Redistributing central bank profits & losses across the eurosystem: The eurosystem's monetary income

Cesa a o, Se gio; Feb e o, Eladio; Pan elopoulos, Geo ge
Wo king Pape
Redis ibu ing cen al bank p o i s & losses ac oss he
eu osys em: The eu osys em's mone a y income
FMM Wo king Pape , No. 104
P o ided in Coope a ion wi h:
Mac oeconomic Policy Ins i u e (IMK) a he Hans Boeckle Founda ion
Sugges ed Ci a ion: Cesa a o, Se gio; Feb e o, Eladio; Pan elopoulos, Geo ge (2024) : Redis ibu ing
cen al bank p o i s & losses ac oss he eu osys em: The eu osys em's mone a y income, FMM
Wo king Pape , No. 104, Hans-Böckle -S i ung, Mac oeconomic Policy Ins i u e (IMK), Fo um o
Mac oeconomics and Mac oeconomic Policies (FMM), Düsseldo
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/301005
S anda d-Nu zungsbedingungen:
Die Dokumen e au EconS o dü en zu eigenen wissenscha lichen
Zwecken und zum P i a geb auch gespeiche und kopie we den.
Sie dü en die Dokumen e nich ü ö en liche ode komme zielle
Zwecke e iel äl igen, ö en lich auss ellen, ö en lich zugänglich
machen, e eiben ode ande wei ig nu zen.
So e n die Ve asse die Dokumen e un e Open-Con en -Lizenzen
(insbesonde e CC-Lizenzen) zu Ve ügung ges ell haben soll en,
gel en abweichend on diesen Nu zungsbedingungen die in de do
genann en Lizenz gewäh en Nu zungs ech e.
Te ms o use:
Documen s in EconS o may be sa ed and copied o you pe sonal
and schola ly pu poses.
You a e no o copy documen s o public o comme cial pu poses, o
exhibi he documen s publicly, o make hem publicly a ailable on he
in e ne , o o dis ibu e o o he wise use he documen s in public.
I he documen s ha e been made a ailable unde an Open Con en
Licence (especially C ea i e Commons Licences), you may exe cise
u he usage igh s as speci ied in he indica ed licence.
h ps://c ea i ecommons.o g/licenses/by/4.0/legalcode
FMM WORKING PAPER
No. 104 • July 2024 • Hans-Böckle -S i ung
REDISTRIBUTING CENTRAL BANK
PROFITS & LOSSES ACROSS THE
EU
ROSYSTEM: THE EUROSYSTEM’S
MONETARY INCOME
Se gio Cesa a o
1
, Eladio Feb e o
2
, Geo ge Pan elopoulos
3
ABSTRACT
Na ional Cen al Banks (NCBs) o he Eu osys em pool p o i s and losses ela ed o
mone a y policy ope a ions o o m he Eu osys em’s so-called ‘mone a y income’. This is
hen edis ibu ed – i.e. alloca ed – among NCBs acco ding o espec i e capi al keys ( he
pa icipa ion sha es o each NCB o he ECB’s capi al). Mone a y income has ele ance o
cu en deba es such as ha conce ning he high iscal cos s o an ample ese e egime as
a esul o he abundan ese es banks hold in he deposi acili y o hei espec i e NCBs.
These cos s a e in ac edis ibu ed h ough he alloca ion o mone a y income. None heless,
exac ly how mone a y income is pooled and subsequen ly alloca ed be ween Eu osys em
NCBs emains a he enigma ic. The aim o his pape is o explo e how mone a y income is
bo h pooled and alloca ed. This seems a use ul ask beyond he a o emen ioned deba e o
dissipa e o he puzzling issues like he cos s o TARGET2 imbalances. A mo e de ailed
dissemina ion om he ele an au ho i ies as o he p ocess by which p o i s/losses a e
pooled and subsequen ly alloca ed is howe e in ou iew wa an ed.
—————————
1 Uni e si à di Siena, I aly. Email: se gio.cesa a [email protected] .
2 Uni e si y o Cas illa-La Mancha, Spain.
3 Uni e si y o Newcas le, Aus alia.
1
Redis ibung cen al bank p ofi s & losses ac oss he Eu osys em: he
Eu osys em’s mone a y income
Se gio Cesa ao, Eladio Feb e o, Geo ge Pan elopoulos*
(This e sion June 2024)
Abs ac
Naonal Cen al Banks (NCBs) o he Eu osys em pool p ofi s and losses ela ed o mone a y
policy ope aons o o m he Eu osys em’s so-called ‘mone a y income’. This is hen edis ibu ed
– i.e. alloca ed – among NCBs acco ding o espec e capi al keys ( he pa cipaon sha es o
each NCB o he ECB’s capi al). Mone a y income has ele ance o cu en deba es such as ha
conce ning he high fiscal cos s o an ample ese e egime as a esul o he abundan ese es
banks hold in he deposi acili y o hei espec e NCBs. These cos s a e in ac edis ibu ed
h ough he allocaon o mone a y income. None heless, exac ly how mone a y income is pooled
and subsequen ly alloca ed be ween Eu osys em NCBs emains a he enigmac. The aim o his
pape is o explo e how mone a y income is bo h pooled and alloca ed. This seems a use ul ask
beyond he a o emenoned deba e o dissipa e o he puzzling issues like he cos s o TARGET2
imbalances. A mo e de ailed disseminaon om he ele an au ho ies as o he p ocess by
which p ofi s/losses a e pooled and subsequen ly alloca ed is howe e in ou iew wa an ed.
* Se gio Cesa ao is ull p o esso o Eu opean mone a y and fiscal policy, Dipa men o di economia polica
e s asca, Uni e si à di Siena (I aly); Eladio Feb e o is associa e p o esso o Economics a he Uni e si y o
Caslla-La Mancha (Spain); Geo ge Pan elopoulos, lec u e in economics, Uni e si y o Newcas le (Aus alia).
We hank Se izio Bilancio om he Bank o I aly o ad ice on specific poin s, and Giuseppe Fe e o and Ma c
La oie o some commen s. The esponsibili y o any emaining e o s o misunde s andings a e, o cou se,
exclusi ely ou s. Co esponding au ho : Cesa a[email protected]
2
1. In oducon
Ae he ise o in e es a es in 2022, naonal cen al banks (NCBs) belonging o he Eu osys em –
simila o he o he cen al banks – began paying conside able in e es o banks on he excess
ese es (ER)1 held by he lae in hei espec e NCB’s deposi acili y (DF) as a consequence o
he se e al ‘balance shee policies’ unde aken in o me yea s.2 A he p esen a e, and gi en
cu en s ocks o excess ese es, banks a e ecei ing sligh ly less han €130 billion pe yea . This has
spa ked off some deba e on how o a oid axpaye s paying a hidden subsidy o so s o comme cial
banks (e.g. Eu opean Pa liamen , 2023; De G auwe and Ji, 2023; McCauley and Pin e , 2024; Tucke ,
2022).
A soluon a là De G auwe and Ji (2024) o inc easing he manda o y ese e coefficien and paying
ze o in e es on equi ed ese es (RR) is p oblemac, since ese es a e une enly dis ibu ed
among he eu o a ea ju isdicons.3 As a consequence, a highe ese e coefficien would in all
likelihood ad e sely affec banks in Sou he n Eu ope whe e ese es a e much less abundan han
in No he n ju isdicons. Mo eo e , i would also appea ha he ample ese e egime is he e o
s ay gi en he endu ing la ge demand o ese es om banks (Åbe g e al. 2021; Schnabel 2023,
2024; Al a illa e al. 2024) which comp esses he space o inc eases o minimum ese e
equi emen s (Hudepohl e al, 2024).4
A ela ed issue conce ns he dis ibuon o he high fiscal cos s o he DF. In he Eu osys em p ofi s
and losses associa ed o mone a y policy ope aons a e pooled by all NCBs o o m he Eu osys em’s
so-called ‘mone a y income’, which is hen subsequen ly alloca ed among NCBs acco ding o hei
1 Rese es a e issued by he cen al bank and a e used by comme cial banks o execu e in e bank paymen s,
o p omp ly mee cash wi hd awals by cus ome s, o comply wi h ese e equi emen s (i in place), as a
financial buffe , and o ulfil egulaons conce ning sa e asse s. Cen al Bank Digi al Cu encies (CBDC) would
gi e he gene al public di ec access o his kind o money (Cesa ao and Feb e o, 2023). Fo a u he
discussion o CBDC, see e.g. Pan elopoulos (2024a), Aue e al. (2024).
