The DPI Wall
Why the 2020–2021 vintage funds were always going to hit this ceiling — and which part of the problem is actually solvable
A $75M fund raised in 2021 pays roughly $1.5M annually in management fees during its investment period. Over five years, that’s $7.5M — paid from committed capital before a single cheque goes to a portfolio company. The fund doesn’t invest $75M. It invests $67.5M.
By year five, the investment period ends and the fee steps down to 1.5% on invested capital: roughly $1M annually. Out of that comes salaries, travel, legal, LP reporting, and whatever operational capacity the GP needs to run the firm. There’s nothing left for a structured exit programme. And exits are ether point of the VC game.
The 2020–2021 vintage funds didn’t just face a market correction. They were capitalised for a world that no longer exists — and their fee structures are compressing at exactly the moment exit execution demands are peaking. The funds that raised on $200M+ rounds and 15x paper multiples deployed into a liquidity environment that has since inverted. Five years later, the structural mismatch is arriving as a DPI problem. It was always going to.
What LPs Are Actually Watching
In conversations with fund managers over the past several months, the pattern is consistent: LP re-up appetite has softened, follow-on fundraise timelines are extending, and the managers getting traction are the ones who can point to realised returns — not projected ones.
What’s moved is the evidential standard. DPI — the ratio of cash actually returned to LPs against capital they paid in, not paper multiples, not projected exits, cash — is now the primary signal LPs use to assess whether a GP can actually execute. IRR projections are easy to produce. They’re increasingly easy to ignore.
A fund at year five with 0.3x DPI is having a categorically different conversation with LPs than a fund with 0.7x DPI, even if unrealised NAV looks identical on paper. That 0.4x spread is the difference between a GP who has demonstrated capital velocity and one who is asking LPs to extend their faith on the basis of a model.
For funds facing Fund II fundraises in 2026–2027, that gap is existential. You can’t raise a second fund on a story. You raise it on evidence that the first one is returning capital.
Three Walls
DPI velocity in African VC isn’t constrained by one variable. It’s constrained by three, each one compounding the others.
Wall One: Exit route concentration.
African VC exits cluster heavily around trade sales — credible estimates put strategic acquisitions at 60–70% of realised exits across the continent. IPO infrastructure remains thin outside a narrow set of listings on NSE, JSE, and NGX that rarely suit venture-backed tech companies. Secondary markets, where LP positions or portfolio equity changes hands outside a full exit, are nascent at best.
When the primary route slows, there’s no alternative route absorbing the volume. (Secondary, in every sense.)
The structural consequence: any macro event that delays or reprices trade sale activity hits African VC DPI disproportionately hard, because the fallback mechanisms that exist in more mature markets simply aren’t available at scale here.
Wall Two: Execution infrastructure deficit.
Secondary transactions that should take two weeks take six months. Not because buyers don’t exist. Because the operational plumbing to close them efficiently doesn’t.
This is the wall a lot of GPs don’t see clearly — it looks like a legal problem, a buyer problem, or a company problem depending on which deal just stalled. It’s none of those. It’s a production-line problem being treated as a dealmaking problem, costing funds time and money they can’t afford at year five.
The mid-week follow-up goes deeper on this.
Wall Three: Global capital transmission contraction.
The eurodollar tightening cycle that began in 2022 compressed risk appetite for frontier-market assets in ways still working through the system. Strategic acquirers — the natural buyers in trade sale exits — are paying lower multiples and moving slower through deal processes. Their own capital costs went up. Their boards got more conservative. Their Africa appetite, never deep to begin with, narrowed further.
The 2020–2021 funds deployed into peak global risk appetite. Valuations were high, round sizes were large, growth projections were optimistic. They’ll how have to engineer exits — or try to — in trough risk appetite. The multiple compression isn’t a portfolio quality story. It’s a macro timing story. But the LP doesn’t experience it that way. They experience it as underperformance.
