Returns tell you how far a portfolio went; risk metrics tell you how rough the road was. Capitally measures drawdown, volatility, the Sharpe ratio, Value at Risk, correlation with your benchmark and regional or sector concentration for every row of the Portfolio table — a single position, an account, an asset type, a tag — over the period you have selected, and charts them alongside your benchmarks.
Adding risk metric columns
Risk metric columns are hidden by default. Open the columns menu above the table and switch them on under Risk Metrics; they work on Positions and on every grouped table — Accounts, Types, Currencies, tags and the rest. Column choices are saved per view, so use Apply changes everywhere if you want the same risk columns on every tab.
Every column except the two diversification ones comes in two variants: the plain one, and one marked (Prices). They answer different questions, covered next.
Portfolio or Prices
The plain columns measure what your money actually went through; the (Prices) columns measure how risky the holdings you have today would have been over the whole period. Put both side by side and the gap shows how much your timing — when you bought, sold and rebalanced — changed the risk you took.
- Portfolio (no suffix) — built from the row's own historical return index. The index is time-weighted, so deposits and withdrawals never register as gains or losses, whether the project displays returns as MWR or TWR. Fees, taxes and income are counted or left out according to the returns options of the view.
- Prices — built from the price history of the row's current holdings at today's weights, as if you had held exactly that mix for the whole period. When you bought them does not matter.
Take an ETF bought three months ago, viewed over 5y. Drawdown only sees the three months you have held it. Drawdown (Prices) sees the ETF's full five years, including any crash before you owned it.
On a row grouped by a benchmark, every risk metric reads that benchmark's own price series, so you can compare your portfolio's figures with the index's in the same table. Value at Risk is the exception: an index holds no money to lose, so a benchmark row shows only Value at Risk %.
Drawdown
A drawdown is a decline from a peak. Capitally reports three views of it, all measured within the selected period — a peak reached before the period starts does not count.
- Drawdown — how far the row sits right now below the highest point it reached during the period. It reads 0% on the day of a new peak.
- Deepest Drawdown — the largest peak-to-trough fall in the period. The cell also shows when it started and ended and how many days it lasted; one that has not recovered by the end of the period reads
<start date> – ongoing. - Longest Drawdown — the longest stretch spent below a previous peak, from that peak to the day it was regained or to the end of the period. The cell has the same layout as Deepest Drawdown, and sorting the column orders rows by length in days.
The deepest and the longest drawdown are often different episodes. A sharp −30% crash that recovers in four months is the deepest; a −12% slide that takes two years to climb back from is the longest.
Widen the period to see the full episode
A drawdown that began before your selected period is measured from the period's first day, not from the original peak. If a figure looks milder than you remember, switch to max.
Volatility and Sharpe ratio
Volatility is how much the row's returns swing around their average, annualized — the standard deviation of returns scaled to a year. Sharpe Ratio is the annualized return above the risk-free rate, divided by that volatility: how much extra return each unit of volatility earned. A higher Sharpe ratio means a smoother ride for the same return.
The risk-free rate is the one set for the project in Settings → Advanced Tracking, the same value used to price options. With none set, Capitally uses the 3-month US Treasury bill rate on the last day of the period. If that rate is unavailable, the Sharpe ratio stays blank rather than being calculated against 0%.
Value at Risk
Value at Risk % is the share of the row's value that could be lost over the next 30 days with 95% confidence: in 19 thirty-day stretches out of 20, the loss should stay below it. Value at Risk is the same figure as an amount — the row's value at the end of the period times Value at Risk %.
Capitally uses the parametric model from John Hull's Options, Futures, and Other Derivatives: 1.645 × daily volatility × √30, assuming zero expected change. A holding with 25% annual volatility therefore has a 30-day 95% VaR of about 12% (1.645 × 25% × √(30/365)). Daily volatility comes from the same returns as the Volatility column, so the estimate follows the period you select: a calm year gives a lower VaR than one that includes a crash.
