Friday, April 3, 2020
Get the Morning Brief sent directly to your inbox every Monday to Friday by 6:30 a.m. ET. Subscribe
Coronavirus calls into question data quality
Slowly but surely, we’re getting economic data reflecting the disruption caused by the coronavirus pandemic, which has forced business and consumers to be on lockdown.
While reports like weekly jobless claims and some of the manufacturing surveys have been unambiguously bad, some experts warn that there’s an elevated risk that the quality of the data we’re getting will be extremely low.
Consider the GDP report, published by the Bureau of Economic Analysis. As we know, economists are already employing “guesstimates” to approximate how bad things will get in Q2. (Note: we won’t even get the first official estimate of Q2 GDP until July 30.)
But that’s just where the problems begin.
“[I]t may be difficult for the Bureau of Economic Analysis (BEA) to capture the degree of weakness given the reliance on surveys and historical interpolations,” Bank of America economist Michelle Meyer wrote on Thursday. “During the last recession, it took several releases with downward revisions before the depth of the downturn was understood.”
UBS’s Paul Donovan also dug into this a bit in a note published on Monday.
“Industrial production, some unemployment numbers, inflation numbers, GDP and the various sentiment opinion polls need people to fill in surveys,” Donovan said. “If you are filling in survey forms in a lockdown you are likely to be an unusual person, and possibly not representative.“
The current crisis could put a spotlight on an existing problem that’s been flying under the radar.
“One of the reasons data quality has fallen in recent years is that fewer people fill in surveys,” Donovan added. “Those that do are less likely to answer all the questions. Surveys are a nuisance to do. In the age of email you can barely leave a shop or hotel without being asked to do a survey. People are fed up of answering questions.“
Keep in mind that you don’t exactly get sent to jail for not filling out a survey. And filling out a survey isn’t the kind of distraction people are looking for as their livelihoods are disintegrating.
“In the current crisis, even fewer people are likely to want to fill in a survey,” Donovan argued. “A business owner is not likely to want to answer detailed questions on retail sales in an economic lockdown. Anyone who does take the time to fill in a survey in the middle of this uncertainty is going to be ‘unusual.’ Right now you probably do not want the opinion of anyone who wants to give you their opinion.”
Assuming people do want to respond to these surveys, one has to question what’s being sampled. This is particularly problematic for inflation data, where leisure industries — which are almost totally shutdown — account for a large portion of what’s measured in the consumer price index (CPI).
“If people are not spending, should the prices be counted?“ Donovan asked.
Lest we forget the people conducting those surveys.
“A strain on resources at the BLS itself may cause delays in collecting and processing information,” Barclays’ Blerina Uruçi said in a March 25 research note. She noted that during the U.S. government shutdown of October 2013, the CPI report was delayed and later published using a smaller sample size.
Of course, the CPI report doesn’t consist of just one big number. It covers 200 categories of goods and services, sampled from across the country.
“The BLS collects data from 32 geographical areas and if at least one good quality price quote were not available from at least 50% of its geographical weight that series would not be able to be published,” Uruçi said. “This does on occasion happen for very small weight sub-indices of the CPI, but we are not aware of this rule ever affecting larger components.
“Given the unprecedented nature of the current shock we think it possible for this clause to be triggered and for the BLS to not publish certain subcomponents,” she added. “But we think that the bar to not publishing the overall CPI index is quite high.”
In other words, we’re likely to keep getting these economic reports we’re used to seeing regularly. But buyer beware.
“For a few months, investors need to be very careful about using economic data in investment models,” Donovan said. “If the data is wrong, models will be wrong. Remember the computing adage—garbage in, garbage out.”
What to watch today
8:30 a.m. ET: Change in nonfarm payrolls, March: -100,000 expected, +273,000 in February
8:30 a.m. ET: Unemployment rate, March: 3.8% vs. 3.5% in February
8:30 a.m. ET: Average hourly earnings, month on month: +0.2% vs. +0.3% in February
8:30 a.m. ET: Average hourly earnings, year on year: +3.0% vs. +3.0% in February
9:45 a.m. ET: Markit US services PMI, March final: 38.5 expected, 39.1 prior
10:00 a.m. ET: ISM Non-manufacturing index, March: 43.0 expected, 57.3 in February
7:30 a.m. ET: Constellation Brands (STZ) is expected to report adjusted earnings of 1.65 per share on revenue of $1.836 billion
Tesla delivers better than expected 88,400 vehicles in Q1 [Yahoo Finance]
Chewy reports strong start to 2020 amid COVID-19 outbreak Yahoo Finance]
YAHOO FINANCE HIGHLIGHTS