Here in the UK battery backed PV on residentials isn't really a thing (it made no sense to store it in a battery when you can sell it back to the grid at a large tariff, however the government just killed the feed-in tariff (87% reduction) so the future of PV looks pretty bad).

Kinda pissed about that tbh as we spent nearly a year working on a product for installers and then the government just kicked the legs out of the entire industry.

Their panel costs are say $1, the other costs (construction team, scaffolding, permitting, inverter etc) are say $4. These are rough numbers, per watt.

You actually lease it from them, so you don't own the installation. You either 1) rent the panels, or 2) buy the electricity from the panels. Either way they try to offer you a rate of your current utility - 5% or whatever. Basically it becomes slightly cheaper most of the time, but not by much, locks you into a 20 year contract, and you get the intangible emotional/moral benefit of 'going green'. That's their value offering. This also means you have no installation risk, if the panels break it's their problem, not yours, so things like guarantees on equipment or who has to pay for repairs isn't as big of an issue.

Alright so as they own the stuff, they get all the subsidies. For one there's a 30% tax credit. Say you install $300 of solar, you can reduce your payable taxes by $100, so the effective cost is $200 by now. Then either the $300, or on the remainder ($200), can't recall, you can depreciate in a few years. Say it's the full amount, and your effective tax rate is 30%, you can now reduce your taxable income by $300, of which you save 30% or another $100. Then you can sell the SRECs, after all you generate green energy which is certifiable in solar credits and can be sold to polluting companies. Prices vary but you could get say $50 on a $300 installation for example. So now you've reduced your effective costs to $1 out of every $6 you spent.

Alright so how much does $5 per watt generate in electricity income? Well most of it goes to the user who pays you a market rate - 5%. So ignore the 5% for a moment, the US average is about 13c per kwh. A typical capacity factor for solar is about 15%, so over a 365 day period you can generate 1314 watt per installed watt of capacity, or 1.3 kwh, valued at a little over 17 cents, take off the 5% and you're left with about 16c rounded down, let's say. So for $5, you get a 16c return, but the $5 after all the above tax credits, depreciation and solar credits might be closer to say $1. So SolarCity could make about 16c on the dollar every year, for a 20y contract. That's roughly a 4% return or so.

So now you have a 20 year cashflow contract with a customer, and you can go ahead and sell that on the market as a security with a 4% ROI. This way they can immediately cash out and be cashflow positive and reinvest the money into new installations. This allows them to scale much quicker than growing organically, and this scale gives them economies of scale, and the ability to buy up other companies, even their own manufacturing (as this article is really about) and beat the competition.

In short, there's not a lot of solar technology going on, i.e. totally different from SpaceX or Tesla where there's real engineering innovation. It's mostly clever business models, securitisation of decentralised utility contracts (solar on rooftops) at locked in prices, heavy reliance on subsidies (a reliance which is decreasing every year as it's forced to do).

Anyway above numbers are pretty rough but it gives a sense of the economics. It's a really tricky industry to get a hold of as people throw around (inc. myself sometimes) tons of numbers interchangeably. Like citing capacity in DC/AC makes a difference but is rarely specified. Specifying costs per watt as module costs, installation costs, total costs etc, is often left out. Ownership changes the economics completely for the different parties etc.

Here you can see a chart for the total costs for some players in the space. Ignore the $3 one, it's their chart :p But you can get a sense of the rough numbers: http://www.freecleansolar.com/v/vspfiles/assets/images/lowes...

The SREC market differs wildly from state to state, hard to put a value on it.

Federal subsidies of 30% are phased by 2016's year-end. The depreciation of a solar installation might be a business-only thing and not applicable to consumer-owned solar (which SC doesn't do anyway), I'm not fully familiar with US fiscal policy as a European. Anyway that's the best I can do right now, hope that helps :)