## 6. Global: Does Risk Culture Matter for Startup Exit?

## Source details

**Canonical URL:** [6. Global: Does Risk Culture Matter for Startup Exit?](https://www.imf.org/-/media/files/publications/wp/2024/english/wpiea2024246-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2024/english/wpiea2024246-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2024/english/wpiea2024246-print-pdf.pdf.json)

---

### I. Introduction — research question and approach
- Purpose: examine factors that support startups and venture capital (VC) funding using a novel cross-country database on startups, with a focus on Japan.
- Two empirical exercises:
  - Cross-country (30 large advanced and emerging market economies, 2000–2022) assessing how structural factors (firm dynamism, entrepreneurship) affect capital investment and firm valuations.
  - Firm-level (cross-sectional, end-2023) analysis using PitchBook on startups (young <10 years and backed by early-stage funding) to study how equity funding affects performance, innovation, and exit, and whether cultural traits affect exit probabilities.
- Identification strategy at firm level:
  - Match firms on observables (age, employment, industry) and estimate an endogenous treatment model treating WellFunded as endogenous.
  - Instruments/auxiliary variables for WellFunded: total number of active investors, years since first funding, years since last funding.
- Estimators and specifications:
  - Aggregate: Arellano-Bover/Blundell-Bond linear dynamic panel data estimator with robust standard errors (equation (1)).
  - Firm-level outcome equation: Outcome_i = β X_i + δ WellFunded_i + φ_country-sector + ε_i (equation (2)); WellFunded auxiliary regression: WellFunded_i = β Z_i + u_i (equation (3)).

### II. Stylized facts — Japan’s startup ecosystem
- Scale and composition:
  - Japan has over 10,000 startups.
  - Startups in Japan tend to be smaller compared with the U.S., China, and the U.K.
  - Number of unicorns in Japan: 7 in 2023 (vs 653 in the U.S.).
  - Valuation of unicorns in Japan: about US$9.8 billion in 2023 (vs US$2 trillion in the U.S.).
  - Japan’s unicorns represent just 0.5% of the global total.
  - Total Early-Stage Entrepreneurial Activity (TEA) Index, 2022: Japan 6.4; U.S. 19.2; U.K. 12.9; South Korea 14.9.
- Venture capital funding and geography:
  - Japan’s venture capital equity funding is relatively small as a share of GDP compared with peers (about one tenth of that of the US).
  - Tokyo is among the top 20 cities in the global VC ecosystem (PitchBook rankings); Tokyo accounts for about 80% of total startup funding in Japan.
- Government support:
  - “Startup Development Five-Year Plan” aims to increase the scale of investment in startups by 10-fold by 2027 and focuses on: (1) building human resources and networks; (2) strengthening funding and diversifying exit strategies; (3) promoting open innovation.
  - Japan Investment Corporation launched a 200-billion-yen venture growth fund in 2023 to support later-stage startups and target early-stage startups beyond deep tech and life sciences.
- Industry composition:
  - Most startups operate in information technology (IT), followed by health care, and business products and services (B2B).
  - Japan has a somewhat higher share of IT / SaaS startups and a somewhat lower share of clean technology-related startups compared with the global sample.
- Investor origins and corporate VCs:
  - U.S. investors account for about 50 percent of investors in Japan’s startups between 2010 and 2023.
  - U.K. investors account for about 10 percent; Japanese investors’ share declined to about 5 percent in recent years.
  - Corporate venture capital (CVC) prevalence: Japanese CVCs invested in at least half of all VC deals in Japan between 2015 and 2022, peaking at 62 percent in 2020.

