BEAC

Banque des Etats de

l’Afrique Centrale

BEAC Working Paper

- BWP N° 01/22 -

CEMAC-GLOBAL : Introducing a macroeconomic model for a small

open monetary union of oil-exporting developing countries

ESSIANE Patrick-Nelson Daniel

Economiste
Direction des Etudes, de la Recherche et des
Statistiques

[email protected]

BANQUE DES ETATS DE

L’AFRIQUE CENTRALE
Les opinions émises dans ce document de
736, Avenue Monseigneur
Vogt BP:1917 Yaoundé
travail sont propres à leur (s) auteur (s) et ne
Cameroun représentent pas nécessairement la position de la
Banque des Etats de l’Afrique Centrale.
Tel :(237) 22234030 /
22234060 The opinions expressed in this working
Fax : (237) 22233329 paper are those of the author (s) and don’t
necessarily represent the views of the Central Bank
www.beac.int of Central Africa States.

D I R E C T I O N D E S E T U D E S, D E L A R E C H E R C H E E T D E S S T A T I S T I Q U E S
 CEMAC-GLOBAL : Introducing a macroeconomic model for a small open
monetary union of oil-exporting developing countries z τ

Essiane Patrick-Nelson Daniel ∗

May 2022

Abstract

Evaluating and analyzing the country-specific impacts of the monetary policy and the spillo-
ver effects of domestic shocks is a key policy concern in monetary unions. In order to address the
relative lack of consideration of these issues in the literature on the Central African Economic
and Monetary Community (CEMAC), this paper outlines the key features of a semi-structural
multi-country macroeconomic model of the CEMAC economies called CEMAC-GLOBAL. Here,
every country of the monetary union is modelled as a small open economy, with a single Cen-
tral Bank in charge of the monetary policy. Based on a New Keynesian framework, this model
incorporates (i) a distinction between oil and non-oil sector, (ii) real-financial linkages with cre-
dit cycles, (iii) CEMAC-specific monetary policy features related to the non-conventional fixed
exchange rate regime of this monetary union, (iv) an explicit modelling of liquidity injections
of the Central Bank and (v) national fiscal policies making an arbitrage between closing the
output gap and stabilizing the public debt. The model is estimated with Bayesian methods in
order to improve the empirical performances. Model simulations provide illustrations of (i) how
the Central Bank could respond to various shocks, (ii) the asymmetric effects of the monetary
impulses on CEMAC economies or (iii) the spillover effects of fiscal policy shocks. Historical
decompositions replicate key stylized facts about the recent oil-crisis and COVID-19 pandemic.
Keywords : Semi-structural modeling ; Multi-country models ; spillover effects ; Monetary
Union ; Africa ; Monetary Policy ; Fiscal Policy ; Heterogeneity ; Macroeconomic modeling
JEL Classification : C54 ; E37 ; E52 ; E62 ; O55

z. I am deeply thankful to Junior Maih for the availability, the patience and the priceless comments on
this model. I am also thankful to Carlos de Resende and the African Training Institute staff of the IMF’s
Institute for Capacity Development for having largely inspired the launch of this research project. I am
also full of gratitude to Désiré Avom, Leo Spencer Keungne and Kenneck Massil for their comments and
discussions during the Bank of Central African States’s (BEAC) research seminars. I would like to extend
my sincere thanks to Gabin Afomongono, Jacques Landry Bikai, George Diffo, Blaise Ehowe, Armand
Fossouo, Dorel Maleo Batoumeni, Franck Mba Eyene, Evrard Mounkala, Emile Thierry Mvondo, Francis
Ngomba Bodi, Julie Onomo, Mustafa Taher and my colleagues form the Economic studies, Research and
Statistics Department of the BEAC for their precious advice, in-depth discussions and their availability
during this modelling process. I also wish to thank the participants of 23rd Central Bank Macroeconomic
Modelling Workshop and the 2020 BEAC research seminar for the rich discussions and suggestions. I’d
also like to extend my gratitude to the BEAC’s Director in charge of Research, Mr. Hamadou Abdoulaye,
and the governing board of the BEAC for having allowed the funding of this research.
τ . This is a working paper, and hence it represents research in progress. It is not meant to represent
the position or opinions of the BEAC or its Members, nor the official position of any staff members Any
errors or mistakes are the faults of the author.
∗. Economist, Bank of Central African States (BEAC), Department of Economic Studies, Research
and Statistics. [email protected]

1
 Résumé

L’évaluation et l’analyse du degré d’asymétrie des effets de la politique monétaire ainsi

que des effets de débordement des chocs spécifiques à des pays membres est une préoccupa-
tion essentielle en union monétaire. Face à la relative faible considération de ces enjeux dans
la littérature portant sur la CEMAC, ces travaux présentent les principales caractéristiques
d’un modèle macroéconomique semi-structurel multipays des économies de la CEMAC appelé
CEMAC-GLOBAL. Ce modèle considère chaque pays de la CEMAC comme une petite éco-
nomie ouverte disposant d’une Banque Centrale en charge de la politique monétaire commune.
CEMAC-GLOBAL repose sur un cadre théorique nouveau keynésien incorporant (i) une distinc-
tion entre secteur pétrolier et non pétrolier ainsi que (ii) des liens entre secteur réel et financier
intégrant des cycles du crédit. En outre, le modèle prend en considération (iii) les spécificités
liées au régime de change fixe non conventionnel de la CEMAC, ainsi (iv) qu’une modélisation
explicite des injections de liquidités dans le comportement de la Banque Centrale en plus de la
règle de détermination du taux d’intérêt. Également, le modèle incorpore (v) des politiques bud-
gétaires nationales qui font face à un arbitrage entre annulation de l’écart de production national
et maîtrise de leur niveau d’endettement. Le modèle est estimé avec une approche bayésienne
dans l’optique d’en améliorer les performances empiriques. Les simulations opérées permettent
entre autres d’illustrer (i) comment la Banque Centrale pourrait répondre à différents chocs,
(ii) le degré d’asymétrie des effets de la politique monétaire entre les pays membres ainsi que
(iii) les effets de débordement des chocs de politique budgétaire. En outre, les décompositions
historiques répliquent les principaux faits stylisés relatifs à la crise pétrolière de 2014 ainsi qu’à
la pandémie de COVID-19.
Keywords : Modèles semi-structurels ; Modèle multipays ; Effets de débordement ; Union
Monétaire ; Afrique ; Politique monétaire ; Politique budgétaire ; Hétérogénéités ; Modélisation
macroéconomique
Classification JEL : C54 ; E37 ; E52 ; E62 ; O55

2
 Résumé non-technique

Une Banque Centrale conduit sa politique monétaire dans l’optique d’atteindre des
objectifs macroéconomiques déterminés dans son mandat. A cette fin, elle effectue une
analyse rétrospective et prospective de l’économie de sa zone d’émission ainsi que de
l’environnement international. Sur cette base, elle ajuste ses instruments (taux directeurs,
coefficients de réserves obligatoires, injections de liquidités, etc.) de manière à obtenir la
réaction la plus en adéquation possible de l’économie de sa zone d’émission vis-à-vis de
ses objectifs de politique monétaire.

Exercice habituellement réalisé sur un seul pays, il gagne nettement en complexité

lorsqu’il est conduit dans le cadre d’une union monétaire. L’un des principaux problèmes
dans la conduite de la politique monétaire en union monétaire réside dans le degré plus ou
moins élevé d’hétérogénéité des pays sur laquelle elle s’exerce. En effet, la situation devient
alors semblable à celle d’un pilote de drone devant guider plusieurs appareils en même
temps avec une seule télécommande : sa réussite dépendrait non seulement de la similarité
des drones à sa charge, mais également de la nature et du degré de synchronisation
des évènements qui affecteraient chacun d’eux. C’est la raison pour laquelle les courants
orthodoxes de la théorie économique suggèrent que seuls des pays similaires fassent partie
d’une même union monétaire (cf. Théorie des zones monétaires optimales).

Toutefois, en réalité, il est rare que tous les pays membres d’une union monétaire (voire
même les régions d’un même pays) soient suffisamment proches économiquement, géogra-
phiquement ou institutionnellement pour correspondre à 100% à ce que recommande la
théorie économique standard. En pratique, les États membres d’une union monétaire
cherchent à faire converger leurs structures et politiques économiques de sorte à atténuer
leurs dissemblances et faciliter la conduite des politiques économiques et, en particulier,
celle de la politique monétaire.

L’hétérogénéité des économies membres d’une union monétaire n’étant jamais tota-
lement éliminée, il est donc important d’évaluer non-seulement les potentiels effets asy-
métriques de la politique monétaire sur les économies de ladite union, mais également
d’apprécier les effets de débordement des chocs économiques spécifiques à chaque écono-
mie de l’union sur les autres membres. De la sorte, la Banque Centrale peut non seulement
mieux apprécier l’incidence de son action tant à l’échelle de l’union que sur celle de chaque

3
pays membre, mais également concevoir de meilleures stratégies de réponse lors de la sur-
venance de chocs affectant un sous-ensemble de sa zone d’émission.

Les chercheurs en économie se sont depuis longtemps intéressés à ces problèmes. La

littérature économique a ainsi enregistré de nombreuses contributions dans ce sens, en
particulier depuis l’entame du processus de création de la zone Euro au cours des années
1990. Un accent important a tout spécialement été mis sur la création d’outils suscep-
tibles de répliquer à petite échelle les éléments importants de ces économies à des fins de
simulations de politiques économiques et de prévisions. Certains de ces outils (ou modèles
macroéconomiques) ont tout spécialement été conçus pour évaluer et guider la politique
monétaire dans un environnement relativement complexe.

Toutefois, l’effort de recherche en matière de modélisation macroéconomique en union

monétaire s’est essentiellement concentré sur les économies avancées (zone Euro en tête),
s’intéressant relativement peu aux unions monétaires des pays en développement. La Com-
munauté Économique et Monétaire des États de l’Afrique Centrale (CEMAC), pourtant
l’une des plus anciennes unions monétaires de l’Histoire, a ainsi fait l’objet de peu de tra-
vaux cherchant à la modéliser à des fins d’analyse des effets asymétriques de sa politique
monétaire ainsi que des effets de débordement des chocs spécifiques à des pays membres
de l’union monétaire. Le présent travail ambitionne d’adresser ce problème.

En particulier, cet article présente les principales caractéristiques du modèle CEMAC-

GLOBAL, un modèle macroéconomique semi-structurel multipays conçu pour les éco-
nomies de la CEMAC (Cameroun, République Centrafricaine, Congo, Gabon, Guinée
Équatoriale et Tchad). Il s’agit d’un modèle d’équilibre général dynamique stochastique
(DSGE) dont la base repose sur les fondements théorique de l’école des nouveaux key-
nésiens, mais dont plusieurs extensions ont été incorporées pour permettre de prendre
en compte des caractéristiques spécifiques aux économies de la CEMAC. Ainsi, il inclut
(i) une distinction entre secteur pétrolier et non pétrolier, (ii) une inter-relation entre
sphère économique réelle et financière, (iii) des spécificités liées au régime de change fixe
non conventionnel de la CEMAC, (iv) une prise en compte explicite du comportement
d’injections de liquidités de la Banque Centrale ainsi (v) que des politiques budgétaires
nationales faisant face à un arbitrage entre atteinte du potentiel de leurs économies na-
tionales respectives et maîtrise de leur niveau d’endettement.

4
 Afin de faire correspondre le comportement empirique du modèle CEMAC-GLOBAL
aux principaux faits stylisés des économies de la CEMAC, il est nécessaire de donner des
valeurs aux paramètres des différentes équations qui le composent. Dans cette optique, une
approche séquentielle a été adoptée. Tout d’abord, la technique de la limited information
impulse response matching a été utilisée pour un premier calibrage des paramètres relatifs
à chacun des pays de la CEMAC. Cette technique vise à déterminer une combinaison de
paramètres qui permet de répliquer qualitativement et, dans la mesure du possible, quanti-
tativement, le comportement des économies en terme de réponses à différents chocs. Pour
y parvenir, le modélisateur doit en général se reposer sur des travaux antérieurs ayant
partiellement adressé certains aspects du modèle, les résultats de modèles similaires por-
tant sur d’autres pays, ou encore sa connaissance des économies qu’il modélise. Une fois
les différentes combinaisons des paramètres pertinentes obtenues, celles-ci sont considé-
rées comme priors d’une estimation bayésienne. Schématiquement, l’estimation bayésienne
consiste à combiner l’information a priori sur les paramètres à estimer et l’information
contenue dans les données empiriques des économies étudiées. La combinaison de ces deux
sources d’informations permet d’obtenir les valeurs dites "a posteriori" des paramètres du
modèle. Celles-ci sont une sorte de "mise à jour" de la connaissance des économies étu-
diées sur la base des données utilisées pour l’estimation. Dans le cadre de ces travaux, une
estimation bayésienne est conduite sur des versions mono-pays de CEMAC-GLOBAL,
afin d’alléger la charge de calcul. L’estimation repose sur un algorithme d’optimisation
par biomimétisme répliquant le comportement de recherche de nourriture d’un essaim
d’abeilles mellifères. Cette technique est particulièrement intéressante pour l’optimisation
de fonctions complexes à de nombreuses dimensions comme celle relative à la maximisa-
tion de la distribution a posteriori des paramètres du modèle CEMAC-GLOBAL. Une
fois ces paramètres obtenus, les versions estimées mono-pays sont consolidées pour obtenir
le modèle CEMAC-GLOBAL estimé.

Afin d’apprécier la pertinence du modèle estimé, des analyses de fonctions de réponses

impulsionnelles suite à des chocs de (i) politique monétaire, (ii) de demande domestique,
(iii) sur les prix du pétrole, (iv) des biens alimentaires mondiaux ainsi que (v) de politique
budgétaire ont été réalisées. Les principaux résultats montrent que la politique monétaire a
des effets asymétriques dans la CEMAC, avec un impact relativement plus important dans
les pays dotés d’un secteur pétrolier moins proéminent. En outre, les effets de débordement

5
des chocs spécifiques aux pays semble relativement faible, et sont dépendants de la taille
des pays d’où ils surviennent. Également, les premières simulations montrent que des
hausses de dépenses publiques coordonnées ont plus d’impact sur l’activité globale de la
CEMAC que des choc budgétaires isolés, bien qu’ayant un effet négatif plus important
sur les réserves de change et augmentant les primes de risque pays. Compte-tenu de la
taille importante du modèle, il est évident qu’un nombre plus important de chocs et de
simulations peuvent être simulés, ce qui en fait un outil intéressant pour les analyses de
politique économique.

Sur le plan empirique, des décompositions historiques ainsi que des prévisions intra-
échantillon ont été réalisées sur la base de données trimestrielles officielles publiées par
la Banque des Etats de l’Afrique Centrale, les Institutions en Charge des Statistiques
des pays de la CEMAC, le Fonds Monétaire International, la Banque Centrale Euro-
péenne ainsi que la Réserve Fédérale Américaine allant de 2006 à 2020. Globalement,
les principaux faits stylisés relatifs à la crise pétrolière de 2014 ou encore l’incidence de
la pandémie de COVID-19 ont été répliqués. En outre, les performances en matière de
prévision à moyen terme sont intéressantes, suggérant une certaine utilité en matière de
prévisions macroéconomiques pour les pays de la CEMAC.

Comme tous les modèles macroéconomiques, CEMAC-GLOBAL est un outil appelé à

constamment évoluer tant dans sa structure que sa paramétrisation. Il s’inscrit donc dans
un processus de modélisation toujours en cours. Plusieurs pistes d’amélioration subsistent,
notamment relativement à la modélisation de l’économie étrangère, à une désagrégation
de son bloc d’analyse de l’inflation, à la prise en compte d’une structure par terme des
taux d’intérêts ou encore à l’intégration de changements de régimes. Enfin, il est à noter
que CEMAC-GLOBAL constitue une sorte de module de base pouvant, avec une relative
facilité (du fait de son caractère semi-structurel), intégrer des modules supplémentaires liés
par exemple à des analyses de l’emploi, des agrégats de monnaie ou encore des projections
sectorielles. N’ayant pas pour objectif l’exhaustivité (ce qui n’est d’aillaurs pas souvent
recommandé en matière de modélisation macroéconomique), ces différents aspects ne sont
pas abordés dans les présents travaux. Ils pourraient toutefois être adressés dans des
recherches ultérieures.

6
 Contents
Introduction 8

1 A brief overview of CEMAC Economies 12

1.1 A small open monetary union of oil-exporting developing countries . . . . . . . . . . . . . . . . . . . . . . . 12
1.2 An unconventional hard peg regime . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
1.3 Low financial development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
1.4 An Oil-dependent fiscal policy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

2 Intuitive Overview of the model 16

3 Model Equations 19
3.1 Notations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.2 Production equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.3 Phillips Curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.4 Fiscal policy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.5 Exchange rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.6 Financial sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.7 Central Bank . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
3.8 Interbank interest rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.9 Foreign Economy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

4 Parametrization 31
4.1 Calibration of balanced growth path parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
4.2 Trade and sectoral weights . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
4.3 Estimation of structural parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

5 Analysis of some shocks 34

5.1 Monetary policy shocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
5.2 Domestic demand shocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
5.3 Oil price shock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
5.4 International food price shock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
5.5 Fiscal policy shocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

6 Historical decompositions 44

7 Future extensions 46

Conclusion 46

References 48

Appendix 56

7
 Introduction

One of the most important economic policy questions when analyzing a monetary
union is whether the common monetary policy has the same effects in all member states.
This is because differences in the transmission of the common monetary policy to the
member economies of a monetary union could lead to asymmetries in the evolution of
business cycles and thus make the monetary policy decision-making process more difficult
Clausen and Hayo (2006). In addition, the policymaker in a monetary union needs to
evaluate the spillover effects of country-specific shocks to design the best stabilization
policy response. Although these issues have received a fair amount of attention in the
literature on currency unions in developed countries (mainly the Eurozone), few studies
have focused on these phenomena in developing country currency unions. In particular, the
CEMAC 1 , despite being among the oldest monetary unions in the world, has not received
much attention 2 , especially concerning the construction of macroeconomic models able
to conduct such analysis.

To address this issue, this paper introduces CEMAC-GLOBAL, a multi-country semi-

structural macroeconomic model for policy analysis and forecasting in the CEMAC 3 .
In addition, this modelling work is in line with several reforms engaged since the early
2010’s by the Bank of Central African States (hereafter BEAC 4 ) in order to improve its
analytical framework for monetary policy purposes (see Mvondo (2019b) for a discussion).

One of the key aspects of these reforms is the improvement of the macroeconomic mo-
deling framework. The historical tool for policy analysis at BEAC is based on the financial
programming framework of the IMF. However, since the beginning of the 2010s, several
studies conducted by BEAC’s staff have intended to propose new forecasting and simu-
lation tools mainly based on (i) time series analysis 5 and (ii) DSGE modelling. On this
last point, the studies were essentially focused on analyzing the monetary policy trans-
1. French acronym for Communauté Economique et Monétaire d’Afrique Centrale (Economic and
Monetary Community of Central African States)
2. This same issue in the "twin" monetary union of the Franc Zone, the West African Economic and
Monetary Union (UEMOA for the French acronym standing for Union Economique et monétaires Ouest
Africaine) could also be analyzed in future papers.
3. The country members are : Cameroon, Central African Republic, Congo, Gabon, Equatorial Guinea,
and Chad.
4. French acronym for Banque des Etats de l’Afrique Centrale.
5. For papers based on time series analysis, see for example Fossouo Kamga (2015), Bikai and Kenkouo
(2019), Bikai and Essiane (2017) or Bikai and Ngomba Bodi (2019).

8
mission channels and the effectiveness of BEAC monetary policy (Keungne K. and al.,
2016; Mvondo, 2019b).

However, most of these studies have considered the CEMAC as a single homogeneous
entity (Keungne K. and al., 2016; Mvondo, 2019b,a), a collection of countries analysed
separately (Fossouo Kamga, 2015) or both (Bikai and Essiane, 2017; Bikai and Mbohou
Mama, 2018; Bikai and Ngomba Bodi, 2019; Bikai and Kenkouo, 2019). These approaches
are quite limited to study the spillover effects of the monetary policy and various other
shocks through the member states of a monetary union. More, an issue with this ap-
proach is the relative inability to provide country-specific forecasts and simulations that
are coherent with monetary-union level forecasts and simulations. Then, the need for a
comprehensive multi-country modeling approach for policy analysis becomes critical. The
objective of this paper is to address this issue.

The CEMAC economy has its own specificities which constitute an interesting case
for macroeconomic modelling. First, it’s a monetary union of developing countries. There
is a quite rich literature on semi-structural macroeconomic modelling in the Eurozone
(Clausen and Hayo, 2006; Dieppe and al., 2018; Lang and Welz, 2018; Georgiadis and al.,
2021) 6 , which is a monetary union of developed economies, but few existing papers exten-
sively focus on monetary unions of developing countries. Second, most CEMAC countries
are oil exporters, which implies a slightly different behavior of these countries when facing
oil price shocks compared to oil importers. In particular, the fiscal policies of CEMAC
countries are strongly dependent on oil revenues, which in turn makes debt sustainability
heavily dependent on oil sector shocks (Wiegand, 2004). Third, CEMAC economies, like
many developing countries, record a low level of financial development, which has conse-
quences on the monetary policy transmission, especially with the low effectiveness of the
interest rate transmission channel (Mishra and Montiel, 2013; Mishra and al., 2014; Bikai
and Kenkouo, 2019). Fourth, the CEMAC exchange rate regime is an unconventional fixed
exchange rate regime where the Central Bank doesn’t intervene in the Forex to stabilize
the parity of the currency. This situation leads to a different design of the monetary po-
licy framework in a fixed exchange regime. Here, there is still space for monetary policy
intervention on the economy beyond the currency stability purposes, contrary to standard
6. The literature on semi-structural models in the developed world also incorporates several papers
focusing on one country like in Lemoine and al. (2019)

9
fixed exchange regime frameworks (Mvondo, 2019a).

The model developed in this study is a semi-structural DSGE model. Following the
seminal works of Poloz and al. (1994); Black and al. (1994); Coletti and al. (1996), it
consists of a reduced-form New Keynesian general equilibrium model expanded with some
ad-hoc features (Bokan and Ravnik, 2018). Its structure is largely inspired by the GPM6
of Carabenciov and al. (2013, 2008a) and ECB-Global of Dieppe and al. (2018). Like these
works, it’s a multi-country model that considers real-financial linkages in a New Keynesian
framework. But, contrary to GPM6 and ECB-Global, the model presents a small open
monetary union where all the member states are explicitly modelled. CEMAC-Global
is also inspired by the works of Hlédik and al. (2018) and Chafik (2019) which build
semi-structural models for oil-exporting countries in a fixed exchange rate regime.

Following the key features of the CEMAC economy, CEMAC-GLOBAL model de-
parts from these previous models in several ways. First, the specific features of the non-
conventional fixed exchange rate regime of CEMAC are explicitly modeled, to better
replicate the policy implications of the CFA Franc (XAF) currency arrangements. Se-
cond, it develops a more detailed monetary policy transmission mechanism beyond the
traditional interest rate transmission by including liquidity injections and credit cycles.
Third, the model distinguishes an oil sector and a non-oil sector and links them through
fiscal policy. Fourth, the monetary policy rules (interest rate and liquidity injections) inte-
grate the currency stability and the financial stability objectives, according to the current
policy orientations of the BEAC.

