Yann Magnan says evergreen funds and retail investors are pushing private-market valuations to mimic public markets for real-time transactions.
73 Strings' AI extraction technology reaches close to 95% accuracy, pulling data from board decks and quarterly reports.
Magnan says AI agents won't fully replace human valuers soon because regulators, auditors and 401(k) owners must be comfortable.
Private asset valuations leave Excel as retail money arrives
Private asset valuation is moving out of Excel because the money behind it changed. Evergreen funds and retail investors want to enter and exit at any time, so valuations now support transactions rather than quarterly reporting and must resemble public market pricing. Yann Magnan co-founded 73 Strings in 2020 to put that work into a governed, auditable framework, where technology structures the data and people keep the judgment.
Takeaways
Retail capital forces valuations to mimic public markets. Evergreen fund inflows accelerated dramatically over the past year. Investors want to enter and exit at almost any time, so valuations now support transactions rather than quarterly reporting to institutional LPs. Without public-market-like pricing they cannot transact.
More liquidity and transparency is the price. Businesses need private capital and retail investors have capital, so some liquidity is the cost, Magnan says. He personally treats more liquidity and transparency as good, conceding it is debatable, and expects 401(k) plans to open to retail investors.
Technology's real leverage is governed data. Valuation has run on Excel and email for roughly 40 years, which is costly and inefficient. The larger gain is collecting and structuring portfolio company data in an auditable, traceable environment, leaving human judgment for genuine choices.
No single methodology, but consistency wins. Ten analysts produce ten different valuations. What varies is how methodologies are applied. Consistency over time matters most, especially to regulators, and a GP cannot change methodology without explaining and justifying it. Magnan calls valuation art and science.
Data is the firm's DNA, not a moat. A GP with 100 portfolio companies draws more intelligence than one with 20 to 25, especially when data points are labelled consistently. Magnan would not discard smaller managers, whose repeated deal experience gives them valuable human expertise.
AI will not replace the valuation human yet. Regulators and auditors must be comfortable with the process, and AI is probabilistic. Agents can still roll valuations forward overnight. Magnan thinks AI is probably already better than humans at parts of the work.
The missing piece is GP-to-LP data sharing. Magnan is surprised that data produced in one format on the GP side gets turned into a PDF for LPs. Sharing it digitally is the next step, and he frames interconnection as ultimately a question of trust.
Where valuation is leaving Excel
The US market is the most advanced at moving valuation work out of Excel, with the largest GPs adopting software platforms.
Tap a name to go there · tap a line to see why they are linked
On this episode 2
Show 1
The UK is moving fast with a similar mindset to the US.
Continental Europe is slower, with more old-school processes still in place.
73 Strings opened a Singapore office weeks before this episode, citing massive APAC demand.