The US economic expansion has just become the second longest on record. If it continues beyond mid-2019, it will be number one. Its longevity is probably due to a mixture of circumstances, judgement and luck. The severity of the recession following the global financial crisis (GFC), coupled with the slowness of the subsequent recovery, has played a part. Regulatory reforms to the financial sector, implemented in response to the GFC, may also have contributed. And the fact that the US hasn’t been hit by any shocks of sufficient size to knock it badly off course has certainly helped. But as the economy heads towards that all-time record in 2019, what could bring it back down to earth? And are we capable of predicting the end before it’s upon us?
Oil price shocks, rising interest rates and asset bubbles tend to figure high in any list of recession triggers.
But while large and unanticipated rises in oil prices contributed to recessions in the 1970s and 1980s, they have had a smaller impact recently. This reflects the fact that oil now has less of an influence on either producer or consumer price inflation, and therefore tends to squeeze firms’ margins and real incomes less than in the past. As a result, it has a smaller chance of tipping the economy into recession.
Monetary policy tightening, in contrast, has been just as much a feature of the run-up to recent recessions as past ones. One study points to rate rises as a proximate cause of 29 of the 45 recessions the G7 economies have experienced since 1960. Not only that, but a significant tightening of monetary policy occurred in the three years preceding every US and UK recession during the past 58 years.1 This makes intuitive sense: textbook descriptions of the business cycle outline a period of expansion that comes to an end as spare capacity is used up, inflation rises and monetary policy is tightened. This slowing in activity, intended to weaken inflationary pressures, can end in recession.
3 reasons why reality is more complex than textbooks
First, on eight occasions since 1960, interest rates actually fell in the run-up to recession. This may have reflected (ultimately ineffectual) attempts to fend off a downturn.
Second, policymakers’ motivation for raising interest rates has changed over time – so the underlying causes of recession have shifted, even if the monetary policy response has not. During the 1960s, policymakers tended to raise rates to address current-account imbalances and exchange-rate pressures. In the 1980s, it was high and rising inflation. More recently, asset price bubbles have become more of a concern – particularly in the housing market, where 11 of 12 housing-market busts have been followed by recession within three years. Given the role that property plays in both banking sector assets and household wealth, this is hardly surprising.
Finally, recessions often have multiple causes: 70 different triggers were identified for the 45 recessions in the study. Sometimes these multiple causes may be coincidental. On other occasions, they are linked – as in the GFC, which featured tightening monetary policy, a burst US housing bubble and an international banking crisis.
So what lessons can we learn to help predict the next recession?
Focusing our attention on prospective inflation pressures – by looking at wage trends, or at surveys of firms’ price-setting behaviour and households’ inflation expectations – may be necessary, but it is not sufficient. In a world where independent central banks have adopted targets and pre-emptive tightening, and where inflation may have become more subdued for long-term reasons to do with “casualised” labour markets and an “Amazon effect” that has increased price transparency, we need to monitor other aspects of economic behaviour, too.
We have therefore incorporated a number of measures of economic leverage and potential imbalance – including credit growth, residential investment, mortgage servicing costs and corporate profit growth – into our analysis of recession risks. Our models suggest that there is currently little pressure from such imbalances, but this may change as we look to 2020 and beyond. There are, however, huge uncertainties around the results of such models. For example, while our analysis suggests that there is a 66% chance that the next recession will strike between January 2020 and June 2023, there is a 95% chance that it could happen any time over the next seven years. Such broad timescales are not particularly informative when it comes to business planning. Furthermore, such models cannot “see” the latest US fiscal stimulus, nor the Federal Reserve’s likely response, because the stimulus is unprecedented at such a late stage in the economic cycle. So the recession risks could be underestimated.
All is not lost, however. In the past, analysts looked to the term spread - the difference between long-term and short-term interest rates – to help predict recessions. This approach had fallen out of favour, against a backdrop of ultra-low interest rates and quantitative easing. But recent research by the San Francisco Federal Reserve suggests that the term spread – or yield curve - remains a strikingly accurate predictor, even under current circumstances. After all, every U.S. recession in the past 60 years was preceded by a negative term spread, or inverted yield curve. Furthermore, an inverted yield curve was always followed by an economic slowdown and, except for one time, by a recession. The US yield curve is certainly flat at present, with the term spread falling to its lowest level in a decade. But it has not inverted. For now, our baseline view remains that the US economy will break that 2019 expansion record.
For further insights and opinions on global macro events and investment strategies, please visit our blog