Finance, Accounting and Business Analysis

Volume 7 Issue 1, 2025

http://faba.bg/

ISSN 2603-5324

DOI: https://doi.org/10.37075/FABA.2025.1.06

Asymmetric Impact of Interest Rate on Economic Growth in Kenya

Talknice Saungweme

* , Glenda Maluleke

, Nicholas M. Odhiambo

Department of Economics, University of South Africa, South Africa

* Corresponding author

Info Articles

Abstract

History Article:

Submitted 30 January 2025

Revised 17 April 2025

Accepted 28 April 2025

Purpose: This study re-examines the relationship between interest rates

and economic growth, focusing on the asymmetric effects of lending

interest rates on Kenya's economic performance.

Design/Methodology/Approach: The study applied the nonlinear

autoregressive distributed lag (NARDL) model to ascertain the distinct

impacts of positive and negative interest rate shocks on economic growth

in both the short and long run. It uses yearly time series data spanning

the years 1980-2021.

Findings: The results of the cointegration tests found evidence

supporting the existence of an asymmetric long-run relationship, while

the Wald test results show that there is a long-run and short-run

asymmetry link between interest rates and economic growth in Kenya.

On average, positive changes in lending interest rates have no significant

impact on economic growth in Kenya, both in the short and long run.

However, negative interest rate shocks spur economic growth in the

short run but impede growth in the long run.

Research Limitations/Implication: The study is limited to the Kenyan

context and the dataset range of 1980–2021. Future research could

explore thresholds for optimal interest rate levels and include a broader

range of countries for comparative analysis.

Originality/Value: This study uniquely applies the NARDL framework

to Kenya, providing new insights into the asymmetric impact of interest

rates on economic growth.

Paper Type: Research Paper.

Keywords:

Economic growth, interest

rates, Kenya, NARDL

JEL: C32, E43, O42

Address Correspondence:

E-mail: talknice2009@gmail.com

malulg@unisa.ac.za

odhianm@unisa.ac.za

T. Saungweme, G. Maluleke, N. Odhiambo / Finance, Accounting and Business Analysis, Volume 7, Issue 1, 2025

INTRODUCTION

Introduction, theoretical and empirical literature synthesis

Beginning with the Asian Tiger economies and shifting to newly industrialised economies, such as

China, South Africa, India, Russia and Brazil, the most powerful and successful strategy for poverty

reduction is through sustained economic growth emanating from aggressive and rapid innovation and

industrialisation (Chatterjee and Naka 2022; United Nations Industrial Development Organisation 2020).

These two groups of countries pursued both export-oriented and import substitution strategies, while

advancing innovative production technologies (Liu et al. 2023). While both monetary and nonmonetary

factors can be drivers of economic growth, nonmonetary factors have received most of the attention in

research studies. This insight serves as the motivation for this paper, which has two primary goals. The first

is to thoroughly review and document the relationship between lending interest rates and economic growth

in Kenya. Lending interest rates can have a significant impact on the rate and trajectory of economic growth

by influencing the magnitude and return of investment, as well as the scope and composition of both saving

and consumption. Infrastructure development, industrialisation, institutional investors, mutual funds, and

the corporate sector are all exposed to risks stemming from interest rate volatility (Olasehinde-Williams et

al. 2024).

The second is to bridge the gap in interest rate modelling by applying a model that can be used to

quantify and comprehend the nature of the link between interest rates and growth in Kenya. Specifically,

this study investigates the asymmetric impact of interest rates on economic growth in Kenya using time

series data from 1980 to 2021. The nonlinear ARDL method captures the positive and negative changes

asymmetrically and the short- and long-run dynamics of interest rates on economic growth in Kenya,

enabling a more precise analysis across different economic conditions (see Saungweme et al. 2024; Shin et

al. 2014). To the best of our knowledge, this is the first analysis of its kind conducted in Kenya, and it is

unique since it employs an advanced estimation procedure that takes into account the asymmetrical

characteristics of lending interest rates.

Therefore, the primary goals of this study are to complement previous growth research on Kenya and

to support ongoing reforms in the areas of monetary, economic, and financial policy. These reforms are

essential to preserving macroeconomic stability, preserving debt sustainability, strengthening market

confidence, and enhancing the achievement of Kenya's medium-term growth prospects (International

Monetary Fund/IMF 2024; Odhiambo and Saungweme 2023a; Saungweme and Odhiambo 2021). In

addition, Kenya’s economy faces unique structural challenges and external shocks, such as fluctuating global

interest rates and capital flows (IMF 2024). Therefore, understanding how interest rate fluctuations affect

growth can provide insights into optimal policy decisions for sustainable development.

From a theoretical standpoint, there are multiple opposing hypotheses about the relationship between

interest rates and economic growth. The first is a cogent explanation of the boom-bust pattern offered by the

Austrian school of economic thought. That is, low interest rates from the central bank would encourage

investment bubbles, which would then lead to a burst in asset prices, a financial crisis, and a severe recession

(Foldvary 2015). The rate of interest is interpreted by the Austrian school as reflecting a methodical

discounting of future values. The Austrian hypothesis states that the relationship between interest rates and

economic growth typically revolves around the time preference issue. For example, increased productivity

could encourage people to invest more now, making present-day investing more preferred over future

investment (Holmes 2011). Furthermore, the market for loanable funds—funds that are accessible for

borrowing—determines the interest rate (Foldvary 2015). Borrowers will be able to access more funding for

consumption and investment at reduced interest rates. In general, Austrian economics holds that a central

bank's manipulation of money and interest rates is what causes recessions; the best way to prevent these

controls is to let the money supply and interest rates be determined by free market forces in money and

banking.

