The yield curve, cycle and cyclicals
The US recovery has been the second longest on record, but 6pp less than its
‘average’ historical magnitude. Based on the factors below, we conclude that a
US recession is unlikely until Q3 2020. Typically, recessions are preceded by:
■ Yield curve inversions, which normally lead a recession by 11 months. We
would expect an inverted yield curve in mid-19, implying recession in mid-20;
■ Wage growth rising to 3.5%: This would hit margins and more importantly
force the Fed into a tight policy. We continue to believe full employment will
be reached at an unemployment rate of c.3.5%, and thus do not believe
this level of wage growth will occur until H2 19. Historical experience
suggests a recession a year later (H2 2020);
■ Output gaps closing: Normally a recession occurs 2.25 years after the
output gap closes (implying Q2 2020);
■ A margin squeeze: Normally a recession occurs 2.7 years after margins
have peaked, although neither NIPA nor bottom-up margins have peaked.
The eventual margin squeeze is likely to be caused by wage growth
accelerating to 3.5% or over-investment, of which there are few signs so far;
■ Deleveraging: 63% of the increase in global private sector leverage since
2008 has come from China, but we see no reasons for China to deleverage
in a disruptive manner given net foreign assets, government debt to GDP of
just 23%, a loan to deposit ratio of 88% and 30% LTVs. The risk is US
corporate leverage, as net debt to EBITDA ex tech is close to its previous
peak. However, FCF is still higher than total payouts: recessions tend to
occur two years after the corporate sector moves into deficit;
■ Other factors: Inventory cycles are much better managed and we think
central banks will continue to risk doing too little rather than too much
tightening.
Implications: 1) Remain overweight equities. In the past 50 years (with the
exception of 1987, when the equity risk premium hit a very low 1% compared
with 5.1% now), equities have never peaked more than 13 months ahead of a
recession. A de-rating typically only occurs in the 6 months ahead of a
recession. 2) High vol is here to stay, sell leverage: A flat yield curve leads
volatility by 2 years, so high vol looks here to stay. This means credit (which
correlates to vol) is more risky than equity. Sell highly leveraged companies,
with low FCF yield and negative earnings revisions (stock screen inside). 3)
Stay underweight non-financial cyclicals ex tech. Non-financial cyclicals ex
tech have already priced in PMIs of c65, are expensive, have outperformed for
longer than any other periods and are over-owned. We favour lowly leveraged
defensives (see screens inside). 4) At this stage of yield curve flattening,
normally software and energy are the best performing sectors, while
consumer staples and utilities perform the worst.