Content uploaded by Ahmed Kchikeche
Author content
All content in this area was uploaded by Ahmed Kchikeche on Oct 11, 2024
Content may be subject to copyright.
22
Ekonomika ISSN 1392-1258 eISSN 2424-6166
2024, vol. 103(3), pp. 22–39 DOI: https://doi.org/10.15388/Ekon.2024.103.3.2
External Shocks, Asset Prices, and Credit
Dynamics in Morocco: Insights from
Disequilibrium Models
Ahmed Kchikeche*
Mohammed V University of Rabat, Morocco
ORCID: https://orcid.org/0000-0003-1928-3381
Email: ahmed_kchikeche@um5.ac.ma
Driss Mafamane
Mohammed V University of Rabat, Morocco
Orcid : https://orcid.org/0009-0008-4796-9886
Email: d.mafamane@um5r.ac.ma
Abstract. We investigate external shocks and asset price’s impact on the slowdown of business and household
credit in Morocco using disequilibrium models. The results show that banks’ y to quality, driven by a simul-
taneous decline in interest margins and borrower creditworthiness, is a key factor behind the slowdown of
credit supply. On the demand side, slower growth and saturated housing demand have contributed to reduced
borrowing and repayment capacity of borrowers. Furthermore, external shocks are transmitted to credit supply
through foreign deposits and households’ credit demand through remittances. Additionally, stocks and residential
real estate asset prices are closely tied to credit demand. These ndings suggest that addressing bank credit
barriers could stimulate economic growth. To do so, policymakers may consider employing unconventional
monetary policy tools to eectively manage the transmission channels of external shocks and asset prices to
bank credit dynamics.
Keywords: Bank credit; Asset prices; External shocks; Disequilibrium model.
1. Introduction
Over the last decades, Morocco has experienced a substantial slowdown in economic growth
(Sadok et al., 2022). This slowdown has coincided with an unprecedented decline in bank
credit to the private sector.1 In developing countries, bank credit and economic growth are
closely linked. Given the importance of bank credit as a source of nancing in these countries
(Kchikeche et al., 2024), investigating credit slowdowns has crucial policy implications, as
1 As Appendix 1 shows, the average quarterly growth of real business and household credit in Morocco has
slowed down signicantly during the last decade.
Received: 31/05/2024. Revised: 22/06/2024. Accepted: 07/07/2024
Copyright © 2024 Ahmed Kchikeche, Driss Mafamane. Published by Vilnius University Press
This is an Open Access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use,
distribution, and reproduction in any medium, provided the original author and source are credited.
Contents lists available at Vilnius University Press
* Correspondent author.
Ahmed Kchikeche, Driss Mafamane. External Shocks, Asset Prices, and Credit Dynamics in Morocco
23
determining their causes helps to choose the appropriate policy tools for intervention. For
example, monetary policy is better equipped to deal with credit supply while scal policy
is more appropriate for stimulating credit demand (Tamini & Petey, 2021).
The literature examining credit slowdowns often neglects the impact of some key
factors on bank credit dynamics. For instance, while previous studies emphasized the
impact of economic and nancial crises on the real sector (Abere & Akinbobola, 2020;
Rodríguez et al., 2023), studies on the transmission of external shocks to credit growth
are scarce (Krishnamurthy & Muir, 2017; Mamonov et al., 2020; Silalahi et al., 2012).
Furthermore, while several authors examined the impact of credit growth on asset prices
(Gerdesmeier et al., 2010; Mora, 2008; Singh & Nadkarni, 2020), studies on the eect
of asset prices on bank credit are lacking (Frommel & Karagyozova, 2008; Gupta et al.,
2022; Pouvelle, 2012).
Examining the impact of these factors requires identifying their transmission channels
to credit supply and demand. However, empirically distinguishing credit supply from
demand is challenging since they can only be jointly observed at equilibrium (Stiglitz &
Weiss, 1981). To deal with this identication problem, disequilibrium models developed
by Maddala & Nelson (1974) are used to estimate supply and demand. Studies employing
these models explain credit slowdowns, verify credit rationing, and investigate specic
bank credit determinants (Barajas & Steiner, 2002; Herrera et al., 2013; Oulidi & Allain,
2009; Pazarbasioglu, 1997; Tamini & Petey, 2021). Thus, this class of models constitutes
an appropriate tool to ll this gap in the literature.
Our study contributes to the existing literature by employing disequilibrium models
to simultaneously identify the impact of external shocks and asset prices on the supply
and demand for business and household credit. In addition to examining the traditional
determinants of bank credit, our approach investigates the underlying causes of credit
slowdowns by highlighting the transmission channels through which shocks to foreign
capital inows, liquidity, remittances, and changes in expected stock and residential
property prices inuence credit dynamics. To do so, the rst section discusses the relevant
literature, the second one describes our data and methodology, and the third one presents
and discusses our results.
2. Literature Review
Examining credit slowdowns was conducted through various methods, from regression
analysis (Bernanke et al., 1991), to survey-based studies (Ito & Pereira da Silva, 1999).
These methods, however, fail to distinguish supply from demand (Tamini & Petey, 2021).
According to Stiglitz & Weiss (1981), credit markets are characterized by disequilibri-
um, as imperfections prevent interest rates from equating supply and demand. Thus, banks
use nonprice terms to allocate credit supply at levels that can be below credit demand
at the prevailing interest rate. Conversely, the endogenous money theory stipulates that
banks are not constrained by the availability of funding and only face soft constraints
related to protability and prudential regulations (Ábel & Mérő, 2023).
ISSN 1392-1258 eISSN 2424-6166 Ekonomika. 2024, vol. 103(3)
24
Consequently, distinguishing supply from demand is necessary to properly identify
the causes of credit slowdowns (Jiménez et al., 2012). While microeconomic loan-level
data can help dierentiate supply from demand factors, data on loan applications are often
available, especially in developing countries.
