Re ue In e na ionale de la Reche che Scien i ique
(Re ue-IRS)
ISSN: 2958-8413
Vol. 3, No. 6, No emb e 2025
This is an open access a icle unde he CC BY-NC-ND license.
h p://www. e ue-i s.com
6646
Global Oil Shocks and In la ion in Mo occo
Chocs pé olie s mondiaux e in la ion au Ma oc
Me yem Balamou1, Ayad El baz2
Bouchaib Doukkali Uni e si y, Na ional School o Business and Managemen , LERSEM, El Jadida ,Mo occo.
Abs ac : This a icle explo es he e olu ion o he co ela ion be ween global oil p ices and in la ion in
Mo occo, wi h special a en ion o he uel subsidy e o ms ha ook place be ween 2012 and 2015. By using
olling eg ession analysis and s uc u al b eaks es s, i has been ound ha he e was no any ema kable
b eakpoin du ing he e o ms. This esul sugges s ha ools ela ed o p ice s abiliza ion o mone a y policies
we e e icien in mi iga ing in la iona y p essu es in he sho e m. The e has been a sub le ansi ion in he
beginning o he 2000s o co espond wi h he de egula ion o he pe oleum indus y in gene al and he
mac oeconomic adjus men e o ms. The absence o any ema kable b eakpoin since 2015 migh be ela ed o
inadequa e da a, delay in he ansmission p ocess o economic luc ua ion, o he global educ ion in oil p ices.
These esul s highligh he complex in e ac ion be ween a ia ions in oil p ices and in la iona y p ocesses and
he p ominen e ec o ins i u ional and global ac o s in he in la iona y p ocess in he Mo occan economy.
Résumé : Ce a icle examine l'é olu ion de la co éla ion en e les p ix mondiaux du pé ole e l'in la ion au
Ma oc, en po an une a en ion pa iculiè e aux é o mes des sub en ions su les ca bu an s mises en œu e
en e 2012 e 2015. G âce à l'u ilisa ion de l'analyse pa ég ession glissan e e des es s de up u es s uc u elles,
il a é é cons a é qu'aucun poin de up u e signi ica i n'es su enu du an la pé iode des é o mes. Ce ésul a
suggè e que les ou ils liés à la s abilisa ion des p ix ou aux poli iques moné ai es on é é e icaces pou a énue
les p essions in la ionnis es à cou e me. Une ansi ion sub ile a é é obse ée au débu des années 2000, en
lien a ec la dé églemen a ion du sec eu pé olie e les é o mes d'ajus emen mac oéconomique. L'absence de
up u e ma quée depuis 2015 pou ai ê e a ibuée à des données insu isan es, à un e a d dans le p ocessus
de ansmission des luc ua ions économiques, ou à la baisse mondiale des p ix du pé ole. Ces ésul a s me en
en lumiè e l'in e ac ion complexe en e les a ia ions des p ix du pé ole e les p ocessus in la ionnis es, ainsi
que le ôle dé e minan des ac eu s ins i u ionnels e des dynamiques mondiales dans l'é olu ion de l'in la ion
au Ma oc.
Keywo ds: In e na ional oil p ices; Mo occan in la ion; S uc u al b eak; Rolling eg ession; Ene gy sec o
libe aliza ion; P ice pass- h ough; Mac oeconomic shocks.
Mo s-clés : P ix in e na ionaux du pé ole ; In la ion au Ma oc ; Rup u e s uc u elle ; Rég essions glissan es ;
Libé alisa ion du sec eu éne gé ique ; T ansmission des p ix ; Chocs mac oéconomiques.
