Stick a fork in the now proven wrong theory that plunging gas prices would boost consumer spending. Why? Because 4 months after the full impact of tumbling gas price was said to become apparent, consumer spending is not only not picking up, it is in fact slowing down more, especially in those places where there was snow in the winter, and gasp, where oil price actually fell the most!
Ridiculous? No. This is what the latest Bank of America card data survey reveals. To wit:
"Consumer spending remained sluggish in February, according to the Bank of America card data. After netting out volatile gasoline and autos, spending on Bank of America credit and debit cards was unchanged MoM seasonally adjusted. On a YoY basis, sales are still up a solid 5.4%, but that reflects favorable comps from last year."
How is it possible that everyone could have been so wrong? Simple: for the second year in a row, nobody could possibly anticipate that the "gas savings" boost to consumer spending in the early winter would be fully (and then some) offset by an utterly inconceivable heavy snowfall in January and February. Because seasonal adjustments are there only for show, and to smooth out goalseeked lines presumably, not to - you know - adjust for the seasons.
The BofA scapegoating is just humor in its most purest:
We can explain the weak MoM rate, in part, from the harsh winter weather in February. Excessive snowfall in parts of the country left many households stuck at home, reducing spend on their credit and debit cards. Of course, we attempt to control for seasonality, but it is difficult to capture abnormal weather patterns.
Wait what: seasonally-adjusted data is unable to capture the seasons? Please do go on...
But it gets better, when BofA tries to prove that it wasn't its fault its models predicted something totally different than what actually happened.
We can test this hypothesis by examining spending by the largest metropolitan areas (MSAs). As we show in the Chart of the Month, retail spending ex-gasoline and autos tumbled in Dallas, Charlotte and Boston. To control for noise in the data, we take the three-month moving average – simply looking at the change in February, however, would reveal a particularly sharp decline in Boston. The Boston MSA had record snowfall, while Charlotte was hit by late winter storms.
But the absolute punchline is this: recall that the primary thesis of the "plunging oil is unambiguously good" herd of penguins was that lower oil and thus gas would boost consumer spending. Well, apparently that is the case everywhere else... except where the oil plunge is strongest!
Dallas also dealt with unseasonably poor weather, but we also suspect the weakness could be a reaction to the decline in oil prices, which has curbed investment in the oil industry, hurting the local economy.
And back to square one, as BofA concludes its models were right; it was reality that was wrong.
If not for the unfortunate weather, we think consumer spending would have been stronger. The backdrop is favorable: solid job growth, improving consumer sentiment and savings from lower gasoline prices. We remain constructive, forecasting a healthy acceleration in consumer spending in the coming months.
With that load of humor behind us, here is something actually useful: actual data charted:
[9]
And the hits just keep on coming, because the retail sales ex autos and gas was the worst since January of 2014!
[10]
- Looking past the effect of gasoline prices still reveals a slowing trend in consumer spending ex-autos, as can be seen by the blue bars in the chart.
- After a solid gain in October, sales ex-gasoline and autos have weakened. The data in February was particularly sluggish even relative to the trend.
- We have been surprised by the weak performance of spending ex-gasoline and autos given the improving consumer backdrop.
But... but.. the model said so...
- The trend in consumer spending has slowed across the sectors, as measured by the three-month average of the MoM SA change.
- Restaurant sales still look relatively stronger, with sales at fast casual dining and quick service restaurants remaining positive, albeit slower than the average monthly growth rate last year.
- Sales at furniture and clothing stores have weakened, reflecting a decline in demand for bigger ticket items.
[11]
- The trend in retail spending, as defined by sales ex-autos, ex-gasoline, continued to weaken in Texas. Sales are running at 2.8% yoy in February, a notable slowdown from the pace of 4.5% last summer.
- For the first time since late-2010, spending in Texas is running at a rate that is slower than the rest of the country.
- We have seen other signs of weakness in Texas – initial jobless claims, for example, have spiked. We are concerned that the decline in oil prices has already started to weigh on the local economy.
[12]
- The seasonally adjusted retail sales ex-autos aggregate from the BAC internal data was unchanged MoM in February. This left the three-month moving average to fall 0.8% mom SA.
- Nearly all of the recent contraction can be explained by a decline in spending at gasoline stations as a result of falling prices over the past several months.
- BAC card spending has trended very closely with the comparable measure from the Census Bureau, suggesting the Census data, released on Thursday, should show a similar trajectory.
* * *
But wait, surely the money "saved" from the lack of spending on gas, if only for the past 3 months now that gas prices are screaming higher, was spent on something? Yes, precisely. And we explained precisely what that something was in "Here Is What Americans Spent Their "Gas Savings" On [13]"
[14]
And with that we consider the "mysterious" case of the missing US spending, closed.
