Real Estate Mathematics

Friday, August 18, 2006

Relation between Price Changes and Inventory Changes



At an intellectual level when inventory increases, price is expected to decrease and if inventory decreases, price is expected to increase. There are exceptions of course, for example if there is large scale immigration or emmigration, or changes in interest rates.

There is also seasonality involved in prices, so using the classic trick to reduce seasonality, we divide the current price by the price a year ago (same with inventory levels). This means that "1" becomes significant --- it's where there is no (or little) change and you would expect both price and inventory to become static/stable.

We will use King County (Seattle) to illustrate this.

As you see, we have the line being pretty stable, and following the expected pattern:

  • If Inventory is below 1, price is above 1
  • If Inventory is above 1, price is below 1

What I want to look next is the lag factor between the two.
Looking at the bottom red circle, we see that inventory started to fall in month 27 and price started to climb in month 31 -- approximately a 4 month lag. At month 47, there was a drop in inventory and in month 50 we see a major jump in price, a 3 month lag. At month 51 there was a jump in inventory (or at least a reporting of such) and in month 57 there was a major drop in price (a 5 month lag). In that same month, there was the start of a run up in inventory rate of change and in month 60, the drop in price rate started to decrease..

In short, rate of price change is lagging inventory by about four months. So what does that say about the next four months of prices in Seattle? They should be pretty flat (that is the same prices as seen a year ago) and may actually start to decline.

Thursday, August 17, 2006

King County (Seattle) July Figures


In the prior blogs I described various indicators. In this blog I will look at some real current numbers, starting with King County / Seattle Washington.

If you plot our Active Listings you will see a chart like the one below.

One of the first problem is that Active Listings is seasonal (X11) and can have a lot of noise. If you compare the current numbers against their values a year ago (i.e. as a ratio), then a lot of noise disappears and you can quickly see a nice "sink". There are two blips... one that is a peak -- and one year, it's echo, a trough - ignoring this, we end up with a very smooth curve.

If you look at the other counties around Seattle, you see a similar pattern but with the depth being much less and the current values going higher. In short, the distant suburbs around Seattle have far more house-inventory building than Seattle --- perhaps people trying to sell and move closer because of rising gasoline costs...





It is interesting to note that most of the counties have the same data oddity that King County has on the same date - it may have been some form of reporting error in the MLS system or some other event.

Trend Lines

Let us add trend lines to the above charts to show how clean the pattern is. The King County follows the numbers well -- and the two hiccups balanced each other out. What is interesting is that the last three values are further away from the line then at any time (except for the hiccups). It appears that there has been an extra influence in the market to get more listings -- Interest Rates? Gas Prices?

For the other countrie, we see the bottom occurring much earlier. It almost appears that the folks in King Country were able to hold on longer to the good market, and are not surrendering to the regional reality.


So, what can we conclude --- properties are going to get harder and harder to sell, i.e. they will stay on the market a longer and longer period. This brings us to another Indicator: Days on Market, this is also a long lag indicator because it is not obtained until the house closes. That is, if the Days on Market is 4 months, it reflects the situation four months ago, not now. A new listing now may have Days on Market being 9 month.

Additionally, Days on Market have another problem, it is below the expected days to sell -- because the properties that may be listed for X years are ignored... in short, it's a long lag indicator.

Tuesday, August 15, 2006

Puget Sound Real Estate - Example
In this lesson, we will look at the June 06 numbers for Washington State. You can see them at: http://www.nwmls.com/discover/library/statistics/recaps/Recap2006/Jun06Recaps.pdf
All but two countys show increases of inventory, none in the Seattle area. This suggests that the market is souring, houses will take longer to sell. With 50% more houses, you would expect it to take 50% longer -- no such luck.

With increase of inventory, there is usually a drop in number of houses being sold. From the same report we can actually see this clearly, and introduce our four and fifth indicators:
* Pending Sales
* Closed Sales





Pending Sales Indicator


If you take pending sales and compare them with the prior year, you will see that sales are falling for most area. The major exception are again outside of the Seattle area.


Pending sales is a short lag indicator -- it reflects sales that just happened or which happened in the last few months.







Closed Sales Indicator


This is another indicator reflecting sales that closed this month. The price was set several months ago and is thus lagging. As you can see, every county is having less sales closed. There are less people buying. This is usually some percentage of Pending Sales. The percentage may change over time -- and generally not reported. If it was, it would likely be a good indicator.


Looking at the last item,, closed sales, an indicator from the earlier lesson, we see that it is not showing a slow down.
This illustrates the human factor, people are driven by the price and not the other factors.








So what are our indicators...
  1. Change of new listing - leading indicator -- complex to intrepret
  2. Change of Inventory -- leading indicator
  3. Change of Pending Sales - short lag indicator
  4. Change of Closed Sales -- lagging indictor
  5. Change of Sales Price - long lag indicators







A housing market is like the economy - there can be leading indicators of direction and lagging indicators. Many people use current housing prices as their sole indicator not knowing that this is a lagging indicator, and then crying.

For the first lesson, let us look at the numbers available from many multiple listing services as monthly reports. The first number to look at is Inventory of houses for sale. If there is a large inventory then it's a buyer market, if there is a small inventory it is a seller's market.


The image above is from http://www.nwrealtor.com/associations/1563/files/2005%20anreport.pdf and shows the inventory in the Seattle area over three years. The inventory levels dropped from 2003 to 2004, indicating a hot market. In 2005 this pattern continued until October, 2005 when inventory started climbing -- the market was cooling each month more.

What about housing prices... from the same report we see a different situation.
King Country Sales
Although inventory is growing - suggesting a slowing, the sale prices do not reflect this! Why?
There are a few clear reasons:

  • A sale may take 60-120 days to close. What you are seeing are the offer price from month earlier!
  • Buyers tend to be slow to pick up on a change of inventory -- being focused on the price and not the inventory.

This means that sales price is a lagging-indicator that is typically 3-6 months behind the market conditions. House inventory is a leading indicator, it reflects changes almost as soon as the market changes.

Another indicator that is a more complex to intrepretate is new listings. An example is shown below:


What we see is interesting, there was a surge of new listings starting in August 2005. These new listings came on too fast to clear on the market. The problem is why they occurred?

Was it new units that were built for the hot market? Has a magic point happened that caused people to want to capture their gains? Was it people realizing that they were over-extended because of interest rates and wanting to get out?

So what do we have as our first lesson? Three indicators:
  • Change of New Listings (most lead) - complex interpretation
  • Change of Total Inventory (leading) - clean interpretation
  • Change of Sale Price (lagging) - clean interpretation