Ghosh & Ghosh (2000) used aggregated data to distinguish credit supply from demand
by imposing exclusionary restrictions on the demand function, identifying two exclusive
supply variables: the bank’s ability and willingness to lend. The rst variable relies on
liquidity availability and regulatory capital compliance, while the second is inuenced by
banks’ risk perception and asset quality. Barajas & Steiner (2002) focused their analysis
on credit to the private sector in various Latin American countries. Similarly, Herrera
et al. (2013) studied the impact of international capital ows on bank credit dynamics,
showing that contracting capital inows exacerbated the post-2008 credit crunch. Also,
Dumičić & Ljubaj (2018) disaggregate private sector credit into business and household
credit, owing to the heterogeneity of these components. Recent applications of this class of
models are Adair & Adaskou (2020), Ghosh et al. (2023) and Karmelavičius et al. (2022).
Negative exogenous shocks to the economy can reduce lending (Emre Akgündüz et
al., 2021). Although promising, the examined literature often overlooks the impact of
external shocks and asset prices on lending. Silalahi et al. (2012) suggest that foreign
banks transmit nancial crises to developing countries via credit crunches. According to
Aiyar (2011), nancial stress in foreign countries translates into credit crunches through
cuts to foreign sources of funds. Finally, (Dinger & te Kaat, 2020; Mamonov et al., 2020)
show that shocks to capital inows strongly aect credit slowdowns, an impact transmitted
through credit risk (Chen & Li, 2024; Dinger & te Kaat, 2020).
Similarly, few studies examined the impact of asset prices on bank credit. Frommel &
Karagyozova (2008) argue that asset prices aect credit demand through a wealth eect,
as higher asset prices stimulate aggregate and credit demand. (Kiyotaki & Moore, 1997)
suggest that raising asset prices aects credit demand through borrowers’ net worth and
collateral. Gupta et al. (2022) show that higher asset prices used as collateral improves
credit growth. These studies presented some interesting empirical evidence. Frommel
& Karagyozova (2008) show a time-varying relationship between asset prices and
bank lending. Pouvelle (2012) illustrates that while stock prices have a robust eect on
lending, property prices only aect credit growth in periods of nancial stress. Gupta et
al. (2022) show that higher real-estate collateral value improves business credit growth.
To our knowledge, while Oulidi & Allain (2009) explored the role that asset prices
played in an earlier credit crunch (2000–2004), the recent credit slowdown in Morocco
remains understudied. Overall, the literature provides two takeaways. First, the causes of
credit slowdowns are country-specic. Second, the relative role of, and the channel by
which, external shocks and asset prices aect credit needs a thorough examination. Our
study aims to ll these gaps.
Ahmed Kchikeche, Driss Mafamane. External Shocks, Asset Prices, and Credit Dynamics in Morocco
25
3. Data and methods
3.1. Data and specication
We estimate disequilibrium models for business and household credit during the period
2006–2021. The study period is justied by the unavailability of data on lending rates
and property prices before 2006.
3.1.1. The specications of credit supply
Credit supply is determined based on a bank asset portfolio management framework
(Pazarbasioglu, 1997). Accordingly, supply is a function of the protability of the lending
activity (it), balance-sheet constraints (LRt, KRt), the bank’s evaluation of the quality of cur-
rent and future borrowers (
,
(1)
), asset prices (
,
(1)
), and external shocks (LEt).
The specications of the supply functions of business and household credit are similar
and are represented by equation (1) below.
= + + + + +
+
+
+ ×+ + + +
= + +
+ +
+
+
+ + + + +
= + +
+
+ +
+
+ + + +
(3)
(1)
where CSt denotes credit supply, it the real lending rate, Lt the liquid liabilities, KRt the
capital ratio, NPLRt the nonperforming loans ratio, IPIt
e
the expected industrial production,
RRAPt
e
the expected residential real estate prices index, FLt foreign liquidity, T a linear
trend, and D08 and D20 are dummy variables indicating the impact of the 2008 and the
COVID-19 crises, respectively.
We use a real lending rate as a measure of lending revenues (Ponomarenko, 2022), ex-
pecting it to be positively correlated with credit supply. However, information asymmetries
and adverse selection can lead to credit rationing reducing supply (Beyhaghi et al., 2020).
Balance sheet characteristics indirectly measure banks’ ability to supply credit (Al-
tavilla et al., 2021), as they constrain credit availability (Balke et al., 2021). Banks lend
to creditworthy borrowers by creating the necessary funds, without the necessity for
pre-existing deposits (Zolea, 2023). Nevertheless, liquid assets are needed for managing
liquidity risk and meeting regulatory requirements. Thus, banks lend less in case of liquidity
shortages (Thakor & Yu, 2023). We measure banks’ liquid liabilities as the logarithmic
dierence between deposits and reserves, expecting it to positively aect credit supply.
Second, supply is aected by capital’s availability as prudential regulations require a
capital buer proportional to risk-weighted assets. Given that information asymmetries
increase the cost of raising capital, undercapitalized banks can increase the capital ratio
by either reducing supply or shifting it toward less risky borrowers (Fang et al., 2022).
We measure bank capitalization using the logarithm of capital-to-credit ratio, expecting
it to positively aect credit supply.
Risk management requires constant monitoring of the quality of potential borrowers
using internal and external indicators. The NPL ratio usually indicates the quality of a
bank’s loan portfolio and the bank’s risk perception (Naili & Lahrichi, 2022). Accord-
ISSN 1392-1258 eISSN 2424-6166 Ekonomika. 2024, vol. 103(3)
26
ingly, we expect it to negatively aect supply (Tölö & Virén, 2021). Additionally, banks
monitor economic conditions to form expectations about the quality of borrowers (Ma
et al., 2021). We measure these expectations using the expected industrial production
proxied by a 4-period distributed lag of the industrial production index, expecting it to
positively aect supply.
Credit supply is closely linked to asset prices (Min et al., 2023). For instance, raising
real estate asset prices positively aects the net worth of businesses and the book value of
collateral (Gupta et al., 2022). Accordingly, we include the index of expected residential
real estate asset prices, expecting it to positively aect credit supply.
External shocks are transmitted to credit supply by reducing foreign sources of funds
(Aiyar, 2011). We account for external liquidity shocks by including the deposits of
nonresident economic agents with Moroccan banks and the interaction term between this
variable and liquid liabilities. We expect foreign liquidity to have a positive impact on
credit supply, and this impact to be negatively moderated by liquidity availability.
3.1.2. The specication of credit demand
The demand for credit is mainly determined by the cost of credit, borrowers’ expectations
about the state of the economy, asset prices and external shocks. The specication of the
demand function is dierent in the two versions of the model.
