#193 — Cold Take: Canada's Deep Tech "valley of death" is still a problem

Why it matters: BDC Capital's $200M Deep Tech Fund closed in 2024, just four years into a planned 12-to-16-year life. This is a warning for any Canadian founder counting on domestic capital to survive the gap between research and revenue.§

The warning signs were there first

Fund closures rarely happen without notice. Personnel exits from a fund's leadership team are almost never a good sign for founders watching from outside, and BDC's deep tech team had already lost half its staff by the time the closure was announced. There is no announced successor managing the deep tech thesis, even though BDC has signaled a second fund may come, so founders should read this as an unsettled mandate rather than a secure one.§

What BDC actually said

BDC's public statement described the closure as a normal part of reorganizing "to better serve their clients," language that reads as a way to close the fund without a public failure analysis. BDC has reportedly not deprioritized deep tech and is reportedly working on the thesis and mandate for a second fund, so founders and policymakers are being asked to trust round two without any accounting of what went wrong the first time.§

Separate the fund's failure from BDC's failure

The closure looks less like a verdict on deep tech investing and more like a reaction to BDC's own organizational losses, reported around $1B over the two years leading into the announcement. The fund itself may not have underperformed on its own terms. Founders should weigh this news for what caused it, BDC's balance sheet stress, rather than treating it as a market signal that deep tech theses don't work.§

Timelines: two separate problems

Traditional VC runs a 10-year cycle with liquidity pressure starting around year five to seven, a model suited to biotech and B2B SaaS but poorly matched to quantum computing, fusion, or AGI-scale timelines. BDC's fund had two distinct timeline failures worth separating:§

• Structural: a 16-year maximum lifespan (12 years plus a 4-year extension) still puts a hard ceiling on genuinely disruptive R&D, and quantum computing has been "close" for decades
• Operational: the fund was shut down at year four, so no one will ever know whether the 16-year structure would have worked, because the parent organization's instability ended the experiment early

If BDC launches a second fund, founders and policymakers should push for limits on the parent org's ability to defund the fund in response to external pressure such as elections, trade disputes, or internal reorgs. Deep tech capital needs to survive full market cycles and political change untouched.§

Accountability should cut both ways

In most limited-partner structures, an LP who fails to fund a capital call forfeits their entire stake. BDC functions as the effective LP behind this fund, and its early exit was driven by unrelated losses elsewhere in the organization. The same consequence logic should apply here: reneging on a long-horizon commitment should carry real cost, so reactionary budget shifts aren't cost-free for the institution making them.§

The check size problem ran deeper than "too big"

BDC's $5M minimum check undersells the actual mismatch. Most deep tech companies need closer to $1M to survive pre-seed and cross the valley of death, meaning BDC's floor was off by roughly an order of magnitude rather than simply being on the high end.§

• Most pre-revenue deep tech companies never fit BDC's thesis because they were raising rounds too small to register
• Combined with thin overall deep tech investment in Canada, this left companies to survive the earliest and hardest stretch of the valley of death entirely on their own before BDC could engage
• Passed-on opportunities never appear in a fund's public performance metrics, so the true cost of this mismatch stays invisible in any official report and is likely larger than published figures suggest

The anchor investor problem

BDC's absence risks a chilling effect on upstream investors who treated it as the presumed next check in the chain. If earlier-stage funds underwrote risk partly because they expected BDC to follow on, its exit raises the bar for those earlier checks too, independent of anything specific to the startups involved.§

By the numbers

• $200M: Size of the fund BDC launched in 2020, wound down in 2024 after four years of operation
• $1B: BDC's broader reported losses over the two years leading into the closure
• 12 to 16 years: Fund's intended lifespan including the possible extension, against the 4 years it actually ran
• Roughly $1M: What deep tech companies typically need at pre-seed, against BDC's $5M floor
• 45%: Drop in Canadian pre-seed and seed deals in 2024, while later-stage funding grew 20% over the same period
• 60%: Share of capital Norway's Innovation Fund allocates to early-stage bets, frequently cited as the model to copy

Who's still writing early checks

Fund/Firm	Focus	Why it matters for founders
Next Canada (Montreal)	Early-stage, tech-heavy portfolios	Active patient capital still writing early checks
Quebec's CQR	Early-stage, tech-heavy portfolios	Same profile as Next Canada
ProteinQure (Vancouver)	AI-driven drug discovery	Evidence that patient capital plus a strong team can still deliver returns
Horizon Fusion (Edmonton)	Fusion energy	Same proof point for foundational, long-horizon bets

The founder playbook

• Watch for personnel churn at any fund you're targeting, especially leadership departures with no named successor, and treat it as a leading indicator rather than noise
• Read closure statements skeptically; language about reorganizing to serve clients usually means no failure analysis is coming
• Weigh the cause of a fund's exit before updating your own outlook; a closure tied to a parent's unrelated losses does not mean your sector became less investable
• Size your ask to reality; if you're deep tech pre-seed, you likely need closer to $1M than $5M, and high-minimum funds may never see your round
• If BDC or any institution announces a second fund, ask what changed structurally around lifespan flexibility, check size, and insulation from parent-org politics before treating the new capital as reliable
• Account for anchor investor risk; if a fund you're raising from was counting on BDC-style follow-on, its exit may quietly raise the bar with your earlier-stage investors as well