It will have only one governing section modulus. Regardless of whether the load direction is negative or positive you will have the same location of maximum stress. Presumably this point is in the toes of the channel, since they're furthest from the neutral axis.
There will, however, be a different maximum stress at the tips of the toes and the face of the channel back due to the difference in distance to the neutral axis. You could make a section modulus to represent each of these if you need the stress at this point.
It's also possible that you could have different maximum moment resistances for different load directions if buckling in the individual elements is a failure mechanism. This seems like a really short element in which to have that issue, but I haven't actually checked. You could look at the width to thickness ratios of the individual elements I presume, or investigate them as individual plates with various support conditions if you're concerned.
