It wasn't a prediction, it was about fair value for the market. A 3% difference in one years earnings made an approximate difference of 30% in market value. The article wasn't predicting earnings either it stated that if earnings next years S&P earnings grew by 7% in the aggregate next year the model would show a fair value for the S&P 1050. If earnings grew by 10% in the aggregate next year the model would show a fair for the S&P of 1350. The focus wasn't on predicting the unpredictable it was regarding seemingly small changes in earnings expectations resulting in huge market moves. If we want to focus on something that is unpredictable we should focus on the future earnings prospects of individual companies and trying to predict what their earnings will be [img]/images/smile.gif[/img].