Business credit demand
= + + + + +
+
+
+ ×+ + + +
= + +
+ +
+
+
+ + + + +
= + +
+
+ +
+
+ + + +
(3)
(2)
where BCDt is business credit demand, it the real interest rate, IPIt
e
the expected industrial
production index, OGt the output gap, MASIt
e
the expected Moroccan All Share Index,
RRAPt
e
the expected residential real estate price index, Rmt remittances, FDIt foreign
direct investments, T a linear trend, and D08 and D20 are dummy variables indicating
the impact of the 2008 and the COVID-19 crises, respectively
Business credit demand in equation (2) is driven by its cost and households’ expec-
tations about the state of the economy. Therefore, we include the real lending rate as
the main component of the price of credit (Ponomarenko, 2022), assuming it negatively
aects business credit demand.
Furthermore, we use expected industrial production to represent rms’ expectations
about future cash ows. Favorable economic conditions increase income and improve
rms’ borrowing and repayment capacity (Lian & Ma, 2020). We expect this variable to
positively aect business credit demand.
We also include the output gap following Ikhide (2003). This variable accounts for the
increased demand for credit during periods of falling cashows to maintain production
levels, thus, we expect the sign of this variable to be negative.
Ahmed Kchikeche, Driss Mafamane. External Shocks, Asset Prices, and Credit Dynamics in Morocco
27
We test the impact of external shocks on business credit demand through two trans-
mission channels. First, foreign capital inows increase working capital requirements
and credit demand (Aiyar, 2011). Higher capital inows stimulate credit growth and are
generally associated with credit booms, particularly in countries with less exible ex-
change rate regimes such as Morocco (Magud et al., 2014). Consequently, we expect FDI
to positively aect demand. Second, remittances positively aect aggregate and business
credit demand, so we expect the sign of this variable to be positive.
Furthermore, we examine the role of asset prices using two variables. First, expect-
ed residential real estate asset prices increase the value of residential real estate assets,
improving the net worth of businesses and the value of their collateral (Frommel &
Karagyozova, 2008; Gupta et al., 2022)
Second, the expected Moroccan All Shares Index represents borrowers’ expectations
of economic conditions, as favorable conditions are associated with higher future in-
come. However, higher stock prices indicate an increase in the net worth of rms and,
consequently, their borrowing capacity (Pouvelle, 2012; Varadi, 2024). Finally, the index
represents the opportunity cost of bank nancing, as an improvement in this index can be
interpreted as an improvement in the attractiveness of stock market nancing. Thus, the
sign of this variable is thus ambiguous.
Household credit demand
Household credit demand is presented in equation (3) below.
= + + + + +
+
+
+ ×+ + + +
= + +
+ +
+
+
+ + + + +
= + +
+
+ +
+
+ + + +
(3)
(3)
where HCDt is household credit demand, it the real interest rate, DIt
e
the expected dis-
posable income, INFLt
e
the expected ination rate, ENDTt the indebtedness ratio, RRAPt
e
the expected residential real estate price index, Rmt remittances, T a linear trend, and
D08 and D20 are dummy variables indicating the impact of the 2008 and the COVID-19
crises, respectively
Household credit demand mainly depends on the cost of credit and the borrowing and
repayment capacity of borrowers (Ponomarenko, 2022). Thus, we use the real lending
rate as an indicator of the cost of credit and expect its sign to be negative.
Furthermore, we measure households’ borrowing capacity using expected disposable
income. An anticipated rise in income incentivizes households to increase demand. Con-
versely, in the absence of precautionary balance, a reduction in expected income worsens
the credit constraints of borrowers (Stiglitz & Guzman, 2021). Thus, this variable is
expected to positively aect credit demand.
We include the expected ination rate to measure uncertainty associated with the
expected decline in purchasing power (Tamini & Petey, 2021), as uncertainty negatively
aects credit demand (Ghosh & Ghosh, 2000). Furthermore, we also use the household
indebtedness ratio, measured by dividing national disposable income by household credit.
Since debt does not generate income for households, increasing indebtedness reduces
ISSN 1392-1258 eISSN 2424-6166 Ekonomika. 2024, vol. 103(3)
28
their consumption and repayment capacity. Therefore, we expect the sign of this variable
to be negative.
To measure the impact of asset prices on household credit demand, we use expected
residential real estate asset prices representing housing demand and collateral value.
Changes in residential real estate prices have both a wealth eect and a substitution eect;
rising property prices reduce housing and credit demand. Furthermore, rising residential
property prices may dissuade households from demanding housing credit and push them
toward renting. However, real estate can serve as collateral for a loan application, so rising
property prices improve the value of collateral and credit demand. The expected sign of
this variable is ambiguous.
Finally, we measure the transmission of external exogenous shocks to household credit
demand through remittances. This source of income can serve as a substitute for bank
credit and negatively aects credit demand. Bjuggren et al. (2010) assert that remittances
have a substitution eect on credit demand by oering an alternative to bank nancing.
The expected sign of this variable is negative.
Our data come from various national sources including Bank Al-Maghrib, High
Commissary of Planning, Casablanca Stock Exchange, and the Exchange Oce, as
well as international data sources like the International Monetary Fund and the Bank for
International Settlements.
Since we are mainly interested in long-run credit dynamics, some data transformations
are necessary. First, we eliminate the eect of ination on the levels of our variables by
deating all stock variables using the consumer price index. Second, to isolate the non-
seasonal interactions between our variables, we seasonally adjust them using the X-13
seasonal adjustment procedure. In addition, our variables are in their log-linear form for
the usual statistical reasons. Finally, all variables are standardized to avoid model insta-
bility (Guggisberg & Latshaw, 2020) and to facilitate comparing their relative impact. A
table with the variable names, their calculation formulas, and expected signs, as well as a
graphical representation of their evolution, are presented in Appendices 2 and 3. Finally,
we note that variable handling was conducted using Excel and Eviews software while
the estimations were conducted using the R script provided by Karapanagiotis (2024).
3.2. Empirical method
Our disequilibrium models consist of three equations:
,
(1)
(4)
where Dt is the demand for credit in t, St is the supply of credit in t, XDt and XSt are two
vectors of independent variables of Dt and St, respectively, and uDt and uSt are i.i.d resid-
uals. Furthermore, (6) is the short-side rule equation which, for each period t, associates
the short side (the minimum of Dt and St) with Qt the quantity of observed credit.
