I want to correct my earlier reply. After checking the description of the outlier detection process in PROC ARIMA doc, I realized that the model for level shift is as follows (continuing with the earlier notation): d12(d1(log(Y_t))) = beta*d12(d1(LS_t)) + N_t. That is, the level shift is differenced by the same differencing operator as the response variable. Note that beta*d12(d1(LS_t)) is zero everywhere except at NOV2008, where it is 1, and at NOV2009, where it is -1. This means that in the multiplicative form the model becomes
Y_t = Y_(t-1) * (Y_(t-12) / Y_(t-13)) * M_t * exp(N_t) where M_t is 1 everywhere except at NOV2008, where it is exp(beta), and at NOV2009, where it is exp(-beta). That is, the growth rate gets a jolt of exp(-0.09102) on NOV2008 and another one (in a reverse direction) of exp(0.09102) on NOV2009.
Sorry for this mistake.
... View more