BULGARIAN GOVERNMENT DEBT MONTE CARLO SIMULATION FOR THE PERIOD 2017 2019
Author: Assoc. Prof. Nikolay Atanasov, Doctor of Economics
In the proposed study is made an attempt to integrate the risk of unfavorable increase of government debt of the Republic of Bulgaria in the period 2017 - 2019 with Monte Carlo simulation. Conducted simulation is based on the specifications of regression model of debt with tax revenues as an independent variable. For this purpose was used data for the period 1998 - 2016. The relation between the government debt and the tax revenues was statistically significant (R = 0,789). The availability of significant estimates of regression coefficients was also found using a second order polynomial functional form. Monte Carlo simulation is conducted based on randomly generated expected growth rate of tax revenues for the estimated period 2017-2019. On this base was specified the amount of expected annual tax receipts and using the regression equation - the forecast government debt values. The resulting simulation annual average debt (in 1000 attempts) and interval of central 68.3 % (based on normal distribution) of cases are close to a significant extent to the estimates of the government's medium-term budget forecast. The conclusions are that the method 'Monte Carlo' can be successfully used to assess the variability of government debt using the tax revenues as independent risk variable.