Ahmed Kchikeche, Driss Mafamane. External Shocks, Asset Prices, and Credit Dynamics in Morocco
29
Given this structure, the probability that an observation Qt is matched to the demand
is given by
,
(1)
(5)
Since Qt can be assigned to the demand equation with probability πt and to the supply
equation with probability 1 – πt, and given fD(Dt) and fS(Dt) the marginal density func-
tions, and FD(Dt) and FS(Dt) the cumulative probability functions of demand and supply,
respectively, the unconditional density of Qt is
,
(1)
(6)
Accordingly, L can be dened as the log-likelihood function:
,
(1)
(7)
Finally, by calculating the rst and second derivatives of L, we can obtain the maxi-
mum likelihood estimates. A complete description of the practical implementation of the
model can be found in (Karapanagiotis, 2024).
As with most macroeconomic stock series, observed credit likely has a unit root. In
the presence of nonstationarity, inference on the signicance of coecients can only be
made if observed credit forms a cointegrating vector with credit supply and demand,
respectively (Ghosh & Ghosh, 2000). To check stationarity, we use the Dickey & Fuller
(1979) unit root test. In addition, we test for cointegration using Johansen’s (1991, 1995)
trace and the maximum eigenvalue statistics.
4. Results
According to Ghosh & Ghosh (2000), inference based on disequilibrium models is only
valid if there is a cointegrating relationship between observed credit and the estimated
quantities of credit supply and demand. The unit root test results in Appendix 4 show
that all endogenous variables are I(1). Furthermore, the cointegration test results in
Appendix 5 show that there is a cointegrating relationship between observed credit
and the estimated quantities of supply and demand in both models. Consequently,
the interpretation of the regression coecients is valid. The estimation results for the
coecients of the determinants of business and household credit are presented in Tables
1 and 2, respectively.
Table 1 shows that the real lending rate positively aects business credit supply.
However, there is no evidence that real interest rates aect credit demand. These results
show that business credit supply is elastic to variations in the lending rate, whereas credit
demand is not.
ISSN 1392-1258 eISSN 2424-6166 Ekonomika. 2024, vol. 103(3)
30
Table 1. Estimation results for business credit disequilibrium model
Business credit demand Business credit supply
Variable Coecient Variable Coecient
C-2.704***
(0.085) C-0.719*
(0.413)
it
-0.056
(0.050) it
0.058***
(0.015)
IPIt
e 0.349***
(0.046) IPIt
e 0.182
(0.225)
MASIt
e -0.121***
(0.020) Lt
0.084***
(0.030)
OGt
-0.021**
(0.009) KRt
-0.105**
(0.050)
RRAP t
a 0.199***
(0.052) NPLRt
-0.381***
(0.028)
Rmd
0.040
(0.033) RRAP t
a 0.001
(0.012)
IDEd
-0.020
(0.032) LEt
0.127***
(0.022)
D08 -0.220**
(0.111) Lt × LEt
0.029
(0.037)
D20 -6.500***
(0.072) D08 0.498***
(0.125)
T0.173***
(0.004) D20 0.501***
(0.104)
T0.011
(0.011)
σt
20.003***
(0.001) σt
20.001***
(0.000)
-2LL -231.504
Note: Standard deviations are in brackets. *, **, *** mean that the coecient of the variable is signicant at
the 10%, 5%, and 1% thresholds, respectively.
Our results also show that balance sheet constraints signicantly impact business credit
supply. Liquidity positively aects business credit supply, demonstrating its importance
for managing liquidity risk. However, the impact of bank equity is contrary to expecta-
tions. By examining the evolution of NPLs in Morocco during the study period, we can
interpret this result by the possibility that faced with the deterioration in the quality of
potential borrowers, Moroccan banks simultaneously increased their capital buer and
reduced business credit supply.
Furthermore, borrower quality is the main determinant of credit supply. Indeed, the
NPL ratio is the largest signicant coecient in the business supply equation, showing its
role in gauging credit risk. Conversely, the coecient for expected industrial production
is insignicant, showing that banks rely on their internal assessment of credit risk rather
than macroeconomic indicators of economic conditions.
Ahmed Kchikeche, Driss Mafamane. External Shocks, Asset Prices, and Credit Dynamics in Morocco
31
On the demand side, expected industrial production positively aects business credit
demand. However, the output gap negatively impacts business credit demand indicating
a precautionary component of it. While both eects seem contradictory, comparing the
magnitude of the coecient demonstrates the pro-cyclicality of business credit demand
as the coecient of the IPIe is bigger.
Our results show a signicant impact of asset prices on business credit. First, property
prices signicantly impact credit demand. Thus, asset prices aect business credit demand
by aecting net worth and collateral value, in support of Gupta’s et al. (2022) ndings.
Second, changes in stock market prices negatively aect business credit demand, showing
that stock market prices aect credit demand via the substitutability of bank and stock
market nancing rather than by aecting net worth, as access to the Morrocan stock ex-
change is still limited to larger, more established rms (Kchikeche & Mafamane, 2024).
Finally, external shocks impact business credit as the dummy variables representing
the 2008 nancial crisis and the COVID-19 lockdown are signicant in the supply and
demand equations. Likewise, while external shocks have no impact on business credit
demand, they are transmitted to business credit supply through foreign liquidity, as sug-
gested by (Aiyar, 2011).
The estimates in Table 2 show that households and business credit supply respond
dierently to their determinants.
For instance, household credit supply is less sensitive to balance sheet constraints,
more aected by their expectations of the economic conditions, and not aected by the
bank’s balance sheet constraints. In addition, household credit supply is strongly aected
by banks’ risk perception, as the coecients of both the internal and external indicators
of borrowers’ quality are signicant. In contrast to business credit supply, banks rely
more on their expectation about economic conditions to gauge the quality of potential
borrowers. As for the impact of asset prices, our results show that expected property
prices do not aect household credit supply. Finally, our estimates suggest that foreign
liquidity positively aects household credit supply, an impact moderated by the level
of domestic liquidity.
