Sample Size: $T$
T represents the sample size or the number of observations in the dataset.
Rotation angle: $\phi_0$
In the bivariate SVAR $u_t = B_0(\phi_0) \epsilon_t$, the reduced-form shocks $u_t$ are a rotation of the structural shocks $\epsilon_t$ by angle $\phi_0$ (in radians).
Rotation angle: $\phi$
In the bivariate SVAR $u_t = B_0(\phi_0) \epsilon_t$, the reduced-form shocks $u_t$ are a rotation of the structural shocks $\epsilon_t$ by an angle of $\phi_0$ (in radians), i.e., $\phi_0 \cdot (180/\pi)$ degrees. On this page, the true rotation is fixed at $\phi_0 = 0.5$. The rotation $\phi$ allows us to compute the innovations as $e_t(B(\phi)) = B(\phi)^{-1} u_t$, which corresponds to rotating the reduced-form shocks by $\phi \cdot (180/\pi)$ degrees.
The identification problem
This page explores the central challenge in SVAR analysis: the identification problem. We'll discuss why identifying structural shocks from reduced-form shocks is difficult and how non-Gaussian properties can help.
Structural Shocks and Innovations
Let's fix the rotation at $\phi_0=0.5$ such that the reduced form shocks are equal to one specific rotation of the structural shocks:
However, in a given application, we neither know the rotation matrix $B(\phi_0)$ nor do we observe the structural shocks $\epsilon_t$. Instead, we only observe the reduced-form shocks $u_t$. Consequently, the question is how can we identify $B(\phi_0)$ and recover the structural shocks $\epsilon_t$ from the reduced-form shocks $u_t$?
We can choose any angle $\phi$ and compute $B(\phi)$:
Interactive Simulation
Move the $\phi$ slider above to see how $B(\phi)$ leads to different innovations $e_t$ equal to rotations of $u_t$:
The question is: how do we know whether we picked $\phi=\phi_0$ and whether our innovations are equal to the structural shocks?
Next Steps
In the following sections, we'll explore two complementary approaches to solving the identification problem using non-Gaussian properties:
- Minimizing the dependencies between the recovered shocks.
- Maximizing the non-Gaussianity of the recovered shocks.
Proceed to Minimizing Dependencies to learn about the first approach.