This modelling approach has several advantages. First, it permits to take into ac-
count the asymmetric effects of monetary policy shocks on the monetary union member
states. Second, it can be used to simulate various country-specific shocks and analyze their
spillover effects in the monetary union. Third, as mentioned by Carabenciov and al. (2013)
and Dieppe and al. (2018), the semi-structural structure adopted for CEMAC-GLOBAL
combines the advantages of fully structural models and those composed of reduced-form
equations. On the one hand, the structural nature of the model makes the equations and
shocks of the model have an explicit economic interpretation. On the other hand, the
reduced-form structure of the model "facilitates modifying the model in a flexible manner
so that it can be adapted relatively straightforwardly" (Dieppe and al., 2018) 7 .
7. As noted by Carabenciov and al. (2013), the semi-structural approach of macroeconomic modelling

10
 Following Carabenciov and al. (2013), CEMAC-GLOBAL parametrization is a com-
bination of calibrated and estimated parameters. This approach is suitable to increase
the empirical performances of the model and replicate key stylized facts of the CEMAC
economy. The main structural parameters and trade-related parameters are calibrated,
while the rest of the parameters are estimated with a bayesian approach.

The main properties of the model are illustrated in this paper by examining a set of
policy-relevant shocks and scenarios. First, monetary policy shocks seem to generate an
asymmetric response in CEMAC, with a less pronounced asymmetric effect when mani-
pulating liquidity injections. In addition, liquidity injections seem to have a lower delay
of impact on CEMAC countries than the interest rate manipulation. Second, domestic
demand shocks have limited spillover effects in the CEMAC, suggesting a low level of bu-
siness cycle synchronization in the monetary union. Third, only oil price shocks seem to
generate a common though heterogeneous response in the monetary union, reflecting the
high vulnerabilities of the member states to external factors. Fourth, the low level of trade
integration diminishes the negative expected impact of non-coordinated country-specific
fiscal policy shocks on the other economies.

Some empirical properties have been enlightened through historical decompositions.

They replicate the key stylized facts about the impact of the oil-price crisis of 2014 and the
Covid-19 pandemic. In addition, they show the importance of supply and foreign shocks
on inflation dynamics. More, the decompositions show asymmetric contributions of the
CEMAC-countries to economic growth.

The rest of the paper proceeds as follows. Section 1 presents some stylized facts about
CEMAC economies, section 2 presents an intuitive overview of CEMAC-Global and sec-
tion 3 presents the model equations in detail. Section 4 summarises the parametriza-
tion approach while sections 5 and 6 respectively analyze some shocks and examine the
historical decompositions of some observed variables. Section 7 mentions some possible
extensions of the model and the last section concludes.
allows the researcher to adopt an intermediary position between a DSGE and purely time series model.

11
1 A brief overview of CEMAC Economies

The CEMAC economy is an economic and monetary union composed of six develo-
ping countries. Created in 1994 8 , this Community is shaped by two main Unions : (i)
the Economic Union of Central Africa (UEAC) and (ii) the Monetary Union of Central
Africa (UMAC). The goal of UEAC is mainly to promote economic cooperation and in-
tegration within the CEMAC Zone, while UMAC structures monetary cooperation and
strengthens coordination between monetary and fiscal policies. It should be noted that
this monetary union exists since 1972 with the adoption of the CFA Franc (XAF) as the
common currency and the Bank of Central African States (BEAC) as the central bank of
the CEMAC area. Thus, UMAC is one of the oldest monetary unions in History.

The CEMAC economy gets some interesting features. Here, we’ll briefly review some
key stylized facts that CEMAC-global tries to capture. Especially, we insist on the fact
that (i) CEMAC is a monetary union of developing countries with (ii) an unconventional
hard peg regime. Besides, the CEMAC experiences a (iii) low financial development and
(iv) quite ineffective fiscal policy. Also, (v) the CEMAC economies are characterized by
a low level of unemployment and systemic underemployment evolving in a large informal
sector.

1.1 A small open monetary union of oil-exporting developing

countries

The member states of the CEMAC 9 are small open developing economies (Figure 1).
All of them, except the Central African Republic (CAR), are oil exporters. Oil and gas
rent represents on average 24% of CEMAC’s average GDP on 2010-2020 (Figure 2b).
8. But the economic cooperation amongst the member states started historically in June 1959 with
the Equatorial Customs Union (UDE for the french acronym), with the Central African Republic, the
Republic of Congo, Gabon, and Chad as member States. In 1961, Cameroon joined the UDE before the
creation of the Economic and Customs Union of Central Africa (UDEAC) in December 1964. Equatorial
Guinea joined UDEAC in march 1984, ten years before the creation of CEMAC.)
9. Cameroon, Central African Republic (CAR), Congo, Gabon, Equatorial Guinea and Chad

12
 Figure 1 – Average GDP per Capita - 2010-2020 (Current USD)

Source : World Bank, 2022 (World Development Indicators)

Figure 2 – Some GDP characteristics of CEMAC countries

(a) Average share of CEMAC countries in CEMAC (b) Share of non oil GDP in global GDP in CEMAC
GDP (2010-2020) (Average 2010-2020)

Source : Bank of Central African States (2022, Monetary Policy Report of first quarter)

This economic dependence on oil rents leads to an important exposure to oil price
volatility. For example, the 2014 oil prices drop had dramatic consequences on fiscal
revenues and, thus, on the economic performances of CEMAC countries.

13
1.2 An unconventional hard peg regime

Contrary to conventional fixed exchange rate regimes with imperfect capital mobility,
the fixed exchange rate regime of CEMAC is ruled by an agreement between the French
treasury and the member states of UMAC. Simply stated, the Central Bank isn’t obliged
to defend the parity of the currency with interventions on the Forex. The French Treasury
guarantees the convertibility of the XAF 10 and, in return, the BEAC has a remunerated
current account opened at the French Treasury where 50% of the international reserves
are on deposit. The main constraint faced by the Central bank here is to maintain the
ratio of the foreign asset to the Central bank’s Short-term liabilities (TCE 11 ) superior
to 20%. But, empirically, the monetary policy authority sets a target quite largely above
20%.

These characteristics lead to important changes in the way the fixed exchange rate
regime is managed and, then, modeled. First, as the nominal exchange rate with the Euro
is "law-determined" instead of "market-determined" there are no direct connexions between
the nominal policy rate and the nominal exchange rate. In other words, the Central bank
is in a fixed exchange rate where it’s not necessary to target the nominal exchange rate :
the monetary policy can still have some autonomy (at least, in the short run). Second,
as there’s imperfect capital mobility, the monetary policy autonomy is (again) reinforced.
Third, the agreement with the French Treasury discourages speculative attacks. The pool
of international reserves of the monetary union and the convertibility guarantee given
by the French Treasury lower the probability of a devaluation (relatively to other fixed
regimes in developing and emerging countries) and then makes speculative attacks more
costly for financial markets.

1.3 Low financial development

As mentioned in the literature, African developing countries have a low level of financial
development. CEMAC countries are no exception, with an access bank rate estimated at
15% in 2015. More, the global financial market capitalization is about 0.4% of CEMAC
GDP in 2018.
10. Currency code of the CFA Franc of CEMAC
11. French acronym for Taux de couverture exterieure de la monnaie (External currency coverage ratio).

14
 Figure 3 – Credit to private sector (% of GDP, average 2010-2018)

Source : World Bank, 2022 (World Development Indicators)

As a consequence, the bank lending channel 12 is generally considered the prominent

transmission channel of monetary policy in CEMAC (Bikai and Essiane, 2017; Bikai and
Kenkouo, 2019). However, even if the banking sector is one of the main financing sources
for the private sector in CEMAC, the credit to GDP ratio is relatively low (Figure 3). Thus,
more than 80% of credit to the private sector is captured by big firms, which reveals some
limited lending opportunities for small and medium enterprises (SMEs) (Fouopi Djiogap
and Ngomsi, 2012; Mandiefe Piabuo and al., 2015).

1.4 An Oil-dependent fiscal policy

Fiscal policies in CEMAC economies are heavily dependent on the dynamics of the oil
sector. In most cases, oil production is realized by foreign consortiums that share the oil
rent with the governments.
12. See Dwarkasing and al. (2017) for a literature review on banking lending channel of monetary
policy.

15
 Figure 4 – CEMAC fiscal revenues, CEMAC public expenses and oil prices

Source : IMF, World Economic Outlook (2022) and Bank of Central African States (2022)

The 2010 decade is an illustration of the structural dependence on oil rent of the
public sector in CEMAC. It has been characterized by a depletion of oil resources in
many CEMAC countries, which led to a diminution of global oil production, especially in
Equatorial Guinea. But the main shock that affected the CEMAC Economy has been the
oil price drop in 2014, which dramatically impacted the value of oil exports.

As a consequence, CEMAC countries experienced simultaneous fiscal stress which has

conducted the governments of CEMAC to adopt fiscal consolidation plans with the IMF
since 2017.

2 Intuitive Overview of the model

The model described in this paper could be classified as a semi-structural gap model
that takes into account the main features of a Neo-Keynesian model, following the ap-
proach of Poloz and al. (1994), Berg and al. (2006) and Carabenciov and al. (2008b). The
model also integrates specific equations in order to capture empirical characteristics of
the CEMAC economy.

First, the model is constructed as a small open monetary union of oil-exporting eco-
nomies. The six countries of CEMAC 13 are modeled symmetrically except for the Central
13. Cameroon, Central African Republic, Congo, Gabon, Equatorial Guinea, Chad

16
African Republic, which is not an oil producer. Each country has six blocs : (i) production,
(ii) prices, (iii) fiscal policy, (iv) financial sector, and (vi) exchange rate and interest rate
parity. A single Central Bank conducts monetary policy with its impulses on the policy
interest rate and liquidity injections on a simplified interbank market.

Second, the specificities of the CFA Franc peg regime are captured with (i) a fixed
nominal exchange rate with Euro and (ii) a monetary policy rule which sets the inter-
est rate to manage international reserves dynamics 14 beyond traditional output-inflation
trade-off.

Third, the model includes real-financial linkages. In particular, following the approach
proposed by Bianchi and Bigio (2022), the model explicitly includes the credit channel of
monetary policy, which is also prominent in the CEMAC economy. The interbank rate is
influenced by the main policy interest rate, the liquidity injections of the Central Bank
and the bank system stability of CEMAC. The interbank rate in turn affects the real
interest rate, which is a determinant of the credit cycle dynamics. Then, the credit cycle
influences the business cycles, which also affects the credit supply, exerting a feedback
effect on the credit cycle.

Fourth, the model distinguishes an oil output and a non-oil output for every CEMAC
country except the Central African Republic. The oil production dynamics are exogenous,
due to the "offshore" feature of the oil industry in CEMAC. Meanwhile, the oil production
dynamics influence the fiscal oil revenues, which in turn has an impact on the public ex-
penses dynamics. This modelling approach permits to assess quite simply the importance
of the oil prices and production for the fiscal policy, the business cycles and the bank
stability in CEMAC.

Fifth, the model includes some hysteresis effects. This is a useful feature to address
the long-term effects of short-term policies across the monetary union. Hysteresis effect
in CEMAC-GLOBAL goes mainly through the country-specific fiscal policies. Following
the results of Delong and Summers (2012) and Fatás and Summers (2018), the potential
output is affected by the public expenses dynamics, which are themselves affected by the
economic growth (which takes into account both output gap and potential output). This
constitutes a supplementary distinction of CEMAC-Global from most semi-structural and
14. As required by the monetary cooperation agreement between CEMAC countries and France.

17
micro-founded DSGE models in which, as noted by Shaughnessy (2011) usually ignore
hysteresis effects 15 .

For simplification purposes, and considering that CEMAC is a small open economy,
oil prices and foreign output and prices are modeled as exogenous.

Figure 5 describes the typical structure of a country in CEMAC-GLOBAL 16 .

Figure 5 – CEMAC-GLOBAL Scheme : Structure of each CEMAC country

Source : Author

15. Even the orthodox Arrow-Debreu general equilibrium model ignores hysteresis effects, as the equi-
librium is considered as independent of past shocks (Cross, 1993)
16. Except for the Central African Republic, which don’t have an oil sector.

18
3 Model Equations

3.1 Notations

This section describes the key equations of CEMAC-Global. As mentioned above,

countries of the CEMAC economy are modeled symmetrically except the Central African
Republic, which is not an oil producer.

Deviations from trends (or "gaps") are denoted by hats :

Xt
ct = log(
x ) (1)
X̄t

xbt = xt − x̄t (2)

where xt is the log of the level variable Xt and x̄t the log of its trend.

Trends variables are generally modeled as the sum of its own lagged value plus a
quarterly growth rate and a disturbance term that permits a permanent level shift :

¯ + G(x̄t ) + εxt¯t
x̄t = xt−1 (3)

The quarterly growth rate is proportional to its long-term value and, when mentioned,
growth rates of other endogenous variables. This feature is used to link "endogenously"
trends of some variables of the model :

G(x̄)
G(x̄t ) = αgxss + (1 − α)G(H̄)t + εt (4)

Where G(H̄)t is the trend growth rate of the variable (or group of variables) H.
Besides, ∆ refers to the quarterly growth rate and ∆4 to year on year (y-o-y) growth rate.

Following Carabenciov and al. (2008b) and Benes and al. (2017), the distributed lag
of some variables could be considered in the model. This feature is used to replicate a
hump-shaped response of a given variable after a shock and then avoid abrupt reactions

19
of the system. The distributed lag of a given variable x is noted as follows :

DL(xt ) = 0.04xt−1 + 0.08xt−2 + 0.12xt−3 + 0.16xt−4 + 0.20xt−5

+ 0.16xt−6 + 0.12xt−7 + 0.08xt−8 + 0.04xt−9 (5)

3.2 Production equations

The log of headline GDP of country i at time t Yit is the weighted sum of the oil and
non-oil GDP 17 (yitoil and yitnonoil ). These two subcomponents are themselves the sum of a
"gap" and a trend component. We then have :

yit = shoil oil oil nonoil

it yit + (1 − shit )yit (6)

where shoil
it is the share of oil GDP in the headline GDP of country i ∈ [Cameroon,

Central African Republic 18 , Congo, Gabon, Equatorial Guinea and Chad],

and

yitk = ybt k + ytk,ss , with k ∈ [oil, nonoil]. (7)

noil
For each CEMAC country i, the non-oil output gap (ybi,t ) depends on its own lead
and lagged values, and also depends on the domestic credit gap (ccri,t ), the foreign demand
∗ CEM AC
(non-CEMAC countries, ybi,t , and CEMAC countries, ybi,t ), the real exchange rate gap
noil
(zbi,t ), the current public expenditures gap (gbi,t ), and a domestic demand shock (εbyt ).

5
noil
a1,i noil
a1,i noil
ai2 ccri,t a1,i a1,i ∗ CEM AC
X
ybi,t = 0 y bi,t−1 + 1 y bi,t+1 + + 3 z bi,t + 4 [(1 − wi,j )ybi,t + ybi,t ]
j=1

ynoil noil
+ a1,i
5 g bi,t + σib εbyi,t (8)

17. Except for the Central African Republic, where the headline GDP corresponds to the non-oil GDP.
18. As the Central African Republic is not a net oil exporter, this parameters equals to zero in the
calibration of the model

20
 with

5
CEM AC
X
ybi,t = wi,j ybj,t , with j 6= i
j=1

and wi,j the weight of CEMAC country j in foreign trade of CEMAC country i.

Contrary to the typical IS curve described in the literature (Poloz and al. (1994),
Carabenciov and al. (2008a), Carabenciov and al. (2008b), Dieppe and al. (2018) or Benes
and al. (2017)), the output gap here is affected by the credit gap. This feature is adopted
due to the importance of the credit channel in CEMAC and the low financial markets
development 19 . This approach is also inspired by the literature on credit incidence on
business cycle (Bernanke and Gertler (1995), Borio (2012), Huizinga and Laeven (2019)).
Also, from a more empirical perspective, the data availability issue on private interest
rates could be overcome by using credit data, which is available at a monthly frequency.
noil
The growth rate of the non-oil sector potential output (G(y)i,t ) is specified as a
function of its long term value (gyinoil ), its lagged value, the quarterly growth rate of
G(y noil,ss )
public expenditures’s trend (∆gi,t ) 20 and a productivity shock (εt ).

noil noil noil

G(y)noil
G(y)i,t = a3,i 3,i 3,i i 3,i
0 G(y)i,t−1 + (1 − a0 )[(1 − a1 )gy + a1 ∆gi,t ] + σ
G(y)
εt (9)

The oil industry in CEMAC (like in most African oil-exporting countries) is quite dis-
connected from other sectors of the economy. The production is quasi exclusively realized
by foreign multinationals allowed by governments to exploit oil and gas fields (African
Development Bank and African Union, 2009; Diouf and Laporte, 2018). These multina-
tionals operate in the vast majority with imported physical and, often, human capital,
which implies that the oil sector has little connexion with other sectors. In order to simply
oil
capture this characteristic, the oil output (yi,t ) dynamics is exogenous :

oil oil oil y oil y oil

yi,t = yi,t−1 + ∆pri,t + σi i,t εi,ti (10)
19. The global financial market capitalization in CEMAC was about 0.5% of the GDP in 2019, which
is likely to lower the interest rate transmission channel. See https://www.financialafrik.com/2019/
10/29/cemac-la-bvmac-se-dote-dun-plan-daction-ambitieux-et-offensif/
20. See Fatás and Summers (2018) and Cerra and al. (2020).

21
 oil
with the growth rate of oil production ∆pri,t given by :

oil oil oil

∆pri,t = pri,t − pri,t−1 (11)

oil
with oil production (pri,t ) dynamics described by :

oil oil prioil prioil

pri,t = pri,t−1 + G(pr)oil
i,t + σi εi,t (12)

and

1,i 1,i G(pr)oil G(pr)oil

G(pr)oil oil oil
i,t = a0,o G(pr)i,t−1 + (1 − a0,o )gpri + σi
i
εi,t i (13)

where gprioil is the long-term growth rate of the oil production trend of country i.

3.3 Phillips Curve

q
Headline quarter-on-quarter (q-o-q) inflation (πi,t ) modelling follows a forward-backward
Phillips curve. Its dynamics is defined to be a function of the non-oil output gap, the real
f ood
exchange rate gap (zc
i,t ), the quarterly world food price dynamics and (πi,t ) a cost-push
q
shock (επi,t ).

q q 1,i q 1,i f ood q q

πi,t = b1,i 1,i noil
0 πi,t+1 + b1 πi,t−1 + b2 ybi,t + b1,i
3 zci,t + b4 πi,t + σiπ επi,t (14)

3.4 Fiscal policy

Fiscal policy intervention is determined by two opposite objectives. On one side, the
government aims to stabilize the nonoil sector output gap, and, on the other side, it
tries to stabilize the public debt dynamics. Here, the key policy instrument is the fiscal
expenditures 21 .
21. It’s assumed that the government has a little influence on fiscal revenues in the short run, so no
explicit rules are designed to manage them.

22
3.4.1 Public spending gap

The public spending gap (gbi,t ) is a function of its lagged value, the non oil sector
noil
output gap (ybi,t ), the public debt growth (∆DEBTi,t−1 ) and a short term fiscal policy
shock (εbgi,t ) :

g bg
gbi,t = c1,i
0 g bi,t−1 + c1,i
1 y
noil
bi,t + c1,i
2 ∆DEBTi,t−1 + σi εi,t
b
(15)

3.4.2 Public spending trend

The structural public spending growth (G(g i,t )) rate is a function of the nominal
output growth :

q
G(g i,t ) = c3,i oil 3,i noil
0 G(g i,t−1 ) + (1 − rshi )[c1 (G(y)i,t−1 + πi,t )]

3,i G(g) G(g)

+ rshoil oil oil
i [c2 (G(y)i,t−1 + ∆pt−1 + ∆U SDt )] + σi εi,t (16)

q
where rshoil
i is the share of oil revenues in total fiscal revenues of country i, πi,t the

quarterly inflation, ∆poil

t−1 the yearly average of the oil price q-o-q variation and ∆U SDt

the q-o-q variation of the XAF/USD exchange rate.

3.4.3 Public debt

For simplification purposes, the public debt growth (G(debt)i,t ) is a function of the
difference between the public spending growth rate (G(g)i,t ) and the nominal output
growth (G(ny)i,t ) 22 .

G(debt) G(debt)
G(debt)i,t = cd1,i 1,i
0 G(debt)i,t−1 + (1 − cd0 )[G(g)i,t − G(ny)i,t−1 ] + σi εi,t (17)

Thus, equation (18) describes the public debt evolution :

22. The nominal output growth is considered here as a proxy of the fiscal revenues growth.

23
 debti,t = debti,t−1 + G(debt)i,t + cd2,i debt debt
1 ∆U SDt + σi εi,t (18)

3.5 Exchange rate

According to the Franc Zone agreements between France and the CEMAC countries
members, the parity of the CFA Franc (XAF) is fixed. Nevertheless, this fixed exchange
rate parity is not a conventional one. As noted in section 1.2, the evolution of the nominal
exchange rate does not depend on the Central Bank’s interventions in the currency market.
Thus, CEMAC-GLOBAL considers the general specification of the nominal exchange rate
presented by Fall and Sy (2019) (through the uncovered interest rate parity) :

1
st = γ2 (st−1 + ∆sTt /4) + (1 − γ2 )[ζ1 (st−1 + ∆s¯t )
2
+ (1 − ζ1 )Et st+1 + (i∗t − it + P REMt )]/4 + εst (19)

where St is the nominal exchange rate, ∆StT is the desired exchange rate variation,
S̄t is the equilibrium exchange rate variation, i∗t the foreign interest rate, it the domestic
interest rate, Et the expectations term, and P REM the country risk premium. The γ2
coefficient is the level of fixity of the exchange rate. Fall and Sy (2019) as working on the
WAEMU zone 23 set γ2 close to 1, according to the fact that the WAEMU evolves in a
fixed exchange rate regime.

In this paper, γ2 is set to 1 and ∆StT is set to 0, since the evolution of the nominal
exchange is exclusively affected by the authorities. As noted by Benlamine and al. (2018),
the nominal exchange rate "is an exogenous policy choice variable. Without an ad hoc
policy decision, its value does not change." (p.11). Following the Benlamine and al. (2018)
approach, the nominal exchange rate between the XAF and the Euro (Steuro ) is given by :

euro euro
seuro
t = seuro s
t−1 + σi εsi,t (20)
23. The "twin" monetary union of CEMAC, whose have barely the same exchange rate regime agreement
with France.

24
 The real effective exchange rate dynamics (∆zi,t ) is simply a weighted average between
the real exchange rate of the XAF relative to the Euro and the dollar 24 :

euro,q q us,q q
∆zi,t = d3,i euro
0 (∆st + πi,t − πi,t ) + (1 − d3,i usd
0 )(∆st + πi,t − πi,t ) (21)

where ∆seuro
t and ∆susd
t represent the nominal exchange rates dynamics of the XAF
euro,q us,q
relative to the Euro and US Dollar respectively, and πi,t and πi,t represent the q-o-q
inflation rates for the eurozone and the US.

3.6 Financial sector

3.6.1 Country risk premium

The country risk premium (P REMi,t ) is a function of the spread between the domestic
real interest rate (i∗t ) and the foreign interest rate (πt∗ ), the equilibrium real exchange rate
(ziss ), the gap between the actual and the targeted level of international reserves (in
months of exports) (RESMIMPt − resmimp), the evolution of the public debt (∆Debti,t )
and the gap between the current debt-to-GDP ratio (Debtgap
i,t ) and the multilateral limit

of the debt-to-GDP ratio 25 and a risk premium shock (εPi,tREM ).

P REMi,t = d4,i 4,i ∗ ∗ ss

0 P REMi,t−1 + (1 − d0 )[r i,t − (it − πt ) − zi ]

gap
− d4,i 4,i 4,i P REM P REM
1 (RESMIMPt − resmimp) − d2 ∆Debti,t − d3 Debti,t + σi εi,t (22)

3.6.2 Bank Stability

Since the 2008 global financial crisis, the importance of financial stability in economic
literature and macroeconomic policy has arisen significantly (Agénor and al., 2011; Smets,
2014; Bauducco and al., 2014). In Africa and CEMAC in particular, this evolution is
illustrated by changes in macroprudential and monetary policy (see section 3.7) and a
24. For simplification purposes, the other regions have been ignored. Most of the financial international
transaction relations are denominated in Euro and US Dollar in the CEMAC region. Nevertheless, more
currencies could be incorporated into the model in future versions of the model (see section 7)
25. This modelling approach is adopted to considerate both the dynamics and the level of the public
debt in the evolution of the risk premium.

