Sales and Must-Sell Inventory

Submitted by Rich Toscano on April 4, 2007 - 12:12pm

I speculated in an earlier entry that measuring monthly home sales against monthly notices of default would provide a fairly good read on the health of the housing market.

This idea is borne of the observation that excessive inventory does not by itself put much downward pressure on home prices. As long as sellers can hold out for higher prices, they can be depended upon to sit tight and do just that. Prices don't fall substantially until too many motivated sellers -- those who are forced for one reason or another to sell at whatever prices they can get -- enter the market.

But how many motivated sellers is "too many?" That's where the sales volume comes in. If must-sell properties account for just a small number of the homes being sold, the must-sellers are unlikely to have much effect on aggregate prices. But if enough transactions involve highly motivated sellers, prices could suffer.


(category: )

Submitted by FormerSanDiegan on April 4, 2007 - 2:52pm.

Nice chart Rich. Yet another indication that it's time to party like it's 1991.

Submitted by AN on April 4, 2007 - 2:55pm.

If you look at the blue line and compare this peak w/ last peak, the downward trajectory is much steeper this time around. We'll see how deep this rabbit hole goes soon enough.

Submitted by powayseller on April 4, 2007 - 3:31pm.

The chart shows a correlation between the sales/default ratio and dropping prices, but I'm not sure we can say that rising defaults are creating more must-sell inventory.

If you want to know what lowers prices, you'd have to look at seller psychology. That would be the only true measure of must-sell inventory. This is the reason that NODs are not must-sell inventory:

People with NODs often do not want to sell. So what % of NODs is a listing? 5%? 20%? 50%? I don't know. I pay for a subscription to, and I rarely see an NOD listed for sale on the MLS. Rarely. Often I see the same house go in and out of NOD status, as the owner tries to bring the loan current. I've followed homes all the way to auction status, and they are still not listed for sale!

Further, an REO is not an MLS listing right away either. Banks need time to convert REOs to MLS listing. They have to tidy up, get an appraisal, evict the owner if necessary, so there is a lag time from REO to sale. Often the bank sits on the property. Check out the house on Twin Peaks Rd, which was an REO from the 8/06 auction until the 12/06 pending, and closed finally in 2/07.

Typically, there is a 1 year lag from NOD to REO sale, according to the Credit Suisse report Mortgage du Jour.

5% of homes on the MLS now are short sales, but that doesn't seem to put downward pressure on prices, mainly because you've got a bunch of investors and sellers saying, "I NEED..." I NEED to get $520K or $1.1 mil, etc.

Due to seller resistance, foreclosure status is not YET correlated to price reductions.

What will reduce prices? I think the psychology has got to shift. There is still too much optimism. Sellers are still too demanding, with their "I NEED....". A second factor will be rising inventory, and motivated sellers.

The key is that a NOD is not the same as a motivated seller. Just because someone has an NOD doesn't mean they want to sell. Obviously, many are not selling.

I remember Rich once saying that real estate prices are sticky on the way down, and I can really see that happening now. Sellers are so damn stubborn. At a 5-10% price drop per year, this correction could take a decade.

I'm very interested in what will happen when the credit crunch continues, and people actually need a down payment. Today, you can still get 100% financing on a stated income loan, with a FICO above 680 or so. So the credit crunch, while started, is affecting only a small percentage of buyers. How will the credit crunch affect seller willingness to make a deal?

Submitted by sdrealtor on April 4, 2007 - 4:53pm.

Great Chart! What is most surprising to me is how little blue line dipped down and how it went sideways for nearly 10 years. Not only could this downturn last a decade it mostly likely will with much of the losses incurred in real rather than nominal terms just like last time. I'm sure the explosion this time around changes the scale of this graph which minimizes the last downturn but still.....

Submitted by sdrebear on April 4, 2007 - 4:53pm.

PS, I do see your point and you're probably right, but I think Rich already had that idea covered if you read further on "The Voice".

