It's not just the bank that has been mis-forecasting these things. The economy has underperformed in these two respects in every economic model that I know of. When we ask ourselves why that is happening, it is of course our biggest research question. We look into things such as we discussed before, the animal spirits phenomenon, which is how much uncertainty needs to go away before a company will make its investment.
Given what we've been through over these past five years.... Many companies have disappeared; some 9,000 manufacturing firms have disappeared. The ones that survived may have downsized through the course of that period. We're asking when a company like that will be ready to re-expand to meet the new demands coming, let's say, from the United States or some foreign market.
The answer is they need to be more sure today, having been through all of that, than they needed to be five or ten years ago in a similar situation. This is that confidence thing, which is very hard to put your finger on, yet you know intuitively that it's true. You can talk to real people, and they'll tell you that it's true.
Our models don't capture things like that; it's as simple as that. On the export side, there are things now that, as I said, we are able to look at more deeply. We understand which sectors—we understand that they've had long-term problems maintaining their competitiveness—have lost market share in the U.S.
We can point to those and say that now we understand where it is, do we really understand what it is that is going to turn it around? That is what our model suggests will happen, but in the real world, it's real people making real decisions, so historical behaviour has not been a great guide to what we're seeing. And that is an excuse for how models work.