Экспертно-статистический байесовский подход к сценарному политическому прогнозированию
In the book deals with modern methods and models of socio-economic forecasting, the most frequently used in practice. Essential part of economic decisions aimed at obtaining results in the future, so to make the right management decisions need to be reliable socio-economic forecasting, which is impossible without knowledge of methods and models. In connection with this prerequisite training of highly economist and manager is to study their discipline "methods of social and economic forecasting." The second volume contains the basic textbook methods and models used today in the socio-economic forecasting. Consistently provides methods and models short, medium and long-term forecasting how simple models of trends and using factor models. In some groups are identified and methods of forecasting models of evolutionary processes of socio-economic dynamics. The textbook is designed for students of undergraduate academic, but may be useful to undergraduates, postgraduates and doctoral students, as well as practitioners dealing with the forecasting of socio-economic processes.
Introduction. There are a lot of models of bankruptcies prediction, which differ in methods of modeling and in set of factors. These methods are mainly belong to 5 groups: the classic statistical methods, regression analysis, the method of discriminant analysis, logit-analysis methods, methods of fuzzy sets and neural network methods. Combinations of these methods also can take place. The last three groups of methods are currently being developed especially quickly. As for the choice of factors bankruptcy, prevails heuristics. There is no formal methodology for selection and comparison groups of economic indicators to build the model of bankruptcies, as well as effective methods for data preprocessing. In this paper we propose an original method for the choice of indicators, followed by the construction of neural network model diagnostic of bankruptcies based on Bayesian approach.
Methods. The study applied the methods of mathematical statistics, correlation analysis, neural network modeling, methods of knowledge representation in intelligent information technologies.
Results. The developed concept of formalization of choice and comparative evaluation of a system of indicators has created the preconditions for the development of effective neural network model of bankruptcies.
Discussion. The proposed concept was tested for the construction sector of the economy. However, the authors believe that the generality of the approach, the concept and the method may be useful in other industries for a wide range of economic problems, such as the formation of the loan portfolio, an external audit or evaluation of the financial condition of the company.
The paper presents a vector autoregression-based approach to evaluation of fiscal measures. Using both structural and bayesian VAR models, we estimate fiscal multipliers of overall government expenditure and its components: national defense, national economy, education and social policy expenditures. Results are checked for robustness using different lag length and model specifications and different sets of hyperparameters in Bayesian VAR case. The results regarding overall expenditures are quite stable across specifications and correspond well with results obtained in previous research in this field. National defense and social policy multipliers are shown to be negative (in contrast to previous research on Russian data), national economy and education expenditures multipliers are positive. Some policy implications are presented.
The paper illustrates how a Bayesian approach to yield fitting can be implemented in a non-parametric framework with automatic smoothing inferred from the data. It also briefly illustrates the advantages of such an an approach using real data.
The paper uses an infinite dimensional (functional space) approach to inverse problems. Numerical computations are carried out using a Markov Chain Monte-Carlo algorithm with several tweaks to ensure good performance. The model explicitly uses bid-ask spreads to allow for observation errors and provides automatic smoothing based on them.
A non-parametric framework allows to capture complex shapes of zero-coupon yield curves typical for emerging markets. Bayesian approach allows to assess the precision of estimates, which is crucial for some applications. Examples of estimation results are reported for three different bond markets: liquid (German), medium liquidity (Chinese) and illiquid (Russian).
The result shows that infinite-dimensional Bayesian approach to term structure estimation is feasible. Market practitioners could use this approach to gain more insight into interest rates term structure. For example, they could now be able to complement their non-parametric term structure estimates with Bayesian confidence intervals, which would allow them to assess statistical significance of their results.
The model does not require parameter tuning during estimation. It has its own parameters, but they are to be selected during model setup.