TMC is an independent, primarily volunteer organization that relies on ad revenue to cover its operating costs. Please consider whitelisting TMC on your ad blocker and becoming a Supporting Member. For more info: Support TMC
  1. TMC is currently READ ONLY.
    Click here for more info.

TSLA Market Action: 2018 Investor Roundtable

Discussion in 'TSLA Investor Discussions' started by AudubonB, Dec 30, 2017.

Tags:
Thread Status:
Not open for further replies.
  1. Buckminster

    Buckminster Active Member

    Joined:
    Aug 29, 2018
    Messages:
    2,887
    Location:
    UK
    The FUD has got to Mulder and Scully. Clearly Troy has been abducted by aliens and replaced by a shape shifter.

    As previously stated, I am more interested in a blow out Q1 - 7000 steady production is my priority. 80k+ the goal combined with S&P500 inclusion is unstoppable.
    Also, the fact that I have sat in an M3 in the UK suggests that there are a number of M3s sloshing around that Troy is unable to account for somehow.
     
    • Funny x 9
    • Helpful x 1
    • Like x 1
  2. Fact Checking

    Fact Checking Well-Known Member

    Joined:
    Aug 3, 2018
    Messages:
    7,517
    Location:
    Vienna
    #70162 Fact Checking, Dec 25, 2018
    Last edited: Dec 25, 2018
    I have no good estimates for Q4, other than that I think production will be higher than the 53k estimated by @Troy.

    The reasons for my numeric uncertainty:
    • Both @Troy and the Bloomberg tracker is projecting ~53k Q4 production.
    • The LEMUR VINs were allocated in a single huge batch, and there's only been a trickle of new VIN registrations in Q4.
    • For the first time ever during the ramp up Tesla did not provide official Model 3 production guidance for the next quarter.
    • "Days from VIN allocation to delivery" dropped significantly during Q4, which accelerated the drop in the VIN survey data set sample rate.
    • We've heard from carsonight that GF1 was producing battery packs full tilt during most of Q4 except 3 down days since Thanksgiving. (Which might be an end of quarter inventory management measure already.)
    • We heard from Vicky Salvador of a daily Fremont peak of around 1,375 car bodies through the paint shop in a single day, which extrapolates to 9,625 S3X units per week, or 7,625 Model 3's per week peak rate.
    • Elon is tweeting rather happy tweets. If production was below Q3 levels I think he'd be worried a lot. Instead he is tweeting about liquid metal themed stainless steel starships and is making fart app jokes. Not a peep about pain, hurt, Hell or bankwuptcy.
    • There's still "demand levers" left, which Tesla has not pulled in Q4 yet:
      • Still no leasing for Model 3's in the U.S. - which is ~30% of car sales in general.
      • No S/X leasing perks like in Q3, which suggests all S/X's available in Q4 are already spoken for.
      • No Supercharger perks such as "last chance" unlimited Supercharging offers.
    • I.e. everything suggests Tesla is not worried about Q4 demand and not worried about the end of quarter inventory push, whatsoever.
    So, as I said before, I think the Q4 production estimates of 53k by Troy are probably conservative, but I have no good data to support it other than the four circumstantial indicators which were reliable in the past: carsonight, Vicky, Elon's mood and the "demand lever meter".

    So my estimate is 54k-76k for Q4 M3 production and 56k-80k for deliveries, which 30% spread is worse than useless.

    Not advice. :D
     
    • Like x 17
    • Love x 6
    • Informative x 1
  3. EVNow

    EVNow Well-Known Member

    Joined:
    Sep 5, 2009
    Messages:
    9,240
    Location:
    Seattle, WA
    I think 60k is a good ball park figure. More than 5% higher than Q3.
     
    • Like x 3
    • Informative x 1
  4. Mich

    Mich Member

    Joined:
    Feb 19, 2017
    Messages:
    454
    Location:
    Germany
    #70164 Mich, Dec 25, 2018
    Last edited: Dec 25, 2018
    Why does everyone here believes the ramp stalled in Q4? Did I miss something? I didn't read the spreadsheet nor I'm following any VIN tracker, but why the hell should the ramp just more or less stalls at the Q3 level?

