The U.S. auto industry is swimming in data. At the recent press conferences held prior to the opening of the North American International Auto Show, speakers from every OEM recited statistic after statistic to bolster the reputation of their brand or company. At the start of each month, the OEMs provide the press and public with a blitz of sales results. These data can be fascinating and they provide industry followers with a quantitative method by which to understand how the industry and its various players are performing.
But, in both the recent set of press conferences and the monthly press releases summarizing prior month results, the manufacturers are intent on making their results look as good as possible. And in this quest the OEMs occasionally cite data that convey an image slightly at odds with the actual results. Here are a few things to be wary of when looking at automotive sales data:
- A major year-over-year sales increase for a model may be a legitimate result reflecting increased demand for that vehicle, but that increase may be because of exceptionally low results in the prior year. If that is the case, the model now may just be getting back to its "normal" turn rate. Two examples:
- Dodge Durango new registrations November 2011 CYTD are 45,332, up more than 24,000%, but for the same time period in 2010, Durango registrations were just 185.
- Volvo S60 new registrations through 11 months of this year are 18,286, up almost 3,000%, but, this model’s new registrations during the same period a year ago were just 593.
In both cases, the models of a year ago were going through changeover to new designs so their production and availability was non-existent or exceptionally low. The data are correct, but they can give a misleading impression unless one digs beneath the surface.
- A major year-over-year increase also may be because the model was launched part way through the prior year, so the prior year results are for a different (shorter) time period than the results for the current year. This inflates the current year results. Two examples:
- The Buick Regal, through 11 months this year is showing a 318% increase in new vehicle registrations, but it was launched in May of 2010, so its results this year are for 11 months while its registrations in 2010 are for just 8 months. If one compares the May – November results in both years, the Regal still did very well, with a 163% gain, but that is about half the reported results.
- The same logic applies to the Ford Fiesta. It was also launched in May of 2010, so again the year-over-year results are not comparing results for the same time periods. When the time periods are adjusted to compare apples to apples, the Fiesta is again doing well, with registrations up 106%, but this is less than half the 271% increase reported by Ford.
- Occasionally a OEM will report sales results for a given month and compare those with sales in the prior month. This is not comparing apples with apples, as each month of the year is subject to its own unique seasonal influences; the only proper comparison is sales in any given month with sales in the same month of a year ago.
- Low use of incentive data is often cited as evidence of strong natural consumer demand. For example, a manufacturer may say that its average incentive spend per unit in a month was less than that of any other manufacturer. Incentive spend is usually given in a dollar amount. Because of this, the average incentive spend per unit needs to be reviewed relative to the average price of a manufacturer’s product portfolio. A manufacturer with a relatively high average price per model may have a relatively high incentive spend per unit, while an OEM with a lower average price may have a lower incentive spend amount. In both cases, the incentive as a percentage of the transaction price is the same. OEMs with lower average prices tend to mention lower incentive spending per unit, but they can tout this partially because they do not have higher priced products in the portfolio. To overcome this issue, one needs to look at incentive spend as a percentage of average transaction price. This can be done across OEMs and is truly comparing apples to apples.
- Lastly, OEMs will use specific words that include or exclude certain categories, depending on the stories they want to report. For example, a manufacturer may tout that it sells more "midsize sedans" than any other OEM; this is usually a sign that when all "midsize cars" are included, that manufacturer comes up short. Another example of the use of words is when an OEM reports that it has the best performance among "full-line manufacturers"; this implies that that the OEM may not be at the top when all manufacturers are included. The use of words is important and needs to be examined.
It has been said that data are the lifeblood of the auto industry. And there's no question it's fun, interesting and informative to follow the industry via the data. But if one doesn't look beneath the surface, one can get a picture somewhat different from the underlying, natural landscape.
Posted by Tom Libby, Lead Analyst – North American Forecasting, Polk (01.23.2012)