2 This includes he nume ous la ge-scale asse pu chase p og ams (aka Quan a e Easing), in addion o
o he non-con enonal measu es (e.g. Ta ge ed Longe -Te m Refinancing Ope aons) (see e.g. Baglioni,
2023).
3 The maldis ibuon may also conce n diffe en banks (say small and la ge) wi hin a single ju isdicon (see
e.g. F icke, G eppmai , and Paludkiewicz, 2024).
4 Two opposi e opinions on he easibili y o a e u n o a sca ce ese e egime (o ‘co ido sys em’) a e,
espec ely, Bo io (2023) and Al a illa e al. (2024).
3
espec e capi al keys ( he pa cipaon sha es o each NCB o he ECB’s capi al).5 This pooling
includes he in e es paymen s on ese es held in he deposi acilies o NCBs.6 Gi en ha – as
no ed abo e – excess ese es a e maldis ibu ed among he eu o a ea ju isdicons, one may hen
in e ha NCBs in he ju isdicons whe e he p opo on o excess ese es is abo e he coun y’s
capi al key sha e he ela ed ‘excess’ cos s o he DF is-à- is he NCBs whe e he po on o excess
ese es is below he coun y’s capi al key. This leads o he queson o whe he ae he pooling
and allocaon o mone a y income p ocess some NCBs a e indi ec ly subsidizing he banking sys em
o o he ju isdicons. This issue has been ecen ly aised by Baglioni (2024). While we de e a ull
discussion o his issue o u he esea ch, in he p esen pape we deal wi h some p elimina y
me hodological noons conce ning mone a y income ha a e oen sides epped o de e ed o a
li e a u e ha we ound o be a he lacking (e.g. Belhocine e al, 2023; Sonnembe g, 2023). In ou
opinion his p elimina y wo k will shed some ligh on he pooling and sha ing o he cos s o he
deposi acili y, which si s wi hin a complex web o financial flows in ol ed in he Eu osys em’s
mone a y income, he eby allowing a mo e comp ehensi e analysis o he possible exis ence o
unwa an ed Sou h-No h financial flows. As a mae o ac , he p ocess by which he mone a y
incomes o NCBs a e pooled and subsequen ly alloca ed is sll a a he lile-known subjec .7
In secon 2 we shall in oduce he concep o mone a y income, while secon 3 discusses in de ail
he single i ems ha en e in i s calculaon, wi h secons 4 and 5 p o iding some hypo hecal
examples as o how mone a y incomes a e pooled and alloca ed be ween Eu osys em NCBs. Secon
6 akes s ock o he me hodological esul s, wi h secon 7 p o iding an explo a o y analysis o he
5 A NCBs’ sha e o ECB capi al is calcula ed using a key which eflec s he espec e coun y’s sha e in he
o al populaon and g oss domesc p oduc o he EU.
6 B oadly speaking, in he Eu osys em comme cial banks ha e wo majo accoun s a hei espec e NCB
whe e hey hold hei ese es: a cu en accoun in which minimum ese e equi emen s a e ulfilled, and
a deposi acili y, whe e any excess ese es abo e and beyond minimum ese e equi emen s a e kep . Unl
he global financial c isis – in a way ha was unconal o he adional co ido sys em – liquidi y wi hin
he minimum ese e deposi accoun we e ypically emune a ed a a a e highe han ha o he deposi
acili y. Du ing he pas decade o so (ae he adopon o a floo sys em), he emune aon on excess
ese es offe ed by he wo accoun s on ER has oen been idencal, meaning ha banks we e indiffe en as
o whe e o hold hem. Howe e , ae in e es a es on ER held in he DF e u ned om mid-2022 in posi e
e i o y, in mid-2023, he in e es a e paid on minimum ese es was b ough o ze o o somewha educe
he cos s o he ample ese e egime (ECB 2023a).
7 This eminds o he uneasiness o economis s when he Ge man economis We ne Sinn aised he queson
o TARGET2, a opic ha equi ed someme o be ully unde s ood (Bindseil and König, 2012; Cesa ao, 2013;
Feb e o and Uxò 2013). Cesa ao (2023) p o ided a p elimina y explo aon o mone a y income in he
con ex o he deba e on he fiscal cos s o he ample ese e egime.

4
pooling and sha ing o mone a y income in he las en yea s o h ee Eu osys em NCBs. Finally, he
conclusion unde lines he main me hodological esul s and summa ises some implicaons o ou
wo k o he ongoing deba e wi h ega d o he fiscal cos s o excess ese es.
2. Mone a y income
Table 1 shows a syn hesis o he P ofi and Losses Accoun s o he Bank o I aly, he Bank o Spain
and he Bundesbank o he yea s 2021-2023. The (pe haps) mos impo an i em con ained in he
P ofi s & Losses (P&L) accoun s is ‘ne in e es income’ (i em 1). This is he esul o in e es paymen
flows, mos o which a e ela ed o mone a y policy ope aons. In e es income and expenses
ela ed o mone a y ope aons a e hen pooled (i em 5.1), ollowing some specific ules, and
con ibu e owa ds he Eu osys em’s ‘mone a y income’ and hen alloca ed (i em 5.2) acco ding o
capi al key. The ne esul o allocang mone a y income is epo ed as i em 5. The aonale o his
p ocess is o pooling and edis ibung hose in e es flows ha o igina ed by decisions aken a he
Eu osys em le el and whose consequences on NCB’s P&L accoun s mus he e o e be sha ed.8
Table 1 – A selec ed syn hesis o he P&L accoun o h ee Eu osys em NCBs (eu o millions)
Sou ce: CB Annual Accoun s.
The definions o mone a y income p o ided by a ious NCBs a e (expec edly) consis en , so we
mainly e e o he one offe ed by he Bank o I aly (2023). The Eu osys em’s mone a y income is he
8 The Bank o Spain and he Bundesbank ans e he bulk o i em 11 (i.e. final p ofi s) in o al o hei
espec e T easu ies as di idends when he ou come is posi e. By con as as i is pa ly owned by p i a e
ins uonal agen s, he Bank o I aly pays axes ou o i em 11, hen ans e s a p opo on o hose p i a e
owne s and, finally, sends wha emains o he go e nmen .
5
esul o he pooling o NCBs’ own mone a y income. This is de i ed om he p ofi s o losses
calcula ed based on ce ain in e es income and expenses ob ained om a lis o ea ma kable asse s
held by each NCB agains a lis composing i s liabili y base. Bo h lis s ega d ope aons associa ed
o he implemen aon o common mone a y policy decisions, so i is unde s andable ela e p ofi s
and losses and pooled a e ulma ely sha ed. The calculaon o a NCB’s own mone a y income should
be ca e ully disnguished om he NCB’s own P&L accoun , as exemplified in Table 1.
I he analogy helps, he calculaon o a NCB’s own mone a y income is simila o an income ax
decla aon in which some p ofi and losses a e epo ed o a highe au ho i y ha will e en ually
bes ow a posi e o nega e ax (a eba e). Fo ins ance, o he Bank o I aly (2023, p. 73), “ he ne
esul o he allocaon o mone a y income in 2022 [was] equal o €2,375 million (…).9 This was he
diffe ence be ween he mone a y income pooled by he Bank, amounng o a nega e €1,162
million, and ha edis ibu ed [i.e. alloca ed] o he Bank, equal o a posi e €1,213 million”. This
final posi e o nega e ax will con ibu e o he final financial esul (i em 11 o Table 1) o he
NCB modi ying he inial ‘ne in e es income’ (i em 1). As he Deu sche Bundesbank (2024)
explains: “The mone a y income o he naonal cen al banks is inially eflec ed in p ofi and loss
i em 1 ‘Ne in e es income’ [see Table 1], while any unequal allocaon among naonal cen al banks
is balanced ou ia p ofi and loss…‘Ne esul o pooling o mone a y income’”.
Impo an ly, any ans e s o and om he ECB a e effec ua ed as ansc ipons wi hin he TARGET2
paymen sys em.10 Fo ins ance, he pooling o mone a y income by he Bank o Spain o he ECB is
finalised by he o me incu ing a TARGET2 liabili y is-a- is he ECB. Also, i is impo an o no e
ha he ECB’s own p ofi and losses do no con ibu e o he de e minaon o he Eu osys em’s
mone a y income: he ECB jus ealloca es he pooled mone a y income o NCBs acco ding o
9 Ne mone a y income as shown in Table 1 includes €5 million, pe aining o he ecalculaon o amoun s o
p e ious yea s.
