A US recession?

by Steven Bell, Managing Director, Portfolio Manager & Chief Economist, Multi Asset Solutions, BMO GAM

Intangibles may have extended the economic upswing but they haven’t abolished recessions. Indeed, there is an overwhelming consensus, supported by a number of external speakers at our Forum, that the US will slip into recession in 2020.

We disagree. Certainly, a slowdown in US growth is inevitable given the hectic pace at which the economy has expanded this year. Unemployment cannot keep falling by 1 per cent a year, while the boost from President Trump’s hefty tax cuts and fiscal stimulus is likely to fade. It should also be noted that the Fed is probably only part of the way through its cycle of increasing interest rates.

In the past, the US yield curve – the gap between long and short term yields – has been a good predictor of US recessions. In part this reflects a simple correlation. When economic data imply that a recession is imminent, markets will expect interest rates to fall. This would flatten or even invert the yield curve. But this has also been a causal relationship. In the past, a flatter yield curve has compressed bank lending margins. This has led to a downturn in the credit cycle, which has made a recession more likely. This causal link appears to be absent this time around. Lending margins have widened as the yield curve has flattened. Failure to appreciate this may have led many market participants to overstate the chances of a US recession.

Consensus suggests a US recession is likely in 2020. Is this likely?

The Phillips curve shows the relationship between wage inflation and unemployment. There is much debate about the strength of this relationship or, indeed, whether it has broken down completely. It is usually plotted as a scattergram but a clear picture emerges if the dots are joined up, as in AW Phillips’ original 1958 article. In chart 6, the dark blue segment relates to economic downturns and recessions when unemployment is rising. The upswing part of the cycle, when unemployment is falling, is plotted in light blue. The dark segment is clearly steeper than the lighter segment. This suggests that inflationary pressures emerge relatively slowly in economic upswings but diminish rapidly in downturns.

Offsetting these concerns, however, is a benign inflationary picture, which is, in part, a consequence of the way in which intangibles have alleviated capacity constraints. This suggests that, rather than suffering an outright contraction, the US will only experience a modest slowdown in its next downturn.

The charts indicate that if the downturn is sufficient to raise unemployment – not our core view over the next year or so – wage inflation would decline, allowing the Fed to respond quickly. This would reduce the risk of a severe recession.

Forecasters can probably discount one of the main catalysts of the last downturn – the US housing market – from being a cause of the next. Household debt has fallen, serious mortgage delinquencies are on the wane, as is the proportion of subprime loans in overall housing debt. Yet, housing has been a source of weakness this year. This is partly in response to rising mortgage rates but is also a result of the Trump tax reform, which restricted mortgage tax relief, and capped at $10,000 the amount of state and local tax (SALT), much of which is property-related, that can be deducted from federal income tax. This is a useful, if accidental, offset to reflationary pressures stemming from
the Trump tax reform.

As it stands, there is no obvious trigger for the next recession. But if one should arise, perhaps brought about by an energy shock or an overzealous policy reaction by the Fed, we believe it would be mild and short-lived.