According to books I really don't understand, "opportunity cost of capital" is a measure of risk, associated with quantities such as "beta." Apparently its role in corporate finance is akin to that of the infamous "Fair" Isaac credit scoring in personal finance. Businesses believed to have a high opportunity cost are saddled with high (i.e. usurious) cost of capital raising. This manifests as high interest rates in the case of debt financing, or low share price for equity financing. It also means even greater impossibility of dealing with the (expletive deleted) financial sector on one's own terms.
It may be possible to incorporate this concept into pub wan's consumer education mission. Let's assume (as if) a large number of consumers could be persuaded to volunteer information regarding (say) credit cards applied for. Input data are (say) income, net worth, "f"ico score; coupled with output data (data received by the respondent) of credit limit, annual fee, interest rate, intro rate, intro period (in units of time).
Obviously, this example is a gross oversimplification given the number of variables involved, let alone the sheer impossibility of mathematically modeling boilerplate language. The latter difficulty was famously illustrated in a PBS Frontline documentary in which a (Harvard) professor of contract law admitted in an on-air interview that she was completely stumped by her credit card agreement. Our purpose here is to illustrate an example of multi objective optimization.
Let's create a norm spec for such data:
It should be noted that norms (and hence normspecs) are subjective. The input variables (the first 3 columns) are preferenced (signum in maxhi-speak) to represent (from the financial sector's POV) an inefficient frontier among applicants. This should not be taken to imply that the model's intent is to serve the interests of the less creditworthy. The reason for creating this model is not finding out the most marginal-friendly (or least marginal-hostile) lenders, but to empirically determine the "going rate" for "risk," as well as the "going rate" for "tradeoffs" between output variables. This is accomplished by asking the question "how much credit can be secured with how little creditworthiness." This question is relevant to all.
The preference components (see maxhi schema) of the norms given for both input and output variables are pretty much universal. The priority components are arbitrary, and can and should be adjusted to pubwan participant taste. My choices (for the example) reflect a desire to place the highest priorities on the variables I guesstimate to be least "elastic." An actual survey might or might not confirm my guesstimates.