COVID-19 Exposed a Startup Financing Gap: The Most Vulnerable Firms Never Reached the Banks

Access to finance is widely regarded as a determining factor in whether new businesses survive their earliest years. Entrepreneurs need capital to sustain operations, absorb losses and navigate uncertain markets, while banks and investors are often credited with improving business resilience. However, the relationship between financing and survival contains a fundamental ambiguity: businesses may survive because they receive funding, or they may qualify for funding precisely because they have already demonstrated their ability to survive.

A study titled "Financing Structure, the COVID-19 Shock, and Startup Survival: Evidence from a Peripheral Entrepreneurial Ecosystem in Vietnam," published in the Journal of Risk and Financial Management, challenges conventional interpretations of this relationship. Conducted by An Vu Van Nguyen and Ha Hong Nguyen of Tra Vinh University, the research examines 301 startups established during 2021–2022 across five provinces of Vietnam's Mekong Delta, tracking their survival through mid-2025.

The findings reveal that external financing initially appears strongly associated with business survival, but much of this relationship weakens when researchers account for the timing of business failures. The analysis also identifies exposure to COVID-19 disruption as a major correlate of market exit, raising questions about whether existing financial systems reach vulnerable enterprises during their most critical operating period.

Pandemic-born startups expose the limits of conventional business financing

Starting a business involves considerable uncertainty even under stable economic conditions. Entrepreneurs must establish customers, develop operational capabilities and generate revenue before financial reserves become exhausted. The COVID-19 pandemic intensified these challenges, creating an unusual generation of businesses founded during widespread economic disruption rather than before it.

The researchers focus on this generation because businesses established during the pandemic faced different circumstances from existing firms. Established enterprises could potentially rely on previous customers, accumulated assets and banking relationships, while newly founded ventures had to develop these resources amid disrupted demand, restricted mobility and uncertain financing conditions.

Vietnam's Mekong Delta provides a distinctive setting for examining these challenges. Its economy relies heavily on agriculture, aquaculture and processing, while its entrepreneurial ecosystem operates outside the country's principal financial centres. Venture capital is relatively limited, and formal lending represents an important source of external business finance.

The study draws on founder interviews conducted during July and August 2025. Researchers examined financing from personal equity, debt, venture capital, angel investors, government assistance and family-and-friends networks, alongside business characteristics and reported exposure to pandemic disruption.

Of the 301 startups investigated, 114 had exited the market, representing 37.9% of the sample. Among businesses that closed, the median duration from founding to exit was 24.5 months, illustrating how business survival remained uncertain during the first years of operation.

Pandemic exposure produced the sharpest differences in observed outcomes. Among ventures whose founders reported direct disruption from COVID-19 and associated restrictions, 67.1% exited during the observation period. Only 4.3% of startups reporting no direct exposure experienced market exit.

Statistical analysis confirmed a strong relationship between reported pandemic exposure and business failure. In a streamlined model, exposed startups faced substantially higher exit hazards, although estimates became less precise when the researchers introduced a larger set of control variables.

The findings establish pandemic exposure as an important predictor within the sample, but not as definitive proof of causation. Exposure was reported retrospectively by founders, leaving uncertainty about how disruption interacted with underlying business vulnerabilities, financial conditions and other influences on survival.

External funding predicts survival, but the evidence contains a hidden selection problem

The researchers initially found substantial differences between startups according to their financing arrangements. Businesses receiving debt, angel investment and funding from family and friends generally demonstrated lower exit hazards than ventures with less access to these sources.

In a model controlling for founder and organizational characteristics, a one-log-unit increase in debt financing was associated with a 21.6% lower exit hazard. Angel capital was associated with a 19.7% reduction, while family-and-friends funding showed an even stronger negative relationship with business closure.

Venture capital displayed a weaker association in some specifications. By contrast, government financial assistance was not significantly associated with survival, despite reaching a proportion of the surveyed businesses. Greater founder-equity reliance initially corresponded with a higher exit hazard, although subsequent analysis raised doubts about the stability of that finding.

These patterns are consistent with several established explanations of entrepreneurial financing. External funds may provide working capital, extend operating capacity and allow businesses to absorb temporary losses. Lenders may impose financial discipline, while experienced investors can contribute knowledge, oversight and commercial connections.

However, financing also involves selection. Banks generally evaluate repayment capacity before extending credit, while investors assess business prospects before committing funds. Startups obtaining external finance may therefore possess characteristics that already distinguish them from businesses facing imminent closure.

