Hacker News Reader

Comment by annon | Hacker News Reader

Giving the founders 20-to-1 voting rights over regular investors while losing as much money as they are is insane. This really looks like a lemon, and the VC's want to cash out and leave public investors holding the bag.

I don't see how self driving cars are going to help them. Waymo and Tesla are both the farthest along, and they're going to run their own networks - competing against Lyft and Uber.

View on HN

Their share structure prevents them from ending up in ETFs, so one of the largest segments of investors literally won't be able to invest in them. Seems more like VC wanting to cash out and leave employees who can't sell their shares yet holding the bag.

> Their share structure prevents them from ending up in ETFs

I've seen several people say this but I don't understand why, can you explain a bit further?

Dual-class share setups are typically prohibited by most ETFs. That's why companies that do want to have more than one share class will trade them under different tickers ($GOOG vs $GOOGL).

Edit: Prohibited was the wrong word, looked down upon was probably better.

Thanks, can you explain a bit further? (Nerdy curiosity here, I've done a lot of legal work with mutual funds & hedge funds but have no experience with ETFs so I'm both clueless and curious.) Prohibited by what? The prospectus? And what exactly is prohibited that assigning a different ticker gets around the prohibition? SNAP has a similar structure but I don't remember hearing the same buzz about that being an issue for ETFs when they went public.

A lot of it seems to come down to voting power. I linked a few articles that dive into it in this comment: https://news.ycombinator.com/item?id=19341955

Do you knwo why that is? It seems like having the two tickers would have all the same issues.

Agreed, and now that I think about it a bit more the ticker symbol is largely meaningless. Institutional investors trade by CUSIP, not Ticker, and a unique CUSIP is assigned to each security (and each share class of a company's stock is a unique security) regardless of whether they share an issuer.

Different classes always trade with different tickers. When they trade, of course: Google has also a non-traded B class. (By the way, I don’t think the first statement is correct.)

Prohibited was the wrong word, you're right. Institutional investors that run ETFs have become averse to dual class stocks ($SNAP, $LYFT) because they don't have the same voting power per share that they do with other share class setups. Indexes like S&P 500 have also stopped including listings that are dual-class -- this is changing though, see the articles I've linked below.

- https://www.cii.org/dualclass_stock - https://citywireusa.com/professional-buyer/news/crown-duals-... - https://www.washingtonpost.com/business/dual-class-shares/20...

There has been a lot of discussion about this issue in the last years. Dual-class setups (at least for new IPOs) are penalized now by some index providers.

https://citywireusa.com/professional-buyer/news/crown-duals-...

Ahh, thank you. That makes much more sense. So it's not that ETFs are specifically prohibited from purchasing them. Rather, if index providers exclude issuers who have a dual-share structure from their indices then any funds that track to an index wouldn't be purchasing stock in those issuers.

I think it's important to note that many ETFs aren't index-tracking so even though index providers may exclude the issuers with several share classes that doesn't mean the stock will be shunned by all ETFs.

I don't really disagree with you, but I think (or, at least, hope) that there will be room for more than 2-3 players in the "self driving taxi" market.

The thing that would cause that is if there is some requirement for massive amounts of difficult to obtain data that is required to build a competitive system.

I do think data is a competitive advantage right now. But I find it really hard to believe that 10 years from now it will be harder to build a self driving system than today. That's just not how tech is. Building YouTube was a herculean endeavor 15 years ago, but today you can hack a YouTube clone together in hours.

Not just will we have more off the shelf software and hardware, but maybe you can license the data too.

In the end, I don't see the moat around a self driving cab company. If someone has a good app and a single car that operates in my area, why not switch to them?

Sure there are wait times and availability in odd places. But that doesn't inhibit a startup getting early adopters. You could literally just do the same commutes every day and have a profitable business.

Lastly, I expect the variety in vehicles and "mobile spaces" will provide a huge landscape of opportunity that a single company will not be able to fill. Just like there is not one "housing" company, there won't be one mobile housing company either. Too much variation in taste and preference.

Personally, I think that Tesla could own the entire market, regardless of operator, if it were to develop the defacto standard rideshare vehicle, and financed it as such a % of the rideshare market.

See every yellow cab in NYC, as an example, is the same make/model of car.

Make a tesla Y vehicle and have the rideshare payment subsidize the car. So X% of every single ride goes to tesla to finance the car. And guess what they get as defacto: the data.

Let people qualify as a driver and put down a deposit, and the car must meet a quota of rides per month, which is tracked and displayed as the car is simply a 'device' at that point... and must be used to pay it off. Tesla doesnt care where the rider bookings come from, lyft, uber, whatever.

If the car is used for private purposes - then the owner is charged some function of the car's time as his "car payment"

Simple.

Part of the reason rideshare works is because it puts the capex and credit risk on the driver, not the ridesharing company. Taxis are nice and all but that model didn't grow out of the most densely populated markets for a reason.

This model would be better suited for car rental companies, not manufacturers, to utilized unused assets (which I think some have already piloted).

Does my model still not work? Whereby instead of a "car payment" -- a price per mile/slice of your rideshare cost is automatically garnished from your services? And tesla pushes that particular vehicle as a vehicle that must be used as a rideshare vehicle - regardless of if its uber or lyft or whomever - and they reap the data and 'users flocking to their mobile devices'.

Who is paying for the cost of the car up front? That is the person who is carrying the capex and credit risk of this operation.

Even with leases, the lease holder is paying with debt up front and is on the hook to fulfill payments, otherwise they're legally liable and their credit suffers. A rideshare rev split (or affiliate model) doesn't ensure the driver will cover the cost of the car and the manufacturer is on the hook with the risk of a depreciating asset not making money. If you do tie this to credit: - i.e. the driver pays in debt and loses if they don't fulfill rideshare payments - then you're just creating a lease with more strings attached.

> Who is paying for the cost of the car up front? That is the person who is carrying the capex and credit risk of this operation.

Banks, ostensibly. That's what banks do, buy credit risk.

Or Tesla might just take the risk themselves, they continue to be in a reasonable position to raise cash. That's not typical for them though, they've traditionally relied on banks for consumer financing.

This is speculation now, but I get the vibe from Elon that he is interested in a revenue share in the ridesharing world, but that he doesn't think Tesla should be own its own fleet. That puts them into a much more airy fairy financial model, and I think Elon likes the economics of selling cars. Tesla is already a hard enough sell on Wall Street as it is, without a big fleet of cars depreciating on the books.

That is fundamentally not how banks work - they lend money to the manufacturer or the driver, meaning the latter two pay the cost and carry the risk. Banks are not paying the cost.

Tesla has issues with cash and liabilities at the moment, and I highly doubt that any type of scheme like you're describing would be floated by them in the near future.

> Building YouTube was a herculean endeavor 15 years ago, but today you can hack a YouTube clone together in hours.

Ha, good one.

Would your clone have large scale spam, fraud, and abuse systems in place? Would it work on all browsers, mobile devices, TVs, set-top boxes, etc? Is your streaming tech cost-efficient and can it deliver reliability across all regions? Will you be able to respond to DMCAA takedown requests and comply with IP laws across states and countries?

I can go on and on.

Simply put, serving a video over the internet isn't rocket science. Building YouTube...is much more like rocket science.

You don't need those things until after you hit scale, so I wouldn't include them in an initial quote.

There might be but the point is that Lyft and Uber valuation is deflated if you assume they are not the dominant forces in a self-driving car industry.

Reply on news.ycombinator.com