Comment by Someone | Hacker News Reader

Comment by Someone | Hacker News Reader

Parent Full thread

”Some lines are at only 500 cars a week.”

That sounds better than what Tesla claims. http://files.shareholder.com/downloads/ABEA-4CW8X0/549293785... (emphasis added):

”Other lines, such as battery pack assembly, body shop welding and final vehicle assembly, have demonstrated burst builds of about 500 units per week and are ramping up quickly.”

If you can do burst runs of 15 km/hour, that doesn’t mean you can run a three hour marathon.

”The company burned $1.4 billion and has $3.5 billion in cash. Raising money shouldn't be necessary.”

To get that $1.4 billion in revenue, they will need to sell 40k Tesla 3s at $35k each, or 3,000 a week. To get it in margin, even assuming the average model 3 will sell for $50k, they will have to sell 12,000 a week or so (at 20% margin)

=> If things go as planned, I guess they will need about half their money in half a year or so before turning a profit. That doesn’t leave very much in case things go worse than planned (for example, they plan for 25% margin on model 3, but currently don’t have that on their more expensive models, and they plan to sell more of their expensive models, but can decrease production, so, presumably, they have quite a few of them in inventory)

Tesla may have a fantastic future ahead of it, but it also may fail spectacularly fairly soon, or just have a decent future ahead of it that doesn’t warrant its current share price.

Most auto production lines produce nearly 500 cars per shift. Normal cycle times in an auto plant are about one minute. Here's a Mercedes production line.[1] That's a vehicle on a par with Tesla's products.

Something is terribly wrong in Fremont.

[1] https://www.youtube.com/watch?v=fbfTlwm248k

Most auto production lines don't need EV-size batteries. The post above specifically says batteries are the bottleneck, so it isn't really evidence that something is "terribly wrong"

Since batteries are the whole selling point of a Tesla, I'd say it remains evidence that something is "terribly wrong".

It's almost like Mercedes has been doing this for a long time.

That's a moot point though.

If you want to compete in a market, you got to play at the speed/quality/prices of the market -- there is no slack given for being new.

Tesla's profit margins are fine, this is all a question of ramping up to cover demand. Which tends to be much simpler when you are already a huge company.

Their real risk is if the Model 3's reliability is bad or they need a recall thus reducing their profits significantly.

What they've been ramping up is the Model 3's price point.

"The truth is the Model 3 costs $40,000 if you want a standard version with autopilot (an extra $5,000) in black with no other options. If you want a different color, add $1,000. And if you want a longer range ($9,000) to get over 300 miles per charge instead of 220, well now we’re at $50,000. $50,000 for a midrange car."[1]

[1] https://www.theverge.com/2017/7/31/16069960/tesla-model-3-no...

Autopilot is not a standard mid range car feature.

"Enhanced Autopilot adds these new capabilities to the Tesla Autopilot driving experience. Your Tesla will match speed to traffic conditions, keep within a lane, automatically change lanes without requiring driver input, transition from one freeway to another, exit the freeway when your destination is near, self-park when near a parking spot and be summoned to and from your garage."

You can get adaptive cruse control for more reasonable costs, but lane changing takes a lot more awareness around the car.

Seems like there's a lot of slack given for being new. They have had relative ease raising capital, their market cap is still enormous relative to the number of cars produced, and lots of people are still on the waiting list for a car that they won't get for years.

That's just money burned from people investing -- there's no slack from the market, and none of this does much good if you can't compete on the market.

In the end, unless they get profitable, if they burn their stash, they're toast.

They aren't stopping at 500\week. They are admitting it's broken.

That E-class production line video is interesting, it's basically the same assembly strategy Tesla uses: all bits of the engine, drivetrain, suspension etc. is assembled into a "skateboard" in one line, while the body gets fitted with all the trim, interior, doors etc. on another line and then dropped onto the skateboard (which you see from 6:30 and out in the video).

If Tesla is stuck on a battery bottleneck it's just as silly as if Mercedes were stuck on a sheet metal bottleneck. Yes, you need reliable component suppliers that can scale with you as you grow and provide the quality you need. If you don't have that, you have a big problem that vertical integration isn't going to solve.

Well that's what the redesign is meant to fix, no? Supplier can't keep pace so they're either building it in-house or they've designed something that can be supplied easier/by another vendor.

That E-class production line video is interesting, it's basically the same assembly strategy Tesla uses: all bits of the engine, drivetrain, suspension etc. is assembled into a "skateboard" in one line, while the body gets fitted with all the trim, interior, doors etc. on another line and then dropped onto the skateboard

The Mercedes E-class is monocoque, like almost all cars. They build the body/frame and then add the powertrain and wheels. Here's the body/frame and the powertrain/suspension coming together.[1] The powertrain and suspension aren't self-supporting; they're a collection of parts held in a fixture. That's very clear in the video of the parts coming together.

