Secondaries Market Report 2009 — What a Discount Reveals

Private equity and venture funds report values that are appraisals, not prices. A general partner estimates what holdings are worth, subject to a valuation policy and an audit. In the absence of a transaction there is nothing to test the estimate against, and — as the US Venture Capital Report 2008 argues — an untested mark is an assumption about what a valuation would be.

In 2009 that assumption was tested, at scale, for the first and largest time. Limited partners sold fund stakes in the secondary market at prices established by negotiation between buyers and sellers. Those prices were substantially below the reported net asset values of the same stakes.

The temptation is to read the discount as a verdict — that the marks were wrong by the amount of the discount. That reading is too simple, and the decomposition is the analytically useful part. A secondary price reflects at least five things: a view on the underlying assets, the burden of unfunded commitments transferring with the stake, uncertainty about the funds' remaining capital calls, information asymmetry between a general partner and an outside buyer, and — decisively in 2009 — how many buyers were bidding. Only the first is a statement about the marks.

The sellers are the other misread element. The classic image of a secondary sale is a distressed institution dumping assets. In 2009 the largest sellers were substantial institutions with sound balance sheets, selling for the reason described in the US Venture Capital Report 2008: the denominator effect. Public portfolios had collapsed; private marks had barely moved; private allocations were therefore far above policy targets without any private asset changing value. Simultaneously, capital calls continued while distributions had stopped.

So institutions were over-allocated to an asset class that was demanding more money and returning none. Selling stakes solved both problems at once, and it was the only lever available quickly.

The lasting consequence was structural. Having discovered that fund stakes could be sold, institutions kept selling them — in good conditions, for portfolio management rather than for relief. A market created by distress became permanent infrastructure, and that changed what a private fund commitment is.

Why sound institutions became forced sellers

The seller profile matters because it determines how much of the discount reflects the assets.

Three pressures arrived simultaneously, and each alone would have been manageable:

• The denominator effect. Public holdings fell sharply; private marks, appraised and lagged, barely moved. The private share of the portfolio rose well beyond policy limits through arithmetic alone.
• Capital calls continued. Funds in their investment period kept drawing, and those obligations are contractual — a limited partner who fails to meet a call faces severe penalties including forfeiture.
• Distributions stopped. With exit markets closed, the cash that normally funds new calls was not arriving. The self-funding cycle mature private portfolios rely on had broken.

The combination is the problem. An institution over its allocation limit, contractually obliged to send more money to the asset class, receiving nothing back, and needing liquidity for its other obligations.

The available responses were all unattractive:

• Sell public assets to meet calls — but that worsens the denominator effect, since it shrinks the portfolio further.
• Default on capital calls — punitive, reputationally severe, and it forecloses future access to the best managers.
• Sell private stakes — available, immediate, and it fixes the ratio and the call obligation together.

The seller was not making a judgement that private assets were overvalued. They were solving an allocation ratio and a cash-flow obligation, and the stake was the only instrument that solved both.

This distinction determines how to read the price. A transaction between a motivated seller solving a balance-sheet problem and a small number of specialist buyers is not a clean measurement of value. It is a price under duress with almost no competitive tension, and treating it as a pure valuation signal overstates what it says about the assets.

Decomposing the discount

To use the 2009 pricing as evidence, it has to be broken into components, because they behave very differently.

Component one: a view on the underlying assets. The buyer's assessment of whether the reported marks are achievable. In 2009 this was genuinely negative — public comparables had fallen far more than private marks, and some convergence was expected.

Component two: the unfunded commitment. A stake carries an obligation to fund future capital calls. The buyer is not just acquiring assets; they are assuming a liability of uncertain size and timing. In 2009 that liability was frightening — nobody knew how much would be called or whether those investments would be good ones.

Component three: uncertainty, distinct from expected loss. Even a buyer with an unbiased view of value demands compensation for the width of the distribution. In 2009 the distribution was extraordinarily wide.

Component four: information asymmetry. The general partner knows the portfolio; the seller knows more than the buyer; the buyer prices what they cannot verify. The less transparent the fund, the wider the discount, independent of quality.

Component five, and in 2009 the largest: the absence of buyers. Secondary capital was limited and its providers were managing their own problems. A price set with two bidders is a different number from one set with twenty, and this component says nothing whatsoever about the assets.

The practical conclusion: the discount was real evidence that marks were optimistic, and it was not a measurement of by how much. A large share of it was the price of liquidity in a market where liquidity had nearly disappeared — which is the recurring finding of this archive's 2008 reports, arriving in a fifth asset class.

The buyer's problem

Understanding why discounts were so wide requires seeing what the buyer was actually being asked to do.

What the buyer receives: a share of a portfolio of private companies, valued by the manager, which the buyer cannot inspect individually, cannot control, and cannot easily exit.

What the buyer assumes: an obligation to fund unknown future calls, at unknown times, into investments not yet made — by a manager the buyer did not select.

The valuation problem has no clean solution:

• The reported NAV is an appraisal, produced by a party with an interest in it being robust.
• Underlying company information is limited, and much of what matters is not in the reporting.
• The remaining investment period means part of the portfolio does not exist yet, so a share of what is being bought cannot be analysed at all.
• The exit path is entirely dependent on the manager's decisions and on markets reopening.

