One month ago, following the latest negative note from Goldman's auto analyst David Tamberrino, Elon Musk decided to take it personally and in a leaked memo to his employees, said that Goldman is "in for a rude awakening :)"
With Tesla stock sliding 15% since then, it is not exactly clear who has had a more rude awakening, the bulls or bears, but in having kept silent for the past month, Goldman (which has a Sell reco and a $195 PT on TSLA), has come out with another negative note on Tesla, just two days before Tesla's Aug. 1 earnings, in which Goldman details the key investor questions and recent debate on the stock from its conversations with clients.
As Tamberrino writes, "the key focus areas revolve around (1) sustainability of Model 3 production, (2) pace and demand-variants of Model 3 order conversion, (3) margin improvement potential, and (4) FCF burn."
More importantly, for the first time, the Goldman analyst introduces a new analysis on Model 3 sentiment based off social media posts. "This stems from our work on Model 3 order conversion and customer reception."
And the punchline: "Ultimately, we believe the data potentially points to waning customer enthusiasm for the vehicle as order availability and test drives have increased."
We expect another very angry Elon Musk memo to be "leaked" shortly.
* * *
With that out of the way, here is Goldman's Q&A on several key questions relating to the Model 3 roll out:
1. Post the 5k/week Model 3 production achievement in the last week of 2Q, how is the run-rate so far in 3Q?
We track VIN registrations closely for the Model 3, and following the end of 2Q18 the company has registered another 20,700 VINs. With historical conversion from VIN figures to production numbers (as of the end of 2Q18, TSLA had 56k VINs registered and produced 41k Model 3s), we believe this implies that TSLA has produced approx. 15k vehicles in the past four weeks. This essentially equates to a production figure of approx. 4,000/week. We believe questions on the conference call will focus on the sustainability and potential to increase this rate up to (and possibly beyond) the 5,000/week rate – and what the company eventually does with its GA4 assembly line. The sustainability of this rate – particularly without the aid of the GA4 tent assembly line (which produced 20% of the 5,000 vehicles in the last week of 2Q18), is important as it potentially sets the company up for positive FCF in 3Q18 as well as potential for putting incremental capex into Model 3 assembly to increase the production rate to the 10,000/week target – timing for which has been uncertain after the production delays in late 2017/early 2018.
Are reservation holders converting?
As production improves for the Model 3, our conversations with investors have quickly shifted to (1) reservation conversion to orders, (2) demand by Model 3 variants, and (3) new order rates. With TSLA offering Model 3 variants starting at the $49,000 price point – with options that can take the price up to $80,000 – investors question how many of the 420,000 reservation holders will convert at those higher price points (given the vehicle was originally offered with a base price of $35k). Further, investors have questioned the pace of reservation conversion given the company’s advertised wait times for the Model 3 on the website, quoting 2-4 months from order to delivery – suggesting reservation holders may not be converting immediately.
As we have detailed in the past, we believe that the company could convert approx. 23% of its reservation holders to higher price variants —per our price elasticity of demand analysis on the US sedan market (Exhibit 1). This would imply that approx. 100,000 reservation holders could opt for a vehicle with an average selling price around the mid-to-high $60k range. However, this would imply a vast majority of reservation holders could be waiting for availability of a lower price Model 3 variant.
To help further inform our perspective, we discussed initial customer preferences with various Tesla sales personnel in different regions across the US. What we learned is that most areas have recently received (in the past few weeks) a Model 3 that they can offer for test drives. Most stores are still working through reservation holders within their respective areas, but foot traffic to view the vehicle has increased. Interestingly, customers coming to check out the Model 3 arrive in all types of vehicles – with some looking to potentially trade-up from a Camry or a Civic, others looking to swap out of a 3-Series or a C-Class, while others are looking to trade down (with some noting Model S owners are looking to move into a smaller sedan). Altogether, it appears interest in the vehicle still remains high (at least anecdotally) – particularly as the Model 3 began to hit stores. However, what remains unknown – and hopefully communicated with 2Q18’s earnings release, is how new Model 3 order rates are trending.
Lastly, we also looked to gauge sentiment on the Model 3 with an analysis based on daily Twitter and Reddit posts (Exhibit 2). Posts are categorized into basic positives, basic negatives, and neutral buckets. We then subtract the number of negative posts from the positive posts and divide by total posts to develop the sentiment skew within social media. From the data, we see a few trends:
- Weekly posts about the Model 3 are declining: with 2016 averaging 4,400/week, 2017 averaging 3,900/week, and YTD 2018 averaging 3,000/week;
- Positive skew has faded: with 2016 carrying a +18% positive skew, 2017 a 12% positive skew, and YTD 2018 tracking at a 4.5% positive skew;
- With the recent increased availability of the Model 3 and increased production, we note a lack of peak (which would indicate significant buzz around the product) in social media posts;
- Lastly, we find that sentiment has recently declined from a slightly positive skew (i.e., from +5% in 1H18) to a slightly negative skew the past few weeks 0 corresponding with the time period where the vehicle began to hit stores and orders were opened to all.
