A track record is the single most prominent element in almost every fund presentation, and it is among the least reliable. Not because sponsors routinely fabricate results — most do not — but because the way performance is selected, measured, and presented offers many opportunities to shape an impression without stating anything untrue.
Evaluating one is mostly a matter of asking what is not on the page.
Start with what is settled
Full-cycle results — programs that acquired assets, held them, sold them, and returned capital — are the only results where the outcome is known rather than estimated. Everything else is a mark.
The distinction matters enormously in illiquid strategies. A program still holding assets is reporting values the manager produced through a valuation process. Those values may be careful, independently supported, and accurate. They are still estimates, produced by a party with an interest in the estimate, and they will be settled only when someone actually pays.
The practical rule: weight realised results heavily, unrealised marks lightly, and treat a record composed mainly of unrealised marks as substantially unproven — regardless of how strong the numbers look. DPI tells you what came back; RVPI tells you what the manager says is still there.
The selection questions
These are the questions that most often change the picture, and they must be asked explicitly because the answers rarely appear unprompted.
Is this every program? The record should cover all programs the sponsor has managed in the strategy — including ones that were wound down, transferred, restructured, or performed badly. A record showing only continuing or successful programs is a selection, not a history. Ask directly: has any program you have sponsored been excluded from this presentation, and why?
Whose record is it? Performance sometimes travels with a firm when the people who generated it have left, or travels with people who were part of a team without being responsible for the results. The relevant question is whether the individuals who produced the record are the individuals who will manage the new program, and in what roles.
Was the strategy the same? A sponsor with a strong record in one asset class, one geography, or one deal size raising a fund in a different one has a record of limited relevance. Strategy drift is common and is often presented as evolution.
Was the scale the same? A record built deploying modest sums into small transactions may not transfer to a much larger fund, because the investable universe changes. Fund size is one of the more reliable causes of return dispersion between a sponsor past and its future.
Which vintages? Results are heavily influenced by when capital was deployed. A record concentrated in favourable vintage years says less about skill than one spanning a full cycle. A sponsor with no experience of a downturn has an untested record, however good the numbers.
The measurement questions
Gross or net? Gross returns exclude fees and carried interest. Net returns are what investors received. The gap is large, and presentations sometimes show gross prominently with net in a footnote. Only net matters to an investor.
IRR or multiple? IRR is sensitive to cash flow timing; multiples such as MOIC and TVPI ignore time. Each conceals something the other reveals, which is why serious analysis uses both.
Was a subscription line used? Deferring capital calls by borrowing shortens the period investor capital is outstanding and mechanically raises reported IRR without changing a single investment. ILPA has encouraged disclosure of returns both with and without the facility effect. If the presentation does not address it, ask — and expect a manager with nothing to hide to answer readily.
Against what benchmark? A benchmark should be appropriate to the strategy, geography, and vintage. Comparison against a broad equity index tells you little about a private credit fund. Public market equivalent analysis is one of the more honest comparisons available, because it asks what the same cash flows would have earned in public markets.
Who calculated it? Whether the figures were prepared by the manager alone, reviewed by the administrator, or verified by an auditor or an independent party. GIPS compliance, where claimed, has a specific meaning and can be checked.
Reading the losses
The most informative part of a track record is usually the part that went wrong.
Every sponsor with a real history has losses. A record showing none is either very short, very lucky, or incomplete. What matters is what happened around them:
- What went wrong, in the sponsor own account, and does the explanation attribute the outcome to identifiable decisions or entirely to external conditions?
- When was the problem recognised? A manager that identified a deteriorating position early and acted has demonstrated something a manager that marked it at cost until the last moment has not.
- How were investors told? Reporting that becomes vaguer as performance deteriorates is a pattern experienced allocators specifically watch for.
- What changed afterward in process or personnel?
A sponsor that discusses its failures specifically and without defensiveness is providing better evidence of judgment than one presenting an unbroken record of success.
What a track record actually tells you
Less than its prominence suggests, and something real nonetheless.
Against it: teams change, market conditions differ across vintages, strategies drift, scale changes what is investable, and outcomes in any single cycle contain a large component that is not skill. A record is a small sample from a process, measured in an environment that will not repeat.
For it: a record built by the same people, in the same strategy, at a similar scale, across at least one full cycle including a downturn, with losses discussed candidly and results independently verified, is meaningful evidence about how a manager thinks and behaves.
The right use is therefore as evidence about process rather than as a forecast of returns. The most useful diligence question is not “what did you return?” but “walk me through a deal that did not work, from underwriting to exit” — because the answer reveals the reasoning, which is the thing that might actually persist.
This guide is educational and general; it is not investment advice. Past performance does not predict future results.
Frequently Asked Questions
What is a full-cycle track record?
The results of programs that have completed their life — assets acquired, held, sold, and capital returned to investors. Full-cycle results are the only ones where the outcome is settled rather than estimated, which is why they carry disproportionate weight relative to marks on programs still running.
Why are unrealised returns less reliable?
Because they depend on valuations the manager produced rather than on prices anyone paid. In illiquid strategies, valuation is judgment applied within a policy, and a manager has an interest in the result. Unrealised marks are not meaningless, but they are estimates and should be weighted accordingly against realised outcomes.
What is survivorship bias in a track record?
The distortion created when a presented record excludes programs that performed poorly, were wound down, or were transferred elsewhere. The test is whether the record covers every program the sponsor has managed in the strategy, without exclusions, and that question should be asked explicitly rather than assumed.
Does IRR overstate performance?
It can, in specific ways. IRR is sensitive to the timing of cash flows, so use of a subscription line to defer capital calls raises reported IRR without changing the underlying investments. IRR also does not indicate how much capital was deployed. Reviewing IRR alongside multiples such as DPI and TVPI gives a fuller picture than either alone.
How much does a track record actually predict?
Less than its prominence in marketing suggests. Team composition changes, market conditions differ across vintages, strategies drift, and fund size changes what is investable. A track record is evidence about process and judgment rather than a forecast, and it is most informative when the same people are running the same strategy at a similar scale.
Sources
- Global Investment Performance Standards (GIPS) maintained by CFA Institute
- Institutional Limited Partners Association (ILPA) reporting and disclosure guidance
- ASC 820, Fair Value Measurement
