Volume 8, Issue 10, October – 2023 International Journal of Innovative Science and Research Technology

ISSN No:-2456-2165

Macroeconomic Variables and Government Bond

Yields Listed on the Nairobi Securities Exchange
Martin Kilombe Muti1, Dr. Gordon Opuodho2
Department of Economics, Accounting and Finance
Jomo Kenyatta University of Agriculture and Technology.
Nairobi, Kenya

Abstract:- The main objective of the study was to other macroeconomic factors, like unemployment rates
determine the relationship between macroeconomic and government expenditures, impact government bond
variables and government bond yields listed on the yields.
Nairobi Securities Exchange (NSE). The specific
objectives were to investigate the impact of the inflation Keywords:- Government Bond Yields, Inflation Rate,
rate, economic growth rate, foreign direct investment Economic Growth Rate, Exchange Rate, Foreign Direct
(FDI), and exchange rate on government bond yields Investment, Cointegration, Vector Error Correction Model,
listed on the Nairobi Securities Exchange (NSE). The Stationarity.
study adopted a quantitative research design and utilized
secondary data on nineteen,15-year Kenyan government I. INTRODUCTION
bonds listed on the NSE from the 1st quarter of 2007 to
the 1st quarter of 2023, that is a sixteen-year period. The A bond is a loan where the issuer is required to repay
analysis focused on yearly yield variations over the the capital at maturity and compensate the buyer with one or
maturity period of the bonds using data obtained from more future cash flows (Chorafas, 2005). They are
the Central Bank of Kenya, Kenya National Bureau of responsible for repaying lenders or investors the money
Statistics and World Bank. The Vector Error Correction borrowed plus interest over a predetermined period
Model technique was employed to identify the long and (Choudhry, 2006). These cash flow(s) may have a
short run relationships between the macroeconomic predetermined schedule and magnitude or depend on some
factors and government bond yields in EViews. economic variable whose value is typically known a priori.
Diagnostic tests included the Augmented Dickey Fuller Government bonds are types of bonds issued by the federal
test and Johansen Cointegration Tests; to test for or national governments of countries to raise funds while
stationarity and long run relationship between variables rewarding the issuer with annual interest rates over the
respectively. Lag selection was carried out and an maturity period (Ślusarczyk et al., 2020). According to
optimal lag of 1 was selected based on the Akaike McKay and Peters (2019), bond yield refers to interest
information criterion (AIC) and Schwarz information generated from bond investments over the maturity
criterion. VECM was found to be the most suitable period(s).
model since all the time series data of the variables was
found to be stationary upon first difference and there Developing economies in Africa and throughout the
was presence of at least one cointegrating equation. The world are grappling with mounting budget deficits, rising
study established government bond yields were interest rates, and rising inflation rates. As a result, bond
significantly affected by Foreign Direct Investment, issuance remains a crucial part of Kenya's and other
Exchange Rate, and Inflation Rate on the long-term developing countries' deficit finance strategies. For any
while on the short-term the government bond yields investment, investors usually look at the level of returns for
were only affected by the inflation rate. The study also a given risk premium, in this case, investors who buy
established a positive relationship with inflation rates government bonds need to understand government bond
both in the long and short run. The FDI and Exchange yields and potential macroeconomic factors that might
Rate exhibited a positive significant impact on the long significantly affect the bond yield. This forms the
run only. Economic growth did not exhibit any long run foundation of this research.
and short run relationships at five percent significance
level. The research proposed that it is crucial to Macroeconomic variables aid investors by providing a
prioritize government bonds when developing both convenient method of obtaining more information about the
monetary and fiscal policies within the nation. stock market activity as well as determining how businesses
Additionally, it suggested that the government should perform (Syed Jamaludin et al., 2018). The effectiveness of
initiate an extensive awareness campaign regarding the security market is greatly influenced by macroeconomic
government bonds and their associated advantages as a factors. A research study conducted by Adiwibowo and
strategy to boost bond yields. The primary constraint of Sihombing (2020) on the macroeconomic factors affecting
the research was its focus on specific macroeconomic Indonesian government bonds yields revealed that a variety
variables and bonds. To address this, the study suggested of factors affect the yield on government bonds, including
the need for additional empirical investigations into how currency rates, bid-ask gaps, central bank (CB) rates,

