I said I would try to explain what I mean by sophisticated vulgarity in financial modeling, which I will do by imperfect analogy.
Suppose you are thinking of manufacturing tropical fruit salad and you are going to market it in the USA. And suppose there is no tropical fruit salad currently being sold.
Your fruit salad will contain mangos, papayas, passion fruit, coconuts, litchis, kumquats and loquats. Once you quote a price to Whole Foods, your distributor, you will be committed to it, and you need to give them a price. So, you want to figure out how much to plan on charging e for a can of tropical fruit salad which you haven’t actually produced yet. There are a variety of ways you could do it.
1. You could model out how much it would cost to import the raw ingredients – mangos, papayas, lychees, syrup and loquats, etc. –from cost at their source, taking account of shipping, shrinkage, insurance, canning, etc, and determine a fair price, allowing for profit. That’s a direct way to proceed, subject to uncertainty about what these costs will be in the future. In that sense it’s a little akin to Black-Scholes modeling of options, which makes assumption about the future that may not/ will not turn out to be true, and then figures out the cost of synthesizing an option.
2. You could build a deeper model, and instead of accepting the market price for the raw mangos, passion fruit, papayas, lychees and loquats and sugar yourself as a proxy for future fruit prices, you could try to estimate it ab initio, taking into account of the cost of seeds, fertilizer, farmland, your inexperience in raising strange crops, labor etc, and determine the cost that way. That’s deep, but it needs a lot of knowledge you don’t have and will therefore necessarily misestimate. This is a little akin to stochastic volatility models, where you make assumptions about things you really have not much experimental information on. Ambitious, but …