Two large pools of capital now sit on opposite sides of the same market: portfolio-driven sellers rebalancing after public-market moves, and evergreen vehicles holding roughly US$500bn of semi-liquid assets looking for deployment.
Agreeing a price is the short part. What follows — reference dates, consent windows, funding mechanics and register update — determines whether the trade completes and at what effective price.
Most execution risk is priced into the discount. Sellers who can reduce timing and consent uncertainty recover part of it.
Two earlier editions in this series described opposite sides of the same market. One examined why crossover funds and overallocated institutions become structural sellers when public markets fall, regardless of how their private holdings are performing. Another described how evergreen vehicles have become a standing source of demand, with semi-liquid structures holding around US$500bn of assets and 53% of secondary buyers now operating one. What neither described is the part in between: how a willing seller and a willing buyer actually get to a completed transfer.
That gap matters because it is where most of the cost sits. In a listed market, agreement and settlement are essentially the same event. In private markets they are separated by weeks, by other people's consent rights, and by a set of pricing conventions that can move the economics of a trade after the price has been agreed.
Stage one: finding the other side without moving the price
A seller's first problem is information. Signalling that a large position is for sale can itself damage the price, and in single-company trades it can reach the issuer before the seller has decided anything. Sellers therefore approach the market through intermediaries, in stages, with the identity of the seller and the size of the position disclosed progressively rather than at the outset.
Buyers face the mirror problem. A buyer who reveals conviction early invites competition; one who reveals too little cannot get access to the information needed to underwrite. In practice this produces a staged process: an anonymised description of the asset, then a data set under confidentiality, then bids, then a narrower round with the seller's identity and the full documentation. Each stage takes time, and time is the raw material of the discount.
Stage two: pricing against a moving reference
For portfolio trades, price is expressed as a percentage of net asset value rather than as an absolute figure, and NAV is reported with a lag. That creates a convention problem that has real economic consequences. A bid struck at 92% of a 31 March NAV, signed in June and funded in August, refers to a valuation that is five months old by the time money moves.
Three mechanisms bridge the gap, and all three change who bears which risk:
- The reference date. The parties fix the NAV date the price refers to. Distributions the fund makes after that date typically pass to the buyer, and capital calls after it are typically the buyer's obligation — which is why a stale reference date with heavy interim activity produces a different economic outcome from the headline percentage.
- True-up mechanics. Adjustments for interim cash flows between the reference date and closing, so that neither side gains or loses from timing alone.
- Deferred consideration. Part of the price paid at a later date, sometimes with an interest factor. This raises the headline percentage a seller can report while leaving the buyer holding less cash risk, and it is one reason quoted pricing levels are not directly comparable between transactions.
In single-company trades the reference is not NAV but the last primary round, and the same principle applies: the further the market has moved since that round, the more the negotiation is really about which reference point governs.
Table 1: Where a secondary transaction actually spends its time
| Stage | Typical duration | What determines it | Who carries the risk |
|---|---|---|---|
| Preparation and staged marketing | Weeks | Seller's readiness, quality of the data set, number of counterparties approached | Seller — every week of delay is exposure to a moving reference price |
| Bidding and selection | Weeks | Depth of the buyer pool for that asset type; venture portfolios attract fewer bidders than diversified buyout | Seller |
| Documentation and consents | 30–90 days, sometimes longer | Rights of first refusal, board consent, transfer provisions in the constitutional documents | Both — either party can be displaced by a consent holder |
| Funding and register update | Days to weeks after consents | Escrow or delivery-versus-payment arrangements; issuer or transfer agent processing | Buyer — until the register records the transfer, the position is not owned |
Stage three: the consent chain
This is the stage that most often defeats a transaction, and it is entirely outside the control of both parties. In fund transactions, the general partner's consent is required, and sponsors use it: they may decline a transfer to a competitor, require the buyer to meet eligibility standards, or take time the seller does not have. In GP-led processes, existing investors are given a defined window to elect to sell or roll, with industry guidance recommending at least 30 calendar days.
In single-company trades the chain is longer. A transfer notice triggers the company's right of first refusal, commonly 30 days; investor rights may follow; and board consent, joinders and updated share registers complete the sequence. An earlier edition in this series set out why 30 to 90 days is a normal end-to-end timeline and why the delay is contractual rather than technological. The practical consequence is that a buyer and seller can agree everything and still not have a trade.
Stage four: funding and title
Settlement is the least discussed and most consequential stage. Two questions decide the buyer's risk. Does money move against a transfer the issuer or general partner has recognised, or in advance of it? And is the record of ownership updated by the issuer, its transfer agent or the fund administrator, with the buyer receiving evidence of that update?
Where funds move before recognition, the buyer holds a contractual claim rather than an asset for a period. Escrow and delivery-versus-payment arrangements exist to close that window. They are ordinary infrastructure in listed markets and still inconsistently applied in private ones, which is one reason institutional buyers increasingly ask about settlement mechanics before they ask about price.
Why this shows up in the discount
Every stage above is a source of uncertainty, and uncertainty is priced. A seller who must complete within a quarter has less negotiating power than one who can wait. An asset with a thin buyer pool clears wider than one with competitive bidding. A position whose consent path is unclear attracts a wider spread than one where the issuer's position is known in advance. The dispersion the market reports — buyout portfolios clearing far above venture portfolios, top continuation vehicles at or above par while weaker ones trade well below — reflects these operational differences as much as it reflects views on the assets.
The corollary is the useful part. Execution risk is not fixed. Sellers who prepare documentation in advance, establish the consent position before marketing, and offer a clean transfer path recover part of what would otherwise be conceded in price. Buyers who can commit quickly and demonstrate eligibility win access to assets that never reach a broad process.
What the two sides of the market still lack
Crossover sellers and evergreen buyers are, in aggregate, well matched: one needs to reduce private exposure on a timetable, the other has continuous inflows to deploy and a structural preference for assets that are already seasoned. They are not, however, well connected. Sellers reach buyers through relationships and intermediated processes; buyers assess opportunities from data sets assembled ad hoc; and each transaction reconstructs, from scratch, work that was done identically on the last one — verifying the cap table, establishing the consent position, confirming who has authority to transfer.
That repetition is the real inefficiency in this market. It is not a pricing problem and it will not be solved by more price data. It is a process problem, and it is why the operational layer — verified records, known consent positions, standard documentation and settlement that moves money against recognised title — is where the next improvement in this market is likely to come from.