Chapter 11: Sci-Tech Finance Catalysis: Venture Capital and Government Guiding Funds
A solvent firm with a proven product can borrow. A pre-revenue firm with a promising chemistry cannot. This is not a temporary friction to be smoothed away by better credit scoring; it is a structural feature of debt contracts. Lenders are paid a fixed coupon and bear the full downside of failure. An equity investor is paid a share of the upside and can tolerate a portfolio in which most bets fail. Green technology at the early stage is, almost by definition, an equity problem dressed in debt clothing. When policy tries to solve it with credit, it gets either rationing (Chapter 2) or mispricing that shows up later as non-performing loans.
This chapter takes the third link in the chain. Rating-based identification (Chapters 5–6) tells a lender which technology firms are creditworthy and how to price environmental performance. Insurance (Chapters 8–10) buffers the risks that neither rating nor collateral can absorb. Catalysis is the part that supplies equity to firms that have no rating worth the paper it is printed on, no collateral, and no revenue. Venture capital is the private version. Government guiding funds are the public version, and in China they are the dominant institutional form. The question for this chapter is not whether the state should be involved in early-stage green finance—the market-failure argument is strong enough that some involvement is defensible—but whether the specific designs in use actually catalyze private capital or merely displace it.
11.1 Market Failure in Early-Stage Green Technology
Start with a concrete case. A firm is developing a solid-state electrolyte for grid-scale batteries. It has four PhD chemists, a lab, two patents, and a pilot line that will cost $8 million to build. It will not have a saleable product for four years. Its assets are mostly intangible and highly specific: the equipment cannot be redeployed to make anything else, and the patents are worth little to a lender because they are hard to value and harder to sell. Under a standard credit assessment, the firm fails on every dimension that matters. No cash flow. No collateral of the kind a bank can register and seize. No track record. The probability of technical success is perhaps 30 percent; the probability of commercial success conditional on technical success is perhaps 40 percent. A lender pricing this at any coupon consistent with its cost of funds would be underpricing risk by an order of magnitude.
The first failure is the debt-equity mismatch. Debt is structurally unsuited to finance innovation with long, uncertain, back-loaded payoffs. The mismatch is not a market imperfection that better information can fix. It follows from the payoff structure. A loan contract caps the lender's upside at the coupon and exposes it to the full loss on default. For a project with a 12 percent chance of a 10x return and an 88 percent chance of zero, the fair debt coupon does not exist—no coupon compensates for an expected loss of that magnitude unless the recovery value is high, and here it is near zero.
The second failure is information asymmetry of a specific kind. In Chapters 2 and 5 we dealt with the asymmetry between borrower and lender about existing cash flows and environmental performance. Early-stage technology has a different asymmetry: the entrepreneur knows more about the technical trajectory than any outside investor, and the gap cannot be closed by disclosure because the entrepreneur herself does not know whether the chemistry will work. This is not moral hazard in the classic sense (the entrepreneur is not hiding known bad news); it is fundamental uncertainty that no contract can resolve ex ante. Venture capital responds to this with staged financing and active monitoring—invest in tranches, sit on the board, retain the right to abandon. Debt cannot replicate this because the lender has no governance role and no upside to justify the monitoring cost.
The third failure is the public-good component of green innovation. A new battery chemistry or a cheaper electrolyzer generates knowledge spillovers. The firm that bears the R&D cost captures only part of the social return. Private investors therefore underinvest relative to the social optimum even when they are risk-tolerant. This is the classic argument for public R&D subsidy, and it applies with particular force to green technology because the social return includes avoided climate damage that no private investor can monetize. The spillover is not a hypothesis; it is measurable in patent citation patterns, and it is larger for early-stage than for late-stage innovation.
The fourth failure is coordination and ecosystem thinness. Early-stage green technology needs a cluster of complementary institutions: specialized labs, pilot-line facilities, technical talent, and follow-on investors who understand the sector. In a thin market, each of these is undersupplied because each depends on the others. A venture fund that wants to invest in solid-state batteries needs a co-investor who can do technical due diligence; if none exists, the fund passes. This is a coordination failure, and it is the strongest argument for a public institution that can act as an anchor investor and absorb the first-mover cost of building the ecosystem.
