Restated FCA sales show some big swings

DETROIT — A review of restated Fiat Chrysler sales at the model level points to a company that either didn't accurately know what it was selling month to month or was playing fast and loose with its fleet sales.

In its unprecedented restatement a year ago, which revealed that its then-vaunted streak of year-over-year sales gains actually had ended in 2013, FCA US​ downplayed the differences between what it had historically reported as its monthly sales and what a new reporting methodology showed was actually occurring.

To take a deeper look at the figures, the Automotive News Data Center rebuilt the model-level data from the sales totals reported starting in July 2016, comparing figures from the old methodology to the new accounting.

The total cumulative deviations were small — less than 1 percent. But at the model level, which FCA did not disclose at the time, some of the variations were quite large.

For example: FCA and its dealers reported more than 15,000 sales of the Chrysler 200 sedan from July 2015 to June 2016 that the automaker now says didn't occur. It also originally underreported the sales of almost 6,200 Dodge Grand Caravans and about 5,800 Jeep Compasses — sales that appeared only after the switch to its new reporting methodology.

The three vehicles come from different FCA brands, but they have one thing in common: a high percentage of fleet sales.

A year ago and in subsequent statements, FCA said the new counting method provides "the best available estimate of the number of FCA US vehicles sold to end users through the end of a particular month applying a consistent and transparent methodology."

Most of the model-level differences between FCA's originally reported sales totals and its restated sales totals were small, with deviations of less than 1 percent of total sales over the 12-month period for each model. But why was the original reporting for FCA's three fleet queens — the 200, Grand Caravan and Compass — so far off?

An FCA spokesman declined to comment. But a source within FCA with knowledge of the system told Automotive News the issue had to do with the way FCA used fleet sales to daily rental companies as a type of "slush fund" to hit monthly sales goals.

Daily rental fleet vehicle sales can be booked at any time, the source said, speaking on condition of anonymity. There was "no need to deliver or invoice" a daily rental fleet vehicle in order for it to be counted as a sale, the source said, so sales to daily rental fleet operators could be reported as needed.

Indeed, in a lengthy note explaining its switch to the new reporting methodology a year ago, FCA admitted that it had been a "historical practice" at the company and its corporate predecessors to maintain a reserve of fleet vehicles "that had been shipped but not been reported as 'sold' in the monthly sales reports."

FCA said that "while the origin of this practice is unclear and is being looked into, FCA US believes that it was probably originally designed to exclude from the reported sales number vehicles that were in transit to fleet customers, as well as vehicles that were not yet deployed in the field." The company said that the reserve pool ranged in size "and resulted from a subjective assessment at month-end," though FCA maintained that it had always reported fewer sales than the aggregate number of shipped vehicles.

The company's new methodology, started in July 2016, reports U.S. sales totals based on dealer-reported sales minus all unwound transactions recorded through the month, retail "other" sales, including those by dealers in Puerto Rico, and fleet sales that are delivered directly by FCA to the customer or end user.

Monthly sales reporting, though traditional in the industry, "obviously is something that is a bit broken," said Dave Sullivan, senior analyst with AutoPacific. Sullivan says the monthly sales totals are gathered from thousands of sources each month, compiled and reported, and as such have plenty of room for error.

Monthly auto sales are not a useful indicator of a company's health, he said, because they measure dealer and fleet sales for which the automaker already has been paid.

"Production numbers and days supply are the real indicators of company health," Sullivan said.