San Diego Housing Bubble News and Analysis

The question I have regarding your method and comparing it with foreclosure.com # is that, are all the number in foreclosure.com in the MLS? Is there a possibility that because of the slow foreclosure process, many homes are just sitting either empty or in limbo or previous owner living rent free. Also, we really don't know the integrity of foreclosure.com's number, so we will have to take it with a grain of salt as well.

I just looked into a small sample of SFRs that became lender-owned on October 1st. Less than 20% of those (5 out of 32) have been sold since. 30% are currently listed for sale. The remaining houses apparently either didn't make it to the market yet or are pending.

So, it's a long pipeline. You can confirm that by going backwards from your MLS, looking at random REOs sold and determining when they became REOs.

This is a small sample confined to a particular area but may shore up foreclosure.com's data. I am a cheap and arrogant bastard. This combination wont allow me to pay the subscription fee for foreclosure.com (which would cut into my porn website subscription budget) and the arrogance makes me think I can outsmart them. They give you a zestimate on the foreclosure page and a name on the preforeclosure page (they keep them there for a bit as inactive once they are foreclosures), I track both, use the gis and property tax website for the county to figure out the address using the name (tedious because you have to run all the apn's on the street, for big streets I jog by and look for brown lawns). Then I run a grantee search online to verify the nod and not, then run the name in the newspaper legal classifieds to see the actual notice for the N.O.T. and get the debt amount. I have run about a hundred and all of them have fleshed out to be accurate. Furthermore I have found at least a dozen that never went into the MLS but some got for sale signs from out of county real estate offices and I have verified that they have sold to private paties that way, never going into the mls and selling for very low prices, almost making me angry that they didn't go into the MLS. Conversely I have never found them on the MLS without first seeing them on foreclosure.com.

This is confined to a single zip code but may explain the data anamoly, my conclusion is the mls data is faulty, not foreclosure's.

I can only offer my small sample tracked for about a year, I would do the sample in another area to verify my results but that would qualify as an obsessive compulsive disorder, for now I can justify my insanity and I hear that the meds have terrible side effects.

Guys these are good explanations and may indeed help me shore up the data which was really the intent of the post. Late tonite, as that is when I am pretty much confined to longer postings, is when I will look into it. Hopefully more people can throw some ideas into the mix.

SD Realtor

Well, this begs the question. Why would you not list it in the MLS? If you are selling a place in a market that's - being kind - sluggish, why would you limit yourself to anyone going by and seeing a sign? What if you live out of state, for example, moving here from No. Calif. You're not jogging by and seeing a sign. I would think you'd want to get as much exposure as possible and sell the place quickly. This makes no sense to me at all. What the hey!

I'm really just wondering how any banks stay in business. First they give money away to anyone w/a popsicle stand. Then they let people live there for free for a year, then they don't reduce the price of a home that's foreclosed, don't quickly accept offers that are reasonable and now you say they don't list them in the MLS.

I try to learn and I get more confused. I feel like Spock right now.

Hi Adam -- I always appreciate if someone wants to dig into the data better, but I'm not quite sure I understand the issue here because the disparity you cite seems pretty explainable.

One answer is that a disproportionate number of foreclosures are in the low end areas, and a disproportionate number of sales in the high end areas -- so it makes sense that the number of total REOs sitting there per total sales would be lower than the number of SOLD REOs vs sales (ie, more REO inventory is stacking up).

Another answer if I'm understanding is that you are tracking closed REO sales, while I am tracking recorded NODs and NOTs. The lag time between an NOT and REO closing is probably 2 months bare minimum (3 wks to get from NOT to REO, another month or more to close the sale), and probably a lot longer with that. With NOD it's far longer than that by many months.

RE. the data sources: I'm not sure what foreclosure.com is. I get the data from the San Diego Daily Transcript, who gets it from the country recorder's office. I used foreclosureforum.com for the historical data (since they had it in tabular format) but extensively spot checked and found a 100% accuracy rate between foreclosureforum.com and the Transcript. So since both sites are independently getting this from the country, I'm pretty sure they are reporting what they get from the county correctly. I don't have any explanation for why the county numbers would be different from your title database.

Rich

One other thing to keep in mind is that people are very sloppy in entering info into the MLS. I personally looked into buying two REO properties. However, you'd never be able to tell they were REO properties based on the MLS information. (Note: there also tends to be a lot of errors in numbers enterred in the MLS - sold price, sq. foot, etc., whether its sloppiness or deceitfulness on the part of the person entering the data (i.e., the realtors) I do not know).

