A MultiFamily Question on Adjustments

[Hamlet]

Practice dancing backwards perhaps ..... .

Yeah, or if you aren’t athletic and more of the intellectual type you could discover fire and sell it on the internet.

 

[SANDY]

How about we get back to the actual circumstances at hand- Property says that he had adequate comarable sales. He chose to use his grm based data to make one lump sum adjustment in the sca. Underwriter wanted lne item adjustments.
So if he had adequate comparables then why not use them to make the line item adjustments based upon paired sales or some other acceptable sca method? Maybe his sales did not allow for specific line item paired sales info?

 

[SANDY]

Say your subject is a 4 unit property with two 2-bedroom units and two 1-bedroom units. But the only 4-unit comps you have are two with all 2 bedroom units and one with 1 2-bedroom unit and and three 1-bedroom units.

There's probably not enough information there to make supportable adjustments for the unit mix. What to do?

You could peform a rental survey to find out that 2 bedroom units rent for 700 and one bedroom units rent for 600. The incremental rent attributable to a 2nd bedroom is $100.

Rent per month $100
Gross income per year $1,200
5% vacancy -$60
Subtotal $1,140.
Less operating expenses and collection loss at 35% - $399 ($400)
Annual income for second bedroom $740
Capitalize at 8% = $9,250.

or say the GRM is 100. Then the indicator is $10,000 for the bedroom.

What's wrong or misleading with that?

(These are just made up numbers to illustrate the point.)

Greg- Here's what wrong with using the grm to indicate the contributing value for the bedroom. Your example uses sales of properties that are not similar tot he subject in terms of number of bedrooms and those sales likely have other dis-similarities tot he subject. See previous posts on requirements for the development of a grm.

Using a grm also requires some judgement just like paired sales analysis.

There are many times that grms can range from 90-110 and even more of a spread. So if one chooses a grm of 100 and then we apply the bedroom example rent difference of a $100 we get an indicated adjustemnt of $10,000. If we choose a grm of 90 x $100 we get $9,000, grm of 110 we get $11,000 or a 10% margin of error on either side of the 100 grm. Not exactly too much better than a skilled experienced appraiser like yourselves could do with paired sales analysis with reasonable data available.

And that's assuming you haven't messed up or mis-used the data to develop those grm ranges in the first place. Some appraisers are so enamored with numbers and statistics that they actually think that using them or mis-using them is somehow a superior technique-ie.. they leave their common sense locked up and punch in numbers-computer then spit out answer-put in report and claim look how smart me am. For those of you that are challenged-the me am part is a joke. Some of you are pretty thick so I have to explain everything.

 

[SANDY]

You boys really trust that grm number that much to apply it in your sca? Does it account for all those possible differences in age, location etc... of those sales prices and rents you used to develop it?
Hint- That's why you a different method in the sca!!!!!!!!!!!!!!!!!!!!!!!!!!!!
You wanna be lazy though so go ahead of use it again in your sca-not me!

Now I am really ****ed at you boys!!!-Just got a 4-plex to do and my supervisors now want me to include a fourth approach to value-They are calling the Boyd Property Approach-Thanks a lot guys!!

 

[CANative]

I don't think it's a matter of "us" trusting a grm. It's more a matter of developing and reporting credible assignment results.

If you do a rent survey and analyze a dozen 3 bedroom rents and a dozen two bedroom rents you're going to get a pretty good notion of the difference in rental rates between the unit mix. These do not have to be sales.

Once you've reconciled the grm for you're subject from the range of grm's of your sale comps you can apply the amount of rent gain/loss to that reconciled grm. Personally, I don't like to do this because I usually find it is inconsistent or, even worse, overstates the adjustment. My territory has so few 2-4 sales and such a wide variance in property characteristics that everything is a judgment call. But I will use this method if I am forced to. It's better than the alternative which is to do nothing or just offer a guess.

Sandy... You're arguing yourself into a box and each argument is worse than the last. It's not that I am smarter than you or even significantly more experienced. I'm just now struggling to get a grasp on more advanced appraisal concepts and methodology. Frankly, from what I've learned in the last few months I'm kind of embarassed about all the doofus posts I've made over the years.

Appraising houses for lenders is child's play compared to what appraisal can really be. For example, see if you can solve this problem:

Extraction of an investor's expectations: During the confirmation of a sale, the following information is collected:
$2,000,000 sale price
$120,000 Io (level per lease)
Investor expects to earn 10% yield over five years.

If the income is expected to remain level during the holding period, what change in value does this information imply? How would these conclusions be interpreted in terms of a yield rate for the subject?

For extra credit, solve this problem:
How would the answer from the above problem change if the investor quoted a yield rate of 9%?

 

[PropertyEconomics]

Question .. why do investors by rental units?

 

[SANDY]

I don't think it's a matter of "us" trusting a grm. It's more a matter of developing and reporting credible assignment results.

