There are several ways of doing price forecasting. Some people absolutely look at the long-run price of gas and then do an extrapolation of that. There are other people who use, basically, just a rule of thumb: gas is going to go up at 2%, following GDP.
The method we use is a very involved method of the interaction between the supply costs of the upstream supply, transportation tolls, and the change in those tolls based on flow volume, and of more importance is the demand side of the spectrum. How big is the market, or how big could the market go?
When I mentioned converting coal-fired plants to gas-fired plants, there's a huge variability in there in the sense that when you get somewhere above $4.50 per mcf, coal becomes another player, and it's very difficult to get them out of there and not have them come back on you.
LNG exports have a huge implication on the North American price, because if you can get it off, if you can bleed the supply off the basin, then the price will move upwards, which will enhance supply and all that kind of thing. We're not secure in our LNG markets yet. We have, as I mentioned, something around 22 projects that are on the books in North America. We only have one project under construction right now.