These four failures interact. The debt-equity mismatch means banks cannot fill the gap. The fundamental uncertainty means arm's-length equity cannot either, without governance. The spillover means even governance-equipped private equity underinvests. The coordination failure means the ecosystem that would lower the cost of all three is slow to form. The result is a financing gap at the earliest stages—what the literature calls the "valley of death" between basic research and commercial demonstration.
The standard policy response has three instruments: grants (which do not require repayment and are suited to the pre-commercial stage), public venture capital or guiding funds (which take equity and are suited to the pilot-to-commercial stage), and tax incentives (which are suited to firms with taxable income, i.e., not early-stage firms). This chapter is about the second instrument, because it is the one that has grown fastest and the one whose design problems are least understood.
11.2 Government Guiding Funds: Design and Operation
A government guiding fund (政府引导基金, often translated "government guidance fund") is a public fund that invests not directly in firms but in private venture funds, or co-invests alongside them, with the explicit objective of leveraging private capital into a policy-priority sector. The Chinese form is distinctive in scale and in the tightness of the policy constraints attached. The typical structure is a limited partnership in which the guiding fund is a limited partner (LP) and a private fund manager is the general partner (GP). The guiding fund contributes 20–30 percent of the fund's committed capital; the GP raises the rest from private LPs. The fund then invests in portfolio companies, and the guiding fund's return is subordinated or capped to make the deal attractive to private LPs.
The design has four moving parts, and each is a place where the catalysis can fail.
11.2.1 The leverage ratio and the additionality question
The leverage ratio is the number every guiding fund reports first. A fund that contributes 25 percent of committed capital against 75 percent from private LPs claims a nominal leverage of 3:1, and that figure appears in the annual report, the provincial economic work conference, and the fund's own evaluation scorecard. It is almost always misleading. The relevant question is additionality: would the private capital have been invested anyway, in the same firms, at the same stage, without the public money? If the answer is yes, the public capital has crowded out private capital one-for-one and the leverage ratio is an accounting artifact.
Cumming and MacIntosh (2006), studying Canadian evidence, found that public venture capital programs displaced private investment in some segments. The mechanism is straightforward. A private LP deciding whether to commit to a fund compares the fund's expected return to its alternatives. If a public LP is willing to take a subordinated position—accepting a lower return so that private LPs get a higher one—the private LP's expected return rises, and it commits capital it would not otherwise have committed. That is genuine additionality. But if the public LP simply co-invests on the same terms, the private LP's return is unchanged, and the public capital is a substitute for private capital that would have been there anyway. The subsidy has been captured by the private LP, not used to change its decision.
Figure 11.1: Structure of a typical government guiding fund.
The additionality test can be made operational, though it rarely is. Ask the GP: for each portfolio company, would you have invested at the same valuation and the same stage without the guiding fund's commitment? The honest answer is often no, because the guiding fund's commitment raised the fund's total size and allowed the GP to write a larger check or take a riskier position. The dishonest answer is yes, because the GP wants the next guiding fund commitment. The way to get an honest answer is to tie the GP's carry to a counterfactual: compare the fund's portfolio to a matched sample of deals the GP passed on or that were funded by non-guiding funds. Almost no guiding fund does this.
11.2.2 The policy constraints: sector, stage, and geography
Suzhou Industrial Park set up its first guiding fund in 2006 with a mandate that read, in effect, like a provincial five-year plan: a fixed share for biomedicine, a fixed share for nanotechnology, a fixed share for the park's own priority list. The GP had to fill the quotas or return the capital. Fifteen years later the park had a cluster of nanotech firms and a portfolio whose returns were never publicly disclosed. The design question embedded in that mandate—how much sectoral direction a public LP can impose before the GP stops behaving like a venture investor—is the second place catalysis fails.
A green guiding fund typically requires that 60–70 percent of committed capital go to firms in a defined list of green sectors, that a certain share go to early-stage firms, and that a certain share go to firms in a designated region. These constraints are the policy instrument, but they are also the source of the catalysis failure.
Consider a guiding fund that requires 70 percent of capital to go to early-stage green firms in a single province. The GP now faces a constrained portfolio problem. The pool of investable early-stage green firms in one province is small. To deploy the committed capital, the GP must either lower its technical bar or hold capital uninvested. If it lowers the bar, it invests in firms that would not have passed a market test, and the fund's returns fall. If it holds capital, the leverage ratio is not achieved and the policy objective is missed. The GP will choose the first option more often than the second, because its carry depends on deploying capital, and because the guiding fund's evaluation is based on deployment and on the number of firms supported, not on realized returns.
