Following Mo et al. 2010 (page 479), the quantity $E(B-V)$, called the color excess between B and V bands, is equal to:
$$E(B-V) = A_B - A_V = (B - V) - (B - V)_0 = (m_B - m_V) - (M_B- M_V)$$
where $m_B, m_V$ are the apparent magnitude in B and V bands, and $M_B, M_V$ are the absolute magnitude in B and V bands.
Keep in mind the relation between apparent and absolute magnitudes, with $d=10$ pc:
$$m = M - 5 \log(d) + 5$$
If we are at redshift $z>0$, the color excess should be $E(B-V)>0$, because the extinction due to dust is greater in the B band than in the V band, and apparent magnitudes are greater than absolute magnitudes.
I'm wondering if it is correct to have $E(B-V)=0$ if the redshift is $z=0$ and $M=m$. The apparent magnitude is equal to the absolute magnitude and the relation above mentioned should become:
$$E(B-V) = A_B - A_V = (B - V) - (B - V)_0 = (m_B - m_V) - (M_B- M_V) = (M_B- M_V)_0 - (M_B- M_V)_0 = 0$$
Is that correct? Is it possible that we don't have color excess for galaxies at $z=0$, what am I missing?