1F
ixed
effe
cts
Rea
lhou
sepr
ices
2F
ixed
effe
cts
Rea
lequ
ity
pric
es1
Fix
edef
fect
sR
eale
quit
ypr
ices
2F
ixed
effe
cts
Bro
adm
oney
Bro
ader
outp
erfo
rms
G5
No
No
No
No
Yes
Yes
Yes
Yes
Bro
ader
outp
erfo
rms
dom
esti
cY
esN
oN
oN
oN
oN
oN
oN
o
G5
outp
erfo
rms
dom
esti
cY
esN
oY
esY
esN
oN
oN
oN
o
Nar
row
mon
eyB
road
erou
tper
form
sG
5N
oN
oN
oN
oX
XN
oN
o
Bro
ader
outp
erfo
rms
dom
esti
cN
oN
oY
esY
esX
XN
oN
o
G5
outp
erfo
rms
dom
esti
cY
esY
esY
esY
esX
XN
oN
o
Pri
vate
cred
itB
road
erou
tper
form
sG
5Y
esY
esY
esY
esN
oN
oY
esY
es
Bro
ader
outp
erfo
rms
dom
esti
cY
esY
esY
esY
esY
esY
esY
esY
es
G5
outp
erfo
rms
dom
esti
cN
oN
oY
esY
esY
esY
esY
esY
es
Shor
tte
rmra
teB
road
erou
tper
form
sG
5N
oN
oN
oN
oY
esY
esN
oN
o
Bro
ader
outp
erfo
rms
dom
esti
cY
esY
esN
oN
oN
oN
oY
esY
es
G5
outp
erfo
rms
dom
esti
cY
esY
esN
oN
oN
oN
oY
esY
es
Lon
gte
rmra
teB
road
erou
tper
form
sG
5N
oN
oN
oN
oN
oN
oN
oN
o
Bro
ader
outp
erfo
rms
dom
esti
cN
oN
oN
oN
oY
esY
esY
esY
es
G5
outp
erfo
rms
dom
esti
cN
oN
oN
oN
oY
esY
esY
esY
es
Ter
msp
read
Bro
ader
outp
erfo
rms
G5
Yes
Yes
Yes
Yes
No
No
No
No
Bro
ader
outp
erfo
rms
dom
esti
cN
oN
oY
esY
esN
oN
oY
esY
es
G5
outp
erfo
rms
dom
esti
cN
oN
oY
esY
esY
esY
esY
esY
es
Re
sult
sro
bu
stn
ess
che
ck 2
: 1
97
0-2
01
0
Rea
lhou
sepr
ices
1F
ixed
effe
cts
Rea
lhou
sepr
ices
2F
ixed
effe
cts
Rea
lequ
ity
pric
es1
Fix
edef
fect
sR
eale
quit
ypr
ices
2F
ixed
effe
cts
Bro
adm
oney
G5
outp
erfo
rms
dom
esti
cN
oN
oN
oN
oY
esY
esY
esY
es
Nar
row
mon
eyG
5ou
tper
form
sdo
mes
tic
Yes
Yes
Yes
Yes
No
No
No
No
Pri
vate
cred
itG
5ou
tper
form
sdo
mes
tic
No
No
No
No
Yes
Yes
Yes
No
Shor
tte
rmra
teG
5ou
tper
form
sdo
mes
tic
No
No
No
No
No
No
No
No
Lon
gte
rmra
teG
5ou
tper
form
sdo
mes
tic
No
No
No
No
No
No
No
No
Ter
msp
read
G5
outp
erfo
rms
dom
esti
cY
esY
esN
oN
oY
esY
esX
X
Ro
bu
stn
ess
che
cks:
su
mm
ary
•Fo
r G
5 c
ou
ntr
ies,
bro
ad
er
liq
uid
ity m
ea
sure
s
ou
tpe
rfo
rm G
5 a
gg
rega
tes
mo
re o
fte
n.
Su
pe
rio
r
pe
rfo
rma
nce
of
bro
ad
er
me
asu
res
in t
he
fu
ll
sam
ple
of
20
OE
CD
co
un
trie
s n
ot
du
e t
o a
be
tte
r
corr
esp
on
de
nce
be
twe
en
RH
S a
nd
LH
S v
ari
ab
les.
•F
ixe
d e
ffe
cts
pro
bit
est
ima
tio
ns
larg
ely
co
nfi
rm
ba
seli
ne
re
sult
s.
Co
ncl
usi
on
s
•O
ur
fin
din
gs
bro
ad
ly c
on
firm
re
sult
s b
y A
less
ia
nd
De
tke
n(2
01
1).
•H
ow
eve
r, g
lob
al
pri
vate
cre
dit
pe
rfo
rms
be
st a
s d
ev
iati
on
s o
f th
e a
nn
ua
l g
row
th r
ate
fro
m t
ren
d (
als
o d
iffe
ren
t fr
om
B
ori
oe
t a
l.).
•O
ur
resu
lts
cle
arl
y i
llu
stra
te t
he
im
pa
ct o
f fi
na
nci
al
lib
era
liza
tio
n a
nd
glo
ba
liza
tio
n:
the
pe
rfo
rma
nce
of
glo
ba
l (G
5)
liq
uid
ity
me
asu
res
wo
rse
in
th
e l
on
ge
r sa
mp
le t
ha
n i
n
the
la
st d
eca
de
.•
G5
ag
gre
ga
tes
pe
rfo
rm b
ett
er
rela
tive
ly t
o b
roa
de
r m
ea
sure
s. H
ow
eve
r, t
he
ris
ing
im
po
rta
nce
of
the
em
erg
ing
m
ark
et
eco
no
mie
s a
nd
th
e d
ecl
inin
g s
ha
re o
f G
5 i
n g
lob
al
liq
uid
ity
->
th
e c
urr
en
t p
erf
orm
an
ce o
f G
5 m
ea
sure
s m
ay
n
ot
wa
rra
nt
the
ir s
up
eri
or
pe
rfo
rma
nce
in
th
e f
utu
re.
Th
e
ne
ed
fo
r co
nst
ruct
ing
bro
ad
er
liq
uid
ity
me
asu
res
rem
ain
s w
arr
an
ted
.