"Tesla Drops as Autodata Estimates May Sales"

[Todd Burch]

Autodata is estimating May sales at 1,425 Model Ss. How do they get this information? It doesn't seem right.


http://www.thestreet.com/story/11940012/1/tesla-drops-as-autodata-estimates-may-sales.html?puc=CNNMONEY&cm_ven=CNNMONEY

 

[Cosmacelf]

Even if it were accurate, you should take into account that Tesla is putting a lot of cars into their service centers right now, so May may very well be a bit lighter of cars SOLD. They are likely still producing lots of cars though. Anyways, that number still means Tesla will easily meet its 21,000 cars sold in 2013 number.

Buying opportunity

- - - Updated - - -

BTW, Elon has been dropping GIANT HINTS that order rate is higher than the now 21,000 units they said they were going to sell in 2013. So I personally am discounting the usefulness of that Autodata report.

 

[Zextraterrestrial]

The orders seem insane from just being in Menlo for a few minutes. I saw/heard a few people finalizing what they wanted and several on test drives.

 

[CapitalistOppressor]

I haven't bothered to estimate May production yet because there isn't enough fidelity to the VIN's to be really accurate. And I don't think Autodata is plugged into national DMV databases such that they can compile estimates from registration data that quickly. (maybe I'm wrong)

Going off of memory, our April data that was discussed here on TMC gets you ~1325-1425 cars depending on how many loaners were built. I think someone at Autodata just read that and projected forward, the same way folks were doing with reports of 500 cars/week.

A couple of weeks ago every indication was that production was still rolling along at ~400 cars/week, which should get you ~1,800 delivered cars, with up to 100 of those being loaners (we know that process is essentially complete now, but don't know how it is split between April and May). Without having looked at the data 1,425 just doesn't seem credible.

- - - Updated - - -

The orders seem insane from just being in Menlo for a few minutes. I saw/heard a few people finalizing what they wanted and several on test drives.

I was at the 3rd Street Promenade location in Santa Monica recently, and it was buzzing.

- - - Updated - - -

Just glanced into the delivery thread. Just reading the last couple of weeks worth of posts (the data is admittedly sparce) and doing eyeball calculations using nothing but the Mark II Noggin, everything is still consistent with ~1,800 deliveries in May. It is totally inconsistent with any estimate of 1,425 unless you are talking about a production shock like you saw when they switched to Multi-Coat Red, that caused them to jump 400+ VIN's ahead of the sequence.

My guess is that Autodata is just making it all up.

- - - Updated - - -

I should also add that if you look at future delivery dates from folks with a "date certain" (again the data is really sparse) you again see a trend of ~400 cars/week projected forward. Autodata would need to justify their methodology to me before I accepted that over VIN data from actual deliveries.

 

[Johan]

I read this Autodata report and instantly called BS on it. As I saw the stock dropping for no other apparent reason (I know there are a million other reasons) I bought some short-term calls. I think Elon will correct this misinformation at today's stock holder meeting.

 

[yobigd20]

Yea that Autodata report is about as accurate as Faux Business reporter's Elizabeth MacDonald's conjured mystery numbers last Thursday. It's amazing that these people are paid real money. Honestly. If that's all it takes to be a news reporter and all I have to do is make up doody and pull numbers out of my arse, I'm going to quit being a software engineer and start working for faux. lol.

- - - Updated - - -

 

[keys]

Or it might be that they're trying to push the stock price down for their benefit. Not sure if that's legal though?

 

[ItsNotAboutTheMoney]

Or it might be that they're trying to push the stock price down for their benefit. Not sure if that's legal though?

Hmm, puncturing stock? I suspect it's very illegal.

Autodata will have access to quite a bit of Registration and Title data simply because there are states that sell the data or allow it to be sold. I don't think Tesla will correct anything since they only release sales numbers quarterly. They can't fall into the correction trap.

 

[Jeeps17]

... doing eyeball calculations using nothing but the Mark II Noggin ...

A bit OT, but I am curious how you got your Noggin upgraded to Mk2.

:smile:

 

[Zythryn]

I could see it as Europe is gearing up there will be a lag between production and delivery.
Not sure if the timing works out though for that to be the explanation.

 

[Right_Said_Fred]

I could see it as Europe is gearing up there will be a lag between production and delivery.
Not sure if the timing works out though for that to be the explanation.

Production for Europe wasn't supposed to start before June, so that would not be an explanation (for what can now be considered misinformation; comparable VIN-reports still show a production rate of 1,700 cars per month over the last two months).

 

[CapitalistOppressor]

A bit OT, but I am curious how you got your Noggin upgraded to Mk2.

:smile:

I stayed at a Holiday Inn Express last night. (that's the literal truth )

 

[CapitalistOppressor]

Ok, I looked at the Autodata spreadsheet. Still not certain what their methodology is for collecting the data, but I did realize that one key difference between their estimate and our VIN based data is that they are reporting only on the U.S. market, while our data includes Canadian sales (and eventually global sales).

Setting the results for May aside, they show 7,506 U.S. sales so far this year. If I assume 1,700 deliveries in May, and 1,450 in April (I subtracted 100 loaners from May), that is ~3,150 deliveries so far in Q2. Adding the additional ~4,900 in Q1 you get ~8,050 North American deliveries in 2013, with ~544 being Canadian.