    End of Q3 peak rate was 5k/week if I remember correctly. End of Q4 peak rate will be 7500/week or even more. That's +50% I know, peak rate and not sustainable production, but this +50% will be visible in total production was well.

    There can be hickups in the supply chain etc. but who cares? The stock is crazy anyway, no matter how perfect Tesla executes. At least I know the product is perfect, there's no demand issue and I don't care if the stock drops out of nowhere or just because of two weeks low production.

    Don't panik.
     
    • x 6
    • x 1
    • x 1
    • x 1
    • x 1
  5. Fact Checking

    Fact Checking Well-Known Member

    Joined:
    Aug 3, 2018
    Messages:
    7,517
    Location:
    Vienna
    #70165 Fact Checking, Dec 25, 2018
    Last edited: Dec 25, 2018
    Agreed (note that 60k production would be +13% over Q3, which technically is more than +5% higher. ;)):
    • carsonight is predicting 60k-65k production for Q4, based on his GF1 insight.
    • In Q2 Tesla was aiming for a 5k/week peak, in Q3 for a 6k/week peak and for Q4 it's 7k/week. Production was 53.2k in Q3, so if average volume scales with the peak this extrapolates to 53.2*7/6 = 62k Model 3 production in Q4.
    • 65.7k production in Q4 would be the strictest threshold to reach a pro forma 5k/week sustained production rate mentioned by Moody's as a condition for an upgrade.
    • Troy and Bloomberg are predicting 53k.
    Caveats: these are limited data points with significant assumptions.

    Lack of official guidance makes Q4 really hard to predict - but I don't blame Tesla for that, considering the crap SEC is giving them over missing 2017 production guidance ...
     
    • Like x 8
  6. Fact Checking

    Fact Checking Well-Known Member

    Joined:
    Aug 3, 2018
    Messages:
    7,517
    Location:
    Vienna
    So one plausible line of thinking would be:
    • In Q1-Q2-Q3 Tesla had to ramp up aggressively.
    • But fast ramp ups are risky: you stop production for 2-3 days and hope that the upgrade makes up for the lost output and then some.
    • At higher production rates intrusive upgrades are more risky, because lost output is much higher. When you make $75m/week that's an extra risk.
    • Tesla might have decided to "play it safe" in Q4 and earn another 1.4 billion dollars like they did in Q3 and pay off the $400m debt due this quarter and the $920m debt due in March.
    This is a plausible explanation for a "stall" in the ramp-up.

    I do think they made more cars than in Q3, but there's not enough data to make good quantitative predictions, IMHO.
     
    • Like x 7
    • Helpful x 3
    • Informative x 1
  7. BOBTHEJOCKEY

    BOBTHEJOCKEY Member

    Joined:
    Apr 21, 2015
    Messages:
    213
    Location:
    STOCKTON, CA
    I check the Bloomberg car counter website almost everyday and the numbers are almost always slightly over 4,000 a week. This makes no sense given they recently made 1000 cars in a single day.
     
    • Like x 3
  8. BOBTHEJOCKEY

    BOBTHEJOCKEY Member

    Joined:
    Apr 21, 2015
    Messages:
    213
    Location:
    STOCKTON, CA
    Here it is Christmas day and they are still promising deliveries before Jan1. I am assuming everyone knows that Elon Musk is guaranteeing the full $7,500 tax credit. I am guessing they are figuring that a discount of half of the $7,500 tax credit ($3,750) is worth it to guarantee lots of work for the January slow down. Not sure what this means though. Usually when you lower prices you are having a demand problem.
     
    • Disagree x 9
    • Like x 1
  9. Fact Checking

    Fact Checking Well-Known Member

    Joined:
    Aug 3, 2018
    Messages:
    7,517
    Location:
    Vienna
    #70169 Fact Checking, Dec 25, 2018
    Last edited: Dec 25, 2018
    I think one way to trade (deep out of the money) options is to write off the money mentally the moment the contracts are bought. It's a probabilistic instrument with a -100% return (full loss) much more probable.

