2011 forecast, page-11

I think it is invalid to compare it to picking individual horses, when the proposal was around an index, which is akin to a field of horses which race at some track in a given period.

The consideration would be closer to whether the same field at some other period, as a population, will go slower or faster at various pre-determined points on the track, based on a sample of results for an initial segment of a hypothetical time cycle.