I guess hence your qualifier "whether it's likely". I'd find that even less likely than a revolution, but I have no idea what the real actual statistics are. Would be interesting to see data on coups vs resignations vs stayed in power until death or retirement.