Asset Class information is entered quite differently depending on whether you are optimizing asset returns or excess returns. However, if you are optimizing asset returns, the same information is required whether ProVal has calculated an asset-only efficient frontier or you have generated the efficient frontier by an alternative approach and then entered it.

Optimizing Asset Returns

When optimizing asset returns, the Asset class name, expected return, standard deviation, and correlations may be entered on the screen directly for the Asset classes or by using the Populate feature.

If the Class Name, Expected Return, and Standard Deviation are entered directly on the Asset Class screen, you then also need to enter the Correlations directly as well (under a separate topic). The Populate button allows you, instead of entering this data directly, to reference any of the following Capital Market Simulation data:

1. Real returns (input), which refers to inputs of the classic mean-variance simulator.

2. Real returns (simulated), or

3. Nominal returns (simulated).

The simulated real or nominal returns can come from any type of Capital market simulation (classic mean/variance, multi-factor or imported).

The Populate feature populates both the asset class information and the correlations.

The Calculate expected return based on: parameter allows you to define expectations for the statistics calculated from the capital market simulator’s data. If you choose Geometric average return, you are indicating that your best estimate expectation is based on long term compound returns. If you choose Arithmetic average returns, you are indicating that your best estimate expectation is based on annual returns.

Optimizing Excess Returns

When optimizing excess returns, you do not have an option to enter information directly. The Asset Class Information is determined automatically based on the Liability Target and Stochastic Forecast that you select. See the Liability Return and Excess Return Technical Reference article for details.

Select the Liability Target from among the listed accounting and funding liabilities:

• Entry Age Normal Accounting Liability in the U.S. public mode

• Accumulated Benefit Obligation (ABO) in the pension modes

• Projected Benefit Obligation (PBO) in the pension modes or, in OPEB mode, Accumulated Postretirement Benefit Obligation (APBO)

• Expected Benefit Obligation (EBO) in the pension modes or, in OPEB mode, Expected Postretirement Benefit Obligation (EPBO)

• Actuarial Liability in all modes (e.g., unit credit, entry age normal)

• Present Value of Future Benefits in all modes

• Funding Not-at-Risk Liability in the U.S. qualified mode (PPA minimum funding liability)

• Funding At-Risk Liability in the U.S. qualified mode (PPA minimum funding liability)

• RPA ’94 Current Liability in the U.S. qualified mode

• Gateway Current Liability in the U.S. qualified mode

• Solvency Liability in the Canadian registered mode

The available liability targets, as well as the choice list for the Stochastic Forecast parameter, vary according to the ProVal calculation mode of the Project selected at the bottom of the dialog box. The pre-set option for the selected Project is your current Project but you may choose another Project, including one in an entirely different mode.

Note that, even though you may be able to select a particular liability type as the Liability Target, the Efficient Frontier will not run unless the selected Stochastic Forecast includes results for that liability type. Thus, for example, to run the excess return Efficient Frontier based on ABO, PBO or EBO, the selected forecast has to include accounting results. Similarly, to run based on the funding actuarial liability or the present value of future benefits, the selected forecast has to include funding results, and to run the Efficient Frontier based on a current liability or PPA (minimum) funding liability requires a selected forecast that contains results for these liabilities (which, in turn, is determined by the nature of the funding Valuation Assumptions underlying the Stochastic Forecast). In the Canadian registered mode, to run the Efficient Frontier based on solvency liability, the selected forecast must contain funding results and the underlying funding Valuation Assumptions must calculate solvency liability.

The Calculate expected return based on parameter allows you to define expectations for the statistics calculated from the capital market simulator’s data. If you choose Geometric average return, you are indicating that your best estimate expectation is based on long term compound returns. If you choose Arithmetic average return, you are indicating that your best estimate expectation is based on annual returns.

You may also chose to Write liability returns to a file by checking this box. This is useful if you want to use the liability returns directly as an asset class in a Custom type of Capital Market Simulation, presumably to model liability driven investments (LDI). Browse to the csv or text file in which you wish to save the results.

When you click the OK button after selecting a liability target and a stochastic forecast, ProVal re-runs the stochastic forecast to calculate the necessary asset class information inputs for the efficient frontier. These inputs are not displayed at this time but passed automatically and saved. These inputs are available for inspection with the rest of the inputs and results by clicking the View button on the main library dialog box.