Keynesian theory comes second. Keynes' approach to interest rate dynamics is in contradistinction

with loanable funds theory. For Keynes, interest rate dynamics is based on his conception of ontological

uncertainty, liquidity preference, investors’ expectations and animal spirits, financial institutions, financial

markets, and institutional practices (Akram 2021). According to Keynes, the short-term interest rate is

determined by the central bank's policy rate, which then affects the long-term interest rate (Akram 2021).

These long-term interest rates then influence investment, saving and consumption decisions in the economy.

The McKinnon-Shaw hypothesis comes in third. In their original individual works, they contended

that financial policies in developing and emerging economies, including low and restricted interest rates and

restrictive credit management, among other financial repression acts, result in a decrease in savings,

investment and economic growth (Wilson and Odhiambo 2023). McKinnon (1973) studied an economy in

which the vast majority of investors had very limited access to external financing. In his view, savers may

find it more convenient to build up their money in financial assets until they have sufficient funds to invest

T. Saungweme, G. Maluleke, N. Odhiambo / Finance, Accounting and Business Analysis, Volume 7, Issue 1, 2025

in higher-yielding physical assets (Leshoro and Wabiga 2023). Thus, deposits can act as a channel for the

accumulation of capital, rendering deposits and capital complementary assets. The availability of deposits

with positive real rates of return may thus encourage both saving and capital accumulation.

In contrast to McKinnon, Shaw (1973) focused more on external rather than internal financing

options as a fundamental constraint to capital formation. Shaw also underlined the significance of positive

real deposit rates as an incentive to save in financially depressed economies. Shaw (1973) emphasised that

high deposit rates might encourage investment spending by enabling the credit supply to grow in accordance

with the financing requirements of the economy's productive sectors (Iddrisu and Alagidede 2020). Thus,

McKinnon (1973) and Shaw (1973) suggest that low interest rates do not really increase investment and

economic growth (Owusu 2023). After analysing the McKinnon-Shaw arguments, Mohlo (1986) came to

the conclusion that deposits and physical capital complement each other in an intertemporal fashion, with

current deposits being used to fund future investments. This link suggests that higher deposit rates inhibit

investment in the short run but will eventually boost it in the long run.

There is currently little but growing empirical research on the link between interest rates and economic

growth (Leshoro and Wabiga 2023; Adabor 2022). First, Leshoro and Wabiga (2023) looked at how both

positive and negative interest rate shocks affect private investment in South Africa. The study employed

annual time series data from 1971 to 2019 and a nonlinear autoregressive distributed lag technique. The

results indicate that interest rates and private investment exhibit short-run and long-run asymmetric

relationships, with private investment responding differently to negative and positive shocks in interest rates.

Second, by applying the NARDL approach, Adabor (2022) tested the asymmetric impact of lending

interest rates on economic growth in Ghana using yearly time series data covering the period of 1970 to

2019. The study found evidence of long-run and short-run asymmetrical effects of lending on economic

growth in Ghana. The findings further show that positive changes in lending rates generate a decrease of

nearly 0.2% in economic growth while negative changes lead to an increase of about 0.7% in economic

growth.

Other non-asymmetric studies conducted on the interest rate-growth linkage include Lee and Werner

(2023), Shaukat et al. (2019), Awad and Al Karaki (2019). Lee and Werner (2023) analysed the impact of

interest rates on economic growth in 19 industrialised and emerging economies. The analysis used a time-

varying dynamic conditional correlation in a GARCH model and further tested the direction of causation

between the two variables in the studied economies. The results provide evidence consistent with the

conclusion that lowering interest rates is counterproductive when trying to stimulate the economy.

Awad and Al Karaki (2019) examined the impact of bank lending on economic growth in Palestine

using quarterly time series data for the period from 1996 to 2015. The study employed the vector

autoregressive model and vector error correction model, as well as the Granger causality test to test the

underlying relationships. The study found that there was a statistically insignificant relationship between

bank lending and economic growth. Additionally, there is evidence of unidirectional causality that runs from

economic growth to bank lending.

Shaukat et al. (2019) studied the mechanism by which the real interest rate establishes a negative

effect on economic growth in 38 transitory economies. The study applied a dynamic panel data technique

based on the Generalised Method of Moments for the period 1996-2015. The study found that during the

transition period of developing economies, a high real interest rate restricts the economy's potential to grow.

Considering the aforementioned theoretical stances and empirical evidence, the goal of this study is

to empirically test the interest rate-growth relationship tailored to the Kenyan context. The remainder of the

research is arranged as follows: Section 2 provides an overview of interest rates and economic growth trends

in Kenya. Data, methodology, and estimation techniques are provided in Section 3. The empirical analysis

is presented in Section 4, and the main conclusions and policy implications are summed up in Section 5.

Overview of interest rate and economic growth trends in Kenya

Kenya experienced a general balance of payments surplus from 1964 to 1972, with the exception of

1964, 1967, and 1971 (World Bank 2022). Interest rate policy in Kenya was largely dormant throughout this

time. The government managed interest rates by setting minimum savings for all deposit-taking

institutions and minimum lending rates for commercial banks, non-bank financial institutions, and building

societies (Baynham 1989). A variety of internal and external causes, notably inflation brought on by a

significant rise in oil prices and the consequences of a severe national drought in 1973, had a negative impact

on the economy between 1972 and 1981 (IMF 1985). Following the shocks, the government re-examined its

regulatory structure, which resulted in the progressive escalation of restrictions on imports, exports, interest

rates, and domestic pricing.