### III. Data and methodology — sample and variables
- Aggregate dataset:
  - Cross-country panel of 30 large advanced and emerging market economies, 2000–2022.
  - Firm performance and valuations from PitchBook.
  - Structural proxies: entrepreneurship (share of self-employed who are employers), firm dynamism (entry and exit rates), real GDP growth, inflation, average job tenure.
- Firm-level sample:
  - Startups defined as companies <10 years old and backed by early-stage funding at least once.
  - Final sample includes startups from 13 countries: Australia, Canada, France, Germany, India, Italy, Japan, Korea, Singapore, Spain, Sweden, the U.K., and the U.S.
  - Data are cross-sectional as of end-2023.
- Matching and treatment:
  - Coarsened exact matching on country, industry, age, number of employees to pair WellFunded vs not WellFunded startups.
  - WellFunded indicator: 1 if total capital raised per employee is above the country-industry median, 0 otherwise (robustness: 75th vs 25th percentile threshold tested).
- Key econometric specifications and inference:
  - Aggregate: equation (1) with dependent variable Y_{k,t} = mean capital invested or mean valuation, independent variables X_{k,t}.
  - Firm-level: Outcome_i = β X_i + δ WellFunded_i + φ_country-sector + ε_i (equation (2)); identification via endogenous treatment model and auxiliary regression WellFunded_i = β Z_i + u_i (equation (3)).
  - Standard errors clustered at the country level; cov(ε_i, u_i) test reported (χ^2 distribution p-value reported in results).

### IV. Key empirical findings
- Aggregate (cross-country) results:
  - A higher share of entrepreneurship (share of self-employed who are employers) is associated with higher capital investment in firms (Table 2) and higher valuation of firms (Table 3).
  - Better firm dynamism (higher entry and exit rates) is associated with greater valuations and higher capital investment.
  - Japan exhibits relatively low exit rates and a low share of self-employed (who are employers) compared with other OECD countries; entry and exit rates in Japan tend to be lower relative to peers.
- Firm-level results (global sample and Japan subsample):
  - Well-funded startups (total capital raised per employee above country-industry median) perform better on multiple dimensions (Tables 4 and 5):
    - Higher number of total patent documents (log of total patent documents).
    - Larger in terms of number of employees (log of total number of employees).
    - Higher predicted exit probabilities (via IPO or M&A) as reported by PitchBook’s machine-learning estimated probability of exit.
  - Effects are statistically significant and economically meaningful (predicted magnitudes reported in paper tables).
  - Summary statistics (Table 1) — global sample (N = 4,422) and Japan subsample (N = 307):
    - Global sample:
      - Well-Funded mean 0.50; Std.Dev 0.50; Min 0.00; 25th 0.00; Median 1.00; 75th 1.00; Max 1.00; N 4422.
      - Total Patent Documents, log: mean 2.08; Std.Dev 1.08; Min 0.00; 25th 1.39; Median 1.95; 75th 2.83; Max 5.69; N 4422.
      - Total Number of Employees, log: mean 3.15; Std.Dev 1.11; Min 0.00; 25th 2.30; Median 3.14; 75th 3.87; Max 6.80; N 4422.
      - Age, in years, log: mean 1.72; Std.Dev 0.34; Min 0.00; 25th 1.61; Median 1.79; 75th 1.95; Max 2.08; N 4422.
      - Active Number of Investors: mean 9.83; Std.Dev 8.89; Min 1.00; 25th 4.00; Median 8.00; 75th 13.00; Max 1814; N 4422.
      - Years since first financing round: mean 4.80; Std.Dev 1.83; Min 0.00; 25th 3.00; Median 5.00; 75th 6.00; Max 8.00; N 4422.
      - Years since last financing round: mean 0.87; Std.Dev 1.10; Min 0.00; 25th 0.00; Median 1.00; 75th 1.00; Max 5.00; N 4422.
      - Uncertainty Avoidance Index: mean 0.55; Std.Dev 0.18; Min 0.29; 25th 0.46; Median 0.46; 75th 0.48; Max 0.92; N 3628.
      - Power Distance Index: mean 0.45; Std.Dev 0.10; Min 0.31; 25th 0.40; Median 0.40; 75th 0.40; Max 0.77; N 3628.
    - Japan subsample:
      - Well-Funded mean 0.51; Std.Dev 0.50; Min 0.00; 25th 0.00; Median 1.00; 75th 1.00; Max 1.00; N 307.
      - Total Patent Documents, log: mean 1.74; Std.Dev 1.27; Min 0.00; 25th 0.69; Median 1.79; 75th 2.64; Max 4.89; N 307.
      - Total Number of Employees, log: mean 2.89; Std.Dev 1.08; Min 0.69; 25th 2.08; Median 2.71; 75th 3.64; Max 6.21; N 307.
      - Age, in years, log: mean 1.72; Std.Dev 0.34; Min 0.69; 25th 1.61; Median 1.79; 75th 1.95; Max 2.08; N 307.
      - Active Number of Investors: mean 9.96; Std.Dev 6.94; Min 1.00; 25th 5.00; Median 8.00; 75th 13.00; Max 43.00; N 307.
      - Years since first financing round: mean 4.64; Std.Dev 1.76; Min 0.00; 25th 3.00; Median 5.00; 75th 6.00; Max 8.00; N 307.
      - Years since last financing round: mean 1.10; Std.Dev 1.12; Min 0.00; 25th 0.00; Median 1.00; 75th 2.00; Max 5.00; N 307.

