Money Economy
Balance, profit and return answer different questions.
What is the account worth now?
What crossed the account boundary?
What changed through investment performance?
If an investment account grows from ¥500,000 to ¥650,000, has investing earned ¥150,000? With ¥120,000 contributed and ¥30,000 withdrawn during the period, the investment gain is ¥60,000. A balance shows what is there, not why it changed. Regular investors particularly need to separate contributions from performance.
Performance can be expressed as money gained, a time-weighted return or a money-weighted return. No single measure answers every question: investment performance and the experience of contributed money are different perspectives. Using hypothetical records, this guide works through cash flows, distributions, costs and annualisation to make results comparable and checkable rather than merely impressive.[1][2]
What this article covers
Reconcile the account before calculating a return
A basic monetary gain is ending value plus withdrawals minus beginning value and contributions. The opening example gives ¥650,000 + ¥30,000 − ¥500,000 − ¥120,000 = ¥60,000. It combines changes in valuation and income retained within the measured account. Taxes or costs already deducted from its balances are already reflected. Define the account boundary before applying the formula.
Decide whether ending value includes uninvested cash in the account as well as securities. Selling shares for cash does not remove money from the overall account. Looking only at security values would miss the sale proceeds. A whole-account measure needs a consistent boundary that includes its cash.
Align valuation dates and conventions. Combining yesterday’s foreign-market close with tomorrow’s post-deposit cash can create a balance that never existed. When time zones prevent perfectly simultaneous prices, record the convention used. Before debating small performance differences, make sure the figures can legitimately be combined.
Knowing the gain in money does not by itself establish investment efficiency. The same ¥60,000 has a different meaning when earned using ¥1 million for one year and when earned using ¥10 million for one year. Different contribution dates also mean different periods of participation. First reconcile the monetary amounts, then decide which invested capital and which period the return percentage should describe.
External cash flow depends on the measurement boundary
An external flow crosses the boundary of the assets being measured. Selling stocks to buy bonds inside one portfolio is not an external flow. Moving money from a bank account into an investment account is a contribution when measuring the investment account alone, but an internal transfer when measuring both together. Classification follows the boundary.
When combining brokerage accounts, do not count a transfer between them as new overall investment. It is an outflow and inflow in the individual accounts but cancels as an internal transfer in the combined view. Decide how to account for money in transit when neither account displays it temporarily. More accounts make transfer reconciliation increasingly important.
Not every fee payment should be added back as a withdrawal. In a net-of-cost result, investment expenses reduce performance; they differ from cash withdrawn for living expenses. Fees paid outside the account may need separate inclusion depending on the measurement scope. Treating them as ordinary withdrawals and adding them back can overstate net performance.
Use categories such as contribution, withdrawal, internal transfer, cost and investment income. Classify by purpose and boundary rather than merely observing that cash moved. A consistent rule prevents the same event from being treated as income one month and a contribution another. Reliable categorisation comes before sophisticated formulas.
Hypothetical illustration—not data for an actual product, household or company, and not a forecast. Amounts are yen. External contributions are not investment gains.
Do not lose or double count distributions
Suppose a price falls from 100 to 95 while the investor receives a cash distribution of 10. Ignoring tax, costs and additional investment, total value is 105: a 5% total return despite a 5% price decline. A distribution itself does not guarantee profit. Combine the cash received with the remaining investment value.
A distribution retained as cash may already be in the account’s ending value. Adding it again double counts it. If paid outside the measured boundary, omitting that receipt understates the outcome. Track where the distribution went and which balance already includes it.
Reinvesting a distribution buys additional units at that time’s price. Across multiple distributions, simply adding all cash at the end may differ from the actual reinvestment result. Preserve that distinction when comparing a published reinvestment-based return with an investor who received and spent the cash.
A quoted yield may refer only to distributions rather than total return including price changes. A high payout rate alone does not establish strong performance. Align remaining asset value, payments, reinvestment, taxes and costs. Regular cash receipts and an increase in total wealth are separate facts to check.