Between 1974 and 1989, the monetary authorities in Kenya used an administered interest rate

framework. The employment of statutory credit ceilings and minimum savings deposit rates was a crucial

T. Saungweme, G. Maluleke, N. Odhiambo / Finance, Accounting and Business Analysis, Volume 7, Issue 1, 2025

tool for managing market liquidity (IMF 1991a). A revision to the credit ceiling policy between 1989 and

1991 resulted in the maximum rate being split between short-term and long-term lending (IMF 1991a). The

major challenge with controlling interest rates was a widening gap between administered and effective bank

lending rates. For this reason, the monetary authorities have been unable to sell enough quantities of

government securities to limit the growth of the money supply (IMF 1991b).

The government then started a series of economic adjustment programs with the World Bank and

IMF in 1991 and continued them until 1996. The overarching policy framework placed a strong emphasis

on trade liberalisation, interest rate deregulation, and macroeconomic stabilisation (Obrien and Ryan 1999).

In 1991, this framework resulted in a comprehensive liberalisation policy ideology. This philosophy is

supported by the theoretical expectations of the McKinnon-Shaw hypothesis, which postulates that easing

regulatory restrictions on interest rates will increase the volume of funds in deposit-taking

institutions (McKinnon 1973; Shaw 1973).

Regarding economic growth, the years 1963 and 1973 can be regarded as a decade of exceptional

growth, with an average annual real growth rate of 6.7% (World Bank 2022). The favourable weather and

trade conditions for Kenya's commodity exports, along with the successful execution of the import-

substitution program, all contributed to the country's impressive economic performance (Odhiambo and

Saungweme 2023b). Between 1974 and 1979, the country underwent an economic recession marked by

declining terms of trade and rising oil prices (IMF 2022). The period 1980-85 can be described as an era of

macroeconomic imbalance and stabilisation, with low growth. Economic and monetary policy interventions

between 1985 and 1989, which were further enhanced by the 1986 coffee boom, led to an economic rebound.

For the first half of this period, real GDP increased by an average of 5.3% yearly (World Bank 2022). Figure

1 shows interest rates and GDP per capita growth rates for the period 1980 to 2022.

Source: Authors’ compilation using World Bank (2022) data

Figure 1. Interest rate and economic growth trends in Kenya (1980-2022)

Figure 1 largely describes a negative correlation between interest rates and annual growth of GDP

per capita in Kenya. Interest rates rose sharply during times of economic instability, such as 1980–1985,

1989–1992, 1995–1996, 2008, and 2019. Despite the series of economic reforms and monetary reforms,

interest rates in Kenya remained high over the period under review, 1980-2022. These high interest rates on

loans from the banking sector have been perceived by some policymakers as an obstacle to greater

investment, financial inclusion, and economic growth (IMF 2019). As a result, the monetary authorities in

Kenya were compelled to revert to managing interest rates in 2016. As a result, the observable flip in interest

rates from 2016 is consistent with the capping of interest rates in Kenya, which went into effect that same

year (Central Bank of Kenya/CBK 2018). According to the new rule, the maximum lending rate cannot be

more than 4% over the base rate set by the central bank (CBK 2018).

However, the interest rate caps on commercial loans were lifted in 2019, and this is shown by an

upturn in interest rates in Figure 1 (IMF 2021). The goal of eliminating interest rate ceilings in 2019 was to

facilitate greater credit expansion and to stimulate private investment (IMF 2021). Despite the repeal,

lending rates charged by banks have not increased significantly over the prior cap rate. In 2023, as part of an

ongoing set of reforms, the Kenyan central bank established a new interest rate corridor to guide the

-10

-5

1980 1985 1990 1995 2000 2005 2010 2015 2020

Annual %

Interest rates GDP per capita growth

T. Saungweme, G. Maluleke, N. Odhiambo / Finance, Accounting and Business Analysis, Volume 7, Issue 1, 2025

overnight interbank rate and reduce the premium for its discount (IMF 2024). The initiative was undertaken

to further strengthen the monetary policy implementation framework. From the economic growth front, the

notable downward spike in GDP per capita between 1990 and 1992 can be linked to the suspension of the

balance of payments assistance from bilateral donors, while the downward spike in the 2019-20 period is

attributable to the COVID-19 pandemic shock and ongoing volatility in the global financial system (see also

IMF 2022).

METHODOLOGY

Data description

The study used annual time-series data from 1980 to 2021. The availability of reliable and consistent

time-series data for each of the model variables for the study country influenced the selection of the data

range. The data used in this study were obtained from the World Bank's online database. Table 1 presents

the definition of variables, measurements, and expected signs.

Table 1. Definitions of variables and data sources

Variables

Definitions of variables

(Measurements)

A priori expectation

Economic growth (Y)

GDP per capita (constant 2015

US$)

Dependent variable

Interest rate (INT)

Lending interest rate (%)

+/-

Investment (INV)

Gross fixed capital formation (% of

GDP)

Financial development (CRED)

Domestic credit to private sector (%

of GDP)

Trade openness (TO)

Trade (% of GDP)

+/-

Exchange rate (EXC)

Official exchange rate (LCU per

US$, period average)

+/-

Model specification

The baseline model used in this study is defined as follows:





 

󰇛



󰇜

(1)

Where Table 1 defines each variable.

In order to obtain elasticity coefficients on the variables and minimise the impact of outliers, the

variables are converted to logarithms. Therefore, equation (1) is specified as follows:





 



 







 







 







 







 







 



(2)

There is a vast array of literature that attempts to theorise the numerous causes of economic growth.

The core tenet of the Harrod-Domar (H-D) model of economic growth is that increased production levels at

the micro and macro levels are the result of a progressive accumulation of additional physical capital

financed by savings and investments (Nguyen 2023). Subsequent to the H-D model is the Solow-Swan

model, which emphasises the smooth substitution between capital and labour (Nguyen 2023). Early in the

20th century, financial depth and trade openness were recognised as crucial components of economic growth

(Odhiambo and Saungweme 2023a). Nyasha et al. (2021), use GDP per capita as the dependent variable to

facilitate cross-country comparisons of different population sizes. In other words, GDP per capita adjusts

nominal GDP for changes in price levels and population growth.

