Based on my understanding, I don't think you got the first part right -- when SolarCity is building a new system, the installation and panel costs are capitalized (they show up in the cash flow statement as capex), not expensed. Revenues and expenses are then recognized over the life of the lease. Here is the language from the 10-K re: how these costs flow through the financials: Presentation of Cash Flows Associated with Solar Energy Systems The Company classifies cash flows associated with solar energy systems in accordance with ASC 230, Statement of Cash Flows. The Company determines the appropriate classification of cash payments related to solar energy systems depending on the activity that is likely to be the predominant source of cash flows for the item being paid for. Accordingly, the Company presents payments made in a period for costs incurred to install solar energy systems that will be leased to customers, including the payments for cost of the inventory that is utilized in such systems, as investing activities in the consolidated statement of cash flows. Payments made for inventory that will be utilized for solar energy systems that will be sold to customers are presented as cash flows from operations in the consolidated statement of cash flows.
i think the way to tell if nci's are incorporated is to track their operating income forecast down through interest and taxes to end at net income. if the buck stops there and is divided by diluted share count to get to an eps estimate, then nci's are not included in that forecast. in that case, the appropriate metric to compare to see if tesla beat that analyst's operational expectations would be after-tax but before nci net income. for example the ms model that i have shows that for q4 the nci and one time gain was booked as "extraordinary gains (charges)". for q1 no such charges are modeled. the q1 17 net income line shows -326m as the result on a diluted share count of 168m. that gets jonas to -1.94 gaap eps/-1.40 non-gaap. so for my estimates, i would say the proper metric to compare would be jonas's net income of -326m or even operating profit of -163m vs. my net income of -176m or operating profit of -52m. i am suspecting that the robots/quants are not smart enough to understand this distinction. assuming there are meaningful numbers of robot / quant traders that read the report and compare to expectations, there may be an outsized reaction after the report. if such a scenario unfolded, for those who held "excess" shares it might be a good time to sell into that romp, esp. if the exposure was being automatically "replaced" by options going in the money anyway. this discussion is going to change my strategy slightly. going into the report i will substitute some deep in the money calls with the synthetic equivalents: long stock + long put. that way i maintain downside protection and if the report goes as i hope i'll be able to rip off shares to mr roboto and let calls that would have gone in the money by then sub for that exposure. judging by how the recall was covered, i'd expect by morning the short spin machine will be out in full force crying foul. it won't matter in the intermediate term unless they can get the index committee to change their rules.
@surfside, see bottom of pg 32 solarcity 10k, if that affects your opinion... Net Loss Attributable to Noncontrolling Interests and Redeemable Noncontrolling Interests The net loss attributable to noncontrolling interests and redeemable noncontrolling interests represents the share of net loss that was allocated to the investors in the joint venture financing funds. This amount was determined as the change in the investors’ interests in the joint venture financing funds between the beginning and end of each reported period, calculated primarily using the HLBV method, less any capital contributions net of any capital distributions. The calculation depends on the specific contractual liquidation provisions of each joint venture financing fund and is generally affected by, among other factors, the tax attributes allocated to the investors including tax bonus depreciation and ITCs or U.S. Treasury grants in lieu of the ITCs, the existence of guarantees of minimum returns to the investors by us and the allocation of taxable income or losses including provisions that govern the level of deficits that can be funded by the investors in a liquidation scenario. The calculation is also affected by the cost of the assets sold to the joint venture financing funds, which forms the book basis of the net assets allocated to the investors assuming a liquidation scenario. Generally, significant loss allocations to the investors have arisen in situations where there was a significant difference between the fair value and the cost of the assets sold to the joint venture financing funds in a particular period accompanied by the absence of guarantees of minimum returns to the investors by us, since the capital contributions by the investors were based on the fair value of the assets while the calculation is based on the cost of the assets. The existence of guarantees of minimum returns to the investors by us and limits on the level of deficits that the investors are contractually obligated to fund in a liquidation scenario reduce the amount of losses that could be allocated to the investors. In addition, the redeemable noncontrolling interest balance is at least equal to the redemption amount.
