In the format provided by the authors and unedited. COMMENTARY: Emerging clean energy technology investment trends A. Bumpus and S. Comello DOI: 10.1038/NCLIMATE3306 NATURE CLIMATE CHANGE www.nature.com/natureclimatechange 1
Methodology 1. Dataset Creation We used the CB Insights 1 platform to create a dataset of investors, firms and transactions. The focus was clean energy firms and early-stage capital providers, and the transactions between them. 2 The first step was to limit the firms in the dataset to include those that are focused on products and services related to the electricity grid. Instead of a standardized industry classification system such as North American Industry Classification System (NAICS), CB Insights provides its own taxonomy of firms by sector, industry and sub-industry. As such, Supplementary Table 1 provides the sector and industry search criteria used to return the universe of firms examined. Sectors and sub-sectors were chosen after examination of all sectors within the CB Insights and an audit of the kinds of firms/transactions contained within each. This audit employed expert judgment to determine the final set as shown in Supplementary Table 1. 3 Supplementary Table 1: Clean Energy Technology Firm Sectors and Industries Included in Dataset Sector Industry (sub-sector) Energy & Utilities Electric Energy & Utilities Energy Efficiency Energy Storage Energy Trading & Marketing Oil & Gas Production & Exploration Oil & Gas Refining & Distribution Renewables Cleantech Green/Environmental Software Internet Software & Services Mobile Software & Services The second step was to focus only on those transactions between firms and any investor type providing early-stage funding. As mentioned in the Commentary, it is crucial to note that the provision of such funding is not restricted to venture capital firms. As such, and to be clear, Supplementary Table 2 lists the types early-stage capital providers included in our study: Supplementary Table 2: List of Early-Stage Capital Provider Types (Category and Subcategory) Category Subcategory Venture Capital Angel Investor (Group) Angel Investor (Individual) 1 CB Insights is a database subscription focused on venture capital and angel investment transactions and information (https://www.cbinsights.com/). 2 It is noted and acknowledged that CB Insights data service or any other similar investment/transaction database such as ThomsonOne is unable to capture the entire universe of investments, in particular angel investments (as these can go unreported given their size and relative informality). Nonetheless, CB Insights provides a reasonable breadth of transactions for the purposes of this study. 3 It is possible that some sub-sectors that were not chosen could have contained relevant firms and transactions, especially for Wave 2 technologies, which are diverse in industry definition and application. An example of this is the sub-sector titled Internet of things. If this were to be true, then the findings related to Wave 2 may be underrepresented. NATURE CLIMATE CHANGE www.nature.com/natureclimatechange 2
Corporate Incubator/Accelerator Long-term Investor Venture Capital Funds Corporate Venture Capital Funds Private Equity Subsidiary NA Pension Funds Mutual Funds Sovereign Wealth Funds Family Office Funds Non-Profit Foundation Funds Endowments Further, Supplementary Table 3 lists the kinds of early-stage capital providers included in the dataset. Supplementary Table 3: List of Innovation Capital Types Type Description Angel/Pre-Seed Investments made by individuals (or groups of individuals) in personal capacities Seed Initial investments made by professional investors on behalf of themselves or clients Series A Start-up capital provided by professional investors on behalf of themselves or clients Series B, C, D Early stage capital provided by professional investors on behalf of themselves or clients Series E, F, G Expansion capital provided by professional investors on behalf of themselves or clients Unattributed VC Venture capital funding where the specific type was unavailable in the data Applying the sector/industry filter to narrow the field of firms, and the capital type filter to specify admissible early-stage capital transactions resulted in a dataset of 2,129 private firms and 6,769 individual transactions, bounded by the dates January 1, 2000 to December 31, 2015 inclusive. The year 2000 was chosen as the start date because this was generally the timeframe where early stage capital providers increased their interest in clean energy firms. 4 However, given that the number of transactions related to the sectors used in this study was negligible before 2003, we chose to begin our analysis in 2003. All data was accessed February 7 th, 2016. 