Risk analysis

Sharpe ratio

The Sharpe ratio measures the excess return earned per unit of total volatility.

It is calculated as the annualized portfolio return above the risk-free rate, divided by the annualized standard deviation of returns. A higher Sharpe ratio indicates better risk-adjusted performance. A ratio below zero means the portfolio underperformed the risk-free rate.

All Sharpe calculations use TWR-adjusted daily returns, which strip out the effect of deposits and withdrawals, and assume 252 trading days per year. The default risk-free rate is 2% per annum.

Sortino ratio

The Sortino ratio is a variant of the Sharpe ratio that only penalizes downside volatility.

Rather than dividing by the standard deviation of all returns, it divides excess return by the standard deviation of negative returns only. This makes it a more precise measure of the risk that investors actually care about – the risk of losing money – without penalizing upside volatility.

As with the Sharpe ratio, the default risk-free rate is 2% per annum, and calculations use 252 trading days per year.

Volatility

Volatility measures the degree of variation in portfolio returns over time, expressed as an annualized standard deviation.

It is calculated from TWR-adjusted daily returns and annualized by multiplying by √252. Higher volatility indicates a wider range of daily outcomes and greater return uncertainty. Volatility does not distinguish between upside and downside variation.

Rolling metrics

Rolling Sharpe, rolling Sortino, and rolling volatility show how each metric evolves over time using a sliding window of returns.

At each date, the metric is recalculated using only the returns from the preceding window (default: 252 trading days). This reveals whether risk-adjusted performance is improving or deteriorating, and identifies periods of elevated or suppressed risk.

You need at least one full window of return history. If the portfolio has fewer days of history than the selected window, the chart will not be displayed.

Value at Risk (VaR)

Value at Risk estimates the maximum expected loss over a single day at a given confidence level.

VaR is calculated at the 95% confidence level using the historical simulation method: it is the fifth percentile of the observed TWR-adjusted daily return distribution. A VaR of −2% means that on 95% of days, the portfolio did not lose more than 2% of its value.

VaR does not describe the magnitude of losses in the worst 5% of cases.

Conditional Value at Risk (CVaR)

Conditional Value at Risk, also known as Expected Shortfall, measures the average loss in the worst-case tail of the return distribution.

Where VaR sets a threshold, CVaR answers: given that the threshold was breached, what was the average loss? It is calculated as the mean of all TWR-adjusted daily returns that fall below the 95% VaR threshold.

CVaR is a more complete measure of tail risk than VaR alone

Maximum drawdown

Maximum drawdown is the largest peak-to-trough decline in portfolio value observed over the selected period, expressed as a percentage.

It is derived from the cumulative TWR return series. The worst drawdowns table shows the top drawdown episodes ranked by severity, including the start date, the trough (valley) date, the recovery date if the portfolio subsequently reached a new high, and the total duration in days. Drawdowns that have not yet recovered show no recovery date.

The underwater chart shows the portfolio’s percentage decline from its most recent all-time high at each point in time.

Beta

Beta measures the sensitivity of portfolio returns to movements in a selected benchmark.

It is calculated as the covariance of the portfolio’s TWR-adjusted daily returns with the benchmark’s daily returns, divided by the variance of the benchmark returns. A beta of 1 means the portfolio tends to move in line with the benchmark; a beta above 1 indicates amplified moves; a beta below 1 indicates lower sensitivity.

Beta is only shown when a benchmark is selected. The benchmark is valued using its historical end-of-day prices aligned to the portfolio’s equity curve dates.

Correlation heatmap

The correlation heatmap shows the pairwise Pearson correlation of daily price returns between all currently held positions.

It includes only positions with a positive net quantity. Assets are excluded if their price history spans fewer than 30 days (configurable). Correlation is computed on the overlapping date range where both assets have price observations; missing dates are excluded rather than filled.

Values range from −1 (perfectly inverse) to +1 (perfectly correlated). Assets with correlation close to zero provide the greatest diversification benefit relative to each other.