Determinants of Credit Allocation for Photovoltaic Projects: Research Outline and Preliminary Findings from Conjoint Experiments with German Financing Experts

Abstract

An increasing concern for photovoltaic (PV) projects is a shortage of capital in the medium term. PV project financing is an urgent topic due to a fundamental lack of knowledge about how debt capital providers evaluate this class of infrastructure projects. Moreover, after recent turbulences on the financial markets it is unclear how the availability of capital for renewable energy projects will develop. In this context financing models are decisive especially for medium- and large-scale PV projects, since large shares of debt capital are needed: Practitioners assume debt ratios of up to 80% or even 90%. As there is only limited academic research on project characteristics and their impact on credit allocations, the focus is on lenders’ preferences for different project types. The explorative research approach therefore addresses the following question: Which types of PV projects do lenders prefer to finance? We try to answer this question by conducting an Adaptive Choice Based Conjoint experiment (ACBC) with German experts in renewable energy project financing (banks, savings banks, consultants, project developers). To get robust and relevant results we clearly define our research scope: medium- and large scale ground-mounted installations being subject to the German Renewable Energy Sources Act (EEG) as of 2009. Initially, we supposed that from a lender’s perspective two attributes might be of utmost importance: First, Debt Service Cover Ratio (DSCR), indicating the security of loan repayment. Second, a hardly discerned but assumingly important aspect is the project initiator who is the central stakeholder during project development. This attribute is conceptualized by referring to possible project initiators’ business models. Examples are vertically integrated PV manufacturers, financial investors, service providers, and regional or multinational utilities. To provide for an as authentic as possible conjoint experiment, six attributes with 22 attribute levels in total were developed through literature studies and expert interviews. The six attributes are: DSCR, system capacity, brand/low cost technology, project initiator, maintenance concept, and equity ratio. Our survey, taking place from January to March 2010, is still online as this working paper is being written; thus, results are preliminary. We report from a sample size of 1.240 choice tasks conducted by 30 interviewees; 40 to 50 experts are expected to participate in the end. So far, the empirical data do not support our assumptions regarding the outstanding importance of DSCR and project initiator. Instead, our preliminary results are suggestive of a “brand bias” in lenders’ decisions.