For forty years, the famous Powerlaw has structured the logic of venture capital. Some exceptional “unicorn” type outputs were supposed to compensate for a massive failure rate among the financed startups. This statistical asymmetry – where 1 % of financed companies generate 90 % of returns – was sanctuarized by Silicon Valley, erected as a model, taught in business schools, and replicated in all innovation hubs.
But this paradigm now shows its limits. Not for philosophical reasons, but because it no longer produces, structurally, expected yields.
A model based on extreme rarity
The traditional functioning of the VC is based on a portfolio where the majority of startups fail. In fact:
-
- Approximately 70 % Startups financed in pre-SEED or Seed never get up for series A.
- Only 2 to 3 % become “outliers”, that is to say companies with explosive growth, often overvalued, sometimes profitable.
- Less than 1 % reach the status of unicorn (valuation> 1MD $).
It is on these very rare cases that the hopes of multiple X10 to X100 rest allow a fund to restore its commitments to LPS (Limited Partners) with yield.
Problem: the environment in which this logic has prospered – bullish market, abundant capital, interest rate for zero – has disappeared.
The Power Law no longer delivers
Rimmer ventures have been stagnating for more than a decade. Several analyzes, including the databases of Pitchbook and Cambridge Associates, show that theAverage IRR of VC funds (all generations combined) struggles to exceed that of technological public indices, once the costs are taken into account.
Furthermore, the scarcity of significant exits (IPO or Acquisition with high valuation) reduces the effective capacity of the gains. Many “unicorns” created since 2016 have never delivered liquid output.
The performance is therefore increasingly concentrated in a handful of historical funds, leaving the others in chronic underperformance.
A structurally different context
The VC model as it was designed in the 1980s no longer corresponds to the current dynamics:
-
- The industrialization of no-code and AI Allows you to launch a business with little initial capital.
- The logic of growth without profitencouraged by VCS, is increasingly questioned in a post-zirp context.
- Private markets have been hypertrophiedwith extended liquidity cycles and a drastic drop in ompos.
- The profile of the founders evolves : they are more experienced, less inclined to dilute quickly, and oriented towards profitability.
In this context, targeting an aggregate portfolio yield only based on “shots” amounts to ignoring 95 % of the emerging entrepreneurial landscape.
Towards an alternative risk architecture
An alternative model emerges: that of a VC Post-Unitedno longer based on the statistical exception, but on the moderate recurrence and structural viability.
Take a fictitious scenario:
| Model | Portfolio structure (100 startups) | Total return |
|---|---|---|
| Power Law | 1x (10x) + 4x (2x) + 15 Breakeven + 80 to zero | ~ 1.55x |
| Post-United | 10x (5x) + 90x (2x) | ~ 2.3x |
This scenario is based on a hypothesis of high survival rate and Recurrent average outingsmore realistic in a world where M&A of 10 to 100 million euros are much more frequent than omPos at 10 billion.
This model nevertheless requires a deep relegation of the VC:
-
- Smaller fundsadapted to average outings.
- Cash-Flow-oriented selection strategies, no longer Tam only.
- More balanced value sharing mechanisms with the founders.
- Evergreen or semi-liquid vehiclesallowing to extend the time horizons.
Change objectives to change structure
This new deal is not an anti-tech utopia. It is a rational response to the evolution of tools (producing costs less), behaviors (the founders target freedom more than blitzscale), and markets (exits are concentrated in the median).
He invites to develop the very definition of success: no longer a statistical rarity, but a systemic robustnesswhere venture capital no longer seeks exceptions at all costs, but to build solid bases for an enlarged entrepreneurial fabric.
