A Multiproduct Model Decomposition of Components of Russian Gross Domestic Product

<p>The paper proposes a procedure for multiproduct model decomposition of macroeconomic statistics with the use of Russian GDP data and its components by expenditure. The problem of preparation of the data for the further usage in macroeconomic modeling is thoroughly investigated, and the main step here is seasonal adjustment. A set of axioms, describing the necessary properties of a seasonal adjustment procedure, is described. The impossibility of simultaneous fulfillment of additivity and invariance to deflation requirements is shown, therefore, one of these properties is to be chosen and the decision is determined by the characteristics of the problem solved. The most popular today seasonal adjustment procedures do not meet the invariance to deflation requirement, so an alternative procedure is proposed. Its properties are investigated and compared with the properties of other procedures.</p><p>&nbsp;A method of multiproduct decomposition is described. In contrast to preceding decomposition procedures, it does not link one of the model products to export or import, so the solution can searched for in a wider class and the restrictions on export and import deflators can be removed. A theoretical background is provided and the methodology of computation on the real data is described, as well as the computation results based on Russian macroeconomic data. The resulting two-product decomposition is not just an intermediate step in the macroeconomic models estimation. It also provides some interesting and valuable results concerning the behavior of macroeconomic agents. The paper proposes a method of decomposition of the change of stocks, which could not be decomposed by earlier versions of decomposition procedure. A three-product decomposition is proposed to demonstrate the flexibility of the procedure.&nbsp;</p>