To me the cost approach is about building a substitute property.
Builders charge a price to build and that price we break down and call "cost". But it isn't the cost, as in "the actual expense to produce", it is the cost as in "the amount you would have to spend to have it produced for you". That is price and like any price profit is built in. On that note I will have to disagree with the Geiger principal because entreprenurial profits should already be built into the "cost" to produce to the typical consumer.
When times get tough and builders want to stay in business, they have to do the same thing the rest of us do and that is lower their profit margins in order to draw in buyers.
This reduction in price found in your immediate market, relative to what they could charge in another area of the pricing market (as defined by the cost guide) is measured as economic depreciation. As a more concrete example, putting it into appraiser terms, I cover two counties and both are just as easy for me because I live close to the border. My fee for both counties was always "X", but in the one county where the bigger city is located, if you want to get the work given the surpluss of appraisers, you will have to reduce your fee to meet the competition, which is basically "X minus Y". In the other county I can still charge "X", so "X - Y" is not the new cost to the consumer, but "Y" is the adjustment to cost to reflect the economic conditions.
The difference is that in building a house there is the added component of land.
I find land values drop at a faster rate than economic depreciation. When in an area that is similar to your location what I do is I isolate all the land sales available and I look to find what I have the best data on. I then compare those land sales from the past year to the same type of land sales from the year before. In my last assignment, the average price was $85,000 from the prior year and it was $55,000 in the current year. I then look at the houses that sold on that type of land in both years, and in this same example I saw a drop in value from $275,000 to $215,000, the result was a $60,000 drop with about $30,000, or half, attributable to the land. My house was a $220,000 house, which based on local rates of decline it would have been worth about $268,000 in the prior year. Because my lot sales were different (that is the subject was on a 60 X 100 lot and the lots I had been looking at were 1/4 to 1/2 acre), I considered the loss using allocation. I viewed my $48,000 drop in value due to $24,000 (50%) from the land and a $24,000 (50%) from economic depreciation due to the sluggish market and how it affects builder pricing.
If I do not have sales available in order to do the math, or if I do not have newer sales from which I can work with, then building the substitute property is not even a theoretical option, and the cost approach is not applicable.
That's how I do. I am sure it isn't perfect, but it seems to work out.
