The levelized cost of solar photovoltaic (PV) electricity is lower than conventional sources in a growing number of markets in the U.S. Yet many projects cannot get off the ground because of limited access to funding. Given the stable, long-term cash-flow profiles of existing power purchase agreements (PPA), solar developers are pursuing the use of solar asset-backed securities (ABS) to alleviate this problem. SolarCity has already issued the first solar ABS in the U.S., a $54.4-million deal that closed last November. More issuance could spawn a large and efficient market, with strong returns for investors and cheaper, more available funding for developers.
But many developers are too small to go it alone. We found that the key to a viable securitization market for solar is aggregating assets and that is possible only through standardization, which would facilitate data collection and help make up for the sector’s short track record. Aggregation would facilitate securitization, which significantly cuts financing costs, helps the smaller firms with the toughest financing challenges, and extends the asset class’s geographic reach.
Originating and Pooling Assets
Actions taken by originators during the contracting process impact the rate of early contract termination, which, in turn, greatly influences the cash flow profiles of potential asset pools. When firms are responsible for adequate credit enhancement, high quality asset pools, with low rates of early contract termination, generate more financial capital. But the current inconsistency of contracts and contracting procedures among firms makes it difficult to either aggregate assets or systematically improve contract quality. The lack of standardization carries a steep cost.
Our parameterized securitization model (2) suggests that by aggregating assets in a solar ABS firms can reduce financing cost by 5%-13%. These cost savings cannot be widely realized without standardization. Despite rapid growth in the third-party ownership market, no single firm now owns sufficient power purchase agreements (PPAs) to issue an asset-backed of market-preferred size. Inter-firm pooling would bring in smaller developers with currently poor access to funding as well as extend the geographic footprint of an asset pool.
Standardization would facilitate the data collection that is vital given the short history and evolving use of third-party ownership. Access to open data improves the potential for statistical analysis by increasing observation points and expanding diversity within the sample. Data analysis could provide feedback that would improve credit screening and contracting procedures, reducing early contract termination rates. Figure 1, below, illustrates the impact of this change. Cash flows from a pool of similar contracts quickly diverge as asset quality declines. Figure 2 depicts the impact of these disparate cash flow profiles on the value of an issued security.
Reducing default rates increases the value of the ABS regardless of asset quality. Our model-based estimates suggest that a 1% decrease in early contract terminations can increase the amount of capital raised through securitization by 4%-8%. The benefit of improving asset quality is greatest when both investor returns and contract termination rates are low. These potential gains give firms a sharp incentive to adopt standardization.
Asset Evaluation by Rating Agency
For a securitization to hit investment grade, a credit rating agency assigns the required credit enhancement based on the risk of the collateral. Usually an agency uses historical data to forecast asset default rates, but this data is not yet available for solar PPAs. Available proxy data does not fully capture industry specific risks, which reduces the accuracy of a rater’s risk assessment. Uncertain risk raises the level of required credit enhancement, raising the cost of financing for the issuer.
When rater inaccuracy is expressed as excess credit enhancement it has the same impact on security prices as the rate of contract default. Using overcollateralization to provide credit enhancement means the par value of issued securities is lower than the value of the collateral - the more excess credit enhancement required, the less capital can be raised from the asset pool.
When overcollateralization is used to provide credit enhancement, we found that an increase in excess required credit enhancement from nothing to 1% is estimated to add 0.33%-0.51% to the cost of financing. A similar change from 7%-8% increases financing costs by 0.97%-1.06%. In each case, rater inaccuracy is also more costly when the quality of the asset pool is low.
Data collection can improve rater accuracy. However, with contracts lasting 20 years, collecting data will be a long process. Another way to reduce the required enhancement and therefore cut costs for the issuer would be for a government entity to guarantee deals by talking the first-loss hit. This strategy was also used by the Department of Energy in 2005 to help bring innovative, clean energy technologies to market through its 1703 Loan Guarantee Program.
Purchase of Solar ABS
The return required by investors is the primary determinant of financing costs. This is important in the nascent solar ABS market where factors like illiquid assets tend to increase the required investor returns. Our paper details the factor combinations that lower financing costs (2). Our estimates show that raising investor returns by 1% adds 1.01%-1.16% to the cost of financing.
The larger increases occur when excess credit enhancement magnifies the impact of higher returns. Achieving sufficient scale to create liquidity would reduce required returns, but this is difficult given the prevalent use of tax equity financing. Under this structure the financier must be recognized as the owner of the asset during the recapture period in order to claim tax benefits. It is possible to utilize securitization before the end of this recapture period by attaining loans backed by developers’ equity positions. Even with this option, liquidity cannot be added quickly.
Higher rates of return also reduce the amount of capital that can be raised from a particular asset pool. Based on our model, a 1% increase in investor return reduces the amount of capital raised by the securitization by 3.9%-8.1%. Increases in investor returns have a larger effect on asset prices when the quality of the asset pool is high.
Despite these obstacles, by pooling residential PPAs solar developers would be able to generate low-risk assets that provide competitive returns to investors, while at the same time reducing developers’ financing costs.
This is a summary of “Securitization of residential solar photovoltaic assets: Costs, risks and uncertainty”, Energy Policy, 67, pp.488498 (2014). Theresa Alafita is a visiting scholar, and Joshua Pearce an associate professor, at Michigan Technological University.