The Arithmetic of Survival
Back to that $75M fund. $67.5M deployed across roughly 67 portfolio companies after management fees. To hit 0.7x DPI by year six, the fund needs to distribute $52.5M — calculated on the full $75M paid-in capital. Management fees are a cost of running the fund, not a return of capital. LPs paid in $75M. $7.5M of that funded operations. The remaining $67.5M went to work. But LPs measure their return against every dollar they wired — which means the fee load doesn’t reduce the return hurdle. It basically raises the effective return required on every dollar that actually reached a portfolio company.
Assume the fund has already returned $20M through earlier trade sales and partial secondaries — no single outsized exit required, just the ordinary progression of a maturing portfolio, an acqui-hire here, a strategic buy-out by a regional player there. That leaves $32.5M the fund still needs to deliver — inside a window that’s already narrowing.
Across 67 portfolio companies, assuming a 60% loss or write-down rate — conservative for this vintage given deployment valuations — realised value is possible in roughly 27 companies. By year five, entry stakes of 10–15% have been diluted through follow-on rounds, ESOP expansions, and bridge conversions to approximately 5–7% per position. Secondary buyers won’t pay last-round NAV. In African VC, thin buyer pools, jurisdiction risk, and absent auction dynamics produce discounts of 25–35% to last-round valuation. A buyer acquiring a minority stake in a private African company with no public price signal, limited resale options, and ROFR mechanics to navigate isn’t paying full price. The discount is the cost of all that friction, crystallised into a single negotiated percentage.
The proceeds math on a representative transaction: $40M company, 30% secondary discount, 6% diluted stake, 60% position sold. Effective valuation: $28M. Proceeds: $28M × 6% × 60% = approximately $1M — ranging from $500k at the lower bound to $1.8M at the upper, depending on company size, discount negotiated, and position size sold.
At $1M per transaction, closing $32.5M through secondaries alone requires 33 transactions. No fund executes 33 secondary transactions in 12–18 months. SAFE positions — the simplest to transfer, fastest to close, and cleanest on jurisdiction since they don’t carry the shareholder rights that trigger ROFR or pre-emption protocols the way equity does — are the natural entry point for building execution muscle. But their notional size means they contribute to operational credibility more than DPI.
The DPI engine is trade sales: full position exits, no secondary discount, negotiated multiples against strategic value. A realistic 12–18 month programme combines 3–5 trade sale processes running in parallel with 8–12 secondary transactions on smaller positions. Under conservative assumptions, that programme returns $15M–$22M — closing roughly 60% of the $32.5M gap.
The remainder requires structural tools. A continuation vehicle — which transfers selected portfolio assets into a new fund structure, giving existing LPs the option to cash out while new capital comes in to hold the position longer — is one route. LP negotiation on timeline extension is another. A single larger trade sale that shifts the trajectory is a third. They’re the actual toolkit — and the GP who maps this arithmetic before the programme starts can sequence the right tools in the right order, approach LPs on extensions from a position of transparency rather than necessity, and preserve the relationship capital that Fund II depends on.
What’s Actually Solvable
The market problem — thin exit routes, compressed multiples, scarce strategic acquirers — has no thirty-day fix. The structural problem — continuation vehicles, LP timeline negotiation — requires relationship capital and legal architecture that takes months to put in place, not weeks.
The operations problem is the exception.
At current execution velocity, 3 to 6 months per secondary transaction, even the secondary component of a realistic exit programme stretches across years when run sequentially. At optimised execution — 2 to 4 weeks per transaction — the same programme takes months. That difference doesn’t change how many exits the fund needs. It determines how many are achievable inside the window that still matters for Fund II.
The funds that navigate this vintage cycle won’t necessarily have the most attractively valued portfolios or the most patient LPs. They’ll be the ones that treated exit execution as an operations problem early enough to build the infrastructure while management fees still covered the cost — and while there was still time to run the process before the LP clock made every conversation harder.
That window is closing for most 2021 vintage funds. For 2022 and 2023 vintage funds, it’s still open. Whether the lesson travels fast enough is a different question.
African VC has spent the last decade building deal flow infrastructure. The exit infrastructure gap is where the next decade’s performance divergence happens.
If your fund is in its fourth or fifth year and exit execution is slower than your LP update cadence, the market isn’t the whole explanation. The plumbing is part of it — and unlike the market, the plumbing is fixable.
I work with African GPs on exactly this.