The importance of jump clustering is widely recognized in the financial market. We use the Hawkes process to capture jump propagation risks and study their role in modeling the Volatility Index (VIX) term structure. Applying the joint estimation approach to the S&P 500 index and the VIX term structure, we find that incorporating jump propagation risks is important to reconcile the dynamics of joint data and to model the VIX term structure, especially when the market is highly volatile. The long‐term variance factor further improves the description of the VIX term structure for low and medium market volatility levels.