25
greater need for bank (in)stability indicators (see Bikai and Mbohou Mama (2018) for a
discussion).

For simplicity purposes, bank stability is affected through 2 main channels : the public
debt dynamics (see Das and al. (2010) and Assoumou (2017)) and the headline economic
growth 26 .

ST ABi,t = h40,i ST ABi,t−1 + (1 − h40,i )[−h41,i ∆Debti,t−1 +

(1 − h41,i )∆yi,t−1 ] + σiST AB εST

i,t
AB
(23)

3.6.3 Bank Lending Tightness

Following Carabenciov and al. (2008a,b, 2013) or Benes and al. (2017), the model
integrates a bank lending tightness (BLTi,t ) indicator, which captures the country-specific
constraints on credit to private sector 27 . Here, the bank lending tightness is influenced by
noil
(i) the expected non-oil sector’s output gap (ybi,t+1 ), (ii) the country-specific bank stability
dynamics (∆ST ABi,t−1 ), (iii) the evolution of the interbank interest rate (∆iinterbank
i,t ), (iv)
the progression of the central bank liquidity injections (∆cbinji,t ) and (v) a credit supply
shock (εBLT
i,t ).

noil
BLTi,t = h10,i BLTi,t−1 + (1 − h10,i )[−h11,i ybi,t+1 − h12,i ∆ST ABi,t−1

+ h13,i ∆iinterbank
i,t + h14,i ∆cbinji,t ] + σiBLT εBLT
i,t (24)
26. The rationale behind this modelling approach is quite simple : the banks mainly are exposed to
credit risk, which in turn are highly dependent on the economic dynamics of the oil and non-oil sector.
It is well documented that the economic conditions are key determinants of the quality of bank’s credit
portfolios and the evolution of non-performing loans (NPL) in the economy (Fofack, 2005; Beck and al.,
2013). These factors, in turn, tend to exacerbate bank instability (de Bandt and Hartmann, 2000; Benoit
and al., 2016). For simplicity purposes, CEMAC-GLOBAL links directly the economic growth dynamics
to bank stability. Obviously, in future versions of the model, NPL could be explicitly modeled.
27. According to Abuka and al. (2019)’s results, the bank lending tightness indicator influences both
volumes and rates of credit in CEMAC-global.

26
3.6.4 Credit

The credit gap ccri,t is influenced by its lagged value, the bank lending tightness indi-
cator BLTi,t , and the real interest rate gap rbi,t and :

ccri,t = h20,i ccri,t−1 + (1 − h20,i )[−h21,i BLTi,t − h22,i rbi,t ] (25)

and the growth rate of credit to private sector’s trend (G(cri,t )) is linked to the nominal
headline economic growth rate G(ny)i,t .

G(cr) G(cr)
G(cri,t ) = h30,i G(cri,t−1 ) + (1 − h30,i )(h31,i G(ny)i,t ) + σi εi,t (26)

3.7 Central Bank

The Bank of Central African States 28 (BEAC 29 ) sets its nominal interest rate ac-
cording to its monetary policy objectives : (i) the "internal" monetary stability objective
(price stability) and (ii) the "external" monetary stability objective (Banque des Etats de
l’Afrique Centrale, 2017). This latter objective is implied by the particular exchange rate
regime of the monetary union (see section 1.2).

Also, the Central Bank also monitors bank stability 30 . In 2017, the monetary authority
created a special lending facility to help banks facing liquidity stress 31 . The liquidity injec-
tions of the Central Bank can be seen as a relevant tool for financial stability purposes 32 .
Beyond this objective, liquidity injections are used by the Central Bank to minimize the
gap between the interbank market rate and the nominal policy rate.
28. The Central Bank of CEMAC.
29. Acronym for Banque des Etats de l’Afrique Centrale
30. see Caruana and al. (2011) for a more general discussion on the evolution of central banking in
Africa
31. Decision N°17/CPM/2017 of September 17, 2017 (available here https://www.beac.int/
wp-content/uploads/2016/10/Decision_12_CPM_2017.pdf)
32. This is even more accurate with the sharp increase of liquidity injections (+108.3 % for 2020Q1) in
the interbank market after the Covid-19 shock.

27
3.7.1 Nominal Policy rate

It should be noted here that there is a vivid debate amongst the policymakers and ana-
lysts about the more appropriate specification of the policy interest rate rule in CEMAC
and, more globally Franc Zone. In particular, there is no consensus about the way the
external stability objective of the Central Bank should be modelled. Strictly, the indicator
considered in the BEAC mandate is the external coverage rate of the money (ECRM) 33 ,
which is a ratio between the foreign assets and the current liabilities of the CEMAC eco-
nomy 34 . But the literature on the interest rate rule of CEMAC shows that alternative
indicators could be considered to capture the external monetary objective of the Central
Bank. Some papers integrate into the monetary policy rule the net foreign assets (Bikai
and Mbohou Mama (2016)), others consider the international reserves (Mvondo, 2019a,
2020) and, in a few cases, some papers simply ignore the external monetary stability ob-
jective (Keungne K. and al., 2016; Mvondo, 2019b). When looking at the "twin" monetary
union of the CEMAC - the West African Economic and Monetary Union (WAEMU) 35
-, researchers are also considering the external monetary objective in various ways. But
the most common feature is to simply ignore the external coverage rate of the money
(ECRM) or any other external monetary stability-related aggregate or indicator (Fall and
Sy, 2019; Diaw and Sall, 2018; Assemien and al., 2019). Instead, the focus is mainly on the
exchange rate stabilization process and the level of capital controls (Fall and Sy, 2019) 36 .

In this version of the CEMAC-GLOBAL model, the policy interest rate rule departs
from specifications founded in the literature in three ways. First, we consider the interna-
tional reserves in months of imports as the indicator of external monetary stability. This
choice is guided by the monetary policy strategy stated in the monetary policy reports
of BEAC (Banque des Etats de l’Afrique Centrale, 2021) : "From an operational perspec-
tive, the external monetary stability is achieved when the level of international reserves
represents 60% of the Central Bank short term liabilities, which is equivalent to at least
3 months of goods and services imports." (p.6). Second, the monetary policy rule only
considers the non-oil output gap as the activity indicator. This specification is in line with
33. "Taux de couverture extérieure de la monnaie" in french.
34. Which could be proxied by the broad money, see Mvondo (2018b) for a discussion.
35. Which members are Benin, Burkina Faso, Côte d’Ivoire, Guinea-Bissau, Mali, Niger, Senegal and
Togo.
36. which seems to be a general approach for fixed exchange rate regimes for scholars from the Inter-
national Monetary Fund (see Benlamine and al. (2018); Hamiani and al. (2020)).

28
the fact that the monetary policy transmission process hardly affects the oil sector evolu-
tion, which is mainly driven by exogenous factors : CEMAC countries are price-takers on
the global oil market, and the oil production capacities are given by the country-specific
natural resources endowment. These features should lead the central bank to be focu-
sed on the behavior of the non-oil sector business cycle when conducting its monetary
policy 37 . Third, the monetary policy targets the annual average inflation, instead of the
q-o-q inflation as considered by Bikai and Mbohou Mama (2016), Keungne K. and al.
(2016) and Mvondo (2019b). This choice is guided by the official objectives of the Central
Bank Banque des Etats de l’Afrique Centrale (2017) and the operational framework of
the monetary policy as clarified in the BEAC’s monetary policy reports Banque des Etats
de l’Afrique Centrale (2021).

Thus, in CEMAC-GLOBAL, the Central bank sets the nominal policy interest rate,
iPt OLICY , considering three main objectives : (i) Price stabilization, (ii) International re-
serves and (iii) output stabilization. In addition, in the long run, the policy rate is equal
to the neutral interest rate of the CEMAC region (iN,CEM
t
AC 38
) .

iPt OLICY = e10 iPt−1

OLICY
+ (1 − e10 )[iN,CEM
t
AC 4m,CEM AC
+ e11 (πt+1 − π)
P OLICY P OLICY
+ e12 ybtnoil + e13 (RESMIMPt − resmimp)] + σ i εit (27)

4m,CEM AC
where πt+1 is the yearly average of the CEMAC inflation, π is the inflation "tar-
get" of the Central Bank 39 , RESMIMPt is the level of international reserves (in months
of imports of goods and services) and resmimp is the Central Bank target of the inter-
national reserves (in months of imports of goods and services).
37. This specification of the monetary policy rule should not let the reader think that the oil sector
dynamics does not affect the behavior of the Central Bank. CEMAC-GLOBAL is structured in such a way
that the oil and non-oil sectors interact. For example, the non-oil output gap is affected by the oil sector
through fiscal policy. In addition, the fluctuations of the international reserves (which are considered in
the monetary policy rule) depend on the oil exports
38. which is a linear combination of the country-specific neutral interest rates.
39. Stricto sensu, the mandate of the BEAC does not define an inflation target. In practice, the Mo-
netary Policy Committee takes into consideration the inflation criteria of the CEMAC multilateral sur-
veillance framework (yearly average of 3%).

29
3.7.2 Liquidity injections

The Central bank sets its liquidity to minimize the spread between the interbank
interest rate and the policy rate (Mvondo, 2018a) 40 , manage liquidity in the bank system
and address financial stability issues (Banque des Etats de l’Afrique Centrale, 2019, 2021).
Here, liquidity injections (IN BCRt ) are set according to the spread between the policy
rate and the interbank interest rate, and the global bank stability of the monetary union.

CBIN JECt = i20 CBIN JECt−1 + (1 − i20 )[i21 (ItIN T ERBAN K − ItP OLICY )

− i22 ∆ST ABtCEM AC ] + σ CBIN JEC εCBIN

t
JEC
(28)

where

6
ST ABtCEM AC =
X
shi ST ABi,t (29)
i=1

P6
with shi the share of country in CEMAC real GDP and shi ∈ [0, 1]∀i and i=1 shi = 1.

3.8 Interbank interest rate

The interbank interest rate evolves according to the nominal policy rate (ItP OLICY ),
the Central Bank liquidity injections (IN BCRt ) and the bank stability index of CEMAC
(ST ABtCEM AC ).

ItIN T ERBAN K = i10 ItIN T ERBAN K + (1 − i10 )[ItP OLICY − i11 CBIN JECt
IN T ERBAN K IN T ERBAN K
− i12 ST ABtCEM AC ] + σ I εIt (30)
40. Again, the preamble of the monetary policy reports of BEAC (Banque des Etats de l’Afrique
Centrale, 2021) states that "To reach its objective, the approach chosen by the BEAC is to intervene on
the Money market, through open market operations, either by liquidity withdrawal or liquidity injections,
in order to drive the interbank interest rate around the main policy interest rate." (p.6)

30
3.9 Foreign Economy

For simplicity, and as CEMAC could be considered as a small open economy, the
foreign economy is a collection of exogenous variables 41 . The foreign demand gap, the
nominal exchange rates of commercial partners, the world price food index, and the oil
price are random walks. The nominal foreign interest rates and the foreign inflation rates
are given by AR(1) processes with constant terms.

4 Parametrization

Due to the large dimension of CEMAC-Global, a mixture of calibration and Bayesian

estimation is envisioned to parametrize the model (see Carabenciov and al. (2013) for
a discussion on this parametrization approach). Following Dieppe and al. (2018), the
parameters are divided into three groups : (i) those defining the balanced growth path
the nonstationary variables, (ii) trade and sectoral weights, and (iii) parameters with a
structural interpretation.

The first two groups are calibrated with historical data of CEMAC Countries, and the
third end fourth ones are estimated through a 3-steps procedure (see section 4.3).

Model solving and simulations are carried out in RISE Toolbox (Maih, 2019).

4.1 Calibration of balanced growth path parameters

CEMAC-GLOBAL is a non stationary model. As a consequence, key parameters for

the dynamics of the model are the term growth rate of the non stationary variable’s trends
(cf. parameter equation 4). They are typically computed as follow :

¯ (∆Xt ))]
θ̂(G(x̄t )) = median[HP (31)

¯ () is an operator for the trend component of the Hodrick-Prescott filter and

where HP
θ̂() is the symbol for calibrated value.
41. A key future extension of the model could be the improvement of this bloc. See section 7

31
4.2 Trade and sectoral weights

All the bilateral trade weights are computed with the international trade data from
the Atlas of Economic Complexity (The Growth Lab at Harvard University, 2019) 42 :

(importsi,j + exportsi,j )
wi,j = P5 (32)
j=1 ((importsi,j + exportsi,j ))

4.3 Estimation of structural parameters

The estimation of structural and shocks-related parameters follows a 3-steps procedure

(Figure 6).

Figure 6 – Estimation procedure of structural parameters of CEMAC-GLOBAL

4.3.1 Country-specific calibration through "Limited information" impulse res-

ponse matching

The first step of the parametrization of CEMAC-GLOBAL is the calibration of the

country-specific versions of CEMAC-GLOBAL through the limited information impulse
42. See Table 10.

32
response matching approach, described in Dieppe and al. (2018). In Dieppe and al. (2018),
the parameters with a structural economic interpretation is calibrated using the limited
information impulse response matching technique. This technique permits finding a confi-
guration of the parameter values that are consistent with findings of empirical and theo-
retical literature concerning the economy studied, especially in terms of the plausibility
of the impulse response functions for standard shocks. This approach is implemented in
two key steps. First, initial values are set by drawing on the literature on semi-structural
models as Carabenciov and al. (2008a, 2013); Dieppe and al. (2018); Benes and al. (2017);
Hlédik and al. (2018); Chafik (2019) as well as fully-structural models with a focus on
emerging and developing countries such as Liu and al. (2009), Peiris and Saxegaard (2010),
Berg and al. (2015), Alpanda and al. (2011) and Mvondo (2019b). In addition, some sty-
lized facts of CEMAC countries have been considered during the calibration exercise such
as low price elasticity of foreign trade (Essiane and Ngomba Bodi (2018)), low fiscal mul-
tipliers ((Bikai and al., 2017)) or high inertia of inflation ((Fossouo Kamga, 2015)). The
second step consists of fine-tuning the values of the key parameters and assessing the
plausibility of the impulse response functions. Following Dieppe and al. (2018), the para-
meters are chosen to informally minimize the distance between impulse-response functions
implied by CEMAC-Global and those found in the literature. However, this step appears
particularly tricky in the case of CEMAC-GLOBAL, due to the relative lack of empiri-
cal papers on the CEMAC which analyze the macroeconomic relations presented in the
model. In particular, the spillover effects of the monetary and fiscal policy in CEMAC
are barely analyzed, analyzing the empirical plausibility of the impulse responses quite
difficult 43 .

4.3.2 Bayesian estimation

As noted by Carabenciov and al. (2008a), the strength of bayesian estimated models
resides in the combination of reasonable fit to data, coherent structural results from
a theoretical perspective, and quite sound insights and results for policy decisions. In
the case of CEMAC-GLOBAL, parameters’ priors are given by the calibration exercise
conducted in the first step with limited information impulse response matching approach.
43. Nevertheless, this situation is challenging and could be a good incentive for further research in this
particular domain.

33
 Then estimation is conducted, using the bees algorithm developed by Yuce and al.
(2013) 44 as an optimization algorithm to obtain the posterior mode and standard devia-
tions. The bees algorithm is a computational model of swarm intelligence which mimics
the foraging behavior of honey bees to find the optimal solution to a given optimization
problem. This algorithm is particularly suitable when facing a high dimension optimiza-
tion problem, which is the case with a large number of model parameters. The estimation
procedure has been implemented with 10000 iterations and 20 scout bees (see Yuce and al.
(2013), p.652). The data used for the estimation are presented in the appendix.

4.3.3 Parametrization of the multicountry version of CEMAC-GLOBAL

Following the approach of Carabenciov and al. (2013), the estimated modes of the
country-specific estimation step are used to parametrize the multi-country version of
CEMAC-GLOBAL. Then the common parameters (especially the monetary policy-related
parameters) are calibrated following the "limited information" impulse response matching
to replicate the key stylized facts of the CEMAC region. The parameters are detailed in
the appendix.

5 Analysis of some shocks

In order to assess key characteristics of CEMAC-GLOBAL, some key shocks will be

analyzed here : (i) Monetary policy shocks, (ii) Domestic demand shocks, (iii) oil price
shock (iv) foreign demand shock, (v) international food price shock, and (vi) fiscal policy
shocks.

5.1 Monetary policy shocks

Monetary policy shocks are analyzed through 2 instruments : policy interest rate 45
and liquidity injections.

When the Central Bank raises its main policy interest rate, the level of liquidity
44. Coded by Maih (2019) in the RISE toolbox.
45. Reflecting the manipulation of the TIAO (french acronym for "Taux d’intérêt d’appels d’offres"),
which is the principal policy rate of BEAC

34
injections is lowered in order to reduce the supply of money in the interbank market
and exert pressure on the interbank interest rate to minimize the interbank spread. This
leads to a tightening of the bank credit to the private sector, which negatively affects the
output gap (and the economic growth) and the inflation (figure 7). In addition, a lower
demand, combined with a depreciation of the real exchange rate, leads to an increase in
international reserves 46 .
Figure 7 – 100 basis points policy rate shock : Aggregated impulse responses functions of some CEMAC
variables

Monetary Policy rate - CEMAC Central10

Bank
-3 liquidity injections to GDP ratio Credit
10 -4 growth (y-o-y) - CEMAC
0.01 0 0

0.008
-2 -1
0.006
-4 -2
0.004
-6
0.002 -3

0 -8
2 4 6 8 10 12 2 4 6 8 10 12 2 4 6 8 10 12
Quarters Quarters Quarters
Real
10 GDP
-4 growth rate - CEMAC (y-o-y) CEMAC
-4
10 average annual inflation rate Variation 10 of -3international reserves - CEMAC (y-o-y)
0
0 1

-0.5 0.8

0.6
-1 -1
0.4

-1.5 0.2
-2
0
2 4 6 8 10 12 2 4 6 8 10 12 2 4 6 8 10 12
Quarters Quarters Quarters
Bank
10 -5 Stability index - CEMAC Real effective
10 -4 exchange rate - CEMAC (y-o-y) Public
10 -5 expenses - (y-o-y growth,%)
0
2
15
-1
10
-2 1
5
-3
0
0
2 4 6 8 10 12 2 4 6 8 10 12 2 4 6 8 10 12
Quarters Quarters Quarters

Source : Author

When looking at the disaggregated effects of monetary policy shock, the results show
an asymmetric response of the CEMAC countries (Figure 8). The amplitude of the shocks
is quite different, with Congo having a greater reaction to the monetary policy tightening.
The key explanation here is the low level of effectiveness of the fiscal policy reaction to
interest rates hike. As estimated fiscal multipliers are smaller than the average of CEMAC
countries, fiscal policy stimulus does not generate enough growth to mitigate the effect
of the monetary policy tightening. Thus, the combined effect of the recession and the
increase in public expenses deteriorate the debt situation in Congo more than in the
46. The intuition behind this dynamics is that a depreciation of the real exchange rate tends to boost
exports through price competitiveness effect (see Bahmani-Oskooee and Ratha (2004)).

35
other CEMAC countries.
Figure 8 – 100 basis points policy rate shock : Disaggregated impulse response functions

Interest rates dynamic (%) Central Bank liquidity injections to annualized GDP ratio (%) Bank lending tightness indicator
1
Monetary Policy rate CEMAC 0.03
Interbank rate -0.4 0.02
0.5 Neutral Interest rate, CEMAC 0.01
-0.6
0
0 -0.8 -0.01
2 4 6 8 10 12 14 2 4 6 8 10 12 14 2 4 6 8 10 12 14
Quarters Quarters Quarters
-3
10Bank credit dynamics (q-o-q,%) 10 -3 Real GDP growth (q-o-q,%) Average inflation (%)
5 5
0 -0.01
0
-5 -0.02
-5
-10 -0.03
2 4 6 8 10 12 14 2 4 6 8 10 12 14 2 4 6 8 10 12 14
Quarters Quarters Quarters
International -4 -3
10reserves (y-o-y variation, months of imports) 10 Country Risk Premium (%) Real effective exchange rates (y-o-y variations,%)
10
5 0.03

5 CEMAC 0.02
0
0.01
0 -5
2 4 6 8 10 12 14 2 4 6 8 10 12 14 2 4 6 8 10 12 14
Quarters Quarters Quarters
10 -3 Bank Stability Index 10 -3 Public expenses growth (%) 10 -3 Public debt growth (%)
2 5 10
0 5
0
-2
-4 -5 0
-6 -10 -5
2 4 6 8 10 12 14 2 4 6 8 10 12 14 2 4 6 8 10 12 14
Quarters Quarters Quarters

Cameroon Central African Republic Congo Gabon Equatorial Guinea Chad

Source : Author

When the monetary policy shock starts from the central bank liquidity injections
(Figure 9), the transmission scheme to credit is similar, but the effect on inflation and
international reserves seems more rapid, with a less heterogeneous impact on the different
CEMAC countries (except for Congo). This result is similar to those of Bikai and Essiane
(2017) and Bikai and Kenkouo (2019), who noted that the liquidity channel seems to be
more effective than the interest rate channel in CEMAC.

36
 Figure 9 – Central bank liquidity shock : Disaggregated impulse response functions

Interest rates dynamic (%) Central Bank liquidity injections to annualized GDP ratio (%) Bank lending tightness indicator
0.08
0.06 Monetary Policy rate 0.02
-0.2
0.04 Interbank rate 0.01
0.02 Neutral Interest rate, CEMAC -0.4
CEMAC 0
0
2 4 6 8 10 12 14 2 4 6 8 10 12 14 2 4 6 8 10 12 14
Quarters Quarters Quarters
-3
10Bank credit dynamics (q-o-q,%) 10 -3 Real GDP growth (q-o-q,%) 10 -3 Average inflation (%)
0 2
0 0
-5 -2
-5
-10 -4
-6 -10
2 4 6 8 10 12 14 2 4 6 8 10 12 14 2 4 6 8 10 12 14
Quarters Quarters Quarters
-4
International
10reserves (y-o-y variation, months of imports) 10 -3 Country Risk Premium (%) 10 -3
Real effective exchange rates (y-o-y variations,%)
4 2 10
2 5
0
0 0
-2 CEMAC -2
-5
2 4 6 8 10 12 14 2 4 6 8 10 12 14 2 4 6 8 10 12 14
Quarters Quarters Quarters
10 -3 Bank Stability Index 10 -3 Public expenses growth (%) 10 -3 Public debt growth (%)
2 4 4
2 2
0 0 0
-2
-2 -2
-4
2 4 6 8 10 12 14 2 4 6 8 10 12 14 2 4 6 8 10 12 14
Quarters Quarters Quarters

Cameroon Central African Republic Congo Gabon Equatorial Guinea Chad

Note : For comparability purposes, the size of the liquidity shock has been calibrated in order to have the same negative
variation of the liquidity injections at time 1 than the policy interest rate shock described in figure 8.
Source : Author

Another key message here is that monetary policy can still have some effectiveness
despite the fixed exchange rate regime of the CFA zone with the eurozone. Contrary to
the conventional understanding of monetary policy under a fixed exchange rate regime
(Mundell, 1963; Swoboda, 1973; Frankel and al., 2004; Obstfeld and al., 2005; Rose, 2011),
the Central Bank in CEMAC still has some ability to manipulate its nominal interest rate
for stabilization purposes. The key factor behind this characteristic is the unconventional
peg regime permitted by the legal agreement between the French treasury and the CFA
countries. This agreement (i) exempts the Central Bank to sterilize its interventions on
the Forex to maintain the fixed exchange rate and (ii) protects the monetary union from
speculative attacks. Also, the capital controls applied by the monetary authority reinforce
this autonomy of the monetary policy.

5.2 Domestic demand shocks

Two domestic demand shocks are simulated : (i) Cameroon’s demand shock (Figure
10) and (ii) Central African Republic’s demand shock (Figure 11). These two countries

37
are respectively the largest and the smallest economies of CEMAC. Two main findings
could be highlighted.

