Rates as of 04:00 GMT
Market Recap
If you ever wanted proof that a) currencies are not the stock price of a country, and b) right now currencies aren’t reacting to economic indicators in the way one might usually expect, yesterday was the perfect example. One is hard-pressed to sum up just how bad yesterday’s economic data was for the US. In short, it was the worst ever in every category: retail sales had their worst-ever drop, industrial production had its worst-ever peacetime drop, the Empire State manufacturing index had its worst-ever drop to its lowest recorded level (more than double the previous low), the NAHB housing market index had its worst-ever drop (4x as big as the previous worst drop), etc etc. Nonetheless, the dollar was the best-performing currency as risk aversion gripped the market.
On the other hand, Australia announced a surprisingly good set of employment data: the number of jobs rose by 5.9k (vs expectations of -30k job losses) and the unemployment rate only rose a bit to 5.2% from 5.1% (5.4% expected). This probably was due to the timing of the survey. Note too that this had zero effect on AUD, which fell sharply anyway.
AUD wasn’t the worst-performing currency of the day however. That “honor” went to CAD, which is getting hit both by the high dollar and the low oil price (see below). The Bank of Canada yesterday held rates steady but announced two more purchase programs and said it’s ready to adjust the scale or duration of its purchase programs if necessary. It estimated that real GDP will be 1%-3% lower in Q1 and a whopping 15%-20% lower in Q2. Terrible figures, but Canada probably isn’t alone in this by any means, although the Canadian economy’s sensitivity to oil prices may make its downturn greater than some other countries’.
Speaking of which, oil continued to fall as the demand shock implied by the dreadful US economic data outweighed any conceivable hope from the recent OPEC+ cuts. Contrary to what I said yesterday, the spread between WTI and Brent is now below $4 out until the end of next year, so it’s not particularly economic for US producers to ship their oil to Europe. This will mean further shutdown of US oil production.
The saying in the oil market is that “low prices cure low prices,” and this is one example: when prices are too low, producers have no choice but to stop producing. Lower production eventually pushes prices back up. The question is, can oil producers cut back production enough to meet the collapse in demand? And what will be the long-term effect of this? An oil well isn’t like an auto factory, where you can turn out the lights and send everyone home and then a month later just come and start everything up again. Shutting down an oil well can cause permanent damage to the well. The industry last saw such widespread shut-ins, as they’re called in the oil industry, back in 1986. In some cases, crude oil has in the past actually gone negative – one producer paid someone (not much, 50 cents a barrel) to take it off their hands instead of shutting-in the well.
One positive note: although the cases of the virus worldwide rose above 2mn, there are signs that the curve is flattening: it took 13 days for the number to double to 2mn, whereas it took only eight days to double from 500k to 1mn. This should be seen as somewhat of a victory, but a minor one: it’s still true that the number of cases continues to increase, even if it is at a slower pace. And a lower percentage increase doesn’t necessarily mean a lower absolute number every day. We are still a long way from back to normal. Germany yesterday showed us what the “new normal” will look like as it announced a cautious exit strategy from its lockdown, which was imposed on 22 March. Some smaller shops will reopen next week and schools will gradually restart in early May However, restaurants, bars and hotels will remain closed until further notice while large events, such as soccer matches or concerts, will remain forbidden at least until end-August at the earliest. These guidelines suggest to me that we are in for a prolonged period of low economic activity.
In the US, Trump indicated that he would unveil today guidelines on a similar lifting of restrictions. However, since it isn’t up to him (contrary to what he says) but rather up to the governors of each state, this is only a suggestion.
Today’s market
Eurozone industrial production, which was showing signs of another downturn in December, recovered strongly in January. It’s expected to show no change from January’s level in February, which actually would be a fairly good response as it would mean the Eurozone maintained January’s gains. Alas, with Italy shutting down its main industrial region and the rest of Europe following along shortly thereafter, I would doubt if we’ll see the gains maintained in March.
US housing starts and building permits for March are expected to be lower, of course, but nothing dramatic. They’re expected to fall back to the level of last summer, when people were not particularly thinking that the US economy was likely to collapse.
To put this in the longer perspective, it would still be a relatively high figure for the post-Global Financial Crisis economy. Moreover, the expected fall of 292k in starts during the month is nothing like the record -597k fall in March 1984, when of course the US housing stock was much lower to begin with. Similarly, the expected 152k decline in permits would be less than half the record 419k in February 1990. We’ll have to see though how well housing holds up.
