Weighted sales reconciliation methods

[Michigan CG]

Not only in the reconciliation but with ALL adjustments. It's an opinion.

You need to SUPPORT your opinion otherwise your opinion is worthless.

 

[mark hurlock]

Us "average appraisers" grasp the concept just fine. We just don't agree with it. The % given to each comp is not an exact science, and the use of percentages implies a level of precision that just doesn't exist. The AVM crowd loves this concept, as the AVM is trying to portray appraising as only a math problem.

Obviously the average appraiser does not grasp the concept which is why they continue to argue over exact sciences and mathematical calculations rather than the reasoned opinion that leads them to their value conclusions.

The reconciliation of three numerical values into a single “precise” value conclusion needs to be supported.

 

[residentialguy]

The reconciliation of three numerical values into a single “precise” value conclusion needs to be supported.

Not by math

 

[Joyce Potts]

Obviously the average appraiser does not grasp the concept which is why they continue to argue over exact sciences and mathematical calculations rather than the reasoned opinion that leads them to their value conclusions.

The reconciliation of three numerical values into a single “precise” value conclusion needs to be supported.

The WinTotal program will weight those comparables having the least amount of gross and net adjustments based on a formula which the appraiser can certainly ignore or tweak if they choose too. The problem is too many appraisers took liberties with the "it's my opinion" aspect for too long.

I contend that both a statistical (macro) and Sales Comparison (micro) approach should be done all all appraisals. If the average/median price or any kind of value derived by a regression analysis has a reasonable disparity (assuming the data wasn't manipulated in a certain direction for either approach) then I call that a compelling valuation. If, however, the statistical analyses via average/median or regression has a large difference, then your property is either not an ideal candidate for a statistical approach because it's either atypical OR the market data doesn't supply a large enough sampling. So run the data multiple ways and RECONCILE why or why they're aren't within a reasonable range.

Or just cherry pick the closest selling comps based on price, lowest comps or highest comps and hope an underwriter or reviewer lets it slide cause it's just your opinion, right?

 

[J Grant]

The whole appraisal report is what supports your value opinion. Sticking in a precise percent for each comp at the very end just precisely shows what comp got what weight, but imo it is window dressing....how do you decide what comp gets 35% and what gets 38% and so on? Win total does it automatically by weighting the adjustments, and it is a good reference, the final value opinion is supposed to be from the appraiser, per their jugement taking into account market conditions, trends , quality of subject and comps, etc.

 

[TJSum]

The WinTotal program will weight those comparables having the least amount of gross and net adjustments based on a formula which the appraiser can certainly ignore or tweak if they choose too.

Obviously the average appraiser does not grasp the concept which is why they continue to argue over exact sciences and mathematical calculations rather than the reasoned opinion that leads them to their value conclusions.

The reconciliation of three numerical values into a single “precise” value conclusion needs to be supported.

Your faulty assumption is that the degree of adjustment is the only factor when arriving at a value. There are other important factors when supporting your value opinion. You are trying to turn an art into some sort of AVM math problem.

 

[J Grant]

Statistics can support an adjustment or a trend but relying on them for value is the provence of Zillow and AVM's . They can prouce a value, but it can be different than the traditionally developed MVO where the judgement of the appraiser interacts with the market data.

They are two different approaches that can yield very different results. IF a client wants an additional stat value , go for it, but we are signing off that we opined our value, derived from a number of sources, including statistics, mathmatical adjustments, quality comparisons, analyzing market trends, talking to market particpants, inspecting and observing the subject and comps. IT is our judgement on each report which of the sources are most relevant and credible toward the value development.

Appraisers have to be imersed in the market to understand the behaviour and motivations and spending patterns of the market participants.

Buyers do not set out to buy the statistically median price house in a neighborhood, nor do they get excited and decide to pay more because a house represents a certain data point.

 

[Metamorphic]

Comp 1: $555,470 x .45 = 249,962
Comp 2: $501,680 x.25 = 125,420
Comp 3: $505,120 x .15 = 75,768
Comp 4: $452,320 x.15 = 67,848

You all are missing the forest for the trees.

The problem is is that he has not finished the sales comparison approach. There's a $103k range of adjusted values.....20%.... seriously!?! I mean come on. You cannot tell me that there is not some valuable amenity that has not properly considered, or that perhaps a non-market transfer is thrown in there; something is blowing up that range. Typical buyers just dont shotgun their purchases plus or minus 10%. You can't say 20% is just the normal background variation due to buyer caprice, and unadjustable intangibles?

The whole point of the SCA is that if you compare properties that compete in the same market and adjust, based on the market, for the differences between those properties, the adjusted values will converge at some number. In theory, if you had perfect knowledge of market reactions and had fully cross competitive and fully market sales you should be able to adjust to a single number. In reality, you really should be able to get to within 5% on all but the most complex properties. If the closest you can get is 20% you'd be better off just treating the whole SCA qualitatively as the quantitate approach is not particularly helpful.

In the case of the OPs numbers, I would have called the Listing Agent of the one that adjusted to $555k and said "I looks like you got a really great price for your clients on that sale. When I compare it with similar sales it looks like you got about $50k more than I would have expected. Was there something extra special about this property that caused the buyers to pay that much? Tell me what I'm missing here?" Then you shut up and let the agent talk. Seriously...9 times out of 10 when you do this the agent will reveal something about the transaction that was not in MLS and not evident from the street that will nicely explain the difference. When you're calling on the property that looks like it sold too cheep you want to call the Selling Agent...not the listing agent. "Hey, it looks like you got a really great price for your buyers......yadda yadda yadda." Listing agents like to bag about why they got top dollar. Selling agents like to brag about how they helped their clients negotiate the price down. Usually with the low adjusted price sales there was some defect identified late in the transaction, or some amount of undisclosed seller distress that drove the price down. With the high priced sales, you usually find out that was some undisclosed feature (view, upgrades) or concessions or personal property that earned the higher price. When you adjust based on this new information your adjusted values will tighten up.

As far as the OP's initial question. When you're sitting there with 4 comps that adjusted to within a few percent of one another, the reviewer doesn't care much at all how you reconciled because the dollar values become insignificant. With the 4 adjusted comps you presented, an error on your part could wipe out the buyers 10% down payment TIMES TWO, if your reconciliation logic was flawed. The reviewer was spot on to grill you about your reconciliation.

 

[J Grant]

That is a very good point, M, his range of adjusted values is huge for a modest price property. Paric among lower priced properties, unless there is a compelling reason or property oddity, the adjusted value range ( and hopfeully unadjusted prices should be much tighter). I noticed that when he first posted but did not address it, glad that you did as a valuable insight.

 

[AnonApprsr]

You need to SUPPORT your opinion otherwise your opinion is worthless.

It's actually worth up to a million dollars insured.


Meta, I'd blame the fact that the lender wants 3 or 4 sales when maybe only two sales are truly "Comparable." Sometimes you only have one!