سرمایه گذاری موجودی و مشخصات بخشی در برخی از کشورهای عضو سازمان همکاری و توسعه

This paper discusses the effects of sectoral structure on the long run macroeconomic inventory behaviour of national economies. Data on 15 OECD countries are included in the analysis, which is based on correlation and cluster analysis methodologies. The study is part of a long-term research project exploring factors influencing the inventory behaviour of national economies. First, we introduce some basic characteristics of macroeconomic inventory formation in the 15 OECD countries. We argue that our previous results on the existence of specific characteristic features of macroeconomic inventory investment are justified, hence it makes sense to study the factors influencing these features. We then examine the contribution of various sectors to the production of in the countries involved and the relationship between sectoral structure and inventory intensity (annual inventory change/Gross Value Added). We find that the high share of agriculture and manufacturing increases inventory intensity, that the increasing share of services has a negative effect and that the role of construction and trade is not obvious. The relatively low stability of the statistical results warns us to be cautious with our judgements. Further, case-by-case analysis would be required to obtain more solid results.

Our paper focuses on some macroeconomic characteristics of the inventory behaviour of national economies. This research field is largely unexplored. There is only one country in which it is adequately studied: the United States, from which a rich set of research results is available (for some recent studies, see Ehemann, 2004, Hirsch, 1996, Humphreys, 2001, Humphreys et al., 2001, Irvine, 2003a, Irvine, 2003b, Irvine, 2005, Irvine and Schuh, 2005 and Stern, 2001). There are very few country-specific studies (for exceptions, see Dimelis and Lyriotaki, 2007 and Guariglia and Mateut, 2010) and, as logically follows, there are very few cross-country or USA-EU comparisons (Dimelis, 2001 and West, 2002). It should be noted, however, that there are interesting and recurring debates about the macroeconomic roles of inventories, especially in connection with business cycles (see Blinder and Maccini, 1991, Lovell, 1994, Chikán et al., 1994, Dimelis, 2001 and Malgarini, 2007). Our interest in macroeconomic inventory behaviour dates back to the 1980s. At the First International Symposium on Inventories in 1980, a paper on the effect of the general state of the economy on inventory investment and structure was presented (Chikán, 1981). Since that time, we have largely focused on international comparisons of macroeconomic inventory behaviour. As in the paper by Chikán and Horváth, 1999, we compared trends in 88 countries to determine the connection between the inventory intensity (inventory investment/GDP) of the individual countries and various components of GDP. The current series of papers began in 2003 (Chikán and Tátrai, 2003), when we started to examine data from 14 OECD countries. (The data are probably the most reliable inventory data available). Since that time, we have used the annual inventory investment/GDP ratio to measure national inventory intensity. In the 2003 paper, we analysed the ratio’s relationship with various measures and the development, growth and fluctuation of GDP. In our research, we have used regression analysis and multivariable statistical methods. We have examined almost 20 hypothesis and found interesting results, such as the following: Inventories relative to GDP tend to decrease in developed countries after the early 1980s. • The inventory characteristics of developed countries converge; • however, no general regression model that uses components of GDP usage (fixed capital investment, exports, etc.) as independent variables can be found to describe the inventory behaviour of various countries. • Even though the tendency of the inventory/GDP ratio to decrease and converge appears to be valid across countries, the reasons for this tendency appear to vary not only by country but also by time period. In this paper, a new dimension of the factors that influence inventory investments is analysed: the sectoral structure of the economy. In recent years, we have conducted a series of examinations of macroeconomic inventory behaviour using multivariate statistical methods. Results of our research (the special characteristics of which are methodological: we consequently use multivariate statistical methods) have been reported in a series of papers presented at the biannual symposia of the International Society for Inventory Research (ISIR). For the latest paper in the series, with references to earlier papers, see Chikán and Kovács, 2009. In this paper, we continue in the same methodological tradition, but take a different perspective in our analysis: in earlier papers, we analysed national inventory behaviour in terms of output (GDP use), but we now turn to input for our analysis (GDP production). We take steps to understand the relationship between the sectoral structure of various economies and their inventory behaviour. As stated earlier, we used OECD data for the analysis—this is the relatively richest and most coherent international database for macroeconomic inventory analysis. Even this database has limitations, however, which restricted our options during the analysis. The database is accessible on the OECD website (www.oecd.org 〈www.oecd.org/〉). We believe we have obtained some interesting results even though several of our attempts to find relationships between various sectoral characteristics of the studied economies and their inventories have failed. Our main hypothesis is that the sectoral structure of economies and changes in this structure influence inventory investments. We test this hypothesis by examining the relationship between the sectoral structure and inventories from various angles.

The relationship between economic sectors and the ratio of inventory investment per Gross Value Added in developed countries is not at all straightforward. We made many various calculations that exposed some characteristics of that connection, but a number of questions remain unanswered. The hypothesis that there is a relationship between sectoral structure and inventory intensity can be accepted—it has been proven by our cluster analysis results and the correlation analysis as well. We can also say that higher the proportion of manufacturing and agriculture in the production of GVA, the higher the inventory intensity of the country. We cannot make similarly definite statements about the role of construction and trade. Within manufacturing, the chemical, mechanical, metallurgy and electronics industries are important: their major share leads to a relatively higher inventory level. The analysis also supports the otherwise trivial conclusion that the main factor lowering inventories in the long run is the increasing share of services. We did not find any significant correlation between annual changes in inventory investments and sectoral structure. This is not surprising because structural changes have longer-term features than the factors that change annual inventories. However, it is shown that, in the long run, sectoral changes do accompany changes in inventory behaviour. For more conclusive results, further research, including the examination of national case studies, is required.