Thanks for explaining your thoughts. However, I would come to a different reasoning here: If we assume that any production of Model S from mid-November onwards was diverted to NA, wouldn't it makes sense to fill only(!) orders for Europe / Asia now? In other words, there must be some orders + loaners + demo cars etc. that are urgently required in EU+Asia now. So if we assume that about 1.5 months of orders from outside NA are "backlogged" then it makes sense - in my mind - that order/delivery times in NA jump by approximately 2 months as soon as they "switch" production to non NA cars.
Mid-Q1 you could then start to maintain a somewhat more balanced mix of deliveries between NA + rest of the world.
Again, I'm not saying it HAS to be like this, but it seems a plausible explanation in my mind.
Also here I'm not quite sure it HAS to be exactly as you outlined: there could well be an overall drop in production for a week or two while the factory "gets ready to have the two cars on the line" - or let me rephrase this: I don't see how the first weeks of Model X production immediately lead to a boost in total cars made per week - so I could well see a slight reduction in Model S cars made while Model X ramps up.
I'm not trying to be difficult: I see how the data shifts in funny ways. And I can see how there maybe new pockets of demand opened up. But I don't think the whole jump is as big a puzzle to me.
I do not see the scenario of overseas/NA production you've outlined as a plausible one. It essentially boils down to producing exclusively overseas cars for 6 weeks starting in second half of December. What this means in real terms is that there will be no NA cars delivered in January 2016 - improbable IMO. Here is the
link for InsideEVs scorecard. As you can see, depending on how front loaded production of overseas cars during the particular quarter was, the NA deliveries for the first month of the quarter were ranging from approx 1,100 (Jan 2015) to 1,900 (October 2015). So assumption of zero NA cars manufactured in second half of December/first half of January is just not accurate. As you can see from the above, the historical range of NA cars delivered during the first month of the quarter is 1,100 to 1,900 cars.
To continue with the quantitative analysis, during the second month of a quarter, according to the InsideEV scorecard NA deliveries ranged from 1,500 (Feb 2015) to 3,200 (Nov 2015).
Since wait time for NA was moved out to second half of February, even if we assume historically low deliveries from what is outlined above, the conservative estimate for NA deliveries in January-first half of February 2016 would be 1,100+0.5x1,500=2,600. So this number represents a conservative estimate of NA deliveries during the first six weeks of 2016. To look at this number another way, it represents about 36% of the assumed production rate of about 1200 MS/week. Given that according to the latest we heard from Tesla, NA orders comprise approximately 55% of total, this is confirmation that the historical data used above is indeed representative for a quarter with production allocation front loaded for overseas cars.
What above means is that we can conservatively expect that at least 2,600 MS will be delivered during the first 6 weeks of 2016. What this also means is that for Tesla to move estimated delivery time for NA cars to end of February, they had to receive about 2,600 NA orders, all within the span of one week. This is unrealistically high as the second half of 2015 historical average for a week worth of NA orders is only about 0.55 x 1200 = 660 cars! So the incoming rate of orders either jumped to quadruple of the expected rate, or there is some other anomaly at hand, hence my conclusion regarding the puzzle.
To sum the above, the “plausible” scenario you’ve outlined to explain the “puzzle” requires zero NA deliveries of MS in the first six weeks of the 2016 – not even remotely possible scenario in my opinion. I am still convinced that we are experiencing something unusual here – the puzzle remains unsolved as far as I am concerned. One probability I was wondering about is some one time unusual occurrence – for example a large fleet order of MS.