### Availability of funding — measured impacts (global and Japan)
- Well-funded startups have 1.5 times more employees and exit with 43 percentage points higher probability, compared to non-well-funded startups.
- Availability of funding is associated with a larger number of patents for startups (impact on patents appears larger and more precisely estimated in the Japan subsample).
- For the Japan sample, the impact of availability of funding on the number of patents seems larger and more precisely estimated, while the estimated impact on other outcome variables does not seem to be materially different from the global sample.

### Endogeneity and auxiliary evidence
- Evidence of endogeneity:
  - The p-value of the null hypothesis of uncorrelated error terms in the main and auxiliary regressions is strongly rejected in the majority of specifications (the p-value of the χ^2 statistic reported in the last row of the upper panel of Table 4).
- Instruments:
  - The number of active investors is statistically significant in predicting whether or not a startup is well-funded (lower panel of Table 4).
- Identification:
  - Endogenous treatment model estimates of equation (2) are used (global sample and Japan sample reported separately). Each specification includes country-sector fixed effects; standard errors clustered at the country level.

### Role of risk-taking culture
- Cultural proxies used: Hofstede (2013) indices — (i) uncertainty avoidance; and (ii) power distance.
- Hypothesis: a greater positive impact of availability of funding on startup performance in countries that reward risk-taking behavior.
- Results:
  - Predicted impact of availability of funding on IPO exit is higher in countries with less uncertainty avoidance.
  - Predicted impact of availability of funding on M&A exit is higher in countries with less power distance.
- Estimates presented in Table 6 (endogenous treatment model estimates of equation (4)); country-sector fixed effects included; standard errors clustered at the country level.

### Caveats and data limitations
- Data are cross-sectional; unable to track startups’ performance following funding rounds.
- Vintage and survivor bias: analysis studies a particular vintage of the data (end-2023) that includes surviving startups; data for firms that had exited before end-2023 is not available, hence survivor bias remains (direction of bias not evident).
- Study focuses on availability of funding and does not explore the role of cost, conditionality, or particular types of investors (for example, differences between investor types), due to data limitations.
- Identification is mainly within country-industry, which sharpens identification but overlooks potential variation across countries and/or industries.
- Analysis does not resolve whether private equity funding complements or substitutes other types of financing (for example, debt financing).

