August 13th, 2012
The Reuters consensus estimate for U.S. retail sales growth in July calls for a 0.3% monthly advance, insufficient to fully offset the 0.5% decline between May and June but the first increase in four months. The median among primary dealers is even lower, however, at 0.2%. While the range from 72 contributing economists extends from -0.2% to +0.6%, only three primary dealers submitted a headline retail sales estimate above +0.3%, while 11 estimates were called for growth of 0.2% or lower. After three straight negative readings and two downside misses, a moderate rebound is not just a welcome relief but a necessary condition for many dealers’ outlooks calling for acceleration in Q3 and H2 economic output.
For economists at J.P. Morgan Economic Research, retail sales is this week’s key upcoming release. “The most important release bearing on current-quarter growth will be July retail sales. The forecast looks for a 0.2% increase in total sales and a 0.3% increase in core retail sales that, if realized, would be above the average for the past three months (-0.1%) but below the average for the previous year (0.4%). Digging into the sources of the supposed gain, JPM wrote, “The decline in gasoline prices in the prior few months has weighed down the nominal values for sales at gasoline stations in recent reports, but we think the increase in gasoline prices in July (using our own seasonal adjustment) will push up sales at gasoline stations during the month (+0.3%). We forecast that retail sales at motor vehicles and parts dealers declined 0.7% in July largely based on the decline in unit vehicle sales already reported by the BEA for the month.” Excluding autos and gasoline, JPM estimates retail sales increased 0.4% in July. Data on chain store sales were fairly upbeat for the month, and we believe sales at building materials dealers increased 0.7% in July after plunging the prior three months.”
Economists at Nomura Economics Research (e: 0.0%) agree that July retail sales data will set the tone for Q3 spending but they think that tone is unexciting. “We expect lower gas prices and fairly flat auto sales in July to contribute to weakness in headline retail sales. Moreover, growth in chain store sales remained moderate over the month.” Nomura’s forecast does not improve much after stripping out the contribution from autos, though without gasoline, autos and building materials, they’re looking for the best rise in four months. Their preview continued, “Excluding autos we expect retail sales to have increased by 0.1% (compared with a decrease of 0.4% in June) but we forecast an increase of 0.6% in so-called ‘control’ retail sales in July.”
Barclays Research wrote in this week’s Global Economics Weekly, “We are looking for a stronger tone to core sales data to emerge relative to Q2, consistent with a modest pickup in real consumption growth. We expect retail sales to increase by 0.2% m/m in July. Within this we expect a small increase in gasoline sales (in line with the modest rise in prices during the month) to be partly offset by slightly softer auto sales (consistent with the drop in unit sales reported by the main suppliers). Elsewhere, we are also looking for a solid 0.3% gain in ‘core’ sales. This would be the strongest since March and consistent with our forecast for consumption growth to pick up somewhat in Q3 relative to Q2.”
The AlphaWise Macro Team at Morgan Stanley US Economics Research wrote Monday that real-time data are closer to the ex-autos consensus estimate (e: +0.3%) than to their own desk forecast (e: +0.5%) submitted on Friday. “Based on our Real-Time Tracker, we expect Census Bureau retail sales (ex-autos) to show a 23-bp increase in July, versus consensus of +30 bps. Ex-auto ex-gas, our model suggests a 34-bp increase versus consensus of +50 bps.” The deviation from consensus (i.e., data surprise) may generate a reaction in 10-year Treasury note prices but the team is not issuing a call. “Our retail ex-auto model’s standard error is 32 bps. We make calls and trade recommendations when there is a delta from consensus of more than 30 bps. Because we are only seven basis points below consensus, we do not have a call.” But they did provide a schedule of expected market impacts. “Our tick data analysis of Retail Sales beats / misses (5-year history) suggests an 11-bp average price reaction (or 1.3-bp average yield reaction) in 10-year US Treasuries.”
Citi Research supposes some of the weakness in consumer spending in the second quarter “may reflect the larger share of discretionary categories in retailing versus overall PCE where the mild winter may have robbed sales from the spring season. With the recent lift in financial conditions and more widespread coverage of stabilizing home prices, perceptions about personal finances may have brightened in the past month and July’s better employment news ought to help as well. This sets the stage for a potentially pivotal reading on retail sales in the coming week. We expect a rebound in sales for July of about 0.3% for both total and core helped in part by healthy gains at large retail outlets, buttressed by a rise in gasoline prices”. As expressed by other dealers, a moderate gain in July retail sales is crucial to Citi’s thinking about broader near-term trends. “Three straight declines are rare but do pop up in an economic expansion given seasonal sensitivities and sampling problems. Nonetheless, a fourth consecutive decline, especially in core categories would raise serious questions: We would likely have to brace for weaker output, rising jobless claims and slowing in employment beyond August.”
Deutsche Bank Fixed Income Research tempered its otherwise upbeat assessment. “Our forecast of second half real GDP growth outperforming the first half of the year is largely contingent on an improvement in consumer spending, which is why Tuesday’s report on July retail sales will be of elevated significance. Consumer spending slowed in Q2 largely due to slowing consumption of motor vehicles and parts (-10.7%).” DBGMR points out, however, that spending on services (+1.8%) grew at the fastest pace in a year. They believe the pullback in spending on autos is temporary “because commercial banks are making it easier for consumers to obtain auto financing and households are saying that it is currently a good time to purchase a motor vehicle.” DBGMR similarly believes the near-term outlook for non-auto consumption has recently improved and the snapback in chain-store sales bodes particularly well for ex-autos July retail sales. “According to our calculations, July total chain-store sales were up +8.5% compared to the same period last year. This is the second strongest growth rate this year, behind March’s +8.8% increase. Importantly, July chain-store sales are up appreciably relative to June, which experienced just a +4.2% y-o-y advance. In general, the trend in chain-store sales tends to be highly correlated with retail sales excluding autos; the correlation coefficient between the two series stands at a high 0.71. We conservatively estimate a sequential increase in July ex-autos retail sales of 0.5%. While we think the risks are tilted toward an even stronger number, we are hesitant to forecast a larger gain for July retail sales, because this series is prone to significant revisions in subsequent releases. It is not unusual for large gains in the monthly retail sales data to only become evident after revisions.”
Nearly all dealers consider U.S. July retail sales to be the key domestic data release this week, and with the possible exception of euro zone GDP data also out Tuesday, retail sales may be the most significant item on this calendar week. Disagreement about the role of gasoline prices explains some of the variation in headline estimates but there are similar divisions on the core and control group measures. Regardless, following three straight monthly declines in headline and ex-autos measures, a moderate bounce in retail sales is more than long-awaited, it is required for dealers to maintain their already modest Q3 and H2 growth outlooks.