Hi,

is it possible to plot the response to a *negative* 1 standard deviation shock? How?

Anton

Hi,

Yes it’s possible. You just have to define an auxiliary variable. For instance, if **epsilon** is your exogenous variable, you can create an endogenous variable **EPSILON** such that EPSILON = -epsilon.

With an approximation at order one, the response to a negative shock is minus the response to a positive shock.

Best,

Stéphane.

[quote=“nakov”]Hi,

is it possible to plot the response to a *negative* 1 standard deviation shock? How?

Anton[/quote]

Hi, thanks,

I was referring to 2nd order approximation in which responses to shocks can be asymmetric. But I guess a similar trick may be applied - the shock process can be redefined such that the innovation enters with a minus sign. In that case a positive 1 standard deviation of the innovation would be like a negative shock, am I right?

Yes, that is the idea

Best

Michel

Thanks a lot!

Anton

[quote=“MichelJuillard”]Yes, that is the idea

Best

Michel[/quote]