Portfolio Performance Analytics: 7 Essential Metrics, One Worked Example

To judge how your portfolio is really doing, you need a handful of numbers, not one: your absolute return, an annual growth rate (CAGR), a time-weighted return to judge the investments, a money-weighted return (XIRR) to judge your own result, volatility, maximum drawdown, a risk-adjusted score such as the Sharpe ratio, and a comparison with a fair benchmark. Each one answers a different question. Used together, they tell you whether you were paid enough for the risk you took.

This guide explains each metric in plain English and works through all of them on one example portfolio, so you can see how the same three years can produce returns of 21.6%, 6.7%, 7.8% and 8.1%, and why every one of those numbers is "correct".

The example portfolio

Sam starts investing on 1 January 2023 and adds money once a year. Amounts are in US dollars, but every formula works the same in pounds, Singapore dollars, rupees or dirhams.

DateMoney inValue after that yearReturn that year
1 Jan 2023$100,000$112,279 (31 Dec 2023)+12.28%
1 Jan 2024$20,000$139,461 (31 Dec 2024)+5.43%
1 Jan 2025$20,000$170,200 (31 Dec 2025)+6.73%
Total$140,000$170,200

Inside each year the portfolio moved quarter by quarter. These were the quarterly returns:

  • 2023: +5%, -3%, +4%, +6%
  • 2024: -8%, +2%, +7%, +5%
  • 2025: +3%, -6%, +4%, +6%

Chaining each year's four quarters gives the yearly returns in the table. For 2023: 1.05 x 0.97 x 1.04 x 1.06 = 1.1228, so +12.28%.

1. Absolute return: how much did I make?

Formula: (current value - total money put in) / total money put in.

Sam's gain is $170,200 - $140,000 = $30,200. Absolute return = $30,200 / $140,000 = 21.6%.

Good for: a quick answer to "am I up or down?"

Weak spot: it ignores time. 21.6% over three years is very different from 21.6% over one year. It also treats the $20,000 added last January as if it had been invested for the full three years.

2. CAGR: the steady yearly growth rate

Formula: (ending value / starting value)1 / years - 1.

CAGR (compound annual growth rate) is the single yearly rate that would turn the starting value into the ending value. It is only valid when no money was added or taken out during the period.

For Sam's first year, nothing was added: ($112,279 / $100,000)1 - 1 = 12.28%. That is a clean CAGR.

Over three years it breaks. If you plug in the total invested, ($170,200 / $140,000)1/3 - 1 = 6.7% a year. That understates Sam's result, because it pretends all $140,000 was invested from day one when $40,000 arrived later. If you plug in the first deposit, ($170,200 / $100,000)1/3 - 1 = 19.4% a year, which overstates it badly, because the $40,000 of new money is counted as growth. When money moves in and out, use the next two measures instead.

3. Time-weighted return (TWR): how good were the investments?

Method: measure the return of each period between cash flows, then chain them together. Deposits and withdrawals do not affect the result.

Sam's three yearly returns chain like this: 1.1228 x 1.0543 x 1.0673 = 1.2635. That is a cumulative return of 26.35%, or 1.26351/3 - 1 = 8.11% a year.

TWR answers: "how well did this portfolio, or this manager, perform, regardless of when money was added?" That is why it is the standard for judging fund managers. The CFA Institute's GIPS standards require investment firms to present time-weighted returns in most cases. In India, SEBI requires portfolio managers (PMS) to show the time-weighted return of each strategy next to its benchmark when they publish performance.

Weak spot: it needs a portfolio value on every date that money moved. If you only know start and end values, you cannot calculate it precisely.

4. XIRR (money-weighted return): how did I actually do?

Method: XIRR finds the single yearly rate that makes all your dated cash flows add up to zero. Money in counts as negative, today's value counts as positive:

  • 1 Jan 2023: -$100,000
  • 1 Jan 2024: -$20,000
  • 1 Jan 2025: -$20,000
  • 31 Dec 2025: +$170,200

Solving gives an XIRR of 7.83% a year. In a spreadsheet, put the amounts in one column and the dates in the next, then use =XIRR(amounts, dates).