First, CEMAC-GLOBAL reproduces typical demand shock reaction identified in the

literature (see Benes and al. (2017); Benlamine and al. (2018)). After the demand shock,
output and inflation rise and international reserves fall, leading to a hike in the nominal
policy interest rate. A higher level of economic activity improves bank stability and sti-
mulates credit, in line with real-financial linkages literature (see Bernanke and al. (1999)).
Here, the fiscal policy is counter-cyclical, with public spending decreasing after a positive
demand shock 47 .

Second, the spillover effects are quite small, even though the Cameroonian shock has
a larger effect on the other CEMAC countries than its Central African Republic’s coun-
terpart. This could be explained by the low level of trade integration in the monetary
union (see Table 10). However, the impact of country-specific demand shocks on larger
economies seems to have a positive impact in the short-run on these economies, and a
negative impact on the other countries of the region. This is a consequence of the negative
impact of the demand shock on international reserves, which increases the risk of external
monetary instability in the monetary union. The monetary policy tightening that follows
this sequence of responses contributes to bringing the system back to the balanced growth
path.
47. This result could appear counter-intuitive, considering several trends in the literature on fiscal
policy cyclicality in developing countries (see Gavin and Perotti (1997); Kaminsky and al. (2004); Talvi
and Végh (2005); Carmignani (2010); Bouanza and al. (2018)), but this model shows that the fiscal policy
response depends on the nature of the shock that hits the economies of CEMAC (cf. section 5.3)

38
 Figure 10 – Cameroon’s demand shock : Disaggregated impulse response functions

Interest rates dynamic (%) Central Bank liquidity injections to annualized GDP ratio (%) Bank lending tightness indicator
0.02 0.02
-0.005 CEMAC
Monetary Policy rate 0
-0.01
0.01 Interbank rate
Neutral Interest rate, CEMAC
-0.015 -0.02
-0.02 -0.04
0
2 4 6 8 10 12 14 2 4 6 8 10 12 14 2 4 6 8 10 12 14
Quarters Quarters Quarters
Bank credit dynamics (q-o-q,%) Real GDP growth (q-o-q,%) Average inflation (%)
0.06
0.5 0.1
0.04
0.02 0 0.05
0 0
-0.5
-0.02 -0.05
2 4 6 8 10 12 14 2 4 6 8 10 12 14 2 4 6 8 10 12 14
Quarters Quarters Quarters
International -3
10reserves (y-o-y variation, months of imports) Country Risk Premium (%) Real effective exchange rates (y-o-y variations,%)
0.02 0.05
2
0
0 0
CEMAC -0.05
-2
-4 -0.02 -0.1
-6 -0.15
2 4 6 8 10 12 14 2 4 6 8 10 12 14 2 4 6 8 10 12 14
Quarters Quarters Quarters
Bank Stability Index Public expenses growth (%) Public debt growth (%)
0.2
0.1 0
0.1 0
-0.05
-0.1
-0.2 -0.1
0
-0.3 -0.15
2 4 6 8 10 12 14 2 4 6 8 10 12 14 2 4 6 8 10 12 14
Quarters Quarters Quarters

Cameroon Central African Republic Congo Gabon Equatorial Guinea Chad

Note :The shock is calibrated to simulate a 100 percentage point increase of the non-oil output gap.
Source : Author

Figure 11 – Central African Republic Demand shock : Disaggregated impulse response functions

10 -4 Interest rates dynamic (%) -3

Central Bank10liquidity injections to annualized GDP ratio (%) Bank lending tightness indicator
15 -0.5 CEMAC 0
Monetary Policy rate
10 -1 -0.05
Interbank rate
-1.5 -0.1
5 Neutral Interest rate, CEMAC
-0.15
0 -2
2 4 6 8 10 12 14 2 4 6 8 10 12 14 2 4 6 8 10 12 14
Quarters Quarters Quarters
Bank credit dynamics (q-o-q,%) Real GDP growth (q-o-q,%) Average inflation (%)
0.2 0.1
0.6
0.1 0.4 0.05
0.2
0
0 0

2 4 6 8 10 12 14 2 4 6 8 10 12 14 2 4 6 8 10 12 14
Quarters Quarters Quarters
International -4
10reserves (y-o-y variation, months of imports) Country Risk Premium (%) Real effective exchange rates (y-o-y variations,%)
0.05 0.05
2
0 0 0
CEMAC
-2 -0.05 -0.05
-4 -0.1
-0.1
2 4 6 8 10 12 14 2 4 6 8 10 12 14 2 4 6 8 10 12 14
Quarters Quarters Quarters
Bank Stability Index Public expenses growth (%) Public debt growth (%)
0.2 0.1
0 0
0.1
-0.1 -0.1
0 -0.2 -0.2

2 4 6 8 10 12 14 2 4 6 8 10 12 14 2 4 6 8 10 12 14
Quarters Quarters Quarters

Cameroon Central African Republic Congo Gabon Equatorial Guinea Chad

Source : Author

39
5.3 Oil price shock

Three main results could be highlighted after a 10 percent increase shock in the oil
price (Figure 12).

First, an increase in the oil price leads to higher growth and inflation in oil-exporting
economies. A higher level of oil prices broadens fiscal revenues, thus extending the fiscal
policy space of oil-exporting countries. This larger policy space facilitates the expansion
of public expenses, reduces the country risk premiums of the oil-exporting economies,
and ease the credit conditions of most countries in the monetary union. These combined
effects boost the economic activity in the non-oil sector. This transmission scheme is in
line with several results in the literature on the procyclicality of fiscal policy in Africa or
oil-exporting developing economies (see Thornton (2008); Ilzetzki (2011); Mbemba (2011);
Berg and al. (2015); Bikai (2015); Bova and al. (2016); Chuku and al. (2018)).

Second, the results show that monetary policy reaction to an oil price shock is pro-
cyclical. This reaction stems from the external constraint faced by the Central bank due
to its policy objectives. The unconventional fixed exchange regime of CEMAC conducts
the Central Bank to target a level of international reserves (see Mvondo (2019a); Ngomba
Bodi (2022)). An increase in international reserves after a positive oil price shock leads
to a monetary easing, which exacerbates the procyclical behavior of the banking system
and the fiscal policy.

Third, an increase in the oil price has a positive, though relatively limited, impact
on the banking sector in Central African Republic, although it’s not an oil exporter
country. The monetary easing that follows the oil price shock, combined with the lower
risk premium of the monetary union, lead the banking sector to ease credit conditions in
Central African Republic.

40
 Figure 12 – Response to a 10 % increase in the Oil price

Interest rates dynamic (%) Central Bank liquidity injections to annualized GDP ratio (%) -3
10Bank lending tightness indicator
0.15 0
-0.05
0.1 -5
Monetary Policy rate
-0.1 Interbank rate -10
0.05
Neutral Interest rate, CEMAC CEMAC -15
-0.15
2 4 6 8 10 12 14 2 4 6 8 10 12 14 2 4 6 8 10 12 14
Quarters Quarters Quarters
Bank credit dynamics (q-o-q,%) Real GDP growth (q-o-q,%) 10 -3 Average inflation (%)
0.8 20
0.05
0.6 0.04
0.4 0.03 10
0.02
0.2 0.01
0
2 4 6 8 10 12 14 2 4 6 8 10 12 14 2 4 6 8 10 12 14
Quarters Quarters Quarters
International reserves (y-o-y variation, months of imports) Country Risk Premium (%) 10 -3
Real effective exchange rates (y-o-y variations,%)

0.15 CEMAC
0
0.1 -0.5
-10
0.05
0 -20
-1
2 4 6 8 10 12 14 2 4 6 8 10 12 14 2 4 6 8 10 12 14
Quarters Quarters Quarters
Bank Stability Index Public expenses growth (%) Public debt growth (%)
1.5 0.5
0.06
1 0
0.04
0.02 0.5 -0.5
0 0
2 4 6 8 10 12 14 2 4 6 8 10 12 14 2 4 6 8 10 12 14
Quarters Quarters Quarters

Cameroon Central African Republic Congo Gabon Equatorial Guinea Chad

Source : Author

5.4 International food price shock

Like many developing economies, CEMAC countries are significantly dependent on

food imports, making food price shock a key policy concern. A food price shock in
CEMAC-GLOBAL increases headline inflation and reduces the level of international re-
serves (more expensive imports) (Figure 13). In front of the monetary instability induced
by the increase in international food prices, the Central Bank tightens its policy by in-
creasing the interest rate and cutting the liquidity injections, to bring back the inflation
its the target. This sequence of responses tends to have a recessive effect on the economy,
increase the risk premium and generate instability in the banking system. In addition,
the Central African Republic seems to be the most harmed economy of CEMAC after
an increase in international food prices. This behavior could be explained by a narrower
supply of domestic food, which exposes the country to international food price shocks
more than other CEMAC countries.

41
 Figure 13 – International food price shock : Disaggregated impulse response functions

Interest rates dynamic (%) Central Bank liquidity injections to annualized GDP ratio (%) Bank lending tightness indicator
1.5
0.2 0 CEMAC
0.1 1
Monetary Policy rate -0.1
0 Interbank rate 0.5
-0.1 Neutral Interest rate, CEMAC -0.2

2 4 6 8 10 12 14 2 4 6 8 10 12 14 2 4 6 8 10 12 14
Quarters Quarters Quarters
Bank credit dynamics (q-o-q,%) Real GDP growth (q-o-q,%) Average inflation (%)
0.2 0.2 2
0 1
0
-0.2
0
-0.2 -0.4
-1
2 4 6 8 10 12 14 2 4 6 8 10 12 14 2 4 6 8 10 12 14
Quarters Quarters Quarters
International reserves (y-o-y variation, months of imports) Country Risk Premium (%) Real effective exchange rates (y-o-y variations,%)
0.8 1
0.02 0.6
0
0.01 0.4
-1
0 0.2
CEMAC -2
-0.01 0
2 4 6 8 10 12 14 2 4 6 8 10 12 14 2 4 6 8 10 12 14
Quarters Quarters Quarters
Bank Stability Index Public expenses growth (%) Public debt growth (%)
0 0.8 0.8
0.6 0.6
-0.1 0.4 0.4
-0.2 0.2 0.2
0 0
-0.3 -0.2 -0.2
-0.4
2 4 6 8 10 12 14 2 4 6 8 10 12 14 2 4 6 8 10 12 14
Quarters Quarters Quarters

Cameroon Central African Republic Congo Gabon Equatorial Guinea Chad

Source : Author

5.5 Fiscal policy shocks

Fiscal policy in a monetary union is a key issue in economic literature both on deve-
loped and developing countries 48 . In particular, a key concern is the level of coordination
of fiscal policy in monetary unions. Two fiscal policy scenarios has been simulated : (i) A
single country-specific shock (Figure 14) and (ii) a coordinated fiscal policy shock (Figure
15).

Country-specific fiscal stimulus increases demand, following the same scheme described
in section 5.2, except for public debt, country risk premium, and international reserves.
Here, the fiscal stimulus in Cameroon increases public debt and the Cameroonian risk
premium. The expansion of external debt temporarily increases the level of international
reserves and contributes to mitigating the monetary policy tightening induced by the
fiscal policy stimulus. Again, the spillover effects is low, with a noticeable effect only on
the Central African Republic.
48. See Avom (2007) for a discussion on CEMAC economies.

42
 Figure 14 – Cameroon’s fiscal policy shock : Disaggregated impulse response functions

10 -3 Interest rates dynamic (%) -3

Central Bank10liquidity injections to annualized GDP ratio (%) -3
10Bank lending tightness indicator
3
CEMAC 0
2 Monetary Policy rate -1
Interbank rate -5
1 -2
Neutral Interest rate, CEMAC
0 -3 -10
2 4 6 8 10 12 14 2 4 6 8 10 12 14 2 4 6 8 10 12 14
Quarters Quarters Quarters
-3
10Bank credit dynamics (q-o-q,%) Real GDP growth (q-o-q,%) 10 -3 Average inflation (%)
8 0.06 20
6 0.04
10
4 0.02
2 0 0
0 -0.02
2 4 6 8 10 12 14 2 4 6 8 10 12 14 2 4 6 8 10 12 14
Quarters Quarters Quarters
International -3
10reserves (y-o-y variation, months of imports) Country Risk Premium (%) Real effective exchange rates (y-o-y variations,%)
0.04 0.01
2 CEMAC
0
1 0.02
-0.01
0
0 -0.02
-1
2 4 6 8 10 12 14 2 4 6 8 10 12 14 2 4 6 8 10 12 14
Quarters Quarters Quarters
10 -3 Bank Stability Index Public expenses growth (%) Public debt growth (%)
10 0.15
0.5
5 0.1
0 0 0.05
0
-5 -0.5
2 4 6 8 10 12 14 2 4 6 8 10 12 14 2 4 6 8 10 12 14
Quarters Quarters Quarters

Cameroon Central African Republic Congo Gabon Equatorial Guinea Chad

Source : Author

The coordinated fiscal shock impulses have the same qualitative evolution as the
country-specific ones. This result is not surprising, considering the low level of trade
and financial integration in CEMAC 49 .
49. Another possible reason could be found in the structure of the model, which only connects the
outputs gaps of the countries of the monetary union. A future extension of the model could address
this issue by integrating also the trends. Nevertheless, the spillover effects could be a priori similar,
considering the low level of trade integration in CEMAC.

43
 Figure 15 – Coordinated fiscal policy shock : Disaggregated impulse response functions

Source : Author

6 Historical decompositions

The modelling approach of CEMAC-GLOBAL permits analyzing the contribution of

the shocks to the variables’ dynamics through both geographic and economic perspectives.
This is an advantage if the policy analyst wants, for example, to understand if the policy
interest rate evolution responds to symmetric or asymmetric dynamics of the economies
of a monetary union.

The decomposition of the main policy interest rate of BEAC 50 shows that the country-
specific economic shocks affected the TIAO in the same direction during the 2010 decade
(Figure 16). In the early 2010s, CEMAC countries positively contributed to the evolution
of the policy interest rate, in a context of financial conditions easing and a greater demand
affecting the economic activity (Figure 31 page 117), inflation (Figure 32 page 118) and
international reserves (Figure 33 page 119). Then, a second phase followed, characterized
by a negative contribution to TIAO of every CEMAC country. This phase regrouped
50. The TIAO, see section 5.1

44
two very different economic situations. First, the favorable profile of oil prices before the
second semester of 2014 significantly contributed to the evolution of international reserves,
thus relaxing the external constraint faced by the Central Bank. In addition, the low level
of inflation during this period didn’t encourage the Central Bank to tighten its monetary
policy. Second, the reversal of the oil price dynamics after the 2014 shock deteriorated
the economic growth in CEMAC (Figure 31). Combined with inflation still low in the
region, the monetary policy maintained its accommodative stance till late 2016. Then
the third phase started, with positive contributions to interest rate dynamics of every
CEMAC country, caused by their negative impact on international reserves evolution and
the impact of the COVID-19 pandemic.

This kind of analysis could be conducted for every variable of the model. For parsimony
purposes, only a few historical decompositions are presented in this paper (see appendix,
page 117).

Figure 16 – Historical decomposition of monetary policy rate (variations, y-o-y)

0.03
0.03

MonPolicy
Cameroon
CAR
Congo 0.02
0.02 Gabon
Guinea
Chad
Foreign
OilPrices
0.01
CapitalFlows
0.01 others
init

0
0

-0.01

-0.01

Demand
Supply
-0.02 MonetaryPolicy
OilRevenues
-0.02 ForeignShocks
CapitalFlows
FinancialShocks
-0.03 others
init

-0.03
1

3
Q

Q
10

11

13

14

16

17

19

20

10

11

13

14

16

17

19

20
20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

Source : Author

45
7 Future extensions

CEMAC-GLOBAL is an ongoing modelling project and the version presented in this

paper is a baseline specification. Although this version of CEMAC-GLOBAL is useful for
spillover analysis of several shocks in CEMAC and the study of the asymmetric effect of
monetary policy shocks, it could be improved in several ways. First, a more structural
specification of the foreign economy could be implemented, following Carabenciov and al.
(2013) or Dieppe and al. (2018). This approach would improve the simulation capacities of
the model concerning foreign shocks. Second, the model could disaggregate the inflation
dynamics to analyze the evolution of core, food and energy inflation more explicitly.
Considering the rising issues concerning inflation since the COVID-19 pandemics, this
refinement of the inflation analysis could be interesting (see Benlamine and al. (2018)).
Third, the model could take into consideration the term structure of the interest rates,
in order to integrate yield curves analysis in the simulations, following Dieppe and al.
(2018) and Benes and al. (2017). Fourth, this version of CEMAC-GLOBAL does not
explicitly consider the degree of capital control in the monetary policy rule, contrary to
Benlamine and al. (2018). Even if the Central Bank does not explicitly apply strict capital
control, including this feature could serve for policy simulation purposes. Fifth, the model
could include regime-switching features, to reflect various macroeconomic regimes that
could lead to different economic policy behaviors, as noted in Mvondo (2019a). These
limitations can be addressed in future extensions of the model.

Conclusion

This paper introduced a semi-structural macroeconomic model for policy analysis, risk
assessment, and forecasting in the CEMAC monetary union. It departs from previously
known models of the CEMAC economy mainly by modelling every country of the mo-
netary union, which allows studying the asymmetric effects of the monetary policy and
the spillover effect of various shocks like national fiscal policy shocks. The model also (i)
captures the features of the unconventional fixed exchange rate regime of the CEMAC, (ii)
distinguishes the oil and non-oil output with an interaction between these sectors through
fiscal policy, (iii) specifies a detailed monetary policy transmission by considering liqui-

46
dity injections and a simple interbank interest rate dynamics, (iv) includes real-financial
linkages with a credit cycle dynamics affecting the business cycle and (v) integrates some
hysteresis effects with fiscal stimulus depending on the headline economic growth.

The potential use of the model has been illustrated by discussing country-specific
and monetary-union level responses to monetary policy shocks, demand shocks, oil price
shock, foreign demand shock, international food price shock, country-specific and coordi-
nated fiscal shocks and credit supply shock. In addition, the empirical performance has
been assessed through the historical decomposition of the GDP growth, inflation, interna-
tional reserves, and bank credit. These simulations showed that monetary policy has an
asymmetric effect on CEMAC countries, with a greater impact on countries with a less
prominent oil sector. In addition, the size of the spillover effects seems to increase with
the size of the country where the shock occurs. Also, coordinated fiscal impulses have a
greater positive impact on the CEMAC output than country-specific ones, however with
a negative impact on international reserves and the country risk premium. Obviously, a
vast variety of simulations and scenarios can be analyzed with CEMAC-GLOBAL, and
numerous studies could explore specific policy issues based on this model.

CEMAC-GLOBAL is still an ongoing modelling process and future versions could

address progressively its limitations, among which are the lack of granularity of inflation
dynamics, the oversimplified foreign sector modelling, or the absence of regime-switching
behavior.

47
 References
Abuka, C., Alinda, R. K., Minoiu, C., Peydró, J. L. and Presbitero, A. F. (2019) ; « Monetary
policy and bank lending in developing countries : Loan applications, rates, and real effects » :
Journal of Development Economics, 139(November 2018), p. 185–202

African Development Bank and African Union (2009) ; Oil and gas in Africa : Oxford University
Press, 231 p.

Agénor, P.-R., da Silva, L. A. P. and da Silva, L. A. P. (2011) ; « Macroeconomic Stability,

Financial Stability, and Monetary Policy Rules » : Ferdi Working Papers, (29)

Alpanda, S., Kotzé, K. and Wolgom, G. (2011) ; « Forecasting Performance Of An Estima-

ted Dsge Model For The South African Economy » : South African Journal of Economics,
79(March), p. 50–67

Assemien, A., Esso, L. J., Kanga, K. D. and Esso, L. J. (2019) ; « Can Monetary Policy Influence
Employment ? The case of West African States » : Revue d’Economie Politique, 129(5), p.
777–813

Assoumou, O. (2017) ; « Public debt and financial stability : The case of economic community
and monetary union of central Africa (EMCCA) » : Journal of Economics and International
Finance, 9(9), p. 89–94

Avom, D. (2007) ; « La coordination des politiques budgétaires dans une union monétaire :
l’expérience récente des pays de la CEMAC » : Revue Tiers Monde, 4(192), p. 871–893

Bahmani-Oskooee, M. and Ratha, A. (2004) ; « The J-curve : A literature review » : Applied

Economics, 36(13), p. 1377–1398

de Bandt, O. and Hartmann, P. (2000) ; « Systemic risk : A survey » : SSRN Electronic Journal

Banque des Etats de l’Afrique Centrale (2017) ; « Statuts de la Banque des Etats de l’Afrique
Centrale » :

Banque des Etats de l’Afrique Centrale (2019) ; « Bulletin du Marché Monétaire » : Décembre

Banque des Etats de l’Afrique Centrale (2021) ; « Rapport sur la Politique Monétaire (Décembre
2021) » : Rapport technique, Banque des Etats de l’Afrique Centrale

Bauducco, S., Christiano, L. J. and Raddatz, C. E. (2014) ; « Macroeconomic and financial

stability : an overview » : Series on Central Banking Analysis and Economic Policies, (19)

48
Beck, R., Jakubik, P. and Piloiu, A. (2013) ; « Non-performing loans : What matters in addition
to the economic cycle ? » : ECB Working Paper Series, (1515)

Benes, J., Clinton, K., George, A., Gupta, P., John, J., Kamenik, O., Laxton, D., Mitra, P.,
Nadhanael, G., Portillo, R., Wang, H. and Zhang, F. (2017) ; « Quarterly Projection Model
for India : Key Elements and Properties » : IMF Working Paper Series, (WP/17/33)

Benlamine, M., Bulíř, A., Farouki, M., Horváth, Á., Hossaini, F., Szilágyi, K., Taamouti, M.
and Vávra, D. (2018) ; « Morocco : A Practical Approach to Monetary Policy Analysis in a
Country with Capital Controls » : IMF Working Paper Series, (WP/18/27)

Benoit, S., Colliard, J.-E., Hurlin, C. and Pérignon, C. (2016) ; « Where the risks lie a survey
on systemic risk » : Review of Finance, (p. 1–44)

Berg, A., Karam, P. and Laxton, D. (2006) ; « Practical Model-Based Monetary Policy Analysis
— A How-To Guide » : IMF Working Paper Series, (WP/06/81)

Berg, A., Yang, S.-C. S. and Zanna, L.-F. (2015) ; « Modeling African Economies : A DSGE
Approach » : Dans Monga, C. and Lin, J. Y. (Rédacteurs) The Oxford Handbook of Africa
and Economics, tome 1 : Context, (p. 1–30)

Bernanke, B. S. and Gertler, M. (1995) ; « Inside the Black Box : The Credit Channel of Monetary
Policy Transmission » : Journal of Economic Perspectives, 9(4), p. 27–48

Bernanke, B. S., Gertler, M. and Gilchrist, S. (1999) ; « The financial accelerator in a quantitative
business cycle framework » : Handbook of Macroeconomics, 1(PART C), p. 1341–1393

Bianchi, J. and Bigio, S. (2022) ; « Banks, Liquidity Management, and Monetary Policy » :
Econometrica, 90(1), p. 391–454

Bikai, J. L. (2015) ; « Règles de surveillance multilatérales et pro cyclicité de la politique bud-

gétaire dans la zone CEMAC » : BEAC Working Paper Series, 2015(BWP N°07/15)

Bikai, J. L. and Essiane, P.-N. D. (2017) ; « Politique monétaire, stabilité monétaire et croissance
économique dans la CEMAC : une approche SVAR bayésienne » : BEAC Working Paper
Series, (BWP 08/17)

Bikai, J. L. and Kenkouo, G. A. (2019) ; « Analysis and evaluation of the transmission channels
of monetary policy in CEMAC : A SPVAR-X approach » : Mondes en développement, 185(1),
p. 109–132