My explanation would mimic Rich's. I will add a short run through some numbers though. January REO sales correlate with November or earlier NOTs. November NOTs correlate with July NODs (a quick regression shows that NODs lag NOTs by 4 months more accurately than 3 months).
There were 2033 NODs in July 07. Of these, about 50% become NOTs, meaning there should be about 1,000 foreclosures potentially being sold in January. This is what is TRYING to be sold. My question is what's a standard ratio of what sells vs. what is for sale? Maybe 1/4, if those listings are priced aggressively? 1000*.25=251
This last ratio is key. Even if there was 1 foreclosure per sale, that only means maybe 1 REO sale per 4 other sales.
I believe the data is correct, but the assumptions/projected implications are shortsighted.

I am not a conspiracy theorist, but doesnt it seem that a listing agent would want to leave REO outa the MLS listing? They are there to sell that property for the maximum possible. Their comissions and the banks $ recovery all depend on them selling fast, and at full price. Anything labeled "REO" is gonna get hit with a -10% off listing offer in this market, if not more. My poor and unMLS accessing memory tells me that this happened even with some of the few nice houses that have come up REO. So shouldnt the bias be to "accidently" not release that info as long as possible?

And dont tell me that agents are following all the rules about disclosure. How many times have you seen a house at 100+ days, or even 200+ days on market get removed from MLS, only to be relisted at 0 days on market the next day and have a note saying "priced to move.... get it before its gone". Note the list price didnt move a wink and it's still the same agent. (This is usually used in conjunction with the standard "we are expecting 3 offers tomorrow, so bid now" line. It is kinda hard to use that line if the house has been sitting for 8 months).

So, I guess if I was gonna question data, I would look to where the observed bias and inconstiancies in reporting are first. Especially if multiple independent spot checks of the .com data sets seems to hold up.

January REO sales correlate with November or earlier NOTs.

Much earlier. There seems to be a 2-3 month delay just bringing NOTs on the market. Per Rich's data there is a 12 month supply of houses on the market. We'll be optimistic and say that banks are being aggressive in pricing and REOs sit on the market an average of 4 months before selling. Finally let's say it takes a month for the sale to close.

Adding all together, January REO sales should correlate with June NOTs.

I've seen a good number of listings that were REOs but were not advertised that way. Some of those included a clue about a motivated seller, but others didn't.

That's one advantage to checking the current ownership on every listing. It's time consuming, but it does have results.

I've seen a good number of listings that were REOs but were not advertised that way. Some of those included a clue about a motivated seller, but others didn't.

Sometimes the property is described as "lender owned" (not "bank owned").

Sometimes the only clue that the property is a REO is a statement like in #071093843

"The seller has never occupied the property and is exempt from supplying the buyer with a tds. This home is being sold in its current as is condition. "

with variations (#086011794)

"Property sold as is ... Seller is disclosure exempt."

That's one advantage to checking the current ownership on every listing. It's time consuming, but it does have results.

How does one go about looking up the current ownership on a listing?

Hi Rich -

I think ultimately what I am trying to get to is the true number of bank owned properties that are being sold. Using the MLS as my database I realize that is a very poor vehicle to gain insight but it is all I have for the limited time I have to look into things. Just to note I did my numbers based on all San Diego county so that should smooth out regional disparities in the county.

I know that yes agents are sloppy or lazy but REO or bank owned properties are usually indicated. I would argue only a small sample are not.

I do better understand your analysis is from a NOD/NOT perspective. Also what bolsters your thoughts/numbers further are distressed sales from those who are not even NODDED or NOTTED yet. Also per esmiths input I would agree that REO sales should lookback much farther then the 3-4 months time. His numbers are dead on there.

I guess what I am really trying to find is the correlation of the number of foreclosures reported by foreclosureforum.com to the number of closed reo sales I see on the MLS. So if I look at the NOD numbers in 2007 and then the % that went to foreclosure, I should be seeing about 500 closings per month that would be REOs. This is a generously low number given the lagtime of the pipeline. That is I looked at the early spring numbers of NODs and the % number and then would expect to see roughly that many closed REO sales down the line.

I need to look at the data more tonite but can you see what I am trying to rough out? Yes there may be sloppiness in an agent not putting in the string I am looking for and such and I need to reread the other entries more closely.

It just is gnawing at me and I want to figure out the difference.

I hate it when I cannot explain the data....