The lending interest rate is the primary independent variable in this study. Theoretically, interest rates

are regarded as the costs of borrowing investment and consumption funds from financial institutions. Bank

lending rates are expected to either have a positive or negative relationship with economic growth. A

decrease in lending rates is expected to induce borrowing for investment and consumption, hence leading to

economic growth (Adabor 2022; Foldvary 2015). However, an increase in lending interest rates can lead to

a decline in economic growth as it increases the cost of borrowing, which might discourage borrowing for

investment and consumption, which is needed to grow the economy.

The incorporated control variables in model 1, namely investment, financial development, trade

T. Saungweme, G. Maluleke, N. Odhiambo / Finance, Accounting and Business Analysis, Volume 7, Issue 1, 2025

openness, and exchange rate, are buttressed by both theory and prior empirical evidence; hence, it is

anticipated that their coefficients will be statistically significant. The Solow-Swan model states that

investment variations have an impact on total production up until steady-state per capita income is reached.

This suggests that while investment has a significant role in determining growth in the short term, its impact

on national output is essentially neutral over the long term. Empirical studies consistent with a positive

relationship between investment and economic growth include Odhiambo and Saungweme (2023b) and

Ibrahimov et al. (2023).

The inclusion of financial development in the model is due to the perceived positive spillover effects

in an economy reliant on knowledge and technology, which ultimately leads to economic growth (Giri et

al. 2023). Furthermore, the underlying theory supporting the inclusion of trade openness in the baseline

model argues that trade openness either amplifies or attenuates the effects of globalisation, which in turn

encourages technology transfer and foreign direct investment, which eventually influences economic growth

(Chen et al. 2020). However, Baliamoune-Lutz and Ndikumana (2007), in their panel data analysis from 39

African countries covering the period 1975-2001, found evidence in support of a negative relationship

between trade openness and economic growth. This relationship was caused by weak institutions in the

studied economies.

The role of exchange rates in influencing economic growth cannot be understated, particularly in an

analysis of a commodity-exporting country, such as Kenya. Overvalued currencies can be linked to

macroeconomic instability, rent-seeking and corruption, unsustainable high current account deficits, foreign

exchange shortages, and balance of payments crises (Rodrik 2008). All these factors are detrimental to

economic growth.

Estimation techniques

This study applies a nonlinear autoregressive distributed lag (NARDL) model. The NARDL

model explicitly captures the short-run and long-run asymmetries (positive and negative shocks) in lending

interest rates on economic growth—a major advantage over linear vector autoregressive models. The

cumulative dynamic multipliers of the NARDL model further explain the speed with which economic

growth returns to equilibrium following a negative or positive shock in lending interest rates (Shin et al.

2014). The superiority of NARDL over other competing techniques is its ability to give reliable coefficients

even in small samples, account for short-run volatilities and structural break problems in the data, account

for endogeneity among all the variables, and its applicability to data with mixed orders of integration of at

most one (Shin et al. 2014).

Thus, following Shin et al. (2014), interest rates can be decomposed into partial sums of positive

changes and negative changes, making it possible to examine the marginal impact of the two components

on economic growth in Kenya. This gives the following expression:





 









 









 



 



(3)

where:







  









  

󰇛







󰇜





(4)







  









  󰇛



󰇜





(5)

Where change, Z is a set of control variables, 



is white noise error term.

Using equations (4) and (5), the NARDL framework as defined by Shin et al. (2014) is specified as:





 



 









































































 











 











 







 











 











 







 























 



(6)

Where   is optimal lag order, 



is constant, 























are short-run

T. Saungweme, G. Maluleke, N. Odhiambo / Finance, Accounting and Business Analysis, Volume 7, Issue 1, 2025

coefficients, 











are short-run asymmetric distributed lag parameters, 























are

long-run coefficients, 











are long-run asymmetric distributed lag parameters, 



is white noise error

term, t is time period and l is natural logarithmic transformation. All other variables are as defined in Table

Cointegration in a NARDL model setting is ascertained by comparing the computed F-statistic to the

upper and lower critical bounds from Pesaran et al. (2001) critical values. The asymmetric impact of lending

interest rates on economic growth in Kenya is ascertained if 





 





. If this condition holds, then the

following error correction model (ECM) is specified:





 



 









































































 











 







 













(7)

Where is coefficient of the error term and ECM is error correction term.

To confirm convergence to long-run equilibrium following a shock or short-term disequilibrium, the

coefficient of the error correction term () is anticipated to be negative and statistically significant, lying

between 0 and 1.

The current paper makes use of time series data, so it is necessary to pre-test each variable for unit

root in order to prevent spurious regressions and to determine the order of integration for each variable.

According to Pesaran et al. (2001) and Shin et al. (2014), the NARDL model requires that no variable be

integrated of an order higher than one. To distinctly determine the order of integration, the paper uses three

techniques, namely, the Dickey-Fuller Generalised Least Square (DF-GLS), Phillips-Perron (PP) and Zivot-

Andrews (ZAURoot) techniques. The paper incorporates the ZAURoot technique so as to correct for

structural breaks and, therefore, correctly determine the order of integration among the variables.

After ascertaining the order of integration for each variable, the paper then conducts a cointegration

test to ascertain the applicability of the NARDL process. The paper also undertakes nonlinearity tests in the

series using the BDS test. The null hypothesis of linearity, under various BDS dimensions (m = 2, 3, 4, 5, 6),

is put to the test. The Wald test is also included in this paper to check for both short-run and long-run

asymmetries. After undertaking these preliminary checks, the paper proceeds to estimate the asymmetrical

long- and short-run effects of lending interest rates on economic growth using equations 5 and 6. Finally,

the paper carries out post-diagnostic tests, including the recursive CUSUM and CUSUMSQ tests, to check

the null hypothesis that the parameters are unstable. The dynamic multiplier tests further show graphically

the rate of response of economic growth to positive and negative variations in interest rates (see Shin et al.,

2014).

