Musk’s pants on fire: September was a record high? Definitely not in terms of Model S sales

September high

September high

(Preface: I know, the Daily Kanban looks like a Tesla fanzine lately. Gomen nasai, but there’s just too much out there to be passed over.)

On October 27, 2014, the Wall Street Journal wrote that sales of Tesla’s Model S are “declining in its home market.” The Journal said that “through September, Tesla sold 10,335 Model S sedans in its home market, down 26% compared with the same period in 2013.” At Tesla, Musk disagreed.

In the rest of the auto business, an errant writer would have been ignored, for fear a reaction could create more waves. Or the reporter would get a phone call to set matters straight (I get those all the time.) Tesla is not like the rest of the auto business, therefore,  the Wall Street Journal got an angry tweet from Elon Musk himself:

tweetchart

Musk’s tweet appeared to reverse a serious downtrend of the TSLA stock. With the all-cl;ear from the boss, TSLA went north again, cheered-on by mondo retweets and favorites from the disciples of the Church of St. Musk. They overlooked that Musk did not refute the Journal’s claim of year-to-date sales being down. Pulling the collective wool over his Twitter followers, Musk said that “September was a record high WW and up 65% year-over-year in North America.” He said September, not the 9 months through September. Easy for him to claim, because Tesla does not release monthly numbers – and who’s got the data to disprove him?

Shockingly, those data are readily available. From Ward’s, where the WSJ got them. Or from JATO, where the Daily Kanban found what everyone should be looking for. Send them a check, they send you the data. Speaking of which, let’s do a sanity check. The Journal’s data, sourced from respectable Ward’s, did show 10,335 Model S sold through September. JATO shows 10,068 Model S registrations in North America for the same period. This indicates that the data are independently sourced, and that they are coming to a very similar results. With a 40 year professional background in car counting, I have high confidence both in JATO and Wards.

Now on to matters that deserve less confidence, let’s look at Musk’s claims of record highs worldwide in September.
Teslaworld-wideregs

Sorry, JATO’s data don’t agree with Musk. JATO shows that worldwide, registrations peaked in March 2014, and that they were down ever since. JATO shows September registrations of about half of the March peak. Tesla’s worldwide trend has been down for most of the year.

(Looking at the chart, one can’t help but notice a strange tendency among new Tesla owners to register their vehicle in the last month of the quarter, whereupon registrations crash in the following month. Just a thought.)

Musk was very wise warning not to read too much into a single month.  However, he should be less patronizing. Don’t treat analysts like children. Most have spent more time in this business than Musk, they are grown up enough to do the required data smoothing. Even numbskulls like us can add a helpful 3 month moving average. However, in absence of data from Tesla, analysts will get their numbers elsewhere. Musk has yet to disrupt the ability to count.

Now to Musk’s claim that the September was “up 65% year-over-year in North America.”

TeslaregsNA
Again, JATO’s data don’t bear that out. JATO has NA registrations down 30 percent in September. Let’s give Musk the benefit of the doubt that he meant “3 months ending in September,” and that he couldn’t fit it in Twitter’s 140 characters. Again, no joy. JATO has the third quarter of 2014 down by 24 percent compared to the same period in 2013.

Hucksters like Musk can get away with telling tall stories, as long as reporters don’t do what they are supposed to do, namely ask questions. Or as long as Mark Rogowsky writes for Forbes, saying Musk’s 65% up claim must be true, and the WSJ must be wrong, because Musk said so. And after all, Musk “would be liable for some securities law violations if his number were shown to be invalid.“ Interesting logic. I’ll try that next time I’m in court.  “I can prove my innocence: If I would have done it, I could be sent to jail!”

As long Tesla refuses to hand out reliable data, the Daily Kanban simply must note that registration data supplied by sources respected throughout the auto industry paint a grossly different picture of Tesla than the tweets of its CEO. We won’t try to speculate about possible reasons for the divergence. In the industry, there always is a difference between reported deliveries to dealers and hard registrations. Dealers sit on inventory, sometimes for a long time. However, Tesla is make-to-order. Musk famously said: “We have no inventory.”
reconciling

In the auto business, the numbers for produced and registered units should equalize over time, especially on a global level, save for a small overhang of crushed cars, lemons, grey imports to markets without data monitoring, etc. Interestingly, at Tesla there is such a happy equilibrium, before 2014. According to Tesla data, approximately 25,000 Tesla cars were made before 2014. JATO shows registrations for 24,837.

Only in 2014 did deliveries and registrations gap seriously. Coincidentally or not, they did so when registrations were in a sudden downtrend. In the industry, large gaps are caused by overproduction and cars sitting on dealer lots. This should not happen at made-to-order, no inventory Tesla. So before shooting the messenger and JATO’s data, ask yourself why the data were good in 2012 and 2013, and why should they suddenly go bad in 2014. If one is, like Rogowsky, “not exactly clear how Wards complied that data,” pick up the phone and ask.

And write “compiled.” Complied would be a Freudian slip.

After doing all that complicated math (which would have been easier and of higher precision, would Tesla be more forthcoming with data), what we know is this:

  • Musk’s claims of record sales in September cannot be independently confirmed.
  • Tesla registrations have peaked globally in the first quarter of 2014
  • As of September, more than 4,000 cars were reported as sold, but not as registered.
  • About half of the missing cars supposedly are in China, no idea where. Where the other half is, nobody seems to knows either.

As for the errant Chinese Teslas, there is a readily accepted theory bouncing around in the echo chambers of the Internet: they are being hoarded by Chinese “scalpers” who stash them away for resale at a mark-up. Even at Tesla, that story has been discounted. Scalpers have “not been a significant factor in China,” Simon Sproule told the LA Times when he still was spokesman at Tesla. Actually, there should be more Model S registered in China than officially imported by Tesla. There has been a serious amount of grey importing, a matter so pressing that Tesla denied warranty coverage on rogue Model S brought to China. These cars should add to Chinese registration numbers, and they should be missing in California from where they are brought in.