Search Fund Returns vs Private Market Benchmarks

Search fund returns vs private market benchmarks cannot establish which investment will serve an LP better when losses, missing outcomes and selection limits differ across the evidence. Start by identifying whose experience each source observes, then separate study returns from annual benchmark results. The research can help frame questions about an SMB allocation, but these figures do not support an asset-class ranking or a forecast.

Start with the population and missing outcomes

A search can end without an acquisition, and an acquired business can lose capital. Research readers need to know where those outcomes appear before interpreting a return headline. A sample of completed acquisitions answers a different question from a sample that also follows unsuccessful searches. Neither description alone tells you how an investor's entire commitment performed.

Selection begins with who could enter the dataset. Survivorship concerns arise when the population being observed omits failed or discontinued efforts. Missing outcomes create a further question: did researchers lose contact, exclude incomplete records or wait for an outcome that has not occurred? These possibilities require different explanations. A missing result should remain unknown, rather than being counted as either a success or a loss.

Unfinished investments present another problem. An investment still being held may have an estimated value but no final sale result. Including that value can help describe the portfolio at an observation date, while leaving uncertainty about what investors will ultimately receive. Excluding unfinished investments changes the population. Neither treatment makes the open question disappear.

Andy Allaway describes ad hoc follow-up in his Empire Flippers interview on The SMB Investor podcast. The narrow lesson from Allaway's account is to ask whose outcomes became visible and whose did not. A buyer who returns to describe an experience supplies an observation; that account does not establish the experience of everyone who bought a business.

That observation is a reading question, not evidence that academic researchers or benchmark providers use the same collection methods. Ask each source how it obtains results, handles nonresponses and treats unfinished holdings. Avoid attaching a generic bias label to every dataset before inspecting its methods.

Separate study returns from annual benchmarks

The registered evidence describes different populations and measurement windows. Keeping those descriptions attached to the figures prevents a familiar mistake: treating a long-running study result as if it were earned during the same calendar period as a market benchmark.

The Stanford GSB Search Fund Study reports an aggregate pre-tax IRR of 33.9 percent as of Dec 31, 2025¹. It covers first-time, investor-funded core searches in the United States and Canada launched since 1984¹. The aggregate pools investor cash flows across funds, including unsuccessful searches, operating companies and exits; it is not a calendar-year observation. Operating holdings contribute estimated value rather than final sale proceeds. The study excludes follow-on financing; its fee basis is not specified.

The Yale SOM case How are Search Fund Investors Really Faring?, by Lazier, Thomas and Wasserstein, reports a median MOIC of 1.0x across investor-backed search decisions, including broken searches². Its observations come from participating investors across funds. For operating deals, the case uses current marks supplied by investors; the reported returns are gross of GP fees. The observation window and investor-specific tax basis remain unresolved. Keep those fields unknown rather than borrowing conventions from another source.

The IESE International Search Fund Study reports an international aggregate IRR of 18.1 percent³. Its return calculation uses only a subset of concluded first-time core searches outside the United States and Canada: it includes failed searches and eligible operating and exited companies, but excludes some acquisitions with too little operating history or insufficient data. Follow-on financing was excluded for operating companies but included for exits, as the IESE methodology explains. The study uses reported market values for operating companies and calculates returns on a pretax cash-flow basis. The aggregate IRR spans the included investment histories through the study cutoff rather than a single calendar year. The fee basis remains unspecified.

Cambridge Associates reports calendar 2025 US buyout returns of 7.6 percent and US venture capital returns of 21.1 percent⁴. These are annual observations from its benchmark populations, distinct from the search-fund study aggregates. The provider describes private index returns as pooled horizon IRRs, net of fees, expenses and carried interest; the annual measurements can reflect remaining portfolio value as well as cash flows. Investor-specific tax treatment is not established here. These conventions appear in the benchmark commentary.

A shared percentage sign does not align the evidence. Study aggregate IRRs combine cash flows over the included investment histories; their precise measurement windows need to be read from each study. An annual benchmark measures a defined calendar window, potentially across holdings at different stages. Comparing the displayed rates as if they shared the same starting point would create an unsupported conclusion about relative performance.

Compare what the measures support

Use the measurement label to determine what question a result can answer. The distinctions below apply before deciding whether a figure belongs beside another figure in an investment discussion.

Measure	What it describes	What it leaves unanswered
A study aggregate IRR combines the study's eligible cash flows and values.	It describes their timing-sensitive combined result under the study method.	It does not identify the typical company or establish a particular LP's experience.
The Yale median MOIC describes the middle investor-backed search decision, including broken searches.	It helps locate the center of that sample's multiples.	It does not express an annual rate or establish when cash reached an investor.
An annual private-market benchmark describes a defined reporting window.	It provides period-specific context for its covered population.	It does not establish the outcome of a new commitment held through its life.

