What matters for private investment in renewable energy?

2.1. Public expenditure

Most of the empirical literature studying government R&D directed to renewable energy technologies focuses on the impact on innovations and technology development (Plank & Doblinger, Citation2018) rather than on private investments. R&D of less mature renewable energy technologies is often financed from public sources. Investments at earlier stages of technology development carry greater risks and are thus less attractive for private investors. By investing in R&D of renewable energy, governments aim to boost private investments at the technology R&D stage or at later stages of technology maturity, such as manufacturing scaling-up, rollout, and acquisition (Figure 3). The literature on public investment in renewable energy, however, lacks evidence on the crowding-in and crowding-out effects on private investment in renewable energy due to government investment. Moreover, Deleidi et al. (Citation2020) show that the terms ‘crowding in’ and ‘crowding out’ may be inappropriate in sectoral studies such as renewable energy. These authors show that public direct investments in renewable energy are effective in mobilizing private investments. Our paper delves deeper into this issue by examining different sources of private investment financing, as well as a larger sample size and more timely data. Based on this strand of the literature, our first hypothesis to be tested is:

2.2. Renewable energy policies

Polzin et al. (Citation2019) provides a review of empirical literature measuring the impact of policies on investments in renewable energy. They demonstrate that policies effective in mobilizing private finance are those which address risk and return simultaneously. Thus policies need to be designed that match expected returns with acceptable levels of risk (TheCityUK and Imperial College Business School, Citation2019a, Citation2019b). Since expected returns with acceptable levels of risk differ across sources of financing, it follows that the impact of policies may vary across sources of financing.

In the case of Asia, Chang et al. (Citation2016) conducted a quantitative assessment of renewable energy policies aimed at promoting investment in the 16 East Asia Summit countries. They did so by using an index across five key areas that can be informative for prospective investors, namely: (i) market (whether the policies help to create a renewable energy market); (ii) uncertainty (whether policies reduce uncertainty); (iii) profitability (whether policies promote an environment for profitability); (iv) technology (whether policies support technology adoption); and (v) financial resources (whether policies promote funding availability). Overall, they find heterogeneity in the results, with PRC, India, Republic of Korea, and Japan as among the highest-ranking Asian countries in terms of the impact of their renewable energy policies. Our paper is related to the work of Chang et al. (Citation2016). The aim of our paper is to quantitatively estimate the effect of renewable energy policies on private investment for a sample of 13 economies including a sub-set of Asia. Other work on examining renewable energy policy effectiveness in Asia includes Bakhtyar et al. (Citation2013), which focuses on FITs for the cases of Indonesia and the Philippines. Earlier work by Chen et al. (Citation2014) noted that less conservative efforts in development of the renewable energy sector is needed in the economies of East Asia. Related work on the effectiveness of renewable energy policies for various Asian economies have also been carried out by Mouraviev (Citation2021) in Kazakhstan; Schmid (Citation2012) in India; Shen and Luo (Citation2015) and Wang et al. (Citation2010) in PRC; Toan et al. (Citation2011) in Vietnam; and Tongsopit and Greacen (Citation2013) in Thailand. Also related to this, but focusing on North America and Europe, Bürer and Wüstenhagen (Citation2009) find that FITs are the most important driver of private investment in renewable energy. Based on a survey of 60 investor professionals, their results are more pronounced for Europe, where there is higher exposure to clean energy.

Two of the most popular economic policy instruments are FITs and tax incentives. FITs have played a major role in increasing renewable energy capacity in Europe (Davies & Diaz-Rainey, Citation2011), particularly in Germany, Spain, and France, as well as in Asia, notably in Japan and the Republic of Korea. FITs are pre-determined tariffs that a government can commit to for purchasing renewable energy by signing fixed long-term contracts with renewable energy suppliers. FITs are designed to attract private investment in renewable energy sources by providing a guaranteed compensation above the electricity market price.

