CVC Survival Effect
Whether corporate backing lowers failure, measured rather than assumed
European startups with corporate backing go bankrupt at 1.24% against 2.58% without, a raw comparison on one secondary source; LATAM has no equivalent, so build the matched study.
The rationale
The European figure is doing work it cannot support. Hello Tomorrow and the EIC report a 1.24% bankruptcy rate for CVC-backed startups against 2.58% for those without, roughly half, on a single secondary source and as a raw comparison. Raw rates cannot separate the effect of corporate money from the selection that precedes it, because corporate investors pick companies that already look like survivors. LATAM has no version of this comparison at all, in a region where only about 40% of corporates run any venture activity against roughly 90% in the US.
The study is a matched design rather than a rate comparison. Pair each CVC-backed venture with non-CVC ventures of the same vertical, stage, vintage and prior funding, then estimate hazard ratios for failure and for time to next round, with procurement conversion as a second outcome. It is the standing instrument the CVC Bridge observatory is meant to house, so collection becomes a by-product of matchmaking rather than a separate research project, and every pitch day adds cohort records to the panel.
The honest constraint is sample size. Dealroom counted US$36.5m of CVC into LATAM deep tech in 2024, about 70% below the 2022 peak of US$119m, so a single vintage will not support a hazard model. The design has to pool cohorts across 2018 to 2024, accept wide intervals and publish them. A study that reports a confident ratio off a hundred companies would be worse than no study, because the ratio would be quoted for a decade after the caveats were lost.
The evidence
What it unlocks
Evidence on whether corporate backing actually lowers failure in LATAM, matched rather than raw, so founders and CVCs price the relationship honestly.
It has been done before
Europe, the EIC corporate-startup collaboration program
The European figure exists because the engagements were registered first. Hello Tomorrow and the European Innovation Council report more than 1,500 startup-corporate engagements with over 120 corporations since 2017, and it is that recorded population, published in July 2025, from which the 1.24% against 2.58% bankruptcy comparison could be computed at all. The measurement followed the infrastructure, in that order.
First moves
- 0–6 months
Fix the cohort record at intake
Add eight mandatory fields to CVC Bridge pitch-day intake, vertical, stage, vintage, prior funding, corporate investor, pilot or procurement status, round date and company status, so every future cohort is analysable without retrospective data collection.
CVC Bridge / coalition secretariat + regional accelerators - 6–18 months
Reconstruct the 2018 to 2024 back-book
Assemble the historic cohort from Dealroom, LAVCA and registry records, match each CVC-backed venture to non-CVC controls on vertical, stage, vintage and prior funding, and publish the matched dataset and the matching code before publishing any result.
LADP research team + LAVCA + university research partners - 18–36 months
Publish hazard ratios with intervals and a replication file
Report failure and time-to-next-round hazard ratios by vertical with confidence intervals, name the cells too thin to report, and release the full replication package alongside the finding.
CVC × Deep Tech Observatory + IDB Lab
How we would know it worked
- Share of CVC Bridge pitch-day participants with a complete eight-field intake record, tracked quarterly by the observatory, above 90%.
- Number of matched pairs in the panel, published per release, reaching 300 or more across the 2018 to 2024 cohorts.
- Median time from first corporate contact to signed term sheet, tracked quarterly by the observatory, falling below six months.
- Verticals reportable at or above the pre-registered minimum cell size, counted per release, rising from zero.
What could go wrong
Selection is mistaken for effect
Report a ratio without the matching and it will be quoted as proof that corporate money saves companies, when the likelier reading is that corporate investors pick companies that were going to survive anyway. Contain it by publishing the matched dataset and matching code alongside every headline, and by stating on the same surface that the design measures association under matching, not causation.
The population is too small to answer the question
At US$36.5m a year the LATAM deep tech CVC universe may not produce enough failures to estimate a ratio with useful precision. Contain it by pre-registering the minimum cell size, pooling cohorts across years, and being willing to publish the finding that the region cannot yet answer the question.
Who acts
Rests on
09 Corporate Venture Capital A neutral bridge to corporate capital, and a way to measure it
14 Research & Visibility An open data commons and a funding catalogue
The data behind it
Corporate venture capital into the region’s deep tech peaked at USD 119M in 2022, then cooled with the wider funding market. The base is set for the CVC Bridge the report proposes.
As published in the report, September 2025.
SourceDealroom.co, Deep Tech Overview: Latin America
Methodology: Corporate venture capital into Latin American deep tech, in millions of US dollars, from Dealroom as cited in the report. 2021 is reported as not available and is left empty, because a zero there would claim a year with no corporate venture capital at all. The series peaked at $119m in 2022 and fell about 70% by 2024, which still leaves it around four times the 2020 base. It has not been extended past 2024.
| Year | Value |
|---|---|
| 2020 | $9.0M |
| 2022 | $119.0M |
| 2023 | $73.9M |
| 2024 | $36.5M |
Global corporate venturing grew 2.8× and corporate–deep tech collaborations 4.2×; 71% of firms expect deep tech to weigh more in their portfolios.
As published in the report, September 2025.
Methodology: Growth multiples for corporate venturing from IESE's 2021 open innovation study, measured over 2017 to 2021. These are multiples across that whole period, not a year-by-year series, and the 71% figure is the share of surveyed firms expecting deep tech to weigh more in their portfolios over the following five years, which is why the unit is recorded as mixed. Context not in these numbers: Latin American corporate venture activity roughly doubled between 2020 and 2023, and AI is now its leading theme at about 41% of funds.
| Category | Value |
|---|---|
| Overall corporate venturing growth | 2.8× |
| Corporate–deep tech collaborations | 4.2× |
| Of firms expect deep tech to grow in their portfolio | 71% |