Start with a period that has no external flows
With no external flows and all income retained within the boundary, the basic return is ending value divided by beginning value minus one. ¥1 million becoming ¥1.08 million is 8%. Income paid outside the boundary needs appropriate treatment. Applying the same formula unchanged to a contribution-funded account mistakes new capital for gains.
Link subperiod returns by multiplying growth factors rather than adding percentages. A 10% gain followed by a 10% loss gives 1.10 × 0.90 − 1 = −1%. The loss applies to the increased balance. This linking principle also underlies measurement that splits an account at external flows.
Specify which capital a gain is divided by. If another ¥1 million is added a year after the initial ¥1 million, ¥2 million was not invested from the beginning. A simple gain divided by cumulative contributions ignores time invested. It may be a useful rough ratio, but should not be labelled or compared as a time-adjusted return.
Equal percentage returns can produce different gains: 8% of ¥1 million is ¥80,000, while 8% of ¥100,000 is ¥8,000. Amounts matter for progress toward a financial goal; percentages help compare different scales. Keep balance, contributions, gain, period and return in separate fields rather than treating one as the sole correct measure.
The opening example: starting balance ¥500,000, ending balance ¥650,000, deposits ¥120,000 and withdrawals ¥30,000. This isolates a gain amount, not a date-sensitive return rate.
Read the assumptions and explanation →
Time-weighted return links investment performance between flows
Time-weighted return splits the period at external cash flows and links subperiod returns. It aims to separate performance from how much the investor contributes at different times. Exact calculation requires valuations around those flows. Month-end balances alone may not adequately adjust for a large mid-month contribution.[1]
Suppose ¥100,000 grows 10% to ¥110,000. A ¥900,000 contribution then raises the balance to ¥1,010,000, after which a 10% decline leaves ¥909,000. The subperiod returns are ¥110,000 ÷ ¥100,000 − 1 = 10%, and ¥909,000 ÷ ¥1,010,000 − 1 = −10%. The contribution itself is not treated as a gain.
Linking gives a −1% time-weighted return. Yet total contributed capital is ¥1 million, so the monetary loss is ¥91,000. Those figures are not contradictory. The gain occurred on a small balance and the loss on a much larger one, making the later decline more influential on the investor’s money.
The name does not mean taking a simple average of monthly returns. It means geometrically linking growth factors across subperiods. Unequal interval lengths do not justify an equal-weight arithmetic average either. Reports labelled time-weighted may use different valuation frequencies or approximations, so check the methodology.
The order of returns changes the experience of contributed money
Reverse the return order. The initial ¥100,000 falls 10% to ¥90,000; adding ¥900,000 gives ¥990,000, which then rises 10% to ¥1,089,000. Time-weighted return remains 0.90 × 1.10 − 1 = −1%, but the monetary gain on ¥1 million contributed is ¥89,000. The increase now occurs when more money is invested.
This comparison does not imply that contributions should be perfectly timed. Wages, bonuses, living costs and major payments generate flows for reasons unrelated to market forecasts. Calling every well-timed contribution skill and every poorly timed one failure is simplistic. The point is that investment performance and the experience of an investor’s money can differ.
Even fixed regular contributions gradually enlarge the balance, so later returns can move more money than earlier ones. A volatile final year in a mature savings portfolio may dominate monetary results relative to the smaller early balances. More contribution dates do not eliminate timing’s effect. Retain the record of how much was invested when.
Money-weighted return can summarise that cash-flow experience in one rate. It includes the size and timing of contributions and withdrawals rather than isolating only the investments. It is neither universally better nor worse than time-weighted return. Choose according to whether the question concerns the investments’ path or the money committed to it.[1][2]
Money-weighted return uses both dates and amounts
Money-weighted return finds a rate that reconciles invested money with receipts at their respective times. Spreadsheet XIRR is one method for irregularly dated cash flows. It discounts each flow over its particular interval and seeks an annual rate making their sum zero. It is not simply monetary profit divided by cumulative contributions.[3]
From the investor’s perspective, money committed is negative and money received is positive. Opening assets are entered as a negative value as though invested on the measurement start date. Ending assets enter as a positive terminal value even if unsold. This closes the measurement period; it does not require an actual sale.
