Hacker News Reader

Comment by gamblor956 | Hacker News Reader

The thread was flagged because the pro-Tesla crowd on HN felt that the article was inaccurate regarding the amount of money Tesla lost on each car, and flagged it over wording.

Tesla makes money on each car it sells (sales exceeds cost of goods sold), but once fixed costs and other expenses like R&D and capital investments are included, it has an enormous loss, so on average it loses about $4000 to $17000 per car (depending on whether you use Unicorn Accounting or Generally Accepted Accounting Practices). The article reported this as Tesla losing $4000 per car, which is accurate in a general/layman's sense but is technically inaccurate.

View on HN

Not to re-hash the argument again but it's not quite an accurate representation since for each car that Tesla sells that $4k number goes down.

They've broken out the per-unit cost for a Model S and they're posted ~22% profit(feel free to fact-check me on this, I might be a tad off) excluding ZEV credits.

So if Tesla wasn't production constrained(and building out the Gigafactory) they'd be turning a profit by the same logic.

From the Shareholder letter:

"Total Q2 gross margin was 23.4% on a non-GAAP basis and 22.3% on a GAAP basis."

http://files.shareholder.com/downloads/ABEA-4CW8X0/522530778...

Thanks, that's the number I was trying to find and had seen before.

Gross margin is the income from sales less cost of goods sold, before items like R&D, salaries, capital improvements, etc., are considered. As I said above, Tesla has a positive gross margin, but a major net loss. When that net loss is averaged to units sold (which is a common metric for measuring financial performance in the automotive industry), Tesla has a loss of approximately $4000 to $17000 per car. This doesn't mean that Tesla is actually losing money each time a car is sold, it simply means that Tesla isn't selling enough cars to achieve net profitability.

If Tesla were to sell more cars, its average loss per car would go down (and eventually become average profit per car). Unfortunately, Tesla announced that it's cutting production and sales estimates for the remainder of the year, so that's not going to happen in 2015.

Do you know if that's because Tesla is not selling enough cars, or the market demand is just not there? China used to be thought of as the big market eventually but with everything going on there, it seems that's not a good assumption anymore.

The utility of a hybrid is just much higher. You get the gas saving in city drives, and the range on the odd weekend trips. The latest Prius can be charged too: http://www.toyota.com/prius-plug-in-hybrid/ and it cost less than 50% of the cheapest Tesla. The only thing you don't get in these Prius is the look factor, though they look cool enough for many people.

>When that net loss is averaged to units sold (which is a common metric for measuring financial performance in the automotive industry)

I agree the metric is common, but it is an invalid comparison for Tesla. The other car companies have mature product lines. The model S is their only production line, and they are spending money on development for the Model X and the Model 3.

If any car company were developing 2 vehicles for every 1 they have in production, then it would be a valid comparison. Comparison metrics have to be evaluated in the context of the company's overall life cycle. The art of valuing companies lies in choosing the relevant metrics to compare.

Tesla has lowered their current year estimates from 55,000 to 50,000 to 55,000, due to uncertainty about the Model X go live. And it hit their stock price... However, they have not changed their long term goal or time frame for get their factory to maximum capacity of 500,000 vehicles around 2020.

On the other hand, Tesla's accounting only figures on spending a few hundred dollars per car to pay for free, unlimited SuperCharger access for the lifetime of the car. Which is frankly absurd. You put a free, unlimited SuperCharger in somebody's backyard and of course they're going to use the heck out of it; it's like getting paid $15/hr to read the internet or a book or whatever else you do while charging. That's free money.

At 22% profit without credits that means with credits they're making say at least $40k profit per car sold. This is like 1/4 year of sales... but they need ever more money, now? Why? Maybe Model X reservations are cannibalizing S sales? Or they need to throw money at 3 because the competition is going to have an equivalent car out sooner? Or so battery factory comes online before LG Chem's new production?

This doesn't add up for me. It seems to me either their accounting is not accurate or they are seriously far behind the competition.

Tesla charges $2,000 for lifetime supercharger access. I don't think screwing that up is going to cause them big financial problems: it's just not that much money, and they can always raise the price for future buyers.

Help me out here, 2014 Q3 says $21.9m deferred for SuperCharger electricity minus $10.3m as of previous year, so they deferred $11.6m from 2014 sales of ~22,000 cars. That's ~$500/car accounted for the lifetime SuperCharger costs. Tell me where I'm wrong.

At $10 a charge (12 cent kWh) that's 50 charges per car for its lifetime. That seems like a pretty low estimate to me on number of charges, especially as the network expands.

The difference to your $2000 works out to $33m (more for the full year). That seems like a significant amount to me.

There's also the capital cost of installing superchargers to take into account (which only supports your argument further).

That may be why the superchargers aren't in anyone's backyard.

What does Unicorn Accounting mean? A general term for bullshit accounting? Ala Groupon calling it's sales team's salary a capital expense in it's IPO filing?

The classic phrase is "EBITDA", or "earnings before interest, taxes, depreciation, and amortization". This is sometimes referred to as "earnings before all the bad stuff". Then there are "extraordinary items", which somehow are always in the direction that makes the balance sheet look better.

In the days when AOL sent out disks with their program, it came out years later that they were treating that marketing effort as a capital expense and amortizing those throwaway disks over many years. AOL actually became profitable about six years after it said it did.

I used to have a system which tried to analyze financial statements automatically. I have a database of several hundred euphemisms for "net loss". ("Net income after extraordinary item", etc.) It's embarrassing. The SEC's standardization on XBRL has helped a little with that.

Tesla is an industrial company, and accounting for industrial companies is well understood. This includes growing industrial companies. Upfront expenditures for an industrial company result in real assets, which can be valued. I look forward to reading the prospectus for the IPO.

That's not really correct. EBITDA is not earnings before bad stuff - that stuff is all % of your income and/or write downs that have no impact to the business, so financial folks like to exclude it as noise (since all it really does is reduce your taxes). Examples:

- You bought a company and now it's worthless - You made money and now you have to pay taxes on it - You bought a bunch of computers and now they're three years old and worth less than when you bought them - You bought something 10 years ago, and, instead of paying for it all up front, you paid over time

I'm going to have to start using the "unicorn accounting" phrase (vs. GAAP). Love it.

It's nice but it kinda hints that GAAP is "the right thing", which is a bit too optimistic IMO.

Particularly when you look at any company which is growing very fast. SolarCity would be an example where the company is growing exponentially, but the returns are spread over 20 years instead of being realized quarterly.

Well, accounting is not supposed to tell how valuable a company is, just what its balance sheet is, right? That people then take the balance sheet and invent ways to mechanically arrive at a valuation based on that is not GAAP's or any other accounting rules' fault.

The trouble with any accounting rules is that it's hard to convert everything on the balance sheet into the same unit. If you don't convert it all, it's really hard to compare balance sheets. If you do convert it all, this conversion may be unfair to one balance sheet more than another. Or at least that's how I understand it...

One example is pricing stock option grants so that you can then add them up with other expenses. I'm sincerely not sure how I'd do that. Here's a critique of what GAAP mandates:

http://www.cypress.com/documentation/ceo-articles/dr-rodgers...

In this case it is a bit unfair.

The difference is mainly about leasing agreements for Tesla cars and how much such a lease is worth to Tesla. GAAP was/is definitely too pessimistic. Tesla's own accounting may be too optimistic. I believe Tesla's guess is pretty good. We'll see.

Needs more symmetry, I think: Unicorn Accounting Principles (UAP), by analogy with Generally Accepted Accounting Principles (GAAP). There should also be UFRS (for IFRS).

Reply on news.ycombinator.com