Price discovery and volatility spillovers in the interest rate derivatives market

The interest rate derivatives market plays a crucial role in fostering bond market development and managing interest rate risk. In China, this market is rapidly expanding, with instruments like interest rate swaps and treasury bond futures becoming increasingly prominent. This study addresses the critical question of price discovery efficiency and volatility spillover within this evolving market. Effective price discovery, the process of incorporating new information into security prices, is essential for informed investment decisions. Analyzing the volatility spillover structure provides insights into asset pricing and portfolio management. Focusing on treasury bond futures and interest rate swaps—the most prevalent instruments in China's interest rate derivatives market—this paper empirically examines their price discovery function and their impact on the treasury bond spot market. The research contributes to the existing literature by analyzing the price discovery efficiency of China's interest rate derivatives market and exploring the volatility spillover structure among treasury bond futures, interest rate swaps, and the treasury bond spot market. This study's findings are particularly relevant given the relatively nascent stage of China's interest rate derivatives market, offering valuable insights for other emerging markets.

Existing research on price discovery often focuses on the transaction cost hypothesis, suggesting that markets with lower transaction costs, such as futures markets, tend to lead in price discovery. Several studies have confirmed the leading role of treasury bond futures markets in price discovery, demonstrating that futures prices often precede spot prices. Research on interest rate swaps has largely centered on pricing models. Regarding volatility spillovers, various methodologies, including bivariate heterogeneous autoregressive models, EGARCH techniques, and VAR-GARCH models, have been employed to analyze the transmission of volatility between different markets. The Diebold and Yilmaz (DY) spillover index model, which measures spillover relationships in multiple markets, has gained popularity. While some research exists on the relationship between government bond spot, futures, and interest rate swap markets in China, a comprehensive analysis of price discovery and volatility spillover incorporating both treasury bond futures and interest rate swaps remains limited. This study bridges this gap by investigating these dynamics in China's context.

This paper employs two primary models for empirical analysis: the information share (IS) model and the Diebold and Yilmaz (DY) spillover index model. Data includes daily closing prices of five-year treasury bond futures (TF), five-year interest rate swaps (IRS) based on FR007, and the China Securities Index aggregate bond index representing the spot market (S). The data period spans January 4, 2016, to December 31, 2020. To mitigate heteroscedasticity, logarithmic transformations were applied to the price series (InF, InIS, InS), and logarithmic returns were calculated (RF, RIS, RS). **Information Share Model:** A vector error correction (VEC) model is first established to capture the long-term equilibrium relationships between the markets. This model is transformed into a vector moving average (VMA) form to calculate information shares, representing the contribution of each market to price discovery. The study considers both cases with and without correlation between innovation items, using the Cholesky decomposition method to obtain upper and lower bounds for information shares. The average of these bounds is used as an estimate of the information share. The component share (CS) model, also based on the VEC model, is used for further analysis of price discovery contributions. **Spillover Index Model:** A vector autoregressive (VAR) model is employed to analyze volatility spillovers. To address the issue of variable ordering dependence in variance decomposition, the generalized VAR framework of Koop, Pesaran, and Potter (KPPS) is adopted. The H-step-ahead forecast error variance decomposition (FEVD) is used to calculate the spillover index. The study calculates the total spillover index, directional spillover indices, and net spillover indices to comprehensively understand the volatility transmission between markets. A rolling window approach (100-day rolling sample) is applied to assess the dynamic changes in volatility spillovers over time. In addition to the primary models, the study also conducts a Phillips-Perron (PP) test for stationarity, a CUSUM test for parameter stability, a Bai-Perron multiple breakpoint test for structural breaks, and a Johansen cointegration test to assess long-term relationships between variables.

**Price Discovery:** The information share model's full-sample results indicate that treasury bond futures and interest rate swaps exhibit stronger price discovery capabilities than the spot market. This finding is consistent across both IS and CS models. However, when the sample is segmented based on three identified structural breaks in the spot market (October 25, 2016; January 22, 2018; and April 3, 2020), the price discovery contributions of derivatives display dynamic changes. In certain periods, the spot market surpasses the derivatives markets in price discovery. **Volatility Spillover:** The static spillover analysis reveals a total spillover index of 49%, suggesting significant volatility transmission among the three markets. Treasury bond futures exhibit the highest spillover to other markets (51.1%), while the spot market receives the largest spillover from other markets (54.5%). The net spillover analysis shows that treasury bond futures have a net positive spillover effect, while the spot market experiences a net negative spillover effect. Dynamic spillover analysis using a 100-day rolling window shows an upward trend in the total spillover index over time, indicating a strengthening linkage between the markets. The directional spillover indices reveal that treasury bond futures and interest rate swaps increasingly impact other markets, particularly the spot market, indicating that the influence of the derivatives market has strengthened.

The findings suggest that while China's interest rate derivatives markets generally contribute significantly to price discovery, this contribution is not consistently dominant across all periods. The dynamic nature of price discovery, as revealed by the structural break analysis, underscores the impact of market events and policy changes. The significant volatility spillovers from both treasury bond futures and interest rate swaps to the spot market highlight the interconnectedness of these markets and the potential for risk contagion. The consistently net positive spillover from treasury bond futures indicates that this market acts as a significant driver of volatility in the overall interest rate landscape. The results generally support the notion that a well-developed derivatives market contributes to market efficiency and price discovery, but also highlights the need to consider the dynamic interactions and potential for risk transmission.

This paper demonstrates that treasury bond futures and interest rate swaps play a vital role in price discovery within China's interest rate market, though their influence is dynamic. The significant volatility spillover from derivatives to the spot market underscores market interconnectivity. Policy implications include promoting synergy between treasury bond futures and interest rate swaps through product diversification and enhanced regulatory oversight to mitigate systemic risk. Future research could incorporate macroeconomic factors and a wider range of interest rate derivatives to provide a more comprehensive understanding.

The study's limitations include the focus on only five-year treasury bond futures and the exclusion of macroeconomic factors. Future research could expand the analysis by including various maturities of treasury bond futures and incorporating macroeconomic variables to gain a more complete understanding of the market dynamics.