Credit Models: Apply observation weights for PD
You can use observation weights for training data in credit models for probability
of default (PD). The following PD objects have a name-value argument for
WeightsVar:
The following PD functions include a new output column:
DiscMeasure output in modelDiscrimination includes a
WeightedCount column.
CalData output in modelCalibration includes a
WeightCount column.
For an example using WeightsVar, see Create Weighted Lifetime PD Model.
Market Risk: Include dates, plots, and exceptions in VaR backtests
You can use additional value-at-risk (VaR) capabilities with the varbacktest object that include visualizations, exceptions reports,
time windowing, and data-appending. In addition, the varbacktest object accepts a Time name-value
argument. Using a varbacktest object, you can:
Create an object from a varbacktest object that
holds a smaller time window or selected VaR vectors using the select function.
Visualize returns, VaR vectors, and exceptions using the plot function.
Construct a list of exception dates, losses, and corresponding VaR
levels for a given VaR vector using the exceptions function.
Add portfolio and VaR data to a given varbacktest
object using the append function.
For more information, see VaR Backtesting Workflow.
Example: Forecast mortality trends using Lee-Carter model
The Forecast Mortality Trends Using Lee-Carter Model example shows how to forecast trends in mortality using the Lee-Carter method with an ARIMA model.
Example: VaR and ES backtesting for equity portfolio
The Estimate Expected Shortfall for Asset Portfolios example shows how to compute the expected shortfall (ES) for a portfolio of equity positions.
The Estimate VaR for Equity Portfolio Using Parametric Methods example shows how to estimate the value at risk for a portfolio of equity positions using two parametric methods.
Modelscape: Using the Modelscape API
To learn how to work programmatically with Modelscape™ resources such as lifecycles, model versions, and reviews, see this new workflow example: Using the Modelscape API.