There is a discussion of this already on these boards
See the thread:
Insulating House Paint
Since posting in that thread, I have found out that it technically may be possible to have a coating with enough additive, and put on a coating thick enough to have some sort of (extremely small) measured effect, it is absolutely impossible for a paint coating to have enough insulating properties to have an "R" value (how insulating properties are measured) of even "1", or even a significant portion of "1"
Please keep any more comments in the thread linked above and good luck