This is a principal-agent problem with multiple principals, and the literature on common agency (Dixit, Grossman, and Helpman, 1997; Martimort, 1996) describes it precisely. The GP is the agent. The private LPs and the guiding fund are two principals with different objective functions. The private LPs want risk-adjusted return. The guiding fund wants sectoral and regional targeting. The GP's compensation is set by the private LPs (carry and management fee) but its access to capital is controlled by the guiding fund. When the two objectives conflict, the GP will favor whichever principal controls its next fund. In practice, that is the guiding fund, because the private LPs are dispersed and the guiding fund is a repeat player.
The result is a fund that looks like a private VC fund in its legal form and behaves like a policy instrument in its portfolio. The private LPs have been used as a source of capital, not as a source of discipline. The catalysis has run in reverse: public objectives have been imposed on private capital, rather than private discipline being imposed on public money.
11.2.3 Compensation and the return cap
In 2015 the Shenzhen government revised the compensation rules for its guiding funds. The old rules paid managers a fixed management fee and a modest performance bonus tied to the number of firms supported; the new rules introduced a carry tied to realized returns above a hurdle, with the guiding fund's own return capped so that the excess flowed to the private LPs. The revision was presented as a technical fix. It was actually an admission that the earlier contract had produced the behavior the contract paid for: managers who deployed capital quickly, supported many firms, and did not much care whether the firms succeeded.
The standard private VC contract is "2 and 20": a 2 percent management fee and a 20 percent carried interest on profits above a hurdle rate (typically 8 percent). The guiding fund can alter this in two ways. It can cap the GP's carry, or it can cap the guiding fund's own return and distribute the excess to the GP or to the private LPs.
The return cap is the more interesting instrument. A guiding fund that caps its own return at, say, 8 percent and gives the excess to the private LPs is effectively subsidizing the private LPs' return. This raises the private LPs' expected return and can generate genuine additionality, because the private LPs now face a better risk-return tradeoff than they would in an unsubsidized fund. The subsidy is transparent, and its cost is the difference between the guiding fund's capped return and the return it would have earned on the same capital elsewhere.
But the return cap creates a new problem: it weakens the GP's incentive to maximize fund returns. If the guiding fund's return is capped, the marginal dollar of fund profit goes to the private LPs, not to the guiding fund. The GP's carry is calculated on total fund profit, so the GP still has an incentive to maximize returns. The private LPs do too. The guiding fund does not, because its return is capped. This is a minor problem when the fund is profitable, but it becomes acute when the fund is marginal: the guiding fund has no incentive to push for a workout or a follow-on investment that would raise the fund's return above the cap. The public capital becomes passive at exactly the moment when active governance matters most.
Jaaskelainen, Maula, and Murray (2007), studying Finnish hybrid VC funds, found that compensation structures in publicly and privately funded funds differ systematically, and that the differences affect investment behavior. Publicly funded managers with fixed salaries and no carry invest more conservatively and exit less aggressively. The lesson is that the compensation contract, not the ownership structure, determines behavior. A guiding fund that wants private-style behavior must pay private-style compensation, and that is politically difficult because the carry can be large.
11.2.4 Governance and the evaluation problem
Hefei's guiding fund published an evaluation framework in 2021 that scored its GPs on five indicators: leverage ratio, number of firms supported, share of portfolio in the priority sectors, number of firms that received follow-on investment, and number of firms that listed. Realized return was not among them. A GP who read that scorecard would have known exactly what to optimize for, and it was not the return to the public capital.
The Chinese guiding fund model typically involves a state-owned asset management company as the fund's administrative body, a selection committee that chooses the GP, and an annual evaluation that scores the GP on a set of indicators. The indicators usually include the leverage ratio, the number of firms supported, the sectoral composition of the portfolio, and the number of firms that subsequently listed or received follow-on investment.
The evaluation problem is that these indicators measure activity, not catalysis. A fund that supports 50 firms and leverages 3:1 private capital scores well, even if all 50 firms would have been funded anyway and the private capital was crowded out. A fund that supports 10 firms and leverages 1.5:1 scores poorly, even if those 10 firms are the ones that would not have been funded otherwise and the 1.5:1 is genuine additionality. The indicators are chosen because they are easy to measure, not because they are informative. This is a version of Goodhart's law: when a measure becomes a target, it ceases to be a good measure.
