Return and statistical data

This evaluation contains the annual return and several statistical metrics for a single instrument. It is shown when a row in the correlation matrix is expanded. The same evaluation is also available in the expanded table row of the quote data feed view, where you open it through the section “Return and statistical data”.

  • Period: The return is computed over the entire span of available end-of-day prices. The period for the statistical metrics can be narrowed via the date fields of the correlation matrix. The quote data feed view does not offer these date fields, so the entire span is always evaluated there.
  • Client and instrument currency: The currency of the instrument may differ from the client’s currency (the main currency). The results are therefore presented in two columns — once in the instrument’s currency and once in the main currency. For currency pairs and Forex the main-currency column is not produced.

Return

The calculations are based on the last available closing prices within each year — ideally from December 31 of the previous year to December 31 of the year under evaluation. If dividends are available for the instrument they are added in, with the ex-date determining the year allocation. When the instrument currency differs from the main currency, GT uses historical exchange rates to produce the second column in the main currency.

Return per calendar year

The return is computed for every fully available calendar year and for the current year. The result is a percentage total performance that reflects both the price change and any dividends received.

Annualized return

The annualized return condenses the individual yearly returns via the geometric mean (Compound Annual Growth Rate). GT computes it for the standard horizons of 1, 3, 5 and 10 years, plus an additional entry covering the entire available period if that is longer than 10 years. The current year is excluded so an incomplete year does not distort the multi-year values.

Statistical metrics

The hierarchy table groups the statistical metrics by sampling period (daily, monthly, annual). Each sampling period exposes one or more properties, presented once in the instrument currency and once in the main currency.

  • Standard deviation — dispersion of the percentage changes. It is the basis for a volatility assessment and is derived from daily, monthly and annual values respectively.
  • Minimum and Maximum — the smallest and largest daily percentage change in the observation window. These two values are only reported on a daily basis.
Annual standard deviation

The annual standard deviation is not computed directly from yearly returns. Instead, it is scaled up from the monthly standard deviation using the factor √12. This makes it possible to annualize the volatility meaningfully even when only a few complete years are available.

Effect of the date restriction

The date fields of the correlation matrix affect every statistical metric — including the Sharpe ratio described in the next section. The return calculation above continues to use the entire span of available closing prices regardless. When you open the evaluation from the quote data feed view the restriction disappears entirely, because that view offers no date fields.

Sharpe ratio

The Sharpe ratio is a risk-adjusted metric that compares excess return against volatility. A higher Sharpe ratio means that the instrument has delivered more return above the risk-free rate per unit of risk taken.

Sharpe = (annualized return − avg risk-free rate) ÷ annualized standard deviation

The metric is shown per horizon in the Annualized return part of the hierarchy table — once in the instrument currency and once in the main currency. When both currencies coincide the two values are identical. When the instrument currency differs, the main-currency variant uses the risk-free rate of the client’s main currency, whereas the instrument-currency variant uses the rate of the instrument’s currency.

The average risk-free rate per horizon is the arithmetic mean of the daily rate values in the window [today − number of years, today]. If the window contains no data point at all, GT falls back to the last known rate before the end date (carry-forward).

Prerequisite: Risk-free rate mapping

The calculation needs a valid row in the Risk-free rate mapping for each currency involved. If a mapping is missing — or if the linked security has no historical closing prices yet — the Sharpe ratio remains empty in the corresponding column.

Current simplification

All horizons currently share the same annualized standard deviation — the value shown one row above in the statistics table. A horizon-specific standard deviation would require an extra database query per horizon and is planned as a future refinement.

The Sharpe ratio also stays empty when the annualized standard deviation is zero or non-finite — for example because the closing prices are constant or too few data points are available for a meaningful dispersion calculation.

Only one consumer today

At present the Sharpe ratio is shown only here. Further evaluations that consume the risk-free rate are conceivable but have not yet been implemented.