Cost approach site value (sales lot vs extraction)

[gorillakimchi]

If you have sales, then sales are the best way. Absent sales, then extraction is the best way. Avoid "allocation" from the assessor's records. Very weak way.

There are sales but AVG price/lot is way way low somehow. I understand site(lot) values are not based on price/SF. Subject around 7500/SF and other sales size range 5000SF~10000SF. I can't explain why sales approach value showing $1,300,000 but cost approach value $500,000K. If paired sale lot value can support $1,300,000 I didn't even ask the question. Ex) 5000~7000SF around 700~800K.
I understand & believe paired sale is the best way to determine land value. Find similar land sales(zoning, size) near the subject and extend range & period if needed. Right? That's what I did but I think I missed something really important. There are huge gaps between sales approach value (solid value) vs cost (paired sales) value.

 

[gorillakimchi]

You choose the method, or methods, based on the quality and quantity of available data. If you have plenty of similar land sales, then use Sales Comparison Analysis. If you don't, then you might use extraction, or allocation, or others. You might decide to use several and then reconcile.

Didn't think reconcile Cost approach. Can you briefly explain how you find similar land sales on MLS? I just focus on zoning & lot size near the subject.(sold vacant lot) Is there anything else?
I think I missed something, so my results doesn't support sales approach values.

 

[Joe Flacco]

In most cases you can't value residential land on a per SF basis. The zoning and location of the site are big factors for high value land in addition to size of the lot. Also if you are searching land sales most are likely raw land.

In LA you probably have a lot of infill construction where developers buy improved properties, knock it down and build new. If your market is like that then the best way to get site value is to find those properties with improvements which were razed.

Don't search land. Search new construction and search tax records.

 

[gorillakimchi]

In most cases you can't value residential land on a per SF basis. The zoning and location of the site are big factors for high value land in addition to size of the lot. Also if you are searching land sales most are likely raw land.

In LA you probably have a lot of infill construction where developers buy improved properties, knock it down and build new. If your market is like that then the best way to get site value is to find those properties with improvements which were razed.

Don't search land. Search new construction and search tax records

You are right sir. All sold lots are raw land (never built anything). So, I chose C5~C6 improvement(have no much contributory value) comps and extracted site value from it. I think that I did what you advised and that cost approach value somewhat close to sales approach value. So I think it more makes sense than paired land sales site value. However, I can't explain why I chose extraction over paired sales. Many people say paired sale is the first choice and "if there are no vacant lot sales" go for extraction. In my case, have vacant lot sales so I can't just simply say no vacant lot sales. It's lying, so want to explain to the client why I chose extraction over paired during the cost approach. Can you help a bit what's the best way to describe?

 

[Dublin ohio]

You are right sir. All sold lots are raw land (never built anything). So, I chose C5~C6 improvement(have no much contributory value) comps and extracted site value from it. I think that I did what you advised and that cost approach value somewhat close to sales approach value. So I think it more makes sense than paired land sales site value. However, I can't explain why I chose extraction over paired sales. Many people say paired sale is the first choice and "if there are no vacant lot sales" go for extraction. In my case, have vacant lot sales so I can't just simply say no vacant lot sales. It's lying, so want to explain to the client why I chose extraction over paired during the cost approach. Can you help a bit what's the best way to describe?

If you are confident in your sca and depreciated cost of improvements. I would do a deep dive into those vacant site sales. Just to be able to explain why they were not used. From what I have read on this forum in the past. It seems that CA has some "complicated" zoning and building regulations. That could be a factor.

 

[Joe Flacco]

You are right sir. All sold lots are raw land (never built anything). So, I chose C5~C6 improvement(have no much contributory value) comps and extracted site value from it. I think that I did what you advised and that cost approach value somewhat close to sales approach value. So I think it more makes sense than paired land sales site value. However, I can't explain why I chose extraction over paired sales. Many people say paired sale is the first choice and "if there are no vacant lot sales" go for extraction. In my case, have vacant lot sales so I can't just simply say no vacant lot sales. It's lying, so want to explain to the client why I chose extraction over paired during the cost approach. Can you help a bit what's the best way to describe?

