End of period mean in Holt - Winter exponential smoothing

[Quynhnt90]

When I do exponential smoothing a series using Holt method, I see that the last smoothed value is different from what called the "end of period mean".
In EVIEWS, The next forecasted values are actually equal to (end of period mean + h*end of period trend)
The method that I learn provides the formula to calculate forecasted value like this (last smoothed value + h* last value of trend)
So, why is there the difference between the two values in EVIEWS? What actually is the "end of period mean"?