It could well be that with the fall in interest rates, the housing market remains relatively firm. However, banks tend not to lend money to people without jobs, so even if mortgage rates do come down in line with the fall in interest rates, it may well be that this time lower borrowing costs don’t translate into more demand.
However, yesterday’s National Association of Home Builders’ (NAHB) survey was nothing short of disastrous – it was expected to plunge to 55 from 72, instead it collapsed to 30. That 42-point decline is four times the previous worst-ever one-month drop (10.0 points in February 2014). That portends a collapse in home sales.
The Philadephia Fed business outlook survey is expected to show another decline, although not as big as last month’s record 49.4-point fall. Hah! Yesterday’s Empire State manufacturing index was expected to be down 13.5 points to -35, but in fact it plunged 56.7 points to a record low of -78.2 (data back to 2001). Orders, shipments, employment and the average workweek all fell at a record pace.
Now comes the biggie: the weekly US initial jobless claims. Because of the quirks of the survey process for the nonfarm payrolls, this has taken over as the main indicator of the health of the US economy. Even so, it’s fatally distorted by the problems that millions of people have had in filing for unemployment insurance, because the processes and websites for filing simply aren’t built for the level of claims that we’re seeing and can’t handle the traffic. Even the appallingly high level of unemployment that these figures show no doubt grossly underestimates the severity of the real situation.
The median forecast is for an increase of 5.5mn newly unemployed persons Nobody really knows though – estimates range from 2mn to 8mn. And at the low end, there’s only one person at each interval – one at 2mn, one at 2.5mn, two at 3mn, and then there are a number at 4mn. Similar at the high end – one person at 8mn, two at 7mn. The most popular forecast is 6mn.
How would that translate into overall unemployment, and how would that compare with past bouts of unemployment? We’re already at a record number of unemployed persons – the 17.3mn initial claims for unemployment since the beginning of March already exceeds the previous record high for the number of unemployed, 15.3mn in October 2009 (a year or so after Lehman Bros. went bust).
Based on the forecast of 5.5mn more jobless claims this week, that would mean 22.8mn newly unemployed persons. Added to the 5.8mn who were unemployed in February, that would mean 28.6mn unemployed persons or an unemployment rate of around 17.5%. That’s lower than 24.9% peak during the Great Depression, but there’s still one more week of jobless claims to go before the April employment survey.
One curious point is that these jobless figures didn’t affect the stock market, at least they haven’t so far. The last three weeks’ jobless claims figures have all been record bad, and yet stocks have risen on those days anyway:
26 March: claims +3.3mn, S&P 500 +6.2%
2 April: claims +6.6mn, S&P 500 +2.3%
9 April: claims +6.6mn, S&P 500 +1.4%
This demonstrates the stock market’s ability to ignore the present and discount the future. The future that it was discounting was that we were past the peak of the pandemic and looking forward to recovery. As the pandemic drags on though and we enter the peak of bad economic news & bad earnings season, it remains to be seen whether markets can continue to look towards the future with optimism.
Overnight China announces its usual three indicators of retail sales, industrial production and fixed asset investment (FAI). Since these data are for March, we also get the key GDP number.
GDP is expected to plunge (surprise surprise!) – the consensus forecast is for it to flip from +6.0% yoy to -6.0% yoy. That’s the first-ever negative reading for China, albeit the data only goes back to 1992 so it’s hardly a complete historical record.
There is a longer data series of year-to-date GDP growth. That shows what an outlier this figure is over the last 40 years.
The monthly data are expected to be little better. We can’t compare March with February as China doesn’t release February data alone – every year it releases January and February combined, because of the distortion caused by the week-long Lunar New Year holiday, which can be in January some years and February others. For this graph, the lines are the year-to-date yoy change and the dots are the forecast for the yoy change in March. That will give a better idea of the change from February to March.
Doesn’t look so good, does it? The data should show decisively that a “V”-shaped recovery is unlikely. And where China has been is where the rest of the world is probably going.
I don’t usually cover EU auto registrations, which probably isn’t market affecting, but it may give a reasonable indicator of consumer sentiment and consumption in Europe so you may want to look at it when it comes out. No forecast available.