If you do a rent survey and analyze a dozen 3 bedroom rents and a dozen two bedroom rents you're going to get a pretty good notion of the difference in rental rates between the unit mix. These do not have to be sales.

Once you've reconciled the grm for you're subject from the range of grm's of your sale comps you can apply the amount of rent gain/loss to that reconciled grm. Personally, I don't like to do this because I usually find it is inconsistent or, even worse, overstates the adjustment. My territory has so few 2-4 sales and such a wide variance in property characteristics that everything is a judgment call. But I will use this method if I am forced to. It's better than the alternative which is to do nothing or just offer a guess.

Exactly-you still need those sales to generate a grm and if those sales are not really good comparbles then your judgement is required just like it is when using other more traditional sca methods ie.. sensitivity analysis, paired sales etc... I repeat-just because one method's data is weaker than another you don't get to use the same methods in one approach in another approach. Can you say reconcile?

How about this scenario Greg- Take a commercial property assignment that has ohhh maybe 20 comparable sales. Let's say that 8 of those sales were sales by owner operators. Wouldn't those 8 sales make for some good csa data?

For now, try not to interject your particular market situation where you say that you normally do not have good sales data. That is a whole other discussion for later.

Greg-I really think the disagreement we have comes from your particular perspective due to the market you work in, ie.... few good comparables most of the time-in which case I can see the need for alternative but perfectly acceptable methods. But htis whole thread started with PE saying that he had comparable sales but that despite that fact he normally still uses his grm to provide one big lump sum adjustment in the sales comparison grid. The underwriter didn't like that and neither do I. I would assume that PE did not have the level of sales data to calculate certain individual relatively insignificant line item differences but there should have been enough sales data to estimate the major line item adjustments. The rest of the individual adjustments(if applicable) could be adjusted using the sensitivity method or even limited pair sales analysis since we now agree that the income based method also requires some judgement and is rarely ever perfect either. Do you feel me now bro?

 

[PropertyEconomics]

Your total adjustments to a sale Equal $20,000 ...... Use of the GRM method as described in my original post results in an adjustment of $19,875 ..... YUP I can see the point .... definately making individual adjustments makes more sense .... its more accurate .... I concede the arguement .......

Your reporting results are much more credible ......

 

[SANDY]

Your total adjustments to a sale Equal $20,000 ...... Use of the GRM method as described in my original post results in an adjustment of $19,875 ..... YUP I can see the point .... definately making individual adjustments makes more sense .... its more accurate .... I concede the arguement .......

Your reporting results are much more credible ......

Exactly my point-total adjustments by the various methods should be similar and if they aren't it is an indicator that something is wrong with your adjustments in one or more of the approaches. If you use your one method of making adjustments the results would have to be the same. You are unbelievable. I think you should stop posting the obvious on this topic. You messsed up and got called on it-you'll do better next time now that you have learned!!! I understand that it is embarrasing for an underwriter to be rigth and for you to be wrong but it happens. Those former McDonald's cooks are sharp cookies!

 

[hastalavista]

I thought the long-running argument on this thread was in regards to the appropriateness of using a method identified with one approach when analyzing value in another approach?

I brought up several examples where I thought it would be useful:
A. When the market reacts to cost, make the adjustment based on cost in the SCA.
B. When the market reacts to income differences, make the adjustment based on the income difference in the SCA.

I can understand "extracting" an adjustment in the SCA; I can extract an adjustment and not know why the market reacts that way. That happens. The example I used was "appeal"; what may be in favor today can be out of favor tomorrow. Other than matching "like for like" in design, I can't think of a sound way to drill any deeper as to why the market reacts as it does between two styles of design that have similar functional utility.

But in those cases where I can drill deeper, I would want to do so.

If I can correlate income or cost with market reaction to a component of value, then I'm going to use that correlation to consider my market adjustment. And I'm going to cite that correlation as support (basis) for making that adjustment. Wouldn't that give the appraiser (and the user) a higher confidence level about the adjustment being made?

I said before that I think this analysis is extremely useful in justifying when an adjustment is not warranted.
Again, back to a fourplex: Our subject's location appears superior to Comparable A. However, we can determine that superior locations command higher rents. We compare our subject's market rent to that of Comparable A and find no rental difference. Is a location adjustment warranted?
Without analyzing the income, one might make a qualitative argument that a location adjustment is warranted; but in fact the quality of the location that is perceived by the appraiser is not recognized in the market as a value premium.
With analyzing the income, one can make a quantitative argument that that there is no location adjustment warranted. And the decision is supported by market evidence that no premium is awarded.

It doesn't take a giant leap to see that if such an analysis is useful in disproving adjustments, it can also be useful in proving adjustments.

My valuation toolbox has a number of different tools. I don't have it partitioned as SCA, CA and IA where using one precludes the use of the others. They are all grouped as Value Analysis tools and I use the one(s) that best analyze the specific valuation problem.

Nothing misleading, being lazy or fraudulent about it. Not a short cut.