⁵⁴The reciprocal logic would apply in the case of a shock that was known by the econometrician to be perfectly transitory, generating the same serial correlation in predictable consumption growth as in the case of the known-to-be-permanent shock. The only circumstance under which this serial correlation does not arise is when the econometrician has exactly the same beliefs as the consumer about the breakdown of the shock between transitory components. More precisely, it is still the case that the serial correlation coefficient on the predictable component of consumption growth is $(1 - Π)$ . But that predictable component itself is now zero, and $(1- Π )× 0 = 0$ .