Predicting the shape of recovery after a downturn in the business cycle is a favourite parlour game for economists and the coronavirus-induced plunge in business activity has resurrected the usual range of options.
Some forecasters see a fairly rapid and complete recovery once the lockdowns are lifted (V-shaped) while others see a delayed upturn (U-shaped) or a stuttering period of recoveries and renewed downturns (W-shaped).
More ambitious forecasters have raided the semiotic catalogue to propose recoveries shaped like a check-mark, a square root sign or a Nike Swoosh.
But the most likely outcome for the economy and the oil market is an initial fast-but-partial recovery followed by a slower and more difficult phase.
If that proves to be the case, the expansion will look a lot like the one after the recession in 2008/09 – which saw rapid growth in late 2009 and early 2010, followed by more disappointing and uneven performance in 2011/12.
- 1 Gunners flex out at shooting contest
- 2 Ronaldo-less Juventus drop points against Verona ahead of Messi showdown
- 3 More Covid-19 cases reported as medics struggle with admissions
- 4 How Covid-19 has had harsh effects on cancer patients
Covid 19 Time Series
“During a recession, a significant decline in economic activity spreads across the economy and can last from a few months to more than a year,” the US National Bureau of Economic Research (NBER) says.
NBER is a private sector organisation rather than a government agency, but its business cycle dating committee has become the commonly-accepted arbiter of the business cycle in the United States.
The committee normally waits until the existence of a peak or trough is not in doubt and it can assign an accurate peak or trough date.
For this reason, the onset of a recession or recovery is often announced many months after it occurred, in some cases more than a year later.
The last recession started in December 2007 but it was not announced for another 11 months. The last recovery started in June 2009 but was not announced for 15 months.
In some ways, the recent slump in business activity is much easier to date. Most indicators show US activity was still expanding in February but fell sharply in March and even more severely in April.
Coronavirus-induced lockdowns abruptly cut off a mature expansion (128 months) but still looked reasonable underlying in momentum (https://tmsnrt.rs/3caK1ZY).
February (or possibly March) marked the peak of the expansion as well as the onset of the contraction on the standard definition. The trickier question is whether the sudden slump in business activity qualifies as a recession.
The slump in March/April certainly qualifies as “a significant decline in economic activity,” the most significant since the 1930s. But the second part of the definition talks about a decline that “spreads across the economy” and it is less obvious a description of what happened in April.
Recessions are transmitted through an economy as a downturn in business activity, incomes and employment spreads, like an epidemic, from a few households to a wider section of society.
Recessions are normally characterised by second, third and subsequent rounds of business closures, income declines and employment losses as the initial shock is transmitted and amplified.
Like epidemics, recessions are social phenomena, in which the initial shock (or infection) spreads and accelerates through the links between firms and households.
Recessions are therefore social processes in which expectations and narratives play a critical role, which distinguishes them from single or complex events such as natural disasters (earthquakes) and manmade catastrophes (war).
But that is not a good description of what happened in March and April when many businesses were ordered to cease operating as part of virus-control measures.
If the current downturn does qualify as a recession, rather than something else, then it is very different from the recession of 2008/09 or any other recession for a hundred years.
The characterisation of the events of March and April matters because it will determine what happens next in the business cycle.
If it was simply a pause in business activity, the economy might bounce back quickly. The downturn is the deepest but shortest on record and might qualify as a proper recession.
In one extreme example, activities might have peaked in February, declined in March, reached a trough in April, before starting to rise again in May as governments relaxed lockdown for business to open.
On the standard definitions, the recession would have started in February and be over in April, making it just two months long (one reason why this downturn does not fit the normal recession template).
The economy went into the downturn with unprecedented speed. Perhaps it will emerge with equal rapidity.
Policymakers are hoping for an outcome like this. Finance ministries and central banks have provided unprecedented support to businesses and employees in the hope of preventing lingering damage.
Policies have been explicitly designed to put the economy into hibernation or a low-energy state to protect it from lasting harm while the epidemic is suppressed, allowing it to be revived quickly afterwards.
By providing exceptional support, without normal conditionality, governments try to break social linkages and dependencies through which a recession propagates. While health authorities are trying to contain the epidemic through social-distancing, financial institutions focus to avert a recession through economic distancing.
In this strategy, massive lending and income support is being used to loosen the normally tight relationship between business turnover and the state of the economy and between employment and household expenditure.
If the strategy is successful, business could be reactivated quickly in a few months, which would make this an exceptionally brief downturn.
However, for the economic impact of the lockdowns to be quickly and completely reversible, there must be minimal second-round effects on business activity, and few lasting changes in income and spending patterns.
Governments must be successful in severing the web of business-employment-income-spending linkages in the economy, and the post-lockdown economy must look very much like the pre-lockdown one.
In practice, neither of these conditions is likely to be met, which suggests there could be some significant lingering effects from the lockdowns even once they are progressively lifted.
The longer the lockdowns last, the greater the risk of second-round effects on business activity. The more social-distancing required in the post-lockdown world, the greater the painful structural changes.
As lockdowns ease, some effects will be quickly reversed, but others will be much harder to turn back. Some employees currently on furlough will return to work, others are likely to find themselves unemployed.
Some businesses will re-open quickly and completely with little change to turnover. Others will re-open only slowly and incompletely, taking months or years to rebuild. Some will not re-open at all.
The recovery is likely to be characterised by multiple phases. The first re-opening phase will see strong and rapid growth as some businesses reverse the lockdown and reactivate fully or partially, which will translate into rapid increases in petroleum consumption.
But at the end of first phase, business activity and oil use will still be well below pre-lockdown levels. Some businesses will run under capacity and some furloughed workers will be unemployed. This will last for months. The second rebuilding phase is likely to see slower and uneven growth.
The second phase will see businesses and households that have been hit by second-round effects on turnover and employment trying to reconstruct their position.
Rebuilding is tougher than re-opening, and these businesses and households will need much greater support for a longer period.
Even if the recovery starts quickly, economic activity and oil consumption could remain below pre-epidemic levels until 2022.
— John Kemp is a Reuters market analyst. The views expressed are his own.