Our calibrated model is only rough indicator of outcomes. A couple of issues that appear to be particularly salient to interpreting its relevance are whether there is substantial basis risk in home price indices and whether we have modeled moral hazard behavior accurately. We will leave these issues for further research, after describing them.

We define basis risk as the risk that fluctuations in the home price index will not match well fluctuations in the price of the home that are beyond the homeowners’ control. The term basis risk is borrowed from futures market practitioners, who refer to basis risk as the risk that the futures price will not converge on the cash price that is being hedged. If the futures price does not correlate well with the cash price, then the futures market does not allow good hedging. By analogy, if home price indices do not track the individual home prices well, apart from changes that the homeowner deliberately makes in the value, then the various home equity conversion forms will not manage the homeowners’ risk well.

The basis risk for our purposes is fundamentally difficult to measure, since we must measure changes in home prices that are beyond the owners’ control. Thus, measuring basis risk entails measuring price changes that are identified with characteristics of the homes. The problem is that any characteristics of the home that are beyond the control of the homeowner that we also can measure for the purposes of measuring basis risk could by the very fact that we measure them be used to condition the price index, so that they become part of the settlement itself.