Construction of priors using the data

Estimation
1

Hello everyone

I am trying to estimate a model, I know that the construction of the prior distribution must be done without taking into account the data, however, when I estimate my model, my estimated calvo parameter (\theta) is 0.9870, which implies that prices are updated every 73 quarters, which does not seem economically plausible. The prior I specified was a mean of 0.5 with a standard deviation of 0.05. In that case, is it possible or acceptable to modify the prior to obtain plausible values for \theta ? If I modify the standard deviation to 0.035 I get more acceptable values, but at the end, that is using the data to construct the priors

Greetings

2

Philosophically, that is indeed incorrect. That being said, it is quite usual to adjust the prior to be tighter if the posterior still looks implausible. But in that specific case, you may want to find out what is going on in your model. It seems your initial prior was already rather tight. For the posterior to still put such a big mass on prices being completely rigid suggests that there is something strange going on.