Deep Dive - The Cost Approach

[alebrewer]

Depreciated cost says $130 for GLA, regression says $150, I reconciled $145… and so on.

You use depreciated cost to estimate the GLA adjustment factor? Not knocking this at all - just have never thought of doing it. When using depreciated cost, how do you account for diminishing returns? Or do you?

 

[alebrewer]

A cost guide is just a survey. Costs can come from comparable builds in the market. This is the best data to use, in which case the cost data would be derived directly from market participation. Income data can come from surveys as well.

Seems that deriving income data from surveys might be more of a commercial activity? For residential markets, income data is almost always generated from direct rental information... or at least that's my experience.

I think it's great fun to banter about real life stuff that has no black and white solution. Makes us all better.

 

[CGinMN]

You use depreciated cost to estimate the GLA adjustment factor? Not knocking this at all - just have never thought of doing it. When using depreciated cost, how do you account for diminishing returns? Or do you?

It is what you might call a “breakdown” method. Economies of sale are factored in to cost guides. You should talk to Scott Cullen and give Solomon a test drive, maybe it will change your mind, maybe it will just solidify your views. Like any tool, it can be abused if your goal is to cut corners.

Seems that deriving income data from surveys might be more of a commercial activity? For residential markets, income data is almost always generated from direct rental information... or at least that's my experience.

I think it's great fun to banter about real life stuff that has no black and white solution. Makes us all better.

Yeah, just pointing out your beef with the cost approach could be addressed by using actual costs from the market, and it would similarly apply to income.

 

[TerryRohrer]

That’s what I mean.

I can see the argument for that, and I’ve thought about it before and I disagree. If you are using the strongest data for each approach, then it will happen that the approaches are inbred. You generally cannot keep them entirely separate. If you try, you will be weakening an approach. The each approach is going to give you a different value indication to reconcile. Instead, if you intermingle the approaches and use the strongest adjustments, the reconciliation is mostly done at the adjustment level. Depreciated cost says $130 for GLA, regression says $150, I reconciled $145… and so on.

I don't take the view that if my best comp for the sales comparison approach is used there, I can't use it in the income or cost approach. To me, the inbreeding comes from using depreciated cost for sales comparison approach adjustments and the like. But, for farm and ranch appraisals, where the land becomes the focus and the primary value component, that is hard to maintain. I don't know what they teach now, but back in the day, ASFMRA taught how to develop a price per square foot contributory value for each improvement type and quality and condition. It can be presented that way, and often looks cool and sophisticated, but in reality, when you have 7 sales and show meticulous development of the contributory value of ten or twelve improvements, four classes of land, and river front locations, and mountain views, it is every bit as misleading as making 10 adjustments in a residential report based on the 3 sales you relied on.

 

[alebrewer]

You should talk to Scott Cullen and give Solomon a test drive, maybe it will change your mind, maybe it will just solidify your views.

Nah - I'm pretty happy with where I'm at WRT my disdain for the CA. I do really like the CA as an exercise in conceptual mathematics - using the SCA to verify some of the stuff I've developed over the years (curvilinear depreciation schedules for different TEL's, quantifying the weight of updating or remodeling on EA, stuff like that. For me, though, that's pretty much the extent of my relationship with the CA - an exercise in conceptual mathematics.

 

[Ricky LaFleur]

I do not practice using one to complete the other as I think that tends to diminish them all.

While I don't draw any hard and fast lines anywhere, I think this is more a choice than a necessity, primarily due to expediency.

I know what you're advocating for, and I would agree that tweaking each approach just so that they fall in line with each other and look nice is not appropriate methodology. However, it is not the same as deriving information from another approach to make each approach stronger. I was working on some mini-storage facilities recently and surveying the available supply in a given market area. If there was an over-supply, causing a rent ceiling and/or higher vacancy rates, would it or would it not indicate a form of obsolescence when reconstructing depreciated cost in a sale of one of these facilities? This isn't for expedient results (quite obviously that entails a fair amount of time), rather, it may be necessary for credible results.

 

[sputnam]

You use depreciated cost to estimate the GLA adjustment factor? Not knocking this at all - just have never thought of doing it. When using depreciated cost, how do you account for diminishing returns? Or do you?

