We used a Bayesian dynamic factor model (BDFM) to examine the share of variance explained by the European and country factors for 59 regions in Eurozone countries in the period from 1992 to 2020. The BDFM output facilitated the construction of a criterion that enables the assessment of the cost of participation in the European Monetary Union, which is directly related to the optimum currency area theory. Over the examined period, we observed business cycle divergence, with 46 regions experiencing a drop in the share of variance explained by the European factor from 1992–2005 to 2006–2020. However, the analysis over shorter time spans demonstrated that all the regions decoupled from the European business cycle. The results contradict the predictions of "The European Commission View". On the one hand, two predictions stemming from "The Krugman View" are supported by the results: the European regions experienced a slight increase in sectoral specialization, and they experienced business cycle divergence. On the other hand, the data does not support the notion that the ongoing specialization was the underlying cause of this divergence.

我们使用贝叶斯动态因子模型 (BDFM) 来研究 1992 年至 2020 年间欧元区 59 个地区的欧洲和国家因素所解释的方差份额。BDFM 的输出有助于构建能够进行评估的标准参与欧洲货币联盟的成本，这与最优货币区理论直接相关。在研究期间，我们观察到商业周期差异，从 1992-2005 年到 2006-2020 年，46 个地区的方差份额因欧洲因素而下降。然而，对较短时间跨度的分析表明，所有地区都与欧洲商业周期脱钩。结果与“欧盟委员会观点”的预测相矛盾。一方面，“克鲁格曼观点”的两个预测得到了结果的支持：欧洲地区的部门专业化程度略有提高，并且经历了经济周期的分化。另一方面，数据并不支持持续的专业化是造成这种差异的根本原因的观点。