Time to Manage Your Risk

20 October 2020

Factor Analysis Part II

In Part I of this series, we explored the ability to calculate portfolio factor sensitivities using the RiskAPI Add-In. In Part II, we will look at the application of a multi-factor portfolio stress test using the same principles.

Using the Stress-Testing feature of the RiskAPI Add-In, we can set up a multi-factor stress test, applying shocks of varying magnitudes and directions to each factor of interest. The system will evaluate the simultaneous effects of such a shock scenario both on each individual component as well as the portfolio as a whole.

As in Part I, we continue to leverage the "Market Macro" keywords-based feature of the Add-In to quickly generate RiskAPI calculations on portfolio and factor symbols and quantities. The above image shows the output of a multi-factor stress test performed on all 103 Nasdaq 100 components against the same factors used in Part I. These are presented again here:

  • VLUE - the value factor
  • QUAL - the quality factor
  • MTUM - the momentum factor
  • SIZE - the small cap factor
  • STLG - the growth factor

The "Index" and "Index Stress" keywords allow us to specify which set of factors the stress test is being run with, as well as the magnitudes of the shocks applied to each factor. The "Stress Type" keyword defines, via the entry "Index MR", that a multi-factor stress test is being performed (also available are single-factor, underlying prices, implied volatility, and other types). The Index MR stress testing method leverages the system's multiple regression capability to calculate the portfolio's factor sensitivities over the data set and factors specified. It then applies the individual user-defined factor shock values in the process of evaluating the multi-factor shock scenario.

For the stress test in question, the large one-day changes in value experienced on March 16th, 2020 are used. These ranged from -12% in the SPX to as high as 14.6% in the SIZE factor. The system returns three pieces of information as outputs of the scenario:

  1. Stress Price - the price of the portfolio component under the shock scenario
  2. Stress Impact - the simulated P&L of the portfolio component under the shock scenario
  3. Total Stress Impact - the total portfolio simulated P&L under the shock scenario

The same process can be applied to a less intuitive portfolio, leveraging the options valuation capabilities of the system:

Here we see the same multi-factor scenario applied to a present-day "costless-collar" options strategy on AAPL shares. The system automatically performs the factor sensitivity analysis, shock-propagation, and options valuation required to evaluate such a scenario on the option portfolio in question.