10 TARGET2 is in p inciple a paymen plao m and T2 imbalances jus accounng en ies. La o sensu, TARGET2
liabilies and claims may be howe e conside ed as cen al bank money ha bo h he NCBs and he ECB issue
(and accep ) o hei ecip ocal in a-Eu osys em paymen s. In his capaci y TARGET2 claims and liabilies
con ibu e o he ne in e naonal in es men posion (NIIP) o a coun y, and a e eco ded wi hin he
financial accoun balance in he balance o paymen s, in he sub-accoun ‘O he In es men ’. Fo a u he
discussion o NIIP, see e.g. Pan elopoulos (2024b).
6
capi al-key, as no ed abo e. In case he ECB is making losses i migh , howe e , ac a e only a pa al
o a nil edis ibuon.11
3. Ea ma kable asse s and he liabili y base
To calcula e i s local mone a y income o be pooled – ha ulma ely o ms pa o he Eu osys em’s
mone a y income – each NCB e e s o in e es income and expenses ela e o a lis o ‘ea ma kable
asse s’ and o a ‘liabili y base’ bo h e e ing o mone a y policy ope aons. We may no e he e ha
hese in e es income and expenses a e included in i ems 1.1 and 1.2 o Table 1. The p ocess o
pooling and allocaon o he Eu osys em’s mone a y income as eflec ed in he ‘Ne mone a y
income alloca ed’ (i em 5 in Table 1) will modi y he inial impac o hese in e es income and
expenses.
Table 2 de ails he componen s o bo h ea ma kable asse s and liabili y base wi h he espec e
in e es a e a which he income o be pooled is calcula ed. This a e is somemes con enonal and
se equal o he a e on main efinancing ope aons (MRO), and no he a e ac ually pe cei ed (see
e.g. Bank o Spain, 2023, p. 63; Bank o Finland 2023, pp. 212-213). Bo h lis s conce n ope aons
ela ed o mone a y policy (including he smoo h unconing o he paymen sys em).
Table 2 – Ea ma kable asse s and liabili y base
Ea ma kable asse s
Liabili y base
(a) lending o eu o-a ea c edi ins uons
elang o mone a y policy ope aons
(effec e in e es a e on efinancing
ope aons)
(a)
(a’) bankno es in ci culaon (i = 0)
(b) secu ies held o mone a y policy
pu poses (public secu ies: iMRO; co po a e
bonds: effec e in e es a e)
(b’) liabilies o eu o-a ea c edi ins uons
ela ed o mone a y policy ope aons
denomina ed in eu os (effec e in e es a e:
iRR and iDF on equi ed and excess ese es,
espec ely).
(c) in a-Eu osys em claims a ising om he
ans e o ese es o he ECB (iMRO)
(d) ne in a-Eu osys em claims esulng
om TARGET2 ansacons (iMRO)
(d’) ne in a-Eu osys em liabilies esulng
om TARGET2 ansacons (iMRO)
11 Mo eo e , we shall no e in secon 4 ha in a couple o cases he ECB egula es in e es paymen s wi h he
NCBs ia mone a y income.
7
(e) ne in a-Eu osys em claims ela ed o he
allocaon o eu o bankno es wi hin he
Eu osys em (iMRO)
(e’) ne in a-Eu osys em liabilies ela ed o
he allocaon o eu o bankno es wi hin he
Eu osys em (iMRO)
( ) a p e-se amoun o gold holdings and
gold ecei ables in p opo on o each NCB’s
subsc ibed capi al key (i = 0)
Sou ce: Bank o I aly (2024, p. 76; Bank o Spain, 2024, p.58)
As a e e ence ule go e ning he diffe ence be ween wha a NCB con ibu es/pools o he
Eu osys em’s mone a y income and wha i ecei es back, he Bank o I aly (2023, p. 74) sugges s:
“The diffe ence be ween he mone a y income pooled by each NCB and he amoun edis ibu ed
o ha NCB, which may be la ge o smalle , …depends on wo ac o s: (a) he fi s (income effec )
ela es o possible diffe ences be ween NCBs ega ding he in e es income ecei ed on specific
ea ma kable asse s and he in e es expense due on some componen s o he liabili y base; (b)
he second (composion effec ) a ises om he ac ha he amoun s o he abo e asse s and
liabilies in he NCBs’ balance shee s do no gene ally coincide wi h hei capi al keys.”
In o he wo ds, i he composion o ea ma kable asse s and liabili y base o each NCB we e in line
wi h hei espec e capi al key, and i he same in e es a e we e applied on each asse o liabili y,
hen a NCB will pool he exac same amoun o mone a y income ha i will subsequen ly ecei e
back ollowing he mone a y income allocaon p ocess. This is no always he case, as we shall see.
We shall also no e ha he menoned ‘ e e ence ule’ is no ob iously (o only ela ely) applicable
o he cases o he ne in a-Eu osys em claims and liabilies ela ed o TARGET2 (i ems d and d’ o
Table 2) and o he allocaon o bankno es (i ems d and d’) (conce ning he unconing o he
paymen sys em a he han mone a y policy ope aons).
I should finally be no ed ha i is no necessa ily ue ha , espec ely, ea ma kable asse s b ing
p ofi s and he liabili y base causes losses. Fo ins ance, wi h e e ence o he ea ma kable asse s
lis ed in Table 2, i em (a) con ains efinancing ope aons ha in he las decade ha e been associa ed
wi h nega e in e es a es, b inging losses o NCBs; con e sely, bank ese es comp ised in i em
(b’) on he liabili y side ha e been emune a ed a nega e a es again in he las decade, b inging
p ofi s o NCBs.
As menoned, in e es acc ued/paid by a single NCB on ea ma kable asse s and liabilies a e pooled
con ibung owa d he Eu osys em’s mone a y income, and hen alloca ed acco ding o capi al key.
Howe e , an addional ule dic a es ha : “Whe e he alue o an NCB’s ea ma kable asse s exceeds
o alls sho o he alue o i s liabili y base, he diffe ence (i.e. he ‘GAP’) is conside ed o bea (o
14
impo an conclusion he e is ha since no hing is sha ed, he en e ac ual e enues ob ained om
public secu ies emains a he espec e NCBs.
To elucida e he poin , we p o ide a simple example p ecisely ocused on public secu ies wi h he
wo NCBs o he p e ious example (Table 5). In his case, a la ge-scale asse pu chase ope aon is
implemen ed whe e a o al o €300 billion o public secu ies a e pu chased, whe e each NCB
pu chases secu ies in i s own ju isdicon acco ding o capi al key, again 1/3 (NCB-A) and 2/3 (NCB-
B). This ope aon leads o he c eaon o ese es (a liabili y) deposi ed in he deposi acili y whe e
comme cial banks ea n a a e iR ( ese es a e included in he liabili y base – i.e. i em (b), as we shall
see in he nex secon).
Table 5 - A public secu ies pu chase ope aon
NCB-A NCB-B
Bonds: +100 R: +100 Bonds: +200 R: +200
We can now calcula e he i ual ne mone a y income on public secu ies o each NCB ha is
pooled and subsequen ly alloca ed. We assume iMRO = 2% and, o he sake o he a gumen , iR =
0%.22 We ha e:
 NCA-A in e es income o be pooled = 100 x 2% = 2;
 NCA-B in e es income o be pooled = 200 x 2% = 4.
To al mone a y income pooled is = 6. This is edis ibu ed acco ding o capi al key; i.e. 2 o NCB-A
and 4 o NCB-B. Ne mone a y income is he e o e ze o o bo h. In o he wo ds, each NCB con e s
o he pool p ecisely wha i ecei es back om i . In he end, in his case he exe cise o pooling and
allocaon has no effec , and no su p isingly since bo h he income and he composion effec s a e
absen . NCB-A will e ain he income on i s capi al key sha e o public secu ies pu chases,
calcula ed a he ac ual in e es a es, and he same will apply o NCB-B.
An income effec would be p esen i we had conside ed co po a e asse s whose income is pooled
a he ac ual a es which plausibly diffe s among ju isdicons. Ano he case is i he public secu ies
pu chasing ope aon had no been implemen ed acco ding o capi al keys bu , say, o escue a
specific ju isdicon, as in he case o he menoned SMP o o D aghi’s amous OMT 2012 s a emen .
22 To ei e a e, in ou examples we assume ha all he i ems, excep o hose unde sc uny, a e i ele an
(o a ze o-in e es a e is applied o hem).

