Restrictions on risk pricing in dynamic term structure models help connect interest rate changes over time and across different rates. A new method using Bayesian techniques was developed to estimate these models, showing good performance in simulations. The data suggests that only level risk affects prices, and changes in slope impact term premia. By incorporating these restrictions, the model better reflects real-world expectations of future short rates and term premia. This helps explain why short-rate expectations have been surprisingly stable in the past. The models also indicate that a significant portion of the decline in long-term interest rates is due to expectations of future short rates.