Notices of default -- otherwise known as NODs, or the nastygrams that mortgage lenders send to delinquent borrowers -- can serve as an indicator of how much must-sell inventory is out there. It's a rough indicator, mind you. Plenty of defaulting homeowners end up working out their loans, even as some desperate sellers (whose motivation may stem from job loss, relocation, or divorce) enter the market without ever having incurred an NOD. While not a precise measurement, however, the number of NODs delivered in a given month likely correlates very well with the amount of must-sell inventory on the market.

Unless you're a psychic, it would be pretty tough to actually measure the "seller psychology" with any accuracy. However, the NOD's do seem to be a decent "indicator" for must-sell inventory. Certainly not anything exact, but enough to make a reasonable relational chart as Rich did. I think his chart makes a valid enough point.

Submitted by noone on April 4, 2007 - 7:57pm.

Am I reading that right? Currently the number of sales/NOD is 1:1? WOW!

Submitted by greekfire on April 4, 2007 - 9:58pm.

It was only less than a year ago that the mainstream media started to cover (really cover) that the housing market was not going to go up forever. It took another 6 months or so for the media to start to talk about the underlying problems with the housing market, namely subprime mortgages. February 27th hit and that was yet another wake up call, albeit the most forceful one so far. This stuff is only STARTING to hit the mainstream public, the ones who think that RE can only go up. On the flip side, these results were a year or more in advance on this very website.

It is only recently that a good majority of the general public are talking about RE as not being the investment it was believed to be. The psychology IS shifting, it is just taking longer on the way down. There is a mountain of financial data that points to a big market correction. However, one monthly report comes out by some biased organization stating that pending sales are up and VOILA! The economy is back on track and the RE bottom is behind us! You still have hacks like George Chamberlain using the days on market (DOM) indicator as a part of his argument. Remember, there is still money to be made in RE advertising revenues and there are still MANY people that stand to lose from a rapid market correction. Make no mistake about it, the RE bulls are going to fight tooth and nail on the way down.

On another note, it would seem as though there is some historical "competition" between real estate and stocks for investor's money. Doesn't it seem a bit strange that they both depend on each other so much now? I could be reading into it more than I should, but it sounds a bit like 2-card monte to me.

Great chart, Rich.

Submitted by CAwireman on April 5, 2007 - 8:22am.

Psychology is very sticky on the way down!

* PS - very glad to see you posting here!!!

Rich, great way to help make heads or tails of the data. I don't know if you could place a patent on your data charting and manipulation techniques, or if the copyright notices on you site already result in that. But, if no one else is plotting the data in the particular ways you have been, then its certainly worth considering some sort of publication with you stats, facts and figures to help put in on the map.

Who knows your creative methods could some day become part of the industry standard economic bench marks.

Okay, I'll stop gushing now.....

But, as always, great to see this kind of data.

Submitted by slothurston on April 5, 2007 - 11:30am.


Your chart is simply elegant !!!!!

Submitted by itscrowdedinhere on April 5, 2007 - 11:42am.

Curious about everyone's thoughts on how the trend of 'sticky declining prices' might be affected by Web 2.0?

There's so much more information available to the average buyer now, and it gets disseminated so much more quickly. There is an SF blog dedicated to deconstructing (detailed!) and ridiculing overpriced listings in the city. Plus there's mashups galore out there --zip maps, crime stats, coffee shops per capita, commute times, API scores, greenspace...this has GOT to have some effect on the marketplace. What do you think it will be?

Submitted by jg on April 5, 2007 - 12:52pm.

Very nice analysis, Rich.

ps, how do you systematically measure 'investor psychology'? Best way, I think, is to use a reasonable proxy that's already out there. NODs -- threat of proceeding to foreclosure -- sure seems like a plausible one. And, given the nice tight leading (30 month) inverse relationship between NODs and prices, what's not to like about NODs as a leading indicator?

Psychology is nice, but concrete, objective measurements are better. That's why folks downplay reads of consumer confidence and place greater weight on actual consumer spending.

