2014. vol. 18. No. 3

The primary aim of this research is to compare diverse statistical models to predict critical financial state for Russian private small and medium-sized companies belonging to different sectors of economy.

We use the following methods: Linear Discriminant Analysis, Quadratic Discriminant Analysis, Mixture Discriminant Analysis, Logistic Regression, Probit Regression, Tree and Random Forest. Our dataset consists of approximately 1000000 observations from the Ruslana database and covers the period from 2011 to 2012.

Instead of standard definition of default we use the notion of critical financial state which means that we add companies liquidated as a result of legal bankruptcy to those liquidated voluntary.

We study four industries in detail: construction, manufacturing, real estate activities, retail and wholesale trade. Comparing industries, we come up to several compelling conclusions. On the one hand, the difference between sectors is so significant that it cannot be overcome by including several dummy variables but by estimating separate models for each industry.

On the other hand, sectors are similar in several ways. Firstly, importance ranking of regressors is stable among sectors that are analysed. This results in unique optimal set of variables chosen out of six possible alternatives. To add, inclusion of non-financial characteristics impro­ves predictive power greatly. While age of a company and federal region are the key non-finan­cial variables, size of a company is less important, and legal form is the weakest predictor. Secondly, Random Forest outperformed other statistical approaches on all data sets. For this method area under ROC-curve (the applied comparison criterion) reaches up to 3/4 which is the same for all industries.

This research will be of vital importance especially to banks and other credit organisations providing loans to small and medium businesses as well as to state regulators.

Non-financial matters in the operation of an enterprise that influence decision-making process are attracting more and more attention in the contemporary research. One of the most commonly used frameworks is the theory of financial architecture. In a great set of various facets of a financial architecture, most attention is paid to the ownership structure and the composition of the Board of Directors. It has been found that the diversification of the Board (including gender diversification) provides a broader outlook on the development strategy, raises company’s reputation and investment attractiveness.

The main research objective of this paper is to analyze nonfinancial aspects of corporate economic effectiveness taking into account temporal effects and idiosyncrasies of particular companies, as well as non-observable industry-specific and geographical characteristics. The data sample is a panel of Western European companies over a period from 2007 to 2012 based on Bloomberg and Amadeus databases. The multilayer structure of the models allows to control for non-observable variables (causing a lack of uniformity among different companies, countries, industries and time periods). This produces more exact estimates of the effects caused by the factors in question – the number and share of women on the Board, the ownership share of the biggest shareholder and the total share of the three largest shareholders. The particular specification of the model along with controlling for the inhomogeneity of coefficients settles the apparent controversy between theoretical studies and preliminary empirical results. It turns out that the increase of the number (and fraction) of women on the Board of Directors leads to an increase in the effectiveness of the company only up to a certain threshold; further increase tends to decrease the effectiveness. This phenomenon has been observed in most of the specifications of the model and has been controlled for a potential endogeneity by using lagged values instead of current ones. The analysis of marginal effects of the number of women on the Board on the effectiveness of the company shows a negative dependence on the capital stock and financial leverage and a positive dependence on the research and development costs. The dependence on the size of the company has come out inconsistent: the estimates based on the OLS model show constant returns, the estimates based on the hierarchically structured model (controlling for dissimilarities between countries) show increasing returns, and the estimates accounting for the sector dissimilarities show decreasing returns.

Most of the different specifications of the model show no significant impact of the concentration of capital on the strategic effectiveness of companies.

The paper analyses the differences between firms with centralized model of decision-making and firms with decentralized model of decision-making, and studies the factors that affect choice of the decision-making model at the firm. The empirical analysis is based on the large survey ‘European Firms in a Global Economy’ conducted in 2010 in 7 European countries: Germany, United Kingdom, France, Austria, Spain, Italy and Hungary.