On the demand side, while household credit is inelastic to the lending rate, borrowing
and repayment capacity are key determinants of credit demand. Furthermore, higher ex-
pected disposable income encourages household credit demand. Also, the indebtedness
ratio positively aects households’ demand for credit. These results show that household
credit demand is strongly pro-cyclical; borrowers who expect a rise in their disposable
income and have established relationships with banks are more likely to apply for more
loans. Our estimates also show that expected ination does not aect household credit
demand.
Property prices aect household demand through a substitution eect. Our results show
that household credit is negatively aected by residential property prices. Rising property
prices thus discourage housing and credit demand. Finally, remittances negatively aect
household credit demand as they constitute an alternative form of nancing.
ISSN 1392-1258 eISSN 2424-6166 Ekonomika. 2024, vol. 103(3)
32
Table 2. Estimation results for business credit disequilibrium model
Credit demand by households Credit supply to households
Demand Coecient Supply Coecient
C-0.473***
(0.043) C-0.103***
(0.012)
it
-0.031
(0.068) it
0.032***
(0.009)
DIt
e 0.543***
(0.049) IPIt
e 0.522***
(0.029)
INFLt
e -0.118
(0.078) Lt
0.018
(0.014)
RRAP
t
a -0.126***
(0.039) KRt
0.033
(0.027)
ENDTt
0.123***
(0.041) NPLRt
-0.082***
(0.026)
Rmt
-0.154***
(0.037) RRAP
t
a 0.000
(0.010)
D08 -0.310**
(0.144) LEt
0.111***
(0.013)
D20 0.625***
(0.000) LRt × LEt
-0.177***
(0.014)
T0.024***
(0.003) D08 -0.151***
(0.040)
D20 -0.019
(0.032)
T0.015***
(0.001)
σd
20.005***
(0.001) σt
20.000***
(0.000)
-2LL -233.3598
Note: *, **, *** denote that the coecient of the variable is signicant at the 10%, 5% and 1% thresholds,
respectively.
5. Conclusion
We investigate the impact of asset prices and external shocks on business and household
credit between 2006 and 2021 using two disequilibrium models. Our results suggest that
the decline in credit supply is attributed to the ight of banks to quality caused by the
simultaneous deterioration in interest margins and the quality of borrowers. The declining
quality of borrowers is caused by higher indebtedness, saturated housing demand, and
slower economic growth. Our results also show that external shocks aect business and
household credit supply through their impact on nonresident deposits with Moroccan
banks. On the demand side, our results show that remittances harm households’ credit
demand and that stocks and residential property prices impact the demand for business
and household credit.
Ahmed Kchikeche, Driss Mafamane. External Shocks, Asset Prices, and Credit Dynamics in Morocco
33
Our study is not without limitations. Despite disaggregating private-sector credit into
rm and household credit, sectoral or bank-level data could provide more insights into bank
credit dynamics in Morocco. Also, our quantitative focus neglects the potential impact of
structural and institutional factors on credit dynamics in Morocco. For instance, a lack of
nancial innovation can explain credit slowdowns (Lee et al., 2020) by aecting credit
risk (Khan et al., 2021) and SME lending (Hryckiewicz et al., 2023). Further investigation
along these lines is needed.
Despite these limits, our study has important practical applications and policy rec-
ommendations. First, recent evidence by Kchikeche & Khallouk (2021) and Kchikeche
& Mafamane (2024) reveals a causal impact of private-sector credit on GDP in the short
and long run. Thus, the slowdown in credit to the private sector may explain the observed
economic slowdown in Morocco since 2009. Public policy aiming to stimulate economic
growth in Morocco and other developing countries should closely monitor barriers to
private sector nancing, as nancial constraints could constitute bottlenecks to sustained
growth in emerging markets. Policymakers could either investigate barriers to credit growth
or support the development of other nancing providers. Second, our results highlight
the role of banks as transmitters of external shocks to the nancial and real spheres of the
Moroccan economy. As the Moroccan economy and nancial system continue to liberalize
and integrate with the global supply chain, public policy should consider the spillover
eects of asset price uctuations, foreign liquidity drops, and remittances slowdowns on
the nancial constraints of Moroccan rms and households to solidify the resilience of
the Moroccan economy. Finally, we contribute to explaining the disconnect between the
dynamics of interest rates and credit growth (Kchikeche et al., 2024).
Our results suggest that the failure of conventional interest-based monetary policy to
stimulate credit growth in Morocco is due to the relatively weak elasticity of credit supply
to changes in lending rates. Thus, stimulating credit, and consequently, economic growth
may necessitate the development and use of unconventional monetary policy tools.
References
Ábel, I., & Mérő, K. (2023). Endogenous money, soft budget constraint and the banking regulation. Acta
Oeconomica, 73(S1), 99–118. https://doi.org/10.1556/032.2023.00036
Abere, S. S., & Akinbobola, T. O. (2020). External Shocks, Institutional Quality, and Macroeconomic Perfor-
mance in Nigeria. SAGE Open, 10(2). https://doi.org/10.1177/2158244020919518
Adair, P., & Adaskou, M. (2020). Credit Rationing and Mature French SMEs: A Disequilibrium Model. World
Journal of Applied Economics, 6(1), 55–72. https://doi.org/10.22440/wjae.6.1.4
Aiyar, S. (2011). How did the crisis in international funding markets aect bank lending? Balance sheet evidence
from the United Kingdom (Bank of England Working Paper No. 424). https://www.bankofengland.co.uk/
working-paper/2011/how-did-the-crisis-in-international-funding-markets-aect-bank-lending
Altavilla, C., Boucinha, M., Holton, S., & Ongena, S. (2021). Credit Supply and Demand in Unconventional
Times. Journal of Money, Credit and Banking, 53(8), 2071–2098. https://doi.org/10.1111/jmcb.12792
Balke, N. S., Zeng, Z., & Zhang, R. (2021). Identifying credit demand, nancial intermediation, and supply of
funds shocks: A structural VAR approach. North American Journal of Economics and Finance, 56. https://
doi.org/10.1016/j.najef.2021.101375
ISSN 1392-1258 eISSN 2424-6166 Ekonomika. 2024, vol. 103(3)
34
Barajas, A., & Steiner, R. (2002). Why Don’t They Lend? Credit Stagnation in Latin America. IMF Economic
Review, 49(1), 156–184. https://doi.org/https://doi.org/10.2307/3872475
Bernanke, B., Lown, C. S., & Friedman, B. M. (1991). The Credit Crunch. Brookings Papers on Economic
Activity, 2, 205–247. https://www.brookings.edu/articles/the-credit-crunch/
Beyhaghi, M., Firoozi, F., Jalilvand, A., & Samarbakhsh, L. (2020). Components of credit rationing. Journal
of Financial Stability, 50, 100762. https://doi.org/10.1016/j.jfs.2020.100762
Bjuggren, P.-O., Dzansi, J., & Shukur, G. (2010). Remittances and investment (CESIS Electronic Working
Paper No. 216). https://static.sys.kth.se/itm/wp/cesis/cesiswp216.pdf
Chen, R., & Li, J. (2024). How Do Short-Term Cross-Border Capital Flows Aect Bank Risk-Taking? Evi-
dence from China. Emerging Markets Finance and Trade, 60(5), 1064–1076. https://doi.org/10.1080/15
40496X.2023.2267739
Dickey, D., & Fuller, W. (1979). Distribution of the Estimators for Autoregressive Time Series With a Unit
Root. Journal of the American Statistical Association, 74(366), 427. https://doi.org/10.2307/2286348
Dinger, V., & te Kaat, D. M. (2020). Cross-border capital ows and bank risk-taking. Journal of Banking and
Finance, 117. https://doi.org/10.1016/j.jbankn.2020.105842
Dumičić, M., & Ljubaj, I. (2018). Delayed Credit Recovery in Croatia: Supply or Demand Driven? Journal of
Central Banking Theory and Practice, 7(1), 121–144. https://doi.org/10.2478/jcbtp-2018-0006
Emre Akgündüz, Y., Cılasun, S. M., Dursun-de Neef, H. Ö., Hacıhasanoğlu, Y. S., & Yarba, İ. (2021). How
do banks propagate economic shocks? (Central Bank of the Republic of Türkiye Working Paper No.