IJISRT23OCT1684 www.ijisrt.com 1615

Volume 8, Issue 10, October – 2023 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
overnight rates, credit default swaps (CDS) discrepancies, understanding the determinants of these yields is crucial for
and oil prices. Security performance is vital for economic policymakers and investors.
growth, according to Olweny and Omondi (2011), since it
allows for better allocation of funds to productive economic Another study conducted by Ngaruiya and Njuguna
activity, thereby functioning as a barometer of an economy. (2016) focused on macroeconomic factors affecting bond
prices but did not include an analysis of government bond
Poghosyan (2014) discovered that the yield on yields. While bond prices and bond yields are related, they
government bonds has a long-term relationship with both represent different aspects of the bond market. Government
the proportion of public debt to GDP and country level of bond yields are specifically indicative of the return on
economic growth. In the medium term, sovereign bond rates investment for bondholders, while bond prices represent the
in many industrialized nations are projected to face market value of bonds. The failure to account for bond
significant upward pressure due to budget deficits and yields in their analysis limits the understanding of the
public debt (Kumar & Baldacci, 2010). In advanced factors directly impacting the returns and risks associated
economies such as the USA, long-term bond yields have a with government bonds.
positive correlation with government debt and the long-term
yield of US government-issued bonds, both short and long- These research endeavors would contribute to the
term (O'Hara & Zhou, 2021). As the amount of public debt existing body of knowledge, help policymakers in
rises, the price of government securities declines (Bernal et formulating effective fiscal policies, and assist investors in
al., 2016) in the long run, however, Zhou (2021) found a making informed decisions in the Kenyan bond market.
non-linear relationship between the level of government Through an analysis of the impacts of macroeconomic
debt and respective bond yields. factors on government bond yields in Kenya, this study aims
to close the knowledge gaps in this area while addressing
Government bond yields, according to Bhattacharyay recent economic developments such as the COVID-19.
(2013), are adversely correlated with changes in exchange
rate, GDP rate, rates of interest. Economic expansion, III. LITERATURE REVIEW
according to Senga (2018), is a factor in government bond
yields whereby higher rates of economic growth result in In a more recent study, Li and Huang (2021)
less government debt and fewer bond issuances; therefore, investigated the impact of exchange rate movements on
debt and bond issuance decline when the economy grows government bond yields in a sample of Asian economies.
faster. Their research provided additional insights into the
relationship by highlighting a significant positive
II. JUSTIFICATION OF THE STUDY association between exchange rate volatility and
government bond yields. They argued that increased
The existing studies on government bond yields in exchange rate volatility heightened the uncertainty and risk
Kenya have shown limitations and a lack of consideration perception among foreign investors, leading to higher yields
for unique conditions such as the impact of the COVID-19 on government bonds.
pandemic on the economy. These gaps in the literature call
for further research to address these limitations and provide Subhani and Panjwan, (2009) assessed the nexus
a more comprehensive understanding of the factors between the Consumer Price index (CPI) and bond yields of
influencing government bond yields in the Kenyan context. government in Pakistan between July 2001 and 2009
September utilizing secondary monthly data from the State
Nyaga (2014) and Balozi (2017) conducted research on Bank of Pakistan. The research applied an auto-regressive
factors affecting treasury bill uptake in Kenya. However, integrated, moving-average (ARIMA) model for data which
their studies did not directly analyze government bond was timeseries where the consumer price index was the
yields, which are essential in understanding borrowing costs independent variable, and the government bond yield was
for the government and investment decisions made by the dependent variable. The Autoregressive Integrated
individuals and institutions. Furthermore, the use of Moving Average (ARIMA) model results validated the
regression analysis in their research was problematic as it premise that the Consumer Price Index (CPI) had a
failed to capture the long- and short-term effects of the substantial relationship with the yield on government bonds.
macroeconomic variables under consideration. By not
accounting for these effects, the studies fell short in According to Ammer (2018) who conducted research
providing a complete understanding of the relationship that suggests a negative relationship between FDI and bond
between the macroeconomic variables and government bond yields on U.S. bond holdings by foreign investors from 31
yields. countries for the period 2003–2016. They found that a 1%
increase in FDI in a country is associated with a 0.02%
In a related study, Fredrick (2014) examined the decrease in its bond yield. This implies that higher levels of
influence of macroeconomic factors on the growth of the FDI might lead to reduced borrowing costs for the
bond market in sub-Saharan Africa but did not specifically government, as foreign investors' confidence in the country's
consider the impact on government bond yields. This economic prospects can drive demand for government bonds
omission is significant because government bond yields and lower their yields.
directly affect the cost of borrowing for the government, and