It is shown that decentralized firms demonstrate better performance in a spectrum of different areas: they are more involved in the international trade (in particular, in the more complicated forms of international trade), innovative activities, R&D, training of employees. It is also shown that probability of centralization of firm management is significantly influenced by factors related to the distribution of property, as well as by individual characteristics of the CEO. In particular, concentration of property in the hands of individuals or families, belonging of the CEO to the family that controls the firm lead to a considerable increase in the probability of centralization. These results can reflect influence of personal (behavioral) motives of the firm owners or top managers on the choice of the governance model at the firm.

The results hold not only at the whole sample of the survey, but also for each of the countries in the sample. The influence of the features of property distribution and individual characteristics of the CEO on the probability of governance centralization at the firm is strong. For example, according to the predictions of proposed in the paper simple regression model, for the firms where CEO is an individual who controls the firm or a member of the family that controls it, probability of centralization of governance is on average 15 percentage points hig­her. In Austria, where the described effect is most prominent – 38 percentage points higher.

The paper is devoted to the analysis of custom tariff formation process during Russian WTO accession. The authors analyze what factors were the main determinants of inter-indu­stry differences for the Russian WTO import tariff lines. Specifically it is tested whether the industry output, industry import and price elasticity of demand for import were the main determining factors for the industry tariff rates after the transition period of Russian accession. Also, the authors analyze the sensitivity change of import tariffs to the main determinants subject to industry lobbying. The deep theoretical review of economic models analyzing the tariff formation process is observed to form the appropriate model, different econometric methods such as ordinary least squares and the method of instrumental variables is used to test the hypotheses. The empirical study of Grossman – Helpman model leads the authors to the conclusion that inter-industry differences in import tariffs cannot be fully explained by the industry lobbying, industry output and price elasticity of demand for import. The Grossman – Helpman model hypotheses are valid only for certain industries, and subject to very specific assumptions for industry lobbying – for all industries in sample. The main paper result is that on average lobbying industries (ceteris paribus) managed to gain higher tariff protection levels after transition period of Russian WTO accession.

According to current international climate change regime countries are responsible for greenhouse gas emissions, which result from economic activities within national borders, including emissions from producing goods for exports. At the same time imports of carbon intensive goods are not regulated by international agreements.

In this paper carbon emissions embodied in exports and imports of Russia were calculated with the use of inter-country input-output tables from WIOD database. It was revealed that Russia is the second largest exporter of emissions embodied in trade and the large portion of these emissions is directed to developed countries. The reasons for high carbon intensity of Russia’s exports are obsolete technologies (in comparison to developed economies) and the structure of commodity trade: Russian exports primarily fossil fuels and energy-intensive goods and imports products with relatively low energy intensity.

Because of large amount of net exports of carbon intensive goods the current approach to emissions accounting does not suit interests of Russia. On the one hand, Russia, as well as other large net emissions exporters, is interested in the revision of allocation of responsibility between producers and consumers of carbon intensive products. On the other hand, current technological backwardness makes Russia vulnerable to the policy of ‘carbon protectionism’, which can be implemented by its trade partners.

Regression analysis is in wide use in scientific investigation. Fuzzy linear regression is an actively developing area of research since in many real-life situations dependent or independent variables are not given as real numbers. The regression problem with fuzzy data is treated in the literature with different kinds of input-output data.

We consider the model y_i = A + bx_i + ε_i, i = 1,…,n, where A, x₁,…,x_n – fuzzy numbers; b – real number; ε₁,…,ε_n, y₁,…,y_n – fuzzy random variables.

In [Veldyaksov, Shvedov, 2014] A,b estimates were proposed, using ordinary least squares approach. The estimates rely on calculus of variations, and on previous research conducted for the case when A is a crisp (real) number.

Estimate for b is also proposed in [Veldyaksov, Shvedov, 2014]. In first part of this paper, we prove that this estimate is unbiased. We use new fuzzy random variables definition from [Shvedov, 2013].

In second part of this paper we refer to a number of numerical tests to compare confidence intervals for b coefficient, calculated both using traditional approach, and bootstrap approach. We show that the intervals become closer, as number of observations grows. We also propose a procedure for hypothesis testing for b coefficient in regression models with fuzzy data.