49
Bikai, J. L. and Mbohou Mama, M. (2018) ; « Stabilité bancaire et stabilité macroéconomique
dans la CEMAC » : BEAC Working Paper Series, (12/18)

Bikai, J. L. and Ngomba Bodi, F. G. (2019) ; « Les prévisions conditionnelles sont-elles plus
précises que les prévisions inconditionnelles dans les projections de croissance et d’inflation
en zone CEMAC ? » : BEAC Working Paper Series, (03/19), p. 30

Bikai, J. L., Yogo, T. U. and Essiane, P.-N. D. (2017) ; « Quantification du multiplicateur bud-
gétaire dans les pays de la CEMAC » : BEAC Working Paper Series, (BWP 07/17)

Bikai, L. and Mbohou Mama, M. (2016) ; « Une fonction de réaction pour la Banque des Etats
de l’Afrique Centrale dans un contexte de dominance budgétaire » : BEAC Working Paper
Series, (BWP 04/16), p. 1–28

Black, R., Laxton, D., Rose, D. and Tetlow, R. (1994) ; « The Steady-State Model : SSQPM » :
Dans The Bank of Canada’s New Quarterly Projection Model, chapitre The Steady, Bank of
Canada

Bokan, N. and Ravnik, R. (2018) ; « Quarterly Projection Model for Croatia » : Croatian National
Bank Surveys, (S-34)

Borio, C. (2012) ; « The financial cycle and macroeconomics : What have we learnt ? » : BIS
Working Papers, (395)

Bouanza, J. R. F. K., Ngassa, T. C. and Ndinga, M. M. A. (2018) ; « The Effects of Regulatory

Quality on the Pro-Cyclicality of Fiscal Policy in Countries in the Central African Economic
and Monetary Community (CA EMC) » : Modern Economy, 09(07), p. 1228–1246

Bova, E., Medas, P. and Poghosyan, T. (2016) ; « Macroeconomic Stability in Resource-Rich

Countries : The Role of Fiscal Policy » : IMF Working Papers, (WP/16/36)

Carabenciov, I., Ermolaev, I., Freedman, C., Juillard, M., Kamenik, O., Korshunov, D. and
Laxton, D. (2008a) ; « A Small Quarterly Projection Model of the US Economy » : IMF
Working Papers, 08(278), p. 1

Carabenciov, I., Ermolaev, I., Freedman, C., Kamenik, O., Juillard, M., Korshunov, D., Laxton,
D. and Laxton, J. (2008b) ; « A Small Quarterly Multi-Country Projection Model » : IMF
Working Papers, 08(279), p. 1

50
Carabenciov, I., Freedman, C., Garcia-Saltos, R., Laxton, D., Kamenik, O. and Manchev, P.
(2013) ; « GPM6 : The Global Projection Model with 6 Regions » : IMF Working Papers,
13(87), p. 1

Carmignani, F. (2010) ; « Cyclical fiscal policy in Africa » : Journal of Policy Modeling, 32(2),
p. 254–267

Caruana, J., Hawkins, P., Jeanneau, S., Christensen, B. V. V., Christensen, B. V., Rangasamy,
L., Mihaljek, D. and Yunus, M. (2011) ; « Central Banking in Africa : Prospects in a Changing
World » : BIS Papers, (56)

Cerra, V., Fatás, A. and Saxena, S. C. (2020) ; « Hysteresis and Business Cycles » : IMF Working
Paper Series, (WP/20/73)

Chafik, O. (2019) ; « Monetary Policy in oil exporting countries with fixed exchange rate and
open capital account : expectations matter » : MPRA Paper, (92558), p. 0–33

Chuku, C., Simpasa, A. and Oduor, J. (2018) ; « Macroeconomic Consequences of Commodity

Price Fluctuations in African Economies » : African Development Review, 30(4), p. 329–345

Clausen, V. and Hayo, B. (2006) ; « Asymmetric monetary policy effects in EMU » : Applied
Economics, 38(10), p. 1123–1134

Coletti, D., Hunt, B., Rose, D. and Tetlow, R. (1996) ; « The Dynamic Model : QPM » : Dans The
Bank of Canada’s New Quarterly Projection Model, chapitre The Dynami, Bank of Canada

Cross, R. (1993) ; « On the Foundations of Hysteresis in Economic Systems » : Economics and

Philosopy, 9(October 2009), p. 53–74

Das, U. S., Papapioannou, M., Pedras, G., Ahmed, F. and Surti, J. (2010) ; « Managing Public
Debt and its Financial Stability Implications » : IMF Working Papers, 10(280), p. 1

Delong, J. B. and Summers, L. (2012) ; « Fiscal Policy in a Depressed Economy » : Brookings

Papers on Economic Activity, 43(1 (Spring)), p. 233–297

Diaw, A. and Sall, A. K. (2018) ; « Inflation Targeting : What Rule for the Central Bank of West
African States ( BCEAO ) ? » : Revue d’Economie du développement, 26(2), p. 5–49

Dieppe, A., Georgiadis, G., Ricci, M., Robays, I. V. and van Roye, B. (2018) ; « ECB-Global :
introductin ECB’s global macroeconomic model for spillover analysis » : Economic Modelling,
72(June 2018), p. 78–98

51
Diouf, A. and Laporte, B. (2018) ; « Oil contracts and government take : Issues for Senegal and
developing countries » : Journal of Energy and Development, 43(2), p. 213–234

Dwarkasing, M., Dwarkasing, N. and Ongena, S. (2017) ; « The Bank Lending Channel of Mo-
netary Policy : A Review of the Literature and an Agenda for Future Research » : Dans The
Palgrave Handbook of European Banking, (p. 383–407)

Essiane, P.-N. D. and Ngomba Bodi, F. G. (2018) ; « Estimation des élasticités du commerce
extérieur dans des économies en développement riches en ressources naturelles : le cas des
pays de la CEMAC » : BEAC Working Paper Series, (BWP 05/15)

Fall, A. and Sy, D. (2019) ; « La politique monétaire de la BCEAO face au double ancrage :
ciblage d’inflation et ciblage du taux de change » : Revue d’Economie du développement,
27(4), p. 129–175

Fatás, A. and Summers, L. H. (2018) ; « The Permanent Effects of Fiscal Consolidations » :

Journal of International Economics, 112(May 2018)

Fofack, H. (2005) ; « Nonperforming loans in Sub-Saharan Africa : causal analysis and macroe-
conomic implications » : World Bank Policy Research Working Paper Series, (3769)

Fossouo Kamga, A. L. (2015) ; « Etude des propriétés de long terme de l’inflation dans les pays
de la CEMAC : Une approche ARFIMA » : BEAC Working Paper Series, (04/15)

Fouopi Djiogap, C. and Ngomsi, A. (2012) ; « Determinants of Bank Long-Term Lending Beha-
vior in The Central African Economic and Monetary Community ( CEMAC ) » : Review of
Economics and Finance, (May 2012), p. 107–114

Frankel, J., Schmukler, S. L. and Serven, L. (2004) ; « Global transmission of interest rates :
monetary independence and currency regime » : Journal of international Money and Finance,
23(5), p. 701–733

Gavin, M. and Perotti, R. (1997) ; « Fiscal Policy in Latin America » : NBER Macroeconomics
Annual, 12(1997), p. 11–61

Georgiadis, G., Hildebrand, S., Ricci, M., Schumann, B. and Roye, B. V. (2021) ; « ECB-Global
2.0 : a global macroeconomic model with dominant-currency pricing, tariffs and trade diver-
sion » : ECB Working Paper Series, March 2021(2530)

Hamiani, E., Buessing-loercks, M. and Fleuriet, V. (2020) ; « Monetary Policy Under an Ex-
change Rate Anchor » : IMF Working Paper Series, (WP/20/180)

52
Hlédik, T., Musil, K., Rysanek, J. and Jaromir, T. (2018) ; « A macroeconomic Forecasting Model
of the FIxed Exchange Rate Regime for the Oil-Rich Kazakh Economy » : Czech National
Bank Working Paper Series, (11)

Huizinga, H. and Laeven, L. (2019) ; « The Procyclicality of Banking : Evidence from the Euro
Area » : ECB Working Paper Series, (2288)

Ilzetzki, E. (2011) ; « Rent-seeking distortions and fiscal procyclicality » : Journal of Development

Economics, 96(1), p. 30–46

Kaminsky, G. L., Reinhart, C. M. and Végh, C. A. (2004) ; « When It Rains, It Pours : Procyclical
Capital Flows and Macroeconomic Policies » : NBER Macroeconomics Annual, 19(October),
p. 11–53

Keungne K., L. S., Mounkala, E. U. H. and Mukam M., A. (2016) ; « Optimalité de la politique
monétaire de la BEAC depuis 2001 » : BEAC Working Paper Series, (03/16)

Lang, J. H. and Welz, P. (2018) ; « Semi-structural credit gap estimation » : ECB Working Paper
Series, (2194)

Lemoine, M., Turunen, H., Lepetit, A., Zhutova, A., Clerc, P. and Laffargue, J.-p. (2019) ; « The
FR-BDF Model and an Assessment of Monetary Policy Transmission in France » : Banque de
France Working Paper Series, (736)

Liu, G., Gupta, R. and Schaling, E. (2009) ; « A new-keynesian DSGE model for forecasting the
South African Economy » : Journal of Forecasting, 28(5), p. 387–404

Maih, J. (2019) ; « Rationality In Switching Environments (RISE) Toolbox » :

URL https://github.com/jmaih/RISE_toolbox

Mandiefe Piabuo, S., Menjo Baye, F. and Julius, C. T. (2015) ; « Effects of credit constraints on
the productivity of small and medium-sized enterprises in Cameroon » : Journal of Economics
and International Finance, 7(9), p. 204–212

Mbemba, A. (2011) ; « Fiscal Policy in Selected African Countries : The CFA Franc Zone » :
International Journal of Economics and Finance, 3(2), p. 101–110

Mishra, P. and Montiel, P. (2013) ; « How effective is monetary transmission in low-income

countries ? A survey of the empirical evidence » : Economic Systems, 37(2), p. 187–216

53
Mishra, P., Montiel, P., Pedroni, P. and Spilimbergo, A. (2014) ; « Monetary policy and bank len-
ding rates in low-income countries : Heterogeneous panel estimates » : Journal of Development
Economics, 111, p. 117–131

Mundell, R. A. (1963) ; « Capital Mobility and Stabilization Policy under Fixed and Flexible
Exchange Rates » : The Canadian Journal of Economics and Political Science, 29(4), p.
475–485

Mvondo, E. T. (2018a) ; « A la recherche de l’amplitude optimale du corridor des facilités per-

manentes de la BEAC » : BEAC Working Paper Series, (BWP 11/18), p. 1–18

Mvondo, E. T. (2018b) ; « La pertinence du TIAO comme instrument de politique monétaire de

la BEAC : Fondements théoriques et évaluation empirique » : BEAC Working Paper Series,
(03/17)

Mvondo, E. T. (2019a) ; « Oil Shocks and Macroeconomic Effects of Occasionally Binding

Constraint on External Reserves of CEMAC » : International Journal of Business and Eco-
nomic Research, 8(6), p. 422–438

Mvondo, E. T. (2019b) ; « Réformes financières et formation des taux d’intérêt en Afrique Cen-
trale : une approche à travers le modèle DSGE » : Revue Interventions Economiques, (61)

Mvondo, E. T. (2020) ; « Stabilisation et relance macroéconomique post COVID-19 dans la

CEMAC. Quels instruments pour quels effets dans un modèle DSGE ? » : BEAC Working
Paper Series, (01/20)

Ngomba Bodi, F. G. (2022) ; « External Constraints and Procyclicality of the Monetary Policy
of the Bank of Central African States » : BEAC Working Paper Series, (Forthcoming)

Obstfeld, M., Shambaugh, J. C. and Taylor, A. M. (2005) ; « The trilemma in history : tradeoffs
among exchange rates, monetary policies, and capital mobility » : Review of economics and
statistics, 87(3), p. 423–438

Peiris, S. J. and Saxegaard, M. (2010) ; « An estimated dynamic stochastic general equilibrium

model for monetary policy analysis in Mozambique » : IMF Staff Papers, 57(1), p. 256–280

Poloz, S., Rose, D. and Tetlow, R. (1994) ; « The Bank of Canada’s new Quarterly Projection
Model (QPM) : An introduction » : Bank of Canada Review, 20, p. 23–38

Rose, A. K. (2011) ; « Exchange rate regimes in the modern era : fixed, floating, and flaky » :
Journal of Economic Literature, 49(3), p. 652–672

54
Shaughnessy, T. O. (2011) ; « Hysteresis in unemployment » : Oxford Review of Economic Policy,
27(2), p. 312–337

Smets, F. (2014) ; « Financial Stability and Monetary Policy : How Closely Interlinked ? » :
International Journal of Central Banking, 10(2)

Swoboda, A. K. (1973) ; « Monetary Policy under Fixed Exchange Rates : Effectiveness , the
Speed of Adjustment and Proper Use » : Economica, 40(158), p. 136–154

Talvi, E. and Végh, C. (2005) ; « Tax Base Variability and Procyclicality of Fiscal Policy » :
Journal of Development Economics, 78

The Growth Lab at Harvard University (2019) ; « International Trade Data (HS,92) » : Harvard
Dataverse, V2

Thornton, J. (2008) ; « Explaining procyclical fiscal policy in African countries » : Journal of

African Economies, 17(3), p. 451–464

Wiegand, J. (2004) ; « Fiscal Surveillance in a Petro Zone : The Case of the CEMAC » : IMF
Working Papers, 008

Yuce, B., Packianather, M. S., Mastrocinque, E., Pham, D. T., Lambiase, A. and Parade, T.
(2013) ; « Honey Bees Inspired Optimization Method : The Bees Algorithm » : Insects, 4, p.
646–662

55
 Appendix

Data

Table 1 – Description of Data

Observable Variable name Indicator Source Period Observations

∗
ybi,t Foreign output Output gap USA St. Louis Fe- 2000Q1- Cyclical component of the HP-filtered US real GDP
gap deral Reserve 2021Q2
https://fred.
stlouisfed.
56

org/graph/?g=
f1cZ
Continued on next page
 – Continued from previous page

Observable Variable name Indicator Source Period Observations

poil
t Oil prices Log of Brend oil US Energy 2000Q1-
prices (current US Information 2021Q2
dollars per barrel) Administra-
tion, https:
//www.eia.
gov/outlooks/
steo/data.
php?type=
57

tables
i∗t Foreign interest ECB main refi- European Cen- 2000Q1-
rate nancing rate tral Bank 2021Q2
πteu Eurozone Q-o-Q Q-o-Q growth rate International 2000Q2-
inflation rate of the Eurozone Monetary Fund, 2021Q2
consumer price in- International
dex Financial Sta-
tistics database
Continued on next page
 – Continued from previous page

Observable Variable name Indicator Source Period Observations

πtus US Q-o-Q infla- Q-o-Q growth rate International 2000Q2-
tion rate of the US consu- Monetary Fund, 2021Q2
mer price index International
Financial Sta-
tistics database
seur,usd
t Euro-USD ex- Euro-USD ex- International 2000Q1-
change rate change rate Monetary Fund, 2021Q2
International
58

Financial Sta-
tistics database
f ood
πi,t International FAO Food Price Food and 2000Q1-
Food prices Index Agriculture 2021Q2
Organization
of the United
Nations
Continued on next page
 – Continued from previous page

Observable Variable name Indicator Source Period Observations

∆yit Headline real Q-o-Q Real GDP Bank of Central 2000Q1- Real GDP published in BEAC Monetary Policy Re-
economic q-o-q growth of country African States 2021Q2 ports statistical appendix. Chow-Lin temporal disag-
growth of CE- i (BEAC) gregation has been applied on the annual data using
MAC country Quarterly data on broad money (M2) data published
i by the Bank of Central African States
∆yitoil Oil sector Q-o-Q Oil sector Q-o-Q Bank of Central 2000Q1- Q-o-Q growth rate of the real GDP published in
real GDP of coun- real GDP of coun- African States 2021Q2 BEAC Monetary Policy Reports statistical appendix.
try i try i (BEAC) Chow-Lin temporal disaggregation has been applied
59

on the annual data using Quarterly data on broad

money (M2) data published by the Bank of Central
African States

4
πi,t Y-O-Y inflation of Y-o-Y growth rate Bank of Central 2002Q1- Data published by the Bank of Central African States
country i of the Harmoni- African States 2021Q2 based on National Institutions in charge of Statistics
zed Consumer (BEAC) of CEMAC countries.
Price index of
country i

Continued on next page

 – Continued from previous page

Observable Variable name Indicator Source Period Observations

shoil
it Share of oil GDP Share of oil GDP Bank of Central 2000Q1- Indicator computed by the author based on annual
in headline real in headline real African States 2021Q2 data on headline real GDP of country i. The Den-
GDP of country i GDP of country i (BEAC) ton temporal disaggregation method has been used
to obtain the quarterly indicator.

∆gi,t Q-o-Q growth Q-o-Q growth rate Bank of Central 2000Q1- Data published by the Bank of Central African States
rate of nominal of Current public African States 2021Q2 based on National Treasury data. The quarterly data
public expenses of expenses of coun- (BEAC) is obtained by using the Denton temporal disaggre-
60

country i try i gation method on the annual data.

debti,t Debt to GDP ra- External Debt to Bank of Central 2000Q1- Data published by the Bank of Central African States
tio of country i GDP Ratio used African States 2021Q2 based on National Treasury data. The quarterly data
of country i (BEAC) is obtained by using the Denton temporal disaggre-
gation method on the annual data.

∆cri,t Q-o-Q credit Q-o-Q credit to Bank of Central 2006Q1- Author’s calculations based on the monthly Data pu-
growth of country private sector of African States 2021Q2 blished by the Bank of Central African States (Mo-
i country i (BEAC) netary statistics)
Continued on next page
 – Continued from previous page

Observable Variable name Indicator Source Period Observations

RESM IM Pt International Re- International re- Bank of Central 2000Q1- Author’s calculations based on international reserves
serves of CEMAC serves of CEMAC African States 2021Q2 stocks of CEMAC and quarterly data on current im-
in months of im- in months of (BEAC) ports of CEMAC (obtained by temporal disaggrega-
ports exports tion of Annual imports data published by the Bank
of Central African States). The indicator is the ratio
of international reserves at time t and the imports at
time t.
61

IP OLICYt Central Bank Po- Taux d’intérêt Bank of Central 2000Q1-

licy interest rate d’Appels d’Offres African States 2021Q2
(TIAO) (BEAC)

Continued on next page

 – Continued from previous page

Observable Variable name Indicator Source Period Observations

IIN T ERBAN Kt Interbank market Taux d’intérêt Bank of Central 2000Q1- The data on the Interbank rate has been compiled
interest rate moyen pondéré African States 2021Q2 from 2 main sources : (i) the BEAC annual reports
(TIMP) (BEAC) published from 2001 to 2017 and (ii) the monthly
interbank rates statistics of the 2018-2021 period pu-
blished by the Bank of Central African States. The
BEAC annual reports published from 2001 to 2017
present annual average interbank rates (except for
the 2002Q1-2004Q4 period, where monthly data are
62

reported). The Denton temporal disaggregation tech-

nique has been used to obtain quarterly interbank
rates for the 2001Q1-2001Q4 and 2005Q1-2017Q4
periods. For the 2018Q1-2021Q2 period, the quarterly
average of the unsecured interbank rate (Taux Inter-
bancaire Moyen pondéré sur les opérations en blanc
for the french denomination) has been considered.

Continued on next page

 – Continued from previous page

Observable Variable name Indicator Source Period Observations

CBIN JECt Central Bank li- Ratio between the Bank of Central 2000Q1- The net liquidity injections have been computed by
quidity injection net liquidity injec- African States 2021Q2 substracting the liquidity punctions of the Central
to real GDP ratio tions of the Bank (BEAC) from the total liquidity injection of the Central Bank
of Central Afri- on the CEMAC money market. The data has been
can States and the collected from the BEAC website.
real GDP of CE-
MAC
63
 Table 2 – Descriptive statistics of observables data over the estimation period (2006Q1-2019Q4)

Observables Mean Std. Min 25th percentile Median 75th percentile Max
Central Bank liquidity injections to GDP ratio -0.0051 0.05 -0.14 -0.033 0.0012 0.029 0.071
Credit growth (q-o-q) - Central African Republic 0.021 0.075 -0.13 -0.028 0.014 0.055 0.26
Credit growth (q-o-q) - Cameroon 0.023 0.031 -0.039 -0.00011 0.016 0.043 0.12
Credit growth (q-o-q) - Congo 0.048 0.069 -0.094 0.0038 0.045 0.074 0.34
Credit growth (q-o-q) - Gabon 0.019 0.063 -0.1 -0.025 0.004 0.06 0.24
Credit growth (q-o-q) - Equatorial Guinea 0.055 0.17 -0.3 -0.023 0.022 0.11 0.54
Credit growth (q-o-q) - Chad 0.031 0.069 -0.13 -0.0079 0.018 0.072 0.27
Public debt level - Central African Republic 0.35 0.2 0.15 0.22 0.26 0.32 0.86
64

Public debt level - Cameroon 0.14 0.076 0.065 0.077 0.1 0.2 0.29
Public debt level - Congo 0.53 0.27 0.15 0.23 0.56 0.77 0.99
Public debt level - Gabon 0.29 0.097 0.15 0.18 0.28 0.38 0.41
Public debt level - Equatorial Guinea 0.066 0.035 0.0014 0.039 0.078 0.096 0.11
Public debt level - Chad 0.066 0.035 0.0014 0.039 0.078 0.096 0.11
Headline GDP growth - Central African Republic (q-o-q) 0.001 0.039 -0.17 0.0051 0.009 0.013 0.061
Headline GDP growth - Cameroon (q-o-q) 0.011 0.013 -0.026 0.0055 0.01 0.013 0.056
Headline GDP growth - Congo (q-o-q) 0.0054 0.025 -0.063 -0.01 0.0072 0.026 0.062
Headline GDP growth - Gabon (q-o-q) 0.0077 0.012 -0.017 -0.0023 0.01 0.016 0.028
Continued on next page
 – Continued from previous page

Observables Mean Std. Min 25th percentile Median 75th percentile Max
Headline GDP growth - Equatorial Guinea (q-o-q) 0.0017 0.024 -0.057 -0.015 0.0012 0.022 0.048
Headline GDP growth - Chad (q-o-q) 0.0073 0.019 -0.03 -0.0053 0.0045 0.017 0.07
Public expenses (log) - Central African Republic 3.5 0.26 3.1 3.3 3.5 3.7 4.1
Public expenses (log) - Cameroon 6.5 0.39 5.6 6.2 6.5 6.8 7
Public expenses (log) - Congo 6 0.45 5.3 5.7 5.9 6.4 7
Public expenses (log) - Gabon 6 0.28 5.5 5.7 6 6.2 6.4
Public expenses (log) - Equatorial Guinea 6.2 0.47 5.3 5.8 6.4 6.7 6.9
Public expenses (log) - Chad 5.5 0.33 4.8 5.2 5.4 5.8 6
65

Interbank rate - CEMAC 0.039 0.0092 0.02 0.031 0.039 0.046 0.052
Monetary Policy rate - CEMAC 0.038 0.0096 0.025 0.03 0.04 0.043 0.055
Foreign Interest rate 0.013 0.017 0.00073 0.0012 0.0019 0.019 0.053
International food prices variation (percent, q-o-q) 0.0093 0.073 -0.24 -0.023 0.0035 0.03 0.19
Inflation (y-o-y) - Central African Republic 0.048 0.061 -0.038 0.0027 0.035 0.062 0.3
Inflation (y-o-y) - Cameroon 0.023 0.017 -0.0086 0.0099 0.023 0.029 0.068
Inflation (y-o-y) - Congo 0.03 0.024 -0.008 0.013 0.022 0.046 0.11
Inflation (y-o-y) - Gabon 0.023 0.024 -0.017 0.0037 0.024 0.042 0.074
Inflation (y-o-y) - Equatorial Guinea 0.032 0.019 -0.007 0.018 0.029 0.049 0.06
Inflation (y-o-y) - Chad 0.024 0.054 -0.1 -0.019 0.016 0.057 0.19
Continued on next page
 – Continued from previous page

