The time-weighted return (TWR), also called the time-weighted rate of return, measures how an investment portfolio has performed over time while removing the effect of money flowing in and out of it. It strips out the impact of capital calls and distributions so that the figure reflects the performance of the underlying investments alone, not the timing or size of an investor's contributions.
This is what makes TWR useful for judging a manager or a strategy. A portfolio's value can rise simply because an investor added capital, or fall because they withdrew it. Neither tells you whether the investments themselves did well. By isolating market and selection performance, TWR answers a cleaner question: how did the portfolio grow, dollar for dollar, over the period?
TWR is the performance standard for vehicles where the manager does not control the cash flows: public funds, indices, and, in private markets, evergreen or open-ended structures. It contrasts with money-weighted measures such as the internal rate of return (IRR), which deliberately factor those flows in.
The mechanics rest on a single idea: every time money enters or leaves the portfolio, the clock resets.
Each external cash flow (a contribution, a withdrawal, a subscription, a redemption) marks the end of one sub-period and the start of the next. Within each sub-period, the portfolio's return is calculated on its own. The sub-period returns are then multiplied together, or "geometrically linked", to give the return for the whole period. This compounding is why TWR is sometimes called the geometric mean return.
The point of resetting at each flow is to value the portfolio immediately before new money distorts the picture. A capital call or a large deposit changes the size of the portfolio without saying anything about its performance. By cutting the timeline at that moment and measuring each slice separately, TWR keeps those flows out of the result. Sub-period valuations rely on the portfolio's net asset value (NAV) at each cut-off.
The cumulative TWR is the product of one plus each sub-period return, minus one:

Where:
Each sub-period return is calculated as:

The steps are straightforward:
This gives a cumulative return for the whole period. To compare across periods of different lengths, you can annualise it separately: TWR (annualised) = (1 + TWR)^(1/y) − 1, where y is the number of years. Annualisation is an optional final step, not part of the core calculation.
The following figures are illustrative only. Suppose an investor starts the year with a portfolio worth $100,000.
Linking the two sub-periods:
TWR = [(1 + 0.10) × (1 − 0.05)] − 1 = (1.10 × 0.95) − 1 = 0.045 = 4.5%
The $50,000 arrived just before the portfolio declined, but it does not drag the TWR down. That is the whole point: TWR measures how the investments performed in each period, up 10% then down 5%, independent of the fact that a larger sum happened to be exposed to the second, weaker period.
The contribution's timing is exactly what a money-weighted measure such as IRR is designed to capture. Run the same flows through an IRR calculation, with $100,000 invested at the start, a further $50,000 mid-year, and $152,000 at the end, and the result is markedly lower, around 1.6%.
Why the gap? More of the investor's capital was at work during the losing second half. The time-weighted view (4.5%) reflects the portfolio's performance; the money-weighted view (≈1.6%) reflects the investor's actual outcome, given when and how much they put in. Neither is wrong. They answer different questions:
IRR is the same idea as the money-weighted rate of return. For investments with irregular flows, the modified internal rate of return (MIRR) refines it further by applying separate financing and reinvestment rates.
In private markets, the metric a fund reports depends on its structure, and this is where TWR earns its place alongside IRR.
Closed-ended funds report IRR. Capital is called gradually over the investment period, so the timing of those calls and of later distributions matters a great deal. IRR captures that timing, including the J-curve, the early dip when fees and unrealised holdings weigh on returns before exits begin to materialise.
Evergreen and open-ended funds report TWR instead, and for good reason. These structures put capital to work from the first subscription, price their NAV continuously, and reinvest gains inside the vehicle. TWR's underlying assumption of continuous full investment actually holds. Investor subscriptions and redemptions happen at periodic windows on the investor's own timing, so stripping those flows out leaves a clean measure of how the fund itself performed.
That leaves a practical question: how do you compare a closed-ended fund reported on IRR with an evergreen fund reported on TWR? Here, the multiple on invested capital (MOIC), ending value divided by total capital invested, provides a common benchmark that is independent of cash-flow timing. In short: TWR for evergreens, IRR for closed-ended funds, and MOIC as the timing-neutral comparator across both.
Suggested read: Fund structures by liquidity
TWR is the international standard for reporting and comparing investment performance. The Global Investment Performance Standards (GIPS), maintained by the CFA Institute, prescribe it for managed products precisely because it removes the cash-flow effects that managers do not control fully. Market indices are quoted on the same basis, which is what allows a like-for-like comparison between a fund and its benchmark, the same comparability that underpins measures such as the public market equivalent (PME).
For investors, the practical value is consistency. Because TWR is calculated the same way regardless of who contributed what and when, it lets you compare two managers, or a manager against an index, on equal terms.
TWR is not the right tool for every question.
Is a higher time-weighted return better? Generally yes. A higher TWR means stronger compound performance over the period. But always compare it against a relevant benchmark and over the same time horizon, and remember it says nothing about how much capital was at work.
Is TWR annualised? Not by default. The core formula produces a cumulative return for the whole period. Annualisation is a separate, optional step used to compare periods of different lengths.
What is the difference between TWR and CAGR? Both express compound growth. TWR is built to handle external cash flows by linking sub-periods, whereas a compound annual growth rate typically assumes a single sum left untouched between two dates.
Why don't closed-ended private equity funds use TWR? Because the timing of capital calls and distributions is central to those funds' returns, and TWR deliberately removes it. IRR keeps that timing in, which is why it is the standard for drawdown structures.