By providing FITs, governments can promote private investment in renewable energy in two ways. First, by setting tariffs, governments can create fixed rather than uncertain revenue flows, thus reducing the risk of investing in renewable energy. Second, by setting a tariff above the market price, governments can increase the profitability of renewable energy projects. Tax incentives are usually provided as tax credits or tax relief on production, investments, property, sale or import. Tax incentives improve return on investments (Polzin et al., Citation2019). These findings in the literature lead to our second and third hypotheses to be tested in the empirical analysis:

2.3. Technology costs and energy prices

The literature in this field also indicates that private investment in renewable energy is affected by technology costs and energy prices. With lower initial technology costs, the propensity for investment in renewable energy is higher. As noted by Polzin et al. (Citation2019), lower technology costs can help to underpin a more stable investment environment, reducing financial risks for investors. Moreover, higher energy prices of electricity and fossil fuels are likely to be associated with higher investment in renewable energy due to greater price competitiveness of renewable energy and closer to achieving a grid parity (Eyraud et al., Citation2013).

2.4. Source of financing

The sources of financing which renewable energy projects attract depend on the stage of their technological development (Figure 3). Asset finance is the major source of financing for renewable energy, providing funding for renewable energy generation projects, including internal company balance sheets, debt finance, and equity finance (Frankfurt School-UNEP Centre/BNEF, Citation2020). Steffen (Citation2018) investigated financing sources for power plants and show that asset (or project) finance is more important for renewable energy than for fossil fuel-based power plants when investment risks are low. Krupa et al. (Citation2019) also focus on asset finance by using the case of Gulf Cooperation Council. Unlike the existing empirical literature, which focuses on the total investment in renewable energy (Eyraud et al., Citation2013), total private investment in renewable energy (Azhgaliyeva et al., Citation2019; Deleidi et al., Citation2020; Polzin et al., Citation2019), or asset finance (Krupa et al., Citation2019; Steffen, Citation2018), our paper distinguishes between non-overlapping funding sources of private investment (i.e. also including public markets, corporate R&D, venture capital and private equity) (Frankfurt School-UNEP Centre/BNEF, Citation2020). While the above-mentioned empirical studies measured the effect of energy policies on private investment in renewable energy, to the best of our knowledge, our paper is the first empirical study to measure the effect of energy policies on private investment in renewable energy across sources of private financing. The following issues justify our focus on different funding sources of private investment. Sources of financing vary not only across years (Figure 4) and renewable energy sources (i.e. geothermal, small hydro, solar, marine (wave and tidal), and wind), but also across countries (Figure 5). For example, hydro, solar, and wind power are more mature technologies, relying less on corporate R&D, than marine and geothermal power (Figure 6).

The data indicate that sources of financing vary across countries even for the same renewable energy technology. For example, Japan has the largest share of corporate R&D, the US has the largest share of venture capital and private equity, Canada has the largest share of public markets, and the PRC has the largest share of asset finance. Differences across countries can also be explained by country characteristics, including energy policies. While prevailing studies on renewable energy investigate a diverse set of issues and subjects range from unlocking renewable energy investment to the effects of environmental policies on innovation, studies that look at the determinants of the funding sources of private investment in renewable energy projects are scarce, with technology costs also having an important role to play (Polzin et al., Citation2019).

In addition, drawing on the macrofinancial literature on the drivers of private investment, our paper controls for factors relating to different levels of economic, financial, and institutional development across countries. Importantly, we control for the level of economic development and rate of economic growth, as measured by GDP per capita and GDP growth, respectively, which have shown to be strongly correlated with the level and rate of technological advancement (e.g. Faberberg, Citation1987; Justman & Teubal, Citation1991).

This paper also contributes to the broader literature on the drivers of private investment. Typical macroeconomic drivers of private investment would include GDP growth, debt/GDP, and inflation. For example, higher GDP growth has been shown to be positively related to the level of private equity investment (Gompers & Lerner, Citation1998). Similarly, Jeng and Wells (Citation2000) make the point that output growth is positively and significantly related to the demand for venture capital and private equity. This also holds more broadly across other funding sources (e.g. OECD, Citation2016). Other studies stress the role of exchange rate uncertainty and risk aversion as important factors underpinning private investment (e.g. Servén, Citation2006). Beaudry et al. (Citation2001) proxy macroeconomic uncertainty using inflation, citing a negative effect on investment. Indeed, related to this, it may be expected that real interest rates have a significant negative effect on venture capital and private equity investment, given that high real interest rates reduce the propensity to invest in risky assets. This point is reinforced by Gompers and Lerner (Citation1998), who highlight that high-risk investment is negatively associated with interest rates.

