Hacker News Reader

>If I put that same $40,000 in an index fund that got the stock market's historical 7% average, and waited 30 years, I'd have $308,000.

Are you doing the math all in present dollars? Because the return is 7% after inflation but >10% before. So you'd have >500.000 in 2047 dollars. In this case it's important to do that adjustment because you're doing an investment upfront for payoffs over 30 years so the value of money adjustment is extremely important. It should only make your case stronger though if you haven't done that yet.

Energy prices, and thus savings, are certain to increase over that 30 years too. Quite likely at a rate exceeding inflation.

Energy prices have been mostly flat relative to inflation for 40 years, and there's a reasonably good chance that they will start dropping appreciably (at least during peak daylight hours) as solar continues to make major inroads.

That's true in a lot of places but not really in California, where there's a lot of customers for solar. I don't need to go into the reasons why. It's California.

Why so? California's climate is pretty optimal for solar generation and there's a lot of desert waiting to be covered in solar panels.

Solar is not going to have a similar impact on (say) Ireland in the next 30 years.

More solar will likely result in higher grid energy prices, since people will buy a lot less energy while the grid costs the same to maintain.

Solar can also reduce grid costs, by e.g. shaving the yearly peak power demand, which is usually a summer afternoon so it's not clear cut.

Probably the bigger impact is bad incentives for the people building the grid, who are often able to do what they like and get a guaranteed percentage return on top, which pushes them to spend more than they need to. This has been particularly pronounced in Australia I believe.

Usually when there is additional competition (grid owners generate and supply electricity), markets respond by dropping prices.

This might also be why utility companies have been furiously lobbying against rooftop solar subsidies and Barclays have downgraded US utility bonds.

Solar leases use the 2% inflation rate too as a way to make them look more attractive (as does the solar roof calculator) but in reality electricity hasn't been getting more expensive at that rate.

People made similar pronouncements about gasoline a decade ago. $4/g gas is a distant memory.

Not clear on the historic evidence, and especially not clear when you consider people swapping cap ex for op ex.

Increasing renewable energy sources and an abundance of fossil fuels is going to increase energy prices at a rate higher than inflation? I don't think so.

The case can be made even stronger if the consumer's efficiency gap is taken into account.

For example from "Market failures and barriers as a basis for clean energy policies by Marilyn A. Brown" we read:

"Meier and Whittier (1983) studied a case in which consumers were given a choice in stores throughout the United States of two refrigerators that were identical in all respects except two: energy efficiency and price. The energy-efficient model (which saved 410 kilowatt hours per year, more than 25% of energy usage) cost $60 more than the standard model. The energy-efficient model was highly cost-effective in almost all locations of the country. In most regions, it provided an annual return on investment of about 50%. In spite of these favorable economics, which were easily observed by the purchaser, more than half of all purchasers chose the inefficient model. The higher purchase price of the efficient model was presumably the principal barrier to its purchase."

An annual ROI of 50% is left on the table because people have to pay something upfront for benefits later!

Tesla's pricing strategy here is quite blunt. They hope people don't care about the costs, only the looks and the sustainability.

Disclaimer: I don't know how sustainable the manufacturing process of this solution is.

looks and clique value. never underestimate pretentiousness. you can use this to get people to do good things

I would assume (hopefully) that the following is true: my guess is that Tesla is currently executing a similar plan to produce solar products in the same format they produced the cars - expensive, high end up front. Roll your moderate price, high end products with the income, rinse and repeat.

I think if this is the case, they certainly don't want to hint at it incase they lose consumers who will sit back and wait for the cheaper option they ironically wouldn't arrive.

A 7% yield is an overly optimistic number.

Not really. VTSMX has delivered that over 20 years.

Thats the historical average.

Though the market is expensive now resulting in a lower expected return going forward.

You're not deducting income tax on the dividends.

Or they're putting it in a tax-advantaged account.

If you expect either 7% real returns or 10% nominal returns from the stock markets over the next few decades, you're gonna have a bad time - mostly because real interest rates in the 21st century are much lower than in the second half of the 20th.

Current expert thinking is that there is a 4% equity risk premium over cash interest rates. Current cash rates are 1% nominal or -1% real. If you are an optimist like the Fed, cash rates will rise to 3% over the long term - which gets you to 7% nominal/5% real at best.

I wasn't making a prediction on future returns just commenting on the historical returns used. For this case 5-7% should be more than enough as a comparison to the Tesla investment anyway. I'm curious about those risk premium estimates though. Any pointers or sources on where to read more about them?

> Tesla's calculator estimates $40,500 for a roof that generates half our power needs, and should provide $20,400 worth of energy over 30-years.

