This is the 20th annual edition of this report that analyses where current spreads are relative to default risk using historical rating agency data. Since the late 1990s, we have been to both sides of the default and spread spectrum on a few occasions. If these 20 years have taught us anything, it is that default cycles are part of the fabric of financial markets, as are extreme spread ranges. At the moment we remain in a relatively low default world with spreads towards the tighter end of their historical range. We’ve been here before, though, and while such an environment can be long-lasting – and appear bulletproof in real time – it inevitably ends at some point.
We examine four recession-led default cycles (1981-, 1989-, 1998- and 2007-) and two non-recession-led mini default cycles (1985- and 2015-). Fed hiking cycles preceded default cycles by 1.5-3.5 years before the 1981, 1985, 1989 and 2007 cycles. The big miss was in 1998, but this was a shallower start to the default cycle, and the start of the May 1999- hiking cycle occurred well before defaults began to accelerate to elevated levels. So there was still some warning. In this cycle, the Fed started its regular hiking cycle in December 2016.
Before the four recessionary default cycles, the US unemployment rate fell below NAIRU (non-accelerating inflation rate of unemployment) 1-3.5 years before defaults picked up. In this cycle we fell below this level in Q4 2016. Before the same four recessionary default cycles, the US yield curve (YC) inverted 3 months to 24 months before defaults picked up. In this cycle we haven’t inverted yet but the US 2s10s is now at +44bps – the flattest in a decade. Given the Fed hiking cycle is ongoing (DB expects seven hikes to YE 2019), the risk of an inversion is building either through a Fed policy error or a market error where yields don’t rise enough to reflect growth/inflation risks.
So, for the Fed hiking cycle and unemployment rate falling through NAIRU, the timer to the start of the next default cycle has started ticking – and if the last four recessions are anything to go by, the latest a serious default cycle will start is H1 2020. The timer from the yield curve indicator hasn’t yet started ticking, but we’re getting closer and closer to this point as Fed hikes continue and the back end of the curve stays stubbornly subdued. Given the shorter time lag historically between the YC and defaults, an inversion over the next 12 months could still signal a serious default cycle starting in 2020.
Online Document PROD0000000000464268 not found.