Digi al Objec Iden i ie (DOI): h ps://doi.o g/10.5281/zenodo.17680170
Re ue In e na ionale de la Reche che Scien i ique (Re ue-IRS) - ISSN : 2958-8413
h p://www. e ue-i s.com
6647
1 In oduc ion
The link be ween oil p ices in he global ma ke and in la ion in indi idual coun ies is one o he mos essen ial
a eas o unde s and in mac oeconomic s udies, especially in oil-impo ing coun ies. This is because uel is
conside ed one o he mos essen ial p oduc ion ac o s and consump ion i ems in any economy, he eby ensu ing
ha any change in he global ma ke can cause cos -push in la ion in any economy. Ne e heless, one should
ecognize ha he pass- h ough e ec be ween in la ion and oil p ices does no emain consis en in any economy
bu luc ua es because o a ying policies, ins i u ional se ings, and s uc u al shi s in he economy. The case o
Mo occo, which is an oil-impo ing na ion, o e s an in e es ing s udy o his kind.“In Mo occo, whe e he s a e
p o ides uel o he popula ion a hea ily subsidized p ices, i has been explici ly ecognized ha he aim o uel
subsidiza ion was o shield Mo occans om oil p ice ola ili y in he global ma ke .” Fo many yea s, he No h
A ican s a e ope a ed an ex ensi e uel subsidy p og am ha shielded Mo occans om he aga ies o he global
pe oleum ma ke while incu ing la ge budge a y expendi u es in he p ocess. This subsidy policy g adually olled
back be ween 2012 and 2015 h ough he adop ion o policies o de egula e he p ices o uel, po en ially lea ing
in la ion in he coun y ulne able o oil p ice shocks in he global ma ke . The e ec o such policies in making
in la ion in Mo occo ulne able o any change in he global ma ke can be iewed o be close o being de e mina e
in na u e and can hus quali y o be he subjec ma e o he s udy p oposed in his hesis because i has he e ec
o being a he close o being na u al in na u e. “In o he wo ds, i p ices in gene al end o luc ua e in acco dance
wi h ma ke o ces, hey should simila ly luc ua e in acco dance wi h ma ke o ces in Mo occo in spi e o he
subsidy policy because he subsidy policy is expec ed o shield Mo occans om ma ke o ces.” This policy makes
i essen ial o examine he impac o he subsidy policy on he pass- h ough e ec wi hin Mo occo wi h he
pu pose o es ablishing whe he he e has been any s uc u al ans o ma ion in he pos - e o m e a in e ms o he
le el o s uc u al connec i i y be ween he in la iona y p ocess in Mo occo and any luc ua ion ha akes place
in he global ma ke in e ms o pe oleum p ices. “As Hamil on (1983) wisely no ed, “Nea ly e e y pos -wa U.S.
ecession ollowed a majo oil p ice spike,” hus i mly es ablishing oil shocks as de e minan s o in la ion.
Howe e , his pa e n has weakened since he 1970s. Hooke (2002) a gued ha his was because he e was a
policy egime change—namely, he U.S. Fede al Rese e began o ollow an “expec a ion-ancho ed” mone a y
policy ule o o se he in la iona y e ec s o oil shocks. E e since, he e has been a pa adigm shi owa d models
which assume ha policy ins i u ions in he domes ic economy can cushion he e ec o oil shocks. Fo ins ance,
Oyelami e al. (2024) showed ha in he A ican con ex , he pass- h ough e ec can signi ican ly di e based on
indi idual exchange egimes, oil-impo s a us, and subsidy schemes—highligh ing ha “ he in la iona y e ec o
oil is no uni e sal bu a he e y con ex -dependen on local economic and policy undamen als.” Ga ón e al.
(2024) u he cla i ied he na u e o mode n in la iona y ela ionships wi h oil by ad ancing he p inciple o
“in e na ional ulne abili y” o highligh ha “global o ces such as mone a y spillo e s in addi ion o oil ma ke
and supply chain shocks a e a mo e de e mina i e o in la ion in he global economy.” These wo ks highligh
he apidly g owing consensus ha he oil-and in la ion ela ionship emains condi ional a he han linea and hus
emain empe ed by he in e play o domes ic and global o ces.