There is no need to extract anything if the existing improvements were razed. Those are land sales that are suitable for construction.

 

[gorillakimchi]

If you are confident in your sca and depreciated cost of improvements. I would do a deep dive into those vacant site sales. Just to be able to explain why they were not used. From what I have read on this forum in the past. It seems that CA has some "complicated" zoning and building regulations. That could be a factor.

I found raw lot sales around $220,000 (7000SF $31/SF) near the subject. Subject SCA around 1.3 mil. Deprecated cost around 220K. I mean, just a simple calc showing there are 800K gaps. It clearly says that I did something wrong. CA has complicated zoning but subject is just a simple R1 zoning (Residential) and I searched only R1 zoning lot. Thanks for the reply

 

[gorillakimchi]

There is no need to extract anything if the existing improvements were razed. Those are land sales that are suitable for construction.

Ah, It's not razed but close to it. Improvements still have some contributory value but very low (Effective age around 50~55 out of 60) So I extracted site value SFR sales - Depcost of improvement.
As Mr. Glenn explained above, permit fees are expensive here for a new construction so, rehabbers tend to leave foundations and claim it's remodeled. (It's useful to know. Thanks Mr. Glenn

 

[Dublin ohio]

I found raw lot sales around $220,000 (7000SF $31/SF) near the subject. Subject SCA around 1.3 mil. Deprecated cost around 220K. I mean, just a simple calc showing there are 800K gaps. It clearly says that I did something wrong. CA has complicated zoning but subject is just a simple R1 zoning (Residential) and I searched only R1 zoning lot. Thanks for the reply

Sounds like time to use the Sherlock Holmes method (paraphrasing). If you have eliminated the possible. Whatever is left, no matter how improbable. Has to be the answer.

 

[George Hatch]

In my case, there are land sales. As I explained, land sales comps prices are way way lower than the prices that I got from extraction methods.​

Extraction is supposed to be market based. If you're coming up with really different numbers via extraction vs site sales then that literally means you're doing the extraction incorrectly. The most likely reason is the difference in construction costs prior to getting to the depreciation calculation. As I've been telling you, it's common to grossly underestimate the difference in costs by location. As well, getting to the real accrued depreciation is a common stumbling block, too.

Here's an example:

4903 White Court, Torrance 90503 A 2020yb home of 3433sf on a 6014sf subdivision lot, sold for $2,275,000
35283 Rebecca Rd. Yucaipa, 92399; a 2022yb home of 3178sf on a 11,862sf lot, sold for $875k.

These two homes are reasonably comparable in quality but very different locations. There's a $1.4M difference in the prices.

Land values for SFR lots in Yucaipa start at ~$200k, not zero. That means the 3178sf SFR+improvements sold for $675k above the land value. $675k / 3178sf = $212/sf inclusive of the garage, other site improvements, all direct and indirect costs and the developer's profit margin.

You already know about the lot sale in Torrance on Fiat St, 17176sf lot, possibly splittable, that sold as an SFR lot for $935k. The 3433sf house in that town sold for $1.3M more than the site value. $1.3M / 3433sf = $378/sf, also inclusive of everything.

No matter how you cut it, these similar structures had drastically different contributory values above and beyond their respective site values. Some it will be attributable to higher fees/permits between the two jurisdictions, some of it to higher labor costs being charged for the construction, but much of it will be attributable to the higher PROFIT margins which these costs services never adequately cover in the more expensive markets. You have to take these additional factors into account when you're using these residential costs services because they use a fixed profit margin (something like 13%, last time I looked) whereas in real life the profits are the residual between the costs and the sale prices.

Now the developers know how to hide their true profit margins in their construction cost breakdowns, but if you were to compare the cost breakdowns between these two projects you'd be seeing a BIG difference in many of the line items.