15
In his case i , say, he Bank o I aly we e au ho ised o buy a ce ain amoun o domesc public
bonds, wi h he isk sha ed o e he en e Eu osys em, i seems na u al ha he Bank o I aly would
ha e o sha e wi h o he NCBs he ela e e enues.23 Du ing he APP (launched 2015) and PEPP
(launched 2020) he idea was ins ead ha isk o public secu ies emained mos ly naonal. The
modalies and esul s o Table 5 eflec hese las expe iences: no isk and co esponden e enues
a e sha ed on public secu ies. The Deu sche Bundesbank (2021, p. 75) seems o confi m his
in e p e aon when i w i es:
“I is assumed ha no income is gene a ed om […] he co e ed bonds pu chased unde
he CBPP and CBPP2 as well as he go e nmen bonds (including egional go e nmen
bonds and bonds issued by eligible agencies loca ed in he eu o a ea) pu chased unde
he PSPP and PEPP gene a e income commensu a e wi h he applicable main efinancing
a e, as he ECB Go e ning Council has uled ou he possibili y o pooling he isk and
e u ns a ising om hese ins umen s among he naonal cen al banks.”
A pa al excepon o he uni o m composion ule conce ned PEPP pu chases ha , in some
ins ances, could empo a ily de ia e om a NCB’s capi al key o paci y ensions in specific ma ke s
(see ECB, 2023b, 2024a). Looking a Table 3 we may indeed obse e ha in 2023 wha he Bank o
I aly pooled as income om he ‘Secu ies held o mone a y policy pu poses (no subjec o isk
sha ing)’ (€23,984 million), was only oughly e u ned o i (€21,546). In his specific case I aly’s ne
con ibuon o he pool (€2,438) was p esumably due o he sha ing (a a con enonal a e) o
in e es e enues om he (au ho ised) pu chase o domesc go e nmen asse s abo e i s
assignmen .
A u he complicaon ega ds he case in which asse pu chases we e o be implemen ed by a NCB
in o he ju isdicons o he mone a y union, as anspi ed du ing he APP and PEPP. This gene a es
TARGET2 imbalances and is deal wi h below (poin (d)).
(c) In a-Eu osys em claims a ising om he ans e o ese es o he ECB.
This i em e e s o he claims in eu o ha NCBs ha e eco ded on hei balance shee s is-à- is he
ECB o he ans e o o eign ese es (acco ding o capi al key) in i s a ou when he Eu osys em
was es ablished (o when a new NCB en e s he Eu osys em). In summa y, he ECB includes on he
asse s side o i s balance shee he o eign ese es i has ob ained and in coun e pa a eu o-
denomina ed deb is-à- is he NCBs, which in u n epo a claim in hei balance shee . This claim
23 We guess ha he same applied o he Secu ies Ma ke P og amme, he limi ed p og am o pu chases o
go e nmen bonds by Sou he n NCBs launched du ing he cou se o 2010.
16
is included in ea ma kable asse s in he same way as any o he in a-Eu osys em c edi /debi (such
as hose ela ed o TARGET2 and adjus men s ela ed o he issuance o bankno es, as we shall
sho ly see); i has no edis ibu e effec as he income pooled on hese claims (a he con enonal
iMRO a e) and c edi o each NCB a e bo h aligned o i s capi al key. Table 3 shows ha , o ins ance,
in 2023 bo h he Bank o I aly and he Bank o Spain pooled an amoun o income p ecisely equal o
ha edis ibu ed o hem.
(d) Ne in a-Eu osys em claims esulng om TARGET2 ansacons
In a pa cula se o ci cums ances, paymen s ac oss eu o a ea ju isdicons may gene a e T2
imbalances consisng o claims held by some NCBs is-à- is he ECB mi o ed by liabilies by he
emaining NCBs is-à- is he ECB. On hese liabilies NCBs pay mon hly in e es a he p e ailing
ma ginal in e es a e ha he ECB ans e s o he emaining NCBs. The e o e a a fi s glance, i
would be en isioned ha T2 imbalances en ail losses o some NCBs and symme ic p ofi s o he
emaining NCBs. Howe e he ules conce ning he pooling and allocaon o mone a y income
p esc ibe ha a he end o he yea each NCB can deduc (i.e. mus ans e ) i s agg ega e losses
(p ofi s) om ( o) he Eu osys em’s mone a y income. This has wo esul s:
 (1) T2 imbalances ha e no final effec on he NCB own ac ual P&L accoun since he
deducon (addion) om ( o) he NCB’s ans e o he Eu osys em’s mone a y income
cancels ou he inial T2 losses (p ofi s) incu ed o e he yea ;
 (2) In he calculaon o he Eu osys em’s mone a y income he losses b ough by some NCBs
a e p ecisely cancelled ou by he symme ical p ofi s b ough by he emaining NCBs, so he
ne effec on mone a y income is nil (i.e. no hing is edis ibu ed).
As Cou -Thimann (2013, p. 29) au ho i a ely explained when, a decade ago, he con o e sy o e
TARGET2 blew up:
“Ta ge balances a e de ac o no emune a ed wi hin a cohesi e mone a y union. (…) Fi s , i is
impo an o ecall ha he size o dis ibuon o Ta ge balances ha e no impac on he mone a y
income o he indi idual NCBs wi hin he Mone a y Union. Ta ge balances in fi s ins ance bea
mon hly paymen s a he p e ailing ma ginal in e es a e in he main efinancing ope aons (in
ull allo men equal o he main efinancing a e). These in e es paymen s flow om NCBs wi h
Ta ge liabilies ia he ECB o NCBs wi h Ta ge claims. Howe e , a yea -end, when he NCBs
pool hei mone a y income ne o expenses in he con ex o he income-sha ing scheme, hese
in e es paymen s a e aken in o accoun and hus offse . Sll, in he con ex o pe cei ed isk on
he cohesion o he Mone a y Union, he ac ha he Ta ge balances acc ue he mon hly in e es
paymen s migh be seen as emune ang such isk."
17
The aonale o his e en ually ineffec e accounng ule is p o ided in he las passage: in he
case whe e a coun y wi h T2 liabilies lea es he union du ing he yea – he eby eneging i s T2
deb – i would ha e a leas paid an in e es on i s T2 liabilies du ing he yea (which is ‘ e u ned’
i i doesn’ e en ually lea e). Symme ically, he NCB wi h T2 claims ecei es a emune aon o he
isk i has incu ed o e he yea (o seeing i s T2 claims enegaded), and ‘ e u ns i ’ when, a he
end o he yea , he isk has no ma e ialised.
We may modi y he example o Table 5 in o de o in oduce some T2 imbalances (Table 6). In he
modified example NCB-A pu chases domesc go e nmen bonds in a financial ma ke loca ed in
ju isdicon B – whe e in e naonal in es men unds p esumably ope a e – as pa o a public
secu ies pu chase p og amme. In p acce, NCB-B (say he Bundesbank) buys €100 billion o
coun y A public secu ies in i s ju isdicon (say F ank u ) on he behal o NCB-A (say he Bank o
I aly) which ‘pays’ by booking a T2 liabili y. Acco ding o he ECB, his has been he main sou ce o
TARGET2 imbalances ae 2015 (e.g. Eisenschmid e al. 2017).
Table 6 – NCB-A pu chases coun y A bonds in a ma ke loca ed in he ju isdicon B
NCB-A NCB-B
Bonds: +100 T2: +100 T2: +100 R: +100
Bonds: +200 R: +200
Comme cial bank
R: +300 D: +300
We assume ha secu ies offe no e u n and ese es a e no emune a ed, bu ha T2 liabilies
equi e paying say 2% (ou assumed in e es a e on MRO) o e he yea . I paymen s a e made as
‘accounng anno aons’ (i.e. esul in an inc ease in T2 liabilies), on 31 Decembe , he wo NCB
balance shee s would appea as ansc ibed below in Table 7:
Table 7 - NCBs pay/ ecei e a e u n on T2 liabilies/claims, anno ang such paymen in hei P&L
accoun s
NCB-A NCB-B
Bonds: +100 T2: +102
Equi y: -2
T2: +102 R: +100
Equi y +2
Bonds: +200 R: +200
Comme cial bank
R: +300 D: +300
18
We can now calcula e each NCBs’ mone a y income ha mus be pooled. Gi en he hypo hesis, he
only ea ma kable asse (o liabili y) a e TARGET2 claims/liabilies.
Wi h in e es on T2 claims/liabilies se a iMRO = 2%, p ofi s/losses epo ed o he Eu osys em
would be:
 NCB-A= - 2% x T2 = – 2% x 100 = -2 ( his loss co esponds o a new T2 claim o NCB-A)
 NCB-B = 2% x T2 = 2% x 100 = 2 ( his p ofi co esponds o a new T2 liabili y o NCB-B)
The Eu osys em’s mone a y income would he eby equa e o ze o. The e is no hing o sha e he e.