Observables Mean Std. Min 25th percentile Median 75th percentile Max
Foreign Inflation - Eurozone 0.0037 0.0046 -0.0058 0.0007 0.0034 0.006 0.014
Foreign Inflation - USA 0.0048 0.0098 -0.039 0 0.0047 0.012 0.025
Oil prices (100*log) 4.3 0.32 3.5 4.1 4.2 4.7 4.8
International reserves - CEMAC (months of imports) 4.3 1.2 2 3.3 4.8 5.3 6
Nominal Exchange rate (log) 2.8 0 2.8 2.8 2.8 2.8 2.8
Nominal exchange rate EUR/USD (log) -0.24 0.1 -0.45 -0.31 -0.26 -0.13 -0.064
Share of oil GDP in Total GDP - Central African Republic 0 0 0 0 0 0 0
Share of oil GDP in Total GDP - Cameroon 0.063 0.0089 0.05 0.055 0.061 0.07 0.08
66

Share of oil GDP in Total GDP - Congo 0.39 0.062 0.3 0.33 0.41 0.44 0.53
Share of oil GDP in Total GDP - Gabon 0.28 0.056 0.19 0.23 0.25 0.33 0.39
Share of oil GDP in Total GDP - Equatorial Guinea 0.57 0.083 0.43 0.52 0.56 0.64 0.71
Share of oil GDP in Total GDP - Chad 0.14 0.031 0.091 0.12 0.13 0.15 0.22
Foreign Output gap (percent) -0.019 0.018 -0.053 -0.033 -0.019 -0.0011 0.011
Non oil sector Output gap - Central African Republic 0.0014 0.091 -0.25 -0.019 -0.0073 0.028 0.24
Non oil sector Output gap - Cameroon 0.0014 0.012 -0.036 -0.0054 -0.0014 0.0074 0.037
Non oil sector Output gap - Congo 0.0014 0.091 -0.25 -0.019 -0.0073 0.028 0.24
Non oil sector Output gap - Gabon 0.00074 0.041 -0.19 -0.012 0.0038 0.022 0.066
Non oil sector Output gap - Equatorial Guinea 0.0024 0.04 -0.085 -0.027 0.012 0.025 0.078
Continued on next page
 – Continued from previous page

Observables Mean Std. Min 25th percentile Median 75th percentile Max
Non oil sector Output gap - Chad -0.0018 0.027 -0.083 -0.016 0.00044 0.019 0.035

Source : Author’s calculations

67
 Figure 17 – Evolution of common data

Interest rates(%) Central Bank liquidity injections to annualized GDP ratio (%) International reserves (months of imports)
5 5
4 5
0
3 4
Monetary Policy rate -5
2
Interbank rate 3
1 -10
Eurozone Interest rate
2
1

1
Q

Q
06

08

10

12

14

16

18

20

06

08

10

12

14

16

18

20

06

08

10

12

14

16

18

20
20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20
Quarters Quarters Quarters
Foreign prices(q-o-q,%) EUR/USD exchange rate Oil prices (log)
2 -0.1
4.5
0 -0.2

-0.3 4
-2
Eurozone
USA -0.4
3.5
1

1
Q

Q
06

08

10

12

14

16

18

20

06

08

10

12

14

16

18

20

06

08

10

12

14

16

18

20
20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20
Quarters Quarters Quarters
Food prices (q-o-q,%)

10

-10

-20
1

1
Q

Q
06

08

10

12

14

16

18

20
20

20

20

20

20

20

20

20
Quarters

Source : Bank of Central African States, International Monetary Fund, FAO, US Energy Information
Administration, St. Louis Federal Reserve

Figure 18 – Evolution of Cameroon’s observables

Inflation (y-o-y,%) Headline growth rate (q-o-q,%) Share of oil GDP in total GDP (%)
8
6
4 7.5
5

4 7
2
3 6.5
2
0 6
1
5.5
0
-2
1

1
Q

Q
06

08

10

12

14

16

18

20

13

14

15

16

17

18

19

20

06

08

10

12

14

16

18

20
20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

Quarters Quarters Quarters

Public expenses (log) External Public debt to GDP ratio (%) Credit to private sector growth (q-o-q,%)

6.8 10
25
6.6

6.4 20 5
6.2
15
6 0
10
5.8
1

1
Q

Q
06

08

10

12

14

16

18

20

06

08

10

12

14

16

18

20

06

08

10

12

14

16

18

20
20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

Quarters Quarters Quarters

Source : Author’s calculations based on Bank of Central African States data

68
 Figure 19 – Evolution of Central African Republic’s observables

Inflation (y-o-y,%) Headline growth rate (q-o-q,%) Share of oil GDP in total GDP (%)
30 1
5
25
0 0.5
20

15
-5 0
10
-10
5 -0.5

0 -15
-1
1

1
Q

Q
06

08

10

12

14

16

18

20

06

08

10

12

14

16

18

20

06

08

10

12

14

16

18

20
20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20
Quarters Quarters Quarters

Public expenses (log) External Public debt to GDP ratio (%) Credit to private sector growth (q-o-q,%)
4.2 80
20
4 70

60
3.8 10
50
3.6
40 0
3.4 30

3.2 20 -10
1

1
Q

Q
06

08

10

12

14

16

18

20

06

08

10

12

14

16

18

20

06

08

10

12

14

16

18

20
20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20
Quarters Quarters Quarters

Source : Author’s calculations based on Bank of Central African States data

Figure 20 – Evolution of Congo’s observables

Inflation (y-o-y,%) Headline growth rate (q-o-q,%) Share of oil GDP in total GDP (%)
6
10
50
4
8
2 45
6
0
4 40
-2

2 -4 35

0 -6
1

1
Q

Q
06

08

10

12

14

16

18

20

06

08

10

12

14

16

18

20

06

08

10

12

14

16

18

20
20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

Quarters Quarters Quarters

Public expenses (log) External Public debt to GDP ratio (%) Credit to private sector growth (q-o-q,%)
7
30

80
6.5 20

60
10
6
40
0
5.5
20
1

1
Q

Q
06

08

10

12

14

16

18

20

06

08

10

12

14

16

18

20

06

08

10

12

14

16

18

20
20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

Quarters Quarters Quarters

Source : Author’s calculations based on Bank of Central African States data

69
 Figure 21 – Evolution of Gabon’s observables

Inflation (y-o-y,%) Headline growth rate (q-o-q,%) Share of oil GDP in total GDP (%)

6 2
35

4
0
30

2
-2 25
0

-4 20
1

1
Q

Q
06

08

10

12

14

16

18

20

06

08

10

12

14

16

18

20

06

08

10

12

14

16

18

20
20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20
Quarters Quarters Quarters

Public expenses (log) External Public debt to GDP ratio (%) Credit to private sector growth (q-o-q,%)
40
20
6.2 35 15

6 30 10

5
25
5.8
0
20
-5
5.6
15 -10
1

1
Q

Q
06

08

10

12

14

16

18

20

06

08

10

12

14

16

18

20

06

08

10

12

14

16

18

20
20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20
Quarters Quarters Quarters

Source : Author’s calculations based on Bank of Central African States data

Figure 22 – Evolution of Equatorial Guinea’s observables

Inflation (y-o-y,%) Headline growth rate (q-o-q,%) Share of oil GDP in total GDP (%)
4 70
6
2 65
5

4 0
60
3 -2
55
2 -4
50
1 -6

0 -8 45
1

1
Q

Q
06

08

10

12

14

16

18

20

06

08

10

12

14

16

18

20

06

08

10

12

14

16

18

20
20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

Quarters Quarters Quarters

Public expenses (log) External Public debt to GDP ratio (%) Credit to private sector growth (q-o-q,%)
10
40
6.5 8

20
6

6
4 0

2
5.5 -20
1

1
Q

Q
06

08

10

12

14

16

18

20

06

08

10

12

14

16

18

20

06

08

10

12

14

16

18

20
20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

Quarters Quarters Quarters

Source : Author’s calculations based on Bank of Central African States data

70
 Figure 23 – Evolution of Chad’s observables

Inflation (y-o-y,%) Headline growth rate (q-o-q,%) Share of oil GDP in total GDP (%)

6
15 20

10 4 18

5 16
2
14
0
0
12
-5
-2
10
-10
1

1
Q

Q
06

08

10

12

14

16

18

20

06

08

10

12

14

16

18

20

06

08

10

12

14

16

18

20
20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20
Quarters Quarters Quarters

Public expenses (log) External Public debt to GDP ratio (%) Credit to private sector growth (q-o-q,%)
6
26
5.8 20
24
5.6
22 10
5.4
20

5.2 18 0

5 16
-10
14
1

1
Q

Q
06

08

10

12

14

16

18

20

06

08

10

12

14

16

18

20

06

08

10

12

14

16

18

20
20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20
Quarters Quarters Quarters

Source : Author’s calculations based on Bank of Central African States data

71
 Parameter values

Country parameters modes

Table 3 – Country parameters estimated modes

Parameter Labels CEMAC CMR CAR CNG GAB GEQ TCD

a10 Non oil Output gap inertia 0.2 0.087 0.74 0.56 0.31 0.5 0.24
a11 Non oil Output gap expectation 0.085 0.078 0.091 0.11 0.085 0.09 0.095
a12 Financial conditions output gap sensitiveness 0.51 0.4 0.35 0.92 0.57 0.57 0.59
a13 Non oil output gap sensitiveness to exchange rate 0.21 0.22 0.31 0.31 0.19 0.25 0.19
a14 Domestic non oil output gap sensitiveness to foreign output gap 0.56 0.37 0.41 0.31 0.49 0.44 0.36
72

a15 Non oil Output gap sensitiveness to public expenses gap 0.11 0.089 0.11 0.13 0.088 0.08 0.091
a30 Inertia of non oil gdp trend growth 0.75 0.52 0.88 0.92 0.91 0.9 0.89
a31 Non oil GDP reaction to structural public spending growth 0.13 0.077 0.12 0.083 0.22 0.13 0.16
a10,o Inertia of oil production growth 0.6 0.6 0.6 0.6 0.6 0.6 0.6
b10 Weight of lag inflation 0.64 0.66 0.59 0.62 0.6 0.61 0.57
b11 Weight of prices expectations 0.094 0.09 0.1 0.093 0.098 0.092 0.1
b12 Demand price sensitiveness 0.29 0.21 0.063 0.082 0.16 0.2 0.33
b13 Exchange rate price sensitiveness 0.046 0.043 0.063 0.064 0.083 0.068 0.086
b14 Sensitiveness of local inflation to Imported Food prices 0.12 0.12 0.3 0.15 0.17 0.16 0.21
c10 Inertia of public spending gap 0.52 0.52 0.51 0.53 0.52 0.51 0.52
Continued on next page
 – Continued from previous page

Parameter Labels CEMAC CMR CAR CNG GAB GEQ TCD

c11 Weight of output stabilization in fiscal policy rule 0.53 0.35 0.3 0.42 0.52 0.41 0.55
c12 Public spending gap response to public debt 0.088 0.092 0.086 0.094 0.096 0.09 0.092
c30 Inertia of the growth rate of structural public spending 0.31 0.55 0.75 0.7 0.55 0.82 0.64
c31 Structural public spending growth to non oil GDP growth 1.1 0.49 0.37 0.55 0.65 1 0.59
c12 Structural public spending growth to oil GDP growth 0.29 0.39 0.94 0.28 0.23 0.26 0.18
cd10 Public debt growth inertia 0.77 0.79 0.72 0.67 0.74 0.77 0.79
cd21 Sensitiveness of public debt to exchange rate 0.12 0.071 0.099 0.25 0.15 0.19 0.13
d30 Weight of eurozone trade in total trade 0.28 0.36 0.39 0.21 0.31 0.31 0.16
73

d40 Risk premium inertia 0.8 0.8 0.8 0.8 0.8 0.8 0.8
d41 Risk premium sensitiveness to oil price shocks 0.0066 0.0066 0.0066 0.0066 0.0066 0.0066 0.0066
d42 Risk premium sensitiveness to public debt variation 0.2 0.2 0.2 0.2 0.2 0.2 0.2
d43 Risk premium sensitiveness to public debt gap 0.012 0.012 0.012 0.012 0.012 0.012 0.012
debttar Debt target 0.7 0.7 0.7 0.7 0.7 0.7 0.7
e10 Inertia of monetary policy rate 0.9 0.9 0.9 0.9 0.9 0.9 0.9
e11 Inflation in monetary policy rule 0.8 0.8 0.8 0.8 0.8 0.8 0.8
e12 Output gap in monetary policy rule 0.2 0.2 0.2 0.2 0.2 0.2 0.2
e13 International reserves in monetary policy rule 0.0096 0.0096 0.0096 0.0096 0.0096 0.0096 0.0096
e14 Inertia of structural real interest rate 0.8 0.8 0.8 0.8 0.8 0.8 0.8
Continued on next page
 – Continued from previous page

Parameter Labels CEMAC CMR CAR CNG GAB GEQ TCD

e21 Financial openess 0.15 0.15 0.15 0.15 0.15 0.15 0.15
f11 International reserves sensitiveness to economic growth 1.1 1.5 2.1 0.3 4.8 1.1 0.55
f21 International reserves sensitiveness to oil GDP 2.5 3 3 2 3.2 2.4 2.6
f31 International reserves sensitiveness to debt dynamics 3.3 1 0.41 4.3 3.9 1.8 1.6
g01 Inertia of foreign interest rate 0.97 0.97 0.97 0.97 0.97 0.97 0.97
g02 Inertia of foreign output gap 0.63 0.63 0.63 0.63 0.63 0.63 0.63
g03 Inertia of foreign inglation - Eurozone 0.097 0.097 0.097 0.096 0.097 0.097 0.097
g04 Inertia of foreign inglation - USA 0.1 0.1 0.1 0.1 0.1 0.1 0.1
74

g06 Inertia of international food prices variation 0.1 0.1 0.1 0.1 0.1 0.1 0.1
gproil steady state value of oil production growth -0.001 -0.001 0 -0.001 -0.001 -0.001 -0.001
¯
gy Long term value of gdp trend growth 0.0086 0.01 0.0019 0.0019 0.0044 0.014 0.0091
h10 Bank Lending Tightness inertia 0.75 0.72 0.72 0.73 0.73 0.75 0.75
h11 Output expectations effect on credit supply 0.49 0.49 0.49 0.48 0.49 0.48 0.48
h12 Output past dynamics effect on credit supply 0.75 0.76 0.76 0.74 0.77 0.75 0.75
h13 Interbank rate effect on credit supply 0.059 0.059 0.059 0.059 0.059 0.059 0.059
h14 Central Bank liquidity effect on credit suply 0.18 0.18 0.18 0.18 0.18 0.19 0.17
h20 Credit gap inertia 0.56 0.53 0.9 0.83 0.57 0.53 0.49
h21 Bank-lending tightness effect on credit gap 1 1 1 1 1 1 1
Continued on next page
 – Continued from previous page

Parameter Labels CEMAC CMR CAR CNG GAB GEQ TCD

h22 Interest rate effect on credit gap 0.059 0.059 0.059 0.059 0.059 0.06 0.059
h30 Inertia of credit trend growth 0.88 0.79 0.54 0.77 0.62 0.5 0.75
h31 Nominal growth effect on credit trend 0.42 0.39 0.43 0.37 0.45 0.42 0.44
h40 Inertia of Bank Stability Index 0.73 0.72 0.72 0.73 0.71 0.72 0.72
h41 Debt importance in bank stability dynamics 0.097 0.098 0.097 0.097 0.097 0.097 0.097
i10 Inertia of Interbank rate 0.93 0.94 0.99 0.98 0.96 0.96 0.97
i11 Liquidity injection influence on Interbank rate 0.49 0.39 0.17 0.43 0.32 0.38 0.32
i12 Bank stability influence on interbank rate dynamics 0.38 0.44 0.39 0.33 0.4 0.4 0.39
75

i20 Central bank liquidity inertia 0.95 0.95 0.89 0.86 0.88 0.87 0.83
i21 Central bank liquidity reaction to interbank spread 5.6 7 7.5 7.2 7 7.1 7.2
i22 Central bank liquidity reaction to bank instability 0.17 0.17 0.17 0.16 0.17 0.17 0.17
in,∗ Foreign neutral interest rate 0.035 0.035 0.035 0.035 0.035 0.035 0.035
π̄ Inflation target 0.03 0.03 0.03 0.03 0.03 0.03 0.03
π¯∗ world long term foreign inflation (y-o-y) 0.02 0.02 0.02 0.02 0.02 0.02 0.02
¯
poil Oil prices median value 4.1 4.1 4.1 4.1 4.1 4.1 4.1
resmimp Target of international reserves in months of imports 3 3 3 3 3 3 3
rshoil Share of oil revenue on headline fiscal revenues 0.5 0.21 0 0.51 0.66 0.84 0.84
sh share of country GDP in total GDP 1 0.37 0.021 0.14 0.18 0.16 0.16
Continued on next page
 – Continued from previous page

Parameter Labels CEMAC CMR CAR CNG GAB GEQ TCD

σ susd Size of EUR/USD exchange rate shock 0.037 0.037 0.037 0.037 0.037 0.037 0.037
σ ŷ Size of Demand Shock 0.0072 0.012 0.0058 0.0064 0.0069 0.006 0.0074
σ cbinjec Size of Central Bank liquidity injection 0.006 0.012 0.16 0.029 0.024 0.023 0.033
σ cr¯ Size of permament credit supply shock 0.024 0.03 0.074 0.039 0.061 0.12 0.046
σ blt Size of temporary credit supply shock 0.0095 0.0075 0.0072 0.1 0.03 0.031 0.03
σ debt Size of debt level shock 0.0058 0.0043 0.0059 0.0067 0.006 0.006 0.0059
σ ḡ Size of structural fiscal policy shock (level) 0.03 0.006 0.0081 0.007 0.0067 0.0065 0.0063
σ ∆debt Size of public debt growth shock 0.016 0.0072 0.021 0.048 0.022 0.026 0.021
76

σ ĝ Size of short term fiscal policy shock 0.007 0.0055 0.0062 0.0066 0.0062 0.0061 0.0059
σ G(ḡ) Size of structural fiscal policy shock (growth) 0.0081 0.019 0.04 0.049 0.031 0.034 0.029
oil
σ G(pr) size of oil production growth shock 0.007 0.0096 0.019 0.0075 0.019 0.076 0.0079
Continued on next page
 – Continued from previous page

Parameter Labels CEMAC CMR CAR CNG GAB GEQ TCD

σ G(ȳ) Size of growth productivity shock 0.0033 0.0031 0.0036 0.0037 0.0033 0.0034 0.0032
σ IP OLICY size of monetary policy shock 0.003 0.0029 0.0039 0.0033 0.003 0.0031 0.0035
σπ Size of cost push shock 0.0094 0.0071 0.032 0.015 0.012 0.013 0.031
f ood
σπ Size of International food prices shock 0.064 0.064 0.064 0.064 0.064 0.064 0.064
σ poil Size of oil price shock 0.19 0.19 0.19 0.19 0.19 0.19 0.19
σ prem Size of risk premium shock 0.0072 0.0073 0.0073 0.19 0.0073 0.0073 0.0073
proil
σ Size of oil production shock 0.051 0.0086 0.0059 0.16 0.0066 0.0074 0.042
σ r̄ Size of real interest rate trend shock 0.0067 0.0098 0.014 0.0075 0.0072 0.019 0.013
77

σ resmimp Size of international reserves shock 0.57 0.58 0.59 0.61 0.56 0.61 0.6
σ ȳ Size of level productivity shock 0.003 0.0031 0.003 0.0031 0.003 0.003 0.003
oil
σy Size of oil production shock 0.0074 0.0087 0.0059 0.0073 0.0065 0.0097 0.0073
σ z̄ Size of Structural exchange rate shock 0.022 0.013 0.079 0.0052 0.0085 0.0076 0.0074
σ cr¯ Size of credit trend growth shock 0.007 0.011 0.0072 0.024 0.012 0.0072 0.014
σ IIN T ERBAN K Size of interbank risk shock 0.0059 0.0056 0.0058 0.0052 0.0059 0.0057 0.0057
euro
σi Size of foreign interest rate shock 0.0044 0.0044 0.0044 0.0044 0.0044 0.0044 0.0044
euro
σπ Size of foreign inflation - Eurozone 0.0056 0.0056 0.0056 0.0056 0.0056 0.0056 0.0056
us
σπ Size of foreign inflation - USA 0.01 0.01 0.01 0.01 0.01 0.01 0.01
euro
σs Size of nominal exchange rate shock 0.0022 0.0022 0.0022 0.0022 0.0022 0.0022 0.0022
y∗
σ Size of foreign output gap shock 0.017 0.017 0.017 0.017 0.017 0.017 0.017
a
Source : the author, based on CEMAC-GLOBAL estimates
b
The CEMAC column reports the estimated modes of the parameters of the "country-specific" version of the model where CEMAC is considered as a single country
(aggregated data).
 Estimation results

Table 4 – Estimation results - Cameroon

Param name Labels Prior mean Prior std Prior Dist. Estim. Mode Estim. std.
a10 Non oil Output gap inertia 0.081 0.25 beta 0.087 0.078
a11 Non oil Output gap expectation 0.18 0.03 gamma 0.078 0.023
a12 Financial conditions output gap sensitiveness 0.5 0.15 gamma 0.4 0.027
a13 Non oil output gap sensitiveness to exchange rate 0.3 0.1 gamma 0.22 0.1
a14 Domestic non oil output gap sensitiveness to foreign output gap 0.79 0.21 gamma 0.37 0.18
a15 Non oil Output gap sensitiveness to public expenses gap 0.16 0.05 gamma 0.089 0.098
a30 Inertia of non oil gdp trend growth 0.4 0.15 beta 0.52 0.074
78

a31 Non oil GDP reaction to structural public spending growth 0.77 0.12 beta 0.077 0.066
b10 Weight of lag inflation 0.79 0.06 beta 0.66 0.084
b11 Weight of prices expectations 1.1 0.075 gamma 0.09 0.21
b12 Demand price sensitiveness 3 0.36 gamma 0.21 0.28
b13 Exchange rate price sensitiveness 0.34 0.05 gamma 0.043 0.048
b14 Sensitiveness of local inflation to Imported Food prices 0.42 0.15 gamma 0.12 0.27
c10 Inertia of public spending gap 0.78 0.05 beta 0.52 0.025
c11 Weight of output stabilization in fiscal policy rule 2.3 0.21 gamma 0.35 0.29
c12 Public spending gap response to public debt 0.37 0.03 gamma 0.092 0.0037
Continued on next page
 – Continued from previous page

Param name Labels Prior mean Prior std Prior Dist. Estim. Mode Estim. std.
c30 Inertia of the growth rate of structural public spending 0.13 0.1 gamma 0.55 0.14
c31 Structural public spending growth to non oil GDP growth 1.1 0.5 gamma 0.49 0.21
c32 Structural public spending growth to oil GDP growth 1.5 0.25 gamma 0.39 0.22
cd10 Public debt growth inertia 0.63 0.05 beta 0.79 0.097
cd21 Sensitiveness of public debt to exchange rate 3.4 0.25 gamma 0.071 0.53
f11 International reserves sensitiveness to economic growth 66 4 gamma 1.5 6.4
f21 International reserves sensitiveness to oil GDP 2.4 2 gamma 3 1.9
f31 International reserves sensitiveness to debt dynamics 5 2.4 gamma 1 4.9
79

g01 Inertia of foreign interest rate 0.78 0.08 beta 0.97 0.073
g02 Inertia of foreign output gap 0.44 0.08 beta 0.63 0.14
g03 Inertia of foreign inglation - Eurozone 0.26 0.04 beta 0.097 0.021
g06 Inertia of international food prices variation 0.16 0.02 beta 0.1 0.016
h10 Bank Lending Tightness inertia 0.63 0.07 beta 0.72 0.021
h11 Output expectations effect on credit supply 0.43 0.05 gamma 0.49 0.03
h12 Output past dynamics effect on credit supply 0.3 0.16 gamma 0.76 0.11
h13 Interbank rate effect on credit supply 0.03 0.006 gamma 0.059 0.0022
h14 Central Bank liquidity effect on credit suply 0.17 0.06 gamma 0.18 0.02
h20 Credit gap inertia 0.57 0.2 beta 0.53 0.068
Continued on next page
 – Continued from previous page