So I've read and re-read that language about five times now, and I'm trying to wrap my head around what they are suggesting in that language. In layman terms, it seems like they are saying that large losses to non-controlling interests are happening because the fair value of the projects in a liquidation (per HLBV) are significantly lower than the cost of the projects. Is that how you read it? Even taking into account this text regarding the NCI's, I still stand by my statement re: Fallenone's attempt to describe the accounting -- the losses aren't because they are recognizing in the income statement the costs to build out the system -- those costs are capitalized as an asset, as described in the text ("The calculation is also affected by the cost of the assets sold to the joint venture financing funds, which forms the book basis of the net assets"). The losses are based on the difference in the cost of the assets and the fair value in a liquidation (the difference in two ways of valuing the assets). Let me know if you think I'm getting it wrong. surfside
Ok so I've gone through all of the equity research reports that I have access to in detail; some interesting findings -- seems I missed a couple other NCI forecasts in my haste looking through the reports yesterday. But before I can post the details, I really need to know how to make a table. Can someone explain how to make a table? Lol. Feeling pretty technologically challenged at the moment.
Hit "reply" to: 2017 Investor Roundtable:General Discussion and note the coding in the quoted section.
I think it works this way: The NCI's pay FMV for the assets transferred to the SPE but for the liquidation calculation they are allocated only the depreciated cost basis for the assets transferred to the SPE. In completing the Hypothetical Liquidation, the value of the assets on the SPE's balance sheet is reduced so that reduction in asset value flows through the Income Statement as a loss. There may also be a difference in depreciation methods: i.e. accelerated tax basis for HLBV and (management method) straight-line with 30-35 year useful life for the SPE's non-liquidation valuation on its Balance Sheet.
Ok, so here is my compilation of the various equity research analyst Q1 2017 forecasts (I've also noted which ones forecast NCI's): BankPrice TargetQ1 GAAP EPSQ1 non-GAAP EPS Operating Income Net IncomeInclude any NCI's?Comments Barclays$165n/a-1.07 -199 -173 non-GAAP Non/a Deutsche Bank$240-1.41n/a n/a n/a NoHad forecasted +4.19 of EPS for FYE 2017 prior to 3.14.17 J.P. Morgan$185n/a+0.83 -159 134 non-GAAP; 44 GAAP Yes: $1,050 for FYE 2017Was previously forecasting -1.52 of EPS for Q1 before modeling impact of SolarCity Morgan Stanley$305-2.13-1.45 -249 -283 GAAP NoDoes not forecast SolarCity Baird$368n/a-1.23 -230 -198 non-GAAP Non/a UBS$160-0.39n/a Full Year -561 Full Year -477 GAAP Yes: $338 for FYE 2017n/a As I mentioned in my previous post, relative to my comment yesterday re: analysts forecasting NCI's, after looking at J.P. Morgan's research note (instead of their summary model), I saw that they are in fact foreasting NCI's, and their forecast is pretty much in line with Goldman's. UBS is also forecasting NCI's, but only at the level of $338 for the full year. Given the quality (ahem) of Colin Langan's other work on Tesla, that doesn't worry me too much. I posted a couple other comments; I thought it was interesting that Deutsche Bank was forecasting +4.19 up until March 14th (when they released a note updating their model for guidance from the 4Q2016 call). It isn't clear at all from their note the reason for the major change in EPS, particularly given they actually increased their revenue forecast. They did note that their forecast shows Tesla could reach break-even by Q4. Anyways, hope this is helpful. surfside
This is my understanding. This is more complicated than it appears because the systems are depreciated on a straight-line basis for GAAP purposes, and on an accelerated basis for tax purposes. In the early years deprectiation is higher for tax purposes than for GAAP purposes, and in the later years it's the other way around. This difference is (sigh) valuable and if I'm not mistaken, it's considered a GAAP asset (a "tax asset") which is amortized for the HLBV purpose. Clear as mud? I think that's right... but I could be wrong... you can see why most people never dig into this stuff... Anyway the accelerated depreciation is an asset which is "wasting away" very fast, so the value of the VIE drops. Once it's done, the VIE value should stabilize. I'm not at all sure that I've got this right. Anyway, I may be totally wrong, but this is my guess as to the dynamic: Year 1: writedown from cost to FMV on initiation of VIE, largely a loss attributable to SCTY Year 2-5: amortization (sort of) of the tax assets from the accelerated depreciation etc., largely a loss attributable to the NCIs Year 6-(some later year): funny stuff is gone, GAAP profits from the panels, largely attributable to the NCIs until they achieve their "specified return" Year (that later year) - end of lease: profits largely attributable to SCTY, or to the "cash equity investors" if they monetized it
In case your brains aren't hurting enough (note this apparently pre-dates the Tesla restructuring but is after the intent to acquire disclosure) http://files.shareholder.com/downloads/AMDA-14LQRE/2294774888x0x830612/1A32ABBC-4024-44B9-8F81-1B5BD77DD00B/2016.06_SCTY_Investor_Presentation.pdf Note the cash flow graph in slide 22. FWIW, I apologize for contributing to mental Onanism, since I remain firmly in the camp that share price in the latter part of the first week of May is far more dependent on guidance about the M3 being "on track" than positive GAAP earnings from NCI allocation of losses and/or regulatory credit profits.