TARGET2 is also i ele an o he ‘Ne (o g oss) p ofi s o he yea ’ (Table 1) o bo h NCBs since
hei espec e loss o p ofi s (‘posi e o nega e equi y’) – al hough inially coun ed in he ‘ne
in e es income’ (i em 1, Table 1) – a e la e compensa ed by he co esponding deducon in he
‘income decla aon’ o he Eu osys em. De ac o, in e es on TARGET2 claims/liabilies a e he eby
i ele an bo h in he NCBs’ own P&L accoun and o he Eu osys em’s mone a y income: i a NCB
has a T2 liabili y (claim) and makes mon hly paymen s ( ecei es p ofi s) o e he yea , a he end o
he same yea i will deduc hose paymen s (add hose p ofi s) om ( o) he income i con e s o
he Eu osys em’s mone a y income. This is shown below in Table 8. Finally, he TARGET2 losses and
p ofi s con e ed o he Eu osys em’s mone a y income cancel ou so he e is no ‘di idend’ (posi e
o nega e) om TARGET2. Table 8 shows ha he final ou come is analogous o ha o Table 5 in
which pu chases we e done by each NCB in i s own ju isdicon.24
Table 8 - NCBs deduc /add losses/p ofi s om/ o he mone a y income decla ed o he Eu osys em
NCB-A NCB-B
Bonds: +100 T2: +102
Equi y: -2
T2: +102 R: +100
Equi y: +2
T2: +2 (con e al o
losses o MI)
T2: -2 (con e al o
p ofi s o MI)
NET NET
Bonds: +100 T2: +100 T2: +100
Bonds: +200
R: +300
Table 3 (abo e) showed ha in 2023 he Bank o I aly pooled a T2 loss o € 22,769, whe eby i
ecei ed back a ‘ eba e’ o €2,388. Simila esul s a e epo ed by he Bank o Spain. We mus ake
a momen o explain such a esul since we expec ed a ze o ‘ eba e’ (ne - esul ). Le us fi s ecall
ha while he mon hly paymen s on he T2 liabilies we e epo ed among he in e es expenses in
24 I makes sense ha NCB-A should no pay a penal y o ha ing execu ed la ge-scale asse pu chases in he
mos efficien way, e.g. in o he ju isdicons.
19
he Bank o I aly’s ‘Ne in e es income’ (i em 1 o Table 1), he ‘discha ge’ o hese paymen s in he
calculaon o he mone a y income o be pooled wi h he Eu osys em de ac o ze oed hose in e es
expenses o he calculaon o he NCB final financial esul . This is conside ing only he NCBs, as
he e emains a posi e diffe ence, he menoned ‘ eba e’ due o he ECB liabili y posion on
TARGET2.
I mus be ecalled ha he ECB has a debi T2 balance is-à- is he NCBs (no specifically is-à- is
he Banca d'I alia o he Bank o Spain) on which i pays in e es s a he iMRO a e. This in e es income
on TARGET balances is alloca ed on a capi al key basis, wi hin he ne esul o he edis ibuon o
mone a y income. The Bank o I aly in 2023 eco ded, o ins ance, a benefi o 16.85% o he o al
in e es expense paid by he ECB which explains he abo e menoned ‘ eba e’.
The size o he ECB's T2 debi balances a e mainly ela ed o he pu chases o mone a y policy
secu ies ( he ECB pa cipa ed in bo h APP and PEPP by buying 10% o bonds). Eu o a ea
comme cial banks, howe e , ha e accoun elaonships only wi h he NCBs and no wi h he ECB.
As a esul , he pu chases by he ECB we e acili a ed h ough NCBs, acco ding o capi al key, by
booking a T2 liabili y is-à- is each NCB. On hese liabilies he ECB paid an in e es a e ha accoun
o he abo e ‘ eba e’.25 I seems he e o e ha he ECB employs mone a y income o egula e some
paymen s wi h he NCBs.
(e) Ne in a-Eu osys em claims ela ed o he allocaon o eu o bankno es wi hin he Eu osys em.
The issuance o bankno es in he Eu osys em by each NCB should in p inciple be in line wi h hei
espec e capi al key (wi h 8% o he issuance appanage o he ECB). None heless, as he issuance
o bankno es is dependen on he public’s demand (i.e. he demand o bankno es is endogenously
de e mined), whe he a single NCB has issued an amoun o bankno es abo e o below i s capi al
key en lemen is only calculable ex pos .
25 In he no es o he ECB's 2020 balance shee , wi h e e ence o he i em ‘O he liabilies wi hin he
Eu osys em (ne )’, ECB (2021, p. 49) obse es ha “…in 2020 his i em consis ed p edominan ly o he
TARGET2 balances o eu o a ea NCBs is-à- is he ECB. […] The inc ease in he ne TARGET2 liabili y esul ed
mainly om he ne pu chases o secu ies unde he PEPP and he APP, which we e seled ia TARGET2
accoun s.”

20
Any e en ual o e -issuance is penalised, and any unde -issuance ‘ ewa ded’, a he a e applied o
efinancing ope aons.26 Ye as in he case o T2, any possible p ofi (o loss) as a consequence o a
NCB unde -issuing (o e -issuing) bankno es ealised o e he yea is de ac o dele ed, as p ofi s
made by unde -issuing NCBs a e p ecisely offse by he losses incu ed by o e -issuing NCBs, so om
he poin o iew o pooled Eu osys em mone a y income he e is no hing le o edis ibu e once
he pooling and allocaon p ocess is comple e.
Fo ins ance, suppose ha gi en he demand o bankno es in he espec e ju isdicons, NCB-A
issues €80 and NCB-B €220 billion o bankno es. Wi h he espec e capi al keys o 1/3 and 2/3, he
ex-pos en lemen s would be 100 and 200 espec ely. NCB-B will hen pay a penal y du ing he
yea o he Eu osys em on i s o e issue (20€) a he MRO a e (i.e. iMRO) by booking a TARGET2
liabili y and NCB-A will be emune a ed in a symme ical manne due o i s unde issuance (20€) (by
way o posing a TARGET2 claim). Ha ing paid a penal y, NCB-B will epo a loss o he Eu osys em’s
mone a y income, while NCB-A will epo a p ofi . The ne effec o bo h NCB-A and NCB-B is nil.
As in he case o T2, such accounng ules a e en o ced as i a eu o a ea membe we e o lea e he
mone a y union ae ha ing o e -issued eu os du ing he yea ela e o i s espec e capi al key
(in his case, NCB-B), i would ha e a leas paid in e es du ing he yea (which is e u ned i i
doesn’ lea e).
In e esngly, Ge many o e issues bankno es (see e.g. Deu sche Bundesbank, 2022). Fo example, a
he end o 2023, he Deu sche Bundesbank (2024, p. N/A) epo ed ha while in p inciple i could
dis ibu e €377,036 million o bankno es, he alue o he eu o bankno es ac ually issued was
€920,705 million (€543,670 million mo e han i s alloca ed amoun ). On his diffe ence an in e es
a e equal o he iMRO a e was paid by he Bundesbank o e he yea ( ha ia ECB was ans e ed
o unde -issuing NCBs). The Bundesbank subsequen ly pooled his loss ia he pooling o i s
mone a y income. On he o he hand, he Bank o I aly in 2023 issued bankno es below i s
assignmen . Table 3 abo e shows indeed ha he Bank o I aly pooled €1,877 million o mone a y
income wi h ega d o bankno e issuance and ecei ed back a nega e ‘ eba e’ o €-812 million. To
his end, he I alian cen al bank epo ed in i s P&L accoun a loss conce ning he allocaon o
bankno es o €-2,689 million. An e en la ge pos -pooling loss was epo ed by he Bank o Spain
26 ‘The espec e sha e o he o al alue o eu o bankno es in ci culaon due o each cen al bank in he
Eu osys em is calcula ed on he las business day o each mon h in acco dance wi h he key o allocang
eu o bankno es’ (Deu sche Bundesbank, 2022, p. 42).
21
ha pooled €6,933 million and ecei ed back a nega e ‘ eba e’ o €-570 million o a final loss o
€-7,503 million.