So the buyer prices a distribution, not an asset, and demands a margin proportionate to its width. In 2009 every input was at its least certain simultaneously.

This explains the counterintuitive pattern in which better-quality funds sometimes traded at wider discounts. A fund early in its life with a large unfunded commitment carried more uncertainty than a nearly-harvested fund with a known portfolio, regardless of manager quality. The discount tracked structural characteristics — remaining unfunded, vintage, transparency — more closely than it tracked underlying quality, which is the clearest evidence that it was not primarily a valuation judgement.

Why the market became permanent

Secondary trading had existed since the 1980s as a small, faintly disreputable corner. After 2009 it became infrastructure, and the reasons are structural rather than cyclical.

What changed on the seller side:

• The stigma disappeared. Selling had implied distress or a broken relationship with a manager. When large, respected institutions sold for transparent allocation reasons, it became a normal action.
• It was recognised as a portfolio tool. Institutions discovered they could actively manage private allocations — trimming, rebalancing, exiting relationships — rather than holding every commitment to term.
• Programmes became routine. Periodic secondary sales became a standing part of managing a private portfolio, not a response to stress.

What changed on the buyer side:

• Dedicated capital scaled rapidly, as returns achieved on 2009 vintages attracted commitments.
• Pricing capability improved, with specialists building the analytical infrastructure to underwrite portfolios quickly.
• The buyer base broadened, which compressed discounts — the missing-bidder component of 2009 pricing shrank permanently.

What changed for managers:

• Manager attitudes shifted from hostile to accommodating, since a functioning secondary market helps them raise successor funds by giving investors an exit.
• GP-led transactions emerged, in which the manager initiates a transaction to provide liquidity or extend a hold. This inverts the original structure — the manager is now on both sides — and it became the market's largest segment in later years.

The Secondaries Market Report 2019 and Secondaries Market Report 2023 describe that maturity. Its origin is here: the moment the asset class discovered a price and decided it wanted one.

What semi-liquidity changed

The deepest consequence is conceptual, and it is unresolved.

The original bargain. Investors accept that capital is locked for a decade and cannot be withdrawn. In exchange they expect a premium over public markets — the illiquidity premium — justified by the manager's ability to operate without market pressure and to time exits.

What a functioning secondary market does to that bargain:

• The illiquidity is materially reduced. A stake can be sold in weeks at a knowable price.
• The premium is still expected, because it is embedded in how the asset class is marketed and in return targets.
• The valuation convention did not change. Holdings are still appraised, still smoothed, and still reported on a lag.

That combination is uncomfortable. An asset class that is priced as illiquid, valued as if unobservable, and increasingly tradeable is being compensated for a constraint that has partly been removed.

A discount for illiquidity is payment for a real constraint. When the constraint softens and the discount remains, someone is being paid for something they are no longer providing — and it is not obvious who.

The counterargument is fair and worth stating. Secondary liquidity is conditional: it is available in normal conditions and evaporates in stress, which is exactly when it is wanted. On that view the illiquidity premium is still earned, because the liquidity is unreliable — and 2009 itself is the supporting evidence, since that was when discounts were widest and bidders fewest.

Both are partly right, and the tension is live. The Private Equity Report 2015 examines appraisal-based valuation directly, and the Secondaries Market Report 2019 revisits it once the market had matured. What 2009 established is that the question exists at all, because before there was a transaction price there was nothing to compare the appraisal to.

What an allocator could act on

Decompose any secondary price before reading it as a valuation. Underlying view, unfunded liability, uncertainty width, information asymmetry and bidder count are all in the number. Only the first is about the assets, and in stressed markets it is rarely the largest.

Model capital calls against a stressed public portfolio. The 2009 problem was not call size but coincidence — calls arriving while public markets fell and distributions stopped. Commitment capacity should be tested under the conditions in which it will actually be tested.

Track the denominator, not just the numerator. An allocation can breach its limit with no private asset changing value. That is a mechanical consequence of appraisal lag and it will recur in any sharp public drawdown.

Treat appraisal smoothing as deferred volatility. It flatters reported risk and creates the allocation problem exactly when flexibility is scarcest. Smoothing relocates volatility in time; it does not remove it.

Ask what the illiquidity premium is now paying for. If a stake can be sold in weeks, the constraint being compensated has partly gone. Whether the premium should persist depends on how reliable that liquidity is in stress — which is an empirical question with one large data point, and it points the wrong way.

Watch who initiates a GP-led transaction and who prices it. When the manager is on both sides, the conflict is structural rather than incidental, and the protections are procedural. That is a governance question, and it is the market's fastest-growing segment.

What 2009 established

• A transaction-based test of appraisal marks, the only broad one the private asset class has run, and it found them optimistic without establishing by how much.
• The denominator effect as a forced-selling mechanism that operates on sound institutions and requires no view on the assets at all.
• Secondary discounts are mostly not valuation judgements — unfunded liability, uncertainty and bidder scarcity dominate, and they track structural characteristics rather than quality.
• Liquidity, once discovered, became permanent infrastructure, converting a distress channel into a portfolio-management tool.
• Semi-liquidity put the illiquidity premium in question while leaving the appraisal convention untouched — a tension that remains unresolved.