Interesting PDF but it opens up yet more questions. Page 13 projects $0.02 operations & maintenance costs on $0.18 of revenue for future revenue (11%). Page 20 lists actual costs over the last quarters. For example 2016Q1 has $67M aggregate revenue with $14M O&M costs (20%). Why the difference? You have no argument from me there. July production, September deliveries, December ramped to 100k yearly run rate is really the only thing the investment community wants to hear over the next months. And that is exactly what management will tell them. I just happen to like mental gymnastics.
@surfside: thank you sincerely for the hard work you put into this post. i am in the midst of reviewing and tweaking my earnings strategy/exposure but will try to give an intelligent response based on your data today. separately i have heard that jpm just tweaked their model again, not sure if you have access to the latest update. i reached out to someone else i know to see if i can get that information too. thank you again, i know i will be able to take your data and put together a simple table that lets us know which estimates we are beating (and how we beat them) when we report.
You're absolutely welcome -- looking forward to your revised thoughts. Also, I checked and I am not seeing the latest JP Morgan report on Capital IQ; I will post here if/when it shows up. My Capital IQ is funny -- there isn't any rhyme or reason as to why some reports come through and others don't (for example, I get most all of the Morgan Stanley reports, but the one from last week about Tesla Semi still hasn't hit my Capital IQ). surfside
note work in progress... @surfside, do we know if the consensus estimates are based on gaap or non-gaap? for your table, i think the best way to deal with the mess is for us to reduce variables. an easy way is to just look at q1 gaap eps. the non-gaap adjustments for stock based comp should be right around 50c a share. so i'm going to get rid of the non-gaap eps and just change the comments section to note that. let's make another combination, which is to have a gaap income without the nci reversal factored in. that will give us numbers for all the brokers listed. also, something's wrong with morgan stanley. they have -283 for gaap net income, but wth a -2.13 gaap eps figure it implies 132m shares outstanding which is far off. to get to [net income zero nci's] i take [q1 gaap eps] x 163m shares - [included nci's]. for example: for ubs: -0.39 x 163 - 338/4 = -148 please feel free to check the table for consistency. if i've done it correctly, we should be able to compare the [net income zero nci's] to actual results for a measure of operating performance vs. expectations. for those that include nci's we an look at gaap eps. it's good to see jpm's most recent update come up near where i am. i mentioned earlier the problem with morgan stanley's models. there's problems with ubs models too i think, they shouldn't be more bullish than me on net income w/zero nci and then have a $160 price target. i'm very open to being wrong so please feel free to correct. i think what's very clear is that except for those who have studied the consensus situation going into the report, interpreting blindly off first call consensus is going to be a poor decision. BankPrice TargetQ1 GAAP EPSnet income zero nci's Operating Income Net IncomeInclude any NCI's?Commentsluvb2bn/a0.61-176-5298yes: $275 for q1n/aBarclays$165-1.57-256-199 -173 non-GAAP Noforecasted -1.07 non-gaap epsDeutsche Bank$240-1.41-230n/a n/a NoHad forecasted +4.19 of EPS for FYE 2017 prior to 3.14.17J.P. Morgan$1850.33-209-159 134 non-GAAP; 44 GAAP Yes: $1,050 for FYE 2017Was previously forecasting -1.52 of EPS for Q1 before modeling impact of SolarCity. forecasted 0.83 non-gaap eps.Morgan Stanley$305-2.13-283-249 -283 GAAP NoDoes not forecast SolarCity. forecast -1.45 non-gaap epsBaird$368-1.73-230 -198 non-GAAP Noforecast -1.23 non-gaap epsUBS$160-0.39-148Full Year -561 Full Year -477 GAAP Yes: $338 for FYE 2017n/agoldman muppets$160-1.07-466Full Year -864 non-gaap Full Year -1348 nonGAAP Yes: $1166 for full yr 2017working off what appears to be a non-gaap eps figure of -0.57 - figures may be inaccurate need better data