    It also helps to use deep out of the money options: there you start with an instant -20-30% loss due to poor spreads and low liquidity, and the expected trajectory is an even deeper drawdown as you average into a position, as you rarely catch the true bottom of the market.

    This can only be stomached if you literally don't care about the loss. Consider it a rented or leased "chance", with fixed monthly payments.

    The only other method I'm aware of to cope with very high leveraged volatility emotionally is frontal lobe lobotomy. :D
     
    • Like x 4
    • Helpful x 3
    • Funny x 3
    • Informative x 1
  10. Fact Checking

    Fact Checking Well-Known Member

    Joined:
    Aug 3, 2018
    Messages:
    7,517
    Location:
    Vienna
    #70170 Fact Checking, Dec 25, 2018
    Last edited: Dec 25, 2018
    They are not promising that at all: they are only saying that if and only if Tesla says they can deliver an inventory car before Jan 1 they guarantee it.

    Late in the quarter quite a few inventory sales happen across stores, for example in an LA Tesla Showroom a prospective buyer might be told: "There's an inventory M3P where the owner was a no-show in Fremont, can you take delivery in 2 days?". If yes then Tesla ships the car over to LA.

    I.e. Tesla is telling customers that if there's still a car somewhere in inventory that Tesla promises to deliver before January 1 when the order is placed, Tesla will pay if they cannot meet that promise.

    The effect: customers who order in Tesla Stores can safely assume they can get the full credit as long as Tesla promises Q4 delivery - even if delivery happens to slip due to a logistics hickup.

    Alas Tesla has not lowered prices, so the opposite is probably the case.

    Tesla wants to "flush inventory" at the end of Q4, but end of Q4 is special wrt. the tax credit - so Tesla is reassuring all inventory car buyers that their tax credit is safe if they buy an inventory car.
     
    • Helpful x 5
    • Informative x 4
    • Like x 1
  11. AudubonB

    AudubonB One can NOT induce accuracy with precision!

    Joined:
    Mar 24, 2013
    Messages:
    7,951
    If "...worse than useless" is Not Advice, I am with child in waiting to learn just what Advice could be.
     
    • Funny x 2
  12. ZeApelido

    ZeApelido Active Member

    Joined:
    Jun 1, 2016
    Messages:
    2,641
    Location:
    The Peninsula, CA
    Unfortunately for you and Nate Silver, election prediction is a classification problem, not a regression problem. Very common machine learning application.

    That's like saying some autonomous vehicle keeps running through red lights but saying "hey we almost thought it was red!"
     
    • Disagree x 3
    • Funny x 2
  13. EVNow

    EVNow Well-Known Member

    Joined:
    Sep 5, 2009
    Messages:
    9,240
    Location:
    Seattle, WA
    Kramer's "Bear Stearn's Fine" on 3/11/08 ?
     
    • Funny x 2
    • Informative x 1
  14. EVNow

    EVNow Well-Known Member

    Joined:
    Sep 5, 2009
    Messages:
    9,240
    Location:
    Seattle, WA
    #70174 EVNow, Dec 25, 2018
    Last edited: Dec 25, 2018
    Again bad analogy. BTW, do you at least understand Silver doesn't do any turnout predictions ?

    You are predicting the % of votes - not a win/ not -win. You are assigning probability to the various % of votes and then calculating the probability of a win.

    If you are an autonomous car - and your probability calculation of something being a red light is 35%, what do you do ?

    The problem is - most people don't really know what 538 put out before '16. They heard about NYT's Upshot, HuffPost, Sam Wong and others giving over 90% chance to Hillary and then blame Silver for it.

    Look at the mid-terms. Hundreds of predictions - so you can better judge how well the probabilities reflected the reality. Predicted 39 Dem pickups in the House, actual is 40.