Date the earlier example as ¥100,000 invested on 1 January 2025, ¥900,000 on 1 July 2025 and ¥909,000 of terminal value on 1 January 2026. Under XIRR’s 365-day-year convention, the annual rate is about −15.88%. It differs from the −1% time-weighted result because it incorporates the larger amount exposed to the later decline and the time it was invested.
With the same contribution dates and amounts but a terminal value of ¥1,089,000, XIRR is about 16.59%. This is a retrospective summary of the fictional record, not an expected future yield. Exact dates differ slightly from two mathematically equal half-years. Keep dated calculations distinct from equal-period models.
↔ When needed, scroll horizontally within the table.
| Measure | Value | Meaning |
|---|---|---|
| Time-weighted return | −1% | Linked performance across cash-flow boundaries |
| Total contributions | ¥1,000,000 | ¥100,000 + ¥900,000 |
| Profit or loss | −¥91,000 | ¥909,000 − ¥1,000,000 |
Hypothetical illustration—not data for an actual product, household or company, and not a forecast.
Prepare the cash-flow record before using XIRR
Start with date, amount and category. Store dates in a form the calculation recognises, not ambiguous text. Same-day flows can be netted after preserving their classifications. Do not enter historical purchase costs again alongside opening value. For a one-year measurement beginning partway through ownership, the opening valuation is the starting point, not every earlier purchase.
Omitting positive terminal value makes still-held assets disappear from the calculation. If everything was actually sold and the receipt already recorded at the end, do not add the same terminal value again. Reconcile money committed, money received and value remaining before passing the data to a formula.
Microsoft’s XIRR requires at least one positive and one negative amount, with matching dates and amounts. It returns an annual rate, not necessarily the gain percentage over a short observation period. Inappropriate inputs or numerical conditions can produce errors or unstable results. A returned number does not validate cash-flow classification or interpretation.[3]
To check the result, discount the flows to the first date using the returned rate and verify that their sum is close to zero. Rounding may prevent exact zero; a large residual calls for checking signs, dates and terminal value. Reconcile the rate with the monetary record, especially when a large loss appears alongside an unexplained positive rate. Verification matters more than memorising the function name.
Some cash-flow patterns do not have one useful rate
An internal rate of return does not always provide one clear answer. When cash-flow signs change repeatedly—for example, an investment pays out and is followed by another obligation—multiple rates can satisfy the equation. A simple pattern of contributions followed by terminal value differs from a project including later liabilities.
Consider three annual cash flows: −100, +230 and −132. The expression −100 + 230 ÷ 1.1 − 132 ÷ 1.1² equals zero, so 10% is a solution. Using 1.2 also gives zero, making 20% another solution. This fictional non-conventional pattern does not justify choosing whichever answer looks better.
Do not change the starting guess merely to select a flattering result. Show the cash-flow pattern. A present-value comparison at an independently chosen discount rate may be useful, but that rate also requires a purpose and assumptions. For a beginner’s records, clear gains, payments, receipts and remaining obligations can be more useful than forcing an ambiguous rate.
Difficult-to-value assets pose another issue. For rarely traded holdings, uncertain realisable terminal value can dominate the return. Exact mathematics cannot remove uncertain inputs. Where valuation is a range, compare results under different terminal values. More decimal places do not make the valuation more reliable.
Annualisation is a scale conversion, not a forecast
A 20% gain over two years annualises to √1.20 − 1, about 9.54%, not the simple 10% obtained by division. It is the constant annual rate producing the same terminal value. It does not say both actual years earned 9.54%. Annualisation is a summary for comparing durations.[1]
A one-month 3% gain, scaled by assuming the same multiplier twelve times, annualises to 1.03¹² − 1, about 42.58%. It is not a forecast of next year’s profit. Short observations can generate extreme-looking annualised figures. Always retain the actual observation period and monetary outcome alongside the rate.
