The exuberant images this weekend of lines-around-the-block at Apple stores were met with triumphant flashing red headlines this morning when Apple announced the sale of more than 10 million iPhone 6 and 6 Plus models (more than expected). Typically, new product launches do not move the needle on aggregate US economic data. Apple’s iPhone has been the most notable exception, with past launches occasionally having a substantial effect on core retail sales. However, Goldman notes, with the launch of the new iPhone 6/6+ this month, estimates (based on historical data) of a 0.1 to 0.7ppt boost to September core retail sales is highly uncertain due to seasonal adjustments that have been highly erratic, and could easily take a big bite out of the Apple effect.
- *APPLE HAS SOLD OVER 10M NEW IPHONE 6 AND IPHONE 6 PLUS MODELS
- *APPLE SAYS SALES FOR IPHONE 6 & IPHONE 6 PLUS EXCEEDED VIEWS
- *APPLE'S COOK: COULD HAVE SOLD MORE IPHONES W/ GREATER SUPPLY
- *APPLE SAYS SALES FOR IPHONE 6 & IPHONE 6 PLUS EXCEEDED VIEWS
Via Goldman Sachs,
Typically, new product launches do not move the needle on aggregate US economic data. Apple’s iPhone has been the most notable exception, with past launches occasionally having a substantial effect on core retail sales. In particular, iPhone sales show up in two categories of the report: electronics stores (representing in-store sales) and "nonstore retailers" (representing online sales). Exhibit 1 shows the behavior of these two categories of retail sales during the release month for the iPhone 4S, iPhone 5, and iPhone 5S/5C, both on an initial print (i.e. as-reported) basis, as well as the final revised estimates. The iPhone 4S and 5 launches showed up clearly, particularly in the initial print data. The iPhone 5S/5C launch did not appear to boost retail sales as notably, perhaps due to (1) the lack of a preorder period for the 5S, and (2) the fact that the 5C was not as popular as initially hoped, with many retailer orders heading into inventory.
[10]
Preorders for the iPhone 6/6+—which began on September 12—have been record-breaking, at a reported 4 million units in the first 24 hours (up from 2 million for the iPhone 5). Our IT hardware equity analysts project 8 million units for first weekend sales, which will probably be released on September 22. Exhibit 2 shows how past company sales metrics have mapped into the initial print retail sales data. Because the Census Bureau does not require standardization of how retailers report cell phone sales (by retail price, discounted price with a cellular contract, etc.), it is not possible to directly estimate the retail sales effect with a simple price x quantity calculation. In addition, preorders and opening weekend sales are reported on a global, rather than US-specific, basis. However, using the simple units to sales "translation" from the iPhone 4S and the iPhone 5, we would expect a boost of about seven-tenths of a percentage point to the September core retail sales figure. On the other end of the spectrum, looking at the experience of the 5S/5C—likely less comparable than the other examples to the 6/6+ rollout—one might expect a small boost of only one-tenth of a percentage point.
[11]
Of course, once first weekend sales are available, we may be able to refine this figure. Notably, China will not be included in first weekend sales due to delays with regulatory approval, meaning that any given unit level of sales probably represents more US sales than historical comparisons would suggest. However, the level of sales reported by Apple is not the only variable in play. The retail sales data are reported on a seasonally-adjusted basis, and the idiosyncrasies of seasonal adjustment can often be a source of volatility. Specifically, the expected September seasonal for nonstore retail has been highly erratic (Exhibit 3). Although the Census Bureau provides indicative seasonal factors for data not yet released—shown as striped columns at the far right of the exhibit—these are subject to considerable revision at the time the data is reported.
[12]
Nonetheless, the indicative seasonal adjustments provided by Census do look like a tougher hurdle this year than in prior years. This could reflect a number of factors: (1) the tendency for fall product launches from Apple may be starting to be incorporated into "normal" seasonality, (2) the early timing of Labor Day could be a factor, or (3) shifting seasonality in areas completely unrelated to the iPhone could be driving the higher expected seasonal gain. The first possibility would represent the most downside risk to our earlier estimate of a 0.1 to 0.7% boost to September core retail sales. At this point, it is too early to have an explicit forecast for the September retail sales report, and we will refine our estimate of the iPhone effect further in coming weeks, but our early analysis points to September upside relative to the recent trend on core retail sales.