Why he €-812 million and €570 million esiduals o he Bank o I aly and he Bank o Spain,
espec ely? As much as in he case o TARGET2 esiduals, he e should be ze o esiduals. We mus
e e again o he ole o he ECB. Al hough he ECB is alloca ed an 8% sha e o he o al alue o he
eu o bankno es in ci culaon, he ECB does no echnically pu bankno es in o ci culaon. Bankno es
a e always pu in o ci culaon by he NCBs ha ha e a elaonship wi h he eu o a ea banking
sys em. The alue o bankno es in ci culaon ound in he balance shee s o he NCBs is a
con enonally adjus ed alue: i he o al alue o bankno es in ci culaon o he whole Eu osys em
is 100, 8% is con enonally alloca ed o he ECB and he emaining 92% is alloca ed o he NCBs in
p opo on o hei espec e capi al keys. The diffe ence be ween he amoun o bankno es issued
by each NCB acco ding o capi al key and he ac ual amoun issued is offse by in a-Eu osys em
claims/liabilies (diffe en om TARGET2 balances bu always emune a ed a he MRO a e). The
ela e p ofi s o losses a e hen pooled and sha ed ia mone a y income, an ope aon ha in
p acce cancels hem ou . The ‘ esidual’ conce ns he ECB ha , agains he 8% o bankno es on i s
balance shee , ma u es a claim agains he NCBs o he same amoun . Agains his claim, all NCBs
pay in e es o he ECB (€-4,817m in 2023) acco ding o capi al key; he Bank o I aly's sha e (capi al
key o 16.85%), o ins ance, co esponds p ecisely o a paymen o €-812m.
( ) A p e-se amoun o gold holdings and gold ecei ables in p opo on o each NCB’s subsc ibed
capi al key.
I was decided o include gold among ea ma kable asse s, o a o al amoun fixed o he en e
Eu osys em and b oken down o each NCB acco ding o he capi al key. The e o e, since o he
pu poses o mone a y income calculaon each NCB includes an amoun o gold (and gold-equi alen
claims ans e ed o he ECB) aligned wi h he capi al key, he e a e no edis ibu e effec s – he
mo e so since, as seen in Table 2 “Gold is no conside ed o gene a e in e es ” (Bank o I aly, 2023,
p. 74).
5. Calculaon o ne mone a y income: liabili y base
(a’) Bankno es in ci culaon
Refinancing ope aons and he o he sou ces o liquidi y ha we find among he ea ma kable asse s
(poin (a)) mus be ma ched by co esponding liabilies. Refinancing ope aons c ea e base money
22
ha consis o bankno es and ese es. Bankno es do no gene a e any so o in e es a e pe se
so his i em is i ele an o he ac ual o maon o mone a y income (none heless as di ulged
abo e, he une en allocaon o bankno es among NCBs leads o in e es paymen s, bu he ela ed
p ofi s and losses a e la e ze oed h ough pooling e c.).
(b’) Liabilies o eu o-a ea c edi ins uons ela ed o mone a y policy ope aons denomina ed
in eu os
This i em conce ns he o he componen o base money consisng o manda o y (o equi ed)
ese es (RR) and excess ese es (ER). In a classical co ido sys em (o sca ce ese e egime), RR
a e held by banks a hei espec e NCB, and yield an in e es a e equi alen o ha applied o
MROs, while excess ese es a e no mally held in he deposi acili y (DF) bu yield a lowe in e es
a e.
This applies also o a floo sys em (o ample ese e egime, see ECB, 2024b), bu a iaons a e
possible. As al eady no ed, o ins ance, gi en he high e u ns banks we e ecei ing on excess
ese es held in he deposi acili y in he Eu osys em, he ECB since July 2023 b ough he
emune aon o RR o ze o so as o educe ‘ he o e all amoun o in e es ha needs o be paid on
ese es in o de o implemen he app op ia e s ance’ (ECB, 2023a) – a ‘mini ie ing‘ a la De G auwe
(2023). Be his as i may, he losses NCBs bea on emune a ed ese es a e pooled and sha ed.
Impo an ly, he dis ibu ion o excess ese es and ela ed losses a e no howe e in line wi h
espec i e capi al keys. And since he in e es a e applied is uni o m, a composi ion e ec is ac i e
he e, bu no o an income effec .
Le us ake he example o Table 6 ( ep oduced below as Table 9) in which NCB-A pu chases
domesc go e nmen bonds in a financial ma ke loca ed in a o eign ju isdicon (i.e. in ju isdicon
B) as pa o a public secu ies pu chase p og amme. Fo simplici y we suppose ha manda o y
ese es do no exis and ha all ese es a e deposi ed in a deposi acili y ha yield a posi e
in e es a e, say, iDF = 1%. As usual we neglec he o he i ems (o assume hey yield a ze o-in e es
a e).
Table 9 – NCB-B bea s he cos s o excess ese es
NCB-A NCB_B
Bonds: +100 T2: +100 T2: +100 R(DF): +100
Bonds: +200 R(DF): +200
Comme cial bank
23
R: +300 DF: +300
We can now calcula e each NCBs’ mone a y income o be pooled. Gi en he hypo hesis, he only
liabili y base’s componen p oducing an income (o loss) is he deposi acili y. I he DF yields a
e u n o comme cial banks o , say iDF = 1%, each NCB’s mone a y expenses o be pooled would
he e o e be:
 NCB-A = - 1% x R(DF) = – 1% x 0 = 0
 NCB-B = - 1% x R(DF) = -1% x 300 = -3
The Eu osys em’s mone a y income would equa e o -3, which will be edis ibu ed acco ding o
capi al keys; i.e. NCB-A would ecei e -1 and NCB-B -2. NCB-A will ecei e a ne mone a y income o
-1 – 0 = - 1. Hence, e en hough NCB-A would ini ially epo in i s P&L accoun ze o in e es income,
ollowing he pooling and alloca ion p ocess i would incu a ne mone a y income (a loss) o -1. By
con as , NCB-B will pos a ne mone a y income o -2 – (-3) = + 1 in he a e ma h o he pooling
and alloca ion p ocess; i.e. NCB-B epo ed in he P&L accoun an in e es expense o -3, bu a e
he posi i e esul o he ne mone a y income will see he loss educed o -2. In his way, he cos
o he excess liquidi y is e-p opo ioned so ha each NCB bea s i s k-key sha e (in he pe iod 2015
o mid-2022 a nega i e in e es a e penalised excess ese es, so ha NCBs collec ed a posi i e
income; his was hen pooled and alloca ed as pe he usual p ocess).
(d’) Ne in a-Eu osys em liabilies esulng om TARGET2 ansacons
This i em has al eady been deal wi h in he ea ma kable asse side.
(e’) Ne in a-Eu osys em liabilies ela ed o he allocaon o eu o bankno es wi hin he
Eu osys em
This i em has also al eady been deal wi h in he ea ma kable asse side.
6. Some me hodological esul s
The Eu osys em's mone a y income is he esul o each NCB pooling hei in e es income and
expenses associa ed wi h mone a y policy ope a ions o he unc ioning o he paymen sys em.
This pooled mone a y income is hen sha ed/alloca ed acco ding o each espec i e NCB’s capi al
key. The e will be a edis ibu ion i ei he a composi ion and/o an income e ec a e p esen . A
30
Table 12: Bank o Spain – selec ed da a om he P&L accoun s (millions o eu o)
As is widely known, 2015 o mid-2022 was cha ac e ised by a ze o a e on he MRO, a nega e a e
on he DF in e es a e policy, and by an agg essi e balance shee policy (Ros agno e al., 2021). On
he one hand, nega e in e es a es on excess liquidi y p oduced an in e es income o NCBs. On
he o he hand, his was accompanied by inc easing in e es expenses due o he nega e a es on
longe e m efinancing ope aons ( his cos was pa cula ly acu e be ween 2020 and mid-2022
when seeking ecou se o ECB c edi by banks was pa cula ly a ac e).
O e his pe iod, he Bundesbank (Table 10) mos ly incu ed a nega e ne mone a y income
ollowing he pooling and allocaon p ocess, while ne mone a y income was ins ead posi e o
he o he wo NCBs (Tables 11 and 12). In his same pe iod, in e es paymen s on T2 imbalances
(despi e T2 imbalances inc easing) and on unde /o e -issuance o bankno es had a negligible o nil
ole, gi en ha iMRO was ze o ( he eade should ecognise by now ha , wha e e he le el o iMRO,
he ne impac o TARGET2 on he P&L accoun , once conside ed mone a y income, is e en ually
ze o). The eade will as well acknowledge ha in e es e enues on he no isk-sha ing public
secu ies pu chases p og ams will no lead o any edis ibuon. The menoned pae n o
mone a y income was he likely esul o a ela ely la ge eso o sou he n banks o TLTRO
ope aons (b inging in e es expenses o sou he n NCBs), and o he modalies o APP and PEPP
ha gene a ed abundan ese es in some no he n ju isdicons ( he eby p oducing in e es

31
e enues o no he n NCBs). Th ough he pooling and eallocaon o in e es cos s and e enues,
his double mo emen gene a ed a no h o sou h NCB mone a y income edis ibuon.