In relation to financial development, Black and Gilson (Citation1998) stress the importance of a well-developed and liquid stock market for the overall level of venture capital and private equity investment (see also Cherif & Gazdar, Citation2011; Clarysse et al., Citation2009; Kelly, Citation2012). The price of sovereign risk and sovereign credit rating have also been cited in some studies as important drivers of private investment (e.g. Chen et al., Citation2013). On institutional development, a range of studies – particularly those that focus on emerging markets and developing economies – find a strong role to be played by this factor in attracting private investment (e.g. Bernoth & Colavecchio, Citation2014). The added value in our approach will be to include these country-specific macrofinancial factors in conjunction with determinants that are more energy-specific in driving private investment.

3.1 Data

Our paper comprises data for 13 major economies – Australia, Brazil, Canada, PRC, France, Germany, India, Italy, Japan, Republic of Korea, Spain, the United Kingdom, and the United States – across five renewable energy sources (geothermal, small hydro, solar, marine (wave and tidal), and wind; with project capacity in Table 1), and across four private sources of financing (venture capital and private equity, public markets, asset finance, and corporate R&D). Annual data over the period 2008–2018 is used, with explanatory variables lagged by one period in order to mitigate against endogeneity. The total number of data points in the model is 650 (13 countries × 5 technologies × 10 years). Further details on variable definitions and a correlation matrix are provided in Appendix A (Tables A1 and A2).

Table 2 provides details of descriptive summary statistics for variables used, while Table 3 describes the categorization of these variables across multi-levels over time.

3.2 Methodology

Our analysis comprises four types of private investment (asset finance, corporate R&D, public markets, and venture capital and private equity) in five types of renewable technology (small hydro, geothermal, marine (wave and tidal), solar, and wind) across 13 economies for a period of 11 years (2008–2018). Since the different sources of financing, to a certain extent, may depend on the stage of technological development in the renewable energy sector, we use linear regressions with multi-level fixed effects (country, sector, and time fixed effects) to estimate the determinants of private investment in renewable energy (Correia, Citation2019). Technological advancement and the different stages of technological development across our sample is proxied in this set-up using measures for economic development and growth. Our baseline specification is given in Equation (1), where a linear model with country, sector, and time fixed effects is estimated. (1) $Y_{i,j,t}=\text{β}{X}_{i,j,t-1}+\gamma Z_{i,t-1}+\chi F_{i,t-1}+\mu VI{X}_{t-1}+{\alpha }_i+{\delta }_j+{\lambda }_t+{ϵ}_{i,j,t},$(1) where i=1 to 13 (countries), j=1 to 5 (types of renewable energy) and t=2008 to 2018 (years).

In Equation (1), Y_i,j,t measures the log of private investment in renewable energy technology j in country i at time t. The domestic energy specification factors are captured in the vector X with five variables: log of government R&D, FIT, tax, technology cost, and energy price index. The estimated β coefficients pertaining to each of these components are tested for significance as described earlier in H1 to H5. The domestic macroeconomic and institutional factors are represented in the vector Z with six variables: real GDP growth, GDP per capita, public debt/GDP, fixed capital formation/GDP, inflation, and regulatory quality. Vector F comprises the domestic financial factors: stock market capitalization/GDP, and the long-term interest rate. The VIX controls for global financial market volatility. α_i, δ_j, and λ_t denote country, sector, and time-specific fixed effects, respectively, while ϵ_i,j,t denotes the error term. To delve deeper into the role of regulatory quality on policy effectiveness, we augment our baseline specification with interaction terms of regulatory quality with FITs and tax, respectively. Further, for a robustness check, we have modelled all four sources of investment on the same independent variables with country and sector fixed effects only (Table A3).

