Automatic Detection of Common Long-Term Monetary Policies on Global Exchange Market Using Gabor Analytic Phase Binary Encoder

An application of Signal Processing to time series data analysis is presented here. Gabor Analytic Phase Binary Encoder (GAPBE) is used to retrieve the phase information from the long time series of currency exchange rates. As a result, a binary code is generated for each currency. Then the (dis)similarity between the local trends of any two different currencies is expressed through Hamming distance. A common policy for two different currencies is found when the Hamming distance between the binary codes representing the two given currencies is sufficiently small, but exact interpretation of the similarity scores obtained is left to the specialists. And this is all from the economist’s point of view. On the other hand, those interested in iris recognition should read this paper because the same procedure can be used to encode images of human iris. In this case it is worth to give an example of cheating iris recognition by finding two different irides having too similar iris codes. From this perspective, we show here that the exchange market provide us with such an counter-example in which two different exchange rate variation curves are encoded as too similar binary codes.