One of the most amusing conundrums of the great Polar Vortex economic collapse of the first quarter of 2014, in which Q1 GDP was initially supposed to rise by 2% only for it to plunge by over 2% as a result of harsh snow (in the winter) was the relentless series of economic data "misses" where for some 2 months in a row, nearly all the releases were one after another miss of consensus sell-side expectations (in fact the only time this series of misses has been worse is this year, in 2015!). What happened subsequently is that after every single miss, the economists would simply say "blame it on the weather", making one wonder just what does the word "seasonal" in "seasonal adjusted" stand for.
A perfect example of this is the Twitter feed of DB's permabulllish cheerleader Joe LaVorgna, whose series of tweets blaming the weather for this, that, and everything else, is now enshrined in internet folklore [8]:
[9]
Of course, such things as snowfall in the recent past should make Wall Street's economists - all supposedly smart, intelligent, 7-figure paid creatures - factor that into their upcoming forecasts and adjust these accordingly. Paradoxically, they never did and instead used the weather as an excuse for their gross incompetence at forecasting the future (as they are doing again this year).
At least they did until now, because moments ago that bank that is usually at the forefront of original thought (like how to sell our clients preselected securities that we know will fail and we can short them in advance... and the clients too), had the brilliant idea of looking outside its office at 200 West and realizing that, gasp, it's snowing.
And then a lightbulb went over the giant mollusc head, at which point Goldman had a brilliant idea: "if we have blamed the snow for everything so far, shouldn't we also blame it for the most important data point to come - tomorrow's nonfarm payrolls?"
And sure enough, it released a note prewarning that tomorrow's jobs data will likely be weaker than the consensus of 235K, because, drumroll, it's snowing.
From the report:
We expect nonfarm payroll job growth of 220k in February, below the consensus forecast of 235k. Labor market indicators were mixed in February, and we expect that the effect of four major snowstorms in the month leading into the February survey week will also weigh on payroll growth. We expect a one-tenth decline in the unemployment rate to 5.6%, reversing the increase seen last month. On average hourly earnings, our baseline expectation is for a 0.2% increase, but we see some upside risk from possible statistical distortions related to the severe snowstorms.
Among the factors Goldman lists as arguing for a "weaker report", it lays out the ISM and PMI, job availability, job cuts and of course the "weather."
Weather. Our method for estimating the impact on payrolls of snowstorms and temperature deviations from seasonal norms suggests a hit from harsh weather conditions in February. The Regional Snowfall Index recorded four major snowstorms during this period, and we expect these storms to have a substantial impact on activity. Temperatures remained warmer than usual for the country overall in the four weeks leading into the February survey week. While not relevant for the February report, the final two weeks of February saw the two most extreme (coldest) weekly deviations from average temperatures in the 18-year history of the data, suggesting that weather conditions may be a factor in next month's report as well.
It gets better: according to Goldman there is a close correlation between snowstorms and hourly earnings. Yup. You read that right.
We see potential upside risk to reported average hourly earnings in February from the impact of snowstorms. Similar snowstorms last year distorted reported earnings upward through a composition effect that arises because reductions in paid-for hours predominantly affect lower-paid hourly workers as opposed to higher-paid salaried employees. However, we have not found strong evidence historically that snowstorms have been associated with jumps in reported average hourly earnings.
But why: everyone knows that the most contentious wage negotiations always take place during a blizzard?
Laughter aside, we salute Goldman for at least doing what no other weatherman, pardon, economist on Wall Street is willing to do: i.e., factor in the publicly available data, and realize that sometimes, not too often, reality impacts the BLS' C:\economy\jobs\goalseek.xls spreadsheet.
As for all the other penguins, expect them all to blame the weather, especially if like the ADP report, tomorrow's NFP comes in at or well below 200K. Incidentally based on our own tracker, which also takes into account such things as reality into account, the energy sector collapse, the massive job losses in the shale sector, and west coast port shutdown, and the substantial slowdown in the auto sector now that the subprime boom is officially over, means we expect a jobs print under 100,000K, in line with what final Q1 GDP will be: just about 1.5% or less.
Finally, don't be surprised if the BLS comes up with a "seasonally-adjusted" jobs number of over 300,000 tomorrow. After all, when it comes to the BLS, everyone knows by now that all the numbers are goalseeked and serve merely to fulfill a political agenda, which means Arima X 12 adjustments galore. Indicatively, the average seasonal boost to the unadjusted number over the past decade is 1.5 million jobs.
[10]
Or said otherwise, the actual adjusted jobs number is about 10-15% of just the seasonal adjustment itself!