What is the scope for material cost reduction?
Before (and certainly after) the news reports (from various sources, including Bloomberg and Reuters) that TSLA was looking for refunds from its suppliers, investors were digging into potential material cost reductions for the Model 3. This comes as the expected growth in Model 3 production would drive increased parts orders from suppliers, and potentially kick in price reductions as certain minimum order quantity thresholds were met. This is typical of a supplier/OEM relationship, where (1) economies of scale and/or (2) new business opportunities (i.e., RFPs for new vehicle programs) are used by OEMs to extract price savings from its supply base. Normal annual price downs can range from 1% to 3%, depending on how commoditized a part is – and potentially higher depending on over-supply within the market. Further, ‘quick savings’ granted from suppliers to OEMs can come as a part of new business awards – where suppliers trade-off price on existing contracts in order to help win new business. For TSLA, the company very likely should achieve contractual price downs as its volumes increase – though the volume thresholds are unknown. That said, with original targets for production of 5k/week exiting 2017 and 10k/week exiting 2018, we believe TSLA is likely behind on its potential for volume break-points with its supply base. As a result, the sequential improvements we forecast in 2018/2019 for Model 3 gross margins are largely driven by a reduction in man-hours per vehicle produced – with achievement of targeted 25% margins on the vehicle further out as volumes continue to ramp.
Last but not least, How much cash will the company burn in 2Q?
Lastly, investors question the ending cash balance and the company’s potential need for a capital raise. As most have picked through the delivery release, the large increase of vehicles in transit sticks out as a potential working capital headwind for the quarter. This should be somewhat offset by the favorable cash conversion cycle on US Model 3 deliveries. However, altogether with an incremental 9,000 vehicles in transit at the end of 2Q18, we forecast working capital to be a $170mn headwind for the quarter. This puts our FCF estimate (traditional definition) at a $930mn burn in 2Q18 and an adjusted FCF (TSLA definition, including auto/solar lease added back) burn of $620mn; as a result, we forecast the quarter-ending cash balance to be approx. $2.2bn. Looking ahead to 3Q18, we do forecast working capital to be a major positive (+$590mn from the Model 3) – helping the company show positive FCF; however, we see this unwinding to some degree in 4Q18 and still expect the company to raise capital in order to maintain a cash balance above $2bn.
What does all this mean for Tesla and Goldman's forecasts? This is how the analyst has revised his estimates and assumptions:
- 2Q18 actual deliveries —where Model S/X deliveries were n slightly below our estimates but Model 3 deliveries fell a few thousand short (our 2Q18 delivery recap note);
- Improved Model 3 weekly production in 3Q18 and beyond – we now forecast TSLA sustaining 4,000/week in 3Q, improving to approx. 5,000/week in 4Q;
- Lowered Model S / X deliveries – we now forecast slightly softer demand internationally, primarily as a result of the price increases announced in China; our full year Model S/X forecast of 90k falls below the company’s guidance for 100k;
- Slightly lower gross profit margins – the combination of increased manual labor, the incremental production lines, and FX headwinds from the Euro and RMB;
- Lowered SG&A profile – stemming from the company’s announced headcount reductions, we now forecast a reduction from 1H18 to 2H18 in SG&A expense as well as re-base future expense creep;
- FCF estimates are also lowered for the quarter on a use of working capital – primarily due to the large increase of vehicles in transit at the end of 2Q18; we now forecast TSLA’s adjusted FCF burning $600mn in the quarter – leaving approx. $2.2bn in cash at the end of 2Q18. For the year, we now estimate a $1.5bn cash burn – and continue to forecast a capital raise in 4Q18a
With all that in mind, Goldman writes that "the net result of this is an increase to our 2018 and 2019 EBITDA estimates, but a reduction to 2020 and 2021."
For the next two years, a faster ramp in Model 3 production/deliveries and lower SG&A expense accrue to the bottom line of our forecast (GS EBITDA +11% on average). However, in 2020 and 2021 the lower volume trajectory and margin profile for the S/X slightly reduce our estimates (down an average 6%). For 2018 EBITDA, we are still 15% below FactSet consensus – and our 2019 through 2021 EBITDA is an average 30% below the Street.
But the real take home here is that Musk will only focus on Goldman now using twitter to gauge enthusiasm for the Model 3. As such expect a lot of fake troll accounts to be created somewhere in Myanmar praising the electric vehicle at a rate of 60 tweets per minute, or whatever the clickfarm rate is these days.