Orlando's Outlook: Skunk at a picnic


Bottom line Nominal retail sales in August were surprisingly much weaker-than-expected, and their 0.2% month-over-month (m/m) decline marked the softest results in six months, with downward revisions for June and July. So what went wrong? After a weak June but a good July, whose preliminary m/m gain of 0.6% was the strongest of the year, we were expecting this positive consumer-spending trend to continue into August, contributing to what we thought would be a solid Back-to-School (BTS) season. Like the proverbial skunk at a picnic, however, Hurricanes Harvey and Irma arrived in the last two weeks of August, impacting auto, gasoline and building-material sales. Moreover, we believe that the back-to-back presence of these two “once-in-a-hundred-year” storms created a “CNN effect,” as many Americans were glued to their screens watching the tragic devastation and its aftermath rather than shopping. The Federal Reserve, for example, recently commented that Harvey alone probably reduced GDP by 0.75% in the third quarter.

Aside from the storms, there were two other significant issues that negatively impacted August retail sales results. Amazon’s annual Prime Day on July 11 sent e-commerce sales soaring 1.8% in July, but August plunged 1.1%, providing some payback, as some sales were likely pulled forward. In addition, electronics sales have been sharply negative m/m for four consecutive months through August as some consumers have been patiently waiting for Apple’s new iPhone launch, which finally came this week.

True, this August weakness in consumer spending during the important BTS season raises critical questions about important fourth-quarter holiday spending trends. But we believe that the underlying strength in the economy, marked by solid consumer confidence and a good labor market, will be goosed by post-hurricane rebuilding activity, as well as a resumption of strong electronics and e-commerce sales during Christmas.

BTS is still the best in three years thus far in 2017 August’s surprisingly sloppy 0.2% sequential decline in nominal sales, combined with July’s downwardly revised increase of 0.3%, still equates to average retail sales for the two months of $475.3 billion, a solid 3.3% year-over-year (y/y) increase over the same two months in 2016, which represents the best start to the BTS season in three years. The combination of July and August had risen only 2.3% in 2016 and 2.7% in 2015, but they rose a healthy 4.7% in 2014.

Waiting for September To be sure, we need to see if September’s retail sales begin to rebound from poor August results before we close the books on the 2017 BTS season. Why is BTS 3-months long? Because July’s BTS shopping is for schools that open in August, which is still the heart of the season as students and parents prepare for September class starts. But savvy shoppers typically hold back part of their BTS clothing budget until they see what the popular kids are wearing, and then purchase those items on sale over Labor Day weekend. So while this year’s storms threw a monkey wrench into the typical shopping and spending patterns, we still need to see September results to fully evaluate the overall strength of this year’s BTS season.

BTS remains a window into Christmas sales Barring extreme weather issues—which we’re obviously dealing with this year with Harvey and Irma—Christmas retail sales are usually 80-90% positively correlated with BTS results. The major national retailers and shipping companies have to look through this weather noise to establish their holiday hiring plans now. Target, for example, plans to increase its hiring by 40% this year, adding 100,000 seasonal workers (compared with 70,000 last year), with another 4,500 seasonal workers for its distribution and fulfillment centers (versus 7,500 last year). Macy’s plans to hire 18,000 holiday workers to fulfill online orders this year, a 20% increase. But total holiday hiring will slip to 80,000 this year from 83,000 last year, although Macy’s did close 70 unprofitable stores over the past year. Because consumer spending accounts for 70% of GDP, holiday sales are an important driver of fourth-quarter economic growth.

Importantly, confidence metrics remain strong:

  • Leading Economic Indicator (LEI) has been positive for 11 consecutive months, rising 0.3% in July 2017 m/m to 128.3, a new 58-year cycle high.
  • Conference Board’s Consumer Confidence Index soared to a 16-year high of 124.9 in March 2017, up from a 3-month low of 100.8 in October 2016, and hit a 5-month high of 122.9 in August, its second-highest level since 2000.
  • Michigan Consumer Sentiment Index spiked to a 12-year high of 98.5 in January 2017, up from a 2-year low of 87.2 in October, and it hit a 3-month high of 96.8 in August. September’s preliminary reading, released this morning, slipped to 95.3.
  • The National Federation of Independent Business (NFIB) small-business optimism index roared to a 12-year high of 105.9 in January 2017, up sharply from a cycle trough of 94.1 in September 2016, and it hit a 7-month high of 105.3 in August.

Labor market quirky but solid To be sure, August was a disappointment across the board, as nonfarm payrolls missed consensus estimates by 24,000 jobs (with downward revisions totaling 41,000 jobs in June and July) and household (74,000 jobs) and government hiring (9,000) both fell. But August tends to be the statistically quirkiest month of the year for the labor market, with sizable underperformance when the initial report is released, followed by strong upside revisions during the next two months. Moreover, last month’s weakness was completely inconsistent with broad-based strength elsewhere in the labor market. JOLTS hit an all-time high of 6.17 million job openings in July, August’s ADP report for private hiring was the strongest in five months at 237,000 new jobs, and initial weekly jobless claims for the August survey week at 232,000 remain just off a 44-year cycle low. Harvey and Irma certainly will spike jobless claims temporarily. But the underlying strength of the labor market remains solidly in place.

August surprisingly weak, July solid and June soft Digging into the weeds on this morning’s retail-sales report for August, here are the pertinent details:

  • Nominal retail sales negative in August and June Headline retail sales for August fell 0.2%, compared with consensus expectations for a modest 0.1% increase. July, which had flashed a powerful 0.6% m/m preliminary gain, was revised down this morning to a more modest increase of 0.3%. June’s 0.3% increase was revised down to a final decline of 0.1%. During August, e-commerce sales fell 1.1%, clothing declined 1.0%, electronics dropped 0.7% and building materials lost 0.5%.
  • Ex-auto sales positive August’s core retail sales, which strip out volatile auto results, rose 0.2%, but that was well below consensus estimates for a 0.5% increase. According to the Department of Commerce, auto sales plunged nearly 4% in August on a m/m basis to 16.03 million annualized units, which translates into a 1.6% m/m decrease on dollar volume. July retail sales ex-auto results slipped a tick to an increase of 0.4%. June was revised from a modest gain of 0.1% to a final decline of 0.2%.
  • Ex-autos & gas weak During Hurricane Harvey in Houston, where extreme flooding shuttered some 30% of the nation’s refining capacity, unleaded gasoline prices rose 15% nationally, from $2.33 to $2.67 per gallon, although prices spiked by some 50-70 cents per gallon in some other parts of the country. The higher prices caused dollar volumes at gasoline stations to soar 2.5% during August. So adjusted core retail sales, which strip out both auto and gasoline sales, slipped 0.1% in August, well below the expected increase of 0.3%. July results were unrevised at a strong gain of 0.5%. June was revised down from a gain of 0.3% to a final decline of 0.1%.
  • ‘Control group’ disappoints, pressures GDP Finally, so-called “control” results—which strip out autos, gasoline, building materials and food service, and feed directly into Commerce’s quarterly GDP calculation—fell 0.2% in August, below the expected consensus gain of 0.2%. July’s results were unrevised at a strong increase of 0.6%, its best showing in four months. But June was revised down from a modest gain of 0.1% to a final decline of 0.2%. So the storm-induced, negative control readings from August will ameliorate July’s strength in the third quarter’s GDP calculation. We recently reduced our estimate by a tick to 2.4% to potentially account for the negative impact from the storms. But if we fail to see a rebound in September, then that estimate may have some downside pressure.

Connect with Phil Orlando on LinkedIn