The deeper governance problem is that the guiding fund's ultimate principal is the government, and the government's objective function includes political and industrial-policy goals that are not reducible to a return target. A guiding fund that is told to support the provincial battery industry will support the provincial battery industry, even if the province has no comparative advantage in batteries. The fund's evaluation will not penalize this, because the evaluation is designed by the same government that set the objective. The result is a fund that is accountable to its political principal and not to the market test of returns.
11.3 Catalytic Effects and Evidence
The empirical literature on public venture capital is large and its findings are mixed, which is itself informative. The mixed results are not noise; they reflect real variation in design. The studies that find positive effects tend to study programs with strong additionality mechanisms (subordination, stage targeting, coinvestment with experienced private VCs). The studies that find crowding out tend to study programs that co-invest on the same terms as private capital or that invest directly without a private partner.
11.3.1 The direct-effect studies
Lerner (1999), studying the U.S. Small Business Innovation Research (SBIR) program over 1983–1995, found that SBIR awardees grew faster and were more likely to receive venture capital than matched non-awardees. The SBIR program is a grant program, not an equity fund, but the finding is relevant because it shows that public money can have a certification effect: an SBIR award signals technical quality to private investors, lowering the information asymmetry that deters them. The certification effect is a form of catalysis that does not require the public money to be invested alongside private money; it requires only that the public selection process be informative.
Da Rin, Nicodano, and Sembenelli (2006), studying European countries, found that public policy can create active venture capital markets, but that the effect depends on the policy instrument. Policies that reduce the cost of starting a VC fund (tax incentives, regulatory simplification) are associated with larger VC markets. Policies that directly supply public capital are associated with smaller effects, and in some specifications with crowding out. The interpretation is that the binding constraint in Europe was the supply of fund managers and the regulatory environment, not the supply of capital. Public capital addressed the wrong constraint.
Cumming and MacIntosh (2006) found that Canadian public VC programs crowded out private investment, particularly in later-stage deals. The mechanism was that public funds invested in the same deals as private funds, at the same terms, so the private funds simply reduced their own commitments. The public money did not change the set of firms that were funded; it changed who funded them.
Cumming (2007), studying the Australian Innovation Investment Fund program, found more positive results. The program used a fund-of-funds structure with private managers and a subordinated public position, and it targeted early-stage firms. The design features that distinguished it from the Canadian program were the subordination (which created genuine additionality) and the stage targeting (which addressed the segment where the market failure is strongest).
11.3.2 The Chinese evidence
The scale of the Chinese program makes it the most important case in the world, and the evidence on it is thinner than the scale warrants. By the end of 2023, according to the Zero2IPO Research database, government guiding funds at the national, provincial, and municipal levels had committed more than RMB 6.5 trillion in total fund size, with the guiding funds' own commitments at roughly RMB 2 trillion and the remainder raised from private and state-owned LPs. The funds were concentrated in strategic emerging industries. New energy, new materials, and energy-saving equipment together accounted for a large share of the sectoral allocations, though the exact split varies by province and by year.
The fund names matter because the designs differ. The National Integrated Circuit Industry Investment Fund, known as the Big Fund, was established in 2014 with RMB 138.7 billion in Phase I and RMB 204.15 billion in Phase II (2019); it is a direct-investment fund rather than a fund-of-funds, and its mandate is semiconductors rather than green technology, but it established the template that later green funds copied. The National Green Development Fund, established in 2020 with RMB 88.5 billion, is the largest fund with an explicit green mandate; it invests through a mix of direct equity and sub-funds, with the Ministry of Finance and the major state banks as anchor LPs. The Yangtze River Delta Green Industry Investment Fund, launched in 2021 with a target of RMB 50 billion, is structured as a fund-of-funds and requires its sub-funds to invest at least 60 percent of capital in the delta region. The Guangdong Green Industry Fund, established in 2022 with RMB 30 billion, targets early-stage clean technology and has a stated subordination provision for private LPs. The Hubei Yangtze River Economic Belt Industrial Guidance Fund, established in 2016 with an initial RMB 40 billion, was one of the first provincial funds to write a green sector quota into its mandate.
