Denis - can you spot check my updated Opinion of Site Value Methodology, to make sure it's all in the right order?
"Opinion of Site Value Extraction Methodology: for each comparable closed sale, determine replacement cost of improvements new (RCN) minus depreciation (from a recognized source, Marshall & Swift), add in EI (Entreprenurial Incentive) and as-is value of site improvements (that are not otherwise accounted for in the RCN), subtract from sales price to get site value, divide site value by site size to get value per square foot for the site. Reconcile across comps for a price per square foot of the site and use that to determine and support the opinion of value for the subject's site."
Alin-
I think that is an excellent summary and would certainly meet (from my review experience, far exceeds) the minimum standards for a typical residential assignment.
I'll add one more thing to consider in the valuation of the site; and this may be market-specific....
Assume you have 3 sales and the sites are 10,000sf, 13,000sf, and 9,500sf. It would not be surprising or unusual if, after your analysis, you find the residual (the value that remains and is attributed to the site) of a smaller site to be more than the value of a larger site. The reason is simple: Site value by extraction, as any other analysis, has a degree of uncertainty. We don't know everything, we cannot estimate depreciation precisely (our estimates can, however, be reasonable) and there is always the specific motivations of the individual buyer and seller which affect the sale price.
So, as a consequence, in the example above, the smaller site may have the highest indicated value. Don't let that freak you out.
Assume that the $/sf analysis works out like this:
9,500sf = $15/sf = $142,500
13,000sf = $10/sf = $130,000
10,000sf = $12/sf = $120,000
The average (all indicated prices divided the total lot sf) is $12/sf. Your site area is 12,500/sf.
What's reasonable for a site value?
$12 x 12,500 = $150,000. Your lot may be worth more than the others. I would submit, however, that the "sites" don't trade on a straight $/sf basis. The primary value of a home site is to support the house that is built on it. Assuming all the homes are similar to the subject, another way to value your site is not to do it on the $/sf basis (which is a unit of value) but to consider the total values of the sites. In other words, site values for sites similar to yours range from $120k to $142.5k per site. It would be very reasonable to make that conclusion and then to pick a specific value for your site (say, $130k).
Like I said, the best method depends on the market. In many of my markets, a typical lot size may be 8,000sf and the site value could very well be $800,000 ($100/sf, if we use the $/sf unit of value). In that same neighborhood, my subject may be on a 11,000sf site. Using the $/sf unit of value, my site would calculate out at $1,100,000. It is not likely that the incremental value difference for site utility between the typical lot size of 8,000sf vs. my specific subject at 11,000 sf is worth $300,000k.
More likely, I'll find site sizes ranging from 7,500sf to 12,000sf with values ranging from $800k to $875k. The incremental value difference in site utility is probably closer $50k.
Again, it all depends on your market. And some appraisers believe that site size difference should be based on the $/sf unit of value and not a more nebulous "utility/lump-sum" unit of value. It is up to you to determine what works best for your market and situation; just make sure that once you do your site calculation and conclude a value indication, it makes "sense" in the market.
And don't be afraid to modify the calculated answer/indicator with your appraiser's judgment. The math can take us only so far; it is your judgment which takes the data and synthesizes it to a reasonable/specific point. But it is the math and analysis which makes the process "credible".
Correct methodology/analysis = credibility
Competent appraiser judgment = reasonableness
Combine the two and you should never have a review-issue!
Good luck!