Chaves and Arnott (2012, 2013) used regressions across 30 countries, spanning 60 years, to gauge the linkage of demographic profiles with stock market returns. We have reversed this analysis in order to find the dividend yield that corresponds to those demographic profiles. When we are interpolating within demographic profiles that have been seen before, we can measure the uncertainty of our forecasts. When we are extrapolating past relationships to apply them to demographic profiles that have never been seen before, we cannot. So, extrapolation into uncharted territory is a far more dangerous use of this kind of model. Figure 8 illustrates that the high yields of the 1970s and the low yields of the 2000s were arguably a normal result of demographic pressures.