The Tesla Model S, though, is body on frame. If it can sit on its wheels without a body, as the Tesla Model S can, it's body-on-frame. Here's the Tesla Model S platform sitting on its wheels, with the powertrain installed.[2]

So is the Tesla Model 3.[3]

[1] https://youtu.be/lPKnM9vC6V0?t=395 [2] https://www.youtube.com/watch?v=l8WUIkD7xxI [3] https://jalopnik.com/why-tesla-needs-to-sell-the-model-3-wit...

Relevant Jean-Louis Gassée commentary - https://mondaynote.com/teslas-new-car-smell-315c72c955d3 - referencing BMW Spartanburg's 1500 car per day output, and discussing Tesla's factory.

Based on second hand information from people working with tier-1 Tesla suppliers, it seems that Tesla demanded they set up the Model 3 production in about half the time as other manufacturers. The suppliers knew that the promised dates and production volumes were a total joke but went along anyway just to get the business. The Tesla reality distortion field only goes so far; it doesn't help accelerate the creation of big, complex physical objects like dies and molds.

Apparently the suppliers told Tesla that the dates were impossible, but Tesla management didn't want to hear that and continued to pass on nothing but good news up the chain. It wouldn't surprise me if Elon only recently learned just how bad it is all going.

"To get that $1.4 billion in revenue, they will need to sell 40k Tesla 3s at $35k each, or 3,000 a week. To get it in margin, even assuming the average model 3 will sell for $50k, they will have to sell 12,000 a week or so (at 20% margin)"

It seems like you're assuming that the cash burn will stay at 1.4billion. Expenditure should go down once the production line tooling is complete.

I don't think any of the upfront investment or capital expenditure is included in the 1.4 billion calculation. Thats purely operating loss.

The $1.4bn is $300mn cash flows used in operating activities + $1.1bn capex.

So, can we assume that if the production has to be multiplied by 10, the cash flow used for operating activities should also be multiplied by almost 10, giving expenses of way more than $1.4bn?

No, because a huge cost for them right now is underutilization of assets (factory, tooling, some workers, etc.). Their utilization rate will improve, and so operating costs should not scale linearly with revenue growth. "Should" being the key word.

much of those expenses are one-time cap ex

Tesla originally planned to run two production lines but had to drop to a single one, and they're doing it in an expensive and unusual manner, and a second won't fit in their fremont plant, and they still plan to build out the second line in the same unusual 'cold' manner.

This leads me to expect it's reasonable their burn will continue at the same rate.

Furthermore, it's estimated they won't reach 3k a week till the end of 2018.

https://dailykanban.com/2017/10/source-tesla-responsible-mod...

"Capital expenditures were $1.1 billion in Q3. The majority of capital expenditures were attributable to Model 3 and Gigafactory 1 production capacity increases. "

During the conference call, Musk himself said that CapEx in Q4 could be as high as in Q3.

"Tesla may have a fantastic future ahead of it, but it also may fail spectacularly fairly soon, or just have a decent future ahead of it that doesn’t warrant its current share price."

lol. "it might go up, might go down, or might stay the same!"

Hey at least you know it's not certain to be only one or two of those options.

"If you can do burst runs of 15 km/hour, that doesn’t mean you can run a three hour marathon."

Once you're in 3 hour marathon territory, what you can do for 1 mile is a lot more relevant to what you can do for 26.2 miles than you might think. One of the tent poles of effective marathon/half marathon training is to make your race pace into a drumbeat. You want to feel your race pace in your bones. Very effective workouts look like: take your marathon race pace, and run 5 minutes easy/5 minutes hard (~race pace) for 60, or 10 easy/10 hard for 80.

Tesla is talking about burst builds that last for a week. That sounds like very effective training that will prepare them for race day.

"Tesla is talking about burst builds that last for a week".

Was that in the call? Their statement says "Several manufacturing lines [...] have demonstrated a manufacturing ability in excess of 1,000 units per week during burst builds of short duration. Other lines, [...] have demonstrated burst builds of about 500 units per week...". AFAIK, "short duration" != "a week".

”In Q3, we delivered […] 222 Model 3 vehicles”

Assuming they didn’t produce a lot more than they delivered, if they had even one week in which they produced 500 Model 3s, they also must have had weeks in which they destroyed them.

Probably has something to do with pair particle production [1]: 500 virtual Model 3 is produced along with 500 anti-Model 3 then they are immediately annihilated.

[1] - https://en.wikipedia.org/wiki/Pair_production

Reply on news.ycombinator.com