2 Da a desc ip ion and me hodology
2.1 Da a desc ip ion
This s udy employs da a wi h 34 yea ly pe iods o he na ion o Mo occo, anging om 1990 o 2023. The mos
impo an mac oeconomic a iables, such as in la ion a es, GDP g ow h, pe oleum impo amoun s, and o al
money supply in an economy, a e ex ac ed om he Wo ld Bank eco ds, as speci ied in able 1. C ude oil p ices,
ac ing as mac oeconomic o ces in he global ma ke and being especially signi ican o a c ude oil-impo e
na ion such as Mo occo, a e ex ac ed om he U.S. Ene gy In o ma ion Adminis a ion eco ds. The choice o
hese a iables speci ically has been made wi h he aim o inco po a ing he majo channels h ough which oil
p ices can a ec in la ion.
Re ue In e na ionale de la Reche che Scien i ique (Re ue-IRS) - ISSN : 2958-8413
h p://www. e ue-i s.com
6648
Table 1. De ini ion o Va iables
Va iables
Measu emen
Sou ces
C ude Oil
US dolla s pe ba el
EIA
In la ion
Annual %
Wo ld Bank
B oad Money
% o GDP
Wo ld Bank
GDP G ow h
Annual %
Wo ld Bank
Ene gy Impo a ions
Annual %
Wo ld Bank
Sou ce: Au ho ’s own cons uc ion
2.2 Me hodology
This s udy uses a wo-s ep me hodology o examine whe he he ela ionship be ween oil p ices and in la ion in
Mo occo a ies o e ime: (1) isual inspec ion using olling eg essions, and (2) o mal es ing o s uc u al
b eaks. Examining he changing ela ionship be ween global c ude oil p ices and domes ic in la ion, especially
conside ing he 2012–2015 ene gy subsidy e o ms, is done using he olling eg ession me hodology. In con as
o s a ic models, olling eg ession epea edly es ima es he pass- h ough coe icien o oil p ices, allowing i o
luc ua e o e ime. model ac oss olling windows, which a e mo ing subsamples o ixed leng hs. This me hod
aids in cap u ing he slow s uc u al changes ha could esul om signi ican global shocks o policy changes.
Windows o 8, 10, 15, and 15 ha oll, and 20 yea s a e used o e alua e he esul s' esilience o e a ious ime
pe iods, o e ing bo h sho - and medium- e m insigh s in o he dynamics o oil and in la ion. In la ion is he
dependen a iable in each speci ica ion, and i is eg essed on global c ude oil p ices. O he con ol a iables
include GDP g ow h o accoun o demand-side p essu es, b oad money o e lec mone a y condi ions, and
ene gy impo s o accoun o ex e nal ene gy dependence. The ollowing is he speci ica ion o he olling
eg ession equa ion:
whe e ep esen s he ime- a ying coe icien cap u ing he pass- h ough e ec o oil p ices on in la ion. This
speci ica ion allows us o map ou how he connec ion be ween oil and in la ion has changed o e ime, speci ically
iden i ying i Mo occo became mo e, less, o equally sensi i e o global oil p ice shocks a e uel p ices we e
libe alized. Se e al s uc u al b eak es s, including he Chow, CUSUM, Sup emum-Wald, and Bai-Pe on es s
a e used in he second sec ion o he analysis o look o pa ame e ins abili y. The Chow es de e mines i he
model coe icien s be o e and a e a p ede e mined b eakpoin di e s a is ically signi ican ly he o he h ee es s
check o an unknown b eakpoin and don' make any assump ions abou he b eak da e. Plo ing he cumula i e
sum o ecu si e esiduals o e ime allows he CUSUM (Cumula i e Sum) es o assess he s abili y o he
eg ession pa ame e s. The model is ega ded as s able i he cumula i e sum emains wi hin c i ical bounds;
de ia ions signi y a s uc u al b eak. By es ima ing he model ac oss all po en ial b eakpoin s wi hin a immed
ange and compu ing a Wald s a is ic a each, he Sup emum-Wald es inds a single s uc u al b eak a an
unknown poin in he sample. Pa ame e ins abili y is hen es ed using he la ges o hese s a is ics A no ewo hy
ou come poin s o a s uc u al ac u e. Finally, by di iding he ime se ies in o segmen s and de e mining whe he
he e a e signi ican di e ences in he eg ession pa ame e s be ween hose segmen s, he Bai-Pe on es inds
mul iple s uc u al b eaks. By educing he o al esidual a iance and compa ing he model i bo h be o e and
a e allowing o b eaks, i inds b eakpoin s. When he imp o emen in i (as de e mined by a es s a is ic like
he F-s a is ic) su passes c i ical alues, a s uc u al b eak is deemed s a is ically signi ican . When combined,
hese echniques o e a ho ough e alua ion o s uc u al s abili y by using s a is ical in e ence unde known and
unknown b eakpoin assump ions wi h isual inspec ion om he olling eg essions.
