This article is part of our Real-Money Stock Pick series.

SolarCity (SCTY.DL) is turning out to be quite an exciting stock. I first bought it for my Messed-Up Expectations portfolio (the real-money portfolio I run on behalf of The Motley Fool) early last September. It ran up as much as 196% before falling back down within the last month or so.

With the 2013 financial statements finally released, I've updated my model and now have a sense of whether today's price, down about 25% from its high, is a decent price to add to my position or not. I think it is.

If you recall, SolarCity installs solar panels on the top of homes and businesses and sells the electricity to the customer through a 20-year lease at a rate that is below current and (more important) expected future utility rates. It's been financing operations by bundling the tax benefits from this into securities and selling those to big banks. It's transitioning to bundling the future lease cash flows into securities and selling those instead, similar to what power utilities do.

Because of the weird way the financial statements reflect the company's business, I have to use its metric of retained value (RV), which is the present value of the future lease payments minus certain costs. This would essentially be the value of the business if it just halted new installations and ran out existing contracts. Currently, the market is pricing SolarCity at about 5.4 times the last reported RV of $1.05 billion, which means a fair amount of growth is priced into today's market cap.

Estimating future retained value
Because RV is a discounted cash flow number, but I don't have the annual cash flows SolarCity is using to calculate it, I have to back into calculating RV. My model tries to project what RV will be five years from now by growing MW installed each year (with a declining growth rate) along with growing estimated nominal payments remaining from current and new contracts. From that, comparing how RV has tracked with MW installed over the past three years, and looking at how the company has divided RV between "under energy contract" and "renewal" (it includes cash flows from customers renewing their leases for an additional 10 years, assuming all customers will do so), I can project future RV. Comparing that to today's market cap gives me a sense of where the company stands in valuation.

There are three big levers to this model: The rate of annual growth for MW installed (higher growth leads to higher RV); the percentage of customers renewing their leases for an additional 10 years (the company assumes 100%, I assume 33%); and the market multiple assigned to the future RV (a lower multiple implies less company growth priced in by the market).

The company currently estimates to install 500 MW this year (at the midpoint of 2014 guidance), which is 79% growth over 2013. If I drop that growth rate by 33% per year through 2018, it will install just shy of 1,500 MW in that last year. This implies $5.28 billion in RV in 2018 (with a 33% renewal assumption). If I give that a 1.1 times multiple -- which means the company would pretty much enter "run-off" mode (no more growth, run out current contracts) -- that implies a market cap of $5.8 billion, pretty close to today's $5.73 billion.

Three controlling levers
I mentioned three levers affecting projected RV in my model. Holding all else equal for each of the following scenarios, increasing customer renewal rate to 50% from 33% (still well below the company's 100% assumption), 2018's RV climbs to $6.2 billion. If the growth in MW installed declines by 11% per year instead of 33% (ending at 50% 2018 year-over-year growth instead of 16%), 2018 RV leaps up to $8.89 billion. And if I give it a two times multiple instead of 1.1 times, RV goes to $10.56 billion, almost twice today's market cap.

Of the three, the model is most sensitive to the market multiple to RV. When I first purchased shares, it was at a multiple of 3.5 times. Today, thanks to a higher reported RV and a higher market multiple to that, the share price is much higher. If my model is correct in every respect except for underestimating what the market is willing to price the worth of RV, a purchase today could lead to very satisfying returns.

Of course, my model won't be correct. Hopefully, I've made it conservative enough, though, by cutting the annual MW installed growth rates enough, cutting the assumed lease extension rate enough, and lowering the market multiple to RV enough that most surprises will be to the upside.

I'll be buying more shares of SolarCity for the MUE portfolio as soon as Fool trading rules allow. Feel free to come to the portfolio's discussion board for more insight into what I and other Fools are thinking about this exciting company.