The leverage ratios reported by these funds are typically 3:1 to 5:1. The additionality of the commitments is not known, because no systematic counterfactual evaluation has been published. Zhang Luyang (2010), surveying Chinese guiding funds in their first decade, concluded that the funds had increased the supply of early-stage capital but that their governance and evaluation mechanisms were underdeveloped. The specific problems he identified were the ones described above: the evaluation focused on deployment rather than returns, the policy constraints limited the GP's discretion, and the return caps weakened the GP's incentive to maximize value. Fifteen years later the same problems appear in the audit reports of provincial funds, which is itself a finding: the design has scaled faster than the governance.
11.3.3 The certification channel
Lerner's SBIR finding is the clearest evidence for the certification effect, and later work has refined it. The effect is strongest when the public program is selective (a low acceptance rate) and when the public evaluator has sectoral expertise. A program that funds 60 percent of applicants certifies nothing; a program that funds 10 percent and publishes its technical review criteria certifies a great deal.
For green technology, the certification channel has a specific application. A guiding fund that specializes in green technology can build technical due-diligence capacity that no individual private fund can afford. The fund's investment decision then carries information about the technology's viability. Private investors who lack the capacity can free-ride on the public assessment. This is a public good—the technical assessment is non-rival and non-excludable—and it is undersupplied by the private market. The guiding fund is the natural supplier.
The certification channel also has a dark side. If the guiding fund's selection is politically motivated, its certification is uninformative or misleading. A fund that invests in a firm because it is in a favored province, not because the technology is sound, sends a false signal. Private investors who learn to discount the signal will ignore it, and the certification channel closes. The credibility of the public evaluator is therefore a fragile asset, and it is spent by every politically motivated investment.
11.3.4 The stage-targeting evidence
The stage at which public capital is deployed matters more than the amount. The market failure is concentrated at the earliest stages, where the debt-equity mismatch is most severe and where private VC is thinnest. Public capital deployed at later stages, where private VC is active, is more likely to crowd out. This is the consistent finding across the Canadian, Australian, and European studies.
The implication for green technology is that guiding funds should target the pilot-to-commercial stage, not the growth stage. At the growth stage, private VC and private equity are available, and the public money is a substitute. At the pilot stage, private capital is scarce, and the public money can be catalytic. The Chinese funds have not consistently followed this rule; many are mandated to invest across all stages, and the later-stage investments are the ones that show the least additionality.
11.4 The Catalysis Mechanism, Stated Precisely
It is worth stating the catalysis mechanism in a form that can be tested. The guiding fund's capital is catalytic if it changes the private investor's decision at the margin. Formally, let \(V\) be the private investor's expected value of a portfolio company, \(C\) the cost of the investment, and \(S\) the subsidy provided by the guiding fund (the present value of the subordination, the return cap, or the certification signal). The private investor invests if \(V + S \geq C\). Without the subsidy, the investor invests if \(V \geq C\). The subsidy is catalytic if it moves the investor from not investing to investing, i.e., if \(C - S \leq V < C\).
The subsidy is a pure transfer if \(V \geq C\) without it. In that case, the guiding fund's capital has crowded out private capital, and the leverage ratio overstates the catalysis. The policy problem is that \(V\) is unobservable to the guiding fund. The fund observes the private investor's decision, but not the counterfactual. This is why the additionality question is hard to answer and why most evaluations fall back on the leverage ratio.
The certification signal is a special case. If the guiding fund's investment reveals information about \(V\) that the private investor did not have, the signal raises the private investor's estimate of \(V\). The catalysis is real even if no subsidy is transferred, because the information is new. The size of the effect depends on the credibility of the signal, which depends on the fund's track record and on the independence of its selection process.
The equation is an identity, not an estimate. Its value is that it separates the channels and makes clear that they can have opposite signs. A fund with a strong certification effect and a weak subsidy can be catalytic even if its leverage ratio is low. A fund with a strong subsidy and a weak certification effect can be a pure transfer even if its leverage ratio is high.
11.5 Design Rules for a Green Guiding Fund
The evidence supports a small number of design rules. They are stated as rules because the alternative—a list of best practices—is too vague to be operational.
The first rule is to target the stage where private capital is absent. The fund should invest at the pilot-to-commercial stage, where the debt-equity mismatch is most severe. It should not invest at the growth stage, where private VC is active. The stage target should be written into the fund's mandate and verified in the evaluation. The Australian Innovation Investment Fund is the clean example: it was legislated to invest only in early-stage firms, and its positive evaluation reflects that restriction. The Canadian programs that Cumming and MacIntosh studied had no such restriction, invested across all stages, and crowded out private capital in the later-stage deals where private capital was already present.
