Join our community of smart investors

The Roosevelt moment

Is the developed world heading for a Roosevelt moment? Could be
The Roosevelt moment

This is a tale of two famous quotations, both of which have a bearing on how the developed world will come out of its self-imposed lockdown; whether the re-emergence will be sufficiently confident to justify the resilience of the world’s equity markets these past six weeks (especially that of the US), or whether it will be so cautious, stuttering and faltering as to force equity markets to realign themselves with what the macroeconomic data is saying.

In chronological order, the first quote is perhaps the definitive sentence from Extraordinary Popular Delusions and the Madness of Crowds, Charles Mackay’s mid-19th century essay on financial mania: “Men, it has been well said, think in herds; it will be seen they go mad in herds, while they recover their senses slowly and one by one.” The second is the phrase from Franklin D Roosevelt’s inaugural address as US president in 1933 that “the only thing to fear is fear itself”.

These quotes form a useful juxtaposition because it remains unclear whether the developed world’s economies are nearer the start or the finish of a process that will end in, if you like, a Roosevelt moment. Are the senses of sufficient numbers recovered from the various effects of Covid-19 to permit the rich world’s output in the third quarter to be usefully better than what will be the second quarter’s dreadful figures, thus starting a process of recovery? Or will the nagging uncertainties – made plausible by the virus’s continuing toll – mean that the fearful herd continues to demand shelter and accept the downward economic spiral implied by that?

Table 1 offers a glimpse of the worst such a spiral might look like, taking data from the effects of the Great Depression on both US share prices and the nation’s output. In the 12 quarters covering 1930 to 1932, the Dow Jones Industrial Average managed to fall in nine of them, ending the process 84 per cent off its September 1929 high. Over the same period, US GDP fell three years running so that for every dollar of output at the start of 1930 there was just 75 cents-worth by the end of 1932. The leveraged effect on jobs was even worse. An unemployment rate of 5.0 per cent going into the depression topped out at 23 per cent in 1933.


Table 1: Relentlessly miserable
 Dow Jones Ind AverageUS GDP
 Ch on qtr (%)Ch from high (%)Ch on year (%)
1930 Q115.1-24.9 
1930 Q2-20.9-40.6 
1930 Q3-9.5-46.2 
1930 Q4-19.7-56.8-8.5
1931 Q14.7-54.8 
1931 Q2-12.9-60.6 
1931 Q3-35.7-74.7 
1931 Q4-19.4-79.6-6.4
1932 Q1-5.9-80.8 
1932 Q2-41.5-88.8 
1932 Q367.0-81.2 
1932 Q4-16.3-84.3-12.9
1933 Q1-7.6-85.5 
1933 Q277.1-74.3 
1933 Q3-3.4-75.1 
1933 Q45.4-73.8-1.2
Source: Bureau of Economic Analysis 


In just one month, today’s US jobless rate more than tripled from 4.4 per cent in March to 14.7 per cent in April, its highest level since the Great Depression. Canada followed suit in a muted way – its rate rose from 7.8 per cent in March to 13.0 per cent in April. Among the European economies worst hit by Covid-19 infections, Spain (especially) and Italy have structurally high rates of unemployment anyway, so their rise in joblessness in April may be less dramatic than that of the US. Annoyingly, the UK’s Office for National Statistics only releases labour market data quarterly, so this week’s figures for 2020’s first quarter tell us little about the nation’s dash to the dole. They show unemployment at 3.9 per cent of the workforce, about the level it has been for the past 18 months. However, data for the second quarter may well be around current US levels.

Whether that is the worst it will get chiefly depends on two factors – first, the continuing virulence of Covid-19 measured by its infection-fatality rate (IFR); second, the evolving perception of governments and electorates to the virus’s threat. While these two factors overlap, the second is also independent of the first, quite possibly markedly so. Granted, this may be stating the obvious since perceptions and reality always diverge, but it’s important to understand this.

Thus in Covid-19’s case, the western world under-reacted to its threat then over-reacted because that’s what people do. China, South Korea and Singapore in particular – all scarred by the effects of the Sars epidemic of 2003 – took extremely seriously the mutation of another coronavirus to cause severe respiratory illness. Meanwhile, the west sat back and enjoyed super-spreading events, encouraged by Prime Minister Boris Johnson taking his heavily pregnant girlfriend to a rugby international at Twickenham in March. It was only when northern Italy and Iran started to produce scenes from the apocalypse and epidemiologists at London’s Imperial College did something similar with number crunching that politicians responded. For the most part, dragged by their electorates and by a contagious group-think, they responded with a lockdown – and economic shutdown – that would have defied belief except that it was happening.

Naturally, there were sceptical opinions from day one. Those that counted most were the sceptics among the ‘scientists’ – the virologists, epidemiologists and mathematicians (since epidemiology is mostly a branch of maths). As Sweden pursued its comparatively relaxed response, Johan Giesecke, an emeritus professor at Stockholm’s Karolinska Institute, became the public face of Swedish cool. Infection rates were much higher than health authorities indicated, he said, thus Sweden would have herd immunity by mid-May. A by-product of this was that fatality rates were also much lower; perhaps as slight as one death in 1,000 infections, the same IFR as ’flu.

