A Division of Windham Capital Management, LLC

Asset Liability Analysis

I am interested in adding liabilities to our analysis. How can I approach this using the Windham Software?

There are two approaches to incorporating liabilities. Both techniques will require that you have a proxy time series instrument for the liability in the Windham case file. We can use either

1.  Negative bond return (choosing a bond return with similar duration as the liability), or

2.  Negative bond index, or

3.  Liability Benchmark (an actuary takes the cash flows and constructs a benchmark using fixed income indices)

Method 1 - Surplus Optimization
Create a portfolio that has a short position to the liability instrument. Consider a case where the assets are valued at 100 million and the liabilities are at 120 million, then the short position should be specified as -1.2 in the portfolio. When you optimize you will need to set two constraints. In this example, the “sum of asset weights” found under the constraint tab should be changed to -0.2 (the sum of assets and liabilities). The next constraint is on the “weights and bounds” tab. You should set the min and max weights for the liability to be -1.2. This way you force the optimal portfolio to hold the liability. You can now optimize. Further portfolio statistics and exposure to loss would measure surplus risk. (e.g., the value at risk estimates will be surplus at risk)

Method 2 - Tracking Error Optimization
Create a benchmark that contains the liability instrument only. In the above example the weight would be +1.2. You will probably want to set the min/max constraints in the optimization for the liability to be 0. Select mean tracking error optimization or mean-variance-tracking error optimization. The optimizer will now try to minimize tracking error to the liabilities while maximizing excess return. The relative risk estimates will be relative to the liabilities.

Category:Understanding the Software -> Optimization