Supply and Demand of Housing
The most common way to measure supply and demand of housing is by looking at the number of months it would take for the current supply to be sold given the current rate of purchases. This number is typically about 5 months in an average environment, 4 months in a "sellers market", and 6 months in a "buyers market".
While I think this is probably the best metric, I see a couple of limitations. One is that it is always backwards looking and at least 2 or 3 months behind what current conditions are (granted typically 2 or 3 months doesn't matter very much). Another factor adding to this lag effect is new construction as large neighborhoods may be getting pre sold many months before actually coming onto the market.
Because of these things I've been trying to think of ways to find a more forward looking metric. What I have come up with is very simple, more forward looking, but unfortunately probably lacks much precision. It certainly will not be replacing the days on the market metric anytime soon. When I say simple, I mean it:
Supply: Instead of the number of houses for sale, I think a more forward looking supply metric is the number of housing starts released monthly by Censushttp://www.census.gov/indicator/www/newresconst.pdf By calling starts the supply, I am implying that builders are basing decisions to build houses on sale prices and construction costs (and ignoring inventory of used homes for sale).
Demand: Instead of looking at houses sold, I believe a demand proxy can be the home builders confidence index. This index is simply the specuation of future home buyers, and I believe it is fairly accurate.
Below is a graph of these 2 series, with the scales set to keep them generally matched up. Any comments on this method is appreciated.
While I think this is probably the best metric, I see a couple of limitations. One is that it is always backwards looking and at least 2 or 3 months behind what current conditions are (granted typically 2 or 3 months doesn't matter very much). Another factor adding to this lag effect is new construction as large neighborhoods may be getting pre sold many months before actually coming onto the market.
Because of these things I've been trying to think of ways to find a more forward looking metric. What I have come up with is very simple, more forward looking, but unfortunately probably lacks much precision. It certainly will not be replacing the days on the market metric anytime soon. When I say simple, I mean it:
Supply: Instead of the number of houses for sale, I think a more forward looking supply metric is the number of housing starts released monthly by Censushttp://www.census.gov/indicator/www/newresconst.pdf By calling starts the supply, I am implying that builders are basing decisions to build houses on sale prices and construction costs (and ignoring inventory of used homes for sale).
Demand: Instead of looking at houses sold, I believe a demand proxy can be the home builders confidence index. This index is simply the specuation of future home buyers, and I believe it is fairly accurate.
Below is a graph of these 2 series, with the scales set to keep them generally matched up. Any comments on this method is appreciated.
2 Comments:
I just shot up the 2006 San Diego Home Prices Pretty damn nasty...
I think you should try to correlate the NAHB Housing Market Index to actual home sales. That way, you can see the lead-lag effect of sales to the index.
Interesting thing to note, when index goes down, starts lag by 3-6 months. But when the index goes up, starts almost mirror the index.
Post a Comment
Links to this post:
Create a Link
<< Home