I like SAARs, those awkwardly named Seasonally Adjusted Annualized Selling Rates the U.S. auto industry uses to compare the relative strengths of various months of light vehicle sales.
I prefer SAARs over actual units to compare, say, May of this year against June of 2010. We all know that some years are simply better than others. And in the same way, some months of any year are usually better than others. April is almost always better than any January, for example.
Units have precision. SAARs have flavor. They give us a feel for how a month went.
SAARs are the preferred tool for comparisons and most of the auto industry, including the Automotive News Data Center, uses the same mathematical model as the U.S. Bureau of Economic Analysis does, as a shared standard.
But like any tool, there's maintenance. Especially one that is based on a moving multi-year average of monthly sales.
I mean, we've had a pretty volatile sales climate these past five years.
And every July, the BEA recalculates the seasonal factors, and everybody updates their model. And recalculates past monthly SAARs, this time back to January 2009.
What's all this mean?
Not that much in practical terms, even for the biggest revisions. Most times, it smoothes out the jagged parts of the line charts.
Remember our hot-hot-hot February, the one with the 15.1 million SAAR? Well, the industry still sold 1,149,470 light vehicles. But the new SAAR for February is 14.5 million.
That's still a half-million higher than in January (now recalculated downward to 14.0 million from 14.2 million), a big jump. But not almost a million higher.
Yeah, that kind of change is annoying for an industry that loves absolute values. But it's still only a tool to make comparisons.