Darshan Hiranandani : When to Use DSGE Models for Forecasting and How to Handle Structural Shifts

[darshanhiranandani23]

Hi all,

I’m Darshan Hiranandani, I’ve been studying DSGE models and their use in forecasting, and I’d love some guidance on when they are most effective. While these models are typically used for one-step-ahead forecasts with the assumption of a stable economic environment, I’m wondering about their performance under less stable conditions.

Here are my main questions:

1. Stability of Assumptions: DSGE models assume a stable economic structure, but what happens when this structure changes—whether due to a major shock or regime change? Does the forecast become unreliable, or can the model still provide useful insights?
2. Other Factors: Beyond structural changes, are there specific conditions where DSGE models aren’t suitable for forecasting, such as under extreme shocks, multiple disturbances, or in systems with non-linear behaviors?
3. Adjusting to Change: If the economy undergoes a significant change, how should we approach forecasting with DSGE models? Should we modify the model’s structure, or is it better to use other types of forecasting methods?

I’d love to hear your experiences and suggestions!

Thanks a lot!

Regards
Darshan Hiranandani

[jpfeifer]

As a first step, please see

When is it appropriate to use DSGE models for forecasting?

Forecasts in such cases will not be invalid, but rather may perform poorly relative to the actual data realizations. Fundamentally, forecasting with DSGE models assumes that the model solution is the data generating process. Whenever that assumption is invalid, the the forecasts will deviate from the data for other reasons than just unforeseen shocks. That may be because the model solution is linear but the DGP is not, because the model assumes no structural breaks but there are some in data, e…