RESULTS AND DISCUSSION

Nonlinearity and stationarity results

The study first determines whether the series has a nonlinear relationship. The findings are shown in

Table 2. There is evidence indicating the existence of a nonlinear relationship between the series in Table 2.

This result is confirmed by the BDS test statistics for each variable, which were found to be statistically

significant at 1% across all dimensions. Panels A, B, and C of Table 3 provide a summary of the three

stationarity test results. The results of the DF-GLS test show that investment is stationary at all levels [I(0)],

while GDP per capita, trade openness, and exchange rate are all conclusively stationary after first

differencing [I(1)]. According to the findings of the PP and ZAURoot tests, all series are conclusively

integrated of order one (1). Overall, the stationarity results indicate that the order of integration of the

variables is a mixture of not more than 1. This attests to the appropriateness of using the bounds test to

investigate the long-run relationship between interest rates and GDP per capita in Kenya.

T. Saungweme, G. Maluleke, N. Odhiambo / Finance, Accounting and Business Analysis, Volume 7, Issue 1, 2025

Table 2. BDS Test results for nonlinearity

Variables

BDS Statistic

Dimension 2

Dimension 3

Dimension 4

Dimension 5

Dimension 6

BDS

Statistic

P-

value

BDS

Statistic

P-

value

BDS

Statistic

P-

value

BDS

Statistic

P-

value

BDS

Statistic

P-

value



0.153***

0.000

0.239***

0.000

0.274***

0.000

0.279***

0.000

0.251***

0.000



0.146***

0.000

0.237***

0.000

0.290***

0.000

0.318***

0.000

0.326***

0.000



0.070***

0.000

0.100***

0.000

0.110***

0.000

0.107***

0.000

0.093***

0.000



0.105***

0.000

0.153***

0.000

0.180***

0.000

0.184***

0.000

0.170***

0.010



0.132***

0.000

0.193***

0.000

0.201***

0.000

0.169***

0.000

0.113***

0.000



0.203***

0.000

0.343***

0.000

0.442***

0.000

0.510***

0.000

0.558***

0.000

Notes: *** denote statistical significance at 1% leve1.

Table 3. Stationarity results

Panel A: Dickey-Fuller Generalised Least Square (DF-GLS)

Variable

Level

First Difference

Without Trend

With Trend

Without Trend

With Trend



1.488

-0.434

-4.021***

-4.814***



-1.217

-1.425

-4.591***

-5.611***



-2.774***

-2.980*

___



-1.196

-3.094*

-7.389***

-7.882***



-0.605

-1.974

-6.602***

-6.608***



0.086

-1.073

-4.368***

-5.705***

Panel B: Phillips-Perron (PP)

Level

First Difference

Without Trend

With Trend

Without Trend

With Trend



1.419

-0.323

-3.949***

-4.710***



-1.862

-2.061

-5.432***

-5.606***



-2.732*

-2.890

-11.123***

-11.332***



-1.102

-3.521*

-8.038***

-7.958***



-0.888

-1.848

-6.605***

-6.655***



-3.207**

-1.772

-5.039***

-5.615***

Panel C: Zivot-Andrews (ZAURoot)

Level

First Difference

Without

Trend

Break

With

Trend

Break

Without

Trend

Break

With

Trend

Break



-2.466

1992

-3.339

2000

-5.441***

1991

-5.730***

1990



-2.936

1989

-3.177

1999

-6.836***

1995

-7.043***

1995



-4.198

1996

-4.261

1996

-6.443***

2000

-6.415***

2000



-3.765

2006

-4.018

2006

-6.464***

1991

-7.196***

2013



-3.635

2015

3.814

2010

-6.906***

1988

-7.147***

1995



-4.254

1991

-6.761***

1993

-6.276***

1995

-6.571***

1994

Notes: ** and *** denotes statistical significance at 5% and 1% level.

Cointegration and asymmetric test results

Table 4 displays the outcomes of the cointegration tests carried out utilising the NARDL bounds

testing methodology. The results show that the F-statistic value for the NARDL model is 4.215 and is

statistically significant at the 5% level. This suggests that the variables in the nonlinear model have a

cointegrating correlation. The Wald test results presented in Table 5 firmly reject the null hypothesis of long-

run and short-run symmetry. This is confirmed by the associated long-run and short-run p-value of the Wald

F-statistic (W

), which is statistically significant at 1%, and W

SR,

which is significant at the 10% level. This

finding implies that interest rates have a distinct long-run and short-run asymmetric effect on economic

growth in Kenya.

T. Saungweme, G. Maluleke, N. Odhiambo / Finance, Accounting and Business Analysis, Volume 7, Issue 1, 2025

Table 4. Bounds F-test results for cointegration – NARDL

F-Statistic

Cointegration Status

4.215**

Cointegrated

Pesaran et al.

(2001), p.300,

Table CI(iii)

Case III

Asymptotic critical values for

10%

I(0)

I(1)

I(0)

I(1)

I(0)

I(1)

2.53

3.59

2.87

3.6

4.9

Notes: **denotes statistical significance at 5% level.

Table 5. Wald test results

Test

F-statistic

P-value

Decision

9.201***

0.007

Asymmetric

4.046*

0.056

Asymmetric

Notes: 



is long-run asymmetric test

;

is short-run asymmetric test; *** and * signifies significance at

1% and 10% level.

Long-run and short-run NARDL results

Table 6, panels A and B, presents the long-run and short-run NARDL results, respectively.