This depends quite strongly on where you are. Your numbers are very close to the ones I get if I use my correct address, which is in the Puget Sound area of Washington State.

If I keep everything the same except the address, changing that to one in Merced, California, then it says I'd get $75000 worth of energy over 30 years.

PS: it looks like they try to use the address for more than just finding the location in order to estimate how much sun is available. I tried an address that I had lived at in Pasadena, California, and it did not ask me for the house size and number of floors. Instead, it did the calculation using about 6000 sq ft, which is about the size of the apartment building that is at that address. For my Seattle-area and Merced addresses, it did ask me for the house information.

They are using Google's Project Sunroof[1] to do the modeling. The tool uses satellite images of your roof to create the estimates, and accounts for orientation, shading from trees, etc...

[1] https://www.google.com/get/sunroof

That's amazing. Just seriously amazing.

Funny enough, it was someone's side project at Google.

So was Maps, originally. This is why there's so many fans of the 20% project concept.

A distributed file sharing system (later known as the Internet) was Tim Behrners Lee's side project at CERN for a while too :)

He was only allowed to take his 20% project to 100% after his manager wrote up a justification XD

> later known as the Internet

My turn to be pedantic, I guess, but 80 and 443 are only two of many available ports. The internet is much bigger than just the web, even if lots of people never bother to look beyond their browser.

Pedantry is fine by me when accurate and done in the spirit of learning more nuance about a subject :)

An important clarification

If you ever get to go to Geneva, swing by the CERN exhibit. There's the original fileserver, complete with 'This machine is a server. DO NOT POWER DOWN!!' scrawled on tape stuck over the power button. There's also his paper describing the WWW with the feedback of his manager written across the top: "vague but interesting...".

It's kind of comforting to know that my academic and computing experiences thus far are shared both across the globe and across decades.

Haha "vague but interesting".

Isn't it amazing how big of a disconnect there is between "people who are capable of identifying/executing on big ideas" (aka Tim and his manager)....

and "the projects that businesses/agencies want their employees to do"

Especially when, in hindsight, it's anything but vague.

To be pedantic, he wrote "Vague, but exciting". See http://info.cern.ch/Proposal.html for more info.

Sorry to be a pedant but Maps is not a 20% project, it was an acquisition of a company called Where 2 Technologies. Google Earth was also an acquisition (Keyhole).

GMail however is a 20% project. :)

I'll totally use this when making my next house purchase.

What makes you think that? I'm a little skeptical because Sunroof doesn't cover my address, but I have no problem getting an estimate.

They have a link to sunroof below their estimate. Not sure what happens if you aren't covered by sunroof, but their are a ton of solar estimation tools out there that will give you a rough idea of your potential just based on lat/lon.

Random number generator, I guess. :)

Your electricity is probably a lot cheaper in Washington also and I'm assuming they have local rate information to factor in. Electricity is very expensive in California especially if you use PG&E since they use a tiered rate structure that goes from .12 - .50 cents. And very little of your usage is actually in that 12 cent tier for most people. In states like that I would imagine your solar roof could potentially pay for itself.

Seems off. Seattle gets 1/3rd the sunlight of California.

Isn't the selling point aesthetics? It's very clearly not for everybody, not yet, anyways.

In Colorado it's against the laws for HOAs to ban solar systems, there are some tricks though because they can require certain roofs that you can't really put solar on. Certain tiles and 'presidential' grade roofing which typically cost more than normal shingles as it is. If you live in one of those types of places and want an energy system, then this is what you want. Honestly, I've not heard a lot of complaints about the aesthetics of a conventional solar system, in fact I think some people like it because it's really obvious looking; but if you have the funds and really value the look, then there is an option for you.

It's probably not that different from the market for a Model S vs a Honda Accord. The S is 3+x the cost of the Accord, has fewer fueling stations, can't be serviced nearly as easily, probably is less 'practical' in a handful of other ways but there are still people that buy and drive Model S Tesla cars.

Don't forget the tesla $40,500 includes a roof AND solar. To for a fair comparison you have to include the cost of roof and solar for the next year. Your mentioned other system for $20k that handles 100% of your need doesn't include the roof.

> includes a roof AND solar

...and the cost to remove your current perfectly good roof! I think I smell a bargain!

> Add in a PowerWall and federal tax credit, and I'm still losing $15,400 over 30-years.

Which you probably sink into a second layer of asphalt shingles 15-20 years in (at least up here in upstate NY where there's plenty of rain/snow). Hopefully it's not a tear-off, as that'll cost more.

It's always tearoff. A lay over is saving 2k and shortening the life of brand new roof. Always a bad idea. Unless maybe selling the house soon to a naive buyer.

Wait what, their estimated payback time is 60 YEARS?