### Policy-relevant implications and conclusions
- Strengthen entrepreneurship and firm dynamism:
  - Higher entrepreneurship (share of self-employed who are employers) and more dynamic entry and exit are associated with higher capital investment and firm valuations at the country level.
  - Policies that reduce barriers to entry/exit and encourage entrepreneurship could support capital allocation and firm valuation.
- Support later-stage funding and diversify exit channels:
  - Japan’s startup funding is concentrated in seed and early-stage VC; increasing later-stage funding and diversifying exit options (IPO, M&A) may help scale startups and create more unicorns.
  - The Japan Investment Corporation’s 200-billion-yen venture growth fund (2023) targets later-stage support to create unicorns and expand beyond deep tech and life sciences.
- Leverage corporate VCs while expanding investor base:
  - Corporate VCs play a large role (at least half of VC deals 2015–2022; peak 62 percent in 2020). Broadening the investor base (domestic and international) could improve funding resilience and exit pathways.
- Recognize cultural factors in exit outcomes:
  - The impact of funding on exit probabilities is larger in cultures that tolerate risk taking; cultural and institutional reforms that reduce risk aversion may increase the payoff of funding in terms of successful exits.
- Labor market and structural reforms to support entrepreneurship:
  - A gradual shift away from the lifelong employment system in Japan could encourage startup creation and offer second chances after failure.
  - Reducing labor-market dualism, encouraging merit-based promotions, and facilitating more job mobility can encourage entrepreneurship and are associated with higher capital investment and firm valuations at the country level.
- Increase firm dynamism and reallocate resources:
  - Dynamic firm entry and exit and reduced personal liabilities can encourage entrepreneurship, innovation, and more efficient resource allocation.
  - Gradual reduction of zombie firms could improve allocation of capital and labor to more productive ventures, boosting productivity and growth.

*IMF Working Paper chapter: “6. Global: Does Risk Culture Matter for Startup Exit?” from Riding Unicorns: Startups and Venture Capital in Japan (chapter content as provided).*

### References .............................................................................................................

### References ......................................................................................................................................................... 17

### Tables
- 1. Summary Statistics ......................................................................................................................................... 11
- 2. Cross Country: Capital Investment and Structural Characteristics ................................................................. 12
- 3. Cross-Country: Valuations and Structural Characteristics .............................................................................. 12
- 4. Global: Does Availability of Funding Affect Startup Performance? ................................................................. 14
- 5. Japan: Does Availability of Funding Affect Startup Performance? .................................................................. 14

*Source: wpiea2024246-print-pdf - References .............................................................................................................*

### 6. Global: Does Risk Culture Matter for Startup Exit? ..................................................................

### 6. Global: Does Risk Culture Matter for Startup Exit?

### I. Introduction — research question and approach
- Purpose: examine factors that support startups and venture capital (VC) funding using a novel cross-country database on startups, with a focus on Japan.
- Two empirical exercises:
  - Cross-country (30 large advanced and emerging market economies, 2000–2022) examining how structural factors (firm dynamism, entrepreneurship) affect capital investment and firm valuations.
  - Firm-level (cross-sectional, end-2023) analysis using PitchBook on startups (young <10 years and backed by early-stage funding) to study how equity funding affects performance, innovation, and exit, and whether cultural traits affect exit probabilities.
- Identification strategy at firm level:
  - Match firms on observables (age, employment, industry) and then estimate an endogenous treatment model treating WellFunded as endogenous.
  - Instrumental / auxiliary variables for WellFunded: total number of active investors, years since first funding, years since last funding.
- Estimators and specifications:
  - Aggregate: Arellano-Bover/Blundell-Bond linear dynamic panel data estimator with robust standard errors (equation (1)).
  - Firm-level outcome equation as in equation (2); WellFunded auxiliary regression as in equation (3).

