in a project Laurent Ferrara, we have been studying the properties of financial indicators as predictors of recessions (as defined by the NBER). In addition to term spreads, we also consider financial conditions indices (Arrigoni-Babasu-Venditti, Goldman Sachs), foreign term spreads (a la Ahmed Chin) and the Bank for International Settlements debt service ratio (suggested by Borio-Drehmann-Xia). Slides from the June presentation click here.
figure 1: Probabilistic regressions of term spreads (black), term spreads, FCI, foreign term spreads (red), and term spreads, FCI, foreign term spreads, and debt service ratios (cyan) imply recession probabilities. Estimates from 1985 to September 2023 (assuming no US recession as of September 2023). NBER-defined recession peak-to-trough dates are in gray. Light blue shading indicates out-of-sample periods. Source: NBER and author’s calculations.
Importantly, estimates based on term spreads and/or FCI and foreign term spreads yield a reasonably high probability of a recession in December 2023. Because of the different sample periods, they are not as high as some of the other estimates I have presented before. The most important difference is the debt service ratio of the private non-financial sector, as noted by Borio et al. Reports have roughly the same predictive power as financial cycle variables. This variable is currently only available through the end of 2022, reducing the implied probability to just 10%.
The pseudo-R2 for debt service ratio alone is higher than the term spread alone (or term spread, FCI and foreign term spread).
It’s too early to take too much comfort from these results: debt service ratios are rising rapidly.
(These results differ from June’s results as we now have more recent FCI data. Note that one would not necessarily want to apply the same sample period to Eurozone countries or the UK as we do not have the same level of confidence in these economies The body has not yet begun to decline.)
