Returns that explain themselves.
X-PERFORM computes time- and money-weighted returns with every valuation, shows them gross and net of fees, and breaks them down by position, group and decision – against the benchmark you assign.
- 9
- windows stored with every valuation
- 17
- risk and ratio figures per window
- 3
- attribution methods, each with its own identity
- 12
- charts in the performance report
Three questions a portfolio manager asks
- Question What did we earn, gross and net?
Three return tiers: before management fees, after them, and after custody, audit and administration. The basis is disclosed with every result.
- Question Where did the return come from?
Contribution per position and per asset class, sector, country and currency – adding up to the portfolio return.
- Question Did our decisions beat the benchmark?
Allocation and selection by Brinson-Fachler or geometrically, currency decisions by Karnosky-Singer.
Returns in three tiers, with the basis disclosed.
Time-weighted returns are linked from one valuation to the next, with external flows weighted to the day. The money-weighted return is solved per period. Fees are deducted before the return is computed, so net means net.
- Time-weighted return by linked Modified Dietz
- Money-weighted return per valuation period
- Gross, net and net-total return on every result
- Subscriptions and redemptions converted at the rate of their day
- Annualisation policy set per portfolio
- Local return and currency effect per position
Derivatives are split into underlying, currency and financing, so that their return lands where the exposure is.
Seventeen figures per window, not one Sharpe ratio.
Risk and ratio figures are stored with every valuation for nine windows: one and three months, year to date, one and three years, the last three calendar years and since inception.
- Volatility, downside deviation and maximum drawdown
- Sharpe, Sortino and Calmar ratio
- Tracking error and information ratio
- Beta, alpha, Treynor ratio and M²
- Up and down capture, R²
- Risk-free rate per currency: SARON, SOFR, €STR, SONIA, TONA, SWESTR
If no risk-free series is loaded, the figure is computed with zero – and both screen and report say so.
Benchmarks, assigned over time.
Each portfolio has a primary benchmark and any number of secondary ones, each with its validity period. Benchmark data is never overwritten: every valuation records the rows it read.
- Composite benchmarks with dated weights, nested up to ten levels
- Attribution against each component of a composite
- Another X-FOLIO portfolio as benchmark
- Returns by CSV, by hand or from the data provider
- Periods must match exactly – nothing is interpolated
- Changed benchmark data marks the affected valuations as stale
Contribution: who carried the return, and who cost it.
A position’s contribution is its opening weight times its return, net of what was bought and sold in the period. Groups roll up without a remainder.
- Per position for every period
- Per asset class, sector, country and currency
- Over several periods at group level, linked so that it reconciles exactly
- Works without a benchmark
- Ten largest contributors and detractors in the report
Attribution by three methods.
Choose benchmark, grouping, method and window. Each method meets its own identity – additive, multiplicative, or with currency as a decision of its own.
- Brinson-Fachler
Allocation and selection per sector, country or asset class, adding up to the excess return.
- Geometric
Allocation and selection that multiply to the relative return and chain over periods without any smoothing.
- Karnosky-Singer
Local allocation, local selection and currency allocation for multi-currency portfolios, with a risk-free rate per currency.
Over several periods, effects are linked with the Menchero, Cariño or GRAP method. A group held outside the benchmark is measured against the benchmark’s total return, so no allocation effect appears that nobody decided.
Every return traceable to its inputs.
A performance figure is only as good as the trail behind it. X-PERFORM keeps that trail with the result.
- 01Recalculations are kept
- A new run is stored next to the previous result. Readers see the current one; the earlier one remains as audit trail.
- 02Engine version
- Each result names the calculation version that produced it.
- 03FX audit
- The rates used for every conversion are listed in the report.
- 04Benchmark snapshot
- Each valuation records which benchmark rows it used.
- 05Cash-flow bridge
- Opening value, flows and result reconcile to the closing value, with a check line.
- 06No partial figures
- Missing benchmark data holds the valuation with a to-do instead of producing half a result.
A report in thirteen parts, and the views behind it.
The report goes to clients and the board. The views are for those who have to stand behind the figures.
- PDF report
Performance report
Rendered after every valuation in the report languages of the fund, once its preflight checks have passed.
- Summary with eight key figures, the method and the concentration
- Returns: trailing periods, cumulative and excess charts, calendar years
- Cash-flow bridge with reconciliation
- Largest holdings, ten best and ten weakest contributors
- Contribution by asset class and by currency
- Risk figures over every stored window, with the drawdown chart
- Attribution by sector and by country, and a Karnosky-Singer page
- Monthly returns as a heat grid
- Benchmark and FX audit
- Costs and return tiers, cost ratios per window
- Methodology and a glossary of 21 terms
- On screen
Performance view
Five tabs, per portfolio and per valuation.
- Overview: returns, tiers, values, flows and the risk figures per window
- Drivers: contribution by position and group
- Changes: comparison with an earlier period
- Attribution: benchmark, grouping, method and window to choose
- History: all results, cumulative chart with secondary benchmarks
- On screen
All portfolios
The overview across the house.
- One row per portfolio with returns and risk figures
- Risk and return in one scatter
- Filtered, sorted and exported like every table
- PDF report
Blocks for your own documents
Five performance sections for the report editor of X-REPORT.
- Monthly performance
- Performance summary
- Attribution by sector and by country
- Risk figures
Measured with every valuation.
Performance is not a month-end job. It is committed together with the valuation it belongs to.
- 1 Value
The valuation and its performance are stored in one transaction.
- 2 Measure
Flows are identified; time- and money-weighted returns are computed in three tiers.
- 3 Compare
The benchmark rows of the period are read and recorded.
- 4 Explain
Contribution, attribution and risk figures are stored for nine windows.
- 5 Report
The report is rendered once its preflight checks pass, then delivered.
Methods, by name.
Methods named, not only promised.
Returns
- Linked Modified Dietz
- Money-weighted return per period
- Gross, net, net total
- Annualisation policy per portfolio
- Local return and currency effect
- Dividends counted from the ex-date
Windows
- 1 and 3 months, year to date
- 1 and 3 years
- Last three calendar years
- Since inception
- Report period chosen per portfolio
Risk figures
- Volatility, downside deviation, drawdown
- Sharpe, Sortino, Calmar
- Tracking error, information ratio
- Beta, alpha, Treynor, M²
- Up and down capture, R²
- Source of the risk-free rate shown
Benchmarks
- Primary and secondary, with validity
- Composite, nested to ten levels
- A portfolio as benchmark
- CSV, manual entry, provider
- Index type shown: price, gross, net
- Never overwritten, snapshot per valuation
Contribution and attribution
- Contribution per position and group
- Brinson-Fachler
- Geometric (Burnie-Knowles-Teder)
- Karnosky-Singer, residual stored
- Menchero, Cariño, GRAP linking
- By sector, country, asset class
Costs
- Cost ratio per window and category
- Computed and bank-charged costs kept apart
- Three return tiers
- Cost basis disclosed per result
Not in the module today Fixed-income attribution, GIPS composites and dispersion, return series per share class, and rolling-window analysis.