4. Empirical results

Our empirical results are provided in Table 4. Across our full sample of countries, asset finance has been the dominant financing type for private investment in renewable energy over the past decade. The share accounted for by corporate R&D, public markets, and venture capital and private equity is much smaller by comparison (Figure 4). Nonetheless, an understanding of the factors driving different types of financing can have important implications for efforts aimed at boosting renewable energy investment from private sources. The discussion of the results focuses on a comparison of Asia with that of the full sample of countries.

Government R&D, or public expenditure in R&D of renewable technologies, has a positive impact on investments in renewable energy sources from asset finance and corporate R&D in both the total sample and the sub-sample of Asian economies. Our null hypothesis for H1 from Section 3 is therefore rejected for both samples at the 99% level of statistical significance with respect to asset finance and corporate R&D. The results demonstrate that government R&D is effective in boosting private investments at the technology R&D technology rolling-out stages (Figure 3). We could not find evidence of an impact by government R&D on private investment sourced from public markets and venture capital and private equity (i.e. the null hypothesis for H1 is not rejected for these financing sources), which implies no evidence of the impact on private financing of technology development and manufacturing scale-up. This could be due to the longer duration needed for this impact of government R&D on public markets and venture capital and private equity to be emanated.

For the full sample of 13 economies, government R&D is an important driver of private investment in renewable energy, particularly asset finance-based private investment, whereby a 1% increase in government R&D expenditure increases asset finance-based private investment in renewable energy by 0.40% (Table 4). Government R&D expenditure is also statistically significant and positive for renewable energy financed by corporate R&D (0.28%). A statistically significant effect is not found for financing via public markets or venture capital and private equity. The results showing the positive impact of government R&D on private investments in renewable energy technologies are consistent with the existing literature described earlier, e.g. Deleidi et al. (Citation2020). In our more granular assessment by funding source finds, however, this relationship does not hold across all types of private investment financing. This is an important consideration for policy makers in targeting and channelling public expenditure to leverage greater returns. In particular, given that asset finance and corporate R&D yield significant renewable energy investment returns for a given level of government R&D expenditure, policy makers may focus on driving private investment in R&D via these sources of financing through mechanisms that increase their attractiveness to the private sector.

A similar pattern emerges for the sub-panel of Asian economies, with the largest relative effect on asset finance-based private investment being driven by government R&D, with the magnitude of the effect higher than that compared to the full sample, i.e. a 1% rise in government expenditure on R&D increases private investment from asset-based finance by around 0.56%. As in the case of the full sample, a positive significant effect is also found for corporate R&D as a financing source (0.39%).

Our second hypothesis, H2, tests the impact of FITs on the dependent variables. Rejecting the null hypothesis of no effect, our empirical results show that FITs have a statistically significant positive impact on private investments sourced from public market and venture capital and private equity in both the total sample and sub-sample of Asian economies. In addition, FITs have a significant positive impact on corporate R&D in the total sample, but not in the sub-sample of Asian economies. The results demonstrate that FITs are effective in boosting private investments in less mature renewable energy technologies which are at research, development, and manufacturing scale-up stages. This could be because most FIT rates are differentiated by technology, with greater tariff rates for less mature technologies, such as marine (wave and tidal). This can also explain why the impact of FITs may be stronger for corporate R&D, public market, and venture capital and private equity, which are usually used for financing less mature technologies for development and manufacturing scale-up. Our results showing the positive impact of FITs on private investments in renewable energy are consistent with the literature (Davies & Diaz-Rainey, Citation2011; Polzin et al., Citation2019).

For the total sample, FITs have a statistically significant and positive impact on private investment across financing from corporate R&D, publicly quoted markets, and venture capital and private equity. The implementation of FITs is associated with an increase in investments in renewable energy from corporate R&D by 1.24%, from public markets by 0.86%, and from venture capital and private equity by 0.54%. For the Asian sample, the impact of FITs is dominant in the case of public market-based financing, associated with rises in private investment of around 1.9%. FITs in Asia are associated with private investment rises from venture capital and private equity of 0.57%.