IJISRT23OCT1684 www.ijisrt.com 1616

Volume 8, Issue 10, October – 2023 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
Matei (2021) examined the consequences of the Covid- ΔBYt=β0+β1ΔBYt−1+β2ΔINFt-1+β3ΔFDIt-1+β4ΔEGt-
19 pandemic on sovereign yield spreads in European 1+β5ΔEXRt-1+ λECTt−1+ut………. (1)
Monetary Union countries concerning German government
bonds. Moreover, Matei (2021) also argued that economic Where ΔBYt is the first-differenced bond yield, ΔBYt−1
growth measured by the GDP exerted significant upward is the lagged value of the bond yield, ΔINFt-1, ΔFDIt-1, ΔEGt-
pressure on bond yields of government in the 14 European 1, ΔEXRt-1 are the lagged first-differenced values of; the
Monetary Union nations, throughout the specified time, both inflation rate, FDI, economic growth rate, and exchange
long-run and short-run. rate, respectively. ECTt−1 is the lagged error correction
term.β0 is the constant term, β1 is the coefficient for the
IV. DATA AND METHODOLOGY lagged first difference of bond yield. β2, β3, β4, β5 are the
coefficients for the first differences of the independent
This study was guided by a quantitative research variables (inflation rate, FDI, economic growth rate, and
design. The Causal-comparative research technique was exchange rate, respectively). λ is the coefficient for the error
used to enable the study to specify the type of relationship correction term. ut represents the error term.
between macroeconomic factors that is Inflation Rate,
Foreign Direct Investment, Exchange rate, and Economic  Diagnostic Tests
Growth Rate and how they affect the government bond The study carried out several diagnostic tests on the
yields. The study's goal was to statistically identify long and study variables. The Augmented Dickey-Fuller Test (ADF)
short run relationships between the independent variables was used to test for stationarity, the Johansen Cointegration
(macroeconomic factors) and dependent variables (bond Test was performed to test for the presence of cointegration
yields). The study covered the period of Q1 2007 to Q1 2023 amongst the independent and dependent variables. The lag
where the study utilized yearly secondary time series data to lengths selection was also performed using the Akaike
determine the impacts of inflation rate, Foreign Direct information criterion (AIC), Schwarz Criterion (SC) and
Investment, Exchange rate, and Economic Growth Rate on Hannan- Quinn Criterion (HQ) to select the optimal number
government bond yields listed on the Nairobi Securities of lags for the VECM model.
Exchange in Kenya. The Vector Error Correction Model
(VECM) was utilized to analyze the long run and short run  Empirical Results
relationships. The VECM had the advantage of utilizing
non-stationary time series data and allowed cointegrating  Stationarity Test
relationships. The guiding general model would be The Augmented Dickey-Fuller Test (ADF) was used to
described as: test for stationarity.