Dated annualisation depends on the day-count convention. XIRR’s 365-day basis does not always match a model dividing years into twelve equal months. Such differences may be small, but comparison methods should align. The first safeguard is to keep annualised, cumulative and monthly rates clearly labelled rather than mixing them in one column.
Annualising a brief rebound after a major fall can produce a very high number even while wealth remains below its earlier level. Examine cumulative results, intervening losses and payment needs, not only an attractive annualised window. An annual rate is useful shorthand, not a complete account of the journey.
Choose a benchmark that matches the role
The benchmark affects how success appears. Comparing a stock-and-bond portfolio solely with an equity index compares different risks. That comparison is not forbidden, but the entire gap cannot be called investment skill or failure. Start with the portfolio’s purpose and align the benchmark’s role.
Suppose stocks return 10% and bonds 2% over a year, with a 60/40 starting mix and no intervening trades or flows. The simple combined return is 60% × 10% + 40% × 2% = 6.8%. A similarly allocated account returning 7% is 0.2 percentage points above that reference. Comparing it instead with the stock-only 10% answers a different question.
Align period, currency, distributions and fee treatment. Comparing a net home-currency account return with a gross foreign-currency index introduces factors beyond investment decisions. An index may also be unavailable as a costless real investment. Without matching definitions, subtracting percentages can produce an arithmetically correct but conceptually unhelpful gap.
Avoid selecting a benchmark after seeing results merely because it can be beaten. Record it with the investment policy. If changing circumstances or allocation justify a new benchmark, retain the reason and date. Benchmarks need not be immutable; they should not move solely to flatter performance.
Measurement windows can change the impression
Consider a value moving from 100 to 80 and then to 96. The full-period result is −4%, while the rebound from 80 to 96 is +20%. Both calculations are correct, but the latter alone hides that the original 100 has not been recovered. A favourable window can misrepresent the overall experience.
Small changes in start or end date can alter rankings. Year-to-date, trailing-year and since-inception results answer different questions. Since-inception amounts matter for progress toward a goal; shorter windows can describe recent conditions. Use several purposeful windows rather than treating one period as a complete verdict.
When one product began later, do not compare its short history directly with another’s full history without qualification. Use a common period or state the difference. If unsuccessful products disappear from a list, the survivors alone cannot establish the success rate of the original set. Coverage is part of performance interpretation.
Keep poor months in personal records too. Gaps encourage selective memory of favourable periods. Daily attention is unnecessary, but consistent valuation and cash-flow records permit later checks over different windows. A record for evaluation serves a different purpose from a screen designed to feel reassuring.
A return does not describe the risk endured
Identical final annual returns can conceal very different paths. A decline just before a required payment may cause harm even if value later recovers. Examine drawdowns, liquidity and possible additional obligations alongside the final rate. A profitable result alone does not prove the method was suitable for the investor.
Comparing leveraged and unleveraged strategies on return alone is incomplete. A higher gain rate can come with larger losses or further funding obligations. Results also differ according to whether the denominator is equity capital or total exposure. Understand which capital the percentage refers to and what obligations accompany it.
Risk-adjusted metrics do not resolve everything. Their meaning depends on the risk definition, sample and availability of prices. Beginners may learn more from amounts, rates, duration, observed drawdowns and liquidity terms than from ranking investments by one complex statistic. Unmeasured risk is not zero risk.
Do not infer the quality of a decision solely from its outcome. A well-considered risk can turn out badly, and excessive concentration can succeed by chance. Compare results with the original reasoning, constraints and unexpected developments. Performance calculation supports evaluation; it does not certify future success.
Maintain a record that a busy person can sustain
A sustainable record need not begin with exhaustive detail. Start with dated balances, external flows, distribution destinations and cost treatment. Downloaded statements can help, but classification and duplication still need checking. Required figures can be organised locally without handing account numbers or credentials to another party.