3 Applica ion and Resul s
3.1 Desc ip i e s a is ics
Based on 34 yea ly obse a ions, Table 2 displays he desc ip i e s a is ics o GDP g ow h, in la ion, b oad money
supply, and c ude oil p ices. Acco ding o he analysis, a e age alues o c ude oil we e oughly 52.10, in la ion
was 2.64, b oad money was 86.80, and GDP was 3.59. GDP g ow h luc ua ed he mos among hese a iables,
Re ue In e na ionale de la Reche che Scien i ique (Re ue-IRS) - ISSN : 2958-8413
h p://www. e ue-i s.com
6649
anging om -7.178 o 12.373. Wi h a s anda d de ia ion o 32.33, c ude oil p ices showed he mos ola ili y,
while GDP in la ion s ayed ela i ely s able. Despi e in la ion ge ing close o he de ia ion h eshold (p = 0.056),
he Ja que-Be a no mali y es s la gely con i med he assump ion o a no mal dis ibu ion. Be o e mo ing on o
he dynamic analysis o he oil p ice pass- h ough o in la ion, hese p elimina y s a is ics ga e a undamen al
unde s anding o he da a s uc u e.
Table 2. Desc ip i e s a is ics
S a is ic
C ude Oil
In la ion
B oad Money
GDP G ow h
Mean
52.10265
2.641458
86.79850
3.590188
Median
47.98000
1.719614
94.54207
3.470312
Maximum
111.6300
7.986166
128.8631
12.37288
Minimum
12.76000
0.303386
46.15091
-7.178207
S d. De .
32.33022
2.205386
26.64250
3.976320
Skewness
0.472982
0.991741
-0.127836
-0.504394
Ku osis
1.907361
2.638106
1.523468
4.001091
Ja que-Be a
2.959002
5.758993
3.181145
2.861437
P obabili y
0.227751
0.056163
0.203809
0.239137
Sum
1771.490
89.80957
2951.149
122.0664
Sum Sq. De .
34493.02
160.5029
23424.15
521.7671
Obse a ions
34
34
34
34
Sou ce: Au ho ’s own cons uc ion
The ime se ies beha io o ou key a iables o Mo occo—c ude oil p ices, in la ion, he o al money supply,
and GDP g ow h—is shown g aphically in Figu e 1, which oughly co e s he yea s 1990 o 2023. We can see
se e al di e en pa e ns. C ude oil p ices exhibi signi ican ola ili y, wi h no able spikes peaking be ween 2008
and 2012 and again a e 2020, in e spe sed wi h p ecipi ous declines like he oil p ice mel down o 2014–2016.
When measu ed as a pe cen age o GDP, he b oad money supply shows a dis inc and la gely consis en upwa d
end o e he cou se o he pe iod, sugges ing ei he an expanding inancial sec o o an expansiona y mone a y
policy in ela ion o economic ou pu . Con a ily, in la ion seems o ha e been mos ly con ained and s able du ing
his ime, ypically anging in he low ange (e.g., 0-5%), hough some no able inc eases co espond wi h he la e
spikes in he p ice o c ude oil ha we e seen a e 2020. Finally, GDP g ow h shows no mal cyclical oscilla ions
a ound a posi i e a e age, wi h no able declines no iceable, including one a ound 2020 ha is p obably ela ed o
he COVID-19 pandemic.