There is definitely some low-balling happening. For example there was some blowback over Elon's estimate of "100k-200k" M3's produced in 2017. In the next ER the numbers he used sounded lower, but didn't actually contradict his previous figures. I agree with Brian that the impact of a profitable quarter will not be as big as: 1. News on the M3 production ramp. I believe (speculate) that Elon agrees with this opinion and that is what triggered his optimistic tweet "stormy weather in Shortville." I expect this to be positive, but I'm not sure how definitive the news will be. I expect that it will sufficient to trigger at least a $5-$10 bump in the SP. 2. I don't believe that TE income needs to be substantial or contribute to a positive EPS, to be an SP catalyst. I believe that it only needs to be enough to support credible guidance for good income in the remainder of 2017. I believe that this will happen by the end of Q3 at the latest. Q1 ER? I'm not sure. 3. Similarly for SCTY believe that it's more important to produce sufficient income to demonstrate with credible guidance that over the next 12-18 month's SCTY will make a substantial contribution to revenues, than for Tesla to make a Q1 profit. Of course a Q1 profit, if it happens will also have a positive impact, if it happens in addition to positive M3 ramp news. I think it's too early to trade possible inclusion in the s&p. OTOH I think even if that doesn't happen that there are so many upcoming catalysts that it will be a winning trade, barring any major M3 schedule slips. List of three items: If the M3 ramp is three month's late (which I think is unlikely) that might be good enough for September options, great for November options. TE income becomes material. Future AP releases. I believe that the most likely cause of the delays is they are having unexpected difficulties integrating the new cameras etc. Hopefully once they solve those problems they will get it to the "Paint It Black" level by Q3-Q4. More information on the GF3-4-5.
to be clear my view of positive gaap eps (and therefore even higher non-gaap eps) is that: 1. it brings the s&p addition into focus for sometime in august or september. you can bet the big dogs that follow the s&p 500 will have their ears perked up if we are within a stone's throw of qualifying for the s&p. 2. analysts are far offsides imo - and positive earnings with the chance of another positive quarter will force them to make eps revisions. and how long are these muppets with price targets 30-50% below current prices going to keep those targets in the face of improving earnings? 3. a strongly positive gaap number may mean that q2 can also be positive and there are some shorts who will simply not want to experience multiple quarters of positive eps going into a model 3 launch. i do agree that if they say model 3 is delayed 6 months that would torpedo the effects of positive eps, but don't expect that outcome. i've placed my bets with the idea that the stock should be 330+ after a report along my '++' scenario. i'll have some more commentary later on in the general thread on how positioning was optimized for that outcome, with some built in safeguards for lesser outcomes.
Regarding low-balled guidance: I like the theory that Andrea James (working for Tesla IR) is helping with guidance messaging. Tesla has had a major problem with wierd guidance where no scenario "feels like a win". Promising way too much, making bold claims that don't help and leave them on the hook later, etc. She probably knows very well how to pair promises to actuals, if she knows what the actuals are likely to be. In the past, it felt like if they could make 40,000 cars if everything went perfectly they would guide to 40,000. This time, we pretty much know they left 2016 at a 100k runrate, and they guided to 50k in 1H... It's like they guided to the median case instead of the best case, so that good execution would flow through and be good news. I suspect Andrea is making it her job to make writing ER letters easy, and the best way to do that is to ensure they are good news, and the best way to ensure that is to promise things that *sound* hard but *are easy*. Maybe it isn't her, but she would have the expertise, instincts and opportunity to help the company and investors.