### II. Stylized facts — Japan’s startup ecosystem
- Scale and composition:
  - Japan has over 10,000 startups.
  - Startups in Japan tend to be smaller compared with the U.S., China, and the U.K.
  - Number of unicorns in Japan: 7 in 2023 (vs 653 in the U.S.).
  - Valuation of unicorns in Japan: about US$9.8 billion in 2023 (vs US$2 trillion in the U.S.).
  - Japan’s unicorns represent just 0.5% of the global total.
  - Total Early-Stage Entrepreneurial Activity (TEA) Index, 2022: Japan 6.4; U.S. 19.2; U.K. 12.9; South Korea 14.9.
- Venture capital funding and geography:
  - Japan’s venture capital equity funding is relatively small as a share of GDP compared with peers (about one tenth of that of the US).
  - Tokyo is among the top 20 cities in the global VC ecosystem (PitchBook rankings); Tokyo accounts for about 80% of total startup funding in Japan.
- Government support:
  - “Startup Development Five-Year Plan” aims to increase the scale of investment in startups by 10-fold by 2027 and focuses on: (1) building human resources and networks; (2) strengthening funding and diversifying exit strategies; (3) promoting open innovation.
  - Japan Investment Corporation launched a 200-billion-yen venture growth fund in 2023 to support later-stage startups and target early-stage startups beyond deep tech and life sciences.
- Industry composition:
  - Most startups operate in information technology (IT), followed by health care, and business products and services (B2B). Using an alternative “vertical” classification yields similar results.
  - Japan has a somewhat higher share of IT / SaaS startups and a somewhat lower share of clean technology-related startups compared with the global sample.
- Investor origins and corporate VCs:
  - U.S. investors account for about 50 percent of investors in Japan’s startups between 2010 and 2023.
  - U.K. investors account for about 10 percent; Japanese investors’ share declined to about 5 percent in recent years.
  - Corporate venture capital (CVC) prevalence: Japanese CVCs invested in at least half of all VC deals in Japan between 2015 and 2022, peaking at 62 percent in 2020.

### III. Data and methodology — sample and variables
- Aggregate dataset:
  - Cross-country panel of 30 large advanced and emerging market economies, 2000–2022.
  - Firm performance and valuations from PitchBook.
  - Structural proxies: entrepreneurship (share of self-employed who are employers), firm dynamism (entry and exit rates), real GDP growth, inflation, average job tenure.
- Firm-level sample:
  - Startups defined as companies <10 years old and backed by early-stage funding at least once.
  - Final sample includes startups from 13 countries with largest coverage: Australia, Canada, France, Germany, India, Italy, Japan, Korea, Singapore, Spain, Sweden, the U.K., and the U.S.
  - Data are cross-sectional as of end-2023.
- Matching and treatment:
  - Coarsened exact matching on country, industry, age, number of employees to pair WellFunded vs not WellFunded startups.
  - WellFunded indicator: 1 if total capital raised per employee is above the country-industry median, 0 otherwise (robustness: 75th vs 25th percentile threshold tested).
- Key econometric specifications:
  - Aggregate: equation (1) with dependent variable Y_{k,t} = mean capital invested or mean valuation, independent variables X_{k,t}.
  - Firm-level outcome equation: Outcome_i = β X_i + δ WellFunded_i + φ_country-sector + ε_i (equation (2)); WellFunded_i auxiliary regression: WellFunded_i = β Z_i + u_i (equation (3)).
  - Standard errors clustered at the country level; cov(ε_i, u_i) test reported (χ^2 distribution p-value reported in results).