Table 6. NARDL results - long-run and short-run coefficients

Dependent Variable is Y

Panel A: Long-Run Results

Regressor

Coefficient

T-ratio [p-value]





-0.086

-0.980 [0.339]





0.461***

4.399 [0.000]



-0.043

-0.269 [0.791]



0.066

0.570 [0.575]



0.056

0.480 [0.637]



-0.322***

-4.174 [0.001]

Panel B: Short-Run Results

Regressor

Coefficient

T-ratio [p-value]



4.229***

6.225 [0.000]



0.031***

6.335 [0.000]





0.030

0.721 [0.480]





0.041

1.278 [0.217]





󰇛󰇜

-0.178***

-3.989 [0.001]





󰇛󰇜

-0.120***

-3.297 [0.004]



0.057*

1.817 [0.085]

󰇛󰇜

0.131***

4.327 [0.000]



-0.089***

-3.097 [0.006]

󰇛󰇜

-0.076**

2.408 [0.026]



0.056*

2.052 [0.054]

󰇛󰇜

-0.061**

-2.254 [0.036]



-0.183***

-3.972 [0.001]

󰇛󰇜

-0.567***

-5.904 [0.000]

Panel C: Test statistics

R- Squared

R-Bar-Squared

F-Statistic [Prob]

Normality

Serial Correlation

Heteroscedasticity

Functional Form

0.793

0.693

7.959 [0.000]

0.779 [0.677]

0.969 [0.400]

1.007 [0.493]

0.879 [0.361]

Notes: *, ** and *** denote statistical significance at 10%, 5% and 1% levels, respectively “+” and “-”

denotes positive and negative shocks.

T. Saungweme, G. Maluleke, N. Odhiambo / Finance, Accounting and Business Analysis, Volume 7, Issue 1, 2025

The NARDL results, which are reported in Table 6, indicate that positive changes in interest rates

󰇛



󰇜 have no significant impact on economic growth, irrespective of whether the analysis is conducted

in the short run or in the long run. The results also indicate that negative changes in interest rates 󰇛



󰇜

and economic growth move in the same direction, as evidenced by the positive and statistically significant

coefficient of the negative partial sum of interest rates. This suggests that, other things being equal, negative

shocks to interest rates are likely to be accompanied by a corresponding decline in economic growth in the

long run. One reason for this could be that when the central bank lowers interest rates, economic agents save

less because they will not be earning higher returns on their savings. The low saving rate can lead to a lower

level of investment, which could contribute to a decrease in economic growth.

The results also indicate that there is an inverse relationship between negative changes in lending

interest rates from the preceding period and economic growth in the short run. This is supported by the

coefficient of the partial negative sum of interest rates, which has been found to be negative and statistically

significant. This indicates that decreases in interest rates have the potential to spur economic growth in the

short run, as lower rates can incentivise borrowing for both investment and consumption. The major findings

of this study indicate that negative changes in interest rates have an asymmetrical impact on economic

growth in Kenya, depending on the direction of change and the time scale taken into account. Overall, based

on these conclusions, the study concludes that negative interest rate changes play a significant role in

defining the country's possibilities for future prosperity.

The results of control variables reported in panels A and B show that investment has a significant

positive impact on economic growth in Kenya, only in the short run. This finding is consistent with the

principles of H-D and Solow-Swan models presented in earlier sections. The findings of financial

development point to detrimental effects on economic growth exclusively in the short run, while it is

statistically insignificant in the long run. Trade openness has been shown to promote economic growth in

the short run, but the results also indicate that trade openness from previous periods causes economic

decline. While the adverse effects of financial development may indicate a small financial sector vulnerable

to adverse financial developments in international markets, Baliamoune-Lutz and Ndikumana (2007)

contend that weak institutions to facilitate reciprocal trade benefits are the root cause of the negative

correlation between trade openness and economic growth (see also, IMF, 2024). Furthermore, it was shown

that changes in exchange rates have a negative impact on economic growth in the long and short run. This

implies that a depreciation of the exchange rate in Kenya is likely to boost exports and increase economic

growth.

The dynamic multiplier graph presented in Figure 2 validates the presence of an asymmetric

relationship between interest rates and economic growth. Explicitly, the black dotted line indicates the non-

linear adjustment of economic growth to negative shocks, while the solid black line portrays the adjustment

of economic growth to positive shocks. Overall, the dynamic multiplier reported in Figure 2 shows that the

effects of positive shocks to interest rates are more pronounced than those of negative shocks in the long run.

The CUSUM and CUSUMSQ graphs presented in Figure 3 are within the bounds at a 5% significance

level, implying that the estimated model passes the stability test, confirming the consistency and reliability

of the coefficients.

Figure 2. Dynamic multiplier graph

-1.0

-0.8

-0.6

-0.4

-0.2

0.0

0.2

1 3 5 7 9 11 13 15

Multiplier for LINT(+)

Multiplier for LINT(-)

Asymmetry Plot (with C.I.)

T. Saungweme, G. Maluleke, N. Odhiambo / Finance, Accounting and Business Analysis, Volume 7, Issue 1, 2025

Figure 3a. CUSUM graph Figure 3b. CUSUMSQ graph

CONCLUSION AND RECOMMENDATIONS

The economic and monetary policies of Kenya underwent significant evolution from 1964 to 2023.

