ALTERNATING SCHEME FOR METHOD OF MOMENTS

Abstract

In the financial market, there is always an unexpected issue between measures of dif ferent obligations, stocks, currency. Big financial companies before doing investments are highly interested in exploring the behavior of a certain market. For such analysis, we use different methods which are calling dimension reduction techniques. This work adopted the principal component analysis and maximum mean discrepancy distance to assess ten different bond yields by calculating their changes. In the beginning, we will explain in detail the nature of our data and show some results from the theorem about the Wiener process. After we will apply the classic method and our new (al ternating to PCA) method. In the end, we will compare graphs of each method and conclude the effectiveness of Maximum Mean Discrepancy distance