This is terrible. They've gotta be just trying to snipe rich early-adopter-for-the-status types.

And whats wrong with that? A bunch of rich early-adopter-for-the-status types will get a Telsa roof, Telsa will get better at developing the technology and the price will come down, meanwhile more people are moving to renewable energy sources to power their homes which is good for the environment. Whats the downside here?

Through the tax credits, we're all paying for someone else's pet project.

Since Tesla's products are expensive, it would be a regressive tax even if we all bought enough solar panels for our needs.

The point of the tax credit is to encourage innovation.

Yes, solar roofs are a bad investment NOW, but they might be a good idea in 20 years, once all these businesses have done this research on them.

If solar roof companies are to get an indirect tax credit then:

1) It should be in exchange for preferred stock/bonds in the company; and/or

2) the tax credit should favour lower prices, not higher

> They've gotta be just trying to snipe rich early-adopter-for-the-status types.

Tesla? The company that brought you the Roadster first, then the Model S and X, and only later the Model 3? I'm shocked. ;)

yeah i know aren't these the exact people that should pay for these kinds of things? I think it's a great model. Can't wait for the model 3 version!

All the rich people that bought the expensive Teslas have used up almost all of the tax credits so the majority of model 3 buyers will be SOL

I'm thinking it's early adopters and new buildings. The Tesla roof on a newly built home may make more sense than a retrofit.

Right, noone is acknowledging SolarCity's pre-existing relationships with homebuilders. http://www.solarcity.com/commercial/homebuilders

It's going to be cheaper (marginally) to build something in at design than after.

In 30 years how many times will you have to redo your roof I think one of Tesla solar roofs points was that many houses in the US have to reshingle. Although I still don't think it might be worth it for you.

Shingle warranties are often 20-30 years.

Lots of houses over 100 years old out there. No reason to think a lot of houses today won't still be standing in 100 years.

roofs != houses

Houses need roofs that last. Why put a new roof on every 15-20 years if you can put one on that lasts the lifetime of the house?

Because unless you go with a metal roof that's a pipe dream.

I live in a house built in the early 1900's with the original slate tile roof still on it and it's in decent shape.

I live in a neighborhood full of slate tile roofs and tile replacement is a fairly routine occurrence. Are you positive that no tiles have been replaced in the past 100+ years?

> Houses need roofs that last. Why put a new roof on every 15-20 years if you can put one on that lasts the lifetime of the house?

Please don't seriously suggest Tesla's solar roof will last the lifetime of a house to people.

The real warranty is 30 years for power and weatherization.

Its misleading marketing speak (at best).

https://www.tesla.com/solarroof#order

> Weatherization means that there will be no water leaks or other weather intrusions during the 30 year warranty period that result from our installation.

It will almost certainly require replacing and/or serious repair long before the "lifetime of the house" is up.

Not in CO. Lots of hail damage, roofs get replaced on average around 10y mark. Also, insurance, older roof == higher premium.

> put that same $40,000 in an index fund that got the stock market's historical 7% average, and waited 30 years, I'd have $308,000.

stock market isn't a very safe investment (that is, risk is much higher than putting in a bank), you can't make a comparison against buying some equipment, which is zero risk. A closer comparison is putting the $40,000 in a bank that guarentees interest (or treasury bonds, which is also pretty much risk free like a bank) - you'd get $99,457, which is $59,457 over your initial investment.

I don't think you understand how index funds work...

There is always risk. Maybe the price of electricity changes, or the laws change, so you can't recover costs? (Look at what happened in Nevada.)

Maybe Tesla goes out of business or drops support for this product? Maybe they aren't as durable as planned? Also, a roof is part of a house which is a very illiquid investment, and there are a lot of other factors affecting the price of a house.

But, presumably people aren't buying this just as an investment. If you're going to put a lot of money into a house anyway, maybe a fancy roof isn't so bad.

Well investing in solar panels isn't either, 30 years is a lot of time for such (currently not very efficient) technology. Price of the electricity could go down significantly as current renewable or nuclear generators get improved over time, or perhaps someone will finally figure out how to do the fusion efficiently?

> you can't make a comparison against buying some equipment, which is zero risk.

Equipment isn't zero risk. It could fail and need replacement, or need dramatic repairs. Or the price of electricity could drop dramatically and your investment won't be saving you as much in electricity bills...

I live on top of a hill where we routinely get 50mph+ winds. Explain to me how roofing tiles carry zero risk.

Wouldnt increasing the tag price of your home by 15k to absorb the loss when you sell compensate? Also, in 30 years, 15k will be worth less than 15k today.

That only works if it is worth $15k to the buyer.

Plus you have to be in the market to sell in the next few years (plus accounting for depreciation and the like... roofs only last so long).