The distinction becomes especially important when funding is measured across a venture's operating history. Businesses surviving for longer periods have more opportunities to attract financing, whereas early failures may close before developing the financial relationships necessary to borrow or secure investment.

The researchers explicitly examined this possibility rather than interpreting lower exit hazards as evidence that financing independently prevented failure. Their results show that the apparent protection associated with external capital becomes considerably less convincing once the timing of business survival enters the analysis.

For financial institutions and policymakers, the findings challenge the assumption that observed differences between funded and unfunded enterprises demonstrate the effectiveness of financing itself. Access to capital may signal prior resilience as much as contribute to future resilience.

The first two years reveal why the borrowing advantage may be misleading

The study's most consequential finding emerged from examining which startups obtained financing before the earliest business closures. None of the 36 ventures that exited within 18 months of founding had borrowed. The earliest exit among debt-financed businesses occurred in the twenty-third month.

This timing creates a serious difficulty for interpreting survival statistics. A startup that closes before it has an opportunity to obtain a loan will appear in the unfunded group, while businesses surviving long enough to establish credit relationships are more likely to be classified as borrowers.

The researchers addressed this problem by repeating their analysis among startups that had already survived for 12, 18 and 24 months. This approach allowed them to examine financing relationships after excluding businesses that had closed during earlier operating periods.

The apparent protective association of debt weakened progressively. Among startups surviving for 24 months, the estimated debt hazard ratio reached 1.076, with a confidence interval including no difference in subsequent exit risk. The relationship was no longer statistically significant.

Angel capital and family-and-friends financing displayed similar patterns. Their initially significant associations with lower exit hazards also disappeared among startups surviving beyond the two-year threshold. Venture capital followed a comparable trajectory, although its association remained significant through the earlier survival periods.

These findings suggest that some apparent financing advantages reflect reverse causality: continued business operation may enable access to capital rather than capital necessarily ensuring continued operation. The analysis cannot completely separate financial benefits from lender screening and differences in business quality, but it demonstrates why ignoring financing timing can produce misleading conclusions.

The study also tested whether COVID-19 exposure changed the relationship between financing sources and survival. Statistical tests found no significant evidence that pandemic exposure altered the associations involving founder equity or debt.

However, the researchers acknowledge limited statistical power because only six ventures without reported pandemic exposure exited. The findings therefore cannot establish that financing works identically during crises and normal conditions.

The distinction between association and causation is central to the research. Strong statistical relationships can remain misleading when the chronology of financing and business survival is inadequately represented.

Startup policy needs to reach vulnerable firms before conventional credit becomes available

The findings raise important questions for governments seeking to strengthen entrepreneurship in emerging economies. Financial support may be least accessible during the period when young businesses face their greatest vulnerability, particularly before they develop sufficient operating histories or financial relationships.

Vietnam introduced pandemic-era banking measures allowing lenders to restructure repayment obligations, reduce certain charges and maintain loan classifications for affected borrowers. Such arrangements could potentially relieve financial pressure on businesses already participating in the formal credit system.

However, the researchers could not determine whether these measures improved survival among the sampled ventures. Most startups experiencing early exits had never borrowed, limiting their ability to benefit directly from policies operating through existing credit relationships.

Direct government assistance also showed no statistically significant survival association in the study's unpenalized models. Financial support reached 16.6% of surveyed startups, frequently through competitions, grants and innovation-related programmes. The researchers note that relatively modest assistance and administrative procedures may constrain effectiveness during periods of urgent financial need.

These findings do not establish that government support was ineffective. The observational research cannot isolate programme effects, determine what would have happened without assistance or establish whether alternative forms of financing would have prevented specific closures.

The evidence nevertheless supports examining how assistance is distributed across the startup lifecycle. Programmes relying on established repayment capacity, formal banking relationships or lengthy approval procedures may struggle to reach firms that have not yet developed the qualifications required for conventional financial support.

The research also has important limitations. Its sample covers startups admitted to a regional support network, potentially excluding less-connected entrepreneurs. Financing information was collected retrospectively without precise transaction dates, and the data did not distinguish voluntary closures from businesses forced out by financial distress.

Future investigation would benefit from dated lending records, administrative information on government assistance and detailed explanations of business exits. Such evidence could help determine whether funding improves survival, whether lenders primarily identify stronger businesses and which interventions reach vulnerable ventures before closure.