21/24). https://www.tcmb.gov.tr/wps/wcm/connect/EN/TCMB+EN/Main+Menu/Publications/Research/
Working+Paperss/2021/21-24
Fang, X., Jutrsa, D., Peria, S. M., Presbitero, A. F., & Ratnovski, L. (2022). Bank capital requirements and
lending in emerging markets: The role of bank characteristics and economic conditions. Journal of Banking
and Finance, 135. https://doi.org/10.1016/j.jbankn.2020.105806
Frommel, Michael., & Karagyozova, Kristina. (2008). Bank lending and asset prices: evidence from Bulgaria
(Bulgarian National Bank Discussion Paper No. DP/65/2008). https://library.centralbankmalta.org/cgi-bin/
koha/opac-detail.pl?biblionumber=3326
Gerdesmeier, D., Reimers, H., & Roa, B. (2010). Asset Price Misalignments and the Role of Money and
Credit. International Finance, 13(3), 377–407. https://doi.org/10.1111/j.1468-2362.2010.01272.x
Ghosh, S., & Ghosh, A. (2000). East Asia in the aftermath: Was there a crunch? (Deutsche Bank Research
Note No. 5). https://hdl.handle.net/10419/40258
Ghosh, S., Herwadkar, S., Verma, R., & Gopalakrishnan, P. (2023). Disentangling demand and supply side
determinants of post-GFC credit slowdown: an Indian perspective. Indian Economic Review, 58(S2),
399–421. https://doi.org/10.1007/s41775-023-00177-w
Giuliano, P., & Ruiz-Arranz, M. (2009). Remittances, nancial development, and growth. Journal of Devel-
opment Economics, 90(1), 144–152. https://doi.org/10.1016/j.jdeveco.2008.10.005
Guggisberg, M., & Latshaw, N. (2020). Disequilibrium: Model for markets in disequilibrium. https://cran.r-pro-
ject.org/package=Disequilibrium
Gupta, A., Sapriza, H., & Yankov, V. (2022). The Collateral Channel and Bank Credit. (Board of Governors
of the Federal Reserve System Finance and Economics Discussion Series No. 2022-024). https://doi.
org/10.17016/FEDS.2022.024
Herrera, S., Hurlin, C., & Zaki, C. (2013). Why don’t banks lend to Egypt’s private sector? Economic Mod-
elling, 33(June), 347–356. https://doi.org/10.1016/j.econmod.2013.04.003
Hryckiewicz, A., Korosteleva, J. A., Kozłowski, Ł., Rzepka, M., & Ruomeng, W. (2023). Bank Technological
Innovation and SME Lending: Do We Experience a Transformation in a Bank-SME Relationship? SSRN
Electronic Journal. https://doi.org/10.2139/ssrn.4390485
Ahmed Kchikeche, Driss Mafamane. External Shocks, Asset Prices, and Credit Dynamics in Morocco
35
Ikhide, S. (2003). Was There a Credit Crunch in Namibia Between 1996–2000? Journal of Applied Economics,
6(2), 269–290. https://doi.org/10.1080/15140326.2003.12040595
Ito, T., & Pereira da Silva, L. (1999). The Credit Crunch in Thailand during the 1997-98 crisis : Theoretical
and operational issues the JEXIM survey. Exim Review, 19(2), 1–40. https://www.researchgate.net/pub-
lication/236861255_The_Credit_Crunch_in_Thailand_during_the_1997-98_crisis_Theoretical_and_op-
erational_issues_the_JEXIM_survey
Jiménez, G., Ongena, S., Peydró, J.-L., & Saurina, J. (2012). Credit Supply and Monetary Policy: Identifying
the Bank Balance-Sheet Channel with Loan Applications. American Economic Review, 102(5), 2301–2326.
https://doi.org/10.1257/aer.102.5.2301
Johansen, S. (1991). Estimation and Hypothesis Testing of Cointegration Vectors in Gaussian Vector Autore-
gressive Models. Econometrica, 53(9), 1689–1699. https://doi.org/10.2307/2938278
Johansen, S. (1995). A Statistical Analysis of Cointegration for I(2) Variables. Econometric Theory, 11(1),
25–59. https://doi.org/10.1017/S0266466600009026
Karapanagiotis, P. (2024). The R Package markets: Estimation Methods for Markets in Equilibrium and Dis-
equilibrium. Journal of Statistical Software, 108(2). https://doi.org/10.18637/jss.v108.i02
Karmelavičius, J., Mikaliūnaitė-Jouvanceau, I., & Petrokaitė, A. (2022). Housing and credit misalignments in
a two-market disequilibrium framework. (Bank of Lithuania Occasional Paper No. 42). https://www.lb.lt/
uploads/publications/docs/36380_ac93efa620cef0f0a5e72207fca18012.pdf.