Figu e 1. Visual Analysis o Va iable T ends
Re ue In e na ionale de la Reche che Scien i ique (Re ue-IRS) - ISSN : 2958-8413
h p://www. e ue-i s.com
6650
3.2 S a iona i y es Resul s
We i s used he Augmen ed Dickey-Fulle (ADF) uni oo es o e i y he s a iona i y o ou ime se ies da a
o make su e ou eg ession analysis was sound and o p e en decep i e "spu ious" esul s. Using a 5% c i ical
alue o -1.950, we applied his es o GDP g ow h, in la ion, c ude oil p ices, ene gy impo s, and he b oad
money supply. Rega ding he o de o in eg a ion, he esul s pain ed a mixed pic u e. I was disco e ed ha
b oad money, ene gy impo s, in la ion, and c ude oil p ices we e all non-s a iona y a hei s a ing poin s. Bu
a e being di e enced once, hey all became s a iona y, indica ing ha hey a e in eg a ed o o de one, o I(1).
The GDP g ow h a iable, on he o he hand, was I(0) since i was al eady s a iona y a i s ini ial le el. The e o e,
we used he i s di e ences o all he I(1) a iables and included he al eady s a iona y GDP g ow h a iable in
i s o iginal, undi e enced o m o make su e ha ou subsequen olling eg ession analysis was s a is ically sound
and una ec ed by non-s a iona y ends.
Table 3. Uni Roo Tes
Va iable
Undi e enced
-S a is ic
C i ical Value
1s Di e ence
-S a is ic
C i ical Value
O de o
In eg a ion
C ude Oil
-0.315
-1.950
-5.252
-1.950
I(1)
In la ion
-1.623
-1.950
-3.406
-1.950
I(1)
B oad Money
2.868
-1.950
-2.517
-1.950
I(1)
GDP G ow h
-1.952
-1.950
—
—
I(0)
Ene gy
Impo s
0.712
-1.950
-4.126
-1.950
I(1)
Sou ce: Au ho ’s own cons uc ion
3.3 Rolling eg ession esul s
We used window sizes o 8, 10, 15, and 20 yea s o pe o m olling eg essions on he s a iona y a iables,
employing i s di e ences when equi ed. Reg ession coe icien s along wi h s anda d e o s we e combined wi h
he o iginal da ase o u he analysis using S a a's " olling" command. The o m o he eg ession equa ion was
as ollows:
By adjus ing o GDP g ow h, ene gy impo s, and b oad money, his speci ica ion enabled us o in es iga e he
ela ionship be ween changes in oil p ices and changes in in la ion. Time- a ying oil p ice coe icien s om he
olling eg essions we e hen plo ed o look o possible s uc u al changes and pa e ns h oughou he sample
pe iod.
