Steven Sandler moderated a Global Alts New York 2026 panel on venture capital with Leif Danielsen of Acequia Capital, Heather Hartnett of Human Ventures, Todd Breeden of Jefferies, and Barr Even of Rebalance Capital. The session covered the bifurcation between established and emerging managers, what LPs should actually underwrite in a fund one, and how DPI expectations shift as the exit window reopens.

Venture capital DPI is the metric that separates an allocation that funds an institution’s spending policy from one that only looks good in a quarterly report, and this panel returned to it from four different seats. Sandler opened on bifurcation. The venture market splits between large multi-stage firms and emerging managers, and both capture allocator attention for different reasons. Danielsen made the early-stage case first. He told the room that roughly 72% of venture capital deployed over the past year went into five positions, which means concentration at the top of the market is extreme even by venture standards. What an LP buys in venture is innovation, and innovation starts in the ecosystem long before the mega-rounds. Emerging managers are better at spotting that talent early, and at $100 million to $200 million in fund size, a single outcome still moves the return.

Why venture capital DPI favors emerging managers

Breeden framed the return profile plainly. A major multi-stage fund gets an LP to roughly 2x or 2.5x, and rarely much beyond it. Emerging manager outcomes are far more dispersed, and the reason to accept that dispersion is alignment. A GP raising a fund one or a fund two is not getting rich on management fees. In many cases the firm is breaking even or losing money on them, which means the entire economic outcome depends on crystallizing carry. As Breeden put it, that is the seat an LP wants to be in. When the fee base is large enough to comfortably fund the partnership regardless of performance, the incentive to work for the last turn of multiple weakens.

That alignment translates directly into distributions. If DPI is the North Star metric, Breeden argued, an LP should be allocating to emerging managers, because they are the ones structurally willing to take money off the table as companies scale. A firm without a marquee brand selling down a position along the way is a non-event. The multi-stage firm that led the B or the C and will have its name on the cover of the S-1 usually cannot do the same thing quietly. Long hold periods in venture are a feature rather than a bug, but LPs who need distributions inside a reasonable window should go to the managers whose business depends on getting back to market with the next fund.

Even added the exit-route argument. Early-stage funds are far more likely to realize through M&A than through a listing, and an acquisition at $500 million or $1 billion does not require a receptive IPO window or a lead investor willing to sell. That shortens the DPI cycle independently of what public markets are doing.

What LPs should underwrite in an emerging manager

Even suggested underwriting the GP the way the GP underwrites a founder. Look for determination, for someone who has had a go before and possibly failed, and for genuine depth in the sector they are investing in, because without that depth there is no right to win against better-resourced competitors. From there the job breaks into four distinct skills: sourcing through a differentiated network, picking, portfolio construction, and managing the position through to realization. Almost nobody is best in class at all four, so the diligence question is which two the manager actually owns.

He also pushed back on the idea that emerging managers should be cheap. LPs should demand a risk premium for the business risk they are taking, and the compensating math is real. Writing $5 million into a Series A at a $50 million valuation produces a very different outcome profile than the same check into a mega-fund, though it comes with the constraint that far less capital can be deployed into any single name.

Hartnett reframed the category entirely. She has written about retiring the term emerging manager in favor of alpha generator, on the view that this is close to a separate asset class. These are investors underwriting the next winners a decade before the outcome is visible, insulated from public market sentiment, and the value they capture is informational asymmetry as much as it is access. With something like 4,500 funds in market, she said, LPs need allies with boots on the ground rather than a checkbox diligence form, because no form captures who led a round, why, and what the founder’s backstory was.

Breeden described how Jefferies handles that volume. The team tries to isolate each manager’s superpower, the one thing that manager does better than anyone else it has met, then builds a stable of managers whose powers are complementary. Danielsen offered the question he thinks matters most on both sides: what is our reason to exist, and why do founders choose us over any other group. An emerging manager who cannot answer that has not earned the allocation, and an allocator who cannot answer it about a manager has not finished the work.

Generalist or specialist in venture capital

Conventional wisdom says specialize. The Jefferies data says something narrower. Breeden reported that specialists outperform in biotech and cyber, and that outside those two areas there is no meaningful dispersion in returns between specialist and generalist funds, whether the category is fintech, commerce tech, or consumer. What gives the team confidence in that read is behavioral: most specialist funds in non-bio, non-cyber categories become generalists by fund four anyway, because a larger fund forces a wider aperture.

Even, who runs a sector-focused strategy, made the case that the distinction is softer than it looks. Financial services is roughly 20% of the US economy, so a fintech mandate already spans payments, payroll, insurance, retirement, lending, and savings. Rebalance supports that breadth with venture partners and advisors who are specialists inside each segment, since each one carries its own regulatory map and incumbent set. By fund three, he added, pattern recognition and founder-driven deal flow become cross-sector advantages, which is exactly when sector specialization starts to matter less than track record and network.

What the exit window means for venture capital DPI

On the IPO environment, the panel agreed the market is in uncharted territory. Danielsen argued that early-stage managers sit below the weather. A hurricane can be running above while the discipline stays the same across a ten-year fund life. He noted that his 2021 vintage has already returned just under 20% of the fund with six or seven positions that could still be fund returners, and that the firm starts looking to take capital off the table once a company clears roughly $1 billion in valuation. He also flagged a structural change: companies are being built to scale on far less capital than the traditional A-through-D sequence required, which changes the liquidity landscape for everyone underneath the mega-rounds.

Hartnett expects the listings to help the venture ecosystem even if the public market reaction is bumpier than the current narrative suggests. Liquidity flows downstream, and newly minted wealth from these outcomes tends to arrive as LP capital in the next cycle. Her read on today’s fundraising environment is that it is a buyer’s market for LPs, and that it is likely to shift.

Asked where the underappreciated opportunity sits, the panel avoided the obvious answer. Even pointed to the labor market, where AI disruption of white collar work forces large-scale reskilling and upskilling, and where his thesis on 80 million low and middle income US households with roughly $4 trillion in aggregate purchasing power intersects with workforce technology. Breeden made the contrarian consumer case, arguing that once affordability is addressed, a suite of commerce tools built for agentic purchasing gets unlocked. Hartnett pointed to longevity and the care economy, noting the approaching point at which the US will have more people over 65 than under 18 without the infrastructure to support it. Danielsen came back to the physical world, arguing that the buildout required to support AI is underestimated by orders of magnitude and will run for 30 years or more, combining traditional engineering with trade skills and robotics.

The practical takeaway for allocators is that venture capital DPI is a manager selection question before it is a market timing question. Size emerging manager exposure as a dedicated sleeve with its own risk budget, demand a premium for the business risk in a fund one, diligence the operational side as seriously as the investment thesis, and expect realizations through M&A rather than only through a listing.