The Canadian number is maybe a bit high as a percentage of the market, but Canadian deliveries did not even begin until Christmas 2012, so the vast majority have been delivered in 2013. Looking at the reservation tally, there were ~877 Canadian reservations as of February 1st (including the Sigs), and some small portion (maybe 50-100) were delivered at the end of December. Once you factor in cancellations, 544 Canadian deliveries so far in 2013 looks very credible.

Coming back to the May Autodata report, we know it's U.S. only. If we assume 1,700 total deliveries, and Autodata is only seeing 1,425 U.S. deliveries, that could mean that Tesla just did a big batch (275) for the Canadians, or it could mean a somewhat smaller batch for the Canadians (175), and maybe a somewhat smaller (but still credible) delivery number of ~1,600.

Bottom line is that I am not in a position to question the Autodata numbers, and for 2013 as a whole they are extremely credible.

- - - Updated - - -

I should also point out that if Tesla did a batch of Canadian cars, it would slow the delivery flow compared to the trend (because every Canadian car is by definition far from the plant in Freemont). So a big batch of Canadian cars would look like a slow sales month, almost by definition. If so, you would see that self correct in the following month as deliveries snapped back to trend.

I'm not certain of all the differences between Canadian and U.S. cars, but even minor differences are probably enough to make Tesla prefer big batches over small ones, just so that the work flow is not disrupted as much.

- - - Updated - - -

One more thing, the Autodata numbers for Tesla are an "estimate". Take that for what it is. Here is the spreadsheet for anyone interested -

http://www.motorintelligence.com/fileopen.asp?File=SR_Sales61.xls

 

[adiggs]

Ok, I looked at the Autodata spreadsheet. Still not certain what their methodology is for collecting the data, but I did realize that one key difference between their estimate and our VIN based data is that they are reporting only on the U.S. market, while our data includes Canadian sales (and eventually global sales).

Setting the results for May aside, they show 7,506 U.S. sales so far this year. If I assume 1,700 deliveries in May, and 1,450 in April (I subtracted 100 loaners from May), that is ~3,150 deliveries so far in Q2. Adding the additional ~4,900 in Q1 you get ~8,050 North American deliveries in 2013, with ~544 being Canadian.

The Canadian number is maybe a bit high as a percentage of the market, but Canadian deliveries did not even begin until Christmas 2012, so the vast majority have been delivered in 2013. Looking at the reservation tally, there were ~877 Canadian reservations as of February 1st (including the Sigs), and some small portion (maybe 50-100) were delivered at the end of December. Once you factor in cancellations, 544 Canadian deliveries so far in 2013 looks very credible.

Coming back to the May Autodata report, we know it's U.S. only. If we assume 1,700 total deliveries, and Autodata is only seeing 1,425 U.S. deliveries, that could mean that Tesla just did a big batch (275) for the Canadians, or it could mean a somewhat smaller batch for the Canadians (175), and maybe a somewhat smaller (but still credible) delivery number of ~1,600.

Bottom line is that I am not in a position to question the Autodata numbers, and for 2013 as a whole they are extremely credible.

- - - Updated - - -

I should also point out that if Tesla did a batch of Canadian cars, it would slow the delivery flow compared to the trend (because every Canadian car is by definition far from the plant in Freemont). So a big batch of Canadian cars would look like a slow sales month, almost by definition. If so, you would see that self correct in the following month as deliveries snapped back to trend.

I'm not certain of all the differences between Canadian and U.S. cars, but even minor differences are probably enough to make Tesla prefer big batches over small ones, just so that the work flow is not disrupted as much.

- - - Updated - - -

One more thing, the Autodata numbers for Tesla are an "estimate". Take that for what it is. Here is the spreadsheet for anyone interested -

http://www.motorintelligence.com/fileopen.asp?File=SR_Sales61.xls

One observation CapOpp - in your model, it looks like the loaners are treated as cars that effectively go out of circulation, or at least aren't for sale or get sold. In conversations here in Portland, and from what I've read elsewhere here in the forums, it appears that the loaners are actually selling quite quickly. My own conclusion from what I've read and experienced is that cars built as loaners still sell quite quickly, and that in fact Tesla might still be figuring out what the correct inventory level is to keep loaners in the service centers, much less display cars in the showrooms.

I don't know exactly what to make of it, but subtracting all 100 loaners seems exceedingly pessimistic. Or maybe it's just treated as a constant "inventory" that is being built this month, and then has no impact in further months...

Also worth noting - it's a good problem to have

 

[Mayhemm]

In conversations here in Portland, and from what I've read elsewhere here in the forums, it appears that the loaners are actually selling quite quickly. My own conclusion from what I've read and experienced is that cars built as loaners still sell quite quickly, and that in fact Tesla might still be figuring out what the correct inventory level is to keep loaners in the service centers, much less display cars in the showrooms.

Indeed. I was under the impression that the "loaner" vehicles were selling so quickly that owner's bringing their Model S in for service were often given ICE rental cars because there were no loaners left to give them.

 

[Cosmacelf]

Well here in San Diego, they have all of one Tesla loaner car, so everyone gets an ICE when dropping a car for service.