Short-Term TSLA Price Movements - 2016

[racer26]

hi guys, I know this is TSLA forum, but I want to hear your opinion. I own MBLY stock, bought below $35. The past two weeks and today it has been falling more than 18%. Do you think I should sell now or wait for a rebound?

I went short MBLY (via Sep30 Puts) around $44.17 last week and still have that position open. I think its not done falling yet, but I'm prepared to be wrong on that front. I think a few things have happened with the TSLA/MBLY acrimonious breakup.

MBLY execs wouldn't stop talking about it and shoving their foot further down their throat. They claimed to dispute TSLA's assertions, but didn't actually speak to the most damning accusation TSLA made. That suggests to me that TSLA's side of the story is closer to the truth.

If TSLA is telling the truth, then it suggests that MBLY's modus operandi is to get an OEM hooked on their product, get it designed into the OEM's product, and then start jacking up the price once the OEM is shackled to the MBLY platform. That's a pretty shady way to do business, and will make other OEMs reticent to get in bed with MBLY. I think that doesn't bode well for their future on its own, now that such activities have been made public.

In addition, the v8.0 TSLA update relegating the MBLY sensor to playing second fiddle to Bosch's radar unit seems to be foreshadowing a wider shift. TSLA is the market leader in vehicular autonomy, and if they've decided that radar > visible spectrum camera, you can bet others will soon follow. If EyeQ3/4 is technology no longer needed, that can be replaced with a simple camera module without any integrated processing (as the processing needs to account for the radar feedback, and so needs to be run elsewhere), there goes all of MBLY's advantage.

If MBLY can't adapt quickly, I foresee them being in for a world of hurt. Either way, the damage to their reputation may be irreparable at this point.

 

[stealthology]

5k in transit last quarter was because production ramped late and they couldn't get cars to Europe in time for the end of the quarter

Wasn't this due to the MX seatbelt/seat recall?

 

[esk8mw]

Tesla record stock based compensation in every quarter per their filings, not just in Q1.

Stock based compensation is different than adding to shares outstanding.

My point here is that equity awards (and the corresponding shares outstanding increase as awards from previous years vest/exercise) are typically granted in Q1 of each year at most companies so I would expect a big change in shares outstanding due to awards in Q1, not Q3 (setting aside large sign-on awards like hiring Wheeler, but again we would know about that via 8-Ks). You are correct that equity compensation is typically amortized/expensed over the life of the award and reported as stock-based compensation in filings, though.

 

[Turing]

I went short MBLY (via Sep30 Puts) around $44.17 last week and still have that position open. I think its not done falling yet, but I'm prepared to be wrong on that front. I think a few things have happened with the TSLA/MBLY acrimonious breakup.

MBLY execs wouldn't stop talking about it and shoving their foot further down their throat. They claimed to dispute TSLA's assertions, but didn't actually speak to the most damning accusation TSLA made. That suggests to me that TSLA's side of the story is closer to the truth.

If TSLA is telling the truth, then it suggests that MBLY's modus operandi is to get an OEM hooked on their product, get it designed into the OEM's product, and then start jacking up the price once the OEM is shackled to the MBLY platform. That's a pretty shady way to do business, and will make other OEMs reticent to get in bed with MBLY. I think that doesn't bode well for their future on its own, now that such activities have been made public.

In addition, the v8.0 TSLA update relegating the MBLY sensor to playing second fiddle to Bosch's radar unit seems to be foreshadowing a wider shift. TSLA is the market leader in vehicular autonomy, and if they've decided that radar > visible spectrum camera, you can bet others will soon follow. If EyeQ3/4 is technology no longer needed, that can be replaced with a simple camera module without any integrated processing (as the processing needs to account for the radar feedback, and so needs to be run elsewhere), there goes all of MBLY's advantage.

If MBLY can't adapt quickly, I foresee them being in for a world of hurt. Either way, the damage to their reputation may be irreparable at this point.

Very good points. I think MBLY actions suggest that it is genuinely terrified at the precedent and example Telsa is setting by cutting them out. Made worse by the reality that Tesla has the best driver assist system currently on the road and is on a course to widen its lead.

 

[esk8mw]

Decent write up from Motley Fool summarizing an Aug 31 factory visit: Tesla Factory Tour Reveals Model X Mania -- The Motley Fool

Article confirms 2k/month production in late August and notes an even mix of S and X on the line, with no X production issues. Nothing earth shattering but a good read.