2. Categorization of Wave 1 and Wave 2 Technologies Given the focus of the work was the electricity sector; we filtered out those firms whose industry in oil & gas, transportation or any other line of business (e.g. agri-business, water filtration, etc.). Doing so resulted in a set of 1,129 firms focused on clean energy, electricity sector. The first step was to create the definition of Wave 1 and Wave 2 technologies. Wave 1 consists of technologies that are characterized mainly by the hardware components required for renewable energy generation systems. Such components require immense research, development, deployment and manufacturing capacities over long periods of time to diffuse through the ecosystem. Moreover, such components (though not necessarily their assemblies) become commodities within a supply chain, where the key 4 National Venture Capital Association 2016 Yearbook (http://nvca.org/pressreleases/2016-nvcayearbook-captures-busy-year-for-venture-capital-activity/) NATURE CLIMATE CHANGE www.nature.com/natureclimatechange 3
performance metric is cost minimization, and the opportunities for economic profits are few. By comparison, Wave 2 technologies are characterized by their focus on software applications and demand-side solutions. Admittedly, this is a more diffuse set than that of Wave 1, however the key characteristic of Wave 2 technologies is that of optimization, along demand and cost dimensions. Given that demand-side technologies must meet a wide set of use-cases that engage dimensions that include convenience, timeliness, and quality in addition to cost, there are greater possibilities for economic profits to successful solution providers. Supplementary Table 4: List of Innovation Capital Types Attribute WAVE 1 WAVE 2 Dominant Characteristic Examples Technologies Electricity generation and storage hardware Solar panels Wind turbine blades Biomass digesters Battery cells Inverters Demand, cost optimization, efficiency. Sensors, actuators, controllers and electronics Software (communication, analytics, management, automation) Responsive end-use elements such as dimmable windows or shading, intelligent lighting and smart thermostats The next step was to categorize firms as either Wave 1 or Wave 2, based on the products and/or services being provided by reviewing and assessing each firm s description as provided by CB Insights (usually about one paragraph in length). Following the methodology prescribed in Charmaz, 2014 1 Categorization was performed in a blind-parallel method, where three researchers (the coauthors and one unaffiliated researcher) were provided instructions and independently completed the task. This was done to mitigate selection bias; initial agreement between the 3 sets was 96%. Consensus on the remaining 4% was reached through discussion.. 3. Survey of Clean Energy Incubators in United States Using the US Department of Energy s Incubatenergy 5 network as of December 2015, we created a list of clean energy incubators (see Supplementary Table 5). Through email between February 2015 and March 2015, we administered an email survey to the executive director of each incubator, requesting (i) founding date of incubator, (ii) longitudinal history of all portfolio firms (date enter, date exit, reason(s) for exit) and (iii) firm descriptions. Response rate was 100%. The survey revealed 369 unique firms within the 19 incubators, of which 207 fit the definition of either Wave 1 or Wave 2 technologies (based on the qualitative methodology applied previously). Supplementary Table 5: List of Clean Energy Incubators Surveyed # Name Location Founded 1 New York City Accelerator for Clean and Renewable Economy New York, NY 2009 5 https://incubatenergy.org/ NATURE CLIMATE CHANGE www.nature.com/natureclimatechange 4
(ACRE) 2 Austin Technology Incubator (ATi) Austin, TX 1991 3 Clean Energy Trust Chicago, IL 2010 4 Coalition Energy Chicago, IL 2014 5 CLT Joules Charlotte, NC 2013 6 Cyclotron Road Berkeley, CA 2014 7 Energy Excelerator (EEx) Honolulu, HI 2013 8 Greentown Labs Somerville, MA 2013 9 I-Corps Energy and Transportation Detroit, MI 2002 10 Innovation Incubator (IN 2 ) Golden, CO 2014 11 Innosphere Fort Collins, CO 1998 12 Los Angeles Cleantech Incubator (LACI) Los Angeles, CA 2011 13 Next Energy Detroit, MI 2002 14 Oregon Built Environment and Sustainable Technology Center Portland, OR 2008 (Oregon BEST) 15 Powerhouse Oakland, CA 2013 16 Prospect Silicon Valley Labs (ProspectSV) San Jose, CA 2011 17 Smart Grid Cluster Chicago, IL 2012 18 Silicon Valley Cleantech Incubator (SVCI) Redwood City, CA 2011 19 Mid-West Energy Research Consortium (WERCBench Labs) Milwaukee, WI 2016 4. References 1. Kathy Charmaz. Constructing Grounded Theory. (Sage Publishing, 2014). NATURE CLIMATE CHANGE www.nature.com/natureclimatechange 5