That's a good question. That's very difficult to answer, and it varies quite a bit by mode, obviously LTL and intermodal and some of those modes where you have a greater component of fixed capacity, the level of excess capacity and what types of adjustments would vary quite a bit. But for our largest chunk of revenue on the truckload side, what our transportation leadership has talked about is that if you rewind the clock, 15, 18 months and say that capacity was generally balanced, there was probably fluctuations throughout 2008, but as you get to the end of 2008 when you start to see 17%, 18%, 19% industry drops in freight shipments and demand, you're starting out the year in 2009 with a pretty high degree of excess capacity. And throughout the year in 2009 we talked about that capacity correcting down, large carriers, parking equipment, Small carriers filing bankruptcy, and then prices adjusting downward on the shipper side. So as we get in to the second half of 2009, you start to see price stability on the shipper side. So you've got rates in the marketplace that have adjusted, capacity continuing to gradually correct down. We've heard estimates that make sense to us that I think you said it as well, that there's maybe 3% to 5% still excess capacity. The exact amount of excess capacity varies month-by-month and season-by-season with any year. So, normally January and February are seasonally a little bit slower where you would see kind of excess capacity. So it feels like to us today that there still is plenty of truckload capacity in the marketplace. The issue will become when you get in to the spring, the March and April timeframe, and demand starts to increase, how much of it's been inventory, how much of it is new sale, how much of the capacity leaves us in the February and March timeframe and doesn't renew their operating authority. So it's low to mid-single digits, but, you've got two variables that are pretty aggressive, and the freight demand side of it is much more volatile, and the fluctuation in the demand side or the freight side is probably what will really cause the inflexion point to change.