Submitted by greekfire on April 6, 2007 - 12:10am.

JG - Equally nice chart. Does this inverse relationship exist only for the last downturn (90s), or does it go further back? Do you see any changes to the 30-month timeframe given the huge spike this time around and more widespread access to information (e.g., the Internets)?

Submitted by sdrealtor on April 6, 2007 - 10:43am.

Greed that JG's chart is fabulous. Can't wait for the chart to peak and turn so I can plan get a plan in place for 3 years down the road.

Submitted by rockclimber on April 6, 2007 - 10:58am.

JG, where did you come up with the 30 months? Did you just chart the two and then see where the peaks align with the troughs and visa versa?

Submitted by no_such_reality on April 6, 2007 - 11:29am.

Psychology is very sticky on the way down!

Actually it's not. Last go around, by the end of 1992, everybody knew if they were stuck or not.

Prices were sticky because sellers were stuck. They could make the payments but couldn't afford to sell the house.

This time, more sellers are shaping up to not be able to make the payments and hence, won't have a choice about the house being sold.

Submitted by JC on April 6, 2007 - 12:54pm.

I have a question that I would love some input on. I'm not specifially looking for a REO, but I've looked up a few of those properties on Zillow and other similar sites recently and noticed that almost all of them show very recent sales. Also, in most cases, the sale price was quite a bit higher than the value quoted by the site.

My question is when the banks take over the homes does this count as a "sale" and could this actually artificially inflate home prices in cases where the borrower owed more than the value of the home?

I hope this makes sense!

Thank you!

Submitted by itscrowdedinhere on April 6, 2007 - 4:27pm.

This is pretty amazing. I'd like to put something like this together for Alameda County. Do you know of publicly available data I can use?

Are you including condos or only sfr's? I can search Alameda County records for NOD's pretty easily, but there seem to be 2 kinds--a NOD on a lien (usually on a residence, as far as I can tell), and a NOD on deed of trust. Are these significantly different, enough so that you should filter out the NOD-liens as you construct the chart? Thanks.

Or, jeez, if you have the data available and can make it happen without too much work, I would LOVE to see it for Alameda County.

Submitted by Bugs on April 7, 2007 - 9:10am.


Some of the automated valuation systems (like Zillow) might misinterpret those transactions as sales and it could affect their results. Actually, I should amend that; some of those systems WILL misinterpret those transactions.

Submitted by paranoid on April 7, 2007 - 12:33pm.

can you plot the same figure, but using Trustee Deeds instead of NOD? I think that may give a better view, because TDs are really MUST SELL, better than NODs most of which never go to sale.

Submitted by CAwireman on April 7, 2007 - 4:04pm.


As always, insightful graph. Question, did you shift the NOD data 30 months? Or, are the values "as is" in relation to time?

The NOD data and the Median data contrast very well and I wondered if that was naturally occuring, of if the data needed to be massaged to get that visual effect.


Submitted by powayseller on April 7, 2007 - 11:31pm.

sdrebear, I disagree with the author's statement "however, the number of NODs delivered in a given month likely correlates very well with the amount of must-sell inventory on the market." There is no evidence that I can find to support that statement, and I wonder how Rich came to that conclusion. It would be worthwhile for him to substantiate this somehow.

jg, the correlation that you cite is opposite. Price drops are causing rising NODs, not the other way around.

NODs are lagging indicators. I'm not sure why you used a 30month lagging indicator to predict prices. NODs rise well after a housing market has turned up, and are well known to be lagging.

I suggest that we predict price direction by looking at supply and demand. Now that we have access to inventories, pendings, and sales, we can forget about all those other factors. A few years ago, that data was not available. Prices for any good are determined by supply and demand, and housing is no different. We don't need to look at all the other factors because they are instantaneously reflected: when credit tightens, demand immediately is reduced. When people leave the area, demand is reduced right away. So we are very fortunate to have access to the supply and demand data. I cannot predict the actual prices, but based on supply and demand, I believe we can predict the direction of prices.