Choose a purpose for each measure: contributions and current value for progress, time-weighted results for investment comparisons, money-weighted results for the cash-flow experience. Every metric need not be calculated monthly. Good underlying records allow later analysis; a rate saved without the reasons for flows may be hard to reconstruct.
Check three layers: record consistency, calculation consistency and interpretation. Do balances reconcile, can the result be reproduced, and does it answer the intended question? Correct records and arithmetic still permit a wrong interpretation, such as treating a short-period annualised result as a forecast. Check these layers before trusting visual polish.
Date changes to methods or corrected records. A changed figure may reflect a repaired historical flow rather than new investment performance. The objective is neither self-promotion nor competing over returns, but understanding how money changed, progress toward goals and conditions that may need reconsideration.
Separating balance from performance clarifies the next question
When a balance rises, distinguish contributions, investment gains, distributions and currency effects. When it falls, separate withdrawals from losses. This reduces the urge to label every movement success or failure. Maintaining planned saving is progress in its own right, separate from market performance.
Choose whether a rate should describe the investments or the experience of contributed money. Different time- and money-weighted results can both be valid. Read annualised rates with original periods and amounts. Align allocation, currency, distributions, costs and dates when using benchmarks. This sequence avoids placing every judgement on one attractive number.
Accurate personal measurement also improves questions brought to market analysis. Did markets rise, currencies move, allocation differ or contribution timing dominate? Do not substitute an explanation of the market for an explanation of an individual account. Inspect the investments and flows connecting the two.
Balance locates current wealth, gain states monetary change, and return provides a scale for capital and time. Used according to their roles, they create a more useful record than a dramatic performance display. No forecast is needed to separate flows from results, avoid duplicate costs and align comparisons. Durable performance literacy begins with those basic reconciliations.
Frequently asked questions
Can I use the growth of a regular-investment account as its return?
No: that includes contributions as gains. First reconcile monetary profit from ending value, withdrawals, opening value and contributions. A percentage needs a method handling flow timing. Use time weighting for the investment path and money weighting for the contributed-money experience. Balance, gain and return are distinct.
Which is correct, time-weighted or money-weighted return?
They answer different questions. Time weighting links investment performance while neutralising flow size; money weighting includes amounts and timing. Both can validly differ for one account. Flows driven by wages or life needs mean the gap is not simply skill or failure. Match method to purpose.
Should unsold investments be included in XIRR?
Yes, as positive terminal value on the measurement end date. This closes the period using their value, without recommending an actual sale. Do not add that value again if final sale proceeds are already recorded. Reconcile opening value, intervening flows and terminal value without duplication.
Can I expect the high annual rate calculated from one good month?
No. Annualisation rescales an observed period; it does not forecast repetition. It mathematically repeats the monthly multiplier twelve times, while later markets may differ or lose money. Retain the actual period, gain and risks. Do not make a short observation’s annualised figure a guaranteed planning assumption.
Should every distribution be added to profit?
It depends on whether it is already in the measured balance. Retained account cash must not be added twice; payments outside the boundary must not be omitted. Include changes in remaining asset value as well. Distinguish a published reinvestment return from an investor’s cash-spending experience.
Must I send account information to an outside service to measure performance?
No. Dates, amounts, balances and flow categories can be analysed locally without identifying information. Account numbers and credentials are not required by the formulas. Check storage and transmission when using a tool. A connection to an external service does not itself guarantee correct records or calculations.
References
- CFA InstituteRates and Returns
- CFA InstitutePerformance Reporting: Winning the Scoreless Game
- Microsoft SupportXIRR function
- U.S. Securities and Exchange Commission / Investor.govHow Fees and Expenses Affect Your Investment Portfolio
Numerical examples illustrate mechanisms under stated assumptions; unless expressly identified otherwise, they are not forecasts or results for particular products. This article provides general educational information, not personalized investment or contract recommendations. Rules, taxes, costs and contractual terms vary by jurisdiction and product.