Risk Management

The TWRR module allows you to evaluate the actual performance of the management of a portfolio, sterilizing the distorting effect of investment and disinvestment flows, as happens in the methodologies for calculating the value of the share of funds. Aggregation criteria for the calculation can be freely established by the user, who can analyse the trend of:

• an entire portfolio;
• single holdings;
• metaassets consisting of holdings belonging to one or more portfolios.

TWRR yield calculation and calculation launch can be automated and scheduled as often as desired.
Each aggregate being calculated by the TWRR can be associated with a benchmark (either a market index or an ad-hoc compound benchmark) and a risk-free asset. This allows you to combine return measures with risk, relative performance and risk-adjusted performance measures, including:

• average return;
• money weighted return;
• volatility;
• Sharpe index;
• Sortino index;
• tracking error and tracking error volatility;
• beta, alpha, and Jensen's beta, and their standard deviations.

An essential prerequisite for calculating the TWRR is that market values exists for each holding included in the valuation. In the case of holdings on illiquid bonds or OTC derivatives, the valuation can be carried out using the fair price calculated by Sofia.

The Time Weighted Return module incorporates some Contribution analysis tools: once the TW return calculation procedures have been launched it is possible to break down the return of each portfolio (or metaasset) into the sum of the contributions of subaggregates of holdings.
The analysis can be drilled down to details of the individual holding.
 

It is possible to conduct performance attribution analyses on defined aggregates, highlighting the effect of asset allocation and stock selection and thus identifying the determinants of portfolio over-return with respect to the benchmark.

The basis of Sofia's ex-ante risk analysis modules are the Fair Price calculation functions of bonds and derivatives.
The information needed for the theoretical price calculation is taken from the financial master data of each security (flow dates, indexation rules, flow discount curve, etc).

To support the growing need to price and manage Inflation Linked securities prospectively, it is also possible to calculate the Fair Price for these instruments, associating these securities with the appropriate inflation index and matching the index with the related inflation curve.

The Fair Price calculated by Sofia is written in the price list and can be used to evaluate holdings.
When the theoretical price calculation functions are used by the Stress test and VaR modules, the market curves are replaced with simulated curves.
 

The module allows the analysis of portfolio behaviour in response to user-assumed market shocks.
The risk factors considered are interest rates, stock market indices, exchange rates and credit spreads. The user can apply parallel shift, tilt and twist to interest rate curves. Curves can be modified at the single node level.

The model provides summary information on portfolio loss and analytical data regarding single holdings.
The simulation engine is based on a methodology of full repricing of the portfolio rate component, thus allowing to construct the theoretical price set of each asset after the application of the hypothesized scenario. To carry out the stress analysis of equity instruments it is then possible to vary the prices of the equity stock market indices in percentage terms, through the calculation of the historical equity beta.

It is also possible to vary the exchange rate, considering stress on currency exchange rates.
In addition to the evaluation of the portfolio after the shock, it is also possible to obtain the calculation of its future flows – holding by holding – by indexing future coupons based on the set scenario.

Finally, it is possible to conduct analyses “What if” carried out on portfolios defined by the user and imported into Sofia via simple Excel sheets. The Stress Test module includes functions for calculating the effective duration of bonds and rate derivatives.
 

VaR is available both with parametric calculation methodologies and with historical-simulation methodologies.

The calculation methodology is applicable to both entire portfolios, as well as user-defined metaassets. 

Parametric

The parametric VaR is based on the standard RiskMetrics™ model however combined with the principle of portfolio aggregation, which is why it is not necessary to have the variance-covariance matrices of the risk factors, but only the historical series of the prices of the securities in portfolio.

Historical – simulated

The functionality for calculating Value at Risk with a historical-simulation model, allows evaluating the potential portfolio loss based on the distribution of portfolio losses simulated on scenarios constructed by applying to risk factors the historical variations that occurred in the time period set by the users.

Compared to the parametric VaR method, the evaluation of value at risk is based on the Stress test techniques and Sofia's Fair price engine.

In addition to VaR, Sofia calculates other indicators such as Expexcted Shortfall, Marginal VaR, Incremental VaR, both of the individual holdings and of the metaassets.
Finally, it is possible to conduct analyzes “What if” on external wallets.