Understanding IRR, Gross/Net and Why the Same Fund Can Produce Different Performance Numbers
Fund Performance Begins with Putting Things in Order
You cannot just leave things unfinished. You have to put them in order.
My mother used to say this to me. She passed away from a rare disease before the pandemic.
In the world of investment, “putting things in order” often means evaluating results and making them comparable. And if something is going to be evaluated, most people would probably agree that the evaluation should be as objective as possible and structured in a way that allows meaningful comparison.
The exceptions may be those who scatter capital across investments on little more than instinct, then prefer not to ask what happened because the answer might require an uncomfortable conversation with reality.
This raises a question that sits at the heart of almost everything on this blog, food aside:
What does an objective and comparable assessment of fund performance actually look like?
That was the subject of Podcast Episode 7.
The central message was that the answer depends heavily on the liquidity characteristics of the underlying investment.
An open-ended fund, such as a mutual fund or hedge fund, generally allows investors to enter and exit, subject of course to gates, market liquidity and other restrictions. A closed-ended private markets fund works differently. Investors make commitments during a defined fundraising period, contribute capital when calls are issued and, unless they sell their interests in the secondary market, remain invested until the portfolio has been realised.
Those two structures require fundamentally different approaches to performance measurement.
Unfortunately, forty minutes of one person talking about fund performance is a substantial ask. Even at double speed, it is still twenty minutes. In a world awash with short-form video, that is a fairly rich meal.
In private markets, performance measures broadly fall into three families:
- IRR, which measures the annualised return implied by the timing and size of cash flows
- TVPI, DPI and RVPI, which measure value and distributions as multiples of invested capital
- PME, or Public Market Equivalent, which compares private market performance with a public market benchmark
That sounds manageable.
But do you really think the story ends there?
Of course not. That forty-minute podcast was merely the entrance.
So let us step carefully into the deeper waters of performance measurement, where fund investment analysts occasionally find themselves questioning their life choices.
Before we begin, the principles discussed here can be applied far beyond private equity. They can be relevant to listed equities, buyouts, growth capital, venture capital, real estate, renewable energy, data centres, fund interests, collectables such as Pokémon cards and even the business of buying and reselling items through online marketplaces.
Trying to capture every possible asset class, however, would require so much abstraction that an already demanding article might become unreadable. If something becomes unclear, please resist the urge to lodge a public complaint with the author. A discreet email will do nicely.
First, What Do We Mean by “Performance”?
Let us begin with a familiar question.
Q: If management fees and carried interest did not exist, would we not see the true growth of the investment portfolio and therefore obtain a cleaner measure of the manager’s investment skill?
Broadly speaking, yes.
Fees are part of the price investors pay to gain access to a strategy, so imagining that they did not exist is not always economically meaningful. Nevertheless, if the purpose is to assess the investment team’s judgement, it is reasonable to ask a simpler question:
At what price did the manager buy, and at what value did the manager exit?
To answer that question, we would first construct a set of cash flows containing only the purchases and disposals of the underlying investments.
We might exclude brokerage, transaction charges, placement or intermediary fees and perhaps even the formation, maintenance and administration costs of acquisition vehicles created for individual investments.
That gives us a performance measure before those costs.
If we retain the relevant costs in the cash flows, we obtain a measure after costs.
In simplified terms, this is the foundation of the distinction between:
- Gross performance, before specified fees and expenses
- Net performance, after specified fees and expenses
Once the relevant cash flows have been constructed, we can use them to calculate IRR, investment multiples or, for simpler investment structures, periodic rates of return.
So far, so good.
But that calculation relates to an individual investment.
What happens when we want to calculate the IRR or investment multiple of an entire fund?
The Performance Matrix
The same fund can produce several valid performance figures because performance is being observed along different dimensions:
- LP, fund or GP
- Gross or net
- Leveraged or unleveraged
None of these perspectives is automatically incorrect. The problem begins when different measures are presented or compared as though they answered the same question.
That is the first principle of the performance matrix:
Before asking whether a number is good, ask whose number it is, what has been included and how the cash flows were constructed.
Gross and Net Are Not as Simple as They Look
One approach to calculating fund performance is to treat the fund as a portfolio and aggregate the performance of its individual investments.
That can be useful, but it is not the only approach.
In 2025, ILPA presented methodologies intended to provide a more structured basis for measuring private fund performance. Under the Granular Method, capital calls and distributions are disaggregated according to purpose, including amounts allocated to investments, management fees, fund expenses and carried interest.
Net fund performance is generally calculated from cash flows between the investor and the fund.
For gross performance, however, the treatment of management fees, fund expenses and carried interest must be made explicit. The cash flows are then adjusted according to the stated methodology.
This exposes an important distinction between two forms of “gross” performance.
The first may focus narrowly on the economics of the individual investments, excluding transaction-related costs.