Why it is lower than the TWR of 8.11%: Sam added money just after the best year (2023, +12.28%), so the new money only earned the weaker 2024 and 2025 returns. XIRR weights each year by how much money was invested in it. TWR does not.

Neither number is wrong. TWR judges the investments. XIRR judges your result, including your timing. That is why SEBI also requires portfolio managers to show each client their own XIRR, together with the lowest, highest and median XIRR across all clients in the same strategy.

Weak spot: XIRR is only as good as the dates you give it. If you know a holding's current balance but not when you bought it, any XIRR is a guess.

5. Volatility: how bumpy was the ride?

Formula: the standard deviation of periodic returns, scaled to a year. For quarterly returns, multiply by the square root of 4, which is 2. For monthly returns, multiply by the square root of 12.

Sam's 12 quarterly returns average 2.08% and have a standard deviation of 4.98%. Annual volatility is 4.98% x 2 = about 10.0%. In a spreadsheet: =STDEV.S(returns)*SQRT(4).

Use the time-weighted period returns (the percentages), not the change in account value. The account value jumps when you add money, which is not volatility.

Good for: comparing how much two portfolios swing. A diversified stock portfolio is usually far more volatile than a bond or deposit portfolio.

Weak spot: it treats a sharp rise the same as a sharp fall.

6. Maximum drawdown: what was the worst fall?

Formula: the largest fall from a peak to a later low, before a new peak, as a percentage of the peak.

Build a growth index from the quarterly returns, starting at 1.00. Sam's index reached 1.1228 at the end of 2023, then fell 8% to 1.0330 in the first quarter of 2024. Later it fell 6% in the second quarter of 2025. The maximum drawdown is -8.0%.

Now look at the account balance instead. It went from $112,279 at the end of 2023 to $121,697 at the end of the first quarter of 2024. It looks like a gain, because the $20,000 added on 1 January hid the 8% fall. This is the most common mistake people make with drawdown: new deposits hide losses. Always measure drawdown on returns, not on balances.

Good for: knowing whether you could live through the worst period without selling in a panic.

7. Sharpe ratio: was the return worth the risk?

Formula: (annual return - risk-free rate) / annual volatility.

William Sharpe introduced this measure in 1966 as the "reward-to-variability ratio". The risk-free rate is what you could have earned with no risk, such as a short-term government bill in your currency. Assume 4% for this example:

(8.11% - 4.00%) / 9.96% = 0.41.

Higher is better. A Sharpe ratio means most when you compare two portfolios over the same period, in the same currency, with the same risk-free rate. Sharpe himself pointed out that it ignores how a holding moves with the rest of your money, so do not use it alone.

A related measure: the Sortino ratio uses only downside volatility, so it does not penalise sharp rises.

Benchmark comparison: compared to what?

A return only means something next to a fair alternative. Pick a benchmark that matches what you hold: a broad stock index for a stock portfolio, a mix of stock and bond indices for a balanced one, in the same currency you measure in.

Suppose Sam's fair benchmark returned +15%, +8% and +4% in the three years. Compare year by year:

YearSam (TWR)BenchmarkDifference
202312.28%15.00%-2.72 points
20245.43%8.00%-2.57 points
20256.73%4.00%+2.73 points
3 years, a year8.11%8.91%-0.80 points

The benchmark's three-year figure is 1.15 x 1.08 x 1.04 = 1.2917, or 8.91% a year.

There is a second, very practical test: what if Sam had put the same money into the benchmark on the same dates?

  • $100,000 x 1.15 = $115,000. Add $20,000: $135,000.
  • $135,000 x 1.08 = $145,800. Add $20,000: $165,800.
  • $165,800 x 1.04 = $172,432.

Sam has $170,200, so the simple index alternative would have left Sam $2,232 better off. If Sam's portfolio costs more to run than an index fund, that gap is the real question to ask.