An aggregate and a median are different summaries. A combined calculation can respond to the size and timing of contributions and proceeds. A median locates the middle observation after the source's inclusion rules have been applied. Neither should be described as the other, and neither automatically represents an investor who holds a different collection of investments.

Internal rate of return (IRR) reflects cash-flow timing. Multiple on invested capital (MOIC) relates proceeds and remaining value to the capital invested. A multiple does not contain enough timing information to turn it into an annual return. Dividing, annualizing or otherwise converting the Yale multiple would introduce assumptions that this evidence does not supply.

Valuation also needs its own field. Record whether a result reflects completed exits, estimated remaining value or a combination. If a source summary does not explain the treatment, mark it unknown. Do not silently equate estimated value with distributed cash, or assume that a result labelled pre-tax also specifies its fee treatment.

The existing search fund and traditional PE comparison provides background on the investment models. Here the question is narrower: whether the observations support a common performance comparison. Model labels cannot resolve a mismatch in periods, populations or measurement conventions.

Search fund returns need portfolio context

A portfolio sleeve groups investments by the purpose assigned to them within a wider portfolio. That purpose should be stated before a research headline becomes a reason to add exposure. The SMB sleeve portfolio guide addresses that broader planning question. This article supplies the evidence limits that belong beside it.

We have no matched return series for secondaries, private credit and real assets. Those comparisons therefore remain qualitative. For secondaries, ask what underlying exposure is being compared and where the observed investment sits in its life. For private credit, distinguish a claim on repayment from an ownership interest. For real assets, identify the source of cash flows and what the valuation represents.

These questions describe differences to investigate. They do not establish that any category has lower losses, more dependable distributions or better portfolio fit. Broad labels can contain different exposures, and a missing matched series cannot be repaired with an appealing anecdote or an unrelated headline.

A commitment also creates funding and liquidity questions that an academic return figure cannot answer. Keep those questions in the allocation sizing and commitment pacing discussion. There is no supported allocation percentage, return premium or portfolio forecast to derive from the observations here.

John Steinberg's discussion in Inside the Search Fund Playbook on The SMB Investor podcast supplies a separate attribution boundary. Evidence about the search-fund model belongs to the asset class. This article uses only that qualitative distinction, without carrying over the episode's numerical or correlation claims.

The Stanford, Yale and IESE findings are asset-class or study-sample observations. They are not SMB Investor Network performance or evidence of what a particular opportunity will deliver. Past industry performance does not guarantee future results.

Record the evidence gaps

A useful comparison note can end with an unresolved question. Forcing an answer where the sources do not align creates more confidence than the evidence supports. Keep the limits close to the claim so that another reader can see what remains unverified.

Use this checklist when transferring the observations into your research notes:

• Record the source and whether you read primary research, a provider's commentary or a secondary recap.
• Describe the observed population and the selection rules that are known. Identify any eligibility rules still awaiting verification.
• Separate the observation cutoff from the beginning of the measurement period and the publication date.
• Retain the exact metric, aggregation level and inclusion of broken searches where specified.
• Record how realized proceeds, remaining valuations, missing outcomes and unfinished investments are treated, or mark the treatment unknown.
• State the known fee and tax basis without treating silence as proof of a convention.
• Explain what makes a proposed comparison unresolved, including the absence of a matched series.

The Stanford and IESE primary methodologies identify their populations, cash-flow basis and valuation treatment. The Yale observation window remains unresolved. Cambridge's annual observations do not close those gaps, and its benchmark conventions should not be assigned to the studies. Record each limitation separately; resolving a valuation question would not also resolve population selection or timing.

The LP investment-risk guide provides a broader place to continue reading about risk. Keep this page's conclusion confined to the evidence: the available observations help identify questions, but they do not establish a common return basis across these private-market exposures.

Keep private market benchmark comparisons traceable

Evidence used

• ¹ supplies the Stanford aggregate pre-tax IRR, observation cutoff and core-search population. The registered recap was checked against the Stanford GSB primary study's eligibility and returns methodology.
• ² supplies the Yale median across investor-backed search decisions, including broken searches. Its source is the Yale SOM case by Lazier, Thomas and Wasserstein. The observation window remains unresolved here.
• ³ supplies the IESE international aggregate IRR. The registered secondary account was checked against IESE's primary study for coverage and methodology.
• ⁴ supplies Cambridge Associates' annual buyout and US venture capital benchmark observations. Its source is the provider's calendar-year benchmark commentary. Those observations remain separate from the study results.

Podcast sources

• John Steinberg, Inside the Search Fund Playbook, The SMB Investor podcast. This article paraphrases only the asset-class attribution boundary.
• Andy Allaway, Empire Flippers interview, The SMB Investor podcast. This article paraphrases only the question of whose outcomes were observed.

Record what remains incomparable in your research notes, including the missing period alignment, unresolved fields and absent matched series.

Published in partnership with SMB Investor Network.

Keep the source limits attached to your notes as you continue researching opportunities.