Abstract
Economic crises in countries that prefer a free market economy have increased scientific studies on this issue. Along with the research, definitions, concepts and classifications related to the crisis have also increased. One of the important subheadings related to economic crises is the successful detection of crisis signals. For this, it is possible to use economic indicators dec combining them with statistical modeling techniques. With this study, it is aimed to establish an index with selected real and financial variables in order to keep the pulse of the country's economy in the right
way and to predict a possible crisis. For this purpose, Basic Components Analysis was applied to the compiled and calculated monthly data of thirteen selected variables and five basic components were obtained. The total variability of the components was found to be 79.9%. The five basic components were combined by weighting them with their own variance disclosure ratios and a one-dimensional crisis index was obtained. Time series analysis was applied to the obtained index and prediction was made. In addition, it has been determined that there are structural breaks in the index using the Least Squares-Based Cumulative Total Control Graph method. Between January 2005 and March 2020, five statistically significant breaks in the Turkish economy are: April 2007, December 2008, June 2010, July 2012 and June 2017. The relations of the breaks experienced in the index with the economic and political agenda in line with these dates have been examined and interpreted.