The second may relate to the fund as a whole, where the costs required to establish, maintain, administer and audit the fund may be treated differently.
Those costs can include:
- fund administration
- audit fees
- legal and regulatory costs
- maintenance costs of fund vehicles
- other expenses necessary to operate the investment structure
From the perspective of a fund investor, the portfolio could not exist in its investible form without the fund structure. The costs of maintaining that structure are therefore not merely incidental to the investment. They are part of the environment in which the investment is made.
This is one reason why the performance of a particular deal presented by a GP during fundraising may differ from the fund’s reported gross IRR.
The two figures may both be valid.
They are simply not necessarily describing the same thing.
When the Portfolio Becomes a Fund, the Story Changes
At this point, another ambiguity appears.
We have been using the expression “the fund’s gross IRR”.
But do all investors in the same fund experience the same return?
In a unit trust or corporate fund, different share classes may have different economics. Within the same class, investors may experience the same periodic return, with differences in investor-level outcomes arising primarily from the timing of entry and exit.
Partnership structures are more complicated.
At first glance, the most obvious distinction among partners in a limited partnership is the distinction between the GP and the LPs.
Economically, management fees and carried interest are major sources of difference. That sounds simple until we remember that a GP interest is often not charged management fees in the same way as an LP interest. Charging management fees to the GP through the fund could, among other things, mean that the GP indirectly pays a fee to itself.
Even that apparently narrow fee exemption can produce a material economic difference over a ten-year fund life.
And that is before considering the fact that GP and LP interests differ in rights, obligations, governance and risk.
There are good reasons why strategies involving GP interests and GP-led secondary transactions have emerged. That is another interesting rabbit hole, but one for another day.
The differences are not limited to GP and LP interests.
LPs entering at the first closing and LPs entering at subsequent closings can also experience different cash-flow profiles.
“But does equalisation not make them economically equal?”
Equalisation mechanisms are intended to achieve fairness among investors under the terms of the partnership agreement, taking account of their respective exposure to the fund’s assets and liabilities.
Performance measurement, however, is driven by cash flows.
A first-closing LP may fund a series of capital calls in relatively small instalments as investments are made. A final-closing LP may contribute a much larger amount at once when admitted to the fund.
Where other factors are held constant, IRR tends to rise when the measured investment period becomes shorter. A later entrant may therefore appear to have a higher IRR because its capital has been outstanding for less time.
That does not necessarily mean the later investor received a better deal. The later investor may be required to pay equalisation amounts or additional charges intended to reflect the time and risk already borne by earlier investors.
Fairness and reported IRR are related questions, but they are not the same question.
The Same Fund, a Different IRR
Consider a simplified example.
A first-closing LP contributes through several calls as the portfolio is built. A final-closing LP contributes later, in a more concentrated cash flow. Assume both ultimately participate in the same fund and receive the same broad economic exposure.
The first LP’s capital is outstanding for longer.
The later LP’s measured investment period is shorter.
As a result, the later LP may show a higher IRR, even where the underlying fund assets and eventual proceeds are substantially the same.
This is why “the fund’s IRR” and “the investor’s IRR” must be distinguished.
If the fund has been fully liquidated, the calculation can be based entirely on realised cash flows.
During the life of the fund, however, the final leg used in calculating IRR or investment multiples is not cash. It is the investor’s or fund’s NAV.
And this is where another subtlety begins.
LP, GP and Fund Are Not Looking at the Same Thing
For an LP-level IRR, the ending NAV should reflect the value attributable to that LP interest.
If the portfolio is measured at fair value and an amount of unrealised carried interest is allocated away from the LP, the LP’s ending NAV may be reduced accordingly.
The GP’s NAV may reflect the corresponding entitlement to unrealised carry.
At the total-fund level, that unrealised carry may simply represent a reallocation of value between LP and GP interests. If so, it does not change the fund’s aggregate NAV.
Under that presentation, the fund-level IRR could be expected to fall between the LP-level and GP-level IRRs.
But this depends on the accounting.
The recognition and presentation of unrealised carried interest can vary according to:
- the applicable accounting framework
- the legal form of the fund
- the contractual nature of the carry
- the reporting entity
- the fund’s accounting policies
In some cases, unrealised carry may be recognised as an expense rather than reallocated from the LPs’ capital to the GP’s interest.
If that happens, the earlier analysis no longer applies in the same way.
The LP-level IRR may remain unchanged, while the GP-level IRR is lower because the carry has not been credited to the GP’s capital account. The fund’s total NAV may also be lower because the unrealised carry has been recorded as an expense.
Different accounting presentations can therefore produce different relationships among LP, GP and fund-level performance.
None of this is visible if we look at a single number without asking how the ending NAV was constructed.
The issue becomes even more interesting when we move from net performance to gross performance.
Adding back management fees and carried interest may increase the reported gross IRR or multiple for the LP interest and for the fund. The GP interest, which may not bear management fees on the same basis, could be affected differently.