### IV. Key empirical findings
- Aggregate (cross-country) results:
  - A higher share of entrepreneurship (share of self-employed who are employers) is associated with higher capital investment in firms (Table 2) and higher valuation of firms (Table 3).
  - Better firm dynamism (higher entry and exit rates) is associated with greater valuations and higher capital investment.
  - Japan exhibits relatively low exit rates and a low share of self-employed (who are employers) compared with other OECD countries; entry and exit rates in Japan tend to be lower relative to peers.
- Firm-level results:
  - Well-funded startups (total capital raised per employee above country-industry median) perform better on multiple dimensions (Tables 4 and 5):
    - Higher number of total patent documents (log of total patent documents).
    - Larger in terms of number of employees (log of total number of employees).
    - Higher predicted exit probabilities (via IPO or M&A) as reported by PitchBook’s machine-learning estimated probability of exit.
  - Effects are statistically significant and economically meaningful (predicted magnitudes reported in paper tables).
  - The positive impact of funding on the likelihood of a successful exit appears to be higher in cultures that tolerate risk taking (cultural indices examined include Uncertainty Avoidance Index and Power Distance Index).
- Summary statistics (Table 1) — global sample (N = 4,422) and Japan subsample (N = 307):
  - Global sample:
    - Well-Funded mean 0.50; Std.Dev 0.50; Min 0.00; 25th 0.00; Median 1.00; 75th 1.00; Max 1.00; N 4422.
    - Total Patent Documents, log: mean 2.08; Std.Dev 1.08; Min 0.00; 25th 1.39; Median 1.95; 75th 2.83; Max 5.69; N 4422.
    - Total Number of Employees, log: mean 3.15; Std.Dev 1.11; Min 0.00; 25th 2.30; Median 3.14; 75th 3.87; Max 6.80; N 4422.
    - Age, in years, log: mean 1.72; Std.Dev 0.34; Min 0.00; 25th 1.61; Median 1.79; 75th 1.95; Max 2.08; N 4422.
    - Active Number of Investors: mean 9.83; Std.Dev 8.89; Min 1.00; 25th 4.00; Median 8.00; 75th 13.00; Max 1814; N 4422.
    - Years since first financing round: mean 4.80; Std.Dev 1.83; Min 0.00; 25th 3.00; Median 5.00; 75th 6.00; Max 8.00; N 4422.
    - Years since last financing round: mean 0.87; Std.Dev 1.10; Min 0.00; 25th 0.00; Median 1.00; 75th 1.00; Max 5.00; N 4422.
    - Uncertainty Avoidance Index: mean 0.55; Std.Dev 0.18; Min 0.29; 25th 0.46; Median 0.46; 75th 0.48; Max 0.92; N 3628.
    - Power Distance Index: mean 0.45; Std.Dev 0.10; Min 0.31; 25th 0.40; Median 0.40; 75th 0.40; Max 0.77; N 3628.
  - Japan subsample:
    - Well-Funded mean 0.51; Std.Dev 0.50; Min 0.00; 25th 0.00; Median 1.00; 75th 1.00; Max 1.00; N 307.
    - Total Patent Documents, log: mean 1.74; Std.Dev 1.27; Min 0.00; 25th 0.69; Median 1.79; 75th 2.64; Max 4.89; N 307.
    - Total Number of Employees, log: mean 2.89; Std.Dev 1.08; Min 0.69; 25th 2.08; Median 2.71; 75th 3.64; Max 6.21; N 307.
    - Age, in years, log: mean 1.72; Std.Dev 0.34; Min 0.69; 25th 1.61; Median 1.79; 75th 1.95; Max 2.08; N 307.
    - Active Number of Investors: mean 9.96; Std.Dev 6.94; Min 1.00; 25th 5.00; Median 8.00; 75th 13.00; Max 43.00; N 307.
    - Years since first financing round: mean 4.64; Std.Dev 1.76; Min 0.00; 25th 3.00; Median 5.00; 75th 6.00; Max 8.00; N 307.
    - Years since last financing round: mean 1.10; Std.Dev 1.12; Min 0.00; 25th 0.00; Median 1.00; 75th 2.00; Max 5.00; N 307.

### V. Policy-relevant implications (as reported in the chapter)
- Strengthen entrepreneurship and firm dynamism:
  - Higher entrepreneurship (share of self-employed who are employers) and more dynamic entry and exit are associated with higher capital investment and firm valuations at the country level.
  - Policies that reduce barriers to entry/exit and encourage entrepreneurship could support capital allocation and firm valuation.
- Support later-stage funding and diversify exit channels:
  - Japan’s startup funding is concentrated in seed and early-stage VC; increasing later-stage funding and diversifying exit options (IPO, M&A) may help scale startups and create more unicorns.
  - The Japan Investment Corporation’s 200-billion-yen venture growth fund (2023) targets later-stage support to create unicorns and expand beyond deep tech and life sciences.
- Leverage corporate VCs while expanding investor base:
  - Corporate VCs play a large role (at least half of VC deals 2015–2022; peak 62 percent in 2020). While CVCs provide capital, broadening the investor base (domestic and international) could improve funding resilience and exit pathways.
- Recognize cultural factors in exit outcomes:
  - The impact of funding on exit probabilities is larger in cultures that tolerate risk taking; cultural and institutional reforms that reduce risk aversion may increase the payoff of funding in terms of successful exits.