Re ue In e na ionale de la Reche che Scien i ique (Re ue-IRS) - ISSN : 2958-8413
h p://www. e ue-i s.com
6651
Figu e 2. Es ima ed Rolling Be as o Oil on In la ion (8- and 10-Yea Windows)
The es ima ed coe icien o he change in oil p ice (β_ ) om olling eg essions using 8- and 10-yea windows
is shown in Figu e 2, wi h each es ima e co esponding o he inal obse a ion in he olling sample. The end poin
o each olling window is displayed on he x-axis. Despi e expec a ions o a s uc u al b eak linked o Mo occo’s
ene gy subsidy e o ms om 2012 o 2015, he 8-yea and 10-yea olling eg essions e ealed no disce nible
discon inui y in he oil p ice coe icien du ing ha pe iod. Fo ins ance, du ing he 10-yea window, he coe icien
dec eased h ough 2015 a e only ma ginally inc easing be ween 2011 and 2012 (+0.01). This sugges s ha he
in la iona y e ec s o he e o m migh ha e been oo g adual o be de ec ed by sho -window eg essions, delayed,
o o se by o he economic s abilize s. O he pe iods, on he o he hand, exhibi ed mo e p onounced coe icien
shi s, sugges ing possible s uc u al b eaks. F om 2010 o 2019, he g aphs o bo h windows displayed ela i ely
s able coe icien s; howe e , g ea e ola ili y was obse ed a he beginning and end o he sample. A sign shi
in he oil p ice coe icien ( om nega i e o posi i e) o a yea - o-yea change exceeding 0.04—abo e he ypical
sample a ia ion o 0.01–0.03—was conside ed a po en ial s uc u al b eak. Acco ding o his c i e ion, he
iden i ied b eakpoin s we e 2000–2001 (sign swi ch), 2006 (sha p ise in he 8-yea window; sign swi ch in he
10-yea ), 2009 (mode a e d ops in bo h), and 2022–2023. While mos o he da es did no co espond o clea
domes ic policy changes, he shi a ound 2000 coincided wi h Mo occo’s g adual p i a iza ion and libe aliza ion
o he ene gy sec o in he la e 1990s (Usman & Ameg oud, 2019). Howe e , hey also o e lapped wi h majo
global e en s such as he 2008–2009 inancial c isis and he COVID-19 pandemic, which likely in luenced he
ansmission o oil p ices o Mo occan in la ion h ough ade, expec a ions, and mone a y policy. Du ing hese
pe iods, oil’s in la iona y impac inc eased due o weakened policy bu e s, bu as ex e nal p essu es eased, he
coe icien s ended o decline again. Appa en b eaks a he beginning o end o he sample mos likely e lec ed
edge e ec s a ising om limi ed da a. Nega i e coe icien s in some yea s, sugges ing an in e se oil–in la ion
ela ionship, we e a ibu ed o noise in sho windows, while la ge windows yielded consis en ly posi i e
coe icien s.
Re ue In e na ionale de la Reche che Scien i ique (Re ue-IRS) - ISSN : 2958-8413
h p://www. e ue-i s.com
6652
Figu e 3. Es ima ed Rolling Be as o Oil on In la ion (15- and 20-Yea Windows)
Figu e 3 p esen ed he olling oil p ice coe icien s using 15- and 20-yea windows. The coe icien s we e mo e
s able and consis en ly posi i e compa ed wi h sho e windows. S uc u al b eaks we e de ined as sign swi ches
o coe icien changes ≥ 0.02, since mos yea ly shi s we e smalle . The mos p ominen pa e n was a isible dip
in he coe icien om 2015 o 2021 in bo h g aphs— he coe icien ell by 0.02 in 2015 and emained low un il
ising again in 2022. Al hough he coe icien was expec ed o ise a e he 2012–2015 subsidy e o ms, he decline
likely e lec ed o se ing o ces. These included he e o ms’ o e lap wi h he 2014–2015 global oil p ice collapse
(Ben ou , 2015, p.23), Mo occo’s low-in la ion en i onmen dampening cos pass- h ough (Bo io e al., 2023,
p.11), and ma ke ic ions such as oligopolis ic p icing in Mo occo’s uel sec o (Be elsmann S i ung, 2024,
p.14), which limi ed he ansmission o global oil p ice changes o consume p ices. Toge he , hese ac o s may
ha e delayed he expec ed ise in oil-in la ion sensi i i y ollowing he policy change. The ise in 2022 (+0.02 in
bo h windows) ep esen ed ano he po en ial b eakpoin , consis en wi h he sho -window esul s, and may ha e
e lec ed he delayed ull e ec o he subsidy e o ms as p io o se ing o ces waned. The 15-yea window also
e ealed po en ial b eaks in ea lie yea s: 2005 (+0.04), 2006 (sign swi ch), and 2008 (+0.02). Collec i ely, Figu es
2 and 3 sugges ed ha s uc u al changes in he oil p ice–in la ion ela ionship occu ed in he ea ly 2000s (possibly
ied o Mo occo’s ene gy ma ke libe aliza ion), in 2008–2009 (global inancial c isis), in he pos - e o m pe iod
a e 2015 ( e lec ing dampened cos pass- h ough), and in 2022 (likely ela ed o pos -pandemic shocks and
delayed e o m e ec s). These esul s s emmed om isual inspec ion and we e ea ed as diagnos ic a he han
de ini i e, highligh ing he need o o mal b eak es s o con i m he iming and p esence o s uc u al changes.