Depending on the details of how you apply depreciated cost analysis, it can be useful as an indicator of your GLA adjustment. One trap to keep in mind... when you are adjusting for GLA you should only be adjusting for the area of the dwelling... everything else being equal. That means you are only adjusting for larger or smaller rooms. Kitchens and bathrooms are more expensive and should be factored out... usually by counting and adjusting for the number of plumbing fixtures. A rule of thumb that works okay is to use around 65% +/- of the depreciated per square foot cost. I definitely would only use it as one of several measures... unless there was nothing else.

 

[alebrewer]

Depending on the details of how you apply depreciated cost analysis, it can be useful as an indicator of your GLA adjustment. One trap to keep in mind... when you are adjusting for GLA you should only be adjusting for the area of the dwelling... everything else being equal. That means you are only adjusting for larger or smaller rooms. Kitchens and bathrooms are more expensive and should be factored out... usually by counting and adjusting for the number of plumbing fixtures. A rule of thumb that works okay is to use around 65% +/- of the depreciated per square foot cost. I definitely would only use it as one of several measures... unless there was nothing else.

In residential appraisal, the law of diminishing returns seldom comes into play due to the availability of comparable sales. For highly unique properties, though, I have applied different adjustment factors to different comps - for instance, in a market where there is little demand for homes over ~ 3,000', I might adjust a 4,000' home at half what I would a smaller home. I wouldn't recommend this for FFF work though.

Of course, at the end of the day, the smaller adjustment factor for larger homes is nothing more than a functional adjustment...

 

[sputnam]

In residential appraisal, the law of diminishing returns seldom comes into play due to the availability of comparable sales. For highly unique properties, though, I have applied different adjustment factors to different comps - for instance, in a market where there is little demand for homes over ~ 3,000', I might adjust a 4,000' home at half what I would a smaller home. I wouldn't recommend this for FFF work though

Actually... the rule of thumb that I posted works very well. The % can vary.. but the method works. As I stated, I wouldn't use it as the only indicator.

 

[Terrel L. Shields]

I have never seen an appraiser apply the breakdown method of depreciation in a report, but even if one went to that length, how do you estimate functional obsolescence suffered by a feature not represented in the market?

Unique features are infrequently valued by methods that are readily repeatable. And the cost is at least something you can estimate. And all for what? To guess at its contribution. Because the lack of sales data means you are either stuck with depreciated cost or something I consider even less reliable - a single paired sale.

Example. Years ago, I appraised a new construction. Nice house. Biggest in a small subdivision. But they sold it to a man with the sort of needy woman who wanted to best everyone else. So when the seller who owned more land in the same subd. built an even bigger house with a pool, she insisted the add on to the now 3- or 4-year-old house AND build a new pool since the competition had a pool. She got her way, and I did the proposed construction. The pool and pool room was $80k. (this was 20 years ago.) Fast forward 10 years and they sell. My former assistant appraised it and came to me saying, "This thing is cheap. Too cheap." But again, Sweetie had been bested by an even bigger house, so they were selling this one to build an even larger one. But analyzing the sale, the house either sold cheap or the pool was invisible. The catch? The owner was a contractor who mainly did commercial work but had built a few houses as well. So, did he have a full boat load in the house? Maybe not. Maybe he forfeited any profit to sell below market and move on just to appease the little woman. I donno. But my assistant had to defend an appraisal well above the sales price. I tried to use it as a comp later and ended up giving it no weight. The thing was simply below market.

Compounding the issue is that pools - which are not that uncommon - are so variable in size and style, they are not the simple "$10,000" adjustment 9 of 10 appraisers around here seem to think. And they are an over-improvement but not so much as to be 100% FO in a $400k house.

This is why we get the big bucks - to express our judgment. Otherwise, we could just plug in the numbers and out comes the value.

And don't get me started on outbuildings. They are either invisible or $5000, $10,000, or if huge, maybe $15,000 and no mention of any functional obsolescence.

I'm just saying that there is no way to tell if that adjustment is reasonable or not without going back to the SCA for verification... "accurate"

Nothing in appraisal is "accurate". But depreciated cost at least supports the value estimate. Putting down nothing does not. PFA does not.

most do not properly complete the cost approach because to do so takes more time than the client (FNMA/Fredie in many cases) are willing to pay one for.

Sadly, that is the issue. And compounding that is so many appraisers do it so infrequently as to be stumped by simple issues - like not having a cost book and doing a PFA estimate. Saw someone post a report with $90 as the RCN...Where can you get a house built for $90 a SF?