The second rule is to use subordination, not co-investment on equal terms. The public capital should take a subordinated position, so that the private investor's return is genuinely higher than it would be without the public money. Co-investment on equal terms is a transfer, not a catalysis. The Australian program subordinated the public position; the Canadian programs did not, and the difference shows up in the evaluations. A green fund that wants to be catalytic should write the subordination into the limited partnership agreement and disclose it.
The third rule is to pay private-style compensation. The GP's carry should be tied to fund returns, not to deployment or to the number of firms supported. A GP paid a fixed salary will behave like a bureaucrat, and the fund will underperform. Jaaskelainen, Maula, and Murray (2007) found exactly this in the Finnish hybrid funds: the publicly paid managers invested more conservatively and exited less aggressively than their privately paid counterparts. The Shenzhen revision of 2015 was an attempt to move from the first compensation model to the second, and the fact that it required a formal revision is a reminder that the default public contract is the wrong one.
The fourth rule is to evaluate against a counterfactual. The fund's evaluation should compare its portfolio to a matched sample of non-funded firms. The leverage ratio should be reported alongside the additionality estimate, not instead of it. No Chinese guiding fund currently does this, and the absence is the single largest gap in the evidence base.
The fifth rule is to protect the certification signal. The fund's selection should be independent of political pressure. A fund that invests for regional or industrial-policy reasons should say so explicitly, and its investments should be excluded from the certification signal. The credibility of the signal is the fund's most valuable asset, and it is spent by every politically motivated investment.
The sixth rule is to build technical due-diligence capacity. The fund should employ or contract sectoral experts who can assess the technology. The assessment is a public good, and it is the basis of the certification effect. A fund that relies on financial due diligence alone cannot certify technology.
These rules are not novel. They are the design features that distinguish the programs with positive evaluations from the programs with crowding-out findings. The Chinese guiding funds have adopted some of them and not others. The subordination is common; the counterfactual evaluation is rare; the independent selection is rare; the stage targeting is inconsistent. The pattern is that the rules which cost the fund money or political capital are the ones least adopted, which is what a principal-agent model would predict.
11.6 The Green Premium and the Catalysis Decision
Chapter 4 introduced the green premium as a diagnostic metric: the cost differential between a green technology and its carbon-intensive incumbent, expressed per unit of output or per ton of avoided emissions. The green premium is the central input to the catalysis decision, because it determines the size of the subsidy required to make a green project commercially viable.
Consider a green hydrogen project. The levelized cost of green hydrogen is $5/kg; the levelized cost of gray hydrogen is $2/kg. The green premium is $3/kg. A private investor comparing the two will invest in gray hydrogen unless the green premium is closed by a subsidy, a carbon price, or a long-term offtake contract at a green price. The guiding fund's role is to finance the project through the period when the green premium is positive, in the expectation that learning-by-doing will close it. The fund's return depends on whether the premium closes before the fund exits.
A worked example makes the timing concrete. Suppose the green premium starts at $3/kg and falls by 12 percent a year as cumulative capacity doubles, a learning rate consistent with the observed cost declines in solar PV and lithium-ion batteries. At that rate the premium falls to $1.50/kg in about six years and to $0.75/kg in about twelve. The project's cumulative cash flow turns positive only when the premium falls below the project's operating margin, which in this example happens in year nine. A fund with a standard seven-year VC horizon exits in year seven, three years before the project turns cash-positive, and sells at a discount to the value it created. A fund with a twelve-year horizon captures the turn. The same project, same technology, same management team, produces a loss for the short-horizon fund and a gain for the long-horizon fund, and the difference is entirely the exit date.
The green premium is therefore a timing variable, not a fixed cost. It falls as the technology scales, as manufacturing costs decline, and as the carbon price rises. The fund's investment horizon should match the expected time to premium closure. If the fund exits before the premium closes, it realizes a loss. If it exits after, it realizes a gain. The fund's evaluation should be based on the realized premium closure, not on the project's accounting profit in the first year.