Param name Labels Prior mean Prior std Prior Dist. Estim. Mode Estim. std.
h22 Interest rate effect on credit gap 0.069 0.006 gamma 0.059 0.0052
h30 Inertia of credit trend growth 0.11 0.1 beta 0.79 0.12
h31 Nominal growth effect on credit trend 1.4 0.1 gamma 0.39 0.072
h40 Inertia of Bank Stability Index 0.37 0.07 beta 0.72 0.039
h41 Debt importance in bank stability dynamics 0.064 0.02 beta 0.098 0.019
i10 Inertia of Interbank rate 0.62 0.045 beta 0.94 0.22
i11 Liquidity injection influence on Interbank rate 1.9 0.24 gamma 0.39 1.3
i12 Bank stability influence on interbank rate dynamics 1.2 0.2 gamma 0.44 0.42
80

i20 Central bank liquidity inertia 0.31 0.08 beta 0.95 0.2
i21 Central bank liquidity reaction to interbank spread 10 0.75 gamma 7 1.3
i22 Central bank liquidity reaction to bank instability 0.061 0.08 gamma 0.17 0.031
σ ŷ Size of Demand Shock 1.1 0.005 inv. Gamma 0.012 0.35
σ ȳ Size of level productivity shock 0.66 0.005 inv. Gamma 0.0031 0.084
σ G(ȳ) Size of growth productivity shock 1.2 0.005 inv. Gamma 0.0031 0.071
oil
σy Size of oil production shock 0.85 0.005 inv. Gamma 0.0087 0.16
oil
σ pr Size of oil production shock 1.7 0.005 inv. Gamma 0.0086 0.38
oil
σ G(pr) size of oil production growth shock 0.021 0.005 inv. Gamma 0.0096 0.079
σπ Size of cost push shock 2.8 0.005 inv. Gamma 0.0071 0.059
Continued on next page
 – Continued from previous page

Param name Labels Prior mean Prior std Prior Dist. Estim. Mode Estim. std.
σ ĝ Size of short term fiscal policy shock 0.086 0.005 inv. Gamma 0.0055 0.29
σ G(ȳ) Size of structural fiscal policy shock (level) 1.6 0.005 inv. Gamma 0.006 0.21
σ G(ḡ) Size of structural fiscal policy shock (growth) 1.9 0.005 inv. Gamma 0.019 0.6
σ ∆debt Size of public debt growth shock 2.2 0.005 inv. Gamma 0.0072 0.41
σ debt Size of debt level shock 0.67 0.005 inv. Gamma 0.0043 0.32
σ z̄ Size of Structural exchange rate shock 0.66 0.005 inv. Gamma 0.013 0.24
σ prem Size of risk premium shock 1.4 0.005 inv. Gamma 0.0073 0.046
euro
σs Size of nominal exchange rate shock 2.6 0.005 inv. Gamma 0.0022 0.37
81

σ IP OLICY size of monetary policy shock 0.17 0.005 inv. Gamma 0.0029 0.08
σ r̄ Size of real interest rate trend shock 0.06 0.005 inv. Gamma 0.0098 0.22
σ resmimp Size of international reserves shock 1.3 0.005 inv. Gamma 0.58 0.19
ieuro
σ Size of foreign interest rate shock 0.24 0.003 inv. Gamma 0.0044 0.013
∗
σy Size of foreign output gap shock 0.24 0.003 inv. Gamma 0.017 0.0095
π euro
σ Size of foreign inflation - Eurozone 0.15 0.003 inv. Gamma 0.0056 0.014
σ poil Size of oil price shock 2.8 0.005 inv. Gamma 0.19 0.33
usd
σs Size of EUR/USD exchange rate shock 1.5 0.005 inv. Gamma 0.037 0.13
f ood
σπ Size of International food prices shock 2.6 0.005 inv. Gamma 0.064 0.19
σ blt Size of temporary credit supply shock 2.6 0.005 inv. Gamma 0.0075 0.024
Continued on next page
 – Continued from previous page

Param name Labels Prior mean Prior std Prior Dist. Estim. Mode Estim. std.
σ cr¯ Size of permament credit supply shock 0.018 0.005 inv. Gamma 0.03 0.3
σ cr¯ Size of credit trend growth shock 2.6 0.005 inv. Gamma 0.011 0.055
σ IIN T ERBAN K Size of interbank risk shock 2.6 0.005 inv. Gamma 0.0056 0.6
σ cbinjec Size of Central Bank liquidity injection 2.6 0.005 inv. Gamma 0.012 0.28

Source : Author’s calculations

82
 Table 5 – Estimation results - Central African Republic

Param name Labels Prior mean Prior std Prior Dist. Estim. Mode Estim. std.
a10 Non oil Output gap inertia 0.34 0.25 beta 0.74 0.12
a11 Non oil Output gap expectation 0.25 0.03 gamma 0.091 0.042
a12 Financial conditions output gap sensitiveness 0.56 0.15 gamma 0.35 0.23
a13 Non oil output gap sensitiveness to exchange rate 0.95 0.1 gamma 0.31 0.069
a14 Domestic non oil output gap sensitiveness to foreign output gap 1.1 0.21 gamma 0.41 0.31
a15 Non oil Output gap sensitiveness to public expenses gap 0.22 0.05 gamma 0.11 0.15
a30 Inertia of non oil gdp trend growth 0.76 0.15 beta 0.88 0.14
a31 Non oil GDP reaction to structural public spending growth 0.77 0.12 beta 0.12 0.074
83

b10 Weight of lag inflation 0.35 0.06 beta 0.59 0.048

b11 Weight of prices expectations 0.81 0.075 gamma 0.1 0.095
b12 Demand price sensitiveness 1.9 0.36 gamma 0.063 0.65
b13 Exchange rate price sensitiveness 0.45 0.05 gamma 0.063 0.066
b14 Sensitiveness of local inflation to Imported Food prices 1.3 0.15 gamma 0.3 0.17
c10 Inertia of public spending gap 0.38 0.05 beta 0.51 0.029
c11 Weight of output stabilization in fiscal policy rule 2.7 0.21 gamma 0.3 0.3
c12 Public spending gap response to public debt 0.046 0.03 gamma 0.086 0.018
c30 Inertia of the growth rate of structural public spending 0.51 0.1 gamma 0.75 0.26
Continued on next page
 – Continued from previous page

Param name Labels Prior mean Prior std Prior Dist. Estim. Mode Estim. std.
c31 Structural public spending growth to non oil GDP growth 0.48 0.5 gamma 0.37 0.18
c32 Structural public spending growth to oil GDP growth 1.1 0.25 gamma 0.94 0.3
cd10 Public debt growth inertia 0.65 0.05 beta 0.72 0.1
cd21 Sensitiveness of public debt to exchange rate 1.6 0.25 gamma 0.099 0.32
f11 International reserves sensitiveness to economic growth 33 4 gamma 2.1 4.2
f21 International reserves sensitiveness to oil GDP 21 2 gamma 3 3
f31 International reserves sensitiveness to debt dynamics 44 2.4 gamma 0.41 1.8
g01 Inertia of foreign interest rate 0.89 0.08 beta 0.97 0.14
84

g02 Inertia of foreign output gap 0.36 0.08 beta 0.63 0.057
g03 Inertia of foreign inglation - Eurozone 0.36 0.04 beta 0.097 0.07
g06 Inertia of international food prices variation 0.19 0.02 beta 0.1 0.01
h10 Bank Lending Tightness inertia 0.87 0.07 beta 0.72 0.05
h11 Output expectations effect on credit supply 0.75 0.05 gamma 0.49 0.049
h12 Output past dynamics effect on credit supply 2 0.16 gamma 0.76 0.26
h13 Interbank rate effect on credit supply 0.098 0.006 gamma 0.059 0.0031
h14 Central Bank liquidity effect on credit suply 0.16 0.06 gamma 0.18 0.03
h20 Credit gap inertia 0.49 0.2 beta 0.9 0.13
h22 Interest rate effect on credit gap 0.085 0.006 gamma 0.059 0.0039
Continued on next page
 – Continued from previous page

Param name Labels Prior mean Prior std Prior Dist. Estim. Mode Estim. std.
h30 Inertia of credit trend growth 0.68 0.1 beta 0.54 0.12
h31 Nominal growth effect on credit trend 0.77 0.1 gamma 0.43 0.16
h40 Inertia of Bank Stability Index 0.4 0.07 beta 0.72 0.052
h41 Debt importance in bank stability dynamics 0.15 0.02 beta 0.097 0.029
i10 Inertia of Interbank rate 0.76 0.045 beta 0.99 0.12
i11 Liquidity injection influence on Interbank rate 0.71 0.24 gamma 0.17 0.9
i12 Bank stability influence on interbank rate dynamics 1.3 0.2 gamma 0.39 0.32
i20 Central bank liquidity inertia 0.63 0.08 beta 0.89 0.11
85

i21 Central bank liquidity reaction to interbank spread 4.2 0.75 gamma 7.5 1.4
i22 Central bank liquidity reaction to bank instability 0.77 0.08 gamma 0.17 0.063
σ ŷ Size of Demand Shock 2.3 0.005 inv. Gamma 0.0058 0.21
σ ȳ Size of level productivity shock 1.5 0.005 inv. Gamma 0.003 0.5
σ G(ȳ) Size of growth productivity shock 0.55 0.005 inv. Gamma 0.0036 0.02
y oil
σ Size of oil production shock 1.8 0.005 inv. Gamma 0.0059 0.34
oil
σ pr Size of oil production shock 1.2 0.005 inv. Gamma 0.0059 0.43
oil
σ G(pr) size of oil production growth shock 0.65 0.005 inv. Gamma 0.019 0.43
σπ Size of cost push shock 0.42 0.005 inv. Gamma 0.032 0.23
σ ĝ Size of short term fiscal policy shock 2.1 0.005 inv. Gamma 0.0062 0.24
Continued on next page
 – Continued from previous page

Param name Labels Prior mean Prior std Prior Dist. Estim. Mode Estim. std.
σ G(ȳ) Size of structural fiscal policy shock (level) 0.23 0.005 inv. Gamma 0.0081 0.41
σ G(ḡ) Size of structural fiscal policy shock (growth) 1.2 0.005 inv. Gamma 0.04 0.095
σ ∆debt Size of public debt growth shock 0.59 0.005 inv. Gamma 0.021 0.13
σ debt Size of debt level shock 1.2 0.005 inv. Gamma 0.0059 0.64
σ z̄ Size of Structural exchange rate shock 0.56 0.005 inv. Gamma 0.079 0.29
σ prem Size of risk premium shock 2.2 0.005 inv. Gamma 0.0073 0.26
seuro
σ Size of nominal exchange rate shock 0.76 0.005 inv. Gamma 0.0022 0.29
σ IP OLICY size of monetary policy shock 1.2 0.005 inv. Gamma 0.0039 0.15
86

σ r̄ Size of real interest rate trend shock 0.93 0.005 inv. Gamma 0.014 0.3
σ resmimp Size of international reserves shock 0.77 0.005 inv. Gamma 0.59 0.35
euro
σi Size of foreign interest rate shock 0.14 0.003 inv. Gamma 0.0044 0.022
stdystar Size of foreign output gap shock 0.25 0.003 inv. Gamma 0.017 0.014
euro
σπ Size of foreign inflation - Eurozone 0.19 0.003 inv. Gamma 0.0056 0.014
σ poil Size of oil price shock 2.3 0.005 inv. Gamma 0.19 0.32
usd
σs Size of EUR/USD exchange rate shock 1.1 0.005 inv. Gamma 0.037 0.28
f ood
σπ Size of International food prices shock 1.7 0.005 inv. Gamma 0.064 0.23
σ blt Size of temporary credit supply shock 2 0.005 inv. Gamma 0.0072 0.07
σ cr¯ Size of permament credit supply shock 2.7 0.005 inv. Gamma 0.074 0.25
Continued on next page
 – Continued from previous page

Param name Labels Prior mean Prior std Prior Dist. Estim. Mode Estim. std.
σ cr¯ Size of credit trend growth shock 1 0.005 inv. Gamma 0.0072 0.15
σ IIN T ERBAN K Size of interbank risk shock 2.3 0.005 inv. Gamma 0.0058 0.38
σ cbinjec Size of Central Bank liquidity injection 0.66 0.005 inv. Gamma 0.16 0.41

Source : Author’s calculations

87
 Table 6 – Estimation results - Congo

Param name Labels Prior mean Prior std Prior Dist. Estim. Mode Estim. std.
a10 Non oil Output gap inertia 0.87 0.25 beta 0.56 0.17
a11 Non oil Output gap expectation 0.32 0.03 gamma 0.11 0.097
a12 Financial conditions output gap sensitiveness 1.6 0.15 gamma 0.92 0.27
a13 Non oil output gap sensitiveness to exchange rate 1.2 0.1 gamma 0.31 0.17
a14 Domestic non oil output gap sensitiveness to foreign output gap 0.74 0.21 gamma 0.31 0.15
a15 Non oil Output gap sensitiveness to public expenses gap 0.27 0.05 gamma 0.13 0.13
a30 Inertia of non oil gdp trend growth 0.96 0.15 beta 0.92 0.17
a31 Non oil GDP reaction to structural public spending growth 0.14 0.12 beta 0.083 0.068
88

b10 Weight of lag inflation 0.3 0.06 beta 0.62 0.049

b11 Weight of prices expectations 1.1 0.075 gamma 0.093 0.003
b12 Demand price sensitiveness 0.73 0.36 gamma 0.082 1.2
b13 Exchange rate price sensitiveness 0.6 0.05 gamma 0.064 0.056
b14 Sensitiveness of local inflation to Imported Food prices 1.6 0.15 gamma 0.15 0.12
c10 Inertia of public spending gap 0.28 0.05 beta 0.53 0.054
c11 Weight of output stabilization in fiscal policy rule 3 0.21 gamma 0.42 0.87
c12 Public spending gap response to public debt 0.094 0.03 gamma 0.094 0.004
c30 Inertia of the growth rate of structural public spending 0.41 0.1 gamma 0.7 0.21
Continued on next page
 – Continued from previous page

Param name Labels Prior mean Prior std Prior Dist. Estim. Mode Estim. std.
c31 Structural public spending growth to non oil GDP growth 3.9 0.5 gamma 0.55 0.097
c32 Structural public spending growth to oil GDP growth 1.1 0.25 gamma 0.28 0.58
cd10 Public debt growth inertia 0.63 0.05 beta 0.67 0.063
cd21 Sensitiveness of public debt to exchange rate 2.3 0.25 gamma 0.25 0.054
f11 International reserves sensitiveness to economic growth 58 4 gamma 0.3 0.17
f21 International reserves sensitiveness to oil GDP 4.6 2 gamma 2 3.6
f31 International reserves sensitiveness to debt dynamics 15 2.4 gamma 4.3 10
g01 Inertia of foreign interest rate 0.94 0.08 beta 0.97 0.091
89

g02 Inertia of foreign output gap 0.72 0.08 beta 0.63 0.04
g03 Inertia of foreign inglation - Eurozone 0.3 0.04 beta 0.096 0.082
g06 Inertia of international food prices variation 0.14 0.02 beta 0.1 0.02
h10 Bank Lending Tightness inertia 0.84 0.07 beta 0.73 0.026
h11 Output expectations effect on credit supply 0.77 0.05 gamma 0.48 0.036
h12 Output past dynamics effect on credit supply 0.45 0.16 gamma 0.74 0.2
h13 Interbank rate effect on credit supply 0.076 0.006 gamma 0.059 0.007
h14 Central Bank liquidity effect on credit suply 0.68 0.06 gamma 0.18 0.12
h20 Credit gap inertia 0.75 0.2 beta 0.83 0.049
h22 Interest rate effect on credit gap 0.03 0.006 gamma 0.059 0.0015
Continued on next page
 – Continued from previous page

Param name Labels Prior mean Prior std Prior Dist. Estim. Mode Estim. std.
h30 Inertia of credit trend growth 0.22 0.1 beta 0.77 0.13
h31 Nominal growth effect on credit trend 0.61 0.1 gamma 0.37 0.31
h40 Inertia of Bank Stability Index 0.56 0.07 beta 0.73 0.076
h41 Debt importance in bank stability dynamics 0.25 0.02 beta 0.097 0.017
i10 Inertia of Interbank rate 0.96 0.045 beta 0.98 0.14
i11 Liquidity injection influence on Interbank rate 2.2 0.24 gamma 0.43 0.91
i12 Bank stability influence on interbank rate dynamics 2.2 0.2 gamma 0.33 0.21
i20 Central bank liquidity inertia 0.91 0.08 beta 0.86 0.16
90

i21 Central bank liquidity reaction to interbank spread 7.2 0.75 gamma 7.2 1.2
i22 Central bank liquidity reaction to bank instability 0.25 0.08 gamma 0.16 0.073
σ ŷ Size of Demand Shock 1.5 0.005 inv. Gamma 0.0064 0.018
σ ȳ Size of level productivity shock 2.6 0.005 inv. Gamma 0.0031 0.52
σ G(ȳ) Size of growth productivity shock 0.6 0.005 inv. Gamma 0.0037 0.24
y oil
σ Size of oil production shock 2.8 0.005 inv. Gamma 0.0073 0.081
oil
σ pr Size of oil production shock 1.2 0.005 inv. Gamma 0.16 0.64
oil
σ G(pr) size of oil production growth shock 0.76 0.005 inv. Gamma 0.0075 0.43
σπ Size of cost push shock 2.1 0.005 inv. Gamma 0.015 0.18
σ ĝ Size of short term fiscal policy shock 1.5 0.005 inv. Gamma 0.0066 0.27
Continued on next page
 – Continued from previous page

Param name Labels Prior mean Prior std Prior Dist. Estim. Mode Estim. std.
σ G(ȳ) Size of structural fiscal policy shock (level) 1.1 0.005 inv. Gamma 0.007 0.03
σ G(ḡ) Size of structural fiscal policy shock (growth) 1.1 0.005 inv. Gamma 0.049 0.038
σ ∆debt Size of public debt growth shock 0.62 0.005 inv. Gamma 0.048 0.39
σ debt Size of debt level shock 2.4 0.005 inv. Gamma 0.0067 0.66
σ z̄ Size of Structural exchange rate shock 0.21 0.005 inv. Gamma 0.0052 0.21
σ prem Size of risk premium shock 1 0.005 inv. Gamma 0.19 0.79
seuro
σ Size of nominal exchange rate shock 0.31 0.005 inv. Gamma 0.0022 0.074
σ IP OLICY size of monetary policy shock 2.6 0.005 inv. Gamma 0.0033 0.035
91

σ r̄ Size of real interest rate trend shock 0.12 0.005 inv. Gamma 0.0075 0.32
σ resmimp Size of international reserves shock 0.65 0.005 inv. Gamma 0.61 0.59
euro
σi Size of foreign interest rate shock 0.091 0.003 inv. Gamma 0.0044 0.034
stdystar Size of foreign output gap shock 0.23 0.003 inv. Gamma 0.017 0.0075
euro
σπ Size of foreign inflation - Eurozone 0.12 0.003 inv. Gamma 0.0056 0.008
σ poil Size of oil price shock 0.6 0.005 inv. Gamma 0.19 0.42
usd
σs Size of EUR/USD exchange rate shock 0.9 0.005 inv. Gamma 0.037 0.27
f ood
σπ Size of International food prices shock 2.6 0.005 inv. Gamma 0.064 0.021
σ blt Size of temporary credit supply shock 0.022 0.005 inv. Gamma 0.1 0.47
σ cr¯ Size of permament credit supply shock 0.5 0.005 inv. Gamma 0.039 0.25
Continued on next page
 – Continued from previous page

Param name Labels Prior mean Prior std Prior Dist. Estim. Mode Estim. std.
σ cr¯ Size of credit trend growth shock 2.2 0.005 inv. Gamma 0.024 0.42
σ IIN T ERBAN K Size of interbank risk shock 0.5 0.005 inv. Gamma 0.0052 0.067
σ cbinjec Size of Central Bank liquidity injection 1.4 0.005 inv. Gamma 0.029 0.67

Source : the author

92
 Table 7 – Estimation results - Gabon

Param name Labels Prior mean Prior std Prior Dist. Estim. Mode Estim. std.
a10 Non oil Output gap inertia 0.86 0.25 beta 0.31 0.099
a11 Non oil Output gap expectation 0.23 0.03 gamma 0.085 0.014
a12 Financial conditions output gap sensitiveness 1.5 0.15 gamma 0.57 0.14
a13 Non oil output gap sensitiveness to exchange rate 0.097 0.1 gamma 0.19 0.18
a14 Domestic non oil output gap sensitiveness to foreign output gap 1.1 0.21 gamma 0.49 0.48
a15 Non oil Output gap sensitiveness to public expenses gap 0.7 0.05 gamma 0.088 0.16
a30 Inertia of non oil gdp trend growth 0.72 0.15 beta 0.91 0.12
a31 Non oil GDP reaction to structural public spending growth 0.16 0.12 beta 0.22 0.063
93

b10 Weight of lag inflation 0.52 0.06 beta 0.6 0.044

b11 Weight of prices expectations 1 0.075 gamma 0.098 0.049
b12 Demand price sensitiveness 1.7 0.36 gamma 0.16 0.43
b13 Exchange rate price sensitiveness 0.17 0.05 gamma 0.083 0.077
b14 Sensitiveness of local inflation to Imported Food prices 1.5 0.15 gamma 0.17 0.14
c10 Inertia of public spending gap 0.54 0.05 beta 0.52 0.021
c11 Weight of output stabilization in fiscal policy rule 2.7 0.21 gamma 0.52 0.41
c12 Public spending gap response to public debt 0.14 0.03 gamma 0.096 0.053
c30 Inertia of the growth rate of structural public spending 0.13 0.1 gamma 0.55 0.18
Continued on next page
 – Continued from previous page

Param name Labels Prior mean Prior std Prior Dist. Estim. Mode Estim. std.
c31 Structural public spending growth to non oil GDP growth 2.1 0.5 gamma 0.65 0.25
c32 Structural public spending growth to oil GDP growth 0.9 0.25 gamma 0.23 0.5
cd10 Public debt growth inertia 0.73 0.05 beta 0.74 0.07
cd21 Sensitiveness of public debt to exchange rate 3.7 0.25 gamma 0.15 0.19
f11 International reserves sensitiveness to economic growth 32 4 gamma 4.8 8.6
f21 International reserves sensitiveness to oil GDP 16 2 gamma 3.2 3.2
f31 International reserves sensitiveness to debt dynamics 2.4 2.4 gamma 3.9 9.8
g01 Inertia of foreign interest rate 0.94 0.08 beta 0.97 0.086
94

g02 Inertia of foreign output gap 0.59 0.08 beta 0.63 0.15
g03 Inertia of foreign inglation - Eurozone 0.16 0.04 beta 0.097 0.1
g06 Inertia of international food prices variation 0.12 0.02 beta 0.1 0.035
h10 Bank Lending Tightness inertia 0.36 0.07 beta 0.73 0.035
h11 Output expectations effect on credit supply 0.53 0.05 gamma 0.49 0.029
h12 Output past dynamics effect on credit supply 0.18 0.16 gamma 0.77 0.33
h13 Interbank rate effect on credit supply 0.047 0.006 gamma 0.059 0.00073
h14 Central Bank liquidity effect on credit suply 0.58 0.06 gamma 0.18 0.094
h20 Credit gap inertia 0.3 0.2 beta 0.57 0.21
h22 Interest rate effect on credit gap 0.034 0.006 gamma 0.059 0.01
Continued on next page
 – Continued from previous page

