Performance Attribution
A breakdown of a fund's NAV change into its drivers — price moves, new/exited positions, flows, and fees — so a return can be explained, not just reported.
Definition
Performance attribution breaks the change in NAV over a period into its component drivers, rather than leaving investors with a single headline return figure. A typical decomposition separates the price effect (the gain or loss on positions held throughout the period), the contribution from new positions opened during the period, the contribution from positions exited, net flows (subscriptions minus redemptions, which move NAV without reflecting investment performance), and fees charged against the fund.
Because each of these components is measured in dollars and they are constructed to sum back to the total change in NAV, attribution turns "the fund was up 12% this month" into "of that 12%, most came from price appreciation on existing positions, a smaller amount came from new positions added mid-month, and management fees were a modest drag." A well-built attribution will also carry a small residual — a disclosed plug for timing effects and rounding that could not be cleanly assigned to another bucket — rather than silently forcing components to tie without it.
Attribution is distinct from exposure reporting: exposure describes what the book is positioned in right now, while attribution explains what actually drove the P&L that already happened. The two are complementary — a fund with high concentration in one asset class should expect that class to dominate its attribution, and if it does not, that mismatch itself is worth investigating.
Why it matters
A single return number cannot tell an LP whether performance came from genuine skill (correctly sized, well-timed positions) or simply from being long a rising market, from a handful of new subscriptions arriving right before a good month, or from marking gains that later reverse. Attribution lets an investor separate these stories, and lets a manager demonstrate that the return they are reporting is explained, not asserted.
It also functions as an internal control: if the sum of the attribution components does not reconcile to the actual change in NAV within a small, disclosed residual, that is a signal something in the books — a mispriced position, a missed flow, an unrecorded fee — needs to be found before the number goes to investors.
Building the decomposition
Attribution is computed between two NAV observations — typically two NAV runs a period apart. Each position held at the start is repriced to compute its price effect; positions that did not exist at the start but exist at the end are isolated as new-position contributions; positions that existed at the start but not at the end are isolated as exit contributions; capital movements (subscriptions and redemptions) are pulled from the capital account ledger rather than inferred from NAV movement; and fees are pulled directly from the fee accrual.
The components are constructed so that Price Effect + New Positions + Exits + Net Flows − Fees + Residual equals the actual change in NAV exactly — attribution is a reconciliation exercise as much as an analytical one, and a decomposition that does not tie to the real NAV delta is not trustworthy regardless of how plausible its individual pieces look.
NAV-delta attribution identity
Reconciling a month of NAV movement
A fund opens the month at a NAV of $10,000,000. Over the month: existing positions gain $620,000 from price movement; new positions opened mid-month contribute $150,000; positions exited before month-end cost the fund $40,000; net subscriptions bring in $500,000 of fresh capital; management and performance fees total $45,000; and a $15,000 residual absorbs timing effects from marks taken on different days.
Summing the components gives the actual change in NAV exactly: $620,000 + $150,000 − $40,000 + $500,000 − $45,000 + $15,000 = $1,200,000, so the fund closes the month at $10,000,000 + $1,200,000 = $11,200,000. Note that $500,000 of the $1,200,000 increase is simply new investor capital arriving — an LP who only saw "NAV up $1.2M" without the attribution could easily overstate how much of that was genuine performance.
| Component | Amount |
|---|---|
| Opening NAV | $10,000,000 |
| Price effect | +$620,000 |
| New positions | +$150,000 |
| Exits | −$40,000 |
| Net flows | +$500,000 |
| Fees | −$45,000 |
| Residual | +$15,000 |
| Closing NAV | $11,200,000 |
Common mistakes
Reporting a headline return without separating out net flows — a month with large subscriptions can look like a strong performance month even if the underlying positions barely moved.
Forcing the components to tie to ΔNAV without a disclosed residual, which hides real reconciliation breaks rather than surfacing them — a residual should be small and stated, not silently absorbed into the price-effect bucket.
Attributing P&L to "new positions" and "exits" using the fund's current holdings list rather than what was actually held at each point in the period, which misclassifies trades that were both opened and closed within the same window.
Confusing attribution with exposure analysis — attribution explains realized P&L drivers over a past period; exposure describes point-in-time positioning and says nothing about what caused past returns.
Treating attribution as a substitute for an independent NAV — attribution can only decompose whatever NAV numbers it is given, so an attribution built on a wrong NAV simply produces a wrong, but plausible-looking, decomposition.
In practice
Crypto funds see attribution swing more sharply between periods than traditional strategies, because a single asset can move double digits in a day — a month's "price effect" component can be dominated by two or three days rather than spread evenly, which is worth calling out in commentary alongside the numbers.
Nyx Fund's risk engine computes this decomposition between any two NAV runs directly from the fund's ledger — price effect, new positions, exits, net flows, and fees are pulled from the same records that produced the NAV itself, with the residual disclosed rather than hidden, so the attribution always ties exactly back to the real change in NAV.
Preview what a full monthly investor letter looks like — including the manager-commentary narrative where performance drivers are typically explained in prose — with the free LP report preview before connecting a real book.
Questions, answered
What is performance attribution in a hedge fund?
Performance attribution is a breakdown of the change in a fund's NAV into its underlying drivers — price movement on existing positions, new and exited positions, capital flows, and fees — so a headline return can be explained rather than taken on faith.
Why does attribution separate out net flows?
Subscriptions and redemptions change NAV without reflecting any investment decision. Separating net flows from price effect prevents a month with large new subscriptions from being mistaken for a month of strong trading performance.
What is the residual in a performance attribution?
The residual is a small, disclosed plug that captures timing effects and rounding the other components cannot cleanly absorb — for example, marks taken on slightly different days. A trustworthy attribution states its residual rather than forcing the other components to tie without one.
Is performance attribution the same as exposure analysis?
No. Attribution explains what already happened to NAV and why; exposure analysis describes what the book is positioned in right now. A fund can have concentrated exposure and still show attribution spread across several drivers, depending on which positions actually moved.
Your next LP report,
on autopilot.
14-day free trial · No card required · $999/month after