Loading
WSU Lexicon: Quantitative Methods
The time-weighted return is the compound growth rate of a single unit of currency invested in a portfolio, obtained by chain-linking the holding period returns of each subperiod between external cash flows. It neutralises the size and timing of client contributions and withdrawals, isolating investment performance.
Its counterpart is the money-weighted return, which is the portfolio's internal rate of return.
The entry
The construction proceeds in three steps. First, the portfolio is valued immediately before every external cash flow, which cuts the measurement period into subperiods whose boundaries are defined by contributions and withdrawals rather than by the calendar. Second, a holding period return is computed for each subperiod: ending market value plus any income distributed, less beginning market value, all divided by beginning market value. Third, the subperiod returns are chain-linked by adding one to each, multiplying the resulting growth factors together, and subtracting one from the product.
Because the linking is multiplicative, the result compounds rather than averages, and a large cash flow arriving at a favourable or unfavourable moment cannot inflate or depress the figure. If the total measurement window spans several years, the chain-linked result is converted to an annual figure geometrically by taking the appropriate root of the cumulative growth factor and subtracting one.
The entry
Portfolio managers rarely control when clients add or remove money, so any measure sensitive to cash flow timing would judge them on decisions they did not make. This is why time-weighted returns are the default basis for the performance figures presented in fund factsheets, separately managed account reports and GIPS-compliant composites, and why they are the figures used when ranking managers against a benchmark or against peers. The benchmark itself is a chain-linked index return, so comparability requires the portfolio to be measured the same way.
Daily valuation makes exact subperiod pricing straightforward for liquid portfolios. Where valuations are less frequent, approximations that weight each cash flow by the fraction of the period it was invested are used instead. The main recognised exception is private capital and similar strategies in which the manager controls the drawdown and distribution schedule; there, a money-weighted measure is the more faithful description of performance.
The entry
The two measures answer different questions and both can be correct simultaneously. The time-weighted figure answers how the strategy performed per unit of currency exposed to it throughout the period. The money-weighted figure, an internal rate of return, answers what the specific pattern of investment actually earned, and therefore embeds the investor's allocation timing alongside the manager's security selection.
The practical consequence is a systematic wedge whenever flows are trend-chasing: money arrives after strong runs and leaves after drawdowns, so the capital-weighted experience lags the reported strategy record. In evaluation work the two are used together, with the time-weighted return assessing the manager and the money-weighted return assessing the client relationship or the advisor's guidance on when to commit capital.
The entry
Expect to be given a short sequence of portfolio values with contributions or withdrawals stated at specific dates, and to be asked for the return over the whole window. The mechanical requirements are to place each cash flow correctly relative to the valuation, to avoid treating a contribution as investment gain, to include dividends and interest in the numerator of the relevant subperiod, and to link rather than average.
Two refinements are commonly tested. First, the arithmetic mean of the subperiod returns exceeds the chain-linked result whenever those returns vary, so an arithmetic average is not a valid time-weighted return. Second, annualisation must be geometric; multiplying a quarterly figure by four is incorrect. Candidates should also be able to state in one sentence which measure a question is asking for once it names either the manager's skill or the investor's own outcome.
Common trap
It does not. TWR deliberately strips out the effect of contributions and withdrawals, so it measures the manager's performance per unit invested. The investor's realised experience, which depends on how much was invested and when, is the money-weighted return.
In practice
CGM Focus posted one of the strongest time-weighted records of the 2000s decade, yet the typical shareholder lost money: investors piled in after big gains and redeemed after losses, so the average money-weighted experience was sharply negative.
Candidates ask
See also
One entry from the WSU Lexicon every day. Browse the full archive.