The nearest neighbour method as a test for detecting complex dynamics in financial series. An empirical application

In this paper the nearest neighbour forecasting method is applied to five series taken from the US Stock Market, with the aim of testing whether these display some kind of chaotic dynamics. The main result of this analysis is that the hypothesis may indeed be accepted for all five series. Furthermore, this paper has tried to test the usefulness of the method in terms of the quality of the predictions, with the results obtained being satisfactory in all cases.