Test on yields of equivalently-rated bonds

Purpose
This paper aims to report practice-relevant anomalous investment yield behavior of two types of bonds – Type A, the mainstream bond, and Type B, which is Sukuk – both having similar cash-flow-relevant characteristics.
Design/methodology/approach
Bond valuation theory suggests that yields to investors of similarly rated bonds ought to be same. The authors collected time-series data on A and B bonds, all being coupon-paying bonds with similar rating and similar tenor as two matched samples traded in a bond exchange. To ensure the results are extended to different bond sectors, the data set was separated into treasury bonds as risk-free and corporate bonds as risky ones. The data set was further sub-divided into short-, medium- and long-tenor bonds. As the data straddle the Global Financial Crisis period, the authors use appropriate econometric method to control the possible effect from the crisis.
Findings
The average and median yields on Type A bond are significantly different from those of Type B. The test results show significant and systematic differences: treasury bonds of Type A returns yield lower than treasury bonds of Type B; the yields of corporate mainstream bonds (A) are higher than the yields of Sukuk (B). The authors observe these findings constitute a puzzle, being anomalous to theory.
Originality/value
This paper is original in that it is documenting significant differences in pricing of equivalent bonds. This has both theory and practice implications for fixed-income security market practices. The evidence is very strong to suggest that the identical types of bonds may have missing variable that contributes to the difference. Therefore, further research to identify the missing variable is necessary.