 

[Gerardf]

As I also stated in electrek, To my opinion this probably was the most serious problem of for Tesla :

"(Mobileye) Demanding that Tesla not use data that was collected by its vehicles’ cameras for any purpose other than helping MobilEye develop its products"

Crazy !
Basically that would erease the biggest advantage Tesla has with AutoPilot and developing to higher level self driving, being the data collected continously by 140.000 cars driving around. Data that can be used to feed to the selflearning system , white-/blacklisting to prevent false positives and to improve the maps. The data currently collected by Tesla cars might already have a higher value than the yearly Mobileye R&D spending is PLUS the yearly Mobileye revenue generated by Tesla.

It would not only block Tesla from using the data they collected to improve AP use of other sensor input data like radar, it would also mean that Tesla collected data would directly enable all Tesla competitors to benefit from data collected by Tesla without them compensating Tesla for that.
One could defend that having safer cars sooner will benefit all, but I feel the Tesla competitors should license that data from Tesla, and not get it by means of a Mobileye vendor lock-in.

After reading that, as a Tesla shareholder, I was very happy Tesla dumped Mobileye. I can not imagine other Mobileye customers will accept that demand.

 

[Krugerrand]

Thanks for the character assassination but way ahead of you. Just in case you'd miss it twice I quote my first post again on this subject and put the relevant part in bold. I wasn't the one to bring up this specific analyst. I just corrected someone else's statement on that particular analyst...

 

[TMSE]

Decent write up from Motley Fool summarizing an Aug 31 factory visit: Tesla Factory Tour Reveals Model X Mania -- The Motley Fool

Article confirms 2k/month production in late August and notes an even mix of S and X on the line, with no X production issues. Nothing earth shattering but a good read.

From the article:

Factory changes in the past year-and-a-half revealed a rapidly growing company. The upgrade was most evident in the scale of the new second body assembly line, where an army of recent red robot-arm recruits moved around to perform up to five tasks each. A whopping 580 robots on the body line are building Tesla's Model X SUV, up from the 160 robot arms driving Model S body assembly.

160 robots on 1st line for MS vs. 580 robots on 2nd line for MX. Does it mean the second line is more automated? Or the MX is so complicated that it requires so many more robots to get the same number of cars throughput as for MS?

 

[dakh]

@Gerardf
Look at it from Mobileye perspective: you got a direct competitor that is using your technology in ways you can't control. It's like Apple would be selling iPhones to someone who installs a different OS on them and gains market share with even higher margins. Nobody would do that.

And yes most if not all Mobileye customers seem to be quite happy to just outsource the whole driver aid area to them and let Mobileye collect the data for themselves.

 

[AlMc]

From the article:

260 robots on 1st line for MS vs. 580 robots on 2nd line for MX. Does it mean the second line is more automated? Or the MX is so complicated that it requires so many more robots to get the same number of cars throughput?

I suspect your answer is a little bit of each.

 

[Gerardf]

@Gerardf
Look at it from Mobileye perspective: you got a direct competitor that is using your technology in ways you can't control. It's like Apple would be selling iPhones to someone who installs a different OS on them and gains market share with even higher margins. Nobody would do that.
And yes most if not all Mobileye customers seem to be quite happy to just outsource the whole driver aid area to them and let Mobileye collect the data for themselves.

Well, I have to disagree with that comparison. Mobileye basically wanted to own the data caputered with the camera. I assume google bought a camera as well for their streetview mapping. Imagine the camera manufaturer claiming they own all data google streetview cars collected using it.

P.S. I am also not so sure Apple would have a big problem with what you describe. They do allow installation of windows as a second OS.

 

[dakh]

Well, I have to disagree with that comparison. Mobileye basically wanted to own the data caputered with the camera. I assume google bought a camera as well for their streetview mapping. Imagine the camera manufaturer claiming they own all data google streetview cars collected using it.

Mobileye announced they're gonna be doing fleet learning a while ago and if memory serves OEM's were willing to play. So effectively Tesla wasn't going to become one of their regular clients and they had to part ways. To make good on my Apple analogy, say Google is using ambient phone positioning data to make real-time traffic maps. But is buying up iPhones to then install Android and get the data. This is essentially what's going on.

 

[Gerardf]

Mobileye announced they're gonna be doing fleet learning a while ago and if memory serves OEM's were willing to play. So effectively Tesla wasn't going to become one of their regular clients and they had to part ways. To make good on my Apple analogy, say Google is using ambient phone positioning data to make real-time traffic maps. But is buying up iPhones to then install Android and get the data. This is essentially what's going on.

We will have to agree to disagree. On the whole ownership discussion and on your examples.
I am very happy Tesla did NOT give away one of their biggest assets.

