Austin,
First of all, you notice that none of the SoCal appraisers on this forum have disputed the types of increases that we're talking about here. If you or anyone else is skeptical about these increases, all you have to do is use a national database, find a tract of $100,000 - $300,000 homes anywhere in the region, and compare all of the sale prices from last year to this year's sales. There will be a couple of anomolies in each group, to be sure; but the overall trends should be quite apparent. And no, appraisers did not create these trends by aggressive valuation methods. And I know for a fact that ours is not the only region to experience value increases. Other metro areas have similarly increased. Soon, some areas, maybe including ours, will experience significant price declines. At that point, using a dated sale of 6 months will require a downward adjustment.
You're kidding about a 'built-in' time adjustment, right? We shouldn't be making 'built-in' adjustments of any type without being able to demonstrate that such an adjustment is warranted. I wouldn't abritrarily make an adjustment for a 4th bedroom over a 3-bd home unless there was some indication in that market segment that such an adjustment was necessary. Matter of fact, I wouldn't bother with a time adjustment of any sort in a 4% annual increase market, or even an 8% annual market. C'mon, why should we bother with a 1.5% adjustment that is iffy at best? Better to leave it alone and just use it to demonstrate a trend.
It's true enough that sufficient data to support such an adjustment may not exist within the comparable sales presented on a comparison grid. Genuine 'Matched Pairs' are actually not that common. Nevertheless, data can often be found in the same market segment, albeit with different sized properties or in nearby competing neighborhoods. Sure, these data may not be directly comparable to the subject for the purpose of documenting the appraiser's value for the subject, but they can provide background and context for some of the adjustments.
How about this; when one of your comparable sales has a prior sale within the previous year and there have been no significant improvements. Would that be sufficient to document a change in value based only on the effects of time? We do run into those types of situations here, on an occassional basis. Then there is the comparison of whole sets of data; like comparing sales within a tract of homes from last year to this year's sales. No adjustments, just straight up mean and mode comparisons. This would be a perfect use for one of your regression analyses.
Bottom line, an appraiser will always be looking for the more recent, proximate and similar comparables to use if for no other reason than it makes their job easier. Adjustments are made in a comparison grid only because the appraiser failed to come up with sufficient 'identical' comparables. Sure, time adjustments as a tool can be abused, particularly by the semi-informed. But so can any other tool, like Marshall and Swift, or a DCF analysis tool like Argus, or a Gross Rent Multplier adjustment spreadsheet, or (and especially) a regression analysis. The possibility of abuse should not justify their prohibition.
George Hatch