Line 21 o Tables 10-12 p o ides a ough indicaon o he ne in e es flows o e he pe iod be ween
he local comme cial banks and hei espec e NCBs.29 Because o he easons p o ided in he
p e ious pa ag aph, du ing he pe iod 2015 o mid-2022 he ne in e es flow was a ou able o he
Bundesbank is-à- is domesc banks, whe eas o he wo sou he n NCBs he in e es flow was
a ou able o local banks.
Things changed adically wi h he ise o in e es a es om mid-2022. NCB e enues om
efinancing ope aons e u ned in 2023 o posi e e i o y, pa cula ly o he Bank o I aly gi en
he la ge eso o local banks o TLTROs. The cos s o excess ese es held in he deposi acili y
also became exo bi an . This ou come spa ked off he deba e ega ding he fiscal cos s o he ample
ese e egime (as no ed ea lie in he pape ). Cos s ega ding ne ans e s o banks (as shown in
he las line in Tables 10-12) ha e been pa cula ly one ous o he Bundesbank, modes o he
Bank o Spain, and negligible o he Bank o I aly (whe e ecou se o TLTRO was ex ensi e).
Wi h espec o he esul o pooling and allocaon ia mone a y income o he high DF expenses
(especially o he Bundesbank) and o he TLTRO e enues (pa cula ly o he Bank o I aly) om
mid-2022, NCB mone a y income was edis ibu ed om he sou h o he no h, a leas o I aly,
albei no o Spain, which had an amoun o ese es close o i s capi al key.30
As explained abo e, wi h ega d o T2 imbalances and he misallocaon o bankno es, he p ocess
by which mone a y incomes a e pooled and alloca ed has he uncon, de ac o, o cancelling ou
p ofi s o losses om he NCBs’ P&L accoun s on hese i ems. The eallocaon was, in a sense,
un a ou able o he Bundesbank, pa ally miga ed by a eallocaon o mone a y income in i s
a ou due o he high in e es expense on i s o e -issuance o bankno es.
Despi e sha ing wi h he Bundesbank he high cos s o excess ese es, he pooling o expenses
associa ed wi h T2 liabilies, as well as he al eady menoned olume o ese es held in he deposi
acili y, main ained he Bank o Spain’s ne mone a y income in posi e e i o y. Al hough he Bank
29 This is a ough indica o since he local in e es income and expenses o he NCBs a e hen pooled and
ealloca ed h ough mone a y income (while o cou se symme ic income and expenses o comme cial banks
would no change).
30 Table 3 abo e shows ha in 2023 he Bank o I aly en e ed in i s P&L accoun a final loss, ae pooling and
eallocaon, conce ning he deposi s o banks o €14,123 million, a figu e la ge han i s inial loss o €7,850.
The Bank o Spain epo ed a final loss €7,263 agains an inial loss o €8,160 (Bank o Spain, 2024, p. 64).
32
o I aly sha ed wi h Spain a la ge deb o posion in e ms o T2, a ela ely low amoun o pooled
expenses on bank ese es plus a conspicuous ‘GAP’ b ough I alian ne mone a y income o he
fi s me in o nega e e i o y.31 Fu he , while emaining mu ed o he Bank o I aly, ne ans e s
o banks (line 23) accele a ed in he las wo yea s o he Bank o Spain and especially o he
Bundesbank. All in all, Eu opean banks ha e been, ne e heless, ne winne s is-à- is hei
espec e NCBs (in he sense ha ecen mes ha e mo e han compensa ed o he less
‘ad an ageous’ yea s – i.e. om 2015 o mid-2022).32
All his conside ed, bo h he Bundesbank, he Bank o I aly and he Bank o Spain ha e wi nessed
hei final p ofi s (line 18) wo sening he inial ne in e es income (i.e. p e-pooling, line 1; see also
Table 1). To his end, in 2023 all NCBs eso ed o hei accumula ed financial buffe s o b ing
ans e s o he S a e o ze o (Bundesbank and Bank o Spain), ha emain nega e o he Bank o
I aly once ae a nega e ax con ibuon is conside ed.33 Bo h in 2021 and 2022 o which
Bundesbank da a a e also a ailable, he Bank o Spain and he Bank o I aly appea o ha e pu sued,
in connui y wi h p e ious yea s, a qui e p udenal financial p o ision policy compa ed o F ank u .
Rega ding p ofi ans e s o he S a e (line 20), o e mos o he pe iod conside ed all NCBs pos ed
a ou able esul s, p edominan ly as a esul o APP and PEPP policies; he Bundesbank and he Bank
o I aly could e u n o he S a e he in e es income ea ned on he pu chased secu ies, neu alising
he de ac o fiscal cos s associa ed wi h his po on o public deb ( hough no he Bank o Spain,
wi h pa o he diffe ence unding financial buffe s and ano he pa i s ne balance wi h banks).
None heless in he las wo yea s, such ans e s o he go e nmen ha e d amacally sh unk and
nega e ans e s ha e been a oided (bu no o he Bank o I aly) so a by NCBs e oding hei
accumula ed financial buffe s, as no ed abo e. Conside ing NCBs as pa o he consolida ed public
sec o , his e osion equa es none heless o an o e all loss o public ne weal h (Cecche and
Hilsche 2024, p. 6). The majo ac o unde pinning his esul has o do wi h he imp essi e
emune aon o he la ge s ocks o excess ese es, which mone a y income has spammed o e all
31 This high GAP was possibly due o a high alue o ANFA (see oo no es 12 and 14 abo e). The ‘Annual
a e age Ne Financial Asse s acco ding o ANFA’ in 2023 was (in EUR billion) €111.0 o I aly, €-124.0 o
Ge many, €-22.5 o Spain (sou ce: ECB).
32 I alian banks a e an excepon in 2023 because o hei modes s ock o excess ese es.
33 Final (g oss) p ofi s o he Bank o I aly we e €-7,125 million in 2023. The ax deducon o he bank was
€2,340 million and ecou se o financial buffe s €5,600 million. We he eby ob ain a ne (posi e) p ofi o
€815 million mos o which he Bank o I aly ans e ed o he T easu y. Howe e , wi hou he nega e ax,
ne p ofi s would ha e been € -1,525 as indica ed in Tables 1 and 10.
33
NCBs de e mining some no i ele an subsidies om some NCBs o he banking sys em o o he
ju isdicons.
8. Conclusion
This pape has shown he na u e, composion and aonale behind he p ocess by which mone a y
incomes a e pooled and subsequen ly alloca ed be ween Eu osys em NCBs, which hus a has been
an almos unknown elemen o Eu o a ea mone a y go e nance. Ou con ibuon will be help ul o
dissipa e many ambiguies/omissions p esen in some ecen pape s ha ha e ouched his opic
upon wi h ega d o, o ins ance, p esumed in e es paymen s associa ed wi h TARGET2 imbalances
and o he une en issuance o eu o bankno es. Analogously, we also shed ligh on ela ed
ambiguies, namely su ounding he pooling o in e es income om secu ies pu chased du ing
he cou se o a ious la ge-scale asse pu chase p og ammes. Mo eo e , he pape has cla ified
quesons conce ning he pooling and eallocaon o in e es income and expenses associa ed wi h
efinancing ope aons and excess ese es. This las aspec has a di ec impac on he ongoing
deba e o e he exo bi an fiscal cos s o excess ese es in he Eu o a ea (and elsewhe e), which
has so a been omied a ull discussion as o he ole o how Eu osys em NCB mone a y income’s
a e pooled and alloca ed. While pos poning a comple e policy discussion o his issue, we may a gue
he e ha once he pooling and allocaon o mone a y income is gi en ull conside aon, he une en
cos s o excess ese es due o hei i egula dis ibuon in he Eu o a ea b ings abou a sou h o
no h edis ibuon, as ancipa ed in Cesa ao (2023) and a gued by Baglioni (2024). We also show
ha nega e in e es a es on excess ese es inflic ed on banks in he las decade canno jus y he
p esen huge emune aon which is cu en ly leading o he depleon o NCBs accumula ed
financial buffe s. Taking in o accoun ha banks benefi ed in he same pe iod by nega e a es on
efinancing ope aons, a p elimina y glance o he da a confi ms ha he p esen ne ans e s o
banks is no jusfied a leas om his poin o iew. Indeed, how o econcile a mone a y policy
egime based on ample ese es while simul aneously minimising he fiscal implicaons o excess
ese es is sll an open and challenging queson.
34
Re e ences
Åbe g, P., M. Co si, V. G ossmann-Wi h, T. Hudepohl, Y. Mudde, T. Rosolin, F. Schobe (2021),
Demand o cen al bank ese es and mone a y policy implemen a ion amewo ks: he
case o he Eu osys em, ECB Occasional Pape Se ies No 282 / Sep embe .