Initially characterised by dormant interest rate policies managed through statutory controls, the landscape

shifted by 1991 as internal and external shocks prompted economic adjustments emphasizing trade

liberalisation and interest rate deregulation. Therefore, this study extended the investigation of the

asymmetric impact of lending interest rates on economic growth in Kenya using a nonlinear ARDL model

and annual time-series data spanning the years 1980-2021. Unlike some of the previous studies, which used

linear models, this research employed the NARDL framework to examine the short- and long-run

asymmetric impact of lending interest rates on economic growth in Kenya, providing an intricate

understanding of the interest rate-growth nexus. The findings underscore the existence of an asymmetric

long-run relationship, as evidenced by cointegration and dynamic asymmetry tests. Specifically, while

positive shocks in lending interest rates exhibit no significant impact on economic growth, negative interest

rate changes are shown to spur growth in the short run but impede it in the long run. Short-run benefits of

lower interest rates stem from increased borrowing for productive investments, enhanced consumption, and

the stimulation of export-driven industries such as agriculture and manufacturing, owing to currency

depreciation. Conversely, prolonged interest rate reductions may deter savings and compromise long-term

investment, posing challenges to sustained economic growth. Investment, financial development, trade

openness, and exchange rates were found to play critical roles in moderating the growth trajectory, though

their effects varied across time horizons.

Policy suggestions from this paper are: (1) Policymakers should carefully balance short-term stimulus

policies with long-term sustainability by avoiding extreme interest rate fluctuations. Short-term measures

should focus on reducing lending rates during economic slowdowns to encourage borrowing for investment,

entrepreneurship, and job creation. (2) Given that investment positively affects economic growth in the short

run, particularly with a lag, policymakers should implement short-term measures to stimulate both private

and public investments, such as subsidies and credit facilities. Since investment also contributes to long-term

economic growth, authorities should establish sustainable financing mechanisms, such as public-private

partnerships for infrastructure projects. (3) As credit to the private sector has a negative impact on economic

growth in the short run, financial regulators should closely monitor excessive lending to ensure that credit

allocation supports productive sectors rather than speculative activities. Authorities should promote

responsible lending through macro-prudential policies while also developing robust credit risk management

frameworks and fostering financial literacy to enhance the positive long-run effects of financial

intermediation. (4) Since exchange rate depreciation negatively affects economic growth in both the short

and long run, the central bank should continuously monitor exchange rate fluctuations and, when necessary,

implement appropriate monetary policies - within the bounds of its existing free-floating exchange rate

system - to smooth out extreme and undesirable movements. (5) Export diversification strategies should be

encouraged to reduce vulnerability to external shocks and enhance economic resilience.

Future studies on the subject should extend the analysis to estimate the threshold point of lending

interest rates that would set the country on an optimal growth path. Additionally, the scope of analysis

should be broadened to include other African economies to uncover regional trends and policy implications.

-15

-10

-5

2004 2006 2008 2010 2012 2014 2016 2018 2020

CUSUM 5% Significance

-0.4

0.0

0.4

0.8

1.2

1.6

2004 2006 2008 2010 2012 2014 2016 2018 2020

CUSUM of Squares 5% Significance

T. Saungweme, G. Maluleke, N. Odhiambo / Finance, Accounting and Business Analysis, Volume 7, Issue 1, 2025

REFERENCES

Akram, T. 2021. A Keynesian approach to modelling the long-term interest rate. Levy Economics Institute.

Working Paper No. 988.

Awad, I. M., and M. S. Karaki. 2019. The impact of bank lending on Palestine economic growth: an

econometric analysis of time series data. Financial Innovation, 5(14): 1-21.

https://doi.org/10.1186/s40854-019-0130-8

Baliamoune-Lutz, M., and L. Ndikumana. 2007. The growth effects of openness to trade and the role of

institutions: New evidence from African countries. Economics Department Working Paper Series. 38.

https://doi.org/10.7275/1068933

Baynham, S. 1989. The Kenyan economy: An overview. Africa Insight, 19(4).

https://journals.co.za/doi/pdf/10.10520/AJA02562804_392

Central Bank of Kenya/CBK. 2018. Central Bank of Kenya bank supervision annual report 2018.

https://www.centralbank.go.ke/uploads/banking_sector_annual_reports/1174296311_2018%20

Annual%20Report.pdf

Chatterjee, M., and I. Naka. 2022. Twenty years of BRICS: political and economic transformations through

the lens of land. Oxford Development Studies, 50(1): 2-13.

https://doi.org/10.1080/13600818.2022.2033191

Chen, H., D. O. Hongo, M. W. Ssali, M. S. Nyaranga, and C. W. Nderitu 2020. The asymmetric influence

of financial development on economic growth in Kenya: evidence from NARDL. Sage Open, 1-17.

https://doi.org/10.1177/2158244019894071

Foldvary, F. E. 2015. The Austrian theory of the business cycle. The American Journal of Economics and

Sociology, 74(2): 278-297.

Giri, A. K., G. Mohapatra and B. Debata. 2023. Technological development, financial development, and

economic growth in India: Is there a non-linear and asymmetric relationship? Journal of Economic

and Administrative Sciences, 39(1): 117-133. DOI:10.1108/JEAS-03-2021-0060

Holmes, M. J., J. Otero and T. Panagiotidis. 2011. The term structure of interest rates, the expectations

hypothesis and international financial integration: Evidence from Asian economies. International

Review of Economics and Finance, 20: 679-689. DOI: 10.1016/j.iref.2010.11.021

Ibrahimov, Z., S. Hajiyeva, I. Seyfullayev, U. Mehdiyev and Z. Aliyeva. 2023. The impact of infrastructure

investments on the country’s economic growth. Problems and Perspectives in Management, 21(2): 415-

425. Doi:10.21511/ppm.21(2).2023.39

Iddrisu, A. A., and I. Alagidede. 2020. Is the interest rate setting behaviour of the Bank of Ghana

constrained by high debt levels? African Development Review, 32(3): 459-471.

International Monetary Fund/IMF. 1985. Kenya: The general setting: Financial policy workshops.