Kchikeche, A., El fakir, R., & Mafamane, D. (2024). Can Conventional Monetary Policy Stimulate Bank
Credit? Evidence from a Developing Country. Statistika: Statistics and Economy Journal, 104(2). https://
doi.org/10.54694/stat.2023.54
Kchikeche, A., & Khallouk, O. (2021). On the nexus between economic growth and bank-based nancial
development: evidence from Morocco. Middle East Development Journal, 13(2), 1–20. https://doi.org/1
0.1080/17938120.2021.1930830
Kchikeche, A., & Mafamane, D. (2024). Examining the Interactions of Economic Growth and Bank Credit to
the Private Sector in Morocco: A Causality Analysis. Journal of Economic Cooperation and Development,
45(3). https://www.sesric.org/publications-jecd-articles.php?jec_id=133
Khan, A. B., Fareed, M., Salameh, A. A., & Hussain, H. (2021). Financial Innovation, Sustainable Economic
Growth, and Credit Risk: A Case of the ASEAN Banking Sector. Frontiers in Environmental Science, 9.
https://doi.org/10.3389/fenvs.2021.729922
Kiyotaki, N., & Moore, J. (1997). Credit Cycles. Journal of Political Economy, 105(2), 211–248. https://doi.
org/https://doi.org/10.1086/262072
Krishnamurthy, A., & Muir, T. (2017). How Credit Cycles Across A Financial Crisis (NBER Working Paper
No. 23850). http://www.nber.org/papers/w23850
Lee, C.-C., Wang, C.-W., & Ho, S.-J. (2020). Financial innovation and bank growth: The role of institu-
tional environments. The North American Journal of Economics and Finance, 53, 101195. https://doi.
org/10.1016/j.najef.2020.101195
Lian, C., & Ma, Y. (2020). Anatomy of Corporate Borrowing Constraints. The Quarterly Journal of Economics,
136(1), 229–291. https://doi.org/10.1093/qje/qjaa030
Ma, Yueran, Teodora Paligorova, and José-Luis Peydro (2021). Expectations and Bank Lending. Working
Paper. https://memento.ep.ch/public/upload/les/MaPaper.pdf
Maddala, G. S., & Nelson, F. D. (1974). Maximum Likelihood Methods for Models of Markets in Disequilib-
rium. Econometrica, 42(6), 1013. https://doi.org/10.2307/1914215
Magud, N. E., Reinhart, C. M., & Vesperoni, E. R. (2014). Capital Inows, Exchange Rate Flexibility and
Credit Booms. Review of Development Economics, 18(3), 415–430. https://doi.org/10.1111/rode.12093
Mamonov, M., Pankova, V., Akhmetov, R., & Pestova, A. (2020). Financial Shocks and Credit Cycles. Russian
Journal of Money and Finance, 79(4), 45–74. https://doi.org/10.31477/rjmf.202004.45
ISSN 1392-1258 eISSN 2424-6166 Ekonomika. 2024, vol. 103(3)
36
Min, F., Wen, F., Xu, J., & Wu, N. (2023). Credit supply, house prices, and nancial stability. International
Journal of Finance & Economics, 28(2), 2088–2108. https://doi.org/10.1002/ijfe.2527
Modigliani, F., & Jaee, D. M. (1969). A Theory and Test of Credit Rationing. The American Economic
Review, 59(5), 850–872. https://www.jstor.org/stable/1810681
Mora, N. (2008). The Eect of Bank Credit on Asset Prices: Evidence from the Japanese Real Estate. Journal
of Money, Credit and Banking, 40(1), 57–87. https://doi.org/10.1111/j.1538-4616.2008.00104.x
Naili, M., & Lahrichi, Y. (2022). Banks’ credit risk, systematic determinants and specic factors: recent
evidence from emerging markets. Heliyon, 8(2), e08960. https://doi.org/10.1016/j.heliyon.2022.e08960
Oulidi, N., & Allain, L. (2009). Credit Market in Morocco: A Disequilibrium Approach. (IMF Working Paper
No. WP/09/53). https://doi.org/10.5089/9781451872019.001
Pazarbasioglu, C. (1997). A Credit Crunch? Finland in the Aftermath of the Banking Crisis. Sta Papers -
International Monetary Fund, 44(3), 315. https://doi.org/10.2307/3867562
Ponomarenko, A. (2022). Money creation and banks’ interest rate setting. Journal of Financial Economic
Policy, 14(2), 141–151. https://doi.org/10.1108/JFEP-10-2020-0214
Pouvelle, C. (2012). Bank credit, asset prices and nancial stability: Evidence from French banks (IMF Working
Paper No. WP/12/103). http://www.imf.org/external/pubs/cat/longres.aspx?sk=25874
Rodríguez, G., Vassallo, R., & Castillo B., P. (2023). Eects of external shocks on macroeconomic uctuations
in Pacic Alliance countries. Economic Modelling, 124. https://doi.org/10.1016/j.econmod.2023.106302
Sadok, H., Elfakir, R., & Hakik, I. (2022). Optimal ination threshold and economic growth in Morocco: an
empirical analysis. Journal of Economic Cooperation and Development, 43(4), 103–130. https://jecd.
sesric.org/pdf.php?le=ART22081601-2.pdf
Silalahi, T., Wibowo, W. A., & Nurlian, L. (2012). Impact Of Global Financial Shock To International Bank
Lending In Indonesia. Buletin Ekonomi Moneter Dan Perbankan, 15(2), 75–110. https://doi.org/10.21098/
bemp.v15i2.423
Singh, B., & Nadkarni, A. R. (2020). Role of credit and monetary policy in determining asset prices: Evidence
from emerging market economies. The North American Journal of Economics and Finance, 51, 100874.
https://doi.org/10.1016/j.najef.2018.11.003
Stiglitz, J., & Guzman, M. (2021). The pandemic economic crisis, precautionary behavior, and mobility con-
straints: an application of the dynamic disequilibrium model with randomness. Industrial and Corporate
Change, 30(2), 467-497. https://doi.org/10.1093/icc/dtab012
Stiglitz, J., & Weiss, A. (1981). Credit Rationing in Markets with Imperfect Information. The American Eco-
nomic Review, 75(4). https://www.jstor.org/stable/1802787
Tamini, A., & Petey, J. (2021). Hoarding of reserves in the banking industry: Explaining the African paradox.
