Einige haben es vielleicht noch bei CF gesehen, hier noch ein Update, ist aber auch nicht mehr ganz aktuell aber trotzdem interessant:
MARKET INSTRUMENT FUNCTION
Editor,
I enjoyed the December 2010 S&Carticle on the market instrument functionby Alexander Ershov and AlekseyGerasimov.It appears that the function is not followingor predicting the Dow Jones orS&P as closely as in the past. Is this anormal expectation, or has the formulachanged?
Authors Aleksey Gerasimov andAlexander Ershov reply:
The formula for the market instrumentfunction stays the same. The problem isthat the events were delayed. This delayis 22 weeks.This is a normal situation. In 2008, thedelay was 35 weeks. With the method ofthe market instrument function, we candescribe only the long-term movementof the market. Some of the events mayhappen earlier; some, later. However,this circumstance does not change thegeneral form of the future graph.While using the market instrumentfunction method, it is necessary todetermine the point of the forecast thatcorresponds to the current date. As a rule,the forecast reproduces the market’s finestructure poorly. Thus, the positional relationshipof the curves is determined fromthe maximum of the correlation functionfor the forecasted and actual prices.
In Figure 1, you can see both the forecastedDIA weekly prices and the actualprices. The positional relationship of thecurves corresponds to the maximum of thecorrelation function as of April 2011.
In Figure 2, you can see the forecastedDIA weekly prices and the actual prices.The positional relationship of the curvescorresponds to the maximum of the correlationfunction as of November 2011.From Figure 2, you can see that the priceminimum in the market will happen notin August 2012 but in mid-autumn.
