The term structure of the Yield Curve is one of the most essential tools in financial economics. All agents, both government and private firms, set them as a benchmark for implementing their policy and business decision, respectively. The yield curve can be one of the reasons for prosperous firms/countries or their declines (as evident in 2023)
This paper aims to understand the characteristics of Indonesia’s government bond yield curve, to forecast the curve for 2024 and to assess its future trajectories/volatilities. It's divided into two parts: dissecting and predicting the yield curve (Economic Bulletin – Issue 44) and exploring its interaction with various shocks (Forthcoming).
The importance of understanding the yield curve extends beyond its movement. The spread between long-term and short-term yields, especially the 10-year and 2-year government bonds, provides valuable insights into economic growth and potential recession risks, a concept widely monitored in financial economics.
This paper employs both Machine Learning and Parsimonious models (particularly Dynamic Nelson Siegel based models) to estimate Indonesia's 2024 yield curve.
After thorough examination and comparison, we found that Random Walk model is still ‘the king’ for short horizon (particularly 1 month horizon), while our models exhibit better performance in medium to long horizon (in line with Duffee (2002), Diebold and Li (2006) & Rubín and Ayliffe (2020))
Based on two of our best-performing models, Indonesia’s government bond for 2-year maturity will hover around 5.7% to 6.4%, 5-year maturity at ±6.5%, and 10-year maturity at 6.6% - 6.9%
Predictions for 2024 indicate two possible scenarios: a flattening curve akin to 2023's trend or a normalizing curve. These outcomes are influenced by critical factors such as Policy Rate and Exchange Rate, with the total government debt and its structure playing a significant role.
