Introduction
Count the buyers at the end of a limited partner (LP) portfolio sale and you learn how the seller's advisor settled its central trade-off. Jefferies' review of the first half of 2025 found that the average LP transaction closed with 1.9 buyers, down from 2.3 in the second half of 2024, as secondaries funds grew through evergreen vehicles and closed-end funds above $20 billion that could write larger cheques. Behind every count above one sits a mosaic: a portfolio divided among several buyers because the best price for each fund came from a different bidder.
A mosaic usually raises the headline. It also adds agreements, counterparties, conditions, and the chance that unwanted funds stay with the seller. The private capital advisory (PCA) banker has to judge whether the extra points of net asset value (NAV) survive those costs, a comparison of risk-adjusted proceeds rather than headline percentages.
Why a Portfolio Breaks Into Pieces
Buyers bid on portfolios but price them fund by fund, and their fund-level knowledge is uneven. A buyer that already holds a manager's earlier funds can forecast their cash flows with more confidence than a generalist, and a lower required return becomes a higher price. Strategy specialists add a second layer, pricing their own strategy more aggressively than a generalist pricing the same funds inside a mixed book, as pricing by strategy explains. The same Jefferies review called mosaic solutions a meaningful part of the market, particularly for multi-strategy portfolios, where specialists drive outlier pricing.
- Mosaic Sale
A secondary sale in which a seller's portfolio of fund interests is divided among several buyers, each acquiring the funds for which it made the winning bid, instead of being sold whole to one buyer. Each buyer usually signs its own purchase agreement.
The logic is the one behind a sum-of-the-parts valuation of a conglomerate: each piece is worth most to its best owner, and one owner of the whole rarely pays every piece's best price.
When One Buyer Pays More for the Whole
The reasoning also runs the other way. Buyers accept lower return targets on diversified books than on single positions, as pricing LP interests explains, so a large buyer can pay a diversification premium for a whole portfolio that no slice would earn alone. Scale buyers, the evergreen vehicles and large funds behind the falling buyer count, can also absorb a whole book in one trade. Structure pushes the same way: Jefferies' 2024 full-year review found that structured transactions were often bought by a single buyer, which helped single-buyer solutions reach 63% of LP deals that year.
Building the Winning Combination
Final bids arrive in three shapes, which the advisor sets on one grid:
- Whole-portfolio bids for everything, sometimes offered only for everything.
- Fund-by-fund bids that the buyer will honour in any combination.
- Slice bids for a defined group, such as all the venture funds, often conditional on winning the whole group.
The first shape sets the benchmark, since it is the only offer that clears the whole perimeter with one counterparty.
- Portfolio Bid (Secondaries)
A single offer for every fund interest in a sale perimeter, quoted as one blended percentage of NAV or as fund-level prices valid only if the buyer wins all of them. For the seller, it trades a possibly higher combined price for one counterparty and one negotiation.
Assembling a mosaic is an ordered search, not a matter of taking the highest number for each fund:
Best bid per fund
Take the highest price for each interest. This is a ceiling, not an achievable answer.
Apply buyers' conditions
Remove combinations a buyer will not accept, such as minimum sizes or all-or-nothing slices.
Price the leftovers
Estimate what interests left out of every feasible combination would fetch later, or are worth if held.
Compare with the whole
Set each feasible mosaic against the best portfolio bid after risk and cost.
Go back to bidders
Show portfolio bidders the gap and invite them to close it before signing.
Step two is where most of the theoretical gain disappears: a buyer that bid high on one venture fund may have priced it expecting to win the manager's other funds, and its conditions say so. How each agreement treats an interest that fails to transfer is covered in transfer mechanics and the purchase agreement.
Mosaic Versus Whole Portfolio: The Comparison
The two routes differ on five dimensions, and only the first favours the mosaic by default:
| Dimension | Whole-portfolio bid | Mosaic |
|---|---|---|
| Proceeds | One price, often lower | Best available price per fund |
| Consent risk | One buyer per transfer | Several buyers, some with conditions |
| Timing | One negotiation and closing schedule | Several agreements and closings |
| Unsold funds | The buyer takes everything | Unbid funds stay with the seller |
| Certainty | Rests on one counterparty | Rests on the weakest conditional piece |
Each fund still goes to one buyer, but each buyer brings its own purchase and sale agreement (PSA), counsel, and closing calendar, and a buyer new to a manager may clear consent more slowly. Unbid funds leave the seller with residual administration, from capital calls to reporting, or a later tail-end sale.
Take a seller with 30 funds and $800 million of reference NAV: $520 million in 18 buyout funds, $180 million in eight venture funds, and $100 million in four funds more than ten years old. The best portfolio bid is 87%. Three specialists bid on slices, and the venture bid is all or nothing.
| Piece | NAV | Winning bid | Proceeds |
|---|---|---|---|
| Buyout funds (Buyer B) | $520m | 94% | $488.8m |
| Venture funds (Buyer C) | $180m | 86% | $154.8m |
| Tail-end funds (Buyer D) | $100m | 68% | $68.0m |
| Mosaic total | $800m | 89.0% | $711.6m |
| Portfolio bid (Buyer A) | $800m | 87% | $696.0m |
Risk-Adjusting the Difference
On headline the mosaic leads by $15.6 million. Now suppose a one-in-five chance that a venture GP refuses consent or misses the long-stop date, Buyer C walks, and the venture funds are resold later at 78%, or $140.4 million. The expected cost is 20% of $14.4 million, about $2.9 million. Add $1 million of extra legal and administrative cost for three agreements instead of one, and the mosaic still leads by about $11.7 million. Had Buyer A bid 88.5%, or $708 million, the $3.6 million headline gap would fall short of the $3.9 million expected cost, and the lower headline would be the better sale.
When a Mosaic Is Worth the Extra Work
Published deals rarely show this arithmetic. CPP Investments, which manages the Canada Pension Plan, completed a sale in May 2026 of 33 fund interests to Blackstone Strategic Partners and Ardian for net proceeds of about C$4 billion, naming both buyers but not which funds each took. Whatever the split, a mosaic tends to earn its costs when:
- The portfolio spans strategies or managers that buyers know unevenly, visible as bid dispersion in the first round.
- The book exceeds what one buyer's concentration limits allow at its best price.
- The seller has time for several closings.
A whole-portfolio bid tends to win when the book is mostly buyout from widely held managers, when the seller faces a hard deadline, or when a mosaic would strand the oldest funds. How these choices fit the wider auction is set out in the LP portfolio sale process.
A buyer offering one number for 30 funds is paying more than it would for some and less for others, letting the strong funds carry the weak ones. The advisor's mosaic unbundles that cross-subsidy. Where it is small, as in a homogeneous buyout book, unbundling gains a point or two that the extra work can consume. Where it is large, as in a multi-strategy portfolio with a stale venture tail, the best portfolio bidder is in effect charging the seller for taking funds it does not want, and splitting the book lets specialists pay that difference back.
