The efficiency of treasury bills market in Kenya

Abstract

This study is purposed to test the efficiency of the Kenyan Treasury Bills Market. It tests how accurate forward rates are in predicting the expected spot rates and is founded on the unbiased expectations theory. The data used consist of weekly yields on 91-day and I 82-day T-Bills over the six year period from is February 2002 to 17th March 2008. Using first test of ANOYA we determine whether the 91-day and 182-day Treasury bills are different. We again use ANOYA for 91- Days TB Lag and Forward Rate Lag to test whether the forward rate is equal to the expected spot rate. Finally we run the regression model to find out the change in future spot rate when forward rate changes by 1. This helps determine whether the relationship between future spot rate and forward rate is statistically significant. We find that 91-day and 182 day T-Bills appear different in line with the theory that assets of longer maturity tend to give higher returns as compensation. We also find that forward rate tends to be higher than the comparable spot rate suggesting the existence of forward premiums. The regression co-efficient, P-value of 0.000, show that the relationship IS statistically significant i.e. you can use forward rates to predict future spot rates. The implication is that market players would not achieve much trying to predict future spot rates using the forward rates alone. The CBK should develop a model that incorporates forward rate and other macro-economic factors to predict more accurately the future spot rate; as we find that the forward rates have incremental information for the future changes in the spot exchange rates, given that they move towards the same direction. This would guide investors in their decision to invest in the Kenya Treasury Bills. This study will pay attention to the cross-interest rates and cross-maturity term structures of the forward premium to assess if at all they contain information that can be useful in predicting future spot interest rates. A forward interest rate is the rate one can lock in now for a commitment to buy a one-period bond in the future. This leads naturally to the hypothesis that forward rates forecast future spot (one period) interest rates. Early tests of this hypothesis largely use US Treasury bills, and the results are rather negative. A market is efficient if nobody can obtain extraordinary profit in the long run by using publicly available information (Fama, 1965). Gross (1983) summarize the conditions to be fulfilled if the market is to be efficient: the market is competitive, information is costless to acquire; the market participants have the capacity to effectively use the information; transaction costs is zero; and no non-random innovation between contract time and actual delivery.