Testing the expectations hypothesis and explaining the determinants of term premia: evidence from the Indian money market

ABSTRACT

The pure expectations hypothesis states that the current yields on bonds with different maturities reflect investor expectations of future interest rates. Analysing the short-term inter-bank rates in a Vector Error Correction Model (VECM), the study could reject the pure as well as the general expectations theory in case of the 1 month and 3 months rates but not in case of 14 day rates. The term premia is found to be time-varying. The study attempts to quantify and decompose the term premia, inherent in the money market rates. The study uses the market spreads derived from the swap market, T-Bills, and CD markets to understand the level of decomposition of the term premia. A latent factor model was used to break down the term premia and decompose the same into credit and liquidity risk factors by using information from related money market instruments.

KEYWORDS:

• Expectations hypothesis
• long run restrictions
• term premia
• risk factors
• multiple partial least squares

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹ Financial Benchmarks India Pvt. Ltd – the company which has been set up for disseminating financial benchmarks in India. The computation methodologies are provided on the website https://fbil.org.in/content?p=2103&mq=d.

² The data is available from the authors on request.

³ A cointegrating vector between the 3 months Term Rate and the O/N rate could be estimated only in case of a model without a drift term.

⁴ In a split-sample validation, the different groups are composed of every n^th observation beginning with the first value, followed by every n^th observation beginning with the second, and so on. The value of n was chosen under various scenarios such as 60, 90, and 120 days. The results were consistent in all the scenarios.