At this point, we can identify at least six distinct combinations:
- LP gross
- LP net
- Fund gross
- Fund net
- GP gross
- GP net
Understanding what each figure means is not optional if we intend to compare it with another figure.
For funds with more than one admission date, we must also consider the investor’s entry point, the operation of equalisation and any associated charges.
An investor who argues that equalisation cannot be paid because the investment budget contains only the commitment amount may be missing an important point. Equalisation is intended to compensate for differences in timing and risk.
Funds are finance.
Sooner or later, the numbers matter.
Then Fund Finance Enters the Room
By now, the reader may reasonably feel that the meal has been substantial.
Unfortunately, the modern fund market does not stop here.
Financial engineering has entered the partnership world to address familiar objectives:
- improving capital efficiency
- managing liquidity
- smoothing capital requirements
- coordinating the timing of investor cash flows
In other words, fund finance.
From a performance measurement perspective, fund finance can change the timing at which investors contribute capital and receive distributions.
This is why some observers say that subscription facilities and similar arrangements can “distort” performance.
The underlying investment may be made on the same date and realised at the same value. But if borrowing allows the fund to delay the capital call to investors, the investor-level cash flow begins later.
A shorter measured investment period can increase the investor-level IRR.
The asset did not move.
The cash flow did.
To remove the effect of fund finance, analysts may construct an adjusted cash-flow series by incorporating the financing legs into the relevant capital calls and distributions.
Conceptually, the adjusted cash flows seek to show what might have happened if investors had funded the investment when the fund borrowed and had received a distribution when the financing was repaid.
This allows comparison between:
- reported, or leveraged, IRR and investment multiples
- unleveraged measures based on the underlying timing of investment and realisation
The difference provides an indication of how much the financing altered the reported performance.
The next question is whether that improvement was worth the interest and related financing costs.
Fund finance may also provide operational benefits. More predictable capital calls can make treasury management easier for investors. A proper assessment should consider those benefits as well.
But this is an article about why analysts cry, so let us remain with the performance problem.
Twelve Herbs, Except This Is Not a Curry
We started with six combinations:
- LP, fund or GP
- gross or net
We then added another dimension:
- leveraged or unleveraged
That gives us twelve possible performance perspectives for a single fund.
Twelve herbs in it. No, this is not a curry.
Although by this stage, the reader may feel as though a very large one has just been consumed.
For current purposes, the performance measurement framework may allow us to stop there. In practice, further questions remain.
For example:
- Is recallable return of capital included on a gross or net basis?
- How are bridge periods treated?
- Which expenses are included or excluded?
- How consistently has the methodology been applied?
- Has the methodology changed during the life of the fund?
- Are the source data sufficiently granular to reproduce the calculation?
The deeper we go, the more the performance figure begins to look less like an answer and more like the final output of a long chain of judgements.
What Matters Beyond the Performance Figure
A serious fund analysis cannot stop at performance numbers.
We must also ask:
- What operating model produced the cash flows?
- Does the manager have the systems and data needed to calculate the measures?
- Can the calculations be reproduced?
- Has the methodology been applied consistently over time?
- Can the GP or administrator explain the assumptions behind the figures?
- Can investor relations teams respond clearly when investors ask how the numbers were constructed?
These questions lead directly to operating-model due diligence.
A manager may eventually reach the limits of what can reasonably be performed in-house. The relevant question then becomes whether the GP has appointed service providers capable of delivering reliable data, accounting and reporting throughout the life of the fund.
For a GP, that becomes part of the selection criteria for administrators and other service providers.
For an LP, it becomes part of the assessment of whether the GP works with providers capable of supporting a credible reporting and investor-relations framework.
The quality of a performance figure is inseparable from the quality of the infrastructure that produced it.
What Comes After “Putting Things in Order”
Let us return to my mother’s words.
You cannot just leave things unfinished. You have to put them in order.
An evaluation requires an understanding of context, including the differences between the assumptions behind the reported figure and the assumptions held by the person reading it.
A meaningful comparison requires us to determine how far those contexts and assumptions can be aligned.
That means deciding in advance what information is required, obtaining it and making the necessary adjustments.
Otherwise, a performance figure is only a figure.
The work and cost involved in collecting the data may be wasted. Worse, the investment itself may not be understood or used effectively.
Unfortunately, systems capable of comparing and testing these differences across multiple funds are not yet easy to establish. Nor are tools that implement such analysis necessarily available off the shelf.
That is precisely why the problem is worth addressing.
A performance matrix without context
is just a number.
The path to comparison is not:
Number → Ranking
It is:
Number → Context → Adjustment → Comparison
It turns out that the instruction I heard throughout childhood applies equally well to the investment world.
Clean up after yourself.
Or, in slightly more analytical language:
Put the numbers in order before asking what they mean.
Leave a Reply