This is a different investment logic from standard VC, which exits on a liquidity event (IPO or trade sale) rather than on a technology-cost milestone. A green guiding fund that uses standard VC exit metrics will exit too early, before the premium closes, and will capture only a fraction of the value it created. The fund's mandate should specify the premium-closure milestone as the exit condition, and the compensation should be structured around it.
11.7 The Interaction with Rating and Insurance
The catalysis link does not operate in isolation. It interacts with the identification link (Chapters 5–6) and the buffering link (Chapters 8–10) in ways that determine the overall performance of the chain.
The interaction with rating is through the exit path. A venture-backed green firm eventually needs debt to scale. The rating model developed in Chapter 5 is the instrument that converts the firm's environmental performance into a credit score that a bank can use. Without the rating, the firm is stuck at the equity stage: it can raise VC but cannot borrow, and its growth is constrained by the pace of equity raises. The rating is the bridge from catalysis to scale. A guiding fund that invests in a firm but does not prepare it for rating is investing in a firm that will remain dependent on equity.
The interaction with insurance is through the risk transfer at the project-finance stage. A green project at the commercial stage—a solar plant, a reforestation project, a bioenergy facility—faces risks that equity cannot absorb: weather risk, fire risk, counterparty risk. The insurance instruments of Chapters 8–10 transfer these risks, lowering the project'scost of capital and making it financeable. The guiding fund's investment at the pilot stage is de-risked at the commercial stage by the insurance. Without the insurance, the pilot-stage investment has no exit, because no commercial-stage financier will take the project risk.
The chain is therefore sequential: catalysis supplies equity at the earliest stage, rating supplies debt at the growth stage, insurance supplies risk transfer at the project stage. A gap at any link breaks the chain. The guiding fund is the first link, and its performance determines whether the later links have anything to work with.
11.8 The Limits of Catalysis
The chapter would be incomplete without a candid statement of what catalysis cannot do. Three limits are worth naming.
First, catalysis cannot substitute for a carbon price. The green premium is a function of the carbon price, and no amount of public equity can close a premium that is structurally large. A guiding fund that invests in a technology with a $100/ton green premium and no carbon price is investing in a technology that will not be commercially viable at scale. The fund's money is a bridge to nowhere. The carbon price is the load-bearing policy, and the guiding fund is a complement, not a substitute.
Second, catalysis cannot fix a weak ecosystem. A fund that invests in a region with no technical talent, no suppliers, and no follow-on investors will produce firms that cannot scale. The ecosystem is a precondition, not a product, of the fund. The coordination-failure argument for public investment is real, but it requires the public investor to build the ecosystem (labs, pilot facilities, training) as well as to supply capital. A fund that supplies only capital is solving the wrong problem.
Third, catalysis cannot be evaluated without a counterfactual, and the counterfactual is expensive to construct. The honest position is that most guiding funds do not know whether they are catalytic. They know how much they have deployed and how much private capital they have nominally leveraged. They do not know whether the private capital would have been deployed anyway. This is not a reason to abandon the instrument; it is a reason to build the evaluation capacity before scaling the instrument further.
11.9 An Open Question from the Data
One number from the Chinese program deserves to close the chapter rather than a summary of its claims. The National Green Development Fund, with RMB 88.5 billion under management, has published the sectoral allocation of its first wave of sub-fund commitments but not the realized returns of any portfolio company. The provincial green funds have published leverage ratios but not additionality estimates. The result is a program of unprecedented scale whose catalytic effect is, by the standard the program's own architects would apply to a private fund, unmeasured.
The open question is not whether public equity has a role in early-stage green technology. The market failures in Section 11.1 make that case. The question is whether the specific design in use—sector quotas, regional quotas, return caps, activity-based evaluation—produces the catalysis it claims, or whether it produces a large, well-intentioned transfer from public budgets to private LPs and portfolio firms that would have been funded anyway. Answering it requires the matched-sample evaluation described in Section 11.3.2, applied to at least one of the named funds, with the results published. Until that evaluation exists, the leverage ratio will continue to stand in for the answer, and the ratio measures deployment, not catalysis.
The next chapter takes up the incentive design and performance evaluation of policy-guided green finance in more detail. The question there is how to structure the compensation and the evaluation so that the fund's managers face the same incentives as the private investors they are supposed to catalyze. The answer, as this chapter has suggested, is not to imitate private compensation in form but to align it in substance: pay for returns, measure against a counterfactual, and protect the certification signal from political capture.