For a row holding only long positions the figure is capped at 100%, because a long position cannot lose more than it is worth. A row holding any short position keeps the full model figure, even above 100%, because a short's loss has no ceiling.
What Value at Risk does not tell you
VaR is a threshold, not a worst case: it says nothing about how large a loss beyond it would be. The model also assumes returns follow a normal distribution, while real markets produce extreme days more often than that, so in a crisis losses above VaR come more often than one time in 20. Read it together with Deepest Drawdown, which shows what actually happened.
Correlation with your benchmark
Correlation measures how closely the row's returns move with the project's first benchmark over the selected period: 1 means they move together, 0 that they are unrelated, −1 that they move in opposite directions. An S&P 500 ETF measured against the S&P 500 reads close to 1; a holding that reads near 0 is the one that diversifies you away from the index.
Correlation (Prices) compares today's holdings, at today's weights, against the same benchmark — how closely your current mix tracks the index, whatever you held before. Both need a benchmark on the project; on a chart, correlation is measured against the first benchmark shown on that chart.
Diversification
Regional Diversification and Sector Diversification measure how evenly the row's value is spread across the regions and sectors its holdings are assigned to, using the Herfindahl-Hirschman Index (HHI): the sum of each group's squared share. 1.0 means everything sits in one region or sector, 0.25 is an even split across four, and lower means more diversified. One over the HHI gives the number of equal-sized groups the spread is equivalent to — 0.2 behaves like five.
A portfolio 60% in the United States and 40% in Germany scores 0.6² + 0.4² = 0.52. A fund assigned to several regions contributes to each by its weight, so a world ETF spreads its value across many of them.
Value with no region or sector assigned counts as one group of its own. A row made only of unclassified holdings — property, loans, private investments — therefore reads 1.0. The index is read from the row's value at the end of the period, and stays blank when any holding in the row cannot be priced.
Risk metrics on charts
Risk metrics can be charted as well as tabulated. On the Rate of Return metric, the Drawdown chart preset (between Compare Prices and X-Ray) plots the view's drawdown across the period, with each benchmark you have selected drawn as a dashed line. To chart any other risk metric, edit a chart and pick it from the Risk Metrics group in the metric picker.
Which chart types are on offer depends on the metric:
- Drawdown and Drawdown (Prices) — line charts only.
- Deepest Drawdown and Longest Drawdown, both variants — chart the depth of that drawdown in any non-stacked chart type. A line shows how the deepest drawdown developed through the period, with benchmarks. Time-series bars and swarms give each month's or segment's own deepest drawdown. Bar, pie, X-Ray and heatmap compare groups, and a top-groups breakdown keeps the groups with the worst drawdowns.
- Volatility, Sharpe Ratio, Value at Risk %, Value at Risk, Correlation and the two diversification metrics, with their (Prices) variants — bar, X-Ray and heatmap, compared across groups for the whole period rather than over time. Correlation is offered only on a chart that shows at least one benchmark.
How the numbers are calculated
Returns are sampled daily for periods up to about two years and weekly beyond that. Weekends and holidays that bring no new price are folded into the next real move, so a market that is closed does not count as a calm day and does not flatten the volatility.
Volatility is annualized over 365 calendar days, and uses statistics that account for the length of each interval, so uneven intervals — a long weekend, a holiday gap — do not bias the result.
A position that falls to zero shows a −100% drawdown and no volatility or Sharpe ratio.
The Value at Risk and correlation calculations reproduce the worked VaR examples in Hull's Options, Futures, and Other Derivatives and the example in Excel's CORREL documentation.
Why a fund factsheet shows a different volatility
Fund factsheets usually compute volatility from monthly returns, often over a fixed three- or five-year window. Capitally uses daily or weekly returns over the period you select, which captures swings a month-end snapshot smooths over. Expect the two figures to differ.