The limi ed a ailabili y o highe - equency Mo occan mac o da a mean ha each eg ession included ew
obse a ions, which educed p ecision in es ima ing he coe icien s, widened con idence in e als in hypo hesis
es ing, and could ha e caused noise o appea as s uc u al shi s.
Re ue In e na ionale de la Reche che Scien i ique (Re ue-IRS) - ISSN : 2958-8413
h p://www. e ue-i s.com
6653
Figu e 4. Es ima ed Rolling Be as o Oil on In la ion wi h 95% Con idence Bands (15- and 20-Yea Windows)
To assess he s a is ical signi icance o he olling oil p ice coe icien o e ime, we cons uc ed poin wise 95%
con idence in e als using he -dis ibu ion. Each olling eg ession’s bounds equal he coe icien ± × s anda d
e o , plo ed a ound he coe icien pa h. These in e als show whe he he coe icien a a gi en poin is
signi ican ly di e en om ze o a he 5% le el. Howe e , wi h small sample sizes (all <30), hese in e als may
unde s a e ac ual unce ain y. Figu e 4 shows s a is ical signi icance only o he 15-yea window ending in 2022
and he 20-yea windows ending in 2022 and 2023. The 8- and 10-yea windows show no signi icance (see
Appendix, Figu e A.1). Whe e possible s uc u al b eaks a e sugges ed isually, we ollow up wi h o mal b eak
es s o con i m hei s a is ical alidi y.
4 Discussion s uc u al b eak es esul s
To o mally assess s uc u al changes in he oil p ice coe icien , we conduc ed se e al b eak es s: a Chow es ,
Sup emum Wald es , CUSUM es , and Bai-Pe on es (s. Appendix, Table A.1). Fi s , Chow es s we e un o
selec ed b eak yea s (1999, 2000, 2001, 2008, 2009, 2015, 2022) manually in S a a, using he Chow es o mula
o compa ing he esiduals o sepa a e eg essions o a pooled eg ession:
None o he es ed yea s p oduced s a is ically signi ican esul s a he 5% le el, and he Chow es could no be
un o 2022 due o an insu icien numbe o pos -b eak obse a ions. Nex , a Sup emum Wald es (“es a
sbsingle” in S a a) was pe o med o de ec a single unknown b eakpoin . The es ims 25% o he sample om
bo h ends, lea ing a immed window om 2000 o 2015. The es ima ed b eak yea is 2000, wi h a Sup-Wald
s a is ic indica ing signi icance a he 1% le el and a s uc u al change in he oil-in la ion ela ionship a he
beginning o he 2000s. In con as , he CUSUM es (“es a sbcusum”) ound no s uc u al b eak, as ecu si e
esiduals s ayed wi hin he 95% con idence bands (s. Appendix Figu e A.2). Simila ly, he Bai-Pe on es ,
implemen ed ia he “x b eak” command (a e ins alling “x b eak” and “mo ema a”), also ailed o de ec any
s uc u al b eaks. This es examines whe he he e a e one o mo e b eaks a unknown da es in he eg ession
coe icien s, using he UDmax es and sequen ial F- es s o e alua e i a s uc u al b eak exis s and whe he
in oducing addi ional b eakpoin s imp o es he model i . None o he es s a is ics exceed he co esponding
Re ue In e na ionale de la Reche che Scien i ique (Re ue-IRS) - ISSN : 2958-8413
h p://www. e ue-i s.com
6654
c i ical alues a con en ional signi icance le els. As wi h he Wald es , he da a was immed a 25% om bo h
ends.