Quarterly Review of Economics and Finance, 81, 214–225. https://doi.org/10.1016/j.qref.2021.06.002
Thakor, A., & Yu, E. (2023). Funding Liquidity Creation by Banks (Federal Reserve Bank of Philadelphia
Working Paper No. WP 23-02). https://doi.org/ 10.21799/frbp.wp.2023.02
Tölö, E., & Virén, M. (2021). How much do non-performing loans hinder loan growth in Europe? European
Economic Review, 136, 103773. https://doi.org/10.1016/j.euroecorev.2021.103773
Varadi, A. (2024). Identifying the transmission channels of credit supply shocks to household debt: Price and non-
price eects. European Economic Review, 166, 104747. https://doi.org/10.1016/j.euroecorev.2024.104747
Zhao, Y., & Xu, Z. (2021). The Impact of Cross-Border Capital Flows on the Chinese Banking System. SAGE
Open, 11(2), 215824402110214. https://doi.org/10.1177/21582440211021410
Zolea, R. (2023). A Functional Analysis of the Role of Deposits in the Traditional Banking Industry. Review
of Political Economy, 35(4). https://doi.org/10.1080/09538259.2023.2233870
Ahmed Kchikeche, Driss Mafamane. External Shocks, Asset Prices, and Credit Dynamics in Morocco
37
Appendices
Appendix 1: The evolution of business and household credit in Morocco during the study period.
6. Appendices
Appendix 1: The evolution of business and household credit in
Morocco during the study period.
18,07%
12,57%
-0,11%
-0,91%
1,48%
5,16%
15,66%
7,13%
6,97%
2,32%
3,75%
2,30%
2006-2008 2009-2011 2012-2014 2015-2017 2018-2019 2020-2021
Growth of real business credit Growth of real household credit
Appendix 2: Data description and expected signs.
Variable Formula Expected Sign
itNominal interest rate – smoothed ination rate (+) on supply /
(-) on demand
LtLog(deposits – reserves) (+)
KRtEquity/Loans (+)
NPLRtNonperforming loans/Total loans (-)
IPI
t
e4-period distributed lag of the industrial production index (+)
RRAP
t
e4-period distributed lag of the residential real estate asset
prices index Ambiguous
FLt
Log(deposits of nonresident economic agents with
Moroccan banks) (+)
MASI
t
ethe expected Moroccan All Shares Index Ambiguous
INFL
t
e4-period distributed lag of the ination rate (-)
OGtReal GDP/Potential GDP – 1 (-)
DI
t
e4-period distributed lag of the log(real disposable
income) (+)
ENDTtDisposable income/Household credit (-)
RmtLog(Remittances)/ Log(GDP) (-)
FDItLog(Foreign direct investment)/Log(GDP) (+)
ISSN 1392-1258 eISSN 2424-6166 Ekonomika. 2024, vol. 103(3)
38
Appendix 3: The evolution of variables during the study period.
Appendix 3: The evolution of variables during the study period.
-3
-2
-1
0
1
2
06 08 10 12 14 16 18 20
BC
-3
-2
-1
0
1
2
06 08 10 12 14 16 18 20
HC
-3
-2
-1
0
1
2
3
06 08 10 12 14 16 18 20
I
-2
-1
0
1
2
06 08 10 12 14 16 18 20
L
-2
-1
0
1
2
06 08 10 12 14 16 18 20
K
-2
-1
0
1
2
3
4
06 08 10 12 14 16 18 20
NPLR
-3
-2
-1
0
1
2
06 08 10 12 14 16 18 20
IPI
-3
-2
-1
0
1
2
06 08 10 12 14 16 18 20
DI
-6
-4
-2
0
2
4
06 08 10 12 14 16 18 20
OGAP
-2
-1
0
1
2
3
06 08 10 12 14 16 18 20
INF
-4
-2
0
2
4
06 08 10 12 14 16 18 20
RPPI
-4
-3
-2
-1
0
1
2
06 08 10 12 14 16 18 20
MASI
-2
-1
0
1
2
3
4
06 08 10 12 14 16 18 20
RM
-6
-4
-2
0
2
06 08 10 12 14 16 18 20
FDI
-2
-1
0
1
2
06 08 10 12 14 16 18 20
FL
Appendix 4: Unit root test results.
Variables ADF statistic Decision
BC Level -2.542 I(1)
1st dierence -2.005**
BCS Level -3.036 I(1)
1st dierence -6.802***
BCD Level -1.693 I(1)
1st dierence -7.832***
HC Level -3.479 I(1)
1st dierence -9.586***
BCS Level -2.232 I(1)
1st dierence -7.102***
BCD Level -1.006 I(1)
1st dierence -8.441***
Source: Author’s calculations
Note: *, **, *** mean that the test statistic is above the critical value at the 10%, 5% and 1% thresholds,
respectively.
Ahmed Kchikeche, Driss Mafamane. External Shocks, Asset Prices, and Credit Dynamics in Morocco
39
Appendix 5: Cointegration test results.
Model 1
Cointegration test using the trace statistic
H0 Eigenvalue Trace Critical value at 5% Decision
r = 0 0.306 25.959 15.495 Reject
r ≤ 1 0.051 3.271 3.841 Do not reject
Cointegration test using the maximum eigenvalue statistic
H0 Eigenvalue Max eigenvalue Critical value at 5% Decision
r = 0 0.306 22.689 14.264 Reject
r ≤ 1 0.051 3.271 3.841 Do not reject
Model 2
Cointegration test using the trace statistic
H0 Eigenvalue Trace Critical value at 5% Decision
r = 0 0.253 18.729 15.495 Reject
r ≤ 1 0.015 0.954 3.841 Do not reject
Cointegration test using the maximum eigenvalue statistic
H0 Eigenvalue Max eigenvalue Critical value at 5% Decision
r = 0 0.253 17.775 14.264 Reject
r ≤ 1 0.015 0.954 3.8414 Do not reject
Source: Author’s calculations
