Hacker News Reader

Comment by itsoktocry | Hacker News Reader

>Tesla posted positive free cash flow (CFO - capex) for 4 of the last 5 quarters

Can you explain how their capex is decreasing as they build out more factories and invest in new technology? Seems odd, doesn't it? Yet is sure makes that cash flow number look good. These are the kinda things that analysts consider red flags.

>it the only company with increase in # of deliveries among the 10 largest autos globally

Do you think this is a reasonable comparison when Toyota builds as many cars in a few days as Tesla does in a quarter? Tesla can futz a few thousand cars with a fleet sale or inventory build and drastically change their growth trajectory.

By the way, what's a "delivery"? I don't think I've ever seen them define it. I would assume it means "car delivered to customer", and yet they build inventory. Very opaque.

View on HN

> Can you explain how their capex is decreasing

Capital efficiency. It should not be surprising that a greenfield factory built in China based on a spec you iterated on in Fremont, CA results in a much higher $/Cars/Day, a.k.a capital efficiency. Iterating on a live line in Fremont, CA is significantly more costly. It's like $/sq ft. for renovating your house versus buying new.

Also, look at the flow diagrams they've published on the floor layout and the path through a factory a car takes to go from start to finish in Fremont vs. GF3. This is also why they're moving to Austin.

> By the way, what's a "delivery"? I don't think I've ever seen them define it.

Of course they define it. In their Annual Report, under a section titled "Critical Accounting Policies and Estimates", under a sub-section titled "Automotive Segment - Automotive Sales Revenue";

We recognize revenue on automotive sales upon delivery to the customer, which is when the control of a vehicle transfers. Payments are typically received at the point control transfers or in accordance with payment terms customary to the business.

> and yet they build inventory. Very opaque.

I'm not sure what you're asking here.

flow diagrams they've published on the floor layout and the path through a factory a car takes to go from start to finish in Fremont vs. GF3.

URL? I found a 2016 Fremont planned layout http://digital.olivesoftware.com/Olive/ODN/SanFranciscoChron... and a 2019 Fremont layout https://villanyautosok.hu/wp-content/uploads/2019/09/2019_09... plus GF4 images https://photos.google.com/share/AF1QipOVTM-avTBP0VMbqc3dynh6... No comparison of process flow, which would be very interesting. All I found was https://cdn.shopify.com/s/files/1/0173/8204/7844/articles/Te... which seems to have more to do with combining multiple buildings than altering process.

Yes, your last link was what I was thinking of. It’s page ~16 of 2019 Q4 Letter:

https://ir.tesla.com/static-files/b3cf7f5e-546a-4a65-9888-c9...

It appears to be the same process. Reading junk on the internet the main process innovation they appear to have made appears is large scale die casting to produce entire chassis frames in a single aluminium part. This - like most "innovations" - is less an innovation and more adopting a known process with different tradeoffs, increasing initial capital outlay to obtain enhanced production speed and a simplified production process.

In vehicle manufacturing process engineering, an interesting recent startup is UK brand 'Arrival' who decided to avoid stamped chassis parts entirely and instead assemble from sheet and extrusions (IIRC). They can allegedly assemble a whole van in a 1000m2 'microfactory' and have just brought out a bus concept. https://www.electrive.com/2020/03/20/arrival-moves-into-new-... https://www.pesmedia.com/arrival-bicester-electric-van-manuf... https://www.pesmedia.com/arrival-zero-emissions-bus-17062020... https://en.wikipedia.org/wiki/Arrival_(company) 3B valuation.

>Can you explain how their capex is decreasing as they build out more factories and invest in new technology? Seems odd, doesn't it? Yet is sure makes that cash flow number look good. These are the kinda things that analysts consider red flags.

From 2017 to 2018, they were focused in ramping-up production at the Fremont factory [0]. And compared to Q2 2019 Capex is actually 118% * higher ! * (546m vs 250m), so I can't see how they are slowing investments.

>Do you think this is a reasonable comparison when Toyota builds as many cars in a few days as Tesla does in a quarter? Tesla can futz a few thousand cars with a fleet sale or inventory build and drastically change their growth trajectory.

I do believe they did a tremendous job in ramping-up deliveries (page 18) in such a short period of time). It is 2x the number from just 3 years ago.

>By the way, what's a "delivery"? I don't think I've ever seen them define it. I would assume it means "car delivered to customer", and yet they build inventory. Very opaque.

By delivery Tesla probably needs to recognize revenue according to ASC 606, or when the product is delivered instead of paid. I'm sure theirs auditors must pay close attention to this number.

[0] https://en.wikipedia.org/wiki/Tesla_Factory

>I'm sure theirs auditors must pay close attention to this number.

I'm sure they do.

https://finance.yahoo.com/news/wirecards-auditors-ey-were-vi...

>And compared to Q2 2019 Capex is actually 118% higher ! (546m vs 250m),

And yet they were building a factory in China then.

>capex is decreasing

Could you give more detail on that? The last big capex increase was due to Model Y rollout. Now that most of the lines are completed, it's just replicating what Fermont's Y production line has in Shanghai's.

I do, however, agree that Tesla's definition of 'delivery' is suspect.

>Could you give more detail on that? The last big capex increase was due to Model Y rollout

     Capex:

     2017 - $4 Billion
     2018 - $2.5 Billion
     2019 - $1.5 Billion
     2020 to date - ~$1 Billion

Where was the big increase in capital expenditure?

Reply on news.ycombinator.com