Foreclosures and NODs are interesting just to prove that people are indeed in trouble, and that the bubble has burst, but it has no predictive value. We cannot say that rising foreclosures mean lower prices. Foreclosures are rising much faster than prices are dropping. I think the greatest price drops are due to investors leaving the market last year, causing a 30% drop in demand; investors were a huge part of the market.

Again, foreclosures lag what is actually going on. The 4300 pre-foreclosures in San Diego this month reflect 4300 people who had trouble making their payments in 2006. The people having trouble now are not even counted yet. Those 4300 people would have made decisions to list their home for sale last year, and if they had successfully done so, they would not be in foreclosure now.

I can't wait for prices to drop either. But I am starting to face the fact that it could take a decade for prices to end up to 1999 levels. It's a very very slow process.

Submitted by barnaby33 on April 8, 2007 - 9:18am.

If NOD's are a lagging indicator, they probably aren't predictave of the start of the downturn. They surely are predictive of the end of it. So they are predictive, just not of immediate future pricing.


Submitted by powayseller on April 8, 2007 - 9:58am.

Josh, read my article. Economists do not use lagging indicators, and I have quotes from Cagan, Ratcliff, RealtyTrac, saying foreclosures are lagging indicators.

Foreclosures tell you that homeowners went into distress some time ago, and they could not refinance or sell their way out of the problem.

Eventually, this housing market will turn around, and prices start to rise.

But foreclosures will keep rising after the market has turned around, and those of you relying on NOD data are going to miss the bottom.

Submitted by rseiser on April 8, 2007 - 11:44am.

I am missing something here? jg made the point that NODs would be leading by about 30 months and is therefore a leading indicator. Why are you talking about it being a lagging indicator? Is there any evidence to this as well?

Submitted by HereWeGo on April 9, 2007 - 9:13am.

Outstanding chart, Rich. Normalizing the NOD's by sales certainly seems a fair indicator, based on the last downturn. That said, the linear scale might start to fail a bit (from an analysis perspective) if that ratio continues to drop.

Submitted by Rich Toscano on April 9, 2007 - 9:14am.

OK, I am going to submit a rare comment here now that Powayseller has seen fit to write an entire article on her site about this topic (with discreet statements like “ The Voice of San Diego has it wrong”) in addition to sending my editor a letter “correcting the story” (seriously – “correcting the story”).

As far as I can tell, PS, you have three issues: 1) you claim that NODs do not correlate to distressed inventory, 2) you claim that lower prices cause NODs, and not the other way around, and 3) you claim that foreclosures are a lagging indicator.

The first claim makes little sense. Who is forced to sell? For the most part, it’s people who can’t pay for their homes any more. If they could pay for their homes, they wouldn’t be forced to sell them. Do you honestly believe that there is no correlation between the number of people who can’t pay for their homes (as measurable by NODs) and the number of people who are very motivated to sell? It is so obvious as to be self-evident that these two numbers will move up and down together.

Of course, there is not a one-to-one relationship. But I never said there was, and if you’d taken the time to actually read and comprehend my story before issuing your “corrections,” you’d see that I specifically said that there WASN’T a one-to-one relationship, but that there would obviously be a rough correlation.

Your second argument, that lower prices cause NODs, isn’t an argument at all. Which is to say, you are just stating the obvious. Everyone knows that price appreciation prevents defaults, because homeowners in trouble can monetize home equity. Once prices start falling, this margin of safety goes away and defaults ramp up.

All obvious stuff. But for some reason, you seem to think that because price depreciation can lead to further defaults, that defaults (or more to the point, an increase in the number of people who can’t pay their mortgages) can’t lead to further price depreciation. Why? This is a self-reinforcing relationship… lower prices lead to more defaults lead to lower prices lead to more defaults. It’s the opposite of the virtuous circle we had on the way up. If elements didn’t self-reinforce like this we wouldn’t be prone to the boom/bust cycles to which we obviously ARE prone.