What the numbers say together

MetricSam's figureQuestion it answers
Absolute return21.6% ($30,200)Am I up or down?
Total invested annualised6.7% a yearMisleading when money was added later
Time-weighted return8.11% a yearHow good were the investments?
XIRR7.83% a yearHow did my money actually grow?
Volatilityabout 10.0% a yearHow bumpy was it?
Maximum drawdown-8.0%What was the worst fall?
Sharpe ratio0.41Was the risk worth it?
Versus benchmark-0.80 points a yearWould a simple index have done better?

Sam made a real profit with moderate swings, but trailed a simple benchmark by a small margin. That is a useful conclusion. No single metric would have shown it.

Common mistakes

  • Counting deposits as returns. If your balance grew from $100,000 to $170,200, that is not a 70% return when you added $40,000.
  • Annualising very short periods. A 3% gain in one month is not "42% a year". Annual rates for periods under a year are mostly noise.
  • Measuring drawdown on balances. Deposits hide falls. Use returns.
  • Picking an easy benchmark. A stock portfolio compared with a savings account will always look good.
  • Mixing currencies. A US fund can rise in dollars and fall in pounds or rupees. Measure everything in the currency you will spend.
  • Comparing XIRR with someone else's TWR. They measure different things.
  • Trusting a return built on guessed dates. If the purchase dates are not real, the annual return is not real either.

What Agni Folio shows you

Agni Folio calculates the return figures from your own transactions, converted to your display currency on the server:

  • Absolute return and gain or loss, for each holding and for the whole portfolio.
  • XIRR, for each holding and for the whole portfolio, from your dated purchases, sales and dividends.
  • Period returns for the whole portfolio: 1 day, 1 week, 1 month, year to date, 1 year and since start.
  • A Time-Adjusted figure for each holding: the overall gain annualised over the time since the first purchase. It is not shown for holdings held under 30 days.
  • Charts of your portfolio value over time, and an Annual Returns page.

XIRR is shown only when it is real. XIRR needs real purchase dates. When a holding came from a statement that shows only a closing balance, with no dates of purchase, Agni Folio does not invent a date. That holding shows no XIRR, and the portfolio XIRR leaves it out rather than include a made-up figure. Add the real purchase transactions and its XIRR appears.

Agni Folio does not currently calculate volatility, maximum drawdown, Sharpe ratio or a benchmark comparison. You can work them out from the formulas above in a spreadsheet, using your portfolio's monthly or quarterly returns.

For more, read how the portfolio insights work and how to record transactions so XIRR has real dates. If someone manages money for you, see how to track PMS portfolio performance or track a separately managed account, and how to track all your investments in one place.

Frequently asked questions

What is a good return for a portfolio?

There is no single number. A good return is one that beats a fair benchmark for the same kind of investments, over several years, after fees, without more risk than you can live with. Compare like with like: a stock portfolio against a stock index, in the same currency.

What is the difference between XIRR and CAGR?

CAGR works only when there is one amount at the start and one at the end, with nothing added or taken out. XIRR handles any number of deposits and withdrawals on any dates. With a single deposit and no withdrawals, the two give the same answer.

Should I use time-weighted or money-weighted return?

Use time-weighted return to judge a fund or a manager, because they do not control when you add money. Use money-weighted return (XIRR) to judge how your own money grew. If the two differ a lot, your timing made a big difference, for better or worse.

How do I calculate maximum drawdown in a spreadsheet?

Build a growth index: start at 1, and multiply by (1 + return) for each month or quarter. In the next column, track the running peak, for example =MAX($B$2:B2) copied down. Drawdown for each row is index / peak - 1. The smallest (most negative) value is your maximum drawdown.

What is a good Sharpe ratio?

Higher is better, but the number depends heavily on the period and the risk-free rate you use. Sharpe's own example of a typical stock market gives 0.40. Use the ratio to compare two portfolios over the same years rather than to grade one against a fixed target.

Why is my XIRR negative when my holding is up?

It usually is not, unless the dates are wrong. Check for a purchase dated too early or too late, a missing sale, or a dividend that was recorded twice. A large purchase made just before a fall can also make XIRR lower than the simple return.

Related Articles