The greater role of FITs in underpinning private investment in Asia for public market financing may be explained by the fact that their duration has been much longer in fruition compared to other parts of Europe and the US for certain types of renewable energy. Countries in Asia, such as Republic of Korea and Japan, are at a later stage renewable energy adoption particularly in relation to wind and solar energy, and are already moving to the contracted revenue stream phase. Moreover, it is also important to note that the counterparties to the FITs in Asia have tended to be high quality such as state-owned utilities. Another explanation could be due to greater FIT rates in Asia compared to the rest of countries in the sample. Greater tariff rates are more effective in attracting private investments (Polzin et al., Citation2019). While venture capital and private equity financing is affected by FITs in a similar manner for the Asian sub-sample compared to all countries, corporate R&D financing is not affected by FITs in Asia, and this may be related to differences in market practices and underlying business models. Our results also show that greater regulatory quality does not increase the effectiveness of FITs in attracting private investment. This may seem counterintuitive but may be related to the possibility that countries with lower regulatory quality offer higher FIT rates, making them more effective at attracting private investments, i.e. higher incentives may be required in countries with lower regulatory quality. There may also be greater ease of market access.

Our third hypothesis, H3, that tax incentives (such as investment or production tax credits and tax relief, import duty on renewable energy technologies reduction or exemptions) have no effect on private investment in renewable energy is rejected for some financing sources, but not rejected for others. Tax incentives have a significant positive impact on private investment sourced from public markets in both the total sample and sub-sample of Asian economies. Tax incentives have a significant positive impact on asset finance for the total sample, but no impact in the Asian sub-sample. Also, the result shows a negative impact of tax incentives on corporate R&D in the total sample and on venture capital and private equity in the sub-sample of Asian economies. The mixed impact of tax incentives on private investments is consistent with the literature (Polzin et al., Citation2019). Tax incentives entail greater policy uncertainty compared to FITs, as they do not require a long-term commitment from the government. The negative impact of tax incentives on corporate R&D and venture capital and private equity, which is used for financing research, development and manufacturing scale-up, could be due to exemptions or reductions of import duty on renewable energy technologies. This could incentivize imports of renewable energy technologies, thus disincentivizing investments in corporate R&D and venture capital and private equity for financing technology research, development and manufacturing scale-up. On the other hand, the exemption or reduction of import duty on renewable energy incentivizes private investment sourced from asset finance (in the total sample) and public market (sub-sample of Asian economies) for the import of renewable energy technologies. Thus, the impact of tax incentives on private investments varies across types of tax incentives and sources of financing. Similar to tax incentives, the interaction term, regulatory quality and tax incentives, have a mixed impact on private investment. We also find some evidence that, unlike the case for the full sample of countries, regulatory quality and tax incentives may be important drivers in Asia for stimulating asset-based private investment.

Figure 7 provides an illustration of the relative effects of government expenditure on R&D, FITs, and tax incentives on private investment funding sources for statistically significant coefficients. These categories pertain to hypotheses H1 to H3 formulated earlier in the paper.

On energy specific costs, for the total sample, we find that lower costs of renewable energy technologies increase private investments in renewable energy for asset-based finance (thus rejecting the null hypothesis for H4 of no impact), in line with economic intuition and literature (Polzin et al., Citation2019). In contrast to the full sample model, technology costs in the Asian sample are even more important for private investment. For the Asian sub-sample, a 1% decline in technology costs lead to a rise in asset-based financing for private investment of around 1.50% (compared to 0.91% for the full sample of countries). This overall finding may be related to the fact that cost competitiveness in PRC, for example, is a strong factor that underpins its relative strength in the renewable energy sector. Energy prices are an important factor driving asset-based financing for private investment in the case of the Asian sub-sample (rejecting the null hypothesis for H5), while for the full sample of countries this is a significant factor driving venture capital and private equity based financing. Our results showing the positive impact of energy prices on private investments are consistent with the literature (Eyraud et al., Citation2013).

Our empirical analysis also controls for wider macroeconomic and financial conditions. While we find some statistically significant effects, such as related to the level economic development, the fixed effects incorporated in the model appear to be more important explanatory factors. That said, the level of financial development in the economy (as proxied by stock market capitalization), which is an important factor for consideration by foreign investors given that capital investment is intermediated via the financial sector, exhibits positive and significant effects on investment in some financing segments.