    How FiveThirtyEight’s 2018 Midterm Forecasts Did
     
    • Informative x 5
    • Love x 2
    • Like x 1
  15. Fact Checking

    Fact Checking Well-Known Member

    Joined:
    Aug 3, 2018
    Messages:
    7,517
    Location:
    Vienna
    Actually, making a probabilistic election prognosis is very valuable: anyone making a high leverage bet on the 33% chance Nate Silver saw Trump winning the presidency was earning a lot of money in 2016.

    It's only the MSM news cycle "horse race" narrative that insists on certain predictions - but that's for entertainment and viewership maximization purposes. Life is more complex, and Nate Silver correctly modeling the close outcome was an excellent scientific result.

    Don't fall for the MSM narrative.
     
    • Like x 3
    • Helpful x 2
    • Love x 1
  16. Cartegena

    Cartegena Member

    Joined:
    Jan 1, 2018
    Messages:
    101
    Location:
    Colombia
    Where will those be stamped and painted? Packs from GF-1 or local? S's, X's, 3's and Y's?
     
  17. SOULPEDL

    SOULPEDL Supporting Member

    Joined:
    Jul 25, 2016
    Messages:
    2,797
    Location:
    Arizona
    GM as in General Motors? These acronyms are confusing sometimes.
    Like OT... Off Topic I assume? Could be On Topic too...

    OH!!!! GM = Gross Margins. Took me a while... see what I mean! I guess that was fair considering the topic.

    Yes, more earnings!
     
    • Funny x 3
    • Helpful x 1
  18. Causalien

    Causalien Prime 8 ball Oracle

    Joined:
    Nov 19, 2012
    Messages:
    3,738
    Location:
    Pothead's Republic of Canukstan (PRC)
    I just used the bloodbath to buy back all the calls I sold. Really don't like to pull this move because it reduces my cash flow during a down spike, but it was too good of a deal to pass up.

    Surprisingly, the $24 drop yesterday only affected mid length calls by $1. Looks like option traders are calling this drop a bullshit.
     
    • Informative x 11
    • Like x 2
    • Love x 1
  19. AudubonB

    AudubonB One can NOT induce accuracy with precision!

    Joined:
    Mar 24, 2013
    Messages:
    7,951
    MODERATOR:
    Anyone have a problem with a lot of posts moved to "Politics"?
    There's your Christmas gift from Lord Vetinari: you get to choose whether you want them ported over.
     
    • Like x 9
    • Love x 4
    • Helpful x 2
  20. ZeApelido

    ZeApelido Active Member

    Joined:
    Jun 1, 2016
    Messages:
    2,641
    Location:
    The Peninsula, CA

    No MSM narrative. I work in machine learning / prediction so I understand proper evaluation metrics. Fine he doesn't model voter turnout, he assumes the polls' models are correct. Uh, Even worse?

    Of course most models output probabilities, that doesn't mean it's the metric being evaluated. Nate Silver may claim it is, but most people digesting that information are looking for prediction of the winner. Not how close. Sure for bettors that information is valuable, but not the general public.

    The metric is the winner, not how close. That's why Taleb has demolished Silver. He has no soon in the game.

    Take 2 election models. One is horrible at getting the vote breakdown correct, but nails the final result everytime.

    The other is fantastic at nailing all the details, but messes up some of the final results.

    The first midel is better, for the stated problem, a classification problem.
     
    • Disagree x 3
    • Like x 1
Thread Status:
Not open for further replies.

Share This Page

  • About Us

    Formed in 2006, Tesla Motors Club (TMC) was the first independent online Tesla community. Today it remains the largest and most dynamic community of Tesla enthusiasts. Learn more.
  • Do you value your experience at TMC? Consider becoming a Supporting Member of Tesla Motors Club. As a thank you for your contribution, you'll get nearly no ads in the Community and Groups sections. Additional perks are available depending on the level of contribution. Please visit the Account Upgrades page for more details.


    SUPPORT TMC