Elon also noted that Plaid will be more about performance than range increase. Quite the tweet storm.
Elon Musk on Twitter
I'm just glad that we don't have to go through that annoying thing again here where a couple weeks ago people were adamantly insisting that Plaid was going to have 500-600 miles range :Þ
Tesla fans are so amazingly good at Osbourning themselves.
I didn't make any claims about Q1 S/X rates, I only observed, based on a single unreliable data point, that the S/X Q1 order book might be fine:
Btw, 5-7 weeks has dropped to 4-6 weeks since I wrote that comment.
Yes, because time passes, and we don't live in a static universe. The fact is that it's been end-of-quarter deliveries all quarter - meaning that they've had no surplus capacity for domestic production. The span is also shrinking it; last I checked (several weeks ago), it was a 4-week span.
The pattern is the same sort of pattern we saw last quarter, yet for some reason, you want to interpret it as a sign of weakness.
We also know that the delivery estimates on their webpage are unreliable
In both directions, yes. You can be lucky and they find a car for you early. You can also be unlucky and get yours late.
Today Elon also announced a major Q1 S/X demand boost: a +5% range upgrade. He also announced that the Plaid will have the same range, I.e. countering any Osbourning by the Plaid.
Are you suggesting that Elon should have waited until Q2 with these announcements, because Q1 orders are fine?
Are you suggesting that updating the website to reflect actual ranges, something he already discussed on the earnings call and
was extensively reported on, is some out-of-the-blue pump attempt? Are you suggesting that Elon
should wait until Q2 to announce changes that are already in production? Why? Are you saying that you expect Q2 to be weak and that Tesla should "hide" in-production changes until then?
Where are the price cuts? No, seriously. By early February of last year we'd already had
two price cuts. It's mid-February.... where are they? We had price
increases late last year, and they've not even undone those. In what sort of "demand problem" scenario would there be no change in pricing whatsoever? Margins are superb and could easily support it.
Q1 last year was bad, really bad
And essentially the only thing this quarter shares in common with it is the name "Q1".
FWIW, the "ship loading days" sum of the first 9 ships in the quarter:
- Q1'19: 20.4 days
- Q2'19: 19.1 days
- Q3'19: 25.1 days
- Q4'19: 24.8 days
- Q1'20: 21.5 days
I.e. 5% above Q1'19 levels, but about 15% below Q3 and Q4 levels.
"The first 9 ships" does not reflect the current steady-state; nobody is disputing that this quarter's shipping had a slow start (for very obvious reasons); analyzing that slow start was the
very premise of my post, using the slow start to assess inventory buildup. One might add in ship loading times to adjust the inventory-buildup estimates and determine that the initial lag time and corresponding inventory build should be treated as being greater (I'd consider this a reasonable adjustment), but this should not be used to adjust steady-state shipping rate estimates.
The past, say, 20 days of shipping (same number of ships, 6) has 15,67 loading days, vs 16,95 for the same period last quarter. A 7,6% difference, not a 15% difference, and it's exclusively due to a single ship (Viking Bravery). Asian King is currently loading and looks to be a long load-time ship (only 0,69 days idle at the pier before arrival, doesn't leave for 4 days after arrival), which will lower the difference. Q3 only had 4 ships during this timeperiod because shipping was slower. But if we want to give Q3 more time, and add an extra ship in each direction (Glovis Cosmos and Glovis Tengfei), its total would be 15,86 loading days
There are also other metrics that can be used apart from load times (an imperfect metric - as loading delays actually make the metric look
better) to estimate the number of vehicles - for example, ship sizes (another imperfect metric, since you never know how full they're loading a ship, or whether the ship was contracted more out of convenience than need). Of the six matching ships this quarter, they're all ~200m long, 4 are 32m wide, one is 36m wide, and one an unusually large 38m wide (one of the largest ships Tesla has ever contracted). We have no gross tonnage figures from either of the larger ships and one of the midsize (~200m x ~32m) ships; the other 3 midsize ships were 50,3kt, 58,3kt, and 59,4kt.
For the corresponding ships in Q4, there was one unusually small ship (184m x 30,6m, 44,9kt); three midsize ~200m x ~32m ships (55,7t, 59,2t, 59,8t); and two moderately large (~200m x 35m) ships (65,7t each). Overall, I'd say that ship sizes are roughly a wash.
One must reiterate that RO-ROs aren't the only type of vehicle shipping that Tesla does; see my comments in the original post re: container shipping. As a random example, Iceland is a new market (the market size of Spain / Portugal) which is expected to get its vehicles entirely by container. Or that shipping to AU and NZ by container looks to be quite strong. Container shipping is not as well tracked as well as the RO-ROs, but I see no signs of decline in it, and a potential increase - and indeed, an increase in the fraction of shipping conducted by container vs. RO-RO should be logically expected over time as Tesla expands into smaller markets discontiguous from existing RO-RO markets, and decreases RO-RO shipping to China (indeed, there's only been 1 RO-RO to China this quarter, and it was a short-loading one right at the start; increasingly, Teslas shipped from Fremont to China - S, X, and (for now) higher-end 3s - will be in containers; at this point last quarter, every other RO-RO was to China).