Tesla has been increasing S & X deliveries, in part, by expanding into geographically more remote and thinner markets such as Mexico, South Korea, UAE, Ireland, Portugal, Jordan, etc., all of which require build-out and operation of new sales/service centers and supercharger locations, many of which involve shorter-term operating leases until the market is validated.
Except Tesla is not actually doing those service center expansions. Arguably they should be. But they are in actual fact not.
Other, more established markets such as Denmark, Hong Kong, and the UK where the EV subsidies are contracting still have operating expenses for the embedded infrastructure that do not decline anywhere near as sharply as new sales apparently have.
I'm going to ignore this silliness, because this doesn't change SG&A, it leaves it constant.
Then, there is the promised tripling of Supercharger pedestals before year end 2018. "There are currently 6,124 Superchargers around the world. By the end of 2018 there should be over 18,000 worldwide. Eventually you’ll be able to go anywhere on Earth using the Superchargers," Elon said on Friday night. To the extent site leasehold improvements for expansion are on shorter-term operating lease locations, those costs will also be expensed rather than capitalized. Tesla allocates SC operating costs between COGS and SG&A. The more that is allocated to SG&A, the higher the reported GM%, but it all washes out at the Net Income (Loss) level.
Superchargers don't amount to a hill of beans financially, though. It just doesn't add up to much. Even the capital costs are small, but the operational costs approximate 0. So I'm going to ignore this too.
So far you haven't given a reason to expect SG&A to increase.
The greatest driver of SG&A will obviously be the growth in total deliveries (and servicing)--say from ~76,000 in 2016 to ~350,000 +/- during 2018. It's not just the increase in volume but also the amount of "hand-holding" of new, down-market customers that will be required.
Except that Tesla isn't going to actually do this in 2018. If anything, they've been decreasing the amount of hand-holding. For all of 2018, Tesla will be delivering to early reservation holders -- early adopters. Certainly there will be some administrative overhead involved in handling the customers, but they sure aren't going to do hand-holding. It'll be drop ship and deal; they've made it pretty clear.
Through 6/30/17, Tesla had delivered ~230,000 Ss & Xs world-wide. I'd guesstimate that about 80% were to multiple car households and about 20% were to customers who had previously owned Teslas. Most of the customers in that demographic could readily write a check for their purchase and/or the difference between the Tesla they were returning and their new one. A far greater percentage of M3 customers will be single vehicle households, have un-paid loans/leases on non-Tesla vehicles, and are not nearly as financially astute nor well off as S and X purchasers. IMO, Tesla will have to expend a lot of SG&A to accommodate the new customer demographic to avoid complaints from disgruntled ones who are always more vocal than the more tolerant types.
Tesla is not going to do this in 2018.
Maybe they
WILL get a bunch of complaints from disgruntled customers. That's a very real possibility, and one worth worrying about. But they've made their decision very clear. They aren't even listening to the
existing disgruntled customers (cough cough, USB bugs). There's no way they're going to spend more SG&A on dealing with this sort of thing in 2018 when they haven't been dealing with it at all so far.
I guess I'm saying that they probably
should increase SG&A in 2018 in order to keep their customers happy, but they clearly aren't actually
going to. Maybe in 2019 they will. Or maybe this will become a big risk factor for the company going forward. But it seems pretty clear that for the near future, they're just not spending the money.
As far as R&D growth goes, IMO Tesla cannot stay on schedule with plans for a new MY (on a entirely new platform in an entirely new factory), the Class 8 tractor, the pick-up, the solar shingles etc. without significant growth in R&D during 2018.
That sounds superficially plausible until you think about it. The solar shingles have been spending R&D money at a steady rate for years now and are now heading for the more mature, less-R&D stage. The Model Y is essentially taking over the development position from the Model 3 as the Model 3 moves into the next stage. The class 8 tractor has apparently been in design long enough to have working prototypes, so again, the R&D is already going on. The pickup, which we haven't seen any sign of, is the only one which is likely to actually increase R&D.
I think Tesla's been pulling a bit of a "hidden in plain sight" trick here. I think people have been assuming that the R&D expenses for the last several quarters were almost entirely for Models S, X, and 3, and I suspect that's entirely false and that there's been massive R&D expenditures on the future products already. Basically, I think the elevated R&D levels which you're expecting
have already happened. We'll probably see a mild increase, but I think this quarter's R&D level is a decent approximation of steady state.