Can Adobe claim ownership all the photos you edited with their software. Can an application developer claim ownership of the macros / scripts you write on top of commercial applications ? Of course not.

Mobileye was not even planning to pay for the data Tesla is collecting. Au contraire, they wanted to increase prices while volumes are exponentially growing !

How much does one Mobileye chip and drivers costs ? And how much do you think Google, Uber, Lyft, GM, Volvo etc, etc would be willing to pay for the data Tesla already collected today ?

 

[Fallenone]

From the article:

160 robots on 1st line for MS vs. 580 robots on 2nd line for MX. Does it mean the second line is more automated? Or the MX is so complicated that it requires so many more robots to get the same number of cars throughput as for MS?

Could be the answer of lower margin on the X (in addition to all the production difficulties they had for almost 3 quarters). When I was modeling the costs for S and X, I found the X costs much more than the 5000 difference in price sold to customers. It gives me some worry about the long term margin of X not being able to be equal to the S. Hope I'm wrong.

 

[Spidy]

California removes limit on carpool access for plug-in hybrids

Probably not that great for Tesla. Pretty much had a Monopoly on premium cars that qualified for the HOV lane for some time.

 

[MitchJi]

Could be the answer of lower margin on the X (in addition to all the production difficulties they had for almost 3 quarters). When I was modeling the costs for S and X, I found the X costs much more than the 5000 difference in price sold to customers. It gives me some worry about the long term margin of X not being able to be equal to the S. Hope I'm wrong.

But more robots is a one time cost.

 

[dakh]

We will have to agree to disagree. On the whole ownership discussion and on your examples.
I am very happy Tesla did NOT give away one of their biggest assets.

Can Adobe claim ownership all the photos you edited with their software. Can an application developer claim ownership of the macros / scripts you write on top of commercial applications ? Of course not.

Mobileye was not even planning to pay for the data Tesla is collecting. Au contraire, they wanted to increase prices while volumes are exponentially growing !

How much does one Mobileye chip and drivers costs ? And how much do you think Google, Uber, Lyft, GM, Volvo etc, etc would be willing to pay for the data Tesla already collected today ?

What exactly do we disagree on? MBLY couldn't just tell Tesla to get lost, so they effectively asked them to behave like they want the rest of their customers to behave. All I'm saying is that's not unreasonable at all on MBLY part. The whole point is MBLY is not selling chips and drivers, they're selling Autopilot competitor to OEMs. The value prop is vastly different.

 

[aubreymcfato]

How much does one Mobileye chip and drivers costs? And how much do you think Google, Uber, Lyft, GM, Volvo etc, etc would be willing to pay for the data Tesla already collected today?

I was wondering the other day if maybe this could be a way for Tesla to gain capital: sell data. They maybe read a tipping point in which they are so sure about their competitive advantage that they can make a dump of 1y of data and sell it to the highest bidder or something. Tesla would be gaining another million miles just the day after that. I bet Google and Apple would be willing to pay a lot of money for that.
Is this a very stupid idea?

 

[TMSE]

What exactly do we disagree on? MBLY couldn't just tell Tesla to get lost, so they effectively asked them to behave like they want the rest of their customers to behave. All I'm saying is that's not unreasonable at all on MBLY part. The whole point is MBLY is not selling chips and drivers, they're selling Autopilot competitor to OEMs. The value prop is vastly different.

The whole point, from Tesla's viewpoint, actually is that MBLY is selling just chips and drivers. Tesla does not need or want the MBLY software on top of it.

 

[brian45011]

ESPP administration is totally blind to investors, so I'm not sure. Administration is also all over the place as the Code doesn't mandate that many rules. There's no guarantee they even did a purchase in August. The two most common purchase periods are 3 and 6 month intervals and can occur at any time. Either way, ESPP shares would probably account for very little in terms of shares outstanding. One company I work with (very close in market cap to TSLA) has monthly ESPP purchases and a high participation rate, but the monthly purchase companywide is about $100k/month. Tesla has far more employees than this company, but I still would be very surprised if an ESPP purchase put an appreciable dent in the $110MM of stock difference from the previous filing. I agree that the note redemption is probably the cause.

IIRC, Tesla's ESPP plan is priced during the first week of February and August, and the employees get a 15% discount to the lower share price at the beginning and end of each semi-annual period. (Check out TSLA's share price in early February, 2016.) IMO, ISPPs are a bird's nest on the ground benefit, and smart employees cash out as soon as they can to re-load over the following six months.

I'm not saying ESPP sales were a material factor in the share count increase you noted, but having tried to reconcile share counts increases to Form 4s over multiple quarters I have formed the impression the ISPP is not inconsequential in the share count accretion.