The last issue you are upset about is that foreclosures are a lagging indicator. But lagging to what? I agree they are a lagging indicator of financial distress in that they aren’t filed until people have already been delinquent for months. But so what? In a slow market like RE, they can still give signals about price pressures. The graph clearly shows that inflection points in the sales-per-NOD line have lead those in the price level (which, btw, was all I ever claimed). It’s right there in the chart, yet for some reason you deny it.

All in all, it seems you are either “arguing” against claims I never made or denying verifiable fact.


PS – (That’s postscript, not Powayseller) -- While I’m here, to respond to “paranoid”: a graph with sales-per-NOT shows largely the same phenomenon, except it doesn’t lead by as much and is much harder to read due to the wider volatility of the NOT series.

Submitted by sdrealtor on April 9, 2007 - 11:21am.

One more thing. Her case confuses indicators with he causes. Something can be an excellent indicator of a situation without being a cause of it. In this case we know the many of the causes of the housing bubble, what are are looking for is clear indicators of changes in the market.

Submitted by sdrebear on April 9, 2007 - 12:51pm.

Well, Rich, you answered much more clearly than I could have.

PS - I saw you argue this point several month's ago about NOD's "only" being a lagging indicator and that they have no predictive value what-so-ever. I instantly had an issue with that thought, but never posted on it. Now that you seem to be trying to make a much larger point about it, I guess I'd like to interject a little more.

What's odd about all this is that before, you seemed to get the concept that this bubble would not be bursting due to normal factors such as massive local job loss, or natural disaster as the Bulls have been feeding us for years now. This burst will be the result of the massiveness of this bubble in and of itself collapsing under its own weight.

Part of the collapse is the swing from a virtuous cycle to a vicious cycle where issues that normally lagged the other problems "causing" the collapse are now themselves additional "causes". This is on top of new "causes" bulls never thought of before, like massive loan resets and severe credit tightening. The foreclosures themselves become a strong "third leg" of sorts to further drive down prices. Therefore, predicting (through NOD's and a little math) these foreclosures IS very helpful in forecasting price declines. Will it predict a bottom? Probably not on its own, but seeing the NOD's dry up 3-6 months before the actual foreclosures slow down would be a pretty good clue that things are finally getting better, don't you think? (Kind of sounds like a leading indicator of market health)

If we know that NOD's result in actual foreclosures at a given percentage (and we do) and also know that those foreclosures result in lower prices for those given homes (and we do), then it stands to reason that on a massive scale (which you yourself have predicted for this market), foreclosures can and will have enough economical impact to actually BECOME a driving indicator of FUTURE price reductions.

Just because foreclosures haven't driven decline (wide spread) in the past (due largely to their shear lack of volume), does NOT mean that they can't become that driver in the future. Nothing about how this bubble was created or its pure scale is measurable to past cycles. Things happened on the way up that never happened before. Is it so impossible to predict that there will be a few things on the way down that have never happened as well?

So, again, I'm more confused about how you have come to believe that only "traditional" forces can put leverage on the market (knowing what you now know about THIS market) and also how you believe people would honestly be willing (or able) to wait "a decade" for prices to come back to them.

By the way Rich... Nice write-up in the SD Business Journal!

Submitted by powayseller on April 9, 2007 - 8:08pm.

I think there is some good thinking going on, sprinkled with a little speculation and misunderstanding.

what percentage of homeowners with NODs list their homes for sale?

Of those who list their homes for sale, how much below market price do they list?

What is the average DOM of those homeowners?

What is the % of list price that those NODs sell for?

Once you researched these questions, rather than just speculating about them, do you still think that NODs are causing price drops?

Submitted by robson on April 9, 2007 - 8:41pm.

Not to interrupt the conversation, but on the topic of NOD...Today the SD Daily Transcript refreshed their monthly figure. With 1517 NOD in March, San Diego broke the old record of 1484 in April, 1982. Of course today's figure remains lower than the 1982 figure after deflating for population or housing units. Still, it WAS a 25 year old record.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.