Shortly afterwards, Professor Hendrik Streeck, of the University of Bonn, overlaid some serious research onto this when he published a study of Heinsberg, a town of 40,000 in north-west Germany that had been the victim of a super-spreading event. The study of 600 households in the town found that about 50 per cent of 900 people had been infected and that the fatality rate was 0.37 per cent, or less than four in 1,000. That compares with first indications that Covid-19 – at least in countries with good healthcare – would come with an IFR of about 1 per cent, or 10 times the virulence of ’flu. Professor Streeck also suggests that the caution built into the Heinsberg study means the underlying fatality rate is probably less than three in 1,000 (0.26 per cent, to be precise).

To voices such as these has been added the heavyweight opinion of Professor Michael Levitt of California’s Stanford University, a 2013 Nobel Prize winner for chemistry and on whose thoughts Table 2 and the chart are based. Since February, Professor Levitt, who describes himself chiefly as a data analyst, has been sceptical of the compound growth rates used to project Covid-19’s spread. He says the data does not show compound growth worthy of the name because, for compound growth to be significant, it has to be sustained for a while. Instead, almost as soon as the growth rates in diagnosed cases and deaths have accelerated rapidly from a tiny base, they begin to slow down.


Table 2: Slowdown and lockdown
 Peak 10-day growth rate (%)Rate at the start of lockdown
 DateRate (%)DateRate (%)Days after peak*
*Minus sign means 'n' days before peak; Source: World Health Organisation



Taking the UK, using data from the World Health Organisation, the 10-day compound growth rate in case numbers peaked at 33 per cent on 8 March (see Table 2). In other words, that was the daily growth rate that took recorded cases from just 16 on 27 February to 277 10 days later. Then, 15 days later on 23 March – when the UK went into lockdown and without stringent interventions by the government – the growth rate had slowed to just 20 per cent. But it was hardly as though the virus was running short of potential victims, at least according to official data. Even by 23 March, just 5,040 cases had been logged, so only another 66m or so left to infect.

Perhaps most telling is the case of Iran, arguably the last place where Covid-19 should be running riot – an authoritarian state governed by a semi-functional theocracy, mistrusted by its people, whose healthcare system labours under the cosh of US sanctions. And so it seemed as Iran’s death count mounted in early March. However, its bigger picture is much like the UK’s. By the time the nation went into lockdown, the growth rate in cases had dropped from a peak of 77 per cent to just 16 per cent. Since then the pace has continued to slow and has been 2 per cent or less since mid-April.

Still, as Professor Levitt observes, almost regardless of how a nation’s government tackles the pandemic, its progression has been much the same, with the growth in case numbers starting to fall quickly and usually long before lockdown measures have been put in place.

Sure, there are variations around that; two such are China and the US, both of which went into lockdown before the growth in case numbers peaked. China remains a mystery. Believe the data and, Hubei province aside, the nation hardly saw a Covid-19 outbreak anyway. Granted, no one outside its communist party is likely to believe that or that China has suffered just three deaths from Covid-19 in the past 30 days, as it claims. Yet no one seriously doubts that the virus is under control; that any second waves have been small enough to hush up. Meanwhile, as the chart shows, the US struggled longer than most.

There is also a big problem with the quality of the data; the cliché ‘garbage in, garbage out’ springs to mind. Even so, it remains a mystery why Covid-19 has behaved as it has. Lockdown measures have obviously helped, although it is not clear which ones are effective and which are totemic. Yet there seems to be more to it than that, a gap filled by anthropomorphising metaphors such as “the virus is getting tired” or that “we are getting further away from that bat in Wuhan” – which means that viruses mutate quickly, usually to less lethal versions.

However, what looks increasingly likely is that the main obstacle to economic recovery will not be the virus, but the response of governments and electorates to a perceived threat; the willingness of governments to lift restrictions and the readiness of citizens to respond. Most likely, for some while yet the so-called ‘availability heuristic’ – much loved of behavioural psychologists – will have a strong pull. This is the mental device by which we give outsize importance to things or events that spring most easily to mind. Clearly it will be some while before Covid-19 drops off the list. Yet that also underestimates the pull of normality, the inclination of social animals to be social without, for instance, the slightly spooky pauses of video conferencing.

If this process leaves electorates recovering their senses and less fearful of fear then that must also help justify the resilience of equity markets. Let’s not forget that, at 3303, the FTSE All-Share index is just 22 per cent below the all-time high at which it opened 2020 – a level that seemed incomprehensible just two months ago. Sure, it could yet fall apart – how many false dawns, I wonder, were there during the Great Depression? And it’s not that I am dashing out to buy equities. Why would I? The Bearbull income portfolio remains fully invested. Even so, perhaps we can now get on with equity analysis without the feeling that there is a pesky little virus watching our every move.