Source: SunPower.

Valuing a solar company is a tough job these days. Companies are taking different strategies and holding assets on their balance sheets, and who knows what the solar industry is going to look like 5 or 10 years from now?

As a result, companies like SolarCity (SCTY.DL) and SunEdison (SUNEQ) have started using a measure of value called retained value to show their value to investors. Retained value measures the present value of the future cash flow of projects, which sounds great, but it may actually be a very flawed way to look at these companies.

An average rooftop system like this generates about $10,000 in retained value according to SolarCity. Source: SolarCity.

What's in retained value?
Retained value is basically a present value of all future cash flows a company expects from a solar system. This includes inflows from contract payments minus operating costs and payments to financing partners -- all discounted to today's dollars (usually 6%). In some cases, they even assume a lease will be renewed, adding more value to the system. In short, retained value is the value an investor could expect to capture if the rest of the business was shut down, and they just cash flowed the solar systems for 30 years.

These aren't insignificant figures, either. SolarCity has reported $2.18 billion in retained value as of the third quarter of 2014, and SunEdison says it has retained $918 million in value in the first three quarters of this year.

The figures are impressive, and if you take them at face value, it may seem both SolarCity and SunEdison stock are grossly undervalued. But investors should understand some of the key assumptions built into retained value before taking it at face value.

Retained value assumptions are generous
The biggest problem with retained value is that it assumes nearly everything will go in favor of the solar company for the next 30 years. Here's what I mean.

  • Retained value assumes every customer will pay its contractual obligations over the next 20 years (30 years in some cases). Importantly, this assumes sales of homes with solar leases will see the new owner picking up the old lease, even though they have an option to break the lease (and buy a cheaper system). I think this is questionable at best, especially as systems get older, solar system costs fall, and home buyers become more educated about buying solar. After all, why would I want to take on a 10-year obligation for a 10-year-old solar system when I buy a house?
     
  • Renewal retained value assumes customers will renew their leases in 20 years at a 90% rate of the end of their contract. The assumption is that 90% of customers will renew at their year 20 rate, or 100% of customers will renew at 90% of their year 20 rate. There's no evidence this will be the case, but SolarCity in particular is banking on this. $640 million of SolarCity's $2.18 billion in retained value is based on the renewal assumption. In my opinion, investors should disregard the renewal figure altogether, because there's no evidence customers will renew leases for 20-year-old panels in 20 years, and we have no idea what the industry or solar market will look like then.
     
  • At 6%, the discount rate solar companies are using is generous at best. If interest rates rise, this assumption will be a big risk. To its credit, SolarCity gave a sensitivity analysis for discount rates. If you discount at 8% instead of 6%, retained value under contract falls 15% from $1.54 billion to $1.31 billion.
     
  • The biggest thing investors often don't understand is that retained value doesn't include the costs associated with signing up customers. These are operating costs for sales, administration, executives, etc. These are very real costs to the business that are paid upfront.

These are big assumptions that may not end up being true as the solar industry plays out over the next 20 or 30 years. I haven't even covered high power purchase escalators built into many leases that may make buying power from SolarCity more expensive than the grid in the future. That may make for unhappy customers and an even lower chance of renewing 20-year-old leases.

Utility scale projects like this should perform much more consistently than residential solar systems. Source: SunPower.

Not all retained value is created equal
Another factor to consider when looking at solar companies is who their customers are. SolarCity is primarily selling to residential buyers, while companies like SunEdison or SunPower (SPWR), which don't report retained value, are primarily selling to businesses and utilities. I'd be much more inclined to think a utility will pay its power purchase agreement in 20 years than a homeowner, who may have changed multiple times before the solar system reaches maturity.

This is what investors call credit and/or counterparty risk. Solar companies are counting on someone else to live up to a long-term contractual obligation, but there's risk they won't.

Don't believe everything you see in solar
The bottom line here is that investors need to know what they're signing up for with solar companies. A figure like retained value may be a way to measure value creation, but if you don't know what assumptions are built into the calculation, it's a worthless figure. Given the fact that concepts like residential leases are still relatively new, we don't know how they'll perform long term, so making assumptions 30 years into the future is risky. 

At the very least, these flaws with retained value are something we should all be aware of when considering buying solar stocks.