摘要：For the same macroeconomic variables more predictions can be made, using different forecasting forecasting. The most important step is the choice of the prediction with the highest degree of accuracy, this being used in establishing the governmental policies or the monetary policy by the central bank. We made short run forecasts (January 2012-March 2012) for variables as inflation rate, unemployment rate and interest rate for Romania using techniques like: econometric modeling, exponential smoothing technique and moving average method. In order to improve the forecasts accuracy, we used two empirical strategies: making combined prognosis and building the forecasts based on historical accuracy indicators. The predictions based on exponential smoothing technique have the highest degree of acuracy, being superior to those got applying the strategies of improving the accuracy.
