Strangely enough, all the models converged immediately. This suggests that while the variance between models and reality is large, it is not large enough to compromise the models, yet. This is what I meant when I said that the models are wrong in the right way. The Dow is almost reaching the 1929 divergence.
Visit econocasts.net for free sample trading model downloads.
I did my homework on the DJIA time series and the most recent day there was such a Z-score divergence between actual (2640.18) and predicted (2325.07) was on 1987.10.05.
ReplyDeleteThe divergence went to zero over the next 9 trading days, but on 1987.10.19 the Dow dropped the famous 508 points.
Moving back in time as far as actual vs predicted divergence...we're now at 1938.07.19 when the DJIA was at 143.67 and the predicted value was 125.86. By 1938.09.26 the DJIA was down to 129.91.
ReplyDelete