24.2The Exchange Rate Many of the effects of the mining boom estimated by AUS-M reflect changes in the exchange rate. However, estimating exchange rate responses is difficult. Much of the variation in the data seems to be noise. And systematic responses to macroeconomic variables largely reflect changes in expectations, which are not observable. This makes estimated time series correlations difficult to interpret. In AUS-M the exchange rate is assumed to gradually move to a level that reconciles the trade balance with the savings and investment decisions of households, business and government. This means it increases with the terms of trade, and expected rates of appreciation match interest differentials. These effects are calibrated to be consistent with common views, available research and the long-run equilibrium implicit in the model. They are discussed further in Douglas, Thompson and Downes (1997). Because of its central importance, we decompose the exchange rate change into responses to the different elements of our counterfactual in Figure 5. As can be seen, the exchange rate responds in somewhat similar amounts to the slowing in world growth, the further reduction in minerals prices and the zeroing out of investment residuals. The latter effect reflects the need to offset the long-run effect 2 Our estimate of the increase in income arising from the mining boom is greater than that of Edwards (2014). One important reason for the difference is that we consider the trading gain arising from the higher terms of trade to be an increase in real domestic income. Our estimates of the growth in the volume of GDP encompass the effects of this increased income being spent in line with historical correlations, after allowing for the effects of taxes, net income transfers and so on.