*IMF Working Paper chapter: “6. Global: Does Risk Culture Matter for Startup Exit?” from Riding Unicorns: Startups and Venture Capital in Japan (chapter content as provided).*

### 1.5 times more employees and exit with 43 percentage points higher probability, compared to non-

### Riding Unicorns: Startups and Venture Capital in Japan

### Availability of funding — measured impacts
- Well-funded startups have 1.5 times more employees and exit with 43 percentage points higher probability, compared to non-well-funded startups.
- Availability of funding is associated with a larger number of patents for startups (impact on patents appears larger and more precisely estimated in the Japan subsample).
- For the Japan sample, the impact of availability of funding on the number of patents seems larger and more precisely estimated, while the estimated impact on other outcome variables does not seem to be materially different from the global sample.

### Endogeneity and auxiliary evidence
- The results confirm potential endogeneity of well-fundedness: the p-value of the null hypothesis of uncorrelated error terms in the main and auxiliary regressions is strongly rejected in the majority of specifications (the p-value of the 휒ଶ statistic reported in the last row of the upper panel of Table 4).
- The number of active investors is statistically significant in predicting whether or not a startup is well-funded (lower panel of Table 4).
- Identification strategy: endogenous treatment model estimates of equation (2) are used (global sample and Japan sample reported separately). Each specification includes country-sector fixed effects; standard errors are clustered at the country level.

### Role of risk-taking culture
- The empirical model is augmented with country-level proxies for cultural differences from Hofstede (2013): (i) uncertainty avoidance; and (ii) power distance.
- Hypothesis: a greater positive impact of availability of funding on startup performance in countries that reward risk-taking behavior.
- Results:
  - Predicted impact of availability of funding on IPO exit is higher in countries with less uncertainty avoidance.
  - Predicted impact of availability of funding on M&A exit is higher in countries with less power distance.
- The table presenting these estimates is Table 6 (endogenous treatment model estimates of equation (4)); country-sector fixed effects included; standard errors clustered at the country level.

### Caveats and data limitations
- Data are cross-sectional; the authors were not able to track startups’ performance following funding rounds.
- Vintage and survivor bias: the analysis studies a particular vintage of the data, i.e., end-2023, that includes surviving startups. Data for firms that had exited before end-2023 is not available, hence survivor bias remains (direction of bias not evident).
- The study focuses on the availability of funding and does not explore the role of cost, conditionality, or particular types of investors (for example, differences between investor types), due to data limitations.
- Identification is mainly within country-industry, which sharpens identification but overlooks potential variation across countries and/or industries.
- The analysis does not resolve whether private equity funding complements or substitutes other types of financing (for example, debt financing).
- The authors note that exploiting country-industry variation in the identification strategy partly mitigates potential bias from omitting other funding sources.

- Exact equation text as presented in the source:
  - 푂푢푡푐표푚푒
௜
ൌ훽푋
௜
൅훿푊푒푙푙퐹푢푛푑푒푑
௜
푥푌
௖௢௨௡௧௥௬
൅휑
௖௢௨௡௧௥௬,௦௘௖௧௢௥
൅ 휖
௜
,                ሺ4ሻ

### Policy implications and conclusions
- Better access to equity funding is crucial for startups to grow, innovate, and exit successfully; angel or venture capital investment provides private equity financing and value-added services (for example, operational and market insights).
- Labor market reforms can support entrepreneurship and innovation:
  - A gradual shift away from the lifelong employment system in Japan could encourage startup creation and offer second chances after failure.
  - Reducing labor-market dualism, encouraging merit-based promotions, and facilitating more job mobility can encourage entrepreneurship and are associated with higher capital investment and firm valuations at the country level.
- Increasing firm dynamism can support startups and innovation:
  - Dynamic firm entry and exit and reduced personal liabilities can encourage entrepreneurship, innovation, and more efficient resource allocation.
  - Gradual reduction of zombie firms could improve allocation of capital and labor to more productive ventures, boosting productivity and growth.

*IMF Working Paper: Riding Unicorns: Startups and Venture Capital in Japan — Working Paper No. WP/2024/246*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2024/english/wpiea2024246-print-pdf.pdf_