The Central PA Multi-List (CPML) recently published year-end statistics for residential sales. As is often the case the statistics are somewhat contradictory. On the one hand there is good news: based on average sale price, housing appreciated 1.85% in 2010. The previous year it had depreciated 3.75%. On the other hand sales volume decreased about $132,000,000 (or 10%) and average days-on-market drifted steadily upwards to 102 days in the 4th quarter, the highest number since I started selling real estate in 1995.
From the 2nd quarter to the 3rd quarter the number of homes sold dropped 732 units. This represents a 30% change. Normally about the same number of units sell in the 2nd as the 3rd quarter. With a high degree of confidence we can attribute this abrupt slowdown to the end of the federal government’s homebuyer tax credit program on June 30th. This took away a big incentive for 1st-time buyers; much of the subsequent drop in housing sales occurred in the under-$200,000 market.
Based on the last half of the year about 485 homes are sold each month in the Harrisburg metro area. At year’s end there were 4,502 homes on the market, representing a 9-month supply of homes. A 6-month supply is generally considered a neutral market, i.e., not favoring buyers or sellers.
Already low at the start of 2010, mortgage rates trended lower throughout the year, bottoming out at 4.25% in November. Low rates help all homebuyers: arguably they provide a bigger incentive for high-end buyers than low-end buyers. This is borne out by the numbers. For the year the number of units sold above $300,000 increased 8.4%, while the number of units sold below $300,000 dropped 14%.
The 1.85% increase in the average sale price is somewhat illusory as all the other sales statistics suggest a weak, possibly declining, housing market. According to the Wall Street Journal (“Home Prices Sink Further,” January 31, 2011), citing market research by Zillow.com, declining markets are the norm in cities across the country. Home prices in Philadelphia reportedly fell 8.8% this past year.
In a declining market buyers and sellers often fail to come to terms. Homes just sit on the market or discouraged sellers pull them off the market. So rather than focus on average sale price in central PA, a better gauge is probably home inventory and days-on-market. When these numbers start to drop it likely signals a true change in market direction.
Bottom line: the Harrisburg housing market is still in the doldrums, where it has been since housing prices peaked in 2007.
Showing posts with label Average House Price. Show all posts
Showing posts with label Average House Price. Show all posts
Monday, January 31, 2011
Thursday, April 15, 2010
Metro Harrisburg Housing Price Trends
OK fellow real estate wonks…let’s finish the topic of trending house prices.
As noted in an earlier post Harrisburg is not one of the twenty metro areas in the S&P/Case-Shiller Index. Presumably its small size would make it difficult to collect a sufficient number of repeat sales to have a statistically significant sample size without either extending the collection period and/or expanding the metro area, both of which would make the resulting index less useful to someone selling homes in Harrisburg.
Therefore, in small metro areas, such as Harrisburg, PA, the multi-list is typically the best (or only) source of housing market statistics. My local multi-list, the Central PA Multi-List, publishes mean (average) and median house prices quarterly.
The historical residential sales figures I have collected date back to 1991. The statistics reported in 1991 were rudimentary—units sold, sales volume, and average price. Gradually the Central PA Multi-List added more statistics—average days on market, number of active listings, median sale price, and breakdowns by county and price range. Today a motivated real estate wonk can download the entire multi-list database to Excel and analyze to his/her heart’s content.
For reasons never made clear (to me anyways) the Central PA Multi-List based all its sales statistics through 2006 on settled and pending (under contract) properties in any quarter or year, presumably using the list price of pending properties. Inclusion of pending properties doesn’t make sense as a) actual sale price is historically 97% of the list price at the time a property goes under contract, and b) some deals fall through. All of which suggests that pending properties would tend to artificially inflate the average and median house price figures. Curiously, when the Central PA Multi-List recomputed sales statistics for the years 2002 through 2006 without pending properties, the average and median increased--go figure!
The chart above shows trendlines for Harrisburg housing sales from 1991 to present. The old-style average sale price (including pending sales) is the blue line from 1991 to 2006. The new-style average sale price (excluding pending sales) is the red line from 2002 to 2009. Note slight discrepancy between the two sets of average sale prices in the overlap years 2002 to 2006.
The Central PA Multi-List has only published the median sale price from 2002 to the present. I have previously argued the median is more useful than the average as a measure of "typicalness" for housing. Here it is displayed as the green line from 2002 to present. All median house prices exclude pending sales (thankfully). The new-style average and new-style median trendlines closely follow each other separated by about $25,000.
As noted in an earlier post Harrisburg is not one of the twenty metro areas in the S&P/Case-Shiller Index. Presumably its small size would make it difficult to collect a sufficient number of repeat sales to have a statistically significant sample size without either extending the collection period and/or expanding the metro area, both of which would make the resulting index less useful to someone selling homes in Harrisburg.
Therefore, in small metro areas, such as Harrisburg, PA, the multi-list is typically the best (or only) source of housing market statistics. My local multi-list, the Central PA Multi-List, publishes mean (average) and median house prices quarterly.
The historical residential sales figures I have collected date back to 1991. The statistics reported in 1991 were rudimentary—units sold, sales volume, and average price. Gradually the Central PA Multi-List added more statistics—average days on market, number of active listings, median sale price, and breakdowns by county and price range. Today a motivated real estate wonk can download the entire multi-list database to Excel and analyze to his/her heart’s content.
For reasons never made clear (to me anyways) the Central PA Multi-List based all its sales statistics through 2006 on settled and pending (under contract) properties in any quarter or year, presumably using the list price of pending properties. Inclusion of pending properties doesn’t make sense as a) actual sale price is historically 97% of the list price at the time a property goes under contract, and b) some deals fall through. All of which suggests that pending properties would tend to artificially inflate the average and median house price figures. Curiously, when the Central PA Multi-List recomputed sales statistics for the years 2002 through 2006 without pending properties, the average and median increased--go figure!
The chart above shows trendlines for Harrisburg housing sales from 1991 to present. The old-style average sale price (including pending sales) is the blue line from 1991 to 2006. The new-style average sale price (excluding pending sales) is the red line from 2002 to 2009. Note slight discrepancy between the two sets of average sale prices in the overlap years 2002 to 2006.
The Central PA Multi-List has only published the median sale price from 2002 to the present. I have previously argued the median is more useful than the average as a measure of "typicalness" for housing. Here it is displayed as the green line from 2002 to present. All median house prices exclude pending sales (thankfully). The new-style average and new-style median trendlines closely follow each other separated by about $25,000.
Subscribe to:
Posts (Atom)