Table 1 Augmented Dickey-Fuller Test Results

At Level At Level First Difference First Difference
Variable (Intercept) (Trend & Intercept) (Intercept) (Trend & Intercept)
0.0723 0.0645 0.0329 0.0261
Bond Yield (Non-stationary) (Non-stationary) (Stationary) (Stationary)
0.8221 0.0937 0.0140 0.0027
Inflation Rate (Non-stationary) (Non-stationary) (Stationary) (Stationary)
0.72010 0.4390 0.0020 0.0119
Economic Growth Rate (Non-stationary) (Non-stationary) (Stationary) (Stationary)
0.0848 0.3172 0.0414 0.0244
Exchange Rate (Non-stationary) (Non-stationary) (Stationary) (Stationary)
0.0573 0.0819 0.0104 0.0071
Foreign Direct Investment (Non-stationary) (Non-stationary) (Stationary) (Stationary)

In summary, the ADF test outcomes underscore that all these variables were non-stationary at their initial levels, irrespective
of trend inclusion. However, they become stationary when differenced once, with or without a trend. This provides condition
provides a clear need for the implementation of the Vector Error Correction Model.

 Cointegration Test

Table 2 Johansen Cointegration Test Results

Unrestricted Cointegration Rank Test (Trace)
Hypothesized Trace 0.05
No. of CE(s) Eigenvalue Statistic Critical Value Prob.**
None * 0.782720 71.44130 69.81889 0.0369
At most 1 0.706434 47.01620 47.85613 0.0598
At most 2 0.627332 27.40573 29.79707 0.0921
At most 3 0.477295 11.61264 15.49471 0.1765

IJISRT23OCT1684 www.ijisrt.com 1617

Volume 8, Issue 10, October – 2023 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
At most 4 0.074158 1.232834 3.841466 0.2669
Trace test indicates 1 cointegrating eqn(s) at the 0.05 level
* denotes rejection of the hypothesis at the 0.05 level
**MacKinnon-Haug-Michelis (1999) p-values
Unrestricted Cointegration Rank Test (Maximum Eigenvalue)
Hypothesized Max-Eigen 0.05
No. of CE(s) Eigenvalue Statistic Critical Value Prob.**
None * 0.782720 24.42509 33.87687 0.0249
At most 1 0.706434 19.61047 27.58434 0.3687
At most 2 0.627332 15.79309 21.13162 0.2372
At most 3 0.477295 10.37981 14.26460 0.1882
At most 4 0.074158 1.232834 3.841466 0.2669
Max-eigenvalue test indicates 1 cointegration at the 0.05 level
* denotes rejection of the hypothesis at the 0.05 level
**MacKinnon-Haug-Michelis (1999) p-values

In summary, at a 5% significance level, both the Trace and Max-Eigenvalue tests indicated the existence of one cointegrating
equation among the series. This implies that the variables were not moving independently in the long run, but rather, they have a
stable, long-term relationship. The presence of Cointegration and the stationarity of the time series data upon first difference
necessitated the study to adopt a Vector Error Correction Model (VECM).

 Vector Error Correction Model Results

The VECM allowed for the analysis of the long and short run relationships by examining their coefficients.

 Long Run Relationships

Table 3 Summary of T-Statistic of Long Run coefficients.

Variable Coefficient T-statistic(t) Significance
ECONOMIC_GROWTH_RATE (-1) -0.101605 0.84091 Not Significant
FOREIGN_DIRECT_INVESTMENT_INFLOWS_USD(-1) 0.611780 -2.20402 Significant
INFLATION_RATE(-1) 0.179496 -2.36526 Significant
USD_EXCHANGE_RATE(-1) 0.035587 -2.17480 Significant

The derived equation becomes: in FDI are statistically significant and impact the dependent
variable in the long run at a 95% confidence level.
BOND_YIELDt-1= -3.519003 - 0.101605 *
ECONOMIC_GROWTH_RATEt-1 + 0.611780 * The coefficient of 0.179496 showed a positive
FOREIGN_DIRECT_INVESTMENT_INFLOWS_USD t-1 + association between the lagged inflation rate and bond yield
0.179496 * INFLATION_RATEt-1 + 0.035587 * in the long term. With a t-statistic of approximately -
USD_EXCHANGE_RATEt-1 2.36526, changes in the inflation rate are statistically
significant and are likely to affect the dependent variable in
The coefficient of -0.101605 signifies a negative the long run at a 95% confidence level. The coefficient of
relationship between lagged economic growth and bond 0.035587 signifies a positive relationship between the
yield in the long term. However, with a t-statistic of lagged US dollar exchange rate and bond yield over the long
approximately 0.84091, this relationship is not statistically term. The t-statistic for this coefficient, approximately -
significant at the 95% confidence level. The coefficient of 2.17480, exceeds the 1.96 threshold, indicating that changes
0.611780 demonstrates a positive connection between in the exchange rate are statistically significant and are
lagged FDI inflows and bond yield over an extended period. expected to have a notable impact on the dependent variable
The t-statistic for this coefficient, which is approximately - in the long run at a 95% confidence level.
2.20402, exceeds the 1.96 threshold, indicating that changes
 Short Run Relationships

