,
16 tweets,
4 min read
Read on Twitter
1/ I've been tracking EU orders since mid December. I've had a great source of raw data: Karsten's EU M3 tracker. I've done my data validation, correction/sanitization, then used my methods to calculate my order estimates, which made up for probably most errors, lags in raw data.
2/ At the very end of January, strange data started to come in. Noted this to a few ppl in DM's, but ships have just started to approach Zeebrugge, showroom cars appeared along with test drives, so even if I found it strange, I could attribute a few days' uptick to these.
3/ A few days later I've realized that Tesla started to collect money from buyers at least a week before scheduled delivery. As this (final inv's) has inflated invoice numbers in the trackers, which were the source of data, I had to stop providing updates
4/ After some investigations, I've concluded my data to be accurate until 1/30, but not later. Up until this date, trendlines were aligning to actual data, with a few justifiable exceptions (like DE incentives arriving).
5/ So I gave a final update as seen above. I've excahged quite a few DM's with @Tesla_Bear who was questioning these numbers. We both agree that the number the tracker shows now is inflated. But he argues that my estimates could be inflated too.
6/ As I've paid great attention to the trends, verified each and every entry in my model, I thought it was not the case. Then @Tesla_Bear published an excellent thread on the topic, confirming my thesis.
7/ He also correctly pointed out that Switzerland had started to use a new invoice book. Just as France. And possibly 2 others (too few data points to confirm). I've taken these into account.
8/ But @Tesla_Bear still insisted on the possibility of having double invoices earlier too. This tweet was the only one contradicting my thesis.
9/ It's the only one in the EU tracker that shows a lower final invoice # than what I've found to be the largest invoice number for reliable entries. So I dug into the data and spent some time today with this. Used the EU tracker and the Dutch one, which also has invoice #'s.
10/ While I haven't found another such entry in the EU tracker, I've found lower numbers in the Dutch one! I've had no other evidence from the lots of forums I'm following about double invoicing before 1/30. Was I wrong all the way?
11/ We have a lucky entry in both trackers: Frank. First entry in both trackers, deposit invoice #165, final invoice #1546. Took his car on 2/7, so one of the first M3 owners in Europe. Invoice #1546 is smaller than the last one I've used for my Dutch trendlines.
12/ My trendlines still held. What about the smaller final invoice #'s in the Dutch tracker then? They've arrived almost 4 weeks later. So I'm assuming Tesla in the Netherlands has started to use another invoice book also, but that's irrelevant (as being too late) for my calcs.
13/ I have to admit one error though. I've used an invoice (#1692) for calculating totals (but not the trendlines), as too few data was available to check validity of this. Could have been right. Turns out it wasn't. The last correct one then is #1445 for Netherlands.
14/ Thus I'm lowering my guidance by ~250. I still think the single Norwegian invoice must've been an error, as no other data point suggest having double invoices (en masse) earlier. There could have been cancellations, erroneus invoices corrected, but I don't believe it's many
15/ I'm estimating orders on 1/30 having been ~17,500. If then current run-rates remained steady (they were dropping, but deliveries have started, wait tiimes decreased, showroom and test cars available, media buzz is there, recent price cuts!), current orders are ~20,500.
16/ I still think Tesla will have ~22-23k orders by EOQ, will have all of those delivered this Q and have a small inventory in EU.
/fin
$TSLA $TSLAQ
/fin
$TSLA $TSLAQ
