Some thoughts on exchange rate modeling in 2007