摘要：Bond is one of commercial instrument that influence economic sector in Indonesia. Bond transaction can't be made in the market directly, but it has been traded through securities. On average, there are only few bond transactions with the various market prices. Benchmarking is, therefore, needed to be created to determining bond price through yield curve. Through yield curve, the relation between yield of bond with same credit risk (rating) and different time to maturity can be seen. This research is conducted by employing time to maturity to model the yield of some selected corporate bonds with rating of AA and A. Two methods, Nelson Siegel Svensson (NSS) method couple with Levenberg-Marquardt optimization and Cubic Spline Smoothing (CSS) are employed here. These two methods have been applied to data from Indonesian Stock Exchange (IDX) ranging September to November 2008. The results show that CSS give smallest RMSE and MAE. In contrast, Nelson Siegel Svensson reports a model which more parsimony, more easily to be explained, and more adaptable to keep upfoward maturity than CSS. This research takes into account that NSS is better to be to model corporate bonds yield curve than CSS. Another importance conclusion that can be gather is that corporate bonds (with rating of AA and A) yield are hang about under IGSYC in certain period. Its means, that corporate bond market in Indonesia is not good for investment comparing to government bond.