Al a illa, C., M. Ros agno and J. Schumache (2023), ‘DP18581 Ancho ing QT: Liquidi y, c edi and
mone a y policy implemen aon‘, CEPR Discussion Pape No. 18581. CEPR P ess, Pa is &
London.
Aue , S., B anzoli, N., Fe e o, G., Ila i, A., Palazzo, F., e Rainone, E. (2024) CBDC and he banking
sys em, Banca d’I alia- Quesoni di Economia e Finanza (Occasional Pape s), n. 829.
Baglioni, A. S. (2023). Mone a y policy implemen aon: Which "new no mal"? Jou nal o
In e naonal Money and Finance, ol. 141, 1-19.
Baglioni, A. (2024) Implemenng mone a y policy wi h excess ese es: fiscal implicaons o he
eu o a ea, IEP@BU Policy B ie , Ma ch.
Bank o I aly (2023) Annual Accoun s 2022, Rome.
Bank o I aly (2024) Annual Accoun s 2023, Rome..
Bank o Spain (2024), Annual Accoun s 2023, Mad id.
Bank o Finland (2023) Bank o Finland Annual Repo 2022, Helsinki (a ailable online).
Belhocine, N., A.V. Bha ia, and J. F ie (2023) “Raising Ra es wi h a La ge Balance Shee : The
Eu osys em’s Ne Income and i s Fiscal Implica ions,” IMF Wo king Pape WP/23/145.
Bindseil, U., & König, P. J. (2012). TARGET2 and he Eu opean so e eign deb c isis. C edi and Capi al
Ma ke s–K edi und Kapi al, (2), 135-174.
Bo io, C. (2023) Ge ing up om he loo , BIS Wo king Pape s No 1100 May.
Cecche i, S.G., and Hilsche , J. (2024) Fiscal Consequences o Cen al Bank Losses, NBER Wo king
Pape Se ies, n. 32478.
Cesa ao, S. (2013) The implicaons o TARGET2 in he Eu opean balance o paymen c isis and
beyond, Eu opean Jou nal o Economics and Economic Policies: In e enon, 10 (3), pp. 359–
382.
Cesa ao, S. (2020) He e odox Challenges in Economics – Theo ecal Issues and he C isis o he
Eu ozone, Sp inge Na u e Swi ze land, Cham.
Cesa ao, S. (2023) Anno azioni sull’implemen azione della polica mone a ia: ie i, oggi, domani,
Quade ni DEPS n. 904.
Cesa ao, S. and E. Feb e o (2023) Cen al Bank Digi al Cu encies: a p ope eacon o p i a e
digi al money, Re iew o Keynesian Economics, 11 (4).
Co si, M., and Muddle, Y. (eds.) (2022) “The use o he Eu osys em’s mone a y policy ins umen s
and i s mone a y policy implemen aon amewo k in 2020 and 2021”, Eu opean Cen al
Bank Occasional Pape Se ies no. 304, Sep embe .
Cou -Thimann, Philippine (2013) Ta ge Balances and he C isis in he Eu o A ea, Cesi o Fo um,
Volume 14, Special Issue Ap il.
35
De G auwe, P. and Ji, Y. (2023) Mone a y Policies wi hou Gi eaways o Banks, CEPR Discussion
Pape DP18103, Ap il.
De G auwe, Paul and Ji, Yuemei (2024) How o conduc mone a y policies: he ECB in he pas ,
p esen and u u e. Jou nal o In e naonal Money and Finance, 143.
Deu sche Bundesbank (2022) Demand o eu o bankno es issued by he Bundesbank: cu en
de elopmen s, Mon hly Repo Ap il.
Deu sche Bundesbank (2023) Annual Repo 2022, F ank u (pd a ailable online)
Deu sche Bundesbank (2024) Annual Repo 2023, F ank u am Main (only a ailable online)
ECB (2021) Annual Accoun s o he ECB 2020, F ank u am Main.
ECB (2023a) ECB adjus s emune a ion o minimum ese es, P ess elease, 27 July.
ECB (2023b) FAQ on he public sec o pu chase p og amme,
hps://www.ecb.eu opa.eu/mopo/implemen /app/h ml/ecb. aq_pspp.en.h ml
ECB (2024a) FAQ on he pandemic eme gency pu chase p og amme,
h ps://www.ecb.eu opa.eu/mopo/implemen /pepp/h ml/ecb. aq_pepp.en.h ml
ECB (2024b) Changes o he ope aonal amewo k o implemenng mone a y policy, S a emen
by he Go e ning Council 13 Ma ch 2024,
hps://www.ecb.eu opa.eu/p ess/p /da e/2024/h ml/ecb.p 240313~807e240020.en.h ml
Eisenschmid , J., Kedan, D., Schmi z, M., Adalid, R., and Papsdo , P. (2017) The Eu osys em’s asse
pu chase p og amme and TARGET balances, Occasional Pape Se ies 196, Eu opean Cen al
Bank.
Eu opean Pa liamen (2023) Quan a e gh ening in he eu o a ea - Reques ed by he ECON
commiee, Mone a y Dialogue Pape s, Ma ch
Feb e o, E., and J. Uxó (2013) Unde s anding TARGET2 Imbalances om an Endogenous Money
View” Wo king pape DT-DAEF 2013/2, Uni e si y o Caslla-La Mancha.
F icke, D., G eppmai , S. and Paludkiewicz, K. (2024) Excess ese es and mone a y policy gh ening,
Discussion Pape , Deu sche Bundesbank, No 05.
Galli, G. e F. Ne i (2023) h ps://osse a o iocpi.unica .i /ocpi-pubblicazioni-ha-senso-un-alla me-
pe -le-pe di e-delle-banche-cen ali ion o pape s
Hudepohl, T., Pamina, K., Linze , T., Nguyen, B., Sk zypińska, M., & Vaz C uz, L. (2024) How banks
deal wi h declining excess liquidi y, The ECB Blog, 18 June.
McCauley R., Pin e , J. (2024) Un emune a ed ese es in he Eu osys em, pa 1: Heads I win, ails
you lose; pa 2: Tax incidence and deposi elocaon isks, Vox.eu, 15, 16 Jan.
Pan elopoulos, G. (2024a) Na igang some misconcepons in CBDC. Mimeo.
Pan elopoulos, G. (2024b) Can ex e nal sus ainabili y be decoupled om he NIIP? In e naonal
Economics and Economic Policy, (21), 89-116.
Ros agno R., C. Al a illa, G. Ca boni, W. Lemke, R. Mo o, A. Sain Guilhem, J. Yiangou (2021)
Mone a y Policy in Times o C isis: A Tale o Two Decades o he Eu opean Cen al Bank,
Ox o d Uni e si y P ess.
Schnabel, I. (2023) Back o no mal? Balance shee size and in e es a e con ol, Speech gi en a
Columbia Uni e si y, New Yo k, 27 Ma ch 2023

36
Schnabel, I. (2024) The Eu osys em’s ope aonal amewo k, Speech by Isabel Schnabel, Membe o
he Execu e Boa d o he ECB, a he Money Ma ke Con ac G oup meeng F ank u am
Main, 14 Ma ch 2024
Sonnenbe g, N. (2023) ECB S epping on he B ake(S): Mone a y Tigh ening in an Abundan Rese e
Sys em, in Eu opean Pa liamen (2023), pp. 43-77.
Tucke , P. (2022) Quan i a i e easing, mone a y policy implemen a ion, and he public inances,
London: Ins i u e o Fiscal S udies
Imp in
Publishe
Mac oeconomic Policy Ins i u e (IMK) o Hans-Böckle -Founda ion, Geo g-Glock-S . 18,
40474 Düsseldo , Con ac : mm@boeckle .de, h ps://www. mm-mac o.ne
FMM Wo king Pape is an i egula online publica ion se ies a ailable a :
h ps://www.boeckle .de/de/ mm-wo king-pape -22457.h m
The iews exp essed in his pape do no necessa ily e lec hose o he IMK o he Hans-Böckle -Founda ion.
ISSN 2512-8655
This publica ion is licensed unde he C ea i e commons license:
A ibu ion 4.0 In e na ional (CC BY).
P o ided ha he au ho 's name is acknowledged, his license pe mi s he edi ing, ep oduc ion and dis ibu ion o he ma e ial in
any o ma o medium o any pu pose, including comme cial use.
The comple e license ex can be ound he e: h ps://c ea i ecommons.o g/licenses/by/4.0/legalcode
The e ms o he C ea i e Commons License apply o o iginal ma e ial only. The e-use o ma e ial om o he sou ces (ma ked wi h
sou ce) such as g aphs, ables, pho os and ex s may equi e u he pe mission om he copy igh holde .