Well, if you trust the poll here, the people who forecast a 25bp hike outnumber those who forecast a 50bp hike by the rate of 77:9. If this is indeed the market consensus, it probably means that the market will rise slightly in case of the actual 25bp hike and drop substantially in case of a 50bp rise. However, the former outcome 8.55 times more likely than the latter (77 / 9 = 8.55). Futhermore, if the 50bp rise does occur, the subsequent drop will not be 8.55 times bigger than the rise in the event of a 25bp rise. Therefore, the risk/reward odds are on the long side (with a tight stop). Well, if only the markets were that rational as my analysis, and if I had the guts and discipline to follow it, I'd probably be a good trader.