Which credit gap is better at predicting financial crises? A comparison of univariate filters

The credit gap, defined as the deviation of the credit-to-GDP ratio from a one-sided HP-filtered trend, is a useful indicator for predicting financial crises. Basel III therefore suggests that policymakers use it as part of their countercyclical capital buffer frameworks. Hamilton (2018), however, argues that you should never use an HP filter as it results in spurious dynamics, has end-point problems and its typical implementation is at odds with its statistical foundations. Instead he proposes the use of linear projections. Some have also criticised the normalisation by GDP, since gaps will be negatively correlated with output. We agree with these criticisms. Yet, in the absence of clear theoretical foundations, all proposed gaps are but indicators. It is therefore an empirical question which measure performs best as an early warning indicator for crises. We run a horse race using expanding samples on quarterly data from 1970 to 2017 for 41 economies. We find that credit gaps based on linear projections in real time perform poorly when based on country-by-country estimation, and are subject to their own end-point problem. But when we estimate as a panel, and impose the same coefficients on all economies, linear projections perform marginally better than the baseline credit-to-GDP gap, with somewhat larger improvements concentrated in the post-2000 period and for emerging market economies. The practical relevance of the improvement is limited, though. Over a ten year horizon policy makers could expect one less wrong call on average.