International Monetary Fund, 9-23.

https://www.elibrary.imf.org/downloadpdf/book/9780939934379/ch02.pdf

International Monetary Fund/IMF. 1991a. 23 Kenya’s Transition from direct to indirect instruments of

monetary policy, in, The Evolving Role of Central Banks, eds, Downes, P., and Vaez-Zadeh, R.

https://doi.org/10.5089/9781557751850.071

International Monetary Fund/IMF. 1991b. Coordination of financial policy, in, Determinants and Systemic

Consequences of International Capital Flows. IMF. Washington, DC.

International Monetary Fund/IMF. 2019. Do interest rate controls work? Evidence from Kenya. IMF

Working paper. WP/19/119.

International Monetary Fund/IMF. 2021. Kenya 2021 article IV consultation; second reviews under the

extended arrangement under the extended fund facility and under the arrangement under the

extended credit facility. IMF country report no. 21/275.

International Monetary Fund/IMF. 2022. Fourth reviews under the extended arrangement under the

extended fund facility and under the arrangement under the extended credit facility. IMF Country

Report No. 22/382.

International Monetary Fund/IMF. 2024. Kenya 2023 Article IV consultation, sixth reviews under the

extended fund facility and extended credit facility arrangements. IMF Country Report No. 24/13.

Lee, K.S., and R.A. Werner. 2023. Are lower interest rates really associated with higher growth? New

empirical evidence on the interest rate thesis from 19 countries. International Journal of Finance

and Economics, 28(4): 3467-4737. https://doi.org/10.1002/ijfe.263

Leshoro, T.L.A., and P. Wabiga. 2023. The Asymmetric effects of interest rates on private investment in

South Africa. Acta Universitatis Danubius. Œconomica, 19(3): 161–182. https://dj.univ-

danubius.ro/index.php/AUDOE/article/view/2335

Liu, F., X. Zhang, A. S. Adebayo and A. A. Awosusi. 2022. Asymmetric and moderating role of

industrialisation and technological innovation on energy intensity: Evidence from BRICS

T. Saungweme, G. Maluleke, N. Odhiambo / Finance, Accounting and Business Analysis, Volume 7, Issue 1, 2025

economies. Renewable Energy, 198(October), 1364-1372.

https://doi.org/10.1016/j.renene.2022.08.099

McKinnon, R. I. 1973. Money and capital in economic development. Washington, DC: The Brookings

Institution.

Modugu, K. P., and J. Dempere. 2022. Monetary policies and bank lending in developing countries:

evidence from Sub-Sahara Africa. Journal of Economics and Development, 24(3): 217-

229. https://doi.org/10.1108/JED-09-2021-0144.

Mohlo, L.E. 1986. Interest rates, saving, and investment in developing countries: A re-examination of the

McKinnon-Shaw hypotheses. International Monetary Fund.

https://doi.org/10.5089/9781451956726.024

Nguyen, Q. K. 2023. Macroeconomic determinants of economic growth in low- and mid-income countries:

new evidence using a non-parametric approach. Applied Economics Letters, 1-7.

https://doi.org/10.1080/13504851.2023.2283774

Nyasha, S., N. M. Odhiambo and S. A. Asongu. 2021. The impact of tourism development on economic

growth in Sub-Saharan Africa. The European Journal of Development Research, 33(6): 1514-1535.

O’Brien, F. S., and T. C. I. Ryan. 1999. Aid and reform in Africa: Kenya case study. World Bank.

https://policycommons.net/artifacts/1515025/aid-and-reform-in-africa/2189840/

Odhiambo, N. M., and T. Saungweme. 2023a. Financial development and economic growth in Africa, in,

Finance for Sustainable Development in Africa. Routledge: Taylor & Francis Group Publishers.

Odhiambo, N. M. and T. Saungweme. 2023b. Public investment and economic growth in Kenya, Economics,

Management, and Financial Markets, 18(2): 54-76.

Olasehinde-Williams, G., R. Omotosho and F. V. Bekun. 2024. Interest rate volatility and economic growth

in Nigeria: new insight from the quantile autoregressive distributed lag (QARDL) model. Journal

of the Knowledge Economy. https://doi.org/10.1007/s13132-024-01924-x

Owusu, E. L. 2023. Interest rate reforms in African countries, in, Finance for Sustainable Development in Africa.

Routledge: Taylor & Francis Group Publishers.

Pesaran, M., H. Y. Shin, and R. J. Smith. 2001. Bounds testing approaches to the analysis of level

relationships. Journal of Applied Econometrics, 16: 289-326.

Saungweme, T and N. M. Odhiambo. 2021. Inflation and economic growth in Kenya: An empirical

examination. Advances in Decision Sciences (ADS), 25(3): 1-25.

https://doi.org/10.47654/v25y2021i3p1-25

Shaukat., B., Q. Zhu and M. I. Khan. 2019. Real interest rate and economic growth: A statistical exploration

for transitory economies. Physica A: Statistical Mechanics and Its Applications, 534: 122193.

https://doi.org/10.1016/j.physa.2019.122193

Shaw, E. 1973. Financial deepening in economic development. New York: Oxford University Press.

Shin, Y., B. Yu, M. Greenwood-Nimmo. 2014. Modelling asymmetric cointegration and dynamic multipliers in a

nonlinear ARDL framework. In: Sickles, R., Horrace, W. (eds). Springer, New York, NY.

https://doi.org/10.1007/978-1-4899-8008-3_9

United Nations Industrial Development Organisation. 2020. Industrialisation as the driver of sustained prosperity.

Vienna. UNIDO.

Wilson, M. K., and N. M. Odhiambo. 2023. Financial Reforms and Financial Development in Africa:

Evidence from Selected SSA Countries, in, Finance for Sustainable Development in Africa. Routledge:

Taylor & Francis Group Publishers.

World Bank. 2022. World Bank online database. Washington DC: World Bank.