SELF-AFFINITY OF RECORDS OF FINANCIAL INDEXES
One-dimensional profiles of financial indexes can be characterized by the statistical fractal dimension D, estimated through the slopes of the log-log geostatistical semivariograms, assuming that the relationships between the average variances and the time increments are described by a power law. This assumption corresponds to the concept of self-affinity. The industrial Dow Jones index is described by a D=1.332±0.11 indicating the persistence of trends along analyzed business 500 days period, while for almost the same period the German DAX Composite index (D = 1.688±0.026) is dominated by short-range variations, and the British Footsie (D=1.495±0.009) and the Australian Share Price (all ordinaries) (D=1.506±0.008) indexes have a behavior like the fractal Brownian motion (D=1.5). The Nikkei Cash index is characterized by a D=1.474±0.008 which implies that the short-range variation is almost as important as the long-term variation. Ds were evaluated by using scaling arguments, and multifractality was not evidenced.