Are linear models really unuseful to describe business cycle data?

The authors use first differenced logged quarterly series for the GDP of 29 countries and the euro area to assess the need to use nonlinear models to describe business cycle dynamic behaviour. Their approach is model (estimation)-free, based on testing only. The authors aim to maximize power to detect non-linearities and, simultaneously, they purport avoiding the pitfalls of data mining. The evidence the authors find does not support some descriptions because the presence of significant non-linearities is observed for 2/3 of the countries only. Linear models cannot be simply dismissed as they are frequently useful. Contrarily to common knowledge, nonlinear business cycle variation does not seem to be a universal, undisputable and clearly dominant stylized fact. This finding is particularly surprising for the U.S. case. Some support for nonlinear dynamics for some further countries is obtained indirectly, through unit root tests, but this can hardly be invoked to support nonlinearity in classical business cycles.

研究团队采用29个国家与欧元区经对数一阶差分处理的GDP季度序列，以评估使用非线性模型刻画商业周期动态行为的必要性。该研究方法无需模型估计，仅基于检验流程开展。研究团队旨在最大化检测非线性特征的检验功效，同时规避数据挖掘带来的各类陷阱。研究所得的实证结果并不支持部分既有论断，因为仅在三分之二的样本国家中观测到了显著的非线性特征。线性模型不应被轻易摒弃，因其往往具备实用价值。与学界普遍认知相悖的是，非线性商业周期波动并非一项普适性、无争议且占据主导地位的典型化事实，这一结论在美国样本中的表现尤为出人意料。另有部分国家的非线性动态特征可通过单位根检验间接获得佐证，但该结论几乎无法用于支撑经典商业周期中的非线性假设。