bayesian inference

The purpose of this research is Monte Carlo Markov chain simulation under Bayesian inference to identify the parameters affecting earning quality measurement. In this regard, in order to predict the earning behavior of companies and to derive the exact parameters of the model from the Bayesian Markov Monte Carlo (MCMC) technique, which takes cross-sectional heterogeneity into account, an analysis was done by coding in Python. In this research, the earning signals extracted from the financial statements on a quarterly basis for a period of 5 years (2018-2022), for 104 companies admitted to the Tehran Stock Exchange, were collected and analyzed using a new measure of earning quality. Auxiliary variables of accounting comparability, financial leverage, operating cycle, and sales volatility were used to achieve more accurate results, and several statistical performance measures (R2, RMSE, and MSE) were used to evaluate the effectiveness of Bayesian-based forecasting models. The results showed that the proposed criterion of the present study derived from the Bayesian model for training and testing data is well able to predict the quality of earning. The evidence shows that the results of the proposed model are superior to the conventional accrual earning management model, which suggests an error rate of MSE=0.0188 and RMSE=0.1369, respectively. The results of the present research can be used to analyze the portfolio and predict the quality of future earnings of companies using historical data. It can also be used to study factors affecting investment performance.

In order to measure hedge funds operating under the wings of two documented, many financial institutions tend to use the loss distribution approach. But a loss distribution approach requires a large number of internal loss data in order to have the necessary performance, so due to limitations in the database operating losses and the cost of internal loss data collection, in order to increase performance and reliability the operational risk capital should be calculated from other data sources used for operational risk. The biggest challenge facing financial institutions is how to combine different data sources of operational risk. In this regard, expressed in this research has been how to combine a variety of data source. So, in this paper the parameter estimation of frequency of operational risk loss distribution approach using Bayesian inference is explored. In this research, assuming dependencies between data sources, operational risk, the experts and internal loss data is intended. To validate the estimated models for the posterior distribution of numerical tests of goodness of fit is used. In addition, to calculate dependencies between data sources, detailed functions family of Gauss is used. The results indicate that with the assumption of experts dependence between the source data and internal data loss, by increasing the number of predictive parameters, frequency distribution, reduced the value of the parameter distribution, which represents a decrease of profile risk over time.

In previous studies, the normal mixture, as well as the Markov process, were used to model the financial return, separately. In this study, the normal mixture model is extended to the Markov mixture of normals. The mixture weights in every state are considered time-varying and as a function of past observations, so the limit of constant weight assumption is removed. The proposed model is estimated using Bayesian inference and a Gibbs sampling algorithm has been created to compute posterior density. The performance of algorithm is tested with simulation, then a two-state Markov time-varying Mixed Normal-GARCH model (MMN) with one and two components in every state, as well as limited cases (mean zero), were compared by comparison of their likelihood function. Finally, the model is applied to S&P500 and TEPIX daily return and results show that MMN models with two components provide better results than MMN model with one component which is so-called Markov switching GARCH model.