Taken oge he , hese s uc u al b eak es s yield mixed esul s. The Sup emum Wald es iden i ies a b eak a ound
2000, consis en wi h he beha io obse ed in olling eg essions. Howe e , he Chow, CUSUM, and Bai-Pe on
es s do no ind signi ican ins abili y. This disc epancy likely s ems om da a limi a ions and hei e ec ac oss
es s. Speci ically, he Chow es ’s powe educes wi h ewe obse a ions and does no de ec b eaks i
obse a ions be o e/a e he spli yea a e limi ed. Simila ly, he CUSUM es has low powe in small samples
and is weak in de ec ing b eaks nea he bounda ies o he sample (Bonda enko, 2008; El-Shagi & Giesen, 2010),
while he Bai–Pe on es is cons ained by imming ules and equi es longe ime se ies o iden i y mul iple
b eakpoin s (An oshin e al., 2008; Ca ion-i-Sil es e & Kim, 2017). He e i is also impo an o no e ha gi en
a small sample, he Wald es can o e ejec he null hypo hesis o no s uc u al b eak (Guo e al., 2005; Hansen,
2013). O e all, only he Sup emum Wald es shows e idence o a s uc u al b eak in he yea 2000, while indings
o o he yea s sugges no b eaks in he oil-in la ion ela ionship. Howe e , gi en he da a limi a ions, all esul s
emain inconclusi e a his s age due o he small sample size a he han he ue absence o s uc u al change.
5 Conclusion
The isual analysis using olling eg essions (8-20-yea windows) e eals ha oil‑p ice pass‑ h ough o Mo occan
in la ion a ied isibly o e ime. Sho e windows show mo e ola ili y while longe windows yield smoo he ,
consis en ly posi i e e ec s wi h a dip pos ‑2015 (a he end o he subsidy e o ms). The Sup. -Wald es iden i ies
a s a is ically signi ican egime shi a ound he yea 2000, wi h he oil‑in la ion coe icien inc easing a e being
nega i e in 1999, sugges ing a s onge e ec o oil p ice shocks on in la ion. While he e we en' any di ec
e o ms speci ically aimed a oil p ice pass- h ough du ing ha pe iod, i 's impo an o emembe ha Mo occo
was in he mids o a b oade ene gy sec o libe aliza ion. This in ol ed hings like p i a izing uel dis ibu ion
ne wo ks and implemen ing egula o y changes ha enabled independen powe p oduce s o gene a e elec ici y
using impo ed ossil uels (Usman & Ameg oud, 2019). These changes e ec i ely loosened di ec go e nmen
con ol o e how p oduc ion cos s and p icing we e se , and, in doing so, ied domes ic ene gy cos s mo e closely
o in e na ional oil p ices. This likely in ensi ied how global oil shocks ansla ed in o local in la ion. Tha being
said, i 's c ucial o app oach he Wald es esul s wi h a deg ee o ci cumspec ion, especially gi en ha small
samples inhe en ly inc ease he likelihood o a Type I e o . Addi ionally, despi e expec a ions, he 2012–2015
subsidy e o m pe iod did no p oduce a s a is ically con i med s uc u al b eak in oil pass‑ h ough. This esul
sugges s ha e o ms ei he had delayed e ec s, we e o se by concu en global shocks (e.g. oil p ice collapse in
2014–2015), o we e mu ed by ma ke ic ions and o he s abilizing mechanisms. Addi ionally, in mos s uc u al
b eak es s, he da a om 2015 onwa ds was ei he immed o a he e y end o he analyzed sample, educing
s a is ical powe and limi ing he abili y o de ec s uc u al b eaks a hose edge poin s. Finally, we acknowledge
he main limi a ions o his pape - a ela i ely small annual sample (34 obse a ions) due o una ailabili y o
highe equency da a and limi ed s a is ical powe , pa icula ly o he pos ‑ e o m pe iods o in e es . These
cons ain s may obscu e s uc u al changes a he ends o he obse ed pe iod and inc ease isk o Type II (Chow,
CUSUM, Bai-Pe on) o Type I e o s (Wald), dis o ing esul s.
6 Appendix