Table 4 Summary of T-Statistic of Coefficients

Variable Coefficient Standard Error t-Statistic Significance (α = 0.05)
Error Correction Term -0.177673 0.04290 -4.14170 Significant
D(BOND_YIELD(-1)) 0.075835 0.07353 1.03137 Not Significant
D(ECONOMIC_GROWTH_RATE(-1)) -0.046970 0.06166 -0.76181 Not Significant
D(FOREIGN_DIRECT_INVESTMENT_INFLOWS_USD(-1)) -0.214591 0.19732 -1.08754 Not Significant
D(INFLATION_RATE(-1)) 0.156089 0.07985 1.96475 Significant
D(USD_EXCHANGE_RATE(-1)) -0.003305 0.01991 -0.16602 Not Significant

IJISRT23OCT1684 www.ijisrt.com 1618

Volume 8, Issue 10, October – 2023 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
The derived equation is: short run. This is likely due to the long-term nature of most
FDIs, leading to a long-run association with bond yields.
D(BOND_YIELD) = -0.177673*ECTt-1 + The coefficient of the exchange rate is statistically
0.075835D(BOND_YIELDt-1) - significant in the long run with a positive relationship with
0.046970*D(ECONOMIC_GROWTH_RATEt-1) - bond yield, while the short-term effects are insignificant.
0.214591* This suggests that yields react to long-term changes in
D(FOREIGN_DIRECT_INVESTMENT_INFLOWS_USDt- exchange rates rather than short-term fluctuations.
1) + 0.156089*D(INFLATION_RATE t-1) -
0.003305D(USD_EXCHANGE_RATE t-1) + 0.003539 Government bonds should be a focal point in the
development of monetary and fiscal policies. It is crucial
The analysis of the equation for bond yield reveals that interest rates associated with government-issued bonds
several key findings. First, the coefficient on the Error offer attractive yields to encourage higher citizen investment
Correction Term (ECT) is -0.177673, indicating that the in these bonds, while ensuring that such investments do not
system adjusts back to its long-run equilibrium at a rate of negatively impact other securities within the capital market.
17.76733% per year. This coefficient is not only substantial The government should launch an extensive awareness
in magnitude but also statistically significant, with a t- campaign about government bonds and their benefits. The
statistic of -4.14170, exceeding the 1.96 threshold for study also recommends foreign investors to utilize
significance at a 5% confidence level. However, the t- information from this research to better understand how
statistic for the lagged bond yield is 1.03137, falling below macroeconomic variables affect bond yields for efficient
the 1.96 threshold and suggesting that the lagged bond yield portfolio allocation.
may not be statistically significant at the 5% level of
significance in explaining the error correction. In practical REFERENCES
terms, this means that past values of bond yield might not
have a significant impact on the error correction mechanism. [1]. Adiwibowo, P., & Sihombing, P. (2020). Determinant
Likewise, the t-statistic for the lagged economic growth rate of government bond yields. Dinasti International
is -0.76181, indicating that this coefficient is not statistically Journal of Digital Business Management, 1(1), 86–99.
significant at the 5% level. This implies that the lagged https://doi.org/10.31933/dijdbm.v1i1.85
economic growth rate may not be a significant factor in [2]. Bernal, O., Girard, A., & Gnabo, J.-Y. (2016). The
explaining the error correction. The t-statistic for the lagged importance of conflicts of interest in attributing
foreign direct investment inflows in US dollars is -1.08754, sovereign credit ratings. International Review of Law
suggesting that this coefficient is not statistically significant and Economics, 47, 48–66.
at the 5% level. Consequently, the lagged foreign direct https://doi.org/10.1016/j.irle.2016.05.010
investment inflows may not be a significant contributor to [3]. Bhattacharyay, B. N. (2013). Determinants of bond
explaining the error correction. In contrast, the t-statistic for market development in Asia. Journal of Asian
the lagged inflation rate is 1.95475, exceeding the 1.96 Economics, 24, 124–137.
threshold and indicating that the coefficient for the lagged https://doi.org/10.1016/j.asieco.2012.11.002
inflation rate is statistically significant at the 5% level. This [4]. Chorafas, D. N. (1970, January 1). The management of
implies that the past values of inflation rate play a Bond Investments and trading of debt. Elsevier
significant role in explaining the error correction Monographs.
mechanism. The t-statistic for the lagged US dollar https://ideas.repec.org/b/eee/monogr/9780750667265.h
exchange is -0.16602, indicating that this coefficient is not tml
statistically significant at the 5% level, suggesting that past [5]. Chorafas, D. N. (2005). The management of Bond
values of the US dollar exchange rate may not significantly Investments and trading of debt. Elsevier Butterworth-
contribute to explaining the error correction. Heinemann.
[6]. Choudhry, M. (2006). Bonds: A concise guide for
V. SUMMARY, CONCLUSION AND investors. Palgrave Macmillan.
RECOMMENDATIONS [7]. Kumar, M., & Baldacci, E. (2010). Fiscal deficits,
public debt, and sovereign bond yields. IMF Working
The coefficient of inflation rate indicated a long-term Papers, 10(184), 1.
positive relationship with bond yield, and there was also a https://doi.org/10.5089/9781455202188.001
short-term association with the lagged inflation rate. These [8]. McKay, D. R., & Peters, D. A. (2019). Bond basics
coefficients positively influenced the yield in both the long and the yield curve. Plastic Surgery, 27(1), 83–85.
and short run, indicating that an increase in inflation results https://doi.org/10.1177/2292550319830964
in higher yields to compensate investors for inflation. [9]. Nyaga, E. (1970, January 1). Determinants of Treasury
Economic growth rate's coefficient was found to be bonds uptake in Kenya. UoN Digital Repository Home.
insignificant in both the long and short run relationships http://erepository.uonbi.ac.ke/handle/11295/94034?sho
with bond yield. The coefficient was negative for both long w=full
and short runs, possibly due to the economic downturn [10]. O’Hara, M., & Zhou, X. (Alex). (2021). Anatomy of a
caused by the COVID-19 pandemic. The coefficient of FDI liquidity crisis: Corporate bonds in the covid-19 crisis.
was significant in the long run, showing a positive Journal of Financial Economics, 142(1), 46–68.
relationship with bond yield, but it was insignificant in the https://doi.org/10.1016/j.jfineco.2021.05.052

IJISRT23OCT1684 www.ijisrt.com 1619

Volume 8, Issue 10, October – 2023 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
[11]. Poghosyan, T. (2014). Long-run and short-run
determinants of sovereign bond yields in advanced
economies. Economic Systems, 38(1), 100–114.
https://doi.org/10.1016/j.ecosys.2013.07.008
[12]. Senga, C., Cassimon, D., & Essers, D. (2018). Sub-
Saharan African eurobond yields: What really matters
beyond global factors? Review of Development
Finance, 8(1), 49–62.
https://doi.org/10.1016/j.rdf.2018.05.005
[13]. Syed Jamaludin, S. Z., Mahayuddin, S. A., & Hamid,
S. H. (2018). Challenges of integrating affordable and
sustainable housing in Malaysia. IOP Conference
Series: Earth and Environmental Science, 140,
012001. https://doi.org/10.1088/1755-
1315/140/1/012001
[14]. Ślusarczyk, B., Meyer, D. F., & Neethling, J. R.
(2020). An evaluation of the relationship between
government bond yields, exchange rates and other
monetary variables : The South African case. Journal
of Contemporary Management, 17(2), 523–549.
https://doi.org/10.35683/jcm20143.89

IJISRT23OCT1684 www.ijisrt.com 1620