Not out of the woods
The Business Cycle Index (BCI) model, which uses a range of economic and financial variables to estimate the strength of the U.S. economy and to forecast the probability of recession, is still in "risk-off" territory. Although short-term risks remain low, the BCI model estimates the probability of a U.S. recession in 12 months at 30%, which is the warning threshold for leaning out of risky assets. We had staged some initial cyclical downgrades on equities and risk-seeking fixed income last quarter, when the BCI recession probabilities first crossed the warning threshold. If the recession probabilities increase or stay above the warning threshold longer, we may downgrade cycle scores further.
The path forward for the U.S. economy will depend on whether the Federal Reserve can cut rates enough to stabilise China/U.S. trade tremors.
Prolonged trade uncertainty damages global growth and corporate profits. A mild earnings recession, where we see consecutive quarters of negative earnings growth, looks increasingly likely. Historically, firms respond to poor profits by reducing capital expenditure (which they’ve already done) and hiring (which they might be starting to do). Consumer spending remains a pillar of strength for the U.S. economy as the fourth quarter begins, but that may quickly change if the labour market gets hit.
The bond market prices this increased risk through the inverted 10-year/3-month U.S. Treasury yield curve, an indicator which continues to distress the BCI model. An inverted yield curve – provided it’s of sufficient1 duration and magnitude – is historically one of the best predictors of recession. The yield curve has been inverted for more than 100 days and counting, with a maximum inversion of 50 basis points. These are duration and magnitudes consistent with previous pre-recession episodes. There is a chance that the damage is already done, but a path out of the woods involves accommodative monetary policy and a stable trade environment for businesses.
BCI recession probabilities may lower back under the warning threshold if 1) the Fed cuts interest rates enough to un-invert the yield curve over the next couple months, and 2) macroeconomic and financial data do not deteriorate in the meantime. Until then, our models continue to recommend defensiveness as we look toward the fourth quarter.
1Historically, there have been false alarms with the 10-year/3-month U.S. yield inversion, when the inversion was very brief. For example, in September 1998, the yield curve inverted shallowly and sporadically for five days over a 25-day period but did not precede a recession.