Forecasting The Term Structure Of Government Bonds In Kenya

Abstract

n the last two decades, modelling the term structure of interest rates and using the same to re ect the monetary policy framework in the greater scal environment has attracted interest among nancial markets players across the globe. Their interest has been the economy, but broken down, it has been the di erent components in the nancial market that include bonds, bills, mortgages and other forms of securities traded in the market. As such, researchers have embarked on a serious review of the yield curve to model the term structure of interest rates, due to its predictive power and forecasting capabilities of future economic states in an economy. The yield curve has been used over the years to predict and forecast among other economic factors; in ations rates at di erent periods, real economic output, future possibility of periods of recession, and also potential growth. The yield curve has been used in developed, emerging, and developing economies successfully. However, it has been employed in developed and emerging nancial markets than it has been in developing markets. Research shows that the yield curve has been used to model interest rates in American markets; including Colombian and Brazilian market, as well as emerging nancial markets like South Africa and India. Although the concept of forecasting is best discussed from an actuarial perspective, the tenents of Finance and Mathematical Statistics provide very critical grounds for which to begin and build on. Various models have been proposed, especially in the developed and emerging markets. However, few research work has been done in the developing markets that include the Kenyan nancial market. Over time, model-based asset pricing and its evolution over the years to bring in modern day dynamics in nancial markets has become extremely important. As such, nancial institutions o ering xed nancial instruments have relied on estimations, modelling and forecasting to build scenarios and come up with plausible and workable solutions to the uncertain environments they operate in, and the anticipation of changes in future. As such, market players have opted to model term structures of interest rates in predicting yields of nancial instruments/assets in the nancial markets. The models that have been brought forward in the Kenyan context include the no-arbitrage and equilibrium approaches. However,this study proposes a unique model; Nelson-Siegel (1987). Further, this study considers a reparameterized form of the model called by Diebold 3 and Li (2006) which is an advancement of the Nelson-Siegel model. This is the model to be employed to t Kenyan data. Additionally, the study considers the same model but include time varying parameters of regression type. To the best of my knowledge, this combination of models bringing in time varying parameters is one of its kind and the rst to be conducted in the Kenyan scenario. The models of regression type will then be used for forecasting and the results maybe compared with benchmarking models that successfully worked in other application across the globe. The term structure of interest rates is best demonstrated in yield curves. A yield curve basically represents the relationship between yields of a nancial instrument (a bond in our case) and their term, basically described as the maturity. The term structure basically shows the behavior of interest rates in the short-term, medium-term and the long-term. It is usually a plot of interest rates of bonds against maturity in months or years. 1.1 Concept of Bonds Gwalani (2015) de nes a bond as a debt security instrument that is issued by companies or government with an aim of raising money. A coupon bond pays regular instalments and the principal on maturity date. Das, Ericsson and Kalimipalli (2012), notes that bonds are means of raising funds for the government. Government bonds are instruments of nancing a government de cit Ndung’u (2013). Bonds are tools of acquiring loans from individuals and institutions. Bonds are a crucial means of getting funds by the governments. These bonds range from short term to long term bonds depending on the purpose of the bond issuance. Kabua (2011) noted that the rst world countries such USA, and European Countries have the best and most complex bond markets. Jaramillo and Weber (2013) avers that the bond statistics indicates that the world bond market statistics is dominated by the developed countries. For instance America occupied large share of the pie at about 40% of the world value of outstanding domestic bonds; its market comprises mortgage-backed securities, Federal Agency Securities, Corporate Bonds and treasury bills (Myers, 2014). In fact most of Africa’s debt is in foreign currency and is therefore considered as not tradable, and very few countries in Africa have a viable bond market. Particularly, for most African countries the bond market is insigni cant or non-existent, even though Africa has some of the most heavily indebted countries in the world (Yeboah, 2014). Sub-Saharan African countries in developing local bond markets include the di erent debt structures and level of market infrastructure. Kenyan government bonds are traded in the Nairobi security exchange market and have maturities ranging from 1-30 years. The continued growth of the bond market is an indicator of the critical role played by the bond market in raising capital for corporates and funds for the government.