Param name Labels Prior mean Prior std Prior Dist. Estim. Mode Estim. std.
h30 Inertia of credit trend growth 0.5 0.1 beta 0.62 0.083
h31 Nominal growth effect on credit trend 0.2 0.1 gamma 0.45 0.16
h40 Inertia of Bank Stability Index 0.58 0.07 beta 0.71 0.05
h41 Debt importance in bank stability dynamics 0.068 0.02 beta 0.097 0.011
i10 Inertia of Interbank rate 0.66 0.045 beta 0.96 0.19
i11 Liquidity injection influence on Interbank rate 0.66 0.24 gamma 0.32 1.5
i12 Bank stability influence on interbank rate dynamics 1.4 0.2 gamma 0.4 0.51
i20 Central bank liquidity inertia 0.99 0.08 beta 0.88 0.18
95

i21 Central bank liquidity reaction to interbank spread 9 0.75 gamma 7 1.4
i22 Central bank liquidity reaction to bank instability 0.2 0.08 gamma 0.17 0.09
σ ŷ Size of Demand Shock 2.1 0.005 inv. Gamma 0.0069 0.36
σ ȳ Size of level productivity shock 2.5 0.005 inv. Gamma 0.003 0.019
σ G(ȳ) Size of growth productivity shock 1.5 0.005 inv. Gamma 0.0033 0.28
y oil
σ Size of oil production shock 1.3 0.005 inv. Gamma 0.0065 0.32
oil
σ pr Size of oil production shock 0.9 0.005 inv. Gamma 0.0066 0.35
oil
σ G(pr) size of oil production growth shock 0.02 0.005 inv. Gamma 0.019 0.81
σπ Size of cost push shock 0.19 0.005 inv. Gamma 0.012 0.5
σ ĝ Size of short term fiscal policy shock 0.17 0.005 inv. Gamma 0.0062 0.44
Continued on next page
 – Continued from previous page

Param name Labels Prior mean Prior std Prior Dist. Estim. Mode Estim. std.
σ G(ȳ) Size of structural fiscal policy shock (level) 2.1 0.005 inv. Gamma 0.0067 0.35
σ G(ḡ) Size of structural fiscal policy shock (growth) 1.5 0.005 inv. Gamma 0.031 0.34
σ ∆debt Size of public debt growth shock 0.86 0.005 inv. Gamma 0.022 0.33
σ debt Size of debt level shock 2 0.005 inv. Gamma 0.006 0.55
σ z̄ Size of Structural exchange rate shock 1.5 0.005 inv. Gamma 0.0085 0.15
σ prem Size of risk premium shock 0.91 0.005 inv. Gamma 0.0073 0.24
seuro
σ Size of nominal exchange rate shock 1.9 0.005 inv. Gamma 0.0022 0.25
σ IP OLICY size of monetary policy shock 2.6 0.005 inv. Gamma 0.003 0.14
96

σ r̄ Size of real interest rate trend shock 2.6 0.005 inv. Gamma 0.0072 0.54
σ resmimp Size of international reserves shock 0.96 0.005 inv. Gamma 0.56 0.29
euro
σi Size of foreign interest rate shock 0.045 0.003 inv. Gamma 0.0044 0.016
stdystar Size of foreign output gap shock 0.22 0.003 inv. Gamma 0.017 0.053
euro
σπ Size of foreign inflation - Eurozone 0.098 0.003 inv. Gamma 0.0056 0.0015
σ poil Size of oil price shock 0.27 0.005 inv. Gamma 0.19 0.11
usd
σs Size of EUR/USD exchange rate shock 0.11 0.005 inv. Gamma 0.037 0.46
f ood
σπ Size of International food prices shock 1.4 0.005 inv. Gamma 0.064 0.46
σ blt Size of temporary credit supply shock 0.15 0.005 inv. Gamma 0.03 0.26
σ cr¯ Size of permament credit supply shock 2.1 0.005 inv. Gamma 0.061 0.4
Continued on next page
 – Continued from previous page

Param name Labels Prior mean Prior std Prior Dist. Estim. Mode Estim. std.
σ cr¯ Size of credit trend growth shock 0.61 0.005 inv. Gamma 0.012 0.47
σ IIN T ERBAN K Size of interbank risk shock 0.48 0.005 inv. Gamma 0.0059 0.24
σ cbinjec Size of Central Bank liquidity injection 0.55 0.005 inv. Gamma 0.024 0.034

Source : the author

97
 Table 8 – Estimation results - Equatorial Guinea

Param name Labels Prior mean Prior std Prior Dist. Estim. Mode Estim. std.
a10 Non oil Output gap inertia 0.45 0.25 beta 0.5 0.082
a11 Non oil Output gap expectation 0.16 0.03 gamma 0.09 0.064
a12 Financial conditions output gap sensitiveness 1.1 0.15 gamma 0.57 0.12
a13 Non oil output gap sensitiveness to exchange rate 0.38 0.1 gamma 0.25 0.11
a14 Domestic non oil output gap sensitiveness to foreign output gap 0.96 0.21 gamma 0.44 0.33
a15 Non oil Output gap sensitiveness to public expenses gap 0.077 0.05 gamma 0.08 0.17
a30 Inertia of non oil gdp trend growth 0.84 0.15 beta 0.9 0.18
a31 Non oil GDP reaction to structural public spending growth 0.94 0.12 beta 0.13 0.07
98

b10 Weight of lag inflation 0.83 0.06 beta 0.61 0.068

b11 Weight of prices expectations 0.93 0.075 gamma 0.092 0.2
b12 Demand price sensitiveness 0.0021 0.36 gamma 0.2 1.3
b13 Exchange rate price sensitiveness 0.8 0.05 gamma 0.068 0.033
b14 Sensitiveness of local inflation to Imported Food prices 1.8 0.15 gamma 0.16 0.32
c10 Inertia of public spending gap 0.64 0.05 beta 0.51 0.064
c11 Weight of output stabilization in fiscal policy rule 0.8 0.21 gamma 0.41 0.71
c12 Public spending gap response to public debt 0.029 0.03 gamma 0.09 0.069
c30 Inertia of the growth rate of structural public spending 0.43 0.1 gamma 0.82 0.3
Continued on next page
 – Continued from previous page

Param name Labels Prior mean Prior std Prior Dist. Estim. Mode Estim. std.
c31 Structural public spending growth to non oil GDP growth 2.9 0.5 gamma 1 0.4
c32 Structural public spending growth to oil GDP growth 0.23 0.25 gamma 0.26 0.12
cd10 Public debt growth inertia 0.68 0.05 beta 0.77 0.14
cd21 Sensitiveness of public debt to exchange rate 0.41 0.25 gamma 0.19 0.61
f11 International reserves sensitiveness to economic growth 27 4 gamma 1.1 1.2
f21 International reserves sensitiveness to oil GDP 24 2 gamma 2.4 1.7
f31 International reserves sensitiveness to debt dynamics 40 2.4 gamma 1.8 3.9
g01 Inertia of foreign interest rate 0.34 0.08 beta 0.97 0.014
99

g02 Inertia of foreign output gap 0.31 0.08 beta 0.63 0.18
g03 Inertia of foreign inglation - Eurozone 0.19 0.04 beta 0.097 0.0047
g06 Inertia of international food prices variation 0.096 0.02 beta 0.1 0.011
h10 Bank Lending Tightness inertia 0.7 0.07 beta 0.75 0.041
h11 Output expectations effect on credit supply 0.56 0.05 gamma 0.48 0.034
h12 Output past dynamics effect on credit supply 1.8 0.16 gamma 0.75 0.058
h13 Interbank rate effect on credit supply 0.03 0.006 gamma 0.059 0.0062
h14 Central Bank liquidity effect on credit suply 0.83 0.06 gamma 0.19 0.11
h20 Credit gap inertia 0.26 0.2 beta 0.53 0.072
h22 Interest rate effect on credit gap 0.091 0.006 gamma 0.06 0.0039
Continued on next page
 – Continued from previous page

Param name Labels Prior mean Prior std Prior Dist. Estim. Mode Estim. std.
h30 Inertia of credit trend growth 0.29 0.1 beta 0.5 0.089
h31 Nominal growth effect on credit trend 0.79 0.1 gamma 0.42 0.074
h40 Inertia of Bank Stability Index 0.49 0.07 beta 0.72 0.093
h41 Debt importance in bank stability dynamics 0.12 0.02 beta 0.097 0.011
i10 Inertia of Interbank rate 0.72 0.045 beta 0.96 0.21
i11 Liquidity injection influence on Interbank rate 0.85 0.24 gamma 0.38 1.2
i12 Bank stability influence on interbank rate dynamics 2.4 0.2 gamma 0.4 0.1
i20 Central bank liquidity inertia 0.27 0.08 beta 0.87 0.2
100

i21 Central bank liquidity reaction to interbank spread 11 0.75 gamma 7.1 1.5
i22 Central bank liquidity reaction to bank instability 0.22 0.08 gamma 0.17 0.22
σ ŷ Size of Demand Shock 2.7 0.005 inv. Gamma 0.006 0.29
σ ȳ Size of level productivity shock 2.2 0.005 inv. Gamma 0.003 0.45
σ G(ȳ) Size of growth productivity shock 1.2 0.005 inv. Gamma 0.0034 0.4
y oil
σ Size of oil production shock 1.9 0.005 inv. Gamma 0.0097 0.049
oil
σ pr Size of oil production shock 2.4 0.005 inv. Gamma 0.0074 0.4
oil
σ G(pr) size of oil production growth shock 0.77 0.005 inv. Gamma 0.076 0.072
σπ Size of cost push shock 1.4 0.005 inv. Gamma 0.013 0.23
σ ĝ Size of short term fiscal policy shock 1.2 0.005 inv. Gamma 0.0061 0.052
Continued on next page
 – Continued from previous page

Param name Labels Prior mean Prior std Prior Dist. Estim. Mode Estim. std.
σ G(ȳ) Size of structural fiscal policy shock (level) 1.2 0.005 inv. Gamma 0.0065 0.24
σ G(ḡ) Size of structural fiscal policy shock (growth) 0.016 0.005 inv. Gamma 0.034 0.087
σ ∆debt Size of public debt growth shock 0.26 0.005 inv. Gamma 0.026 0.034
σ debt Size of debt level shock 1.2 0.005 inv. Gamma 0.006 0.0093
σ z̄ Size of Structural exchange rate shock 1.6 0.005 inv. Gamma 0.0076 0.2
σ prem Size of risk premium shock 2.6 0.005 inv. Gamma 0.0073 0.19
seuro
σ Size of nominal exchange rate shock 2 0.005 inv. Gamma 0.0022 0.18
σ IP OLICY size of monetary policy shock 0.21 0.005 inv. Gamma 0.0031 0.041
101

σ r̄ Size of real interest rate trend shock 2.2 0.005 inv. Gamma 0.019 0.37
σ resmimp Size of international reserves shock 0.97 0.005 inv. Gamma 0.61 0.32
euro
σi Size of foreign interest rate shock 0.092 0.003 inv. Gamma 0.0044 0.011
stdystar Size of foreign output gap shock 0.21 0.003 inv. Gamma 0.017 0.048
euro
σπ Size of foreign inflation - Eurozone 0.056 0.003 inv. Gamma 0.0056 0.021
σ poil Size of oil price shock 2 0.005 inv. Gamma 0.19 0.2
usd
σs Size of EUR/USD exchange rate shock 2.7 0.005 inv. Gamma 0.037 0.084
f ood
σπ Size of International food prices shock 0.36 0.005 inv. Gamma 0.064 0.1
σ blt Size of temporary credit supply shock 1.6 0.005 inv. Gamma 0.031 0.097
σ cr¯ Size of permament credit supply shock 0.93 0.005 inv. Gamma 0.12 0.54
Continued on next page
 – Continued from previous page

Param name Labels Prior mean Prior std Prior Dist. Estim. Mode Estim. std.
σ cr¯ Size of credit trend growth shock 1.6 0.005 inv. Gamma 0.0072 0.2
σ IIN T ERBAN K Size of interbank risk shock 0.093 0.005 inv. Gamma 0.0057 0.36
σ cbinjec Size of Central Bank liquidity injection 1.2 0.005 inv. Gamma 0.023 0.17

Source : the author

102
 Table 9 – Estimation results - Chad

Param name Labels Prior mean Prior std Prior Dist. Estim. Mode Estim. std.
a10 Non oil Output gap inertia 0.043 0.25 beta 0.24 0.094
a11 Non oil Output gap expectation 0.38 0.03 gamma 0.095 0.084
a12 Financial conditions output gap sensitiveness 0.46 0.15 gamma 0.59 0.12
a13 Non oil output gap sensitiveness to exchange rate 0.32 0.1 gamma 0.19 0.056
a14 Domestic non oil output gap sensitiveness to foreign output gap 1.4 0.21 gamma 0.36 0.29
a15 Non oil Output gap sensitiveness to public expenses gap 0.74 0.05 gamma 0.091 0.054
a30 Inertia of non oil gdp trend growth 0.29 0.15 beta 0.89 0.23
a31 Non oil GDP reaction to structural public spending growth 0.13 0.12 beta 0.16 0.073
103

b10 Weight of lag inflation 0.79 0.06 beta 0.57 0.08

b11 Weight of prices expectations 1 0.075 gamma 0.1 0.18
b12 Demand price sensitiveness 0.053 0.36 gamma 0.33 1
b13 Exchange rate price sensitiveness 0.79 0.05 gamma 0.086 0.097
b14 Sensitiveness of local inflation to Imported Food prices 2 0.15 gamma 0.21 0.16
c10 Inertia of public spending gap 0.67 0.05 beta 0.52 0.045
c11 Weight of output stabilization in fiscal policy rule 1.2 0.21 gamma 0.55 0.39
c12 Public spending gap response to public debt 0.26 0.03 gamma 0.092 0.034
c30 Inertia of the growth rate of structural public spending 0.89 0.1 gamma 0.64 0.23
Continued on next page
 – Continued from previous page

Param name Labels Prior mean Prior std Prior Dist. Estim. Mode Estim. std.
c31 Structural public spending growth to non oil GDP growth 4.7 0.5 gamma 0.59 1
c32 Structural public spending growth to oil GDP growth 2.5 0.25 gamma 0.18 0.26
cd10 Public debt growth inertia 0.33 0.05 beta 0.79 0.13
cd21 Sensitiveness of public debt to exchange rate 1.5 0.25 gamma 0.13 0.81
f11 International reserves sensitiveness to economic growth 0.68 4 gamma 0.55 6.3
f21 International reserves sensitiveness to oil GDP 21 2 gamma 2.6 3.7
f31 International reserves sensitiveness to debt dynamics 31 2.4 gamma 1.6 3.7
g01 Inertia of foreign interest rate 0.84 0.08 beta 0.97 0.13
104

g02 Inertia of foreign output gap 0.57 0.08 beta 0.63 0.16
g03 Inertia of foreign inglation - Eurozone 0.24 0.04 beta 0.097 0.029
g06 Inertia of international food prices variation 0.12 0.02 beta 0.1 0.028
h10 Bank Lending Tightness inertia 0.6 0.07 beta 0.75 0.033
h11 Output expectations effect on credit supply 0.57 0.05 gamma 0.48 0.045
h12 Output past dynamics effect on credit supply 2.3 0.16 gamma 0.75 0.079
h13 Interbank rate effect on credit supply 0.05 0.006 gamma 0.059 0.005
h14 Central Bank liquidity effect on credit suply 0.35 0.06 gamma 0.17 0.023
h20 Credit gap inertia 0.74 0.2 beta 0.49 0.16
h22 Interest rate effect on credit gap 0.1 0.006 gamma 0.059 0.0099
Continued on next page
 – Continued from previous page

Param name Labels Prior mean Prior std Prior Dist. Estim. Mode Estim. std.
h30 Inertia of credit trend growth 0.64 0.1 beta 0.75 0.13
h31 Nominal growth effect on credit trend 0.75 0.1 gamma 0.44 0.096
h40 Inertia of Bank Stability Index 0.9 0.07 beta 0.72 0.093
h41 Debt importance in bank stability dynamics 0.19 0.02 beta 0.097 0.019
i10 Inertia of Interbank rate 0.98 0.045 beta 0.97 0.033
i11 Liquidity injection influence on Interbank rate 3.6 0.24 gamma 0.32 0.28
i12 Bank stability influence on interbank rate dynamics 0.49 0.2 gamma 0.39 0.34
i20 Central bank liquidity inertia 0.62 0.08 beta 0.83 0.054
105

i21 Central bank liquidity reaction to interbank spread 4.1 0.75 gamma 7.2 0.94
i22 Central bank liquidity reaction to bank instability 0.53 0.08 gamma 0.17 0.2
σ ŷ Size of Demand Shock 1.1 0.005 inv. Gamma 0.0074 0.35
σ ȳ Size of level productivity shock 1.7 0.005 inv. Gamma 0.003 0.53
σ G(ȳ) Size of growth productivity shock 1.5 0.005 inv. Gamma 0.0032 0.26
y oil
σ Size of oil production shock 2.3 0.005 inv. Gamma 0.0073 0.24
oil
σ pr Size of oil production shock 2.6 0.005 inv. Gamma 0.042 0.19
oil
σ G(pr) size of oil production growth shock 1.3 0.005 inv. Gamma 0.0079 0.39
σπ Size of cost push shock 2.1 0.005 inv. Gamma 0.031 0.091
σ ĝ Size of short term fiscal policy shock 2 0.005 inv. Gamma 0.0059 0.49
Continued on next page
 – Continued from previous page

Param name Labels Prior mean Prior std Prior Dist. Estim. Mode Estim. std.
σ G(ȳ) Size of structural fiscal policy shock (level) 2.5 0.005 inv. Gamma 0.0063 0.074
σ G(ḡ) Size of structural fiscal policy shock (growth) 0.056 0.005 inv. Gamma 0.029 0.3
σ ∆debt Size of public debt growth shock 0.44 0.005 inv. Gamma 0.021 0.2
σ debt Size of debt level shock 0.35 0.005 inv. Gamma 0.0059 0.58
σ z̄ Size of Structural exchange rate shock 2 0.005 inv. Gamma 0.0074 0.59
σ prem Size of risk premium shock 0.055 0.005 inv. Gamma 0.0073 0.7
seuro
σ Size of nominal exchange rate shock 0.56 0.005 inv. Gamma 0.0022 0.022
σ IP OLICY size of monetary policy shock 0.67 0.005 inv. Gamma 0.0035 0.52
106

σ r̄ Size of real interest rate trend shock 1.8 0.005 inv. Gamma 0.013 0.41
σ resmimp Size of international reserves shock 1.6 0.005 inv. Gamma 0.6 0.28
euro
σi Size of foreign interest rate shock 0.13 0.003 inv. Gamma 0.0044 0.022
stdystar Size of foreign output gap shock 0.062 0.003 inv. Gamma 0.017 0.036
euro
σπ Size of foreign inflation - Eurozone 0.18 0.003 inv. Gamma 0.0056 0.01
σ poil Size of oil price shock 1.8 0.005 inv. Gamma 0.19 0.18
usd
σs Size of EUR/USD exchange rate shock 0.2 0.005 inv. Gamma 0.037 0.46
f ood
σπ Size of International food prices shock 1.4 0.005 inv. Gamma 0.064 0.33
σ blt Size of temporary credit supply shock 0.12 0.005 inv. Gamma 0.03 0.55
σ cr¯ Size of permament credit supply shock 2.7 0.005 inv. Gamma 0.046 0.37
Continued on next page
 – Continued from previous page

Param name Labels Prior mean Prior std Prior Dist. Estim. Mode Estim. std.
σ cr¯ Size of credit trend growth shock 2.6 0.005 inv. Gamma 0.014 0.47
σ IIN T ERBAN K Size of interbank risk shock 1.6 0.005 inv. Gamma 0.0057 0.37
σ cbinjec Size of Central Bank liquidity injection 0.93 0.005 inv. Gamma 0.033 0.1

Source : the author

107
 Table 10 – Intra-CEMAC trade parameters

Cameroon Central Congo Gabon Eq.Guinea Chad

African
Republic

Cameroon - 0,11696 0,00010 0,00792 0,00818 0,03990

Central 0,00449 - 0,00003 0,00001 0,00001 0,00009
African
Republic
Congo 0,00995 0,00853 - 0,02583 0,03080 0,00010
Gabon 0,00573 0,00012 0,00002 - 0,00000 0,00000
Eq.Guinea 0,00463 0,00009 0,00002 0,00000 - 0,00000
Chad 0,00702 0,00039 0,00002 0,00000 0,00000 -

Source : the author

108
 In-sample forecasts

Figure 24 – In-sample unconditional forecasts of some CEMAC variables

109

Source : Author
 Figure 25 – In-sample unconditional forecasts of some Cameroon’s variables
110

Source : Author
 Figure 26 – In-sample unconditional forecasts of some Central African Republic’s variables
111

Source : Author
 Figure 27 – In-sample unconditional forecasts of some Congo’s variables
112

Source : Author
 Figure 28 – In-sample unconditional forecasts of some Gabon’s variables
113

Source : Author
 Figure 29 – In-sample unconditional forecasts of some Equatorial Guinea’s variables
114

Source : Author
 Figure 30 – In-sample unconditional forecasts of some Chad’s variables
115

Source : Author
116
 Historical decompositions

Shocks grouping

Economic Growth

Figure 31 – Historical decomposition of CEMAC economic growth

0.06

0.04
0.04
117

0.02
0.02

0
0

MonPolicy -0.02
-0.02
Cameroon
CAR
Congo
Gabon
Demand
Guinea -0.04 Supply
-0.04 Chad
MonetaryPolicy
Foreign
OilRevenues
OilPrices
ForeignShocks
CapitalFlows
CapitalFlows
others
-0.06 FinancialShocks
init
-0.06 others
init
1

3
Q

Q
10

11

13

14

16

17

19

20

10

11

13

14

16

17

19

20
20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20
Source : Author
 Inflation

Figure 32 – Historical decomposition of CEMAC annual average inflation rate

0.06

0.04

0.04

0.02
0.02

0
0
118

-0.02
-0.02
MonPolicy
Cameroon
CAR
-0.04 Congo
Gabon -0.04
Demand
Guinea
Supply
Chad
MonetaryPolicy
Foreign
-0.06 OilRevenues
OilPrices
ForeignShocks
CapitalFlows
-0.06 CapitalFlows
others
FinancialShocks
init
-0.08 others
init
1

3
Q

Q
10

11

13

14

16

17

19

20

10

11

13

14

16

17

19

20
20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20
Source : Author
 International Reserves

Figure 33 – Historical decomposition of CEMAC international reserves (absolute y-o-y variation)

Demand
2.5 2.5 Supply
MonPolicy
MonetaryPolicy
Cameroon
OilRevenues
CAR
2 ForeignShocks
Congo 2
CapitalFlows
Gabon
FinancialShocks
Guinea
others
1.5 Chad
1.5 init
Foreign
OilPrices
1 CapitalFlows
others 1
init

0.5
0.5

0
119

-0.5
-0.5

-1
-1

-1.5

-1.5

-2

-2
-2.5
1

3
Q

Q
10

11

13

14

16

17

19

20

10

11

13

14

16

17

19

20
20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20
Source : Author
 Credit to private sector

Figure 34 – Historical decomposition of CEMAC credit growth rate (y-o-y)

0.25 0.25 Demand

Supply
MonPolicy
MonetaryPolicy
Cameroon
OilRevenues
CAR
ForeignShocks
0.2 Congo 0.2 CapitalFlows
Gabon
FinancialShocks
Guinea
others
Chad
init
Foreign
0.15 OilPrices 0.15
CapitalFlows
others
init

0.1 0.1
120

0.05 0.05

0
0

-0.05
-0.05

-0.1
-0.1
1

3